Vous êtes sur la page 1sur 456

Ar

chi
vedWebCont
ent Cont
enuWebar
chi

Saving the Twentieth Century: Sauvegarder le XXe siicle :
The Conservation of Modern la conservation des mat6riaux
Materials modernes

Proceedings ofa Conference Les actes de la conf6rence


Symposium '91 - Saving the Twentieth Century Symposium 9l - Sauvegarder le XXe sidcle
Ottawa, Canada Ottawa, Canada
15 to 20 September 1991 du 15 au 20 septembre l99l

Organized by the Canadian Conservation Organis6 par I'Institut canadien de


Institute, Communications Canada conservation, Communications Canada

Edited by Sous la direction de

David W. Grattan David W. Grattan

Canadian Conservation Institute Institut canadien de conservation


Ottawa, Canada Ottawa, Canada

1993 t993
Canadian Cataloguing in Publication Data Donn6es de catalogage avant publication
(Canada)

Symposium '91 - Saving the Twentieth Century Symposium 91 - Sauvegarder le XXe sidcle
(1991 : Ottawa, Ont.) (1991 : Ottawa, Ont.)

Saving the twentieth century : the conservation Saving the twentieth century : the conservation
of modern materials : proceedings of a confer- of modern materials : proceedings of a confer-
ence, Otlawa,Canada, 15 to 20 September ence, Ottawa,Canada, l5 to 20 September
1991 :Sauvegarder le XX" sidcle : la conser- l99l : Sauvegarder le XX" sidcle : la conser-
vation des matdriaux modernes : les actes de vation des mat6riaux modemes : les actes de
la conf6rence, Ottawa (Canada) du 15 au la conf6rence, Ottawa (Canada) du 15 au
20 septembre 1991 20 septembre 1991

Prefatory material and abstracts in English Texte pr6liminaire et r6sum6s en anglais et


and French. en frangais.
Includes bibliographical references. Comprend des r6f6rences bibliographiques.
rsBN 0-660-57854-9 ISBN 0-660-57854-9
DSS cat. no. NM95-5812-1992 No de cat. MAS NM95-5812-1992

l. Museum conservation methods Canada l. Mus6es M6thodes de conservation


Congresses. 2. Museum techniques
- Canada- Canada- -Congrds. 2. Mus6ologie -
Canada
Congresses. 3. Materials
-
Conservation Congrds. 3. Mat6riaux
-
Conservation et
-and restoration Congresses. - I. Grattan, -restauration Congrds. I.- Grattan, David W.
-
David W. II. Canadian Conservation Institute.
-
IL lnstitut canadien de conservation. III. Tife.
lll. Title. IV. Title: Sauvesarder le XXe siecle. IV. Titre : Sauvegarder le XXe sidcle.

AMl4l.S28 1992 069'.53'0971 AM141.S28 1992 069',.53'0971


c92-099400-0F. c93-099400-0F
Proceedings published subsequent to Ce document a 6t6 publi6 d la suite du
Symposium '91 - Saving the Twentieth Symposium 91 - Sauvegarder le XXe sidcle,
Century, organized by the Canadian Conserva- organis6 par I'Institut canadien de conserva-
tion Institute, Communications Canada, and tion, Communications Canada, et pr6sent6
held at the Skyline Hotel, Ottawa, e l'h6tel Skyline, Ottawa, du 15 au
I 5 to 20 September I 991 . 20 septembre 1991.

o Communications Canada, Ottawa, 1993.


o Communications Canada. Ottawa. 1993.

All rights reserved. No part of this publication Tous droits r6serv6s. La reproduction d'un
may be reproduced or transmitted, in any form extrait quelconque de ce livre, par quelque
or by any means, electronic or mechanical, proc6d6 que ce soit, tant 6lectronique que
including photocopying, recording, entering m6canique, ou par photocopie, microfilm,
in an information storage and retrieval system, bande magn6tique, disque ou autre, sans
or otherwise, without prior written permission I'autorisation 6crite de l'6diteur est
ofthe publisher. interdite.

Cover Design: Sophie Georgiev. Graphisme de la couverture : Sophie Georgiev,


Canadian Conservation Institute lnstirut canadien de conservation

PRINTED IN CANADA IMPRIME AU CANADA

Available ftom: Vendu par :

Extension Services Services de diffusion externe


Canadian Conservation Institute Instinrt canadien de conservation
Communications Canada Communications Canada
Ottawa, Ontario Ottawa (Ontario)
KIA OC8 KIA OC8
Acknowledgements Remerciements

Organizing Committee Comitd organisateur

J. Cliff McCawley (chair) J. Cliff McCawley (pr6sident)

Henri Benoit Henri Benoit


Malcolm Bilz Malcolm Bilz
Marie-Claude Corbeil Marie-Claude Corbeil
Valerie Dorge Valerie Dorge
David Grattan David Grattan
Maureen MacDonald Maureen MacDonald
Janet Mason Janet Mason
Wanda McWilliams Wanda McWilliams
Charlotte Newton Charlotte Newton
Jean T6treault Jean T6treault
David Tremain David Tremain
Season Tse Season Tse

The Organizing Committee thanks the speak- Le comit6 organisateur tient d remercier les
ers, session chairs, presenters ofposters, and conf6renciers, les pr6sidents de session et tous
participants in the panel discussion for their les participants d la table ronde pour leur
conhibution to the Symposium. contribution au symposium.

Editorial Committee Comite de rtdaction

David Grattan (chair) David Grattan (pr6sident)

Valerie Dorge Valerie Dorge


Charlotte Newton Charlotte Newton
Jean T6treault Jean T6treault

Production Committee Comite de production

Deborah Robichaud (chair) Deborah Robichaud (pr6sidente)

A.P. Doming A.P. Doming


Sophie Georgiev Sophie Georgiev
David Grattan David Grattan
Sandra LaFortune Sandra LaFortune
Linda Leclerc Linda Leclerc
Publication Publication

The editor thanks the authors for their manu- Le r6dacteur en chef tient ir remercier les au-
scripts and for their kind cooperation during the teurs de leur contribution au projet ainsi que de
editorial process. The editor also acknowledges leur coop6ration. Il d6sire 6galement exprimer
the help of the following people in the prepara- sa gratitude d toutes les personnes qui lui ont
tion of this publication: pr6t6 main-forte :

Caroline Shaughnessy: Caroline Shaughnessy :


English editing R6vision des textes anglais
Andr6 La Rose: Andr6 La Rose :
French editing and translation R6vision des textes frangais et traduction
Sophie Georgiev: Sophie Georgiev :
Design and layout Graphisme et production
Sandra LaFortune and Linda Leclerc: Sandra LaForhure et Linda Leclerc :
proofreading
Editorial assistance and Aide d la r6daction et correction d'6preuves
Valerie Dorge, Charlotte Newton, and Valerie Dorge, Charlotte Newton et
Jean T6treault: Jean T6treault :
Technical review R6vision des textes techniques
Translation Bureau, Departrnent of the Bureau des traductions du
Secretary of State: Secr6tariat d'Etat:
Translation of abstracts Traduction des r6sum6s

Consultation Consultation

We acknowledge the advice and assistance Nous tenons ir remercier certaines personnes
of the following people in plaruring the qui, par leurs conseils et leur aide, ont con-
Symposium: tribu6 d la planification du symposium :

Dr. David J. Carlsson: David J. Carlsson, Ph.D. :


National Research Council Canada, Conseil national de recherches Canada,
Ottawa Ottawa
Karen Graham: Karen Graham :
Canadian War Museum, Ottawa Mus6e canadien de la guerre, Ottawa
Robson Senior: Robson Senior:
National Museum of Science and Mus6e national des sciences et de la
Technology, Ottawa technologie, Ottawa
Dr. David M. Wiles: David M. Wiles, Ph.D. :
Plastichem Consulting, Victoria, B.C. Plastichem Consulting, Victoria (C.-B.)
Introduction Introduction

The conference "Symposium'91 - Saving the La conference < Symposium 9l - Sauvegarder


Twentieth Century" took place in Ottawa in le XX" sidcle : la conservation des mat6riaux
the autumn of I 991 . Organized and hosted modernes > a eu lieu ir Ottawa au cours de
by the Canadian Conservation Institute (Com- l'automne de 1991. Symposium 91, mis sur
munications Canada), Symposium'9 I formed pied par I'Institut canadien de conservation
one of a series of intemational meetings de- (Communications Canada) qui en 6tait 6gale-
voted to specific conservation problems. ment I'hdte, s'inscrivait dans le cadre d'une
s6rie de rencontres internationales consacr6es
Each Symposium addresses topics of major d l'6tude de probldmes de conservation
importance to museums particularly to particuliers.
-
museums in Canada. Topics are selected that
have been generally overlooked and that pose Chacune de ces rencontres porte sur un
new and demanding challenges. In this way, we sujet d'importance pour tous les mus6es
hope to create a better awareness of problems,
-
surtout pour les mus6es du Canada. Et chacun
to stimulate thought, and to encourage research. de ces sujets correspond ir un domaine qui,
The theme for Symposium '91, modem materi- bien que g6n6ralement n6glig6, n'en consti-
als, meets all these criteria. tue pas moins un nouveau d6fi de taille. Les
conf6rences visent donc d sensibiliser davan-
Museums and galleries reflect society, and are tage les gens d certains probldmes, tout en
influenced by opinion and fashion. As material favorisant la r6flexion et en stimulant la recher-
culture becomes more sophisticated, museums che. Symposium 91, qui portait sur les mat6-
collect a much wider range of objects. Museum riaux modernes, respectait tous ces critdres.
curators are, therefore, assembling collections
that include new slmthetic materials, complex Les mus6es et les galeries d'art sont le reflet
electronic circuitry, sophisticated alloys, and de la soci6t6, et ils ne sont dds lors pas d I'abri
many new types of coatings. Consider the de I'opinion et des modes. Au fur et d mesure
following list: compact disc, radio, computer, qu'6volue la culture mat6rielle, les conser-
spacesuit, aircraft, plywood, and plastic doll. vateurs de ces 6tablissements en viennent ir
All are among the rapidly degrading artifacts collectionner une garnme toujours plus vari6e
that were discussed at Symposium '91. d'objets, faits d partir de nouvelles matidres syn-
th6tiques, de c ircuits 6lecfr onique s complexes
"Complex" is perhaps the word that best et d'alliages de haute technicit6, et dont le re-
describes modern objects. They are becoming couvrement comporte lui aussi de nouveaux
increasingly complex in structure, in materials mat6riaux. Les disques compacts, les radios, les
used, and especially in function. Conservators ordinateurs, les combinaisons spatiales, les
are expected to stabilize and conserve artifacts a6ronefs, les contreplaqu6s et les poup6es en
from all periods in history. However, perhaps plastique ne sont que quelques-uns des objets
by inclination and certainly by training, they qui, se d6gradant rapidement, figuraient au pro-
have concentrated on studying past technolo- grarnme de Symposium 91.
gies rather than current ones. Although modern
materials are generally perceived as being Le mot < complexe ) est sans doute celui qui
more stable than traditional substances, permet le mieux de d6crire les objets modemes,
nothing could be further from the truth. Many qui deviennent en effet toujours plus complexes
twentieth-century artifacts are very unstable, quant ir leur structure, aux mat6riaux qui les
and some have problems that are new to conser- composent et, surtout, ir leur fonction. On
vators, such as radioactivity, polymer degrada- s'attend donc d ce que les restaurateurs arrivent
tion, and aluminum corrosion. d stabiliser et d conserver des objets de toutes
les 6poques de I'histoire. Or, peut-€tre par
Conservators are having to cope with these penchant naturel et sans doute en raison de leur
objects using their limited training in materials formation, les restaurateurs se sont consacr6s
science and an ethical code conceived with davantage d l'6tude des techniques du pass6
more traditional artifacts in mind. plut6t que celles du monde contemporain. Et
m6me si I'on a g6n6ralement tendance d croire
The degradation ofrecently acquired artifacts is que les mat6riaux modemes sont plus stables
forcing curators to review collecting policies. que les anciens, rien n'est moins s0r. Nombre
Should museums acquire examples of modem d'objets du XXe sidcle sont en effet fort ins-
artifacts after extensive use has accelerated the tables, et posent aux restaurateurs des probld-
degradation process? Should more emphasis be mes in6dits, qui tiennent entre autres d la
placed on documentation in order to preserve radioactivit6, d la d6gradation des polymdres et
information about the artifact if the artifact ir la corrosion de I'aluminium. Lorsqu'il ont d
itself cannot be preserved? How are we to r6soudre de tels probldmes, les restaurateurs
preserve and maintain this rapidly developing ne disposent donc que d'une formation rudi-
material culture? mentaire dans le domaine de la science des
mat6riaux, et se voient contraints d'appliquer
The aim of Symposium '91 was to approach des rdgles d'6thique qui ont 6t6 congues pour
these issues over a broad fiont. It involved des objets beaucoup plus anciens.
experts in science and technology, both from
within the museum field and outside, and in- La ddgradation des objets r6cemment acquis
cluded many museum professionals. We hoped force les conseryateurs de mus6es d revoir les
to outline the cunent state of knowledge and politiques qui pr6sident d l'6tablissement des
to set future directions for research. As the collections. Les mus6es devraient-ils acqu6rir
following pages show, these goals were accom- des moddles d'objets contemporains qui ont 6t6
plished. As for the conservation of modem grandement utilis6s et dont le processus de
materials, we have only just begun.... d6gradation est d'autant plus avanc6? Vaut-il
mieux privil6gier la documentation, la conser-
Editorial Committee vation de renseignements sur les objets, si
ces derniers ne peuvent pas 6tre conserv6s?
Comment r6ussira-t-on ir conserver les moddles
de cette culture mat6rielle qui 6volue si rapide-
ment, et d maintenir ces collections d'objets?

Ces questions devaient 6tre envisag6es, lors du


Symposium 91, dans le cadre d'une perspective
plus large. Aussi a-t-on fait appel tant d des
sp6cialistes des sciences et de la technique wu-
vrant d I'int6rieur ou d I'ext6rieur des mus6es
qu'aux professionnels de ces 6tablissements.
La conference visait d foumir un apergu des
connaissances acfuelles dans ce domaine, et d
d6cider des orientations que prendra la recher-
che dans le secteur. Les pages suivantes mon-
trent, de fagon 6loquente, que ces objectifs ont
6t6 atteints. Et, pour ce qui est de la conserva-
tion des mat6riaux modemes proprement dite,
force est de constater qu'elle n'en est qu'ir ses
d6buts...

Comite de r6daction
Table of Contents / Table des matiires

1. Modern Materials in collections / utilisation des mat6riaux modernes


au sein des collections

Les mat6riaux modemes au Mus6e de la civilisation d Qu6bec : Sylvie Marcil


un d6fi pass6, pr6sent et futur

Modern Materials in the Collection of the Musie t0


de la civilisation d Qutbec (abstract)

The condition Survey of Sound Recordings at the National Jan Michaels l3


Library of Canada: Implications for Conservation

Les enregistrements sonores d la Bibliothdque nationale du 21


Canada : une dtude d'etut et son incidence sur la conservation htsumd)

Plastics in the Science Museum, London: A Curator's View Susan Mossman 25

Les plastiques au National Museum of Science and Industry 34


le point de vue d'une conservatrice (resumi) -
Membership and Aims of the Plastics Historical Society John Morsan 5I

La composition de la Plastics Histoical Society et ses objectifs 40


(rdsumd)

2. Conservation Policies and Plans / Politiques et projets en matiire de conservation

A Joint Project on the Conservation ofPlastics by John Morgan 43


The Conservation Unit and the Plastics Historical Society

La consentation des plastiques : un projet mixte du Sentice de 50


consentation de la Museums & Galleries Commission et
de la Plastics Historical Society (rtsumt)

Conserving the Science Museum Collections Roger Price and 5l


Anne Moncrieff

La conservation des collections du National Museum 54


ofScience and Industry (rdsumd)
3. History of Technology / Histoire de la technologie

Rubber: Its History, Composition and Prospects for Conservation M.J.R. Loadman 59

Le caoutchouc : son histoire, sa composition et ses IJ


perspectives de conservation (r,!sumi)

Ardil: The Disappearing Fibre? MaryM.Brooks 8l

L'Ardil, lafibre qui disparait? (risumd) 9l

Plastics Found in Archives Alan Calmes 95

Les plastiques prdsents dans les archives (rdsumt) 102

4. Processes of Deterioration / Processus de d6gradation

Changes in Polymeric Materials with Time David M. Wiles 105

La transformation des matdriatn polymdriques avec le temps (resumt) lll


The Physical Aging of Polymeric Materials Christopher W. ll3
McGlinchey

Le vieillissement des matdriaux polymtriques (rdsumi) t20

La pr6vision du comportement d long terme de mat6riaux Jacques Lemaire 123


polymdres synth6tiques d'aprds des exp€riences de
vieillissement artifi ciel

Predicti on of the Long-Term B ehaviour of Synthetic Polyn eri c 133


Materials from Artificial Ageing Experiments (abstract)

Composition Implications of Plastic Artifacts: A Survey of R. Scott Williams 135


Additives and Their Effects on the Longevity of Plastics

La composition des objets en plastique et leur longdvitd (resume) 152

The Nature and Origin of Surface Veneer Checking in Plywood Mark D. Minor 155

Une dtude prdliminaire de I'origine et de la nature des t&


craquelures qui se produisent dans le placage des contreplaques (rdsumd)
5. Case Studies and Specific Problems with Materials/
Etudes de cas et problimes particuliers pos6s par les mat6riaux

Deterioration of Cellulose Nitrate Sculptures Made by Gabo Michele Derrick, 169


and Pevsner Dusan Stulik and
Eugena Ordonez

La dbgradation de sculptures en nitrate de cellulose exdcuttes 181


par Gabo et Pevsner (resumb)

A.W. McCurdy's Developing Tank: Degradation of an Early Plastic Robert D. Stevenson 183

La cuve de dbveloppement d'A.W. McCurdy : un exemple 187


de la ddgradation des premiers plastiques (resumt)

Degradation of Polyurethanes in 2Oth-Century Museum Textiles Nancy Kerr and 189


Jane Batcheller

La ddgradation des polyarithannes prdsents dans les textiles du 204


Xf sidcle conservds dans des musrles lrtsumi)

Treafing Early Regenerated Cellulose Textiles: Two Case Histories Emma Telford 207

Le traitement de textiles anciens en cellulose rdgindrte : 2ll


deux dtudes de cas (resumd)

Treatnent of 2fth-Century Rubberized Multimedia Costume: Clare Stoughton- 213


Conservation of a Mary Quant Raincoat(ca.1967) Harris

Le traitement d'un costume multimedia caoutchouti du Xf siicle : 220


la conservation d'un impermeable Mary Quant datant de
1967 environ ftisumd)

Spacesuits: NASA's Dream Conservator's Nightmare Mary T. Baker and 223


- Ed McManus

Les combinaisons spatiales : un r€ve pour la NASA, mais un 229


cauchemar pour les spdcialistes de la restauration (rtsumd)

Conservation of Paintings on Delaminated Plywood Supports Donald C. Williams 231


and Ann Creager

La conservation de peintures sur supports de contreplaqui 241


ddcollAs (rdsumb)

Chicken Bones and Cardboard: The Conservation of a Anton Rajer and 243
Collection of Eugene Von Bruenchenhein's Art Emil L. Donoval

Les os de poulet et le carton ou la conservation d'une collection 247


d' oeuvres d' Eugene Von B ruenchenhein (resume)
Ensuring a Future for Our Present High-Tech Past: Lessons from Jon Ecklund 249
the ENIAC for the Conservation of Major Electronic Technology and Beth Richwine

Garantir I'avenir des vestiges de notre haute technologie : 259


les leqons d tirer de I'expbrience de I'ENIAC pour assurer
la conseruation des grandes techniques de I'electronique (rdsumi)

Modern Metals in Museum Collections Lorna R. Green and 261


David Thickett

Les mdtaux modernes dans les collections de mus,le (rtsumil 270

Finishes on Aluminium A Conservation Perspective Chris Adams and 273


- David Hallam

L'application de rev€tements sur I'aluminium : un point de vue 284


axi sur la consewation (rdsumd)

Radiation Hazards in Museum Aircraft John Ashton 287

Les risques d'exposition aux radiations provenant d'aeronefs 295


exposds dans les mtndes (rdsumd)

Conservation of One of Alexander Calder's Largest Mobiles Albert Marshall and 301
Shelley Sturman

La restauration d'un des plus grands mobiles d'Alexander Calder (rdsumd) 305

Conservation of Weathering Steel Sculpture John Scott 307

La conseryation de sculptures en acier (rdsumd) 318

6. Testing and Development of Conservation Processes /


Elaboration et mise i I'essai des m6thodes de conservation

An Evaluation of Eleven Adhesives for Repairing Don Sale 325


Poly(methyl methacrylate) Obj ects and Sculpture

L'|valuation de I I adhtsifs en vue de leur utilisation pour la 337


rtparation d'obje* et de sculptures en poly(mtthacrylate
de mLthyle) (rtsumi)

Labelling Plastic Artefacts Julia Fenn 341

La pose des numdros d'enregistrement sur les objets en plastique (rtsumt) 349
Degradation Rates for Some Historic Polymers and the Potential David W. Grattan 351
of Various Conservation Measures for Minimizins Oxidative
Degradation

Le rythme de degradation de certains polymdres d'importance historique 360


et les diverses mesures de consenation qui permettraient de rdduire au
minimum la ddgradation oxydative (risume)

A Field Trial for the Use of Ageless in the Preservation of Rubber Yvonne Shashoua 363
in Museum Collections and Scott Thomsen

L'essai pratique de I'Ageless pour la conservation du caoutchouc dans 371


les collections de musde (resume)

La mise au point d'un traitement cathodique de stabilisation Christian Degrigny 373


de vestiges a6ronautiques immerg6s en alliages d'aluminium

An Electrolytic Treatment for Stabilizing Submerged Aluminium 379


Al loy Aircraft Fragments (abstract)

The Discoloration of Acrylic Dispersion Media James Hamm, 381


Ben Gavett, Mark
Golden, Jim Hayes,
Charles Kelly, John
Messinger, Margaret
Contompasis and
Bruce Suffield

La dbcoloration des medias de dispersion acrvliques (rtsum6) 392

7. Methods of Analysis and Identification / M6thodes d'analyse et d'identification

Practical Pitfalls in the ldentification of Plastics Helen C. Coxon 395

Les dfficultes pratiques liees d I'identification des plastiques (rdsume) 407

The Identification and Characterization of Acrvlic Emulsion Carol Stringari 411


Paint Media and Ellen Pratt

L' identification et la caractdrisation des peintures-imulsions 438


acryliques (rtsumd)
Modern Materials in Collections
Utilisation des mat6riaux modernes au sein des collections
Les mat6riaux modernes au Mus6e de la civilisation ir Qu6bec :
un d6fi pass6o pr6sent et futur

Sylvie Marcil

Muste de la civilisation
Qudbec (Quebec)
Canada

R6sum6 Le Muste ne dispose que de ressources limitdes,


si I'on prend en compte son calendrier d'exposi-
Les mattriaux modernes prennent de plus en tions chargt. Aussi son equipe de conseryation-
plus d'importance dans les collections du Muste restauration doit-elle s'en tenir d la mise en
de la civilisation d Qubbec, qui presente, sous dif- application de mesures prtfoentives et d des inter-
firents thimes, un patrimoine souvent assez ventions minimales. Et il lui tarde de mettre sur
rdcent. Lavaridtd de ces matdriata ne cesse de pied un programme prdventif global pour
croltre caoutchoucs, plastiques de tous genres, ces matdriaux.
-
laques et vernis, metuux et leur composition
-,
nous est souvent inconnue. Ils se retrouvent dans
presque toutes les collections alimentation, Introduction
-
accessoires de mode et de beautd, ameublement,
jeur etjouets, outils et dquipement, transports et En reprenant, au terme du pr6sent symposium,
communications dans des proportions vari6es.
- le sujet de la communication d'introduction,
ir savoir << Les mat6riaux modernes dans le
Qu'ils entrent dans la composition d'objets ex- mus6e >, nous voulons en quelque sorte boucler
posts ou mis en rdserve, ces matdriaux complexes
la boucle. Nous n'aurons pas n6cessairement
posent de nouveaux defis d I'equipe de conserva-
fait le tour de la question mais, en effectuant
tion-restauration, compte tenu du mandat qui a
ainsi un retour au point de d6part, nous serons
ttd confie au Musde par le ministdre des Affaires
culturelles du Qudbec. La prdsente communica- mieux en mesure de d6terminer si cette rencon-
tion Jitit prtcisiment ttat de ces ddfis. tre nous a effectivement permis d'avancer
un peu.
Dans le cas d'expositions, ces d6fis tiennent sur-
tout d des questions de responsabilitd Sans offrir de solutions ni tirer de conclusions
partagee entre deux seruices - qui est
et d'auitude, de des communications pr6c6dentes, le texte qui
-
m€me qu'au statut des objets eux-memes et aux suit pr6sente tout simplement une r6alit6, celle,
conditions ambiantes paniculidres qu' ils exigent. trds commwre, d'une conservatrice-restaura-
Dans le cas des collections, ils se posent, pour trice d'un mus6e qui collectionne l'histoire
ainsi dire, d toutes les 2tapes : depuis I'identifica- en devenir, dans divers secteurs, et qui utilise
tion et la numdrotation des objets ou l'tvaluation diff6rentes th6matiques pour f illustrer. Le
de leur ttat, jusqu'd leur nettoyage, leur traite- contenu t6moigne des d6frs que pr6sente la
ment et leur entreposage, et en passant par les conservation des mat6riaux modemes, pr6-
divers contr6les eux-m€mes tout au-
- des objets
tant que des conditions ambiantes des rtserttes
sents dans un nombre croissant d'obiets de

q ui do iv ent s' effec t u er.


- la collection.
Au fil du texte sont ins6r6es plusieurs photo- services du Mus6e et plus particulidrement
graphies : les figures I d 5 offrent quelques
-
celui des expositions ainsi que quelques
exemples de mat6riaux modernes en exposi-
-,
agences ext6rieures et, d I'occasion, le grand
tion; les figures 6 d 10 sont des exemples de public. Les interventions se font surtout ir la
la d6t6rioration de mat6riaux modemes et les r6serve, dans un atelier 6quip6 ir cette fin.
figures 1l e 13 offrent quelques exemples d'in-
terventions. Il est d noter que si certains objets Les mat6riaux modernes
ont le statut d'objets fabriquds ou proviennent au Mus6e de la civilisation
de collections priv6es, il s'en trouve aussi d'au-
tres qui ne constituent encore que des acces- D6chet hier, tr6sor aujourd'hui, d6chet au-
soires, mais qui pourraient bien 6ventuellement jourd'hui, tr6sor demain. I1 ne fait aucun
devenir des objets de collection. Tous ces doute que les objets modernes, faits de mat6-
mat6riaux se c6toient dans les expositions. riaux modernes, se retrouvent d6jd dans nos
expositions, et qu'ils entrent en grand nombre
Le Mus6e de la civilisation dans nos collections. Leur vari6t6 ne cesse de
croitre : caoutchoucs et plastiques de tous gen-
Le Mus6e de la civilisation est situ6 d Qu6bec res, laques, peintures et vernis, m6taux, cartons,
m€me, au bord du majestueux fleuve Saint- contreplaqu6s, textiles, etc. Ils se retrouvent
Laurent. Il loge dans un immeuble trds r6cent, dans presque tous les secteurs I'alimenta-
qui a ouvert ses portes en 1988. L'6quipe du tion, les accessoires de mode et- de beaut6,
Mus6e est jeune, dynamique et ambitieuse. Elle l'ameublement, I'art populaire, les jeux et les
r6ussit ir attirer quelque 800 000 visiteurs par jouets, les outils et l'6quipement, les transports
ann6e, en faisant en sorte que le Mus6e soit et les communications, etc.
ouvert et accessible d tous.
Ces mat6riaux sont trds complexes, h6t6rogdnes
La r6serve du Mus6e, oi sont entrepos6es les et en constante 6volution, et nous devons donc
collections, est situ6e dans la ville de Vanier, d d6sormais apprendre ir les conserver et 2r les res-
environ 10 km du Mus6e. Une importante par- taurer, au m6me titre que les autres. Il s'agit ld
tie du Service des collections y travaille, dont d'un d6fi de taille pour tous les conservateurs-
les 6quipes charg6es du registrariat, du cata- restaurateurs, mais plus particulidrement pour
logage, de la recherche et des prCts d'objets, ceux qui sont isol6s dans des mus6es oi les
ainsi que celle de la conservation-restauration. collections sont assez diverses et nombreuses.

La conservation-restauration I. Les expositions

L'6quipe de conservation-restauration se Outre les expositions itin6rantes, le Mus6e


compose d'une technicienne en conservation pr6sente une vingtaine d'expositions par ann6e
pr6ventive, Elisabeth Forest, et, depuis mai dans les l0 salles du bdtiment principql, qui
I 989, d'une conservatrice-restauratrice. Des occupent approximativement 5 700 m", et
techniciens en restauration viennent. au besoin, dans deux autres bdtiments, ir savoir la maison
se greffer d cette 6quipe. Le mandat que le Chevalier et I'entrepdt Thibodeau*. Une propor-
ministdre des Affaires culturelles a confi6 au tion marqu6e des objets expos6s se compose de
Mus6e n'est pas un mandat de recherche. Le mat6riaux modemes. Les deux tiers environ des
Mus6e est plutdt charg6 des fonctions pr6ven- objets de chaque exposition sont emprunt6s ou
tives, tandis que le Cenhe de conservation du acquis ir cette fin.
Qu6bec (CCQ) continue d'assumer en grande
partie la fonction curative. Les expositions ont une influence consid6rable
sur le travail de conservation-restauration et
Compte tenu de ce mandat et du roulement sur le d6veloppement des collections, qui com-
des expositions au Mus6e, le travail de conser- portent de plus en plus d'objets cr66s d partir
vation-restauration s'effectue plutdt ir un niveau de mat6riaux modernes. Ces mat6riaux sont
de consultation, de pr6vention ou d'interven- en effet trds pr6sents, car les thdmes qui sont
tion minimale. L'6quipe dessert tous les

4
Figure I : Autopsie d'un sacvert (Photo : Pierre
Soulard, Archives du Mus,le de la civilisation, 348-
PH-6). Figure 3 : Jeux (Photo : Piene Soulard, Archives du
Muste de la civilisation, 341-PH-3).

t:.' @
t;ht::';'
e;;.,*.:|3

Figure 2 : Ephemdre (Photo : Piene Soulard,


Archives du Musde de la civilisation. 330-PH-21).
Figure 4 : Quebec sur glace (Photo : Pierre Soulard,
Archives du Muste de la civilisation, 3 l9-PH-20).
abord6s lors de telles expositions I'alimenta-
tion, la beaut6 et I'environnement,-pour n'en Tous ces objets n'aboutissent pas n6cessaire-
nommer que quelques-uns 6tablissent une ment dans nos collections, mais il faut quand
comparaison entre le pass6 -et le pr6sent, tout en m€me assurer leur conservation pendant leur
explorant les tendances futures. s6jour au Mus6e. Et c'est de ld que proviennent
les diffrcult6s, qui tiennent surtout d des ques-
Cela met donc en valeur un patrimoine sou- tions de responsabilit6 et d'attitude, ou au statut
vent frds r6cent : depuis les ensembles de cos- des objets eux-m6mes. Le Service des expo-
m6tique ou un canot de fibre de verre d'une des sitions est, avec celui des collections, respon-
demidres courses sur glace jusqu'aux systdmes sable des objets. Vu la nature consultative de sa
audiovisuels qui ne sont pas encore sur le fonction, la conservation-restauration n' inter-
march6 chez nous. vient qu'au besoin, d diff6rents moments de la
planification. Un objet r6cent quel qu'en soit
le propri6taire ou la personne - responsable ne
-
abim6 pendant I'installation s'il 6tait coll6,
- La d6marca-
6pingl6 ou peintur6, par exemple.
tion entre I'artefact et I'accessoire n'est d'ail-
leurs pas toujours 6vidente. Il est difficile pour
les conservateurs d'6tablir la valeur d'une ac-
quisition en comparant les co0ts de gestion
et d'enheposage de I'artefact ir ceux de I'objet
neufet en prenant en compte les axes de
d6veloppement de la collection. Mais le d6fi
ne s'arr6te pas ld. Lorsque les objets ont int6gr6
les collections, ce sont d'autres pr6occupations
qui entrent en jeu.

II. Les collections


Actuellement, prds de 65 000 objets sont enhe-
pos6s d la r6serve du Mus6e, dans un espace
d'un peu plus de 3 000 m'. Il est difficile
Figure 5 : Souffrir pour €tre belle (Photo : Piene
d'6tablir le nombre d'objets qui comportent
Soulard, Archives du Musee de la civilisation,
des mat6riaux modernes, car leurs fiches ne
303-PH-21).
font pas toujours 6tat des mat6riaux qui les
commande pas autant de respect qu'un bronze composent, mais ils constituent, de toute 6vi-
de I'Antiquit6, par exemple. De plus, certaines dence, un pourcentage non n6gligeable de cet
pidces de m6me genre ont tantdt le statut d'ob- ensemble. On sait d6jd que les plastiques et les
jets de mus6e, ou d'artefacts, tantOt celui d'ac- caoutchoucs font partie, ir eux seuls, de plus
cessoires, et ne regoivent donc pas toujours de 2 000 de ces objets. Ce nombre augmente
la m6me attention, ce qui porte parfois d rapidement au fur et d mesure que progresse
confusion. la validation des fiches d'objets aux fins d'int6-
gration d la base de donn6es sur vid6odisque,
En ce qui conceme les conditions d'exposi- un systdme dont I'efficacit6 d6pend de la pr6-
tion (la lumidre, la temperature et I'humidit6 cision et de I'exhaustivit6 de I'information qui
relative), on opte plut6t pour la prudence. Le y est vers6e.
Mus6e possdde d'ailleurs un systdme 6volu6 de
contrdle, mais pour les salles d'exposition du Quel que soit le mat6riau, il faut donc I'identi-
bdtiment principal seulement. Les facteurs qui fier, num6roter I'objet, en d6terminer l'6tat, le
posent la plus grande difficult6 sont les pol- nettoyer, le traiter au besoin, I'entreposer dans
luants atrnosph6riques et la compatibilit6 entre les meilleures conditions possibles et, finale-
les mat6riaux entre ceux qui composent l'ob- ment, le contrdler. Beaucoup plus facile d dire
jet lui-m6me, -de m6me qu'entre ces derniers et qu'd faire, surtout lorsqu'il s'agit de mat6riaux
les mat6riaux de construction environnants. De modemes. Pour y arriver, il faut avoir une
plus, il arrive souvent que I'on ne connaisse ni connaissance approfondie de la nature des
la composition des objets ni celle des objets mat6riaux et disposer de bons outils de travail.
environnants, et on ne peut donc pr6voir les La tdche est 6norme et les priorit6s, difliciles
r6actions qu'ils auront dans un montage of ils d 6tablir.
se c6toient.
1. L'identification
Il est par ailleurs possible que, au terme d'une Dans un tel contexte, I'identification est le
exposition, I'objet passe du statut d'accessoire premier probldme qui se pose. L'absence
d celui d'artefact. On essaie toutefois de d6ter- d'identification des mat6riaux rend difficile
miner son statut futur avant le montage, pour le catalogage et les autres 6tapes de la mise en
s'assurer qu'il sera, au besoin, ad6quatement r6serve, de m6me que la r6daction des vignettes
restaur6 et pour 6viter qu'il ne soit inutilement accompagnant les objets.

6
L'identification exige une formation en science
des mat6riaux et un bon sens de l'observation.
MOme si on reconnait la pr6sence d'un
mat6riau moderne, il faudra faire appel ir un
conservateur-restaurateur, qui est souvent le
seul d poss6der la formation qui permettra de
I'identifier. Ce sp6cialiste ne dispose n6an-
moins pas toujours de tous les outils dont il a
besoin pour trouver les r6ponses, et il dewa
donc trouver des moyens simples et rapides
pour y arriver : tests. 6quipements. ouvrages
de r6f6rence, etc. Si on veut que le catalogage '#w
suive le rythme des acquisitions, on ne peut, en
effet, se permettre d'effecfuer une recherche Figure 6 : Ornement de Nodl en forme de
pouss6e pour chaque objet. Aussi une forma- violon (Musde de la civilisation, 89-1679). La laque
tion plus approfondie et continue facilite-t-elle prdsente de grandes fissures, qui laissent paraitre le
verre soffid sous-jacent. (Photo : Elisabeth Forest)
le travail.

2. La num6rotation
Lorsque le mat6riau a 6t6 identifi6, il se peut
que I'on ne puisse pas num6roter I'objet en em-
ployant la m6thode traditionnelle de I'encre et
du vernis qui risquerait de faire fondre le
mat6riau. -C'est le cas de certains plastiques
les nitrates cellulosiques, par exemple -
caoutchoucs, surfaces vernies et m6taux -, laqu6s.
Il est donc bon de faire part de cette mise en
garde aux personnes qui sont charg6es de
cette 6tape, s'ils n'ont pas d6jd regu la
formation pertinente.
Figure 7 : Poupde (Musie de la civilisation, 79-138)
Dans ces cas, nous utilisons parfois des constitude d'un mattriau composite, possiblement
6tiquettes de polyester dactylographi6es, qui une pdte de bois et recouverte d'une couche de
servent habituellement d la num6rotation des plastique non identifie. Ce recouvrement craque
textiles. Nous les fixons ir I'objet avec un et se souldve. (Photo : Elisabeth Forest)
adh6sifsoluble d I'eau en I'occurrence, du
Cellofas - plutdt mal d cer-
mais il adhdre
-, Nous n'avons pas encore, de
taines surfaces.
toute 6vidence, trouv6 la m6thode id6ale.

3.L'6tat de I'objet
Contrairement d ce qu'on aurait pu imaginer, la
plupart des mat6riaux modemes mis en r6serve
au Mus6e paraissent stables et ne montrent pas
de signes de d6t6rioration avanc6e. Une 6va-
luatign trds sommaire a permis d'6tablir que la
majorit6 des mat6riaux serait en trds bon 6tat.
Ceci r6sulte probablement du fait que plusieurs
acquisitions d'objets modernes sont r6centes, et Figure 8 : Poupie (Musde de la civilisation,
qu'on a choisi, dans la mesure du possible, des 88-6495) de caoutchouc rembourr,le. Le caoutchouc
obiets en bon 6tat. sous tension se fissure et laisse paraitre la bounure,
probablenent de laine. (Photo : Elisabeth Forest)
4. Le nettoyage
Le nettoyage demeure I'op6ration la plus
courante d laquelle sera soumise I'objet. Le
temps, les ressources financidres et les donn6es
sont souvent trop limit6s pour pousser plus loin
I'intervention. On s'en tient donc d de simples
tests, ir I'aide de m6thodes et produits usuels.
Il est certain que cette 6tape est plus facile si le
materiau d 6t6 identifie. En cas d'effet negatif,
on se limite au d6poussi6rage.

Lorsque la m6thode de nettoyage a 6t6 d6fi-


Figure 9 : Jeu de parchesi (Musee de la civilisation, nie diff6rentes questions se posent. Jusqu'or)
89-1655). Le papier vemi est ddchird, ainsi que les doit-on aller? Faut-il conserver tous les
bandes de tissu du centre et des c6tds. (Photo : 6l6ments? Doit-on disposer syst6matiquement
Llsabeth I. ore.st )
des contenus d caractdre pharmaceutique ou
alimentaire, par exemple? Et comment doit-on
s'y prendre?

.\rf

Figure l0 : Parapluie (Collection pivde). Les cou-


leurs du tissu ont pdli et le bouton de plastique est ,l/
compbtement Jis surd. (P hoto : Elis abeth Forest) ,,,
't1,"
'\

Les manifestations de d6t6rioration, lorsqu'il


s'en trouve, sont caract6ristiques : les plas-
tiques et les caoutchoucs craquent, durcissent,
jaunissent, suintent; les laques, les peinfures et Figure I 1 : Rasoir (Collection privee). La plupaft
des interventions se limitent d un nettoyage de
les vernis craquent, se souldvent, s'effiitent; les
sudace lorsque le matdriau rdagit bien.
m6taux se cassent, se corrodent, et leurs pla-
cages tombent; les cartons se d6chirent, plient,
se d6forment etjaunisssent; les textiles se
5. Le traitement
tachent et se d6chirent; les contreplaqu6s fen- Un simple nettoyage n'est parfois pas
dent et se d6laminent, etc. Certains mat6riaux suffisant pour assurer la stabilit6 d'un objet.
sont plus touch6s les caoutchoucs et les Si on connait la cause de la d6t6rioration, on
-
plastiques mous, certiaines laques et les cafions
peut parfois intervenir. Il est toutefois difficile
de formuler un traitement et d'en 6valuer les
notamment. L'utilisation qui a 6t6 faite de
l'objet ir I'origine, combin6e d la nature du cons6quences sans avoir de solides connais-
mat6riau, est responsable en grande partie de sances du mat6riau. Les ouwages de r6f6rence,
cette d6terioration. rares, 6pars et souvent trop th6oriques, ne sont
pas d'un grand secours. On se r6fdre plutdt d
des colldgues plus exp6riment6s. On doit
aussi affronter des questions d'6th!que, car il
est difficile d'etablir un principe directeur
d'intervention qui tienne compte de l'intention
du fabricant et des autres consid6rations pour
la conservation.

Comme pour le nettoyage, les quelques inter-


ventions curatives qui s'effectuent au Mus6e
sont limit6es, compte tenu du manque de temps
et de connaissances approfondies. Les r6sultats
sont de ce fait in6gaux. La diversit6 des mat6-
riaux ne nous permet pas de passer en rerv'ue
tous les geffes de traitements qui ont 6t6 mis d
I'essai, mais nous en retiendrons tout de m€me
quelques-uns.

Figure 13 : Le traitement, fait au Centre de


conseryation du Qutbec (CCQ), a permis de
stabiliser la capote de cuirette < American Leather >
avec Ltn entoilage de mousseline de soie colle avec
du BEVA et de reiquilibrer les roues en adaptant
des rubes de ndoprine. (Photo ; Michel Elie. CCQt

Les caoutchoucs wlcanis6s, mousse ou


latex demeurent- non trait6s pour I'instant.
On les- protdge avec du Armor-All, quand
ils sont en bon 6tat, mais on remplace aussi cer-
taines parties manquantes avec des mat6riaux
6quivalents modemes, ce qui est parfois plus
rapide et plus 6conomique, et qui permet mieux
de respecter I'original et son utilisation. Il
faudra voir comment de telles interventions
resisteront aux ann6es.

6. L'entreposage
Determiner les meilleures conditions d'en-
treposage possibles n'est pas toujours facile
Figure 12 : Carrosse (Musee de la civilisation,
89-1655). La capote Atuitfendue et trois des lorsqu'il s'agit de mat6riaux modernes. Bien
cerceaux de caoutchouc manquaient aux jantes. que les rdgles g6n6rales d'entreposage s'ap-
pliquent; il faut d6terminer les besoins par-
ticuliers des mat6riaux en ce qui d trait d
Les meubles, les valises ou les jouets, par exem-
ple, ont souvent un recouvrement qui est fait de I'humidit6, d la temp6rature et ir la lumidre, et d
leur protection contre les polluants atmosph6ri-
textile enduit, de plastique (du vinyle, notam-
ques. L'identification des mat6riaux est 6gale-
ment) ou de carton lamin6 et parfois vemi. On
ment primordiale pour 6tablir leur compatibilit6
r6ussira d refermer les fentes dans les textiles d
avec les autres mat6riaux environnants.
I'aide d'une doublure de soie de polyester col-
l6e avec du BEVA, d faire adh6rer le vinyle au
On a d6jd beaucoup am6lior6 I'entreposage
bois avec du Rhoplex AC-33 et i consolider
g6n6ral au Mus6e, et on continue de le faire,
des cartons d6lamin6s avec du Cellofas.
mais on n'a toujours pas de systdme de contr6le
environnemental adapt6 aux collections, et on de I'ICOM sur les mat6riaux modernes est
ne s'est pas encore pench6 sur les besoins par- une excellente initiative et un grand pas dans la
ticuliers des mat6riaux. Le fait que les objets bonne direction. Il regroupe une foule de ren-
soient souvent compos6s de plusieurs mat6- seignements pertinents. Je suis certaine que le
riaux vient compliquer les choses. L'espace symposium aura 6galement 6t6 profitable. Il
diminue d6jd d rue d'ail avec les acquisitions, s'agit < simplement > de continuer dans cette
surlout quand on doit entreposer s6par6ment direction. Maintenant que les probldmes et
chacun des el6ments de I'objet (un jeu en boite, leurs causes sont mieux connus, les conser-
par exemple, ori l'on retrouverait des pions en vateurs-restaurateurs isoles misent davantage
m6tal et en plastique dans un contenant de sur leurs collegues pour pousser plus loin leur
papier et de carton). Dewait-on enfreposer recherche, et pour publier les r6sultats de leurs
ces 6l6ments par mat6riau, si les mat6riaux travaux, autant les bons que les mauvais.
constituants sont incompatibles? La question
demeure sans r6ponse pour I'instant. Il y a beau- Remerciements
coup d'arguments contre cette option, d'ordre
logistique notamment. J'adresse mes trds sincdres remerciements d
Elisabeth Forest, qui participe activement au
7. Le contr6le travail dont j'ai fait 6tat dans la pr6sente com-
Puisque la plupart des mat6riaux modernes de munication, et qui a ex6cut6 bon nombre des
la collection du Mus6e semblent, pour l'instant, photographies qui ont 6t6 pr6sent6es. Les autres
en bon 6tat, il importe de nous concentrer sur diapositives ont 6t6 tir6es des archives du
I'aspect pr6ventif. Dans un avenir rapproch6, il Mus6e. Je tiens 6galement ir remercier les
faudra d6finir un plan d'6valuation global pour manutentionnaires Ga6tan Gigudre et Lise
ces mat6riaux et en assurer un certain contr6le. Dionne, responsables de la num6rotation et de
Cette 6tape sera facilit6e par le vid6odisque, la mise en r6serve, qui ont emball6 les objets
mais compliqu6e par le fait que ces mat6riaux qui n'ont malheureusement pas pu 6tre pr6sen-
sont dispers6s dans I'ensemble de la collection. t6s en d6monstration.
L'ail averti des conservateurs responsables
de secteurs sera d'une grande utilit6 pour les Note
rep6rer et pour 6tablir les priorit6s de traite-
ment. Encore faudra-t-il trouver des traitements xPour obtenir de plus amples renseignements
qui soient efficaces! techniques. pridre de communiquer avec :

Conclusion Sylvie Marcil


334, rue de la Tourelle
Ilest 6vident qu'on ne peut ignorer I'arriv6e Qu6bec (Qu6bec) clR lC8
continuelle de nouveaux mat6riaux au Mus6e. T6l6phone : (418) 648-1590
Pour un 6tablissement poss6dant une telle quan- T6l6copieur : (418) 529-4195
tit6 et une telle diversit6 d'objets, la conserva-
tion-restauration constitue un d6fi 6norme, et Abstract
les besoins sont proportionnels. Pour bien s'ac-
quitter de cette t6che, il faut s'attaquer globale- Modern Mqterials in the Collection of the Musie
ment au probldme, mOme si cela ne semble pas de la civilisation
6vident de prime abord.
Modern materials have become increasingly im-
Le conservateur-restaurateur ne peut 6tre un portant in the collections of the Musde de la civili-
sp6cialiste de chaque materiau qu'il rencontre. sation, which presents a thematic via,u of recent
Il history. The composition of these artifacts is often
faut qu'il puisse compl6ter sa formation par
unlmown, is becoming increasingly diverse, and
des ateliers ou des s6minaires sp6cialis6s. Il
includes such materials as rubber, plastics, lac-
faut aussi qu'il ait facilement accds ir une infor- quers and vanishes, and metals. These materials
mation regroup6e d I'int6rieur de bibliographies can be found in varying propotlions in almost
choisies, etc. Le bulletin du groupe de travail every area ofthe collection: food, fashion and

10
beauty accessories, fumiture, games and toys, attihtde as much a,s to the status of the artifacts
tools and equipment, transportation, and contmu- themselves and to the particular imbient iondi-
nications. tions thqt require. In the case ofcollections, these
challenges exist at all stages,for example, identi-
Whether one considers the composition of dis-
fiing and numbering the objects; evaluating their
played or of stored artifocrs. th'ese ,o*pi", n,ot"- state; cleaning, treating, and storing them; and
rials present new challenges to the conservation monitoing
team. This is particularlt' true in light of the man- - of the objects themselves as much
as of the ambient storage conditions.
date given to the Museum bv the Quebec Ministry
of C u lt ura I Affa i rs. Th i s p re s e nt a t i o n ou tIin es The Muste de la civilisation has limited re-
these challenges. sources, considering its heavy exhibition sched-
ule. Hence, its team of consentators must aclhere
In the case ofexhibitions, these challenges mainly to preventive measures and use minimal interyen-
relate to issues of responsibility, which is sharecl tion. A global program of prcvention.for these
bv two departments within the museum, and of materials has yet to be established.

il
The Condition Survey of Sound Recordings at the National Library of Canada:
Implications for Conservation

Jan Michaels

The National Library of Canada


Ottawa, Ontario
Canada

Abstract Introduction
The National Library of Canada is the deposit li- The phonograph was invented in 1877 by
braryfor, among other things, published sound Thomas Edison. Like many inventors of
recordings in Canada. In the summer of 1990, as genius, he claimed that it was by "the merest
part ofa three-year program, a condition survey accident":
of sound recordings was conducted of National
Library collections. Separate suryeys of reel-to- I was singing to the mouthpiece of a tele-
reel tapes, cylinders, LPs, 45s and 78s were done. phone, when the vibrations of the wire sent
the fine steel point into my finger. That set
Information was obtained, which will assist in me to thinking. If I could record the actions of
collection management as well as in preservation the point, and then send the point over the
and conservation planning. Shelving methods and same surface afterwards. I saw no reason whv
containers were examined for suitabilin and con- the thing would not talk.
dition. Lignin and pH testi were conducted on
containers made of paper-based materials. As for I tried the experiment, frst on a strip of tele-
the sound recordings themselves, base materials
graph paper and found that the point made an
and oxide lavers were identified. For open reel
alphabet. I shouted the word 'Halloo! Hal-
tape, rub and smell tests were canied out, as well
loo ! ' into the mouthpiece, ran the paper back
as physical examinationfor over 20 categories of
over the steel point, and heard a faint 'Halloo!
damage, including creasing, s tretching, flaking, Halloo!' in retum. I determined to make a ma-
blocking and plasticizer migration. Similar tests chine that would work accurately, and gave
for the other mediawere conducted as appropri- my assistants instructions, telling them what I
ate. For discs, groove wear also was examined. had discovered.

This paper discusses the suruqv's results and They laughed at me. I bet fifteen cigars with
implications. In the short term, our new-found in- one of my assistants, Mr. Adams, that the
formation will help in planning preservation ac- thing would work the first time without a
tivities including re-housing and copying. The break, and won them. That's the whole story.
survey also directs attention to unresolved conser- The discovery came through the pricking of a
vation issues, both technical and ethical. An unre- finger.'
solved ethical issue remains: is the aim to repair
in order to play an item one more time, or to con- From that "merest accident" was born a
serveforposterity? The paper concludes with a technology that has transformed cultural life.
discussion of the National Library's attempts to
come to gips with the problem of establishing a
discipline of sound recording conservation.

13
This transformation is despribed by an anony- major media in the collections and to do some
mous Canadian in Moogk': sound tests, too. In the end, within the allotted
time, only the cylinders, 78s, 45s, LPs and reel-
Records have done more to spread music and to-reel tapes were surveyed. No sound tests
culture to the far-flung comers of Canada, were conducted. A decision on whether or not
and indeed the world, in the first fifty years to complete the survey as originally planned
of its [sic] invention, than personal perform- has not been made. Costs will have to be
ances did in all the centuries that passed be- weighed against the potential value of any
fore. Until the advent of radio, and even for
additional information.
many years after that, the gramophone was
the main source of entefiainment for the
The surveys of the discs and magnetic tapes
settler, the hunter and happer, the isolated
were conducted using a stratified random sam-
farmer, and the little hamlets and communi-
ties in the Canadian north and far west.
ple with approximately 400 items per stratum.
This provides a95oh confidence level. (Further
The National Library, as keeper of the Cana- information on thg sampling method can be
dian published heritage, has been receiving pub- forurd in Bullock.')
lished sound recordings through legal deposit
since 1969. In order to fulfill its mandate, it has In the questionnaire we recorded the physical
purchased or received as gifts or donations tens condition ofthe sound recordings and their con-
of thousands more recordings that were made tainers. Base materials and oxide layers were
prior to legal deposit. The National Library also identified. Over 20 categories of damage to the
has thousands oftapes recorded by or for Cana- recording media were surveyed. pH of the con-
dian composers and performers. As well, there tainers was assessed using Phydrion Insta-
are talking recordings, children's tapes, Cana- check surface pH pencils and lignin content
dian National Institute for the Blind cassettes using the phloroglucinol test.
and educational kit recordings. The Library has
piano rolls, cylinder recordings, 78 rpm records Results and Implications
(78s),45 rpm records (45s),337: rpm records
This paper briefly discusses some results of the
(also known as long-playing records or LPs),
disc and reel-to-reel tape surveys. The cylinder
reel-to-reel magnetic tapes, cassettes, 8-track
survey is not discussed. Detailed analysis of all
cassette tapes, compact discs (CDs), and music
the sound surveys will be available at a later
videos. All told, the National Library of
date.
Canada's Recorded Sound Collection now
holds over 120,000 recordings of Canadiana.
Discs
In the summer of 1989, as the fust step of a
78s: The phonograph was invented in 1877.
new initiative in preservation planning, the
Twenty years later, in 1897, shellac was intro-
National Library undertook the examination of
duced as the major component in the disc. The
the nature and condition ofall its collections.
78 rpm disc was the principal commercial re-
Starting with eight categories of printed items,
cording medium of the 1930s and 1940s."
the surveys continued, evaluating microforms
McWilliams describes these shellac recordinss
and sound recordings. In 1992 the last ofthe
in the following way:
suryeys, manuscripts, will be completed. Once
all surveys are concluded and the analyses fin- The term shellac, as used in record manufac-
ished, the National Library will have a compre- ture. did not mean at that time or sub-
hensive picture of the state of its holdings and sequently
-
a disc made entirely of shellac.
the areas of greatest threat. This will greatly - a convenient way of referring
It was, rather,
facilitate preservation management. to a compound material. Shellac contained
fillers, such as limestone or slate, pigment
In some senses, our sound recording survey (usually carbon black), lubricants, such as
was the least successful of the surveys carried zinc stearate, and binders and modifiers, such
out. Ambitious plans were made to survey all as Congo gum and vinsol.'

t4
Forhrnately, most shellac discs are really Of the inner sleeves, 92oh are originals. Sleeves
very stable. McWilliams goes on to say, composed of paper constitute 55% while those
"Properly formulated shellac cures as it ages of paper with plastic liners make up 260/o. Of
a cross-polymerizatron occurs, which guaran-- the paper sleeves, 97o/o are acid paper and 93Yo
tees good long-term life."o have a surface pH of3 or 4.

However, McWilliams has reported that the Unfornrnately data indicates that all of these re-
composition of shellac compounds deteriorated cordings also are damaged: 74%o are scratched
during the World Wars when shellac supplies and99o/o are dirty. Figure 2 depicts the major
were intemrpted. As a result war-year discs forms of damage to these 3373 rpm recordings.
may be less stable.
45s: The 45 rpm record with its large-diameter
Despite their long-term stability, shellac discs spindle was f-rst introduced by RCA in the
are nonetheless quite brittle and relatively easy 1950s. Currently, these 45s are stored in their
to damage. Results indicate that all of the Na- original outer jackets. Results indicate that99%o
tional Library's collection of 78 rpm recordings of the National Library's outer jackets are acid
are damaged because of dust and dirt. Figure I paper. All of the 45s show some kind of dam-
depicts the major forms of damage to these 78s. age. Figure 3 depicts the major forms of dam-
age to these 45s.

Overall Comments
Records are published with jackets on the
DIRT
outside and sleeves as innerprotectors. Data
SCRATCHES
indicates that99o/o of all types of discs in the
collection, as a group, are in acidic recordjack-
FINGERPRINTS ets. Of the inner liners, 92o/o are originals. Of

STAINS

CHIPS

BTJBBLES
DIRT

GOUGES
SCRATCHES
EDGEFLAKE
FINGERPRINTS

BUBBLES

STAINS

Figure I Damage to 78 rpm recordings. CHIPS

WARPAGf,
LPs: The long-playing disc or LP was first
manufactured in the early 1950s and by the 20 & 60 80 100 120
1960s was ubiquitous. The National Library
collection is not stored on proper shelving or I Pecent with damage

with adequate support. Results indicate that


99%o are still stored in their original outerjack-
ets. Of these outer jackets, 98o/o are composed
of acid paper, and 8lo/o are of ligneous paper. Figure 2 Damage to 331/z rpmrecordings

l5
drag generates enough heat that the plastic
partially melts (though not enough to de-
form), causing a microscopic flow around the
stylus into which dust can be embedded
DIRT
permanently.d

SCRATCIIES
Record player needles cause additional damage
to recordings because of the friction between
FINGERPRINTS
the record grooves and the stylus. The friction
contact is needed to reproduce the signal, but it
STAINS
is also harmful because it causes gradual defor-
B[MBLES
mation or bending of the grooves. The result is
a progressive distortion that builds up from one
playback to the next.'This is why it is recom-
clilPs
mended that a record only be played once in
WARPAGE
any 24-hour period.

20 40 60 80 100 r20 The single greatest problem in the,Breservation


of disc recordings is groove wear.'" This dam-
I Percent with damge
age is not visible except under a microscope.
Another major form of damage to discs, and the
second most common found in the survey, is
scratching. Most scratches occur as a result of
Figure 3 Damage to 45 rpm recordings. use. As playing is one of the major causes of
damage to recordings, a priority must be the
the inner sleeves, 58Yo are paper,97o of which minimizing of damage to originals. Clearly the
are acidic. In general, paper inner sleeves best solution to the potential damage caused by
should be avoided because paper breaks down dust is not to play the recordings. A second-
over time, contaminating the surface and best solution is not to use a stylus to play the re-
grooves of the recording with dust-like paper cordings. This is no longer impossible. Within
debris. Obviously acidity in the paper acceler- the last year a record player has become avail-
ates this breakdown. As well, there is a possibil- able commercially that uses a laser beam rather
ity that ligneous paper can damage recordings.' than a stylus to play the disc. This revolution-
ary piece of equipment has taken almost a dec-
Clearly, we are facing a massive re-sleeving ade to develop. The National Library recently
program. As well, the original jackets will acquired the first such laser turntable outside
require de-acidification and storage space. Japan.

Results indicate that99o/o of the discs are dirty. The ELP laser turntable uses five laser beams
As dust easily can be imbedded permanently to track the record. It can compensate for warp-
into the plastic, it is considered damage for age, discs ofvarious sizes and speeds, certain
purposes ofthe survey. This high level ofdust types ofgroove damage or variable groove
must be related to the inadequate air filtering widths. Even discs that are flakins can be
system in the Library as well as the large num- l
played successfully with this systim.l Use of
ber ofacidic sleeves andjackets. According to this turntable will assist in minimizing damage
St.-Laurent, when a record is played: to the collection.

... only a small point of the stylus is actually Fortunately the long-term prognosis for discs is
making contact with the groove walls. One quite good. Their predicted life expectancy is
and a halfgrams ofstylus pressure on such a
more than a century when stored in ordinary
minute surface translates to several tons
library environments. The Pickett and Lemcoe
ofpressure per square inch. The resulting
report states:

t6
The actual potential storage life with respect Many tape recordings from the 1950s will be
to chemical degradation of an individual disc found on acetate-base tape. Cellulose acetate
is dependent on its exact formulation (includ- is unstable. It will eventually crumble, de-
ing both kinds and amoult of stabilizer and stroying the sound recording, or cause
extender used) and its thermal history prior to patches of the magnetic coating to fall off, de-
acquisition (including processing and mold- stroying areas ofthe recording. Acetate-base
ing). Apparently, small changes in these pa- tapes may hold up well for.years but eventu-
rameters can change the potential storage life ally they will self-destruct.' '
with respect to chemical degradation by sev-
eral decades.'' Polyester or pre-stretched poly(ethylene
terephthalate) (PET) was reserved initially for
Unfortunately evidence does exist that 45s may products exposed to severe conditions, such as
not be so permanent. Very early 45s seem to be those used in the military. Its use spread very
quite stable, but later, polystyrene was com- gradually over a period of 10 years beginning
monly used for 45s due to greater economies in around 1960. Until about 1970. both cellulosg^
manufacture over vinyl. These polystyrene triacetate and polyester tapes were produced.'o
discs are inherently unstable. They fiacture rela-
tively easily." The outer layer of the record sw- Experts project a life expectancy (i.e., the ac-
face can peel offat the area ofcontact like ceptable maintenance of mechanical properties)
old house paint.'' Thoush the National- Li- for cellulose triacetate at approximately 300
brary's coilection of 45s"does not appear to be years, and for PET at several millennia. Under
in worse condition than the other discs, this adverse storage conditions, the polyester base
may be a result of benign environmental condi- appears to be much more stable than the acetate
tions and relatively short life rather than any in- base. Degradation ofcellulose acetate at very
herent longevity in their makeup. Clearly it is high humidity (RH 80%) is extremely rapid.
important to conduct further research before the Cellulose acetate tapes stretch with dampness
45s begin to self-destruct. to produce waviness and contract with dry-
ing.'' This is corroborated in our results: the
Reel-to-Reel Tapes cellulose acetate tapes were reported to be
curled much more frequently than the polyes-
Magnetic recordings have existed since the end ter. Conversely, as the production of acetate
of the l9th century. However, the first success- tape requires the use ofplasticizers, overly dry
ful tape recordings were not demonstrated until storage conditions contribute to the loss of
1935 atthe Berlin Radio Exhibition. Harold these aggnts and the film becomes extremely
Lindsay built the first Ampex machine in 1947, brittle."'As well, cellulose acetate is decom-
successfully introducins the medium to the posed by acids and alkalis.2l This is of great
I5
United States. Low-piiced-urits became concem as results indicated that 93% ofthe
available in the mid-1950s.16 tape storage boxes are acidic. The thin base
layer has a front coating, namely the binder,
There are usually three layers in a magnetic which contains a magnetic recording material,
tape: the base or substrate onto which the which is commonly ferric oxide. Brown er al.
binder and recording material is coated; the decribe these binders:
binder, which bonds the recording material to
the base; and the recording material, which is The magnetic material is embedded in a poly-
capable of being magnetized and contains the meric binder. Common binders are based on
information recorded. polyester polyurethanes. Sometimes there is a
carbon coating on the back, which is also em-
The survey indicated that in the National bedded in a polymeric binder. The carbon
Library's collection 78o/o of the tape bases are coating dissipates electrostatic charge.
polyester, 20Yo are cellulose acetate and only
The polyester pollurethane binders used on
l% is paper. McWilliams states:
tapes are highly cross-linked materials with
complex structures. They can be expected

l7
to have 5-10 times as many ester groups as
urethane groups. Aliphatic esters are used;
these are more susceptible to trydrolytic degra-
dation than the PET substrate." CREASES

OXIDE RUBS OTF


And in a later paper Smith et al. state:
LOSS OF OXIDE

Other materials such as lubricants, adhesives, SCRATCHES


or stabilizers are frequently added to this
TEARS
layer. Polyester polyurethane is subject to
auto catallic hydrolysis, which was expected BRITTLE
to limit the lifetime of the tape. Some types
CTJRL
of magnetic tapes contain chromium oxide
particles, which do affect the chemical degra- STICKINESS
dation mechanisms. This has recently been re-
EDGE WAVINESS
ported by researchers at IBM."
VINEGAR SMELL

In order for the magnetic tape to work properly,


the binder must adhere so frrmly to the base ma-
terial and to the iron oxide that it will resist the I Percent with daroge

stresses ofplayback and storage without craz-


ing, flaking, or peeling. It must maintain these
properties despite chemical degradation and
loss of residual solvent or plasticizer.'- Brown Figure 4 Damage to reel-to-reel tape recordings.
et al. indicate:
Sound recording experts have recommended
Thus the degradation of interest in magnetic that tape should not be stored on plastic rppls.
tapes is primarily that of the binder. On degra- Metal, unslotted reels are recofirmended.'"
dation this softens; adjacent layers oftape These solid, unslotted reels prevent uneven
may stick together or the binder may stick to exposure to the environment. Waites
the recording heads. Ultimately the tape be- explains:
comes unreadable."
Because the hub is the strongest and most sta-
The binder is more sensitive to hvdrolvsis than ble part of the reel, it is the best means of reel
PET.26 The^short estimated useful tapi lifetime support during storage. When the reel is sup-
of20 years'' is directly related to the failure of ported by the hub, there is little if any weight
the binder layer. In some cases, binder break- resting upon its flanges. This protects the
down has been discovered in items only five flanges from problems such as bending and
years old.28 nicks.

The results of the reel-to-reel tape survey were Under no circumstances should a reel be
very worrisome: for 30% of the tapes, the oxide stored resting on its flanges. Paper notes
rubs off; 23o/o cannot be played. Figure 4 indi- about stored data or other sources of
cates damage to the tape recordings. contamination should not be put in the
storage container.3l

The reels onto which tapes are wound can be


another source of damage. Results indicate that The Association for Recorded Sound Collec-
690/o of the National Library tape reels are plas-
tions (ARSC) report also recommends that
tic and 3lo/o are metal. Waites indicates that the liner notes pg stored separately from the sound
reel's "...basic function is to protect the tape recordings." The effect of acid migration on
from damage and contamination. It is often the tapes is reflected in the condition survey re-
reel itself. damaged througl mistreatment. that sults. The pH of paper notes for all items
in turn damages the tape.'Zg where the oxide rubbed off was reported to be
5 or less.

l8
Clearly the potential loss of information that is Pickett and Lemcoe recommend a surveillance
indicated by the results is distressing. Informa- procedure of inspection and test based on stabi-
tion is quickly disappearing in large amounts. lizer exhaustion as the most feasible means of
Unforfunately very little research is being car- determining the need for re-recording these
ried out to allay this worldwide problem. Gerry discs. They say:
Gibson of the Library of Congress said in 1989:
...there is evidence that detectable changes in
With the exception of work now being carried the chemical composition of the record can be
out on magnetic tape for the National Archives used to indicate incipient failure due to chemi-
[in the United States] by the National Bureau of cal deterioration, although more information
Standards (United States et al. 1986), reported is needed to develop analytical techniques
evaluation of Sony's optical Century Media data and surveillance procedures. This aspect of
by the NBS, and of the effects of flre upon the problem might be made the subject ofad-
sound and audiovisual recording supports by the ditional study.rl
French Ministdre de la culhre et de la communi-
cation for the Bibliothdque nationale (Paris) Thirty years later, analysis ofresidual products
(Fontaine 1987), virtually no independent work is being pursued in France by Fontaine with the
is going on at this time on topics directly related intention that these can be used as indicators of
to audio preservation. Further, relatively little is degradation.36
known about the presewation, conservation, or
aging problems or properties of sound rec^ord-
A remedy for another area of concem for tape
ings from directly related scientific strldy."
recordings, binder breakdown, is feasible.
In fact, the 1959 Pickett and Lemcoe study Commercial firms can, through a carefully
Preservation and Storage ofSound Record- controlled time-temperafure cycling process,
reverse the binder breakdown. Kent describes
ings,long out of print, remains the basis of
the technique:
most^o^f the conservation knowledge in the
field."
Heat allows the binder system to rebond
temporarily. The treatment is not permanent,
Conservation Issues but gives at least a 30-day time window in
which to work with the tape. Provided the
Pickett and Lemcoe point out that shellac and time-temperature cycling is done correctly,
poly(vinyl chloride) (PVC) discs deteriorate there is no measurable high-frequency loss,
in opposite ways. Whereas shellac discs will noise increase or increase in print-through.
slowly, progressively become embrittled even Treated tapes appear to play with no diffi-
in a good storage environment, PVC discs suf- culty and no apparent damage. It also appears
fer from an increasingly rapid embrittlement at that the tape couldbe re-treated later for an-
the end of their storage life. They recommend other use period.''
skilled judgment to determine when a disc has
become so embrittled that it should be re- Similar environmentally based binder rejuvena-
recorded. They indicate that: tion programs that reverse tape hydrolysis are
discussed in papeqs^by Sidney Geller'o and
Such embrittlement is often noticed by the Edward Cuddihv."
decrease in flexibility of a disc or by playback
(with good equipment) resulting in disc wear There is a critical need for basic research into
so serious that the powder will dirty a soft both the conservation science and treatment
white cloth wiped across the surface.'" ofsound recordings. The problem is that the
profession of sound recording conservator does
This is reminiscent of a short poem I used to not exist. Even the question ofproper cleaning
know, written by the Danish poet and inventor ofrecordings has not been examined by a
Piet Hein, about making perfect toast: one was conservator or scientist. Gerry Gibson laments,
supposed to put it in the toaster until it began to "What is known is based upon trial and error,
smoke and then toast it for fwo minutes less. not upo.q controlled, objective. scientific
sfudv."au

t9
The National Library is in the initial stages of select items for retention and the remainder
plans for a conservation program for its sound could be dubbed onto an archival medium.
recordings. The intention is to create two posi-
tions for sound recording conservators: one to At this time, however, there are two major prob-
concentrate on disc technologies and the other, lems with this solution of selective retention
magnetic tape technologies. Such positions will and global coplng. There is no proven archival
be challenging. Clearly a strong background in medium for sound carriers.'' In the case of
chemistry will be needed to ensure that non- discs, records may well last longer than any me-
destructive. reversible. ethical conservation dium onto which they are dubbed. Copies them-
techniques are developed. Much scientific re- selves must, over a period of years, be copied.
search will be required before treatment deci- Though tape is most often used for dubbing,
sions can be made. How exactly can we repair Smith et al. report:
a broken 78 so that it can be played again?
How can we simulate grooves so the needle can It has been found that in many cases old tapes
track the record at least one more time? Are we can be read only once. Therefore a tape test-
talking about repair for only one more play or a ing program involving random sampling
more perrnanent repair? If a one-play repair is might eventually destroy a substantial portion
of a tape library without finding bad tapes."
adopted do we immediately reverse the repair
once completed in order to preserve the
The second problem is related to the increased
original?
sophistication of sound systems. A copy made
today is usually dubbed using a computerized
As we have seen, tapes are even more problem-
system that easily filters out the original's
atic. Can anything be done when there is cata-
clicks and pops, adjusts the pitch and removes
strophic failure, that is, when the binder has
background noise. But this adjustrnent in sound
separated from the backing? Can something be
is not reversible on a copy. We still need to
done when the lubricant or plasticizer has
have the original to get as close to the per-
dried, making the tape very brittle?
former's or composer's original intent.
Perhaps the most stimulating and challenging
A major study of preservation needs for sound
debate of this Symposium is that of retention:
recordings was undertaken in 1987 by the
can we, should we, at what price? There are
intemational Association for Recorded Sound
many analogies that can be drawn between
Collections. The results and recommendations,
sound recordings and books: both are mechani-
compiling more than 400 pages, call for action:
cal, they function, they move, they are used.
Both can be damaged through use. Both carry There is a clear and urgent need to preserve
information and are usually mass produced. our surviving heritage of sound recordings.
Both are relatively easily copied: microfilm, Sound recordings have been created and dis-
photocopy, dubbing. And both speak ofa cul- seminated for nearly a century, for the most
ture and its time. That trapper and his cylinder part with little thought for their lasting signifi-
player, the isolated homesteader listening to cance to society. For the most part, recogni-
gramophone recordings in retreat from the sur- tion of their scholarly value has come about
rounding wildemess. a teenager pouring over only recently along with the creation of archi-
his or her favourite 45s. Just as bibliographic val facilities, collections, and preservation
integrity and artifactual value are important projects. ... In the meantime, very large num-
considerations when deciding on whether to re- bers of sound recordings including unique
tain a book after microfilming, similar consid- material -
are rapidly deteriorating.
-
erations are important for sound recordings.
Preservation of archival collections of sound
recordings, both in theory and practice, has
Unquestionably at least a proportion of original only recently begun to receive widespread,
published sound recordings should be kept. serious attention. As with any field of study
Criteria based on rarity, value, examples of dif- which is only now moving through the early
ferent media, major changes to carrier or sound stages of development, the field of audio
recording technologies could be used to help

20
archiving is characterized by widely diver- si le rangement sur 4tagires et les contenants
gent practice, doubt. confusion and a myriad convenaient, et d tvaluer I'etat des obiets ainsi
of questions.*' conservtls. Des tests portant sur la lignine et le
pH des contenants d base de papier ont itt
Nearly all of the conclusions reached by the menis, et I'on a identifi| les matdriaux qui en'
planning study group reflected the lack ofco- trent dans la composition du support des enregis-
trements sonores eux-mAmes, de mAme que ses
ordinated, carefully planned research into audio
couches d'oxyde. Dans le cas des rubans de
preservation and conservation problems.44
magndtophone d bobines, on a effectut des essais
Gerry Gibson has expressed it best:
defrottement et d'odeur, et I'on a tentd d'identi'
Clearly, the solution can not be to endlessly fier plus de 20 genres de dommages, dont le plis-
sement, I'ttirement, I'effitement, le blocage et la
rerecord holdings. We must search to find a
migration du plastifiant. Les autres supports d'en-
more permanent storage media [sic] and to ac-
registrement ont dtd soumis, dans la mesure du
cept an archival format good for 50 or more
possible, d des essais analogues. Dans le cas des
years. Further, we must actively and aggres-
disques, I'usure des sillons a en outre 6td 6valute.
sively work together, since the job is far too
large for any one or two collections to under-
Au cours de la prtsente communication, nous
take. We must carry out coordinated research
traiterons des rdsultats et de I'incidence de cette
into the various factors that affect the long
etude. A court terme, les renseignements nou-
term storage and retrieval of the data and ma-
veaux que nous fournissons faciliteront la planifi-
terials in our collections. We must work to-
gether to build the shared pool ofknowledge
cation des mesures qui seront prises pour assurer
la prdsertation de ces enregistrements leur
which is necessary to prevent premature fail- -
relogement et leur copie notamment. L'etude
ure of the items in our care, and, thus, loss of
the knowledge of our civilization. Only in fait par ailleurs ressortir certains problimes de
conserl)ation qui demeurent non rdsolus, et qui
that manner can we assure that the informa-
sont aussi bien d'ordre technique qu'dthique.
tion that thev carrv will be transmitted to
Ainsi, sur le plan de l'6thique, la questions de-
fut r.e g"noutions.a5
meure de savoir si la rdparation du mtdium doit
d'abord viser d obtenir une audition supplhmen-
It is time that the conservation community took taire de I'enregistrement ou d assurer sa conser'
up the challenge! vation pour la posttrit4. Nous terminerons en
traitant des efforts que ddploie la Bibliothique
R6sum6 nationale en',ae de rtsoudre la dfficulti que
pose la crdation d'une discipline distincte pour
Les enregistrements sonores d Ia Bibliothique la conservation des enregistrements sonores.
nationale du Canada : une dtude d'titat et son
incidence sur la conservation References
C'est d la Bibliothique nationale du Canada que l.Quoted in Moogk, Edward B., Roll Back the
sont archivds, entre autres, les enregistrements
Years: History ofCanadian Recorded Sound
sonores produits au Canada. Au cours de l'6td de
and its Legacy. Genesis to I930 (Ottawa:
1990, les enregistrements sonores conservbs d cet
Atubnssement ontfait I'objet d'une etude d'6tat,
National Library of Canada, 1975) p. 5.
mende dans le cadre d'un programme de trois
Original source, J.B. McClure, ed., Edison and
ans. Les rubans de magnetophone d bobines, His Inventions, (Chicago: Rhodes and McClure
les cylindres, les disques longue durbe, les Publishing Co., 1895).
45 tours et les 78 tours ont tous 6tt soumis d
une dvaluation distincte. 2. Mooglq Roll Back the Years,p.8.

L'information ainsi recueillie facilitera la gestion 3. Bullock, Allison, National Library of


des collections, de m€me que la planification des Canada 1989 Survey ofthe Non-rare Printed
mesures d prendre pour assurer leur prdsertation Collections (Ottawa: National Library of
et leur conservation. On a ainsi cherche d savoir Canada, 1989, in press).

2l
4. McWilliams, Jerry, "Sound Recordings in 15. Gibson. Gerald D.. "Preservation and
Swartzburg," in: Conservation in the Library: Conservation of Sound Recordings."
A Handbook of Use and Care of Traditional Paper presented to the [U.S.] National
and Non-traditional Mqterials, ed. Susan Archives, February 28, 1989,p.4.
Garretson, (Westport: Greenwood Press, 1983)
p.164. 16. McWilliams, "Sound Recordings in
Swartzburg," p.164.
5. McWilliams, Jerry, The Preservation qnd
Restoration of Sound Recordings (Nashville: 17. McWilliams, "Sound Recordings in
American Association for State and Local Swartzburg," p. 169.
History, 1979)p.6.
18. Fontaine, Jean-Marc, Degradation de
6. McWilliams, "Sound Recordings in I' enregistrement magnetique audio (Paris:
Swartzburg," p. 165. Ministdre de la culture et de la communication,
1987) p. 10.
7. David Grattan and Helen Burgess, personal
communication, Canadian Conservation I 9. Fontaine, Degradation de I' enregis trement,
Institute, 1991. p. ll.
8. St.-Laurent, Gilles, "The Care and Handling 20. Fontaine, Degradation de l'enregistrement,
of Recorded Sound Materials," Commission on p.13.
Preservation and Access Report, (Washington,
D.C.: Commission on Preservation and Access, 21. Blank. Sharon. "An Introduction to Plastics
l99l) p. 8. and Rubbers in Collections," Studies in Conser-
vation.35, 1990, p.53.
9. Heckmann, Harold, "Storage and Handling
of Audio and Magnetic Materials," in: 22.Brown, D.W., R.E. Lowry and L.E. Smith,
Preservation of Library Materials, ed. Merrily Prediction of the Longterm Stability of
A. Smith (Paris: Saur, 1987) p. 68. P olyes ter- B as ed Re cordin g Me d ia, NB SIR
82-2530 (Washington, D.C. : National Bureau
10. Gibson, Gerald, "Preseryation of Non-paper ofStandards, June 1982) p. 9.
Materials," in: Conserving and Preserving
Library Materials, Kathryn Luther Henderson 23. Smith, L.E., D.W. Brown and R.E. Lowry,
and William T. Henderson, eds. (Urbana-Cham- Prediction of the Longlerm Stabiliry of
paign: University of Illinois, 1983) p. 103. P olye s t er- B as ed Re cord in g M edia,
(Washington, D.C.: National Bureau of
11. Gilles St.-Laurent, personal communica- Standards, June 1986) pp.2-3.
tion, 1991.
24.Pickett and Lemcoe, Preservation and
12. Pickett, A.G. and M.M. Lemcoe, Storage, p. 56.
Preservation and Storage ofSound Recordings
(Washington: Library of Congress, 1959) 25. Brown, Lowry and Smith, Prediction of
p.30. Long-term Stability, 1982, p. 9.

1 3. McWilliams, Preserttation and Restoration. 26. Brown, Lowry and Smith, Prediction of
p.42. Long-term Stabilin, 1982, p. 2.

14. Alexandrovich, George, "Phono Cartridges 27. Smith, Brown and Lowry, Prediction of
and Communications," Broadcast Engineering, Long-term Stability,l986, p. 1.
August 1982,p.26.
28. Kent. Scott. "Binder Breakdown in
Back-Coated Tapes," Recording Engineer/
Producer, July 1988, p. 80.

22
29. Waites, J.B., "Care, Handling, and 38. Geller, Sidney 8., Care and Handling of
Management of Magnetic Tape," in: Magnetic Computer Magneti c Storage Media, NBS
Tape Recordingfor the Eighties, ed. Ford Special Publication 500-10 I (Washington,
Kalil, NASA Reference Publication 1075 D.C.:U.S. Dept. of Commerce, June 1983)
(NASA, April 1982) p. 50. pp. 94 and I 15.

30. Ad Hoc Subcommittee on the Preservation 39. Cuddihy, Edward F., "Stability and Preser-
of Sound Recordings of the National Archives vation of Magaetic Tape," in: Proceedings of
and Records Administration (NARA) Advisory Conservation in Archives: International
Committee on Preservation. Minutes of meet- Symposium, Ottawa, Canada, May 10-12,
ing July 29-30,1987, Washington, D.C., p. 13. 1988, (Paris: International Council on
Archives, 1989)p.204.
31. Waites, "Care, Handling, and Manage-
ment," p.54. 40. Gibson, "Preservation and Conservation,"
p.7.
32. Association for Recorded Sound Collec-
tions (ARSC), Associated Audio Archives 4l . Association for Recorded Sound Collec-
Committee, Final Pedormance Report. Audio tions (ARSC), Final Pedorrnance Report, p.5.
Preservation: A Planning Study (Washington,
D.C.: ARSC, December 31,1987) p. 54. 42. Smith, Brown, and Lowry,Prediction of
Long-term Stability, 1986, p. 20.
33. Gibson, "Preservation and Conservation,"
p. l. 43. Association for Recorded Sound Collec-
tions (ARSC), Final Performance Report,p. ll.
34. Pickett and Lemcoe, Preseryation and
Storage,p.26. 44. Association for Recorded Sound Collec-
tions (ARSC), Final Perforrnance Report, p.4.
35. Pickett and Lemcoe, Preserttation and
Storage, p. 49. 45. Gibson, "Preservation and Conservation,"
p. 17.
36. Fontaine, Degradation de l' enregistrement,
p.18.

37. Kent, "Binder Breakdown in Back-Coated


Tapes," p. 81.

z)
Plastics in the Science Museum, London:
A Curator's View

Susan Mossman

Collections Division
Science Museum
London, U.K.

Abstract stored separately, and, in cases ofsevere degrada-


tion, are isolatedfrom thosewhich, sofar, ap-
It is only in the fairly
recent past that plastics pear to be "healthy."
have become afeature of many museum collec-
tions. During the lastfve years, curators have Current research into the degradation ofthe earli-
begun to notice that objects made ofplastics de- est cellulosic plastics appears to suggest that in
grade with time, and sometimes very rapidly, 50 years' timefew of these will remain. The cura-
indeed. These provide problemsfor the curator, tor has to consider very carefully whether to dis-
panicularly in a technological museum such as play the most vulnerable types of plastics, which
the Science Museum, where we collect artefacts will then be subjected to quickened rates ofpho-
that may be wholly or partly made of plastics. As todegradation. Should the curator acquire two of
time goes on, our plastics artefacts, with their every plastics object: onefor display and hence
attendant problems of degradation, will only disposable and another to keep in optimum stor-
increase in number. age conditions to give it the longest lift possible?

The Science Museum's collection of plastics Introduction


contains about 1,500 objects, and is ofgreat im-
portance because it contains a ich selection of How long will our plastics collections survive?
historic plastics, ranging from Parkesine, the first This is the question that curators are now hav-
semi-s.vnthetic plastic, to a sample of the earliest ing to ask themselves. It has become evident
polyethvlene. We continue to collect actively in over the last few years that collections ofplas-
thefield of modern plastics materials, in particu- tics might not last forever. They degrade, and
lar composites and biodegradable plastics. In some faster than others. Sadly, it also appears
addition, plastics appear in many of our other
that the most historic and earliest plastics are
collections where they play an increasingly impor-
the ones that are most lulnerable.
tant role, especially in telecommunications, elec-
trical engineering, medicine and transport.
John Morgan explains elsewhere in this publica-
Particular problems we have noted occur in the tion what a polymer is and briefly introduces
earliest semi-synthetic plastics, that is, those con- some of the historic plastics. This paper refers
taining cel lulose nitrate. to many of the same plastics, but from the
point of view of a curator of plastics of varying
lle are in the process of improving the storage of kinds, from the oldest to the most modem,
our plastics, giving due attention to the effects of referring in particular to objects in the Science
photodegradation and high temperatures as well Museum's collections. The paper's flrst part
as ensuring that the various types ofplastics are maintains a roughly chronological framework.

25
The Science Museum's collection of plastics Vulcanite is well represented in the collection
(about 1,500 in number) is rich in the high- that includes many pieces made by Thomas
lights of plastics' history, and many pieces are Hancock, including a plaque of himself, dated
(as far as we can tell) in quite good condition. ca. 1843, and made of rulcanized rubber. Re-
We have obtained specimens from a variety of markable are some ornate vulcanite plaques
sources, ranging from industry to relatives of made by Hancock and sealed up in glass-
the inventors of certain plastics. We are also fronted passe-partout frames, which were dis-
fortunate in owning two large groups of plas- played in the Rubber Exhibition at the Science
tics (together numbering over 500 objects) that Museum n 1929. These have survived in excel-
are not only useful because they cover the full lent condition; possibly linked with the fact that
range of plastics materials from the late l9th their sealed environment has set up a stable mi-
century until ca. 1970, but also because they cro-climate, with no access to oxygen to encour-
contain objects of aesthetic importance. Some age the degradation process (Figure l). Only
of the more important items are addressed in one plaque has begun to fade slightly.
the following discussion.

The collection includes the rubber-based prod-


ucts that predate what is regarded as the first
plastic, Parkesine. Some of the rubber-based
compounds were used in their unmodified
state, such as gutta-percha. These fall into the
category defined as natural plastics.

Gutta-percha is a dark brown substance ob-


tained from the palaquium tree, which is found
in Malaysia. It became a very popular material
in the early Victorian period and was used for a
multiplicity of purposes. The Science Museum
possesses a number of examples of this mate- Figure I Vulcanite plaques, ca. 1843, qre
rial, including an inkstand dated to about 185 I elaborately decorated and in excellent condition.
(moulded to commemorate the use of gutta- This is probably linked to the lack of oxygen
available in their sealed environment.
percha to insulate the first submarine telegraph
cable from England to France in 1850). Gutta-
Perhaps the greatest treasures in the plastics col-
percha possesses excellent insulating proper-
lection are the 87 examples of Parkesine dated
ties, and so was ideal for this purpose. The
from I 855 to I 88 1 . Parkesine, invented by
main problem we have with gutta-percha is that
Alexander Parkes, is generally accepted to be
with age the surface cracks, and the object be-
the first plastic, although at this early date, of
comes brittle and fragile.
course, it was a modified natural product rather
than a fully synthetic material. Parkesine is
Other natural plastics in the collection are hom,
made of cellulose nitrate, which is made from
tortoiseshell and shellac, as well as a very rare
nitric acid, sulphuric acid and cellulose (ob-
material called Bois Durci (literally translated
tained from such sources as cotton flock). It
as "hardened wood") blended from sawdust and
is then mixed with vegetable oils and small
albumen.'
amounts of organic solvents. This mixture
forms a pliable dough that can be moulded into
Certain other rubber-based substances, such as
a variety ofshapes. A collection ofParkesine
vulcanite, are themselves sometimes regarded
objects presented to the Science Museum in the
as semi-synthetic plastics, since they are modifi-
1930s by the family of the inventor contains a
cations ofa nafural substance. In the case of
wide range of items, from crude samples to or-
r.ulcanite, sulphur is added to natural rubber to
nate hair-slides inlaid with silveq brass and
produce a harder material.
mother-of-pearl (Figure 2). Other pieces of
Parkesine also convey religious themes. A

26
remarkable piece is a tiny carved head gf celluloid, combining innovation with sharp mar-
Christ, measuring 2.7 cm in height. A signifi- keting skills. He realized there would be a keen
cant number of these pieces show Alexander market for celluloid combs and wipe-clean
Parkes'skills as a carver. collars and cuffs.

A fibre based on cellulose nitrate, and de-


scribed as artificial silk, was patented by Count
Hilaire de Chardonnet in 1884, and samples
were shown at the Paris Exhibition in 1889.
We have an advertisement for Chardonnet silk
showing dyed and undyed samples that date to
1896 (Figure 3).

7XR-rrFrerTf,b
",:'
srLr,{..
' '-''
The Artificiai Silk Company. Iirnited.
+
Figure 2 Parkesine hair-slides with brass, silver and
mother-of-pearl inlay, ca. 1862. Metal inlay is ltfting
A ,Vr-'rr Btttt'rsn /,rlr''srRr',
due to shrinkage of Parkesine.

Parkes set up the Parkesine Company Ltd.


to manufacture Parkesine. After the company
failed in 1868, his one-time Works Manager,
Daniel Spill, tried to make money out of
Parkes' invention, which Spill renamed Xylo-
nite. The Xylonite Company (set up by Spill in
1869) also produced material made of imitation
ivory (called Ivoride) and coral. The Science
Museum possesses objects made of Ivoride
and imitation coral, which were manufactured
by Spill's company in the 1870s. Of particular
interest is the Ivoride Death's Head walking-
stick handle, which was Spill's own. Samples

Neither Parkesine, Xylonite nor Ivoride were


very satisfactory products due to problems with
flammability. John Wesley Hyatt, an American, Figure 3 Chardonnet silk, 1896. Yellowing;
experimented to develop a more stable product. Iiquidizing; based on cellulose nitrate (artifcial silk
Hyatt made the breakthrough in 1870 when he patented by Count Hilaire de Chardonnet in 1884, of
found that camphor made an excellent solvent which samples were shown at the Paris Exhibiion
and plasticizer for cellulose nitrate. flyatt pat- in 1889).
ented his invention on l2 July 1870' and
called his product Celluloid, the name by Our collection includes samples of other semi-
which most products based on cellulose synthetic plastics, such as the milk-based plas-
nitrate are known today. tic. casein. and the less flammable cellulosic
plastic, cellulose acetate.
Spill then entered painful years of litigation
with Hyatt, who by l872had set up the Cellu- In 1909 came the next important development
loid Manufacturing Company. Spill finally lost in the history of plastics, when a brilliant Bel-
the battle, and Hyatt made a great success of gian chemist, Leo Baekeland, patented the
frst phenolic polymer. When combined with a

27
filler, such as wood flour or cotton flock, phe- Material. This consists of a number of cosmetic
nol-formaldehyde forms an excellent material boxes ofcellulose acetate and later, urea-for-
that can be moulded into a durable plastic maldehyde, made in Paris by the workshops of
with excellent electrical insulating properties. Editions Paris E. Fornells between the years
Baekeland's invention was quickly adapted to 1913 to 1940 (Figures 4 and 5). Eduard For-
many uses, primarily electrical fittings, but was nells Marco, an Andorran by birth, was a mas-
also commonly used for telephones and radio ter carver recognized for his skills and taken on
casings. A point to note is that phenolic plastics as a craftsman by Ren6 Lalique. Fornells was
are dark in colour, due to their fillers. A remark- responsible for the famous cherry box design
able object in our collection is a coffur, in- marked Lalique, made between l9l I and 1913,
vented by James Doleman and made in 1938 of which the Museum also has an example in
by the Ultralite Casket Co. Ltd. The coffin is its collection (Figures 6 andT).
claimed to be the largest phenolic moulding
ever made. It was manufactured from imitation In addition to a wide range of products illushat-
walnut phenolic with a wood flour hller (a spe- ing the applications and particular advantages
cial type of moulding material produced by Ba- of individual plastics materials, the Science
kelite Ltd. of London). Various stories are told Museum collects machinerv used for the
as to why the coffin did not go into large scale
production. According to the inventor's son
(personal communication) it was due to his
father's death in 1944, durngWorld War II.

By 1928, a method had been perfected of cast-


ing phenol-formaldehyde without a filler. The
result was the ability to produce cast phenolics
in bright colours. Cast phenolic was used to imi-
tate natural materials, such as amber and jade.
The Science Museum collection includes a
FADA Radio, dated to the 1940s. This is a
collector's item and made of imitation amber
catalin (cast phenolic).

The mid to late 1920s saw the advent of the Figure 4 Fomells Material: cosmetic boxes of
colourlhl as well as durable and waterproof c e ll ulo s e ac e tat e a nd late r, ur e a-fo rm a I d e hy d e,
plastics made of thiourea-formaldehyde and made in Paris by the workshops o/ Editions
urea-formaldehyde. Identified by such trade Paris E. Fomells, ca. 1913 to 1940.
names as Beetleware and Bandalasta Ware,
these plastics were popular for domestic ware
and picnic sets. They are well represented in
our collection.

With the 1930s, came polyethylene, polysty-


rene, acrylics, and polyamides. We have a sam-
ple of the first polyamide knitted tubing, made
by du Pont in July 1935. Nylon was to be its
successor and would become a runaway suc-
cess as an excellent thread for stockings, and
later in solid form as a material for gears and
bearings.

As well as very practical objects, the plastics


collection includes items important for their
artistic merit. Among the finest is the Fornells
Figure 5 Close-up of Fornells' trade mark.

28
not one of chemistry.a There was a need to de-
velop a reactor that could produce greater quan-
tities without a mnaway reaction leading to an
explosion in the reaction vessel. By 1939, ICI
had developed the first fuIl-scale polyethyl-
ene plant, which was capable of producing
100 tons (based on a 50-litre reactor). ICI
continued to expand their production capacity,
which was necessary, as polyethylene played a
vital role during the Second World War, for ex-
ample, in radar. The nine-litre reaction vessel
was an important step in this development.

The Science Museum is also active in collect-


Figure 6 Cherry box made by Eduard Fornells
ing archive material, for example, trade cata-
Marco (between l9ll and 19I3) shows Lalique
trade mark.
logues, associated technical literature and
sample books. Particularly significant is a gutta-
percha catalogue from the Gutta-Percha Com-
* pany, ca. 1851, which shows a range ofobjects

€{!t
from cherubs to coats of arms moulded in this
material. In addition, the Museum holds a num-
. *-,8 ber of casein button sample cards (part archive

{tt'-
; lf . t*il'
tt- t1
and part object) dated to the first halfofthe
20th century.

Such archive material is useful because it helps


us assess the range ofobjects that were avail-
able at the time the catalogue was produced and
it also helps us identify objects. Work diaries
and private papers often give us insight into the
development of the plastic, working practices,
and problems encountered.
Figure 7 Close-up of trade mark on Lalisue cherrv
box. It is essential when acquiring an object, particu-
larly modem one such as a composite ski
a
production of plastics. The machinery has sepa- boot, to acquire the relevant technical literature
ratg conservation problems (i.e., metal, wood at the same time. A modern Salomon ski boot
and rubber), which are dealt with elsewhere in
(1985) in the collection is made of four injec-
this publication. tion-moulded parts in nylon with polycarbonate
clips; the boot is lined with a polyurethane
An important example of such machinery is a foam sock.
pilot-plant, nine-litre reaction vessel for poly-
ethylene, which was used at Imperial Chemical In the absence of technical literature, it would
Industries (ICI), Winnington, in 1938 to pro- be almost impossible to find out what plastics a
duce the first ton of polyethylene (of which we complicated object is made of, without much
are fortunate to have a sample). ICI had begun
time and expense. Howeveq sometimes manu-
the chemical work that led to polyethylene as facturers are reluctant to reveal the secrets of
early as 19323 andwere finalfy successful in re- their latest invention. Occasionally, we have to
peating the discovery somewhat fornritously in wait for over 50 years before the full story be-
1935. Due to the very high pressures required hind a particular plastic is released.
to produce polyethylene, ICI regarded produc-
ing larger amounts as an engineering problem, Another area to explore and collect is oral his-
tory. Although we have no formal programme

29
in this area at the Science Museum, it is often problem we have with other Parkesine objects
invaluable to record interviews with pioneers of is shrinkage. This is particularly evident with
plastics and later have the text transcribed. decorative combs that have metal inlays. In
some cases, the object has shrunk so much that
On a broader level, plastics materials are not the inlay has popped ou1.
only confrred to the plastics collection; many
other Science Museum collections contain ob-
jects of plastics in combination with other mate-
rials. Some that deserve mention are collections
of electronics, telecommunications, medicine
and aeronautics.

Are plastics worth preserving? We of course


are biased and hope that the brief resum6 of
some of the treasures in our collections are
convincing proof of their value and importance.
Many, in particular the Parkesine and polyethyl-
ene objects, hold great significance for the his-
tory ofthe science and technology ofplastics.

The curator has to make decisions as to what


to collect, both on the grounds of the object's Figure 8 Samples ofdeteriorated Parkesine (thin
significance and longevity. This does not mean films and pen nib), ca. 1862.

that the more l'ulnerable plastics should not be


In the case ofa Parkesine denture, degradation
collected; however, it is important to be aware
was rapid. The denture had been on display for
of their probable lifetime as well as ensure that
all important information is recorded about two years. We noticed that yellowish powder
them upon acquisition. Other ways of recording was deposited around it; when we examined the
denture, it was so brittle, it immediately fell
the shape ofa plastics object should be consid-
into pieces.
ered. Various suggestions of doing this are
explored later.
Other pieces on display appear to be fairly
stable, but these are monitored regularly.
Which Plastics Are Most at Risk?
When a material has been stressed, for example
Cellulose Nitrate-Based Plastics
a cellulose nitrate comb made of imitation tor-
Our immediate concern is with those plastics
toiseshell, it crazes where it has been bent into
based on cellulose nitrate: Parkesine, Ivoride,
shape.
Xylonite and Celluloid, of which we have ex-
amples in our collections. Certain examples
When metal is in contact with cellulose nitrate
have shown characteristic signs of degradation,
objects, in the form ofinlays or hinges for ex-
as also discussed in detail by John Morgan.
ample, it appears to speed up the degradation of
These include sweating, crazing, discolouration
the cellulose nitrate, providing a focus around
(yellowing in the case of imitation ivory cellu-
which degradation begins and then spreads
lose nitrate objects) and the giving off of acidic
throughout the object.
fumes.

Curiously enough, the majority of our Parkes- Other Plastics


ine objects are in good, and sometimes excel- At present our cellulose acetate objects do not
lent, condition. The items that have suffered appear to be exhibiting signs ofdegradation.
most are the thin films and smaller pieces, such However, the Tate Gallery in London has expe-
as a pen nib and a tiny carved head that are rienced problems with sculphrres.made from
both now in fragments (Figure 8). The main cellulose acetate by Naum Gabo.' In addi-
tion, the problems of deteriorating cellulose

30
triacetate filrn are well known.o So we can fore- o The objects in store are kept in the dark. This
see problems with our cellulose acetate objects. is essential as plastics are very susceptible to
photodegradation, particularly when exposed
Currently our casein (also known by such trade to ultraviolet light.
names as Erinoid and Galalith) objects appear
to be in remarkably good condition; however, o Deteriorating plastics, especially cellulosics,
we are well aware of their moisture-retaining
are separated from "healthy" objects.
qualities as well as the danger of warping and
cracking if the humidity levels change too
r Gloves are always worn when handling
radically.
plastics, as fingerprints can indelibly mark
Phenolic plastics (generally referred to as certain types.
Bakelite) appear to be fairly stable unless bro-
ken (when they become susceptible to biologi- o Sticky labels are not applied to objects as we
cal attack due to their fillers made of wood have found that the glue can "eat into" the
flour and cotton flock). The main problem surface.
appears to be with fading (green to brown).
We have to deal with the practicalities of stor-
Cast phenolic plastics (often known as catalin) ing a large number of different plastics as well
are also susceptible to colour change, having a as our other collections (which are themselves
tendency to turn yellow. made of many and varied materials).

Solutions: What Are We Doing If we keep cellulose nitrate-based plastics, such


at the Science Museum? as Parkesine and Celluloid, on open shelving
for ventilation, as is recommended practice,
Thorough Records any gases released may attack other plastics.
The first priority is to record the object fully, Currently most plastics, apart from the largest
both in writing and visually. We are doing this pieces, are wrapped in acid-free tissue and
for the Science Museum's plastics objects. We packed in acid-free boxes. We are well aware
write a detailed description of the object on a this may not be ideal for the cellulosics in
record card, including date and manufacturer's particular.
name. Other space on the card is available for
dimensions, material, and most important of all, A few pieces have been laid out on open
condition. We date the card so that we have a shelves on plasiazote, which appears to be a
report of the item's condition at the time it is relatively stable base. We are monitoring this
recorded. This information will be important arrangement.
as more plastics begin to deteriorate, either by
discolouring, shrinking or by showing various Of course, a separate, well-ventilated room for
other symptoms, such as crazing and cracking. each type of plastic, with a separate room for
those that have begun to degrade, would be
It is important to photograph the object with a ideal. However, every museum has constraints
scale, preferably using colour. We are undertak- on space, and we are no different. Cost is an-
ing a programme that incorporates this practice. other limiting factor.

Stores We might consider using another room at our


We have improved the storage conditions for store in which we might install a ventilated
our plastics, thus: case for those cellulose nitrate objects that are
deteriorating. Even the air for ventilation is
r The plastics are separated by type of material. not without attendant problems: the store is at
Olympia, Kensington, in a busy part of London
and so suffers from air pollution due to heavy
. They are stored in stable conditions that are
traffrc. We cannot allow air to come into the
kept as cool as possible.
case without filtering it or cleansing it in parts replaced as they wear out? Or should the
some way. machinery be preserved in a suspended state,
but in good condition?
Display Material
Whilst we try to protect our plastics in store Moreover, many of the early plastics were
from many of the dangers that could initiate or produced using somewhat rough and ready
accelerate their degradation, what about those methods, which could on occasion be quite
plastics on display? We have to accept that, at dangerous. For example, producing cellulose
present, displaying plastics will shorten nitrate-based plastics, such as celluloid, is po-
their lives. tentially afirehazard. Within the plastics
manufacturing centre at Oyonnax (in La Val
Light levels can be reduced and ultraviolet Plastiques, Jura Mountains, France) is located
light can be excluded by the use ofoptical fil- La Grande Vapeur, a factory that was used to
ters. We have excluded natural light from our produce objects, mainly combs and spectacle
plastics gallery at South Kensington. frames, initially of cellulose nitrate, and later
ofcellulose acetate fromca. 1908. The factory
We also look forward to the results of the work had an in-built sprinkler system as well as a sys-
being carried out by Julia Fenn at the Royal tem of cells with steel doors for each workshop,
Ontario Museum, in setting up a sealed case for so that if fire broke out in one workshop it
plastics, and using a system ofcleansing the at- could be contained.
mosphere using scavengers and absorbents as
well as a method of ensurins the circulation of To be realistic, the factory inspector and the
air within the case.7 Science Museum's safety officer would not be
happy if we tried to reproduce these types of
Using traditional display methods will mean fireproof "cabines" in the Science Museum. In
that in some cases, for example with the cellu- fact, we were not allowed to display a block of
losics, complete breakdown may come very celluloid together with its pilot machinery in
rapidly. Indeed, it is vital for the curator and the 1970s due to fears about the flammability
conservator who are mounting the display of of the plastic.
plastics, to be fully aware of the implications of
their actions. The recipes used to make early plastics were
not standard. Rather like cooking, they varied
It would be most sensible to display only those according to the person making the plastic. We
plastics objects that are duplicates or easily re- have recipe books for casein in the Science
placeable. This revision in curatorial attitudes Museum. Other recipe books exist for cellu-
would have a direct inlluence on collecting poli- loid.* However, records show that the plastics
cies. The most satisfactory solution would be to pioneers spent a fair amount of time unjam-
collect two of everything: one to be kept in ming the machinery when it became clogged
ideal collections in store, and the other to be due to variations in the mixture. Of course to-
kept for display purposes. day, things are very different. Mixtures and
temperatures are precisely measured, and pro-
Solutions to the Problem of Display: vided one has access to the amounts used, it
Pros and Cons would be possible to replicate these. However,
modem machinery is usually on an enornous
Replicas scale, and this might well prove impractical.
A mould could be taken of the object, in order
to make a replica. This might be suitable where
original machinery has survived. However,
the use of the original machinery provides an * Ipswich Public Records Office holds recipe books
ethical dilemma for the curator of technology, for the manufacture of Cellulloid, from BXL Plastics,
as well as practical diffrculties. Should the ma- Wardles Story, Brantham in Essex, the site ofthe old
chinery be maintained in working order and Xvlonite Factorv.
Three-Dimensional Recording visitor would then be able to look at objects not
on display. This would also satisff the visitor
The object could be copied in a more techno- who wants to look at items made of a particular
logical fashion. Three-dimensional photo- plastic, without having to put the objects on
graphy is an option; the object could be display and thus contribute to their degradation.
photographed stereoscopically, so that the Moreover, it would be beneficial to those visi-
picture could be viewed in three dimensions tors with a particular interest. They could select
through special spectacles. This would give only the objects they wished to view for study
a clear idea of shape. Three-dimensional pho- purposes. Again, this practice would cut
tographs could be taken from all importrant down the unnecessary handling ofobjects
angles. Ifthe object degraded, a three- in store.
dimensional record of its form would at least
exist. Such a system is already in use in the However, such a system would have to be used
archaeological world, and in the Victoria and with caution. Visitors will always want to see
Albert Museu;n department of sculpture objects. Though this system will allow visitors
conservation.d to see objects not on display due to lack of
space or fragility of the objects, it will never
A more advanced technique would be to record totally replace the display ofreal objects.
the three-dimensional image electronically.
Such a system is available, for example, the Conclusion
company Tektronix produces a suitable pack-
age. However, all such systems cost money. Today the Science Museum is active in collect-
The Science Museum is currently undertaking ing new plastics materials, specifically those
a programme of photographing objects with engineered for a precise purpose, or those that
the long-term view of transferring the images to have found a new use. For example, recycled
computer; our outstation, the National Museum and biodegradable plastics are an increasingly
of Film, Photography and Television at Brad- important area in which to collect. We have
ford, is considering doing the same for its pho- even collected a flowerpot made of recycled
tographic archives. Another outstation, the plastic. Another recent acquisition is a
National Railway Museum at York, is explor- Biopol bottle, a degradable plastic produced
ing digital storage for its photographic material. by ICI.
Three-dimensional computer imaging is some
way down the ling (on current judgement. As for the rich collections that we have inher-
about five years).' However. if the obiects are ited, we are now becoming increasingly aware
already photographed in this format. the images that we have to put into practice what we leam
could be transferred to computer when suffi- about their stability. We have had to assess
cient funding is available. Currently such a them for condition and probable lifetime. We
proposal is in preparation for the Science have improved our storage; in particular cutting
Museum's collection of plastics. out exposure to light wherever possible, and
maintaining stable storage conditions, specifi -
In the long term, the computerized three- cally of temperature and humidity. We are
dimensional recording of the object would have keeping a careful eye on the more lulnerable
many benefits. The image would be readily plastics, in particular those based on cellulose
available at the touch of a button. It would no nitrate and cellulose acetate. As yet, we have no
longer be always necessary to unpack the ob- ideal solution to the problems of their degrada-
ject from store, which would reduce both unnec- tion but we are exploring new ways of prolong-
essary handling as well as superfluous exposure ing their lives and watching with interest work
to light. Equally, computer terminals could be being done at other institutions in Britain and
put in the plastics gallery and at information abroad, notably in Ontario, Canada.
banks throushout the museuln. The museum

JJ
Acknowledgement accordons ainsi aux ffits de la photodigradation
et des hautes tempdratures toute I'attention qu'ils
The author thanks Sue Cackett and Dr. Derek mtritent, et nous veillons d ce que les divers gen-
Robinson for their helpful comments on this res de plastiques soient mis en nlserve sipard-
paper, and Ben Booth for the information he ment et d ce que les plastiques qui se sont grave-
supplied. ment dtgradds soient isol4s de ceux qui semblent
toujours < intacts D.

All photographs are provided courtesy ofthe


S'ilfaut en croire les rdsultats de la recherche ac-
Trustees of the Science Museum, London.
tuelle sur la dtgradation des premiers plastiques
cellulosiques, il semble qu'il n'en restera quefort
R6sum6 peu dans 50 ans. Aussi les spicialistes de la res-
tauration doivent-ils ne ddcider d'exposer les
Les plastiques au National Museum of plastiques les pltts yulntrables qu'aprds avoir
Science and Industry le point de vae d'ane soigneusement analvst la question, puisque ces
consewatrice - matiires seront alors sujettes d une photodegra-
dation accildrde. Et il convient sans doute de se
Ce n'est qu'assez rbcemment que les plastiques demander s'il nefaudrait pas toujours acqudrir
ontfait leur entrde dans les collections de muste. deux exemplaires du m€me objet en plastique : un
Au cours des cinq dernidres anntes, les specia- premier qui serait exposd on pourrait
listes de la restauration ont par ailleurs com- 4ventuellement disposer
- etetundont
second qui serait
mencd d constater que les objets en plastique se mis en riserve
-
ddgradaient avec le temps, et parfois m€me trds - et qui serait conseryi dans des
conditions optimales pour durer le plus
rapidement. Une telle dtgradation n'est pas sans longtemps possible.
poser des probldmes d ces spdcialistes, surtout
dans un musde technologique comme le National
Museum of Science and Industry, oil les collec-
References
tions comportent des objets qui sont partiellement
ou entidrement constituts de plastique. Le nom-
l. Williamson, C., "Bois Durci," Plastiquarian,
bre des objets de musde en plastique continuera
vol. l, Winter, 1988, p.8.
par ailleurs de s'accroitre, multipliant d'autant
les problimes de ddgradation de ce matdriau. 2.Hyatt, J.W.. "Improvement in Treating and
Moulding Pyroxyline," U.S. Patent no. 105338.
La collection de plastiques du muste regroupe
quelque I 500 objets. Elle est trds importante car 3. Kennedy, C., rcl: The Company That
elle renferme une riche sdlection de plastiques Changed Our Lives (London: Hutchinson,
qui ont une valeur histoique t986) p.64.
- depuis le
P arkes i ne, I e pre mi er p I asti que s emi -synthefi que,
jusqu'd un ichantillon du premier polydthyl?ne. 4. Wilson, Gordon, personal communication,
Et des matidres plastiques modetnes continuent ICI Petrochemicals and Plastics Division
de s'y ajouter
- des plastiques mixtes et
bioddgradables notamment. Les plastiques sont
Licensing Manager, U.K.,9 August 1991.

en outre prtsents dans nombre d'objets des au- 5. Pullen, D. and J. Heumann, "Cellulose
tres collections, puisqu'ils jotrcnt un rdle de plus
Acetate Deterioration in the Sculptures of
en plus grand dans des domaines comme les
Naum Gabo," Modern Organic Materials,The
tilicommunications, le gtnie 4lectrique, la
Scottish Society for Conservation and Restora-
midecine et le transpor"t, pour n'en citer que
quelques-uns. tion, Edinburgh, 1988, pp.57-66.

Les probldmes particuliers que nous avons 6. Edge, M. et al., "Cellulose Acetate: An
relevts se posent dans le cas des premiers plas- Archival Polymer Falls Apart," Modern
tiques semi-svnthttiques, c'est-d-dire ceux qui Organic Materials, The Scottish Society for
renferment du nitrate de cellulose. Conservation and Restoration, Edinburgh,
t988,pp.67-79.
Nous cherchons actuellement d amdliorer la
mise en rdserve de nos plastiques. Nous

.A
)+
7. Fenn, J., "Scavengers for Controlling EUROPE
Combinations of Emissions from Exhibitions Wilhelm Koepp Zellkautschuk GmbH & Co.,
of Mixed Polymers," Polymers in Conserva- D5100 Aachen, Postfach 848,
tion, Congress, Manchester Polyechnic and Hegelsbendenstrasse 20, Germany.
Manchester Museum, July 17-19, 1991.
Companie lnternationale de Plastique
8. Larson, John, "The Three Dimensional Re-
Biod6gradable, 24 Boulevard Princesse
cording ofSculpture," Lecture in the series Tfte
Charlotte, Immeuble Est-Ouest, MC98000
State of the Arl, Royal College of Art/ Victoria
Monte Carlo, Monaco.
and Albert Museum, London, March 15, 1990.

EASTERN EUROPE
9. Booth, B., personal communication, July
BP Chemicals GmbH, Zaunergasse 4,
1991.
A-1030 Vienna, Austria.

NORTH AMERICA
Suppliers
United Foam Plastics Corporation, 172 East
Plastazote: Main Street, Georgetown, Massachusetts
01833-2107, U.S.
U.K.
Polyformes Ltd., Cherrycourt Way, AUSTRALIA
Stanbridge Road, Leight on Buzzard, Dunlop Foam Products Group, P.O. Box l,
LU7 8UH. Mordialloc, Victoria 3 I 95, Australia.

35
Membership and Aims of the
Plastics Historical Societv

John Morgan

P lasticsHistorical Society
Plastics & Rubber Institute
London, U.K.

Abstract and The Conservation Unit (of the Museums


& Galleries Commission) have started a co-
The centenary ofplastics was celebrated in 1962, operative program ofwork into the degradation
yet it is only in comparatively recent years that problems of plastics materials. The international
an interest in collecting plastic mateials has membership of the PHS encompasses a variety of
emerged. In 1968, a History Discussion Circle people, including some from i ndustrv, education
was formed in Britain by the Plastics Institute. and research as well as designers, collectors and
The idea of recording the reminiscences o.l some stafffrom museums and auction houses. Many
of the plastics pioneers, first mooted at the time members have their ovn plastics collections,
of the centenary celebrations, was re-ffirmed. covering a diversitlt of th-emes. The society is thus
However, the Circle was short lived, and it was able to call upon a wide range ofexpertise and ex-
not until 1986 that sfficient general interest in p er i e n ce, and i t s j o urnal Plastiquaria nr c o n t ai n s
plastics had developedfor the Plastics Historical news and articles ofinterest to collectors and
Society (PHS) to become a reality. plastics historians.

The PHS is an independent society, ffiliated to What Is Plastic?


the Plastics & Rubber Institute with which it
shares headquarters. Its aims are "to promote In general terms we all know what plastics are,
the study, preservation and sharing of informa- but a simple definition that includes only plas-
tion on all historical aspects ofplastics, and to tics materials and excludes everything else is
encourage the recording ofany current develop- not easy to find. One dqfurition that has pre-
ment adjudged to be of value to future genera-
viously been suggested' includes the following
tions." A longletm aim is the establishment of a
characteristics:
National Plastics Museum, but in the meantime ef-
fort is being devoted to saving historically impor- o Solid materials
tant records and objects from being consigned to
the rubbish tip. This is especially important when
old, establishedfirms are taken over by someone . Organic, or organic/inorganic polymers
who has little svmpathy for the histoical
signifcance of the company. o Natural, chemically modified natural or
wholly synthetic polymers (or mixtures of
The PHS forms links with other organizations these)
having associated interests,for example, the
Kunsts toffe - Museums -Verein (KMV), a German
o Capable of being compounded with
plastics museum society. In addition, the PHS
colourants, plasticizers, etc.

JI
. Capable of being moulded at some stage in It was not until the 1920s that the word plastics
production was first used as a generic name for the new
materials. During this same period the poly-
There are some arbitrary exclusions, such as meric nature of plastics was elucidated and this
adhesives, fabrics, paints and rubbers, but opin- scientific understanding enabled new polymers
ions on this differ and the Plastics Historical to be researched and to be established in new
Society (PHS) takes a fairly broad view in its ways. Even so, new polymers were still discov-
definition of plastics. ered by accident during other investigations,
but their polymeric nature was immediately rec-
How Old Are Plastics? ognized unlike generations earlier when
polymers- were sometimes produced by experi-
It is generally considered that plastics materials ment, but the sticky residues were discarded as
were first exhibited in England at the Intema- failures. Injection moulding, one of the key
tional Exhibition of 1862 by Alexander Parkes. processes in plastics fabrication, was also devel-
The material was called Parkesine and was oped during the 1920s. This and the various fac-
made from cellulose nitrate, a chemically tors already mentioned are reflected in the
modifi ed natural polymer. sharp rise in the output of plastics, starting in
about 1930 and continuing to the present day,
The roots of the industry, however, lie in the the progression being hindered only slightly by
earlier manipulation of naturally occurring plas- the oil crisis in the 1970s. Figures I and2 re-
tics. Detailed mouldings were being made from spectively show the growth in world plastics
hom as long ago as the 17th century; the vul- production during 1900 to 1940 and 1940 to
canization of rubber was discovered in 1839 1990, from 20 thousand to 100 million tons
and in 1850 gutta-percha was used to protect per year, a five-thousand-fold increase.
and insulate the first submarine telegraph
cables. Also in the 1850s, shellac was com- The fact that plastics were initially regarded as
pounded with wood flour to mould Union cases substitutes no doubt contributed to their poor
to protect and display early photographic im- popular image an image that was reinforced
ages. Albumen, principally from blood, was
- years by the inferior quality
in those and later
also compounded with wood flour to make oftheir design and application a criticism
-
that is sometimes levied even today.
Bois Durci, which was widely used to make
decorative inkwells, plaques and other items.
The Formation of the Plastics
At fnst, the new plastics were mainly regarded Historical Society
as substitutes for naturally occurring materials,
such as shellac, tortoiseshell and ivory. Such A History Discussion Circle was formed in
material was in short supply and was becoming Britain in 1968. but it was short lived and it
technically inadequate for the demands of in-
dustry. At the tum of the 20th century, casein,
a plastics material based on the protein from
milk, was introduced in Germany, and for
many years it was known as artificial horn. At
about the same time the search for a synthetic
substitute for shellac produced the material that
many consider the real starting point of the plas-
tics industry Bakelite. This material was pro-
-
duced by compounding phenol-formaldehyde,
a new and entirely synthetic resin, with wood
flour, the same type of filler that was com-
pounded 60 years earlier with natural shellac
to mould such items as Union cases and mirror Figure I Worldplastics productionfrom 1900 to
and brush backs. 1940.

38
which they have been involved, whereas many
others are motivated by the design possibilities
of plastics compared to other materials. The
warm touch and handling qualities of plastics
are also important to some collectors.

How to best care for a plastics collection is one


question that remains largely unresolved, but an
equally important consideration is the extent to
which plastics should be cleaned, polished or re-
stored. An example is the brown discolouration
that occurs to cast phenolic resins, such as
Catalin and Bakelite, the effect being particu-
Figure 2 World plastics production
larly apparent with blue or white colours. Their
from 1940 to
1990. ready discolouration in natural or artificial light
is one reason why warm colours, especially am-
ber, were more popular. M*y, if not most, cast
was not until alrnost 20 years later that suffi-
phenolic objects now appear much darker in
cient general interest in the historical aspects of
colour, and in many cases are of an entirely
plastics had developed for the Plastics Histori-
different hue to what they were at the time of
cal Society to become a reality.
manufacture. The colour and polish may be re-
stored by an abrasive cleaning process, but at
The PHS, founded in 1986, has established the
the expense of the removal of surface material.
following aims: The same is true of the more familiar moulded
Bakelite, but because of discolouration during
. To promote the study, preservation and moulding, original colours were usually drab.
sharing of information on all historical Nevertheless, colour change does occur the
aspects ofplastics -
original colour can usually be seen by inspecf
ing an unexposed part of the moulding, and one
o To encourage the recording ofany current may be tempted to remove the discolouration.
developments adjudged to be of value to Opinions on the advisability of such treatment
future generations differ widely and debates on these and related
topics are lively and sometimes a little
o To establish a National Plastics Museum controversial.

o To co-operate with other organizations With the aim of addressing these concems, and
interested in the historical aspects ofplastics the more urgent questions about conservation,
the PHS and The Conservation Unit (of the Mu-
The intemational membership of the PHS en- seums & Galleries Commission) have started a
compasses a variety of people, including some co-operative program of work into the degrada-
from all sectors ofthe plastics industry, staff tion problems of plastics materials in museum
from museums and educators, as well as and private collections. (This is discussed in
designers, historians and many others. more detail in a second paper by Morgan in
this publication).
Approximately half of the members are collec-
tors of plastics whose themes range from "any- Activities of the Plastics Historical
thing so long as it's plastic" to specialized Society
collections ofone particular era, one particular
material or a particular type of object. Reasons The PHS is an enthusiastic body of individuals
for collecting are as varied as the objects col- engaged in widely differing aspects of plastics
lected. Those from the industry itself are often history. Some of the topics listed below are
influenced by materials and processes with covered in their discussions.

39
o Recording reminiscences of the "old-timers" ainsi que le paftage d'informations sur l'histoire
who worked on many of the processes and de ces matidres, et defavoriser la consignation de
toute nouvelle ddcouverte qui pounait dtre jugee
materials that have been developed this
importante pour les gdndrations.futures. A long
century
terme, elle envisage de cnler un musie national
des plastiques. Dans I'entre-temps toutefois, elle
o Rescuing records from factories that are met tout en @uvre pour |viter que des dossiers ou
closing des objets d'impoftance du point de vae histori-
que ne soient ditruits. Et ce genre d'action prend
o Researching particular manufacturers, items tout son sens lorsque, par exemple, une entreprise
or materials 4tablie de longue date passe aux mains de gens
qui se prioccupent peu de ce qu'elle reprdsente
sur le plan histoique.
. Ananging exhibitions of plastics
La PHS entretient des relations avec d'autres
o Organizing meetings or visits of interest to organismes qui partagent les m€mes intdr€ts,
plastics historians d o nt I e Kun st st offe - Mus e um s - Ver ei n, une s o ci 6 t d
musdale allemande qui s'inttresse aussi aux
These topics, and many more, are reported in plastiques. Elle s'est de plus unie au Service de
the PHS magazine Plastiquarian together with conservation de la Museums & Galleries
news items and articles of general interest. Commission pour crder un programme qui per-
mettra d'approfondir les probldmes de dtgrada-
R6sum6 tion des matiires plastiques.

La composition de la Plastics Historical Society Parmi les membres diversifi4s de la PHS qui
a ses objectifs proiennent de la communautd internationale,
certains reprdsentent les secteurs de I'industrie,
Si 1962 a marqut le centenaire du plastique, de l'4ducation et de la recherche, tandis que
on n'a pourtant commenci qu'assez ricemment d'autres se recrutent parmi les concepteurs, les
d collectionner des objetsfaits d partir de ce collectionneurs ou le personnel des musdes ou
matiriau. En 1968, le Plastics Institute a crtb des entreprises de vente aux enchires. Par
un cercle de discussion historique en Grande- ailleurs, nombre des membres de la PHS possi-
Bretagne, et on a alors repris I'idde, lancte pour dent leurs propres collections de plastiques, qui
la premiirefois lors des ctltbrations du cen- couvrent toute une gamme de thdmes. C'est
tenaire, de noter les rtminiscences de certains donc dire que la PHS peut faire appel d des com-
pionniers du domaine des plastiques. Ce cercle pdtences et d des exptriences vaides. Son bulle-
n'a toutefois pas fait longfeu, et il faudra atten- rin Plastiquariu contient enfin des nouvelles et
drejusqu'en 1986 pour que I'intdr€t gdntral des articles qui sauront inttresser toute personne
pour les plastiques deienne assez grand pour qui se spdcialise dans la collection des plastiques
amener la criation de la Plastics Histoical ou dans I'histoire de ce matdriau.
Societ"v (PHS).
Reference
Lo PHS est une socitti indtpmdante, ffili4e au
Plastics & Rubber Institute, dont elle partage les 1. Redfarn. C.A.. "What is Plastic?" Plastics
quartiers gtndraux. Ses objectifs sont de promou- and Rubber International, vol. 2, no. 3,1917,
voir I'ttude et la conservation des plastiques, p.134.

40
Conservation Policies and Plans
Politiques et projets en matiire de conservation
A Joint Project on the Conservation of Plastics by
The Conservation Unit and the Plastics Historical Society

John Morgan

Plastics Historical Society


Plastics & Rubber Institute
London, U.K.

Abstract applied, whilst anticipation offuture degradation


problems malt enable the development of
Mounting concern about the deterioration of appr op n a t e prev en ti ve m eas ure s.
early plastics artefacts has led to co-operarion be-
tween The Conservation lJnit of the liuseums & Dffirent plastics degrade in diferent ways, and
Galleries Commission and the Plastics Historical it is therefore necessary to classifi objects accord-
Society ofLondon to devise a program ofwork ing to material. It is particularly important that
aimed at combating the problem. plastics susceptible to auto-catalvtic degradation
mechanisms-the most damagingform of deter, ,-
A step-by-step approach has been adopted, the ration-be identified and kept under obsenation.
first part of which is the publishing ofpreliminary
guidelinesfor the storage and displav ofplastics The progress and preliminary results of this w,ork
objects. These guidelines are based upon our will be presented and discussed.
current understanding of the ageing of these
materials. Introduction
The second part ofthe program is the determina- Museums and private individuals collect in-
tion ofthe nature and extent ofpresent degrada- creasing amounts of plastic, both as artefacts
tion problems so that appropiate priorities for in their own right and as component parts of a
research can be established. A survev encompass- wide range of items. This is an inevitable conse-
ing a w,ide range of histoical collections hai quence of the increasing importance of these
been started, the aims being to: provide details on
materials in our daily lives, starting with the
the o,pe of mateial collected; ascertain the pre-
exhibition of the fust semi-synthetic plastics in
sent condition of objects; determine Ihe atmos-
1862 and culminating with the arrival of "The
pheric conditions under which artefacts are
stored and displayed; and give information about Plastics Age" in 1979, when the volume of plas-
preventive and curative measures alreadlt being tics produced world-wide exceeded that of steel.
/bllowed by conservators and collectors.
Polymer degradation was observed in the l6th
It is already clearfrom the work to date that if cenfury by explorers who were returning with
plastics are Io suruive they will require more in samples of a newly discovered material. Later
the way of preventive consentation than has hith- known as rubber, the material was said to have
erto been given to nrost otlter materials. Proce- "perished" during the long sea voyage. Thomas
dures for arresting detefioration of plastics Hancock studied the degradation of rubber by
alrcady affected need to be established so that measuring the water loss from a sealed rubber
curative techniques may be researched and bag over 30 years, from 1820 to 1850. The

43
result, shown in Figure l, illustrates a tlpical o publish preliminary guidelines for the storage
course of degradation for many polymers; a and display ofplastics
relatively long induction period showing little
change followed by a period of accelerating . carry out a survey of plastics held in U.K.
degradation. In the 1850s problems due to collections
degradation of gutta-percha were encountered
when this material was first used as insulation o collate information relevant to the
on telegraph cables. degradation of plastics in collections

. provide training workshops for conservators


and collectors

o research stabilizing treatments and curative


techniques

The project is still in the early stages, but th9


preliminary guidelines have been published'
and the first of the training seminars took place
in May of 1992. The first stage of the survey is
in the form of a questionnaire, which has been
sent to a wide range of British museums, and
Figure I These data areJrom an experiment to also to several private collectors ofplastics.
determine water loss fi'om a sealed rubber bag. The
experiment was conducted by Thomas Hancock over
a period of 30 years, from I 820 to I 850. The data
Preliminary Findings from the Survey
illustrate a typtcal course of degradation for many
polymen: a relatively long induction period showing First, many private collections are devoted
little change follow'ed by a period of accelerating almost exclusively to plastics while others, for
degradation. example, button or toy collections, contain a
significant proportion of these materials. Quite
Degradation has serious economic implications often the objects are made entirely of plastic. In
to plastics manufacturers who have devoted museum collections, plastics are widely distrib-
much effort to limiting the extent of degrada- uted but often in comparatively small numbers
tion during the service life of their products. Im- and usually form only a part ofthe objects.
proved quality control and the use of stabilizing Very few museums have displays devoted to
additives has enabled plastics to be used in ap- these modern materials. Figure 2 shows the
plications from which they would otherwise be distribution of plastics in collections by period
excluded. However, little attention has so far of manufacture.
been paid to limiting the slow deterioration of
historical plastics, some of which were manu-
facfured before their true nature had been eluci-
dated and when the science of additives was
little more than a "black art."

In recognition of mounting concem about


the deterioration ofplastics in historical collec-
tions, The Conservation Unit (of the Museums
& Galleries Commission of Great Britain) and
the Plastics Historical Society have started a pre l88o- 1900- 19m- 194G 1960- post
18a0 1900 1920 1940 1960 1980 1980
co-operative program of work, the aims of
which are to:
Figure 2 Popularity ofvarious periods in plastics
collections.

44
The survey reveals that approximately half essentially linear or they can form a three-di-
of the museums replying to the questionnaire mensional, crosslinked network of relatively
possess more than 100 objects wholly or partly short chains. The former group comprise the
made from plastics. A slightly smaller propor- thermoplastic materials, that is, they soften and
tion report between 10 and 100, with only a flow when heated; the crosslinked polymers are
few (less than l0%) recording less than 10 or thermosetting, and once formed do not soften
more than I,000 such objects in their care. when heated and will normally char before
melting. Rubbers are a special class of thermo-
The most popular kinds of objects are those for setting polymer having long, flexible polymer
the home, followed closely by personal items chains with few crosslinks.
and toys. Mostly, they date from 1920 to 1960,
but surprisingly, more than one-third of the mu- Some linear polymers, such as polyethylene
seums report holding items manufactured since and nylon are partially crystalline, the degree of
1980. crystallinity varying with purity, processing
conditions and grade ofpolymer. Crystalline re-
Approximately twothirds of the museums gions are less permeable (e.g., to oxygen) than
report some deterioration of plastics, the most amorphous regions and as a consequence, deg-
common being crazing, discolouration or fad- radation may be less pronounced in the more
ing. Figure 3 shows the extent and kinds of crystalline material.
deterioration in plastics collections.
Polymers are also divided into two types,
addition polymers that are formed by "chain
growth" reactions, and condensation polymers
100 that are formed by "step growth" mechanisms.
Thermoplastics may be of either type whereas
o/ thermosetting plastics are almost always
formed from condensation polymers. Exam-
850 ples of addition polymers are polyethylene,
polypropylene, polystyrene, vinyl polymers
and acrylic polymers. Examples of condensa-
tion polymers are polyesters, polycarbonate,
0 nylon and phenol formaldehyde. The two types
Fadrng Bl@m Crang differ structurally in that addition polymers con-
tain only carbon atoms along the backbone of
the polymer chain whereas condensation poly-
Figure 3 Proportion ofcollections reporting various
mers also contain atoms other than carbon (e.g.,
forms of degradation in plastics objects.
oxygen) at regular intervals along the chain.
The two bpes are generally susceptible to dif-
The next stage of the survey will be to visit ferent degradation processes. The hetero-chain
a wide range of collections in order to obtain polymers, for example, may be susceptible to
more detailed information about the composi- hydrolysis chemical reaction with water
tion of objects and materials, particularly those resulting in-decomposition, or scission of
giving cause for concern. The results will be chemical bonds.
used to determine priorities in the final part of
the program, which is the setting of standards Physical Effects
for the care of collections, and the development These effects are associated with loss or
of stabilizing treatments and curative tech- migration of additives, absorption of liquids
niques for plastics materials. or vapours, crazing due to stress or fatigue,
mechanical damage, or excessive heat or cold.
Degradation of Plastics
Physical causes may be responsible for the
Plastics are made from long-chain molecules, following changes:
or polymers. The polymer chains can be

45
o distortion or dimensional change which may eventually lead to cracking, but the
. crazing or cracking maintenance of stable conditions will limit the
o surface deposit (often tacky) rate and, possibly, the extent of such changes.
o changes in flexibility
Plasticizer migration and loss may also occur
(Figure 4 shows typical degradation for cellu- due to changes in miscibility of the poly-
lose nitrate.) Cellulosic materials, casein and mer/plasticizer system brought about by chemi-
nylon are among the plastics that contain sig- cal degradation. In such cases, attempts to
nificant quantities of moisture, and their mois- reverse the process and replace lost plasticizer
ture content can vary with changes in humidity. by re-absorption would be ineffective.
Moisture absorption results in dimensional
changes and consequently the formation of Chemical Effects
stress. Varying humidity levels therefore pre- As with all materials, polymers can react chemi-
sent a low frequency fatigue stress and this is cally by coming into contact with other sub-
partly responsible for the crazing of casein stances. This includes those substances carried
plastics. in the atmosphere, such as oxygen, ozone,
moisture and pollutants. Chemical reactions
that involve scission of the polymer chain are
potentially the most serious. Because of their
network structure, thermosetting polymers can
withstand more chain scission than thermoplas-
tics and are generally found to be more stable.

Chemical effects are nearly always progressive


and irreversible, leading eventually to complete
Figure4 Cellulose nitrate shoehom showing disintegration. They are, therefore, usually
degradation 4,pical of this polymer. more serious than physical effects.

The combination of stress and certain liquids Evidence for chemical effects may be:
or vapours, which would otherwise have little
or no effect, can produce environmental stress- o colour change
cracking. The stress may be an applied stress or o chalkiness or surface bloom
it may be residual from moulding or machining c crazing
operations, or it may have developed with age. o embrittlement with loss of strength
This phenomenon became widely known from o evolution ofdegradation products (often
the premature failure of early polyethylene bot- acidic)
tles containing detergents. Rigid thermoplas- . softening or tackiness
tics, such as polysflrene and poly(methyl
methacrylate), are subject to stress-cracking in Major factors involved in bringing about
the presence of lower alcohols, paraffins, white chemical changes are:
spirit or oils. Acrylic materials intended for
outdoor exposure need to be annealed at a . light, especially ultraviolet
temperature of 80oC in order to minimize risk r heat
from this effect. . stress
o oxygen
Many physical effects can be controlled by the r moisture, including humidif
maintenance of stable conditions of tempera- . ozone and other atmospheric contaminants
ture and humidity and by avoiding mechanical (including those from nearby degrading
stress or contact with liquids and vapours that
objects)
might be absorbed. Migration of plasticizers in . contact with other materials (intentionally
such materials as cellulose acetate and poly(vi-
or by accident)
nyl chloride) result in these plastics becoming . some forms of biological attack
more rigid. This rigidity may cause distortion,

46
Not all of the above are damaging to any par- or synthetic rubbers are susceptible to this type
ticular material, but combinations of two or of degradation, which is potentially very dam-
more may be synergistic (i.e., the combined ef- aging as it can propagate at an accelerating rate
fect is much greater than the sum of the effects throughout the material.
considered individually). Antagonistic effects,
more simply known as "bad neighbours," can Light fades many colourants and discolours
occur between two materials. In this case, the many materials. Colour change can therefore
close proximity of one material adversely af- be a useful and responsive indicator of exces-
fects the stability of another. For example, the sive exposure to light. However, colouring mat-
copper core ofelectrical cables can accelerate ter may also mask early signs of deterioration.
the degradation of plastics insulation. A similar
effect was observed in the early 19th century Sometimes, degradation is conhned to the sur-
by Thomas Hancock the rubber coatings of face, as may be the case with some reactions in-
his waterproof textiles- were prematurely de- itiated by light on opaque materials. This type
composed by colouring matter present on ofdeterioration should be distinguished from
some dyed cotton fabrics. that where the whole of the material is affected,
as the consequences are much less serious.
Hydrolysis by atmospheric moisture is the main
degradation mechanism for the cellulose esters, Conditions to which a material has been
such as cellulose nitrate and cellulose acetate. exposed in the past often play a part in deter-
Hydrolysis plays a significant part in the degra- mining the nature and extent of future deteriora-
dation of polyesters (thermoplastic and thermo- tion. Removal of the object from adverse
setting) and nylon, and can also conhibute to conditions does not necessarily halt the degrada-
the degradation of polymer additives, such as tion processes. Conditions during polymer
plasticizers. The process usually results in the manufacture, compounding or moulding may
formation of acidic by-products, which acceler- also influence future degradation. For example,
ate hydrolysis so that decomposition proceeds during the manufacture of cellulose nitrate and
at an accelerating rate and the build-up ofacid- cellulose acetate polymers, acidic treatments
ity itself can initiate other degradation proc- are used and acidity that is not completely re-
esses. Depending on the polymer, the site of moved by subsequent processing is known to
hydrolysis may be in the polymer backbone (as affect the stability of these polyrners. Degrada-
is the case with many condensation polymers) tion may also be initiated by overheating the
or in side groups. The former is more serious material during moulding.
because it causes chain scission.
Control of Degradation
Hydrolysis is accelerated by acids or alkalies
and it is therefore important to ensure that Early detection of degradation and its likely
traces of such materials are not present in the cause is most important this can only be
vicinity of plastics prone to hydrolysis. -
achieved ifobjects are examined regularly and
their condition noted. Sometimes it may be nec-
Chemical bonds are also broken by other forms essary to conduct physical or chemical tests in
of energy, such as heat, light and mechanical an attempt to monitor the extent of any change.
forces. The broken bonds generally form highly
reactive species, namely free radicals, and the Plastics respond in different ways to the vari-
fate of the polyrner is controlled largely by ous environmental factors, and identification
chemical reactions associated with these. Light, will be necessary in order to appreciate how the
especially from the more energetic ultraviolet material might degrade. The minimum require-
region, causes bond scission in many polymers. ment will be the recognition of those materials
In particular, it may initiate an auto-catalytic that need to be kept under special environ-
form ofdegradation, known as autoxidation, mental conditions. The following tests and ob-
to which hydrocarbon polymers are especially servations may be sufficient for this purpose:
vulnerable. Polyethylene, polypropylene, ny- colour and appearance; handling qualities; hard-
lon, gutta-percha and plastics based on natural ness; density; end use; method of manufacture;

47
trademark; chronology; smell; signs of typical constant. Most plastics absorb moisture to
degradation; and specific tests. some degree and stress may develop as a result
of changing moisture content - stress is a
Confirmatory tests using various analytical contributory factor in many kinds of
procedures may still be necessary, especially degradation.
with objects giving cause for concem. The
large range of plastics now in production makes Ventilation
the task of identification more difficult for Ventilation is important for those materials that
recent material. emit gaseous degradation products. For exam-
ple, the cellulosic materials and poly(vinyl chlo-
Illumination ride) emit acidic products, which as well as
Light has long been recognized as a major fac- accelerating degradation may affect other parts
tor in the deterioration of polymers. It is damag- of the object or other materials and artefacts in
ing to all plastics and its exclusion is the single the vicinity. Metals may be corroded by the
most effective step that can be taken to mini- acidic vapours.
mize degradation. Since ultraviolet (UV) is the
most damaging part of the spectrum of light, it Cleaning
is important not to expose objects to it. Even Plastics materials often become soiled during
with ultraviolet removed, the light levels storage, handling and use. Many are electro-
should be maintained as low as circumstances static in nature and attract dust whilst others
permit. slowly bleed plasticizer to the surface and be-
come tacky. They do not seem to benefit from
Temperature Control the polishing action ofrepeated handling as do
Temperature influences the rate at which some other materials (e.g., horn and bronze).
physico-chemical reactions proceed the Perhaps they lack the right combination of hard-
higher the temperature the faster the -rate. It af- ness and wear characteristics. Only occasion-
fects the rate at which reactants, such as oxygen ally does a plastics object develop an attractive
and moisture, permeate into a material as well appearance or patina from the combined effects
as the rate at which reaction products permeate of ageing, wear and polishing. The most likely
out of the material. The optimum storage tem- plastics to do so are those for which colour
peratures for longevity of the various materials plays only a subtle part of their attraction, for
have yet to be determined. Long-term mainte- example, vulcanite and moulded Bakelite. In
nance of temperatures much below ambient such cases cleaning with a soft, dry brush or
would be expensive and may prove to be of cloth is probably all that is necessary.
little benefit.
It is recommended that plastics (with certain
For all plastics, the provision of a stable tem- exceptions) should be cleaned periodically to
perature is advised especially with objects con- remove surface contamination that may have
taining other materials that may have different built up over a period of time and, particularly,
thermal expansion coefficients. to remove any degradation products that might
accelerate deterioration. The frequency at
Humidity Control which this should be carried out will depend
The required level of humidity will depend upon circumstances, but with cellulosic materi-
upon the particular plastics material. For the als and casein it is recommended that this
cellulosic plastics, relative humidity should, should be done not less than once every five
ideally, be maintained below about 40%o. Ca- years.
sein and nylon become brittle and subject to
stress if their moisture content is allowed to be- Objects should be washed with tepid water con-
come too low, and for these materials a relative taining a small quantity of liquid detergent; the
humidity of 60oh is better. The moisture level use of a soft brush may be necessary for tex-
is less important with most other plastics but tured surfaces. Afterwards, the material should
for all groups it is important that it remains be rinsed with clean water and immediately

48
dried using an absorbent material. Do not soak half immersed in ESBO and left at room tem-
for long periods of time it is much better to perature for about two years. As shown in Fig-
rewash if soiling remains.- Only in rare circum- ure 5, the increase in size ofthe crazedarea
stances and after careful consideration should exposed to the atmosphere indicates continuing
solvents be used. Even if the material is not degradation. The areas immersed in the ESBO
directly attacked by the solvent, stress-crazing have not increased in size. Other objects show-
may result and this might not appear until later. ing initial signs ofdegradation have been
smeared with ESBO and the initial results
Care should be exercised with objects contain- are also encouraging.
ing metals that might corrode, or with hollow
objects that may be difficult to dry on the inner Conclusion
surfaces. Disassembly, if possible, is advised in
such circumstances. Some plastics objects have already undergone
serious changes to their properties and appear-
There may be occasions when objects, perhaps ance. These changes are unlikely to be revers-
soiled with tar or the residue from self-adhesive ible and all that can be done is to attempt to
labels, are difficult to clean with aqueous solu- retain those objects in a stable condition for
tions and in such cases one of the following the future.
solvents should be considered: white spirit, pe-
troleum spirit fuel for cigarette lighters and iso-
propyl alcohol (propan-2-ol). These solvents
possess optimum combinations of cleaning
power and inertness towards most plastics. Use
a cloth moistened with the minimum amount
of solvent. Do not immerse. rub too harshlv or
subject to stress. and do not use on rigid th-er-
moplastics, such as "untoughened" polystyrene
or acrylic because of the risk of stress-crazing.
Even for other materials. a wise precaution is to
test a small area on a hidden part of the mould-
ing. Remember that these solvents are highly
inflammable and toxic.

Degrading cellulose nitrate and cellulose ace- Figure 5 Cellulose nitrate mirror back showing pro-
tate that show crazing or porous areas should nounced degradation in four areos of the perimeter
not be washed. Stabilizing treatments are being at one-quarter and three-quarters ofthe height.
sought and the work on ESBO by Williamson' (Photo: cowtesy of C.J. & M.L. Ililliamson, The
described below shows promising results. Mansion House, Ford, Shrewsbury, U.K.
0743 850267.) Fig. 5a shows the mirror at an
early stage of the degradation. Fig. 5b shows the
ESBO and Cellulose Nitrate mircor at a later stage. The degraded areas in the
Cellulose nitrate is susceptible to hydrolysis by top half of the mirror that were not treated with
atmospheric moisture. Nitric acid is released ESBO continued to grow. The degraded areas in
and this promotes further hydrolysis so that, the lower half, which had been immersed in ESBO,
once started, degradation proceeds at an appeared to remain stable.
accelerating rate. It is possible that the applica-
tion of an acid acceptor, in a form that could mi- Plastics do slowly degrade in the atmosphere
grate into the material, might prove effective in
and their life will be considerably shortened if
retarding degradation by neutralizing the acidic kept under the wrong conditions. Since it is not
by-products. Epoxidized soya bean oil (ESBO) generally possible to eliminate completely the
has been tried. A mirror back, probably from
various environmental factors responsible for
the 1920s, exhibited typical degradation in the the deterioration of plastics, it becomes neces-
thick rim section in four distinct areas. It was sary to control the degree of exposure to them.

49
This will require classification of objects prioitaire. lJne ttude, couvrant wt grand nombre
into groups of materials, an assessment of the de collections historiques, a ainsi ttd amorcie
various risks followed by the provision of ap- dans le but de : founrir des d,ltails quant aux gen-
propriate storage and display conditions. Light, res de mattriaux qui se rett'ouvent dans les collec-
tions; dvaluer l'6tat actuel des objets: ddtenniner
humidity, temperature, ventilation and cleaning
les conditions ambiantes dans lesquelles les ob-
procedures all need to be controlled and
jets sont mis en rtserve et exposts: ffiir de l'in-
monitored.
.fotmation sur les mesures prtventives et curatives
qu'utilisent dtjd les specialistes de la restaura-
It is probably not feasible to provide the best tion et les responsables de collections.
environmental conditions for storing everything
collected and so it will be necessary to optimize Les travaux effectuds rdvilent clairement que,
the use of those facilities that can be made pour durer, les plastiques doivent, comparative'
available. This will require an assessment of ment d la plupaft des autres mattfiaux, Jhire
which materials and objects are at most risk I'objet de plus de mesurcs de conseruation
from deterioration. Such considerations should prtventive. Aus si faut-i I ilaborer des mtthode s
not be excluded when determining the collec- qui permettront d'arrAter la ditdrioration des
plastiques qui sont ddjd touchis
tion policy (i.e., what artefacts are collected).
- et donc effec-
tuer de la recherche dans le domaine des tech-
R6sum6 niques curatives et les mettre en application
tenter de prdvoir les problimes de ddgradation
-, et
La conservationdes plasfiques : un proiet mirte que risquent iventuellement de poser ces matd-
du Sewice de consewalion de la Museums & fiaux et donc ddfnir les mesures pr'4ventives
les
- appropri4es pour les combattre
plus
Galleries Commission et de la Plustics
Histoical Society
Des plastiques dffirents se digradent de manidre
Source grandissante d'inq uietudes, I a dit,lriora- dffirente, et il devient dds lors ndcessaire de clas-
tion des premiers objets de muste en plastique a sifier au prialable les objets en plastique suivant
amend le Service de conseruation de la Museums leur composition. Il importe tout particulidrement
& Galleries Commission et la Plastics Historical de reconnaitre les plastiques qui sont sujets d
Society de Londres d unir leurs elforts pour met- l'action des mdcanismes de dtgradation autocata-
de ddtdrioration la plus
tre sur pied un programme qui ise d enrayer lytique
- soit laetforme
destructrice de les garder sous obseruation.
ce probl'ime. -
Pour v arriver, on a adoptti une approche Nous dtcrirons, dans la prtsente communication,
comportant divers es ttapes. I'dvolution des travaux, tont en fournissant un
apergu de leurs rtsultats prtliminaires.
Ainsi, dans un premier temps, des directives
prdliminaires sur la mise en rtsewe des objets en References
plastique et sur letu" exposition seront publi4es.
Elles seront etubhes enfonction des connats' l. Morgan, J., Conservation of Plastics,
sances que I'on a dijd sur le vieillissentent de An Introductlon (London: Museums &
ce matdriau. Galleries Commission, I 99 I ).

Dans un autre temps, on cherchera d ddterminer 2. Williamson, C.J., Polymers in Conserva-


la nature et I'ampleur des problintes actuels de t n, Intemational Conference, Manchester,
io
dtgradation, afin de dLfinir les secteurs de la U.K., July 1991. (Proceedings to be published).
recherche qui devront €tre considirds defaqon

50
Conserving the Science Museum Collections

Roger Price and Anne Moncrieff

Science Museum
The National Museum of Science and Industry
London, U.K.

Abstract members of conservation staff and are given


basic training.
During 1989 to 1990 a nev,conservation
section was established at the Science Museum Like other museums that collect modent artefocts,
in London. Because itsfacilities are, at present, one oJ'the greatest challenges facing the Science
relativelv modest compared with some other na- Mttseum is how,to deal with the problem of the
tional museurns, the conservation section aims to degradation of modem materials. Although im-
protect the Museum's collections b1'rigorous proved quality ofstorage and displav might
managenent of its resources. help, research into new methods for tteatment
is urgentfu required. In the absence ofits own
In recognition of some damage that might have researchfacilities, the Museum has begun spon-
occurred becattse of the lack of clear guidelines, soring research at properly equipped institutions.
a neu'Code ofConservation Practice is being Bv such collaboration, u,hich need not be prohibi-
drawn up. This will include standardsfor storage tivebt expensive, some progress can be made in
and displat, (draw'ing on the Storer Report, spon- this dfficultfield.
sored jointll' b1' the Science Museum and the Mu-
seums & Galleries Commission), obserttations on In viev, of the diversity of materials collected,
at ailable h'eatntents for various materials, and all of u,hich have their own special problems, it
an explanation ofthe new conservation recording u,ould be sensible to broaden the research. Ide-
svstem. This document will serve as the basis for all1t, 17it should be organized on a world-w,ide
actions that might affect the presentation ofthe basis, with like-minded museums agreeing upon
Museum's collections. some wat' of atoiding duplication of work so that
resources can be more elfectivelv directed to-
In most technical museums, conservation is w'ards solving the problems that beset us.
undertaken bv skilled craft technicians. However,
few ofthem operate in the mainstream ofconser- Introduction
vation practice as it is understood, for example,
in fne arts or archaeology. Largefi, for that rea- Many museums specialize in aspects of indus-
son, some objects have been over restored, which trial and technological history and numerous
ffictively destroys evidence of historical signifi- non-specialist museums now acquire such arte-
cance. To remedy the situation, the Science facts as part of their wider collecting activities.
Museum runs a three-year Conservation oflndus- Among long-established museums of industrial
trial Collections Training Course specifically history, the size and scope of the collection is
designed to meet the needs oftechnicians.
greater than ever. New materials are replacing
Similarly, object-cleaning practices have been
reviewed. Obiect cleaners are nort included as
traditional metals and wood, making the

5l
conservation of industrial artefacts a complex lacking the rigour and discipline of modern
and demanding task. conservation practice. The Forum meets twice
yearly with the Director of the National Mu-
The National Museum of Science and Indus- seum of Science and Industry to discuss ways
try, which includes the Science Museum in in which the conservation of industrial material
London, the National Railway Museum and can be raised to a standard more appropriate for
the National Museum of Photography Film the needs of today and tomorrow.
and Television, is widely acknowledged as the
world's pre-eminent museum devoted to the An early achievement of the Forum is the 1989
history ofscience, technology, industry and publication by J.D. Storer, The Conseryation
medicine. Its collections are the largest, the of Industrial Collections: a Surve-v, sponsored
most comprehensive and the most significant jointly by the Science Museum and the Muse-
anywhere. To ensure that the preservation of ums & Galleries Commission. In this important
these outstanding collections fully meets the re- work, Storer surveys the facilities for conserva-
quirements of modern museum practice, a new tion and storage in 43 museums throughout the
conservation section was established durins country holding industrial and related material,
1989 to 1990. and makes a number of recommendations on
standards that should be met.
Maximizing Our Resources
Another direct result of the Forum's work is
The conservation section's most urgent task the 1989 inauguration of the Conservation of
was to identifu and quantify the major prob- Industrial Collections Training Course, run by
lems that threatened the well-being of the the Science Museum. As its most urgent prior-
Museum's objects so that, by rigorous manage- ity, the Forum advocates staff training in two
ment, the available resources could be deployed distinct areas: frst, the need to recover tradi-
most effectively. Quickly realizing that clear tional engineering craft skills (a need that must
guidelines or procedures were needed, work be and to some extent is being met by appren-
commenced on the compilation of a detailed ticeship to experienced craftspersons in muse-
Code of Conservation Practice. ums and private workshops); and second, the
need for a conservation approach based on the
The Code contains instructions to all staff on methodical study of an object, its history, dete-
how to use the conservation facility. This in- rioration and treahnent, of which a written re-
cludes everyone who is in any way involved in cord must be made.
taking actions that might affect the preservation
of the collections, including curators, designers, This new training course is specially designed
stores officers and building managers. It also to meet the needs of engineering technicians in
contains an explanation ofthe new conserva- the full-time employ of a museum. Over three
tion recording system, environmental standards years they receive 70 days of theoretical instruc-
for storage and display, and guidelines for con- tion during which they gain experience in the
servation staff on materials and techniques. wider issues of conservation. They must also
compile a written portfolio based on work they
Central to this initiative is the work of the have undertaken to meet the requirements of a
Conservation of Industrial Collections Forum. practical syllabus. Their progress is assessed
Founded in 1987, this group consists ofconser- regularly and there are examinations at the end
vators, curators and engineers from national of each year. Successful graduates are awarded
and local authority museums and related institu- a Certificate in the Conservation of Industrial
tions, as well as restorers in private practice. Collections.
For industrial collections in general, the ap-
proach to conservation is not as developed as The Science Museum has always employed
it is in, for example, fine arts or archaeology; staffto clean the objects displayed in the galler-
rather, the care ofindustrial objects is based ies. Staff cleaners have considerab le practical
upon traditional engineering craft skills, often experience but, until now, have had no formal
ofa very high order, but all too frequently training. In recognition of the continued need

5?
for this work, 12 cleaners were assigned to the Objects made from cellulose nitrate are also
conservation section in 1990. Three of them showing signs of deterioration. Because the
have since been promoted to act as supervisors Centre for Archival Polymeric Materials at
for each of three cleaning teams. In the last Manchester Polytechnic was already working
year, they have all had some training, which on the degradation of cellulosic film materials,
was based on standards used by the National the Science Museum provided funding for addi-
Trust. The large number of objects on display tional work. The project seeks to characterize
means that these cleaners will be busy dusting the polymer in both original and artificially
for most of the time, however, they will be aged samples, with the aim of finding methods
integrated more fully into the section and to assess the state ofdeterioration and to slow
given more haining so that they can work down the rate of decay.
with conservators on more specialized
cleaning tasks. The Museum has a few l9th-century cars
whose original upholstery is in poor condition.
Through this endeavor the Science Museum en- To conserve the upholstery rather than restore
sures that the objects on display are presented it with new materials, a joint project was set up
attractively and are also treated with appropri- with the Leather Conservation Centre in North-
ate care. Another important gain from such a ampton. New materials and methods were in-
regular cleaning programme is that the objects vestigated to effect the repair and reinforcement
displayed are subject to frequent detailed obser- of the old leather. So faq certain experimental
vation and any change in their condition can be methods have been successful and have been
recorded and treated where necessary. used on two of the cars.

Our Approach to Degrading Condition surveys on objects in the stores have


Modern Materials confirmed early impressions that some commer-
cial protective coatings are effective in retard-
Like other museums that collect modern arte- ing metal corrosion. The Museum will continue
facts, one ofthe greatest challenges facing the to use these, drawing on the experience of other
Science Museum is how to deal with the prob- users. The Museum is also discussing the
lem of the degradation of modern materials. choice ofoils for lubrication and hopes to spon-
Although improved quality of storage and dis- sor research into the degradation ofoils and
play might help, research into new methods for metal surfaces in collaboration with a univer-
treatnent is needed. With no research facilities sity research department outside London.
of its own, the Museum has begun to sponsor
research at institutions that have the staff and While working with existing research groups
equipment for such work. has the advantage ofhaving access to expertise,
it also has the slight disadvantage ofnot having
The Science Museum has some aluminium complete control over the project and the direc-
alloys, especially castings, that are corroding. tion that it might take. It is not easy to find re-
In collaboration with the Corrosion and Protec- search groups to work on the materials that
tion Centre at the University of Manchester interest the Science Museum; nor are most stu-
Institute of Science and Technology (UMIST), dents prepared to work on old materials. Fur-
the Museum has obtained a Science and Engi- thermore, the problems do not always fit neatly
neering Research Council Award for a Ph.D. into the time available to Ph.D. students. Never-
student to carry out research into this problem. theless, this is an effective way to use limited
The aim is to develop a technique that will resources and with the right partner such co-
reduce the rate of decay to a minimum. The operation can be successful and rewarding.
frst two years ofthis three-year project have
brought some success on simulated objects, There is a very real problem with complex
but the methods have yet to be tested in a objects, such as older computers. An example
real situation. of one approach used by the Science Museum
is the formation of the Computer Conservation

53
Society. Collaborating with the British Com- duplication ofwork so that resources can be
puter Society, the Museum has allocated more effectively directed towards solving the
a curator post and provided facilities for the problems that beset us.
enthusiastic and expert volunteers; the techni-
cians, engineers and computer operators who R6sum6
once actually worked with these same early
computers. La conservation des collections du National
Museum of Science and Industry
The Society first focused on restoring to work-
ing order a Fewanti Pegasus vacuum-tube En 1989-1990, une nouvelle section de conserva-
computer dating from 1958 and an Elliott 803 tion a ttt crtte att National Museunt of Science
germanium-transistor machine dating from the and Industrl, de Londres. Contme ses installations
sont,pour le moment, relativement modestes
early 1960s. This work has been fully docu-
comparativement d celles de certains autres
mented and is regarded as a continuation ofthe
mustes nationaux, cette section a pour mandat
working life of these objects. In the process of de prtsener les collectiotts du musie en assurant
this work young engineers and technicians are une gestiort rigoureuse des ressources.
trained to care for these machines. Though the
computers will only work for a few more years, Il a ttt ,"econnu que l'absence de lignes direc-
the experience enables the Museum to capture tfices claires explique certains des domntages
and record the expertise ofthe designers, build- qui se seraient produits, et l'on travaille donc ac-
ers, maintainers, programmers and users of tuellenrent d la ridaction d'un nouveau rtpeftoire
early computers. des rtigles et usages en matidre de consenation.
Ce document renfennera des normes pour la
The partnership between the Museum and these mise en rdserve et l'exposition des objets qui
computer professionals has also created a pro- s'inspireront du rapport qu'a prdpar6, sous - le
gramme of emulation whereby software is de- patronage mixte du musde et de la Museums &
signed to make a modern computer behave just Galleries Commission, la personne responsable
like the old machines, including their idiosyn- des rtserves
-, des observations sur les traite-
ments clui peuvent €tre appliquds aux divers
crasies. This enables anyone who is interested
mattriaux et des explications au sujet du nouveau
to experience something of the work of the
systdme d'enregistrement auxfins de conserva-
early pioneers. A publication on the work of
tion. Et il servira de document d'orientation
the Computer Conservation Society is planned
lorsque viendra le moment de prendre des me-
for 1992. sures qui pourraient influer sur la conservation
des collections du National Museum of Science
Conclusion and Industry'.

Conservation staff at our Museum keep in Dans la plupart des musies de la technique, les
touch with work going on in other museums by travaux de conservation sont confits d des tech-
attending conferences and through many infor- niciens et techniciennes qualifi,!s et expirimentds.
mal contacts, for example, the ICOM Modem Il s'en trouve ntanmoins peu patmi eux qui tra-
Materials Working Group and The Conserva- vaillent suivant les rigles de consenation qui
tion Unit of the Museums & Galleries Commis- ont cours, par exemple, dans un musde des beaux-
sion (whose Conservation Research Policy arts ou de I'archdologie. Et c'est sans doute ce
Group is keeping a record of conservation re- facteur qui explique le mieux que I'on retrouve
search in the United Kingdom). parfois, dans ces mttsdes, des objets qui ont 6ti
beaucoup tt'op restauris, jusqu'd en perdre pra-
tiquement toute signification historique. Soucieux
In view of the diversity of materials now col-
de faire sa part pour remtdier d cette situation,
lected by museums, all of which have their
le muste a mis sur pied un cours de formation
own special problems, it would be sensible to
de trois ans qui, portant sur la consentation des
broaden research into their deterioration and re- collections indwtielles, est spdcialement conQu
medial measures. Ideally, this should be organ- pour rdpondre aux besoins de ces techniciens et
ized on a world-wide basis, with like-minded techniciennes. Les usages entourant le nettoyage
museums agreeing upon some way of avoiding

54
des objets ont 4galement itd passts en retae, si qu'ffictuent, dans ce domaine, des organismes
bien que des personnes sptcialement affectdes ad,lquatement equipds. Il devient d'is lors possi-
d cette tdche ont 6tt inttgries au personnel de ble, grdce d une telle forme de collaboration, de
conservation, et qu'un cours deformation de rdaliser des progrds dans ce secteur plut6t dffi-
base a tti mis sur pied d leut'intention. cile, et ce, sans que le muste ait d effectuer de
gros ddbours,ls.
Le probldme de la dtgradation des mattriaux
modernes constitue sans doute l'un des plus Compte tenu de la diversitt des matidres qui se
grands d6fis auxquels le musde ait dfaire face d retrouvent dans les collections, et qui prisentent
I'instar des autres musdes qui collectionnent des toutes des probldmes particuliers, il serait
objets modernes. Et s'il est sans doute utile de rai so n n abl e d' 4largir l' ho izon des rec herc he s.
chercher d amdliorer la qualite des methodes de Une telle entrepise devrait iddalement €tre in-
mise en rtserve et d'exposition, il n'en demeure tegrde d un projet de portte mondiale, dans le
pas moins urgent de poursuivre la recherche de cadre duquel les mustes ayant les m€mes intdr€ts
nouvelles m'lthodes de traitement. Ne disposant se seraient mis d'accord sur unefaqon d'iviter
pas d'installations de recherche qui lui permet- les dedoublements, etferaient ainsi en sorte que
traient de poursuivre lui-m€me de tels trava*r, les ressources serent effectivement d la rdsolu-
le musde a commencd d parrainer des recherches tion des probl?mes.

))
History of Technology
Histoire de la technologie
Rubber: Its History, Composition and Prospects
for Conservation

M.J.R. Loadman

Malaysian Rubber Producers' Research Association


Tun Abdul Razak Laboratory
B r i c ke n donbury, H ertford
United Kingdom

Abstract material. Possible causes and effects of surface


changes will be discussed and ways of distinguish-
Everyone living in the "modern" world is familiar ing between them presented.
v'ith rubber and its properties. Perhaps ii is that
ven'familiarin that has bred. if not iontempt. at Knowing v'hat the material is, and from the state
least an unthinking acceptance ofthe material of its surface what has happened to it so far in its
and its position in our civilization. The term ,,ntb- life, consider now what can be done to optimize
ber" is used to descibe any polyner that has or its display life. Wilst there is no one answer for
appears to have elastic properties, and there are all elastomers and their products, many suJfer
many such polymers of both natural and sltnthetic degradation through a series of common phenom-
oigin on the market today. In a "rubber,,prod- ena: oxygen, ozone, heat, light, mechanical work,
uct. the elastomer may well be less than 509o 61 pro-oxidant metals, bacteria and even that
its total mass since many additional chemicali dreaded word "chemistry"; combinations of these
ma1'have been added to produce a seniceable phenomena often have a synergistic effect. An ar-
article, and these in turn may have been selected tempt will be made to weove a path through this
from h un dre ds of p o ss i bi li tie s. minejield so that the best conservation procedure
'of
is chosen for each of the various tvpes elas-
It is not intended, in this paper, to delve deeply tomeric products on display.
into the scientific details of rubber product formu-
lation, but it is essential that, before starting any
conservation process, the conservator is aware of Introduction
the material with which he or she is dealing and
how,inappropiate treatment can do more harm There can be no-one living in the "modem"
than good. The history of elastomeric materials world who is not familiar with rubber and its
will be reviewed, with particular attention to de- properties, but perhaps it is that very familiarity
velopments in the 20th century, which might as-
that has bred, ifnot contempt, at least an un-
sist the conserttator in classtfying and possibly
thinking acceptance of the material and its
dating any elastomeric product. To further these
position in our civilization.
aims, a brief description of some simple methods
of analysing these materials will be included.
The natural material has been used for at least
Visually displeasing changes to the surface ofan 2,000 years. It may, even today, be used "raw"
elastomeric product are a common event and can for crdpe soles of high quality shoes, or mixed
sometimes pres age surface degradation, al though with chemicals in the latex state, prior to hav-
this is not alwal,s the case: some surface changis ing formers dipped into it to produce such
are intended and are beneficial to the life of the articles as baby bottle teats, condoms or

59
surgeons' gloves. The mixed (or compounded) ozone and even that dreaded word "chemistry";
latex may also be treated to produce latex combinations of these phenomena often have a
thread suitable for the finest underwear whilst, synergistic effect. An attempt will be made to
at the other extreme, the dried rubber can be weave a path through this minefield so that the
mixed with more chemicals, often including best conservation procedures may be selected
carbon black, to manufacture the strongest of for the various elastomeric products on display.
engineering products, such as base isolation
units for buildings in earthquake zones, con- The history of natural rubber, or NR as it is usu-
veyor belts and, accounting for by far the great- ally known, from a plaything over 2,000 years
est area ofusage ofelastomers, aircraft and car ago to the founding member of one of the most
tyres. If you doubt the remarkable properties of important classes of materials of modern day is
this material, remember the faith you put in the a fascinating story, unequalled, I believe, for
four 'handprints' ofthe tyres on the road be- any material. The more recent development of
neath your car. synthetic elastomers is equally interesting, aris-
ing as it did from a mixture of the chemist's ba-
In the last 150 years or so, hundreds of chemi- sic desire to mimic and improve on nature and
cals have been added to rubber to modify the the absolute need for a synthetic replacement
polymer itself, or its properties. And numerous for NR by both Germany and the United States
synthetic elastomers have been invented and during the Second World War.
manufactured, put through the same process of
mixing with all the chemicals used with natural Highlighted in an appendix are those events in
rubber, and more. Yet still, for general purpose the history of rubber that seem to me to be im-
applications, and, indeed, many specialized portant, or interesting, in the development of
ones, the sap from the weeping tree, which the rubber industry of today. The list, although
originated in Brazil, gives rubber, a material extensive, is still superficial and very personal,
unequalled in its many useful properties. but even more brief is the summary included
here.
With such a wide variety of end products and
constituent chemicals it is impossible to do Although the oldest rubber known is fossilized
other than generalize, but there are some useful and some 60 million years old, the detailed
observations that can be made. modern history began some 2,000 years ago
with the New World involvement in "rubber"
Materials on display must look good, be they in produced from a "weeping tree." It is probable,
a shop window to encourage sales, or on dis- however, that the earliest rubber came from
play in a museum. Surface deterioration is a Castilloa elastica and not the Hevea brazilien-
common event in both environments and can sls, which produces essentially all the world's
sometimes presage surface degradation, al- natural rubber today.
though this is not always the case; some surface
changes are intended and beneficial to the life Skipping past references from Columbus and
of the material. The possible causes and effects Torquemada, the first real European involve-
of surface changes will be discussed, and ways ment was by the French during the middle of
of distinguishing between them presented. the l8th century, when de la Condamine and
Fresneau attempted, without success, to start
Knowing what the material is, and from the the first manufacturing plants for rubber goods
state of its surface what has happened to it so in Europe. Their main interest was in latex dip-
far in its life, consider now what can be done to ping but they could not ship the latex to Europe
optimize its display life. Many ideas have been without it coagulating a visual and olfactory
put forward, some possible and some definitely -
effect similar to milk curdling!
counter-productive. Whilst there is no one an-
swer for all elastomers and their products, For 50 years little happened, except that
many suffer degradation through a series of Priestly coined the name "rubber" for the mate-
common phenomena: heat, light, mechanical rial that rubbed out pencil marks. Then, within
work, pro-oxidant metals, bacteria, oxygen, a very short period from 1820 to 1839, there

60
was a tremendous growth of interest, when in none succeeded in the long term except
the United Kingdom, Hancock invented his ma- Hevea braziliensis and this has not been bet-
chine to convert lumps of solid rubber into a tered, although the American govemment re-
useable homogeneous gum (a process he called search continued into Guyaule right up until
"pickling" to confuse his competitors), and last year, and during the 1930s the Russians
Macintosh developed his three-layer water- used dandelions as a source material, having
proof fabric. In North America, at the same access to nothing else.
time, a substantial market developed for
dipped rubber shoes, whilst Chaffee invented In 1873 Henry Wickham appeared, and the
his mill and calender, the designs of which story of his shipping 70,000 seeds of Hevea
are basically the same as those in use today. braziliensis to Kew Gardens in London has
Chaffee also founded the first American entered rubber folklore, mostly due to Wick-
rubber company. ham's and others' fireside embellishments of
the diaries and notes that Wickham kept on his
By 1839 the American rubber bubble had burst, trip. What rs known rsthat2,397 seeds germi-
mainly due to the stink of crude unstabilized nated at Kew. and 1.919 of these were sent to
rubber as it decomposed in the heat of summer, Sri Lanka, l8 to Java and "a few" to Singapore.
but also accelerated by the economic collapse Yet little is made of the fact that the British gov-
of the country. In the U.K. Hancock and Macin- ernment forgot to pay freight charges and most
tosh survived due to their company's "Macin- of the seedlings died! Here precise details be-
tosh's Waterproof Double Textures." However, come cloudy, but a month later another 100
the product, although better than anything else, seedlings were sent to Sri Lanka from Kew
was still liable to putrefu and, whilst "state of and yet another 100 were sent in the following
the art," it could notbe consideredperfect. year (1877).

In that same year of 1839, Goodyear discov- In 1878, 22 seedlings were sent from Sri Lanka
ered, by accident, that heating a mix of rubber, to Singapore. Ridley, the Director of the Bo-
white lead and sulphur resulted in a highly tanic Gardens of Singapore and the man who
elastic material that was rubber "cured" of its could fairly claim to be the father of the rubber
problems. It no longer went brittle in the cold plantations in Malaysia, believed that these
and soft in the heat, nor did it seem to putrefr were seedlings from a batch of Brazilian seeds
so easily. Thus the process of heating rubber sent to Kew by another collector, Robert Cross.
with sulphur became known as the "curing" These were the seedlings from which most of
process as well as vulcanization, a name the world's rubber trees have developed. Thus
probably coined somewhat later by an other- it is possible that Cross, not Wickham, should
wise unremembered friend of Hancock's. be thought ofas the true father ofthe natural
Mr. Brockedon. rubber industry. In 1899 the first "plantation
rubber" was shipped from Sri Lanka and the
In 1857 Thomas Hancock published his classic rubber production industry was bom. For an-
guide to the U.K. rubber industry and his illus- other 50 years "wild rubber" from trees, not
trations give some idea of the breadth of uses necessarily Hevea braziliensis, continued to
to which rubber was being put. Not many are come out of Brazil. This was often called "Para
missing from a list of today, since these include rubber," after the port of Para from whence it
air-proofproducts, hoses and tyres, nautical, do- was shipped.
mestic and travel equipment as well as a range
of seals and washers. To the following people, therefore, lies the
credit for founding the rubber industry of to-
In the second halfofthe l9th century explorers day: Goodyear, the idealist and fanatic who
around the world were searching for plants that sacrificed his family, income and health in his
could produce "rubber." These were being attempts to do anything and everything with
shipped to politically and climatically accept- rubber; Hancock, the pragmatic businessman
able locations in the hope of discovering one first, and engineer second, who came to grips
that could be "harvested" profitably. In fact, with the mastication of rubber: Chaffee.

6l
unlmown to most today, but the inventor of the general-purpose elastomers, including NR,
still standard two-speed, two-roll mill and its account for some 80% of the current elastomer
extension, the calender. On the rubber produc- market.
tion side is the famous Wickham (or perhaps
less known, Cross), Hooker and Markham at To my mind, the most important features of
Kew Gardens and Ridley at the Botanic Gar- this summary in the area of conservation are
dens of Singapore. Perhaps, furally, the motor the following:
car should be mentioned, hardly conceived dur-
ing the period covered by those listed above, r Any artefact manufactured before 1840 is not
but the undoubted catalyst to launch the rubber lulcanized.
industry into the 20th cenflrry on an increas-
ingly large scale. e Identification of curatives/additives will help
in dating products manufactured between
In this century a few critical dates can still be 1840 and 1920.
picked out from the great mass of scientific
advancement: o Identification of the polymer is essential
after 1930.
o 1921to 1923 - The development of most
of the chemicals used to "accelerate" and r Non-black filled products were not protected
clean up the chemistry of sulphur with antidegradents before the mid 1950s.
lulcanization
Obviously, more detailed datings may be possi-
o 1925 - Can be considered the start of the ble by considering all ingredients found, and
synthetic rubber industry thus it is worth considering what can be done in
this field. I have not gone into scientific detail
o 1930s - The introduction of amine here, as knowing what can be done, not how to
derivatives as antioxidants do it, is the important thing and analytical de-
tails are not needed to ask the experts to look
o 1933 to 1945 - The development of familiar for something.
synthetics, such as polybutadiene, styrene-
butadiene, nitrile, chloroprene (neoprene), Aspects of Analysis
polyurethane and silicone
Excluding surface effects, which are considered
o 1950s - The introduction ofnon-stainins later, useful information can be obtained
phenolic antioxidants through these three questions:

o 1958 - The introduction of fluoroelastomers o What is the polymeric system?

o 1962 - World production of natural and o What can be said about the cure and
synthetic rubbers each reaches two million protective systems?
tons per annum
r What inorganic materials are present?
o 1990 - Worldoutputof rubberaround
l5 million tons; 2:l synthetic:natural The first is most easily answered by pyrolysis
followed by either infra-red spectroscopy
It is worth noting that, whilst many special (P-IR) or gas chromatography (P-GC). Charac-
elastomers have been developed that can oper- teristic traces that are obtained for each poly-
ate in environments where NR would not func- mer using both techniques and polymer blends
tion satis factorily, none of the general-purpose can be quantified ifthe analytical conditions
elastomers can offer any improvement over are right. Typical patterns for NR are illus-
NR for run-of-the-mill applications. These trated in Figures I and2. These techniques
suffer from one disadvantage in that they

62
cannot be considered non-destructive. Neverthe-
less, P-IR can use only l0 mg whilst P-GC
requires, say, 100 pg and this can usually be
spared from any material. Figure 3 shows a
sample ready for P-GC analysis; the baby
feeder teat gives an idea ofthe scale. The
equipment need not be highly sophisticated I i r'
l

and costs would be 10,000 to 15,000 dollars. .il,


iii, i. l

Examination of the cure and protective systems


is potentially the cheapest of all analytical
techniques, if a gram or two of material can be
spared, since thin layer chromatography (TLC) Figure 2 Pyrolysis gas chromatogram of natural
can be used to separate and visualize virtually rubber. -
all of the candidate chemicals. The sample is ex-
tracted with a suitable solvent, and this solution specific for oxidative degradation of NR, giv-
is then concentrated and spotted onto the ing an orange spot and streak, whilst another en-
bottom of a TLC plate. This is then placed in a ables a distinction to be made between NR and
special tank containing a low volume of a liq- the material that is its nominal synthetic equiva-
uid. which elutes the various components up lent, generical ly call ed " PI" for polyisoprene.
the plate at differing rates. Treatment of the
"developed" plate with particular chemicals pro- If the sample size is limited to milligrams, high
duces variously coloured spots from the compo- performance liquid chromatography (HPLC) is
nents present, which enables identification of the analytical method of choice. This method
the materials with a high degree of specificity, can give as much data, and quantiff it, but costs
provided reference materials or data are avail- more 30,000 dollars upwards!
able. Thus, for a few cents, much useful -
information can be obtained. It may also be It should
be noted here that I have talked only
ofinterest to note that one spray is relatively ofsulphur vulcanization and not ofany other
t1pes. Since the discovery of modem accelera-
tors in 1920, sulphur vulcanization has domi-
nated the methods of crosslinking polyolefins
and probably accounts for over 95o/o of the
market. Peroxide, urethane, and radiation

ll
tl

Figure I Infra-red spectra of a natural rubber Figure 3 Sample being inserted into a pyrolyser coil
pyrolysate (top) and thinfilm (bottom). for P-GC identific ation.

63
crosslinking of polyolefins must be considered be taken to examine a piece Iiee of surface con-
"niche" areas for very specific properties. tamination or, more safely, a section through
Considering the whole elastomer field and the bulk of the material. These instruments are
pseudo-elastomer field, there are two signifi- not cheap but my experience in the U.K. is that
cant exceptions to sulphur wlcanization, the most museums know of their potential and are
pollurethane and silicone rubbers. These have able to persuade either a local university or in-
multi-functional groups in a few molecules of dustrial organization, such as my own, to have
low molar mass that enable three dimensional tu*:,""*u-tn"o''0oe@e
networks to build up during pollmerization.
Here the word "l'ulcanization" is not appropri-
3200
ate whereas "curing" might be. In fact, this OX€s A-4r Pa&b *m!ed aom
Conb'Mi€d L&lDr €atch
2aaa
latter word is used by the resins and plastics in-
dustry quite happily, and reasonably, since their 24

crosslinking processes also cure the sticky and


I

flow problems of the starling materials.


1

12gO I

Finally, any inorganic elements present must aoo I

be considered and this includes any added sul-


phur that would indicate a cured product. For
400
sl ct K
1r
1.1\
simplicigz, and even the chance of fiue non- ot23456789t0
destructive analysis, the scanning electron Figure 5 X-ray spectrum of contaminant with, inset,
microscopy (SEM) with a built-in electron mi- a micrograph of the particle.
croprobe X-ray analyser must be considered.
The X-ray spectrum will immediately identify Such a technique was of instant importance
the elements present and allow a "guesstimate" a few years ago when I was asked to identify
of their levels. Figure 4 shows the spectrum of the "elastic" material found inside a sealed am-
a wlcanizate with zinc and sulphur together phora during a Middle-Eastem excavation (Fig-
with aluminium and silicon, the last two indicat- ure 6). lt was assumed to be a seal of some soft
ing a clay filler. The characteristic peaks ofbar- and, whilst having a brittle skin, was obviously
ium suggest barium sulphate, used to increase "elastic" or rubbery undemeath. A minute
the density of cerlain materials, and there is sliver was taken from within a whorl and identi-
also a trace of iron. In Figure 5 an impuriqz is fied by P-GC as NR. This was exciting as there
identified as rust; it must be emphasized that was no record of rubber being used outside
this is a surface technique with a sample pene- the New World prior to 1500, although with
tration of about 10 microns. This feature is ob- dandelions and the many hundreds of other
viously relevant to the next section; in the rubber-bearing plants aiound. it was not impos-
context of bulk element composition care must sible. SEM X-ray analysis threw the "cat
amongst the pigeons" as this showed both zinc
"t,
and sulphur at "modem" levels of addition.
:"" , , HPLC examination of a further sliver showed
mercaptobenzothiazole (MBT). This placed the
artefact after 1920 and thus raised doubts about
the other artefacts found in the vicinity of the
amphora. It certainly caused local red faces, but
at least prevented a much larger scale of embar-
rassment. There are many different sources of
analytical techniques that can be used to exam-
ine elastomers, and one of which I am co-
author, listed in the bibliography, discusses
these in some detail. Many useful tricks are
Figure 4 Scanning electron microscope documented there and perhaps there is one that
- energ,,
dispersive X-ray spectrum of afilled rubber deserves mention here, as it is cheap! All that is
vulcanizate. required is a press with platens that can be

o+
chemical that is intended to bloom is wax, the
sole reason for its adding being to bloom and
produce an ozone-impelious film on the sur-
face of the rubber product.

Modified blooms result from protective agents


on or near the surface of the rubber reacting
with the environment and undergoing a chemi-
cal change. The non-reacted chemical in the
bulk of the rubber then migrates to the surface
to decrease the concentration gradient, where it
is then further consumed, eventually building
up a protective skin. At this stage the migration
stops until the skin is removed or broken and
"repairs" are needed. Classic examples of this
heated to about 175'C. Hot pressing non-'urrl- are the paraphenylenediamine antidegradents.
canized rubber will result in a smooth film, but
if the rubber is rulcanized it will stretch and Pseudo blooms are what I call dull surfaces,
tear before shrinking to something like its origi- which appear to be suffering from a bloom but
nal size. One important application is in distin- which on closer examination, often under an
guishing between vulcanizates and such electron microscope, are actually suffering
materials as plasticized PVC (arguably called from surface degradation of the rubber. In very
"rubber") and the thermoplastic elastomers, extreme cases, sometimes called "chalking" or
which are taking an increasing share of the
"frosting", the rubber is completely eaten away
current elastomer market. leaving a powdery surface of the inorganic
filler. This has led some people to claim that
Surface Effects fillers can bloom, but this is not possible as
they are insoluble in rubber. They only "come
Examination of the surface of a rubber product to the surface" because the surface is eaten
or artefact immediately reveals whether the sur- away. Also included in this category is surface
face looks good and "respectable" or whether
crazing,to which I will refer later.
there is something wrong with it. If the latter is
true, there could be several reasons and these Surface contamination, staining and
need to be identified before carrying out any discolouration are always problems, particu-
treafinent. The following categories of surface larly with the sort of materials conservators
effects can be defined. But before discussing might be examining. These have to be identi-
each in tum, I must emphasize that it is possi-
fied, possibly by SEM, before taking action to
ble for more than one of these effects to be restore the material, remembering that any
present at the same time.
"clean-up" procedure must take into account
surface effects or defects if fuither damase is
A true bloom is a thin layer on the surface of a
to be avoided.
rubber article of one of the chemicals added
before, or produced during, lulcanization. This The final category, hazing, only applies to
chemical actually migrates from the bulk to the transparent or translucent products and is
surface by a mechanism that will not be dis- caused by either insoluble crystals, particles or
cussed here beyond saying that the chemical immiscible liquid droplets suspended in bulk
has to be soluble to a ceftain extent in the rub-
rubber. These may or may not migrate as well,
ber and be present at a level above that solubil- but little can be done about them. Although not
ity. Chemicals that can bloom are sulphur, zinc a surface effect, it is not always possible to dis-
dialkyldithiocarbamates, mercaptoben zthiazole tinguish between the two with a cursory exami-
and zinc mercaptobenzimidazole. With the pos- nation. Tables I and II illustrate some simple
sible exception ofthe last, they are not intended tests that can be used to provide information
to bloom in a properly formulated product. One

65
on possible surface effects. It is usually possi- area, once again I recommend IhebookThe
ble to find some part of an article that can be Analysis of Rubber and Rubber-like Materials,
experimented on, but as I mentioned earlier, as listed in the bibliography.
more than one effect may be present at any one
time and therefore a degree of care is required Ageing
in drawing any conclusions. The tests them-
selves are perfectly straightforward and do not In any discussion about ageing it is enlight-
require any further elaboration. ening to begin by quoting from Hancock's
classic book The Origin and Progress of the
A more sophisticated examination should be CAOUTCHOUC, or India-rubber Manufacture
carried out ifpossible and it is obvious that the in England, which was published in 1857. The
SEM with X-ray analytical microprobe must be following observation seems to have been
the technique of choice, although some informa- made sometime during 1825 to 1826:
tion is available from chemical spot tests and
multiple internal reflection infra-red spectros- The injurious effects olthe srm's rays upon thin
copy as well as IR microscopy. Once again, films of rubber we discovered and prolrded agarnst
before much damage accrued. All these things are
there are many fricks that can be used to facili-
now cheaply known to those who have followed us
tate examination, but these are beyond the by men leaving our employ and the specificatrons
scope of this paper. For more details in this of our patents; but they had all to be undergone in

Table I
Simple Tests to ldentify "Bloom" Types

1) Did bloom develop in storage?


I

yes
I

I
Surface contamination or inorganic haze
v
2) Does it disappear on heating?

yes Oxidized antidegradant,


Surface degradation,
Basic zinc stearate
I

it
3) Can it be removed with solvent wipe? Basic zinc stearate
ri
vlno
r

I
yes
Try other.o,u"nrl
yes no
I
yes
''l
Try complexing agent-t
I

I I

J lro
V
True bloom <- Haze due to organics

66
our early progress at an enorTnous cost, as well as not prepared (quite reasonably) to compromise
trouble and vexation: and none but those who have
passed through the ordeal can concelve the mortrfi-
this performance so that the product might be
catron experienced during those years. displayed for a few hundred years in a museum.
These effects may also interact so, with the
It would appear that there really is nothing new exception of the protective agents themselves,
under the sm! Unfortunately there is a great dif- none will be discussed in any detail. Their
ference between knowing about something and significance will become apparent as the
circumventing it particularly for the time discussion continues.
-
scale that interests museums. The factors that
can influence the life of an elastomeric product Ageing processes involve three distinct mecha-
fall into fwo categories. which I will cafl "prod- nistic routes that can be identified as follows:
uct characteristics," about which nothing can be
done, and "ageing processes." about which o continuing sulphur .vulcanization chemistry
something can be done.
. shelfageing
Product characteristics describe the product as
it was made and include such variables as: . atmospheric ageing
which elastomer was used; whether the product
is filled or not and if so, whether it is black, col- Sulphur chemistry is particularly important in
oured or white; the vulcanizing system; the pro- lulcanized products that are resisting a distort-
tective system; whether it is properly designed ing force. Chemistry does not stop when the
and manufacfured or has built-in stresses; and vulcanizing temperature is reduced from some
the size and shape. 150'C to ambient it just proceeds more
slowly. One of the-features of a r.ulcanizate is
All of these have an effect on the product's known as its "set" and this is due to the fact that
service and storage lives but unfortunately the C-Sx-C bonds, which form the chemical
there is a conflict in that the manufacturer is crosslinks between polymer chains, can break
concemed with optimizing the product's per- between sulphur atoms and then the loose ends
formance during an acceptable service life with re-attach to either their original partner or an-
some pre-use storage, but the manufacturer is other loose end. Ifthe vulcanizate is under a
distorting force, then, as the bond breaks, the
elasticity of the polymer chain causes it to relax
against the force so that the bond will re-form
Table II further down the chain. The result is a slow
Colour Changes and Their Significance in "slippage" of the lulcanizate, which cannot be
"Bloom" Analysis reversed. This can be minimized by the appro-
1) ls the colour change light-rnduced2 (Mask some
priate choice ofcrosslinking agents, since the
of sample for reference) so-called "effi cient" vulcanizing systems.
Grey/Brown amrne a,/o which are low in sulphur and high in accelera-
Prnk phenolic a,/o
tor, have a much higher proportion of monosul-
2) ls the colour change heat-rnduced? phidic crosslinks than "conventional" cure
lf yes, wtth
systems. As the C-S bond is stronger than the
strckrnesyembnttlement oxdatNedeoradatton
lf yes, wtthout S-S one, it will not break so easily, but, once
strckrnesyembnttlement amrnea,/o there is a wlcanizate, its propensity to set can-
3) ls the colour change patchy?
not be changed. It is essential, therefore, that
Reactron of external contamtnatton wtth vulcanized (and of course non-vulcan ized) arte-
rubber chemtcals facts are stored and displayed with as little dis-
4) Does colour change appear after laundering or torting forces as possible. Vehicles, for
contact wth metals? instance, should not be left on inflated tyres.
Copper/lron reactton wtth dtthocarbamates but should be blocked up and the ryres ieflated
5) Does colour change accompany poor agerng? to a few pounds per square inch (psi). Although
Further evidence for Copper/lron contamt natton the chemical changes will continue in an un-
stressed object and are significant in terms of

67
product usage, they will not significantly affect In real life rulcanizates, particularly when
the appearance ofa display object. black-filled, it seems that the ingress of oxygen
is limited and, apart from the surface 0.5 mm or
Shelf ageing is basically oxidative ageing and, so, the bulk rubber remains in excellent condi-
apart from the obvious influence ofoxygen, the tion. It is this experimental fact that enables
catalytic effects ofheat, light and pro-oxidant conservation scientists to predict with confi-
metals have to be considered. If the chemistry dence service lives well in excess of 100 years
of oxidative degradation was considered in de- for engineering products, such as base isolation
tail it would fill in a lifetime or two and still units for buildings, or bridge bearings that take
leave unanswered questions, but in a simplistic up the expansion and contraction of most mod-
form, it is fair to say that oxidation of a sulphur em bridges.
lrrlcanizate proceeds via at least one chain reac-
tion sequence. This, fundamentally, introduces Light-catalysed oxidation is particularly no-
C-C and C-O-O-C crosslinks between polymer ticed in unfilled or light-coloured objects and
chains as well as C-O-O-C rings within the can result in an inelastic skin which, as it
same polymer chain, whilst another set of thickens, cracks in random directions and pro-
reactions between the sulphur atoms of the duces a pattem known as "crazing" and this is
crosslinks and oxygen breaks the crosslinks. illustrated in Figure 7. In thin sheets the effect
The mechanism of chain scission is still not has been called "light stiffening", whilst in
completely understood but theories have been highly filled articles the degradation can result
proposed that involve both the "in-chain" in complete loss of the resinified elastomer,
C-O-O-C groups and the C-O-O-C crosslinks. ending up with the "chalking" effect that I men-
The reactions between sulphur and oxygen are tioned earlier. Because of the lighrabsorbing
those that eventually lead to sulphuric acid for- properties of black-frlled materials, this effect
mation, which is a particular problem with is not normally seen, although a "bronzing" ef-
ebonites. fect has been claimed to result from light-in-
duced ageing of some very smooth-surfaced,
Therefore, there is alarge range ofboth sequen- black-filled products. I must emphasize that
tial (or chain) reactions and competing reac- light itself does not cause degradation of rub-
tions, and the ones that predominate depend on bers; oxygen must be present and the effect is
factors such as the formulation or composition light-catalysed oxidation. The effect varies with
of the rulcanizate as well as the influences of wavelength, UV being the most dangerous and
heat, light and metal catalysis. Heat ageing bal- red/oranse the least so.
ances the rate of reaction of oxygen with the
elastomer and the rate of diffusion of the oxy-
gen into it. If the temperature is relatively low,
ithas been postulated, for an unprotected elas-
tomer, that diffusion through the elastomer pre-
dominates and therefore there is slow oxidation
throughout the product, but at higher tempera-
fures (e.g., >80'C) the rate of oxidation be-
comes much more important than the rate of
diffusion, so substantial oxidation occurs on
the surface and an oxidized (hard) surface skin
forms. As oxidation continues, the chain break-
down may become more significant and the
hard surface can then soften and tum sticky. To
complicate matters further, under certain condi-
tions this order can be reversed and an initially
sticky degraded surface can harden with further
Figure 7 Crazing pattern induced on a thin white
oxidation.
disc ofrubber.

68
The effects of trace levels of pro-oxidant met- cracks perpendicular to the stress in the side
als, such as copper and manganese, have been wall, whilst the white strip in the side wall,
known for years and were used by Goodyear in which must have had an identical life, shows
I 837 to improve his sticky raw rubber by paint- the random crazing ofoxidative degradation.
ing the mouldings with a concentrated solution
of copper nihate. The "improvement" was actu-
ally a hard oxidized skinl These metals acceler-
ate the oxidative degradation, again forming
brittle skins on both raw and l.ulcanized rub-
bers as well as evenfually degrading the elas-
tomer to a sticky mess, as illushated in the
baby feeder teats in Figure 8. There has been
much dispute as to how much copper an un- \\}
filled r,ulcanizate can reasonably tolerate with-
out specific protection. There is no doubt that it
depends upon the "state" ofthe copper and
"state" probably relates to its chemical form
and degree of solubility. Certainly 10 parls
per million copper is sufficient to destroy a
product, as Figure 8 shows. Figure 9 An aircrafi white wall tyre showing crazing
in the white region and ozone/flex cracking in the
black region.

I have not discussed ageing by work, other than


flex cracking, as in most exhibits this will not
be valid. However, once movement is intro-
duced into an elastomeric product, complica-
tions can occur none of which is beneficial
to the life of the-product. The two dangers for
any dynamic exhibit are first that any protective
layer of oxidized rubber, antidegradent or wax
is cracked, in which case oxidation will pro-
ceed very rapidly from the location of the
crack, and second that a cycle ofwork on a
thick piece of rubber can result in heat build-up
within the rubber, which is eventually dissi-
Figure B Babyfeeder teats, one degraded through pated by catastrophic failure ffom within the
high copper contamination. bulk, as shown in Figure 10. Obviously this
should be considered and investigated prior to
Atmospheric ageing is different from shelf initiating any "dynamic" exhibit. Indeed, in any
ageing in that it is characterized by the attack such exhibit it may well be appropriate to re-
of ozone on the rubber. It is often called flex place the moving elastomeric components with
cracking or ozone cracking as detailed examina-
tions of the types of cracking have shown the

il"'ss
same pattem. It is interesting to note that this
type of cracking occurs more readily at night
than in the day as light-catalysed oxidation
tends to obscure the ozone cracking. Following
from this, ozone cracking will occur in
black-fi lled materials, where light-catalysed
oxidation is minimal. A classic picture in Fig- Figure I0 Heat build-up test pieces showing
ure 9 illustrates an aircraft fzre where the top c a tast rop hic i nte rna I fa i lure.
(black) area shows the typical ozone-induced

69
modem replicas. For a much more exhaustive can have half-a-dozen trade names. The ones I
account of the ageing of rubber, please refer to shall use are those familiar to me in the U.K.,
the Malaysian Rubber Producers' Research but in books, such as the BRMA toxicity guide
Association publication Natural Rubber, to rubber chemicals listed in the bibliography,
Science and Technologt, as detailed in the they can be related to their equivalents from
bibliography. other manufacturing sources.

Conservation Options The main difference between the amines and


phenols is that the former stain to a lesser or
After that brief glimpse of what is going on greater extent and the latter do not again
during the life or storage of an elastomeric within limits (see Table III). -
product, consider now what can be done to pro-
long its display life. In truth, options are re-
stricted to adding protective agents, coating the
product with an impermeable membrane or re-
Table III
Amine and Pherol Antidegradents for Rubber
moving the cause of the trouble oxygen and
ozone. Assuming that the light is- adjusted as in- The amines consist of four basic categories:
dicated, the temperature kept low, and all un-
necessary sources ofstress removed, these keto-amrne good antioxrdants ELEZS
actions still will not stop some of the sulphur condensates
chemistry liom continuing. But, as pointed out qurnoil nes good antroxrdants Flectol H
earlier, the sulphur chemistry will not, by itself. poor ozone resrsrance

harm the exhibit for several centuries. substrtuted good antroxrdants DDA
drphenylamrnes some ozone resrstance Octamrne
substttuted good antroxrdants IPPD
There have been well over 1,000 patents issued p-phenylenediamrnes good antrozonants Wrngstay 100
conceming protective agents for rubber vulcani-
zates antioxidants and antiozonants but The phenols can be divided into four categories:
there - - situ-
are relatively few on the market. The mono-ohenols moderate antroxrdants lonol/BHT
ation is complicated by the many trade names least starnrng
used, but for chemically different materials, the polyhydroxyphenols moderateantroxrdants Wrngstay L
relatrvely non-statntng
number is probably no more than 100. Because
of the cost of introducing new chemicals and bls phenols good antioxidants 2246
possrble starnrng
testing against various consumer regulations, it phenols good antroxidants Santowhtte
throbrs
is not likely that any new wonder material will weak antrozonants crystals
be produced. Current efforts are aimed at meet- slrght stainrng
ing market requirements through blends of
chemicals, but remember, "market" require-
ments are not the same as a museum's! If the intention is to add one of these materials
to a vulcanizate then there is little option other
Essentially the antidegradents in use today can than to dissolve the chemical in a suitable sol-
be divided into two types, those derived from vent, immerse the vulcanizate and wait for dif-
aromatic amines and those from hindered phe- fusion to take place. As long ago as 193 I this
nols. There are a few others such as phosphites, was proposed as a method of giving protection
used with phenols in the plastics area as there is to objects made without added antidegradents,
a synergistic effect, and mercaptobenzimida- but it never found commercial favour and,
zole (MBI), which is effective against metal although I have seen no experimental ageing re-
catalysis. Again, this is used in conjunction sults based on this method, I am not surprised.
with a conventional antidegradent. The usual idea is to use a solvent for the antide-
gradent that will swell the rubber surface and al-
For simplicity, trade names will be used in this low diffusion of the antidegradent into the bulk
section. These are easier than many of the of the rubber. This may be acceptable for a new
chemical names that take a line or two of print, wlcanizate, but if the object is not vulcanized,
but have the disadvantage that one chemical or is degraded so that the surface is structurally

70
weak, it will either dissolve or swell to such an be built up by topical application, provided the
extent that the surface will be severely dam- limitations of cracking, repair and surface
aged. The rates of diffusion of antidegradents finish are appreciated.
are quite slow this is essential to their opera-
tion so they-will also take a long time to The phenolic antidegradents are not considered
- into the rubber and. on this time scale.
diffuse antiozonants, only antioxidants. When oxygen
much damage could be done to the network, attacks a rubber molecule various chain reac-
particularly since the solvent will be extracting tions occur, which result in, among other
whatever it can from the lulcanizate whilst the things, polymer chain breakage as well as the
antidegradent is diffi.rsing in. If this method is insertion of further crosslinks. The phenols of-
used for long-term protection, then a non-sol- fer an altemate path in the chain reaction se-
vent for the rubber should be used, methanol quence, which stops the chain breakage fiom
perhaps, as the antidegradent will still diffirse progressing. They do not stop chain reactions
into the rubber, but there will be less extraction from starting, so their effect is, at best, to slow
from the bulk of existing chemicals. There down the oxidative breakdown by perhaps five
will, of course, still be dissolution of any sol- times. They do not form a protective skin if oxi-
uble layer and, ifthere is a stable surface skin dized on the rubber surface and they need to be
already, either of oxidized rubber or insoluble intimately dispersed/dissolved in the rubber to
oxidized antidegradent residues, this may function.
be impervious to the antidegradent being
added! Any physical barrier will stop or reduce oxy-
gen/ozone diffusion into the rubber and might
If the choice is not to diffirse protective chemi- or might not crack depending on its brittleness.
cals into the unprotected rubber, there is still Again, its capability to self-heal will depend on
the option ofadding a protective surface coat- whether it can migrate from the bulk or
ing, but before considering the various treat- whether it is an added surface layer.
ments that are available, one point must be
made. Antidegradents can be considered to There is also the problem of light coloured arti-
operate by physical (barrier) means, chemical cles and UV catalysed oxidation. As with plas-
reactions, or both. It is crucially important to tics, UV stabilizers can be added to the bulk
understand that the different types ofantide- product or in a surface furish. As mentioned,
gradents work in different ways and thus give light by itself does not degrade rubber, it
visually different surface finishes. catalyses oxidative degradation.

The p-phenylenediamines (PPDs) firnction as Finally, there is the "tart-up" factor. Treatment
both antiozonants and antioxidants and at least with a wax, polish or silicone oil can make a
one of their modes of operation is to migrate or rough degraded surface look beautiful just by
bloom to the surface of a product where they re- reducing the light scattering. The coating mate-
act with the oxygen and/or ozone to form a brit- rial will also have some barrier effect and so
tle skin, which increases in thickness until it might prolong the life a little, but there will be
presents an impermeable barrier to the gas. Any no great long-lasting effect.
damage to the skin, such as by cracking, is re-
paired by further blooming. There are many dif- There is one word of warning about any topical
ferent PPDs and one of the main reasons for application; as mentioned earlier, any surface
selecting a particular one is its solubility in the disruption by solvent swelling/drying will po-
polymer system being protected. This affects tentially be disruptive and could do more harm
the rate of blooming, which is crucial to long- than good. Ifthe product is black and protected
term protection. PPDs oxidize to blue/pur- with PPDs (i.e., post early 1930s) or non-black
ple/black materials and readily stain; this and made after the mid 1950s, thus probably
causes no problems with a black-filled material containing phenolic antioxidants, any protec-
that is not in contact with a light coloured one, tive layer could be washed away by a solvent
but obviously would be catastrophic for, say, a and this would not be a good idea. If the inten-
white thread in underwear. A surface skin can tion is to apply any such application, it should

7l
be the mildest water-based emulsion or a and still perfect, its projected life in the same
relatively dry spray of a solvent-based system. environment must run to many hundreds
ofyears.
Another well-established barrier technique is
the coating of the product with an oil lacquer It is obvious that none of these methods meets
that is then lulcanized. The procedure is de- the conservator's ideal of doing nothing irrecov-
scribed in detail by Gottlob, and Figure I I erable to the samples, but in the chemistry area
shows rubber shoes over 70 years old that were at least, there are few options. lndeed, ifthe
given this finish and that are immaculate today; sample is light-coloured and thin, the additional
they were stored underwater, which may be strengthening supplied by a rulcanized coating
relevant. The electron micrograph illustrated in might be a considerable advantage. After all,
Figure 12 clearly shows the layer oflacquer on old Master works are revarnished. whv not
the surface (running vertically slightly left of old elastomers?
centre) and, as this layer is over 70 years old
Since the degradation to be prevented is caused
by oxygen, and to a lesser extent by ozone, the
obvious answer must be to prevent these gases
from reaching the elastomer. In the absence of
a barrier layer the exhibit must be in an en-
closed vessel that can either be evacuated or
filled with an inert gas. Evacuation is not to be
encouraged because ofboth the potential
danger should the case implode and the possi-
bility of chemicals being volatilized from the
Figure I 1 Rubber overshoes recoveredfrom a ship rubber and condensing on the glass. Nitrogen is
sunk some 70 years ago. cheap and easy to acquire, and a low pressure
gauge. showing that the pressure is one or
fwo psi above atrnospheric pressure, will
give an easy check on leaks.

The shoes discussed earlier are by no means the


only case of rubber being recovered from under-
water. The fyre and inner tube of a Wellington
bomber lost during the Second World War in a
Scottish loch were recovered and the tube in-
flated with no problems. Similarly, a seal
from a hatch of the first British submarine 11ol-
land I , lost during trials in l90l , was found to
be in perfect condition under a layer ofsurface
oxidation. The oldest industrial sample I have
ever examined was a sewer seal, installed in
l86l and removed during the replacement of
rusted iron pipework in 1963. The sewer is il-
lustrated in Figure 13, with the seal just visible
under the rubble as a gentle arc in the fore-
ground. In spite of the distinctly hostile envi-
ronment it had been in, and the surface loss
through abrasion, it was still a good seal and
would have lasted many more years if only
Figure I2 Scanning electron micrograph ofcross- the pipework had lasted. Storage under cold
section through the rubber layer ofthe overshoe water might be a viable preservation option
showing the lacquer as a smooth /ilm about although I have never seen it suggested as a
I 5 pm thick running venically just left of centre. deliberate ploy.

72
The one essential point
is that whatever mode
of preservation is attempt-
ed, the object should be
mounted as it is intended
to be displayed, as soon as
possible after the treafinent
or even before, and then not
touched, flexed or otherwise
distorted. Obviously any
mounting should be as
stress-free as possible.

Conclusion
In summary, therefore,
the only real danger to elas-
tomeric artefacts displayed
in museums is oxygen, al-
though light, heat and ozone
might have a part to play in
Figure I 3 An I86l sewer withJunctional rubber seal arcing across the
some environments.
foreground.
There is no wonder chemi-
Acknowledgement
cal that will prevent oxidation nor is there
- recom-
even one that could be universally The author thanks the Board of the Malaysian
mended for all elastomers is there ever Rubber Producers' Research Association for
likely to be. -nor
permission to attend Svmposium '91 andto
publish this paper.
The best that phenolic antioxidants can do is
slow down oxidative damage by a factor of Figure 7 first appeared in Revue Gtntrale du
about five. What this means depends on the in- Caoutchouc, 3 l, 1954:. 419 and is reproduced
dividual article and its initial life expectancy. with permission of that joumal.
Black products ofa reasonable thickness Figure 9 first appeared tn Transactions Institu-
(i.e., >5 mm) should cause no problems over
tion ofthe Rubber Industry,2l,1945:49 and is
many decades and probably centuries, although reproduced with permission of the Plastics and
the surface may appear to deteriorate margin- Rubber Institute.
ally. There might be a benefit from coating
these products with a mix of wax and an
R6sum6
amine antioxidant.
Le caoutchouc : son histoire, sa composition
Light-coloured or non-vulcanized articles are at et ses perspectives de conservution
great risk and should be barrier protected. This
must be the best option for long-term protec-
Quiconque vit dans ce monde dit < moderne >
tion, and the best barrier is a sealed case and an connaitfort bien le caoutchouc et ses propri'ltds.
inert atmosphere. If the storage temperature can Et c'est sans doute justement parce que cette ma-
be reduced to2"C to 5oC. this could be an tidre nous est sifamilidre que nous I'abordons,
added advantage. sinon avec un certain mtpris, du moins avec une
certaine indifflrence, malgrt la place qu'elle

t)
Appendix
Natural Rubber: History of lts Industrial Development to the 20th Century

lst Millenium Mexico Ball courts/fi gurines holding balls.


B.C.

Aztecs 6th Cent. South Balls, dipped feet -> shoes, coated fabrics.
Mayas & possibly America Pictures copied in National Museum, Mexico.
earlier.

l0th Cent. SW US Ball courts and ball games.


->
Objects in Peabody Museum, Harvard
University.

Columbus 1493 Haiti First European recorded to have seen


rubber balls and retumed some to Europe.

Torquemada t6l5 Mexico Taught Indians how to waterproof cloth


and make dipped goods.

Charles de 1735 to 1740 South In the Andes, described how Indians "milked"
la Condamine America trees for liquid to waterproof fabrics. The
lndians called the tree "HEVA" and the sum
from the liquid "CAHUTSCHU." de la
Condamine christened the "milk" "LATEX."

Fresneau 1743 to 1746 French Realized the potential of the material and
Guiana infected France with enthusiasm for rubber
research. The problem was that latex could
not be shipped to Europe without 'going bad'
and solidiffing. Tree was Hevea braziliensis.

HerrisanV 1761 France Found dry rubber would dissolve in turpentine


Macquer but the resulting dried film was sticky and
soft.

Macquer 1768 France Replaced turps with ether and cast strong
films that were not sticky.

1769 France Made riding boots for Frederick the Great by


multiple dipping process.

Priestly t770 UK Noted that an artist's shop in London sold a


(ofoxygen half inch cube of material for erasing pencil
fame) marks for three shillings. He called it 'INDIA
RUBBER" having found from whence it came.

Roberts/ t790 UK First patent referring to rubber rubber


Dight - oil painting.
solution for treating canvas before

Hummel t 8l3 US Gum elastic vamish first US patent that


mentions rubber. -

75
Thomas 1819 UK First saw rubber.
Hancock

1820 UK First patent for dry rubber; cut strips for


elasticating clothes, braces, etc.

1820 UK Invented his "pickling" machine, which enabled


dry rubber to be worked into a "dough."

1820 US Dipped shoes appeared in the US, made in


South America, exported to Paris gilded,
-
"fashioned" and exported back to America.

by 1823 US Direct imports from Brazil.

Charles 1820 US Born.


Goodyear

Macintosh 1823 UK Realized that if fabric coated with rubber


solution then had another layer of fabric
applied to rubber, the three-layer sandwich was
waterproof and not sticky
-"MACINTOSH."
1825 UK Pitch/rubber solution
->
sheets for coating
ship bottoms, etc.

Faraday 1826 UK Established empirical formula as CsHs.

1825 to 1830 UK Dozens ofuses.

by 1830 US Over 500,000 pairs of rubber overshoes


had been imported.

Chaffee l83l US Rubber/turps/lampblack paint to waterproof


leather.

1832 US Roxbury India Rubber Co. founded (first US


rubber company).

1834 to 1836 US Invented the two-roll, trvo-speed mill


and then enabled it to be heated. Still the
standard mill of today.

Hancock/ 1834 UK Formed Chas. Macintosh and Co.


Macintosh

Chaffee 1835 US Invented the three-roll mill for


"calendering," which is still the basic
procedure today.

Goodyear I 834 to I 83 5 US Became intrigued by rubber


-
some say
obsessed.

76
Hancock 1837 UK Invented the spreader, the standard
coating machine of today.

by 1837 US Economic crisis. Rubber "bubble" burst.

I 839 US Existing rubber industry in the US


finished but 500,000 non-vulcanized shoes
per annum still coming from Brazil.

But in the same vear:

Goodyear I 839 US Left a mix of rubber, sulphur and white lead


on a hot stove and the resulting material was
"CURED" of all its defects. No longer
softened on heatinglhardened on cooling
and lost its stickiness.

l 840 US First commercial vulcanized material rubber


thread for "shirred" cloth. -
Goodyear I 843 US First application for US vulcanization patent.

Hancock 1842 to 1843 UK Identified sulphur in a piece of Goodyear's


cured rubber. Could not duplicate "CURE" as
did not know about white lead, but effectuated
cure with rubber/molten sulphur.

I 843 UK Produced "Hard rubber" (Ebonite) with long


treatment of rubber with molten sulphur.

l 843 UK In November Hancock obtained UK prov.


patent.

Goodyear 1844 UK In February UK patent application refused.

Parkes 1846 UK "Cold cure" process discovered


(sulphur chloride).

Thompson 1 845 UK Patented the pneumatic tyre but no vehicles


suitable to make it a commercial successl

I 851 US Vulcanized rubber shoes being manufactured at


a rate of over 5 million pairs per annurn.

But in the UK there was more interest in vulcanized "Macintosh', material:

Hancock 1857 UK Published his "Personal Narrative. "

Williams 1860 UK Decomposed natural rubber and isolated


isoprene (CsHs).

77
Thompson 1 861 UK First commercial solid vulcanized tyres.

I 861 ? "Hard rubber" frst called Ebonite.

Murphy I 870 US Recognized "oxidation" as cause of


deterioration in rubber.

Collins 1872 UK Commissioned to report on Rubber in Brazil.

Wickham 1813 Brazll Commissioned by Kew to collect seeds of


Hevea braziliensis.

1 875 Brazll Dispatched 70,000 seeds to Kew:2,397


germinated.

1876 Singapore 50 seedlings arrived. Died due to neglect.

Cross? 1877 Singapore 22 seedlings arrived and survived. Basis of


virtually all Asian trees today.

Bouchardat 1879 France Re-polymerized isoprene to "rubber. "

Daimleri l 885 Germany Invented the motor car.


Benz

Ridley 1 888 Singapore Took over Botanic Gardens. Began a one-


man crusade to develop plantations.

Dunlop 1888 UK "Reinvented" the first pneumatic tyre but now


bicycles and vehicles available to use it.

Bartlett/ I 890 UK Developed tyre rim designs essentially


Welch similar to those in use today.

I 899 First plantation rubber shipped from Sri Lanka.

1900 World production of NR approaches 50,000


tons.

Kronstein 1902 Germany Polymerized styrene.

Oenslager 1906 US First chemical to "accelerate" vulcanization


(Aniline) then diphenylthiourea (DPU).

Lebedev l9l0 USSR Polymerized 1,3-butadiene to give a


rubbery material.

l9l0 Wild rubber peaks at about 90,000 tons per


annum. Plantation rubber (Malaysia/Sri Lanka)
reaches 10,000 tons.

78
Ostromislenski 1915 USSR First organic wlcanization systems without
sulphur: nitrobenzene and peroxides.

About this time the importance of zinc oxide


in accelerated cures was appreciated.

Kratz 1920 Diphenylguanidine (DPG).

1920 World production of NR 350,000 tons.

Bruni 1921 Italy Mercaptobenzothiazole (MBT)


Bedford 1921 US independently.

Lorentz 1922 US Thiurams.

Caldwell 1922 US Xanthates.

Bruni 1923 Italy Zinc dithiocarbamates.

Russell 1923 US Patented organic fatty acidlzinc oxide


use in vulcanization.

1924 Germany Fossilized rubber 60 million years old found.

1925 Germany Serious work began on the synthesis of


Synthetic polybutadiene (Buna).

1925 First commercial antioxidants (amines)


introduced. They were staining.

Rosenbaum 1926 US "S;mthetic rubber is dead."

Patrick 1930 US "Thiokol" first commercial synthetic rubber.


- is not a sulphur cure just metal
Note that this
oxides and possibly quinones. -
Semon 1930 US Suggested diffusion after vulcanization of
protective agents into rubbers.

1930 World production of NR 850,000 tons.

Tschunker 1930 Germany Buna N and Buna NN discovered (nitrile


rubbers of 25o/o ACN and 35% ACN).

1930s Introduction of amine derivative antioxidants


(staining).

du Pont 193 I US Duprene. Became polychloroprene (Neoprene).

IG Farben 1932 Germany Sulphenamides adducts of MBT with amines.


-
1932 USSR Russia manufactured SKA followed by SKB
(polybutadiene rubbers). -

79
Tschunker 1933 to 1934 Germany Buna S patented and produced (Styrene
butadiene copolymer).

l 93s USSR Sovprene; equivalent to Neoprene.

Thomas t937 US Butyl rubber.

Union Carbide/ t939 US First use of plasticized PVC as cable sheath.


Goodrich

1939 to 1941 US Nitrile rubber produced.

1940 World production of NR 1.5 million tons.


World production of Synthetic Rubber (SR)
100,000 tons.

1940 USSR SKI polymerized isoprene (synthetic "NR").


-
t943 US GR-S production started (now known as SBR)
Government rubber-styrene styrene
butadiene rubber (similar to ->
Buna S).

Pintin 1943 Germany First patent relating to wethane rubbers.

Dow Corning/ t945 US Silicone rubber. Again not normally a


General Electric sulphur cure, peroxides used.

1950s Inhoduction of Phenolic antioxidants


(non staining).

Last wild rubber exported from Brazil.

Goodrich 1954 US Announced high cis synthetic "NR."

1954 US Introduction of sub stituted


para-phenylenediamine antiozonants.

Firestone I 955 US Low cis synthetic "NR."

du Pont l 958 US Fluoroelastomers.

1960 World production crossover point.


Both NR & SR about 2 million tons.

MRPRA 1970 UK Urethane crosslinking of NR.

1990 World production of NR 5 million tons.


World production of SR l0 million tons.

1990 Perhaps 5 billion Hevea braziliensis


trees worldwide.

80
Ardil: The Disappearing Fibre?

Mary M. Brooks

York Castle Museum


York, Yorkshire
United Kingdom

Abstract new regenerated and synthetic fibres in the


early 20th century. The acetates using cellulose
Ardil is a regenerated proteinjibre produced pulps were well established by the late 1920s.
commercially for clothing and domestic.furnish- Truly synthetic fibres came a little later. Du
ings by Imperial Chemical Industries between Pont developed nylon in the U.S. and in 1939
l95l and 1957. This paper examines the develop- granted ICI a production licence. There was a
ment, characteristics, manufacturing methods real need for suitable fibres and textiles to sat-
and marketing of Ardil. Fev'examples of Ardil
isS both a growing world market and specific
are known, so this paper represents current work
military requirements. *
toward identifying textiles containing Ardil and
makes recommendations for their consewation.
Ardil reprcsents a particular challenge to both However, the cellulose-based fibres did not
the conservator and the curator and highlights yield fabrics that felt warm to wear and there
the importance of researching, identi,fying and was much competition to create an artificial
re co rd i n g manufac tured .fi bre s. wool-like fibre using a protein source. In order
to form a fibre, a polymer must have large
Introduction molecules that are capable of crystallizing,
achieving a reasonable degree oforientation
Ardil is a relatively little-known regenerated and having a high degree ofpolarity to give
protein fibre, which had a short but intriguing intermolecular cohesion. Possible alternative
life as a "wonder fabric," manufactured by Im- fibre sources included casein, a milk-rerived
perial Chemical Industries (ICI) after the Sec- product, waste leather and fish protein. Casein-
ond World War. So far, few named examples based fibres were developed under trade names,
have been identified in British collections. This such as Lanitol (Italy), Fibrolane (Britain) and
may indicate the nature of the long-term stabil-
ity of the fibre or problems in distinguishing
* ln 1942, the Ministry of Economic Warfare in
Ardil flom other fibres and blends unless it is
trade-marked. London was urgently questioning British legations in
Beme and Stockholm regarding the Axis powers tex-
tile supplies: "What is the output by Solanum GmbH
Development of Ardil ofcellulose pulp liom potato tops and other new
forms of cellulose? ... What is the extent of fibre
The development and eventual failure of Ardil production by Germany and Italy respectively from
as a technologically and economically viable fi- Broom plant (Ginstra)?... Your answers to all the
bre involved a complex mix of interacting fac- above may have an important bearing in enemy
tors. There was great interest in developing supply position." '

8l
Aralac (U.S.) and companies in both Europe It is worth noting here that Courtaulds were
and America were experimenting with vegeta- also working on regenerated protein fibres.
ble protein sources. Interest and development They developed Fibrolane A, BC and BX,
was such that in 1946 the American Society for made from case_in, and Fibrolane C, made from
/
Testing Materials proposed that all regenerated peanut protein. The relationship between the
protein-based fibres derived from sources, such two industrial giants was a complex mix of
as casein, peanuts, soya beans or other vegeta- competition and integration. Courtaulds had
ble proteins, should be described using the ge- vast experience in spinning and weaving silk
neric term "azlon." and regenerated viscose mourning crepes, but
relied on chemicals produced by ICI. Con-
It was against this background that W.S. Ast- versely, ICI had little knowledge of fabric
bury of Leeds University and A.C. Chibnall manufacture or consumer-led marketing, but
and K. Bailey of Imperial College, London, dominated the chemicals field. This relation-
approached ICI with the results of their experi- ship was formalized in a 1928 agreement by
ments in producing fibres from vegetable which Courtaulds agreed to abstain from chemi-
protein. Their initia"l research used protein ex- cal manufacture, taking all their supplies from
tracted from hemp,'but by 1936 they had ICI, whilst ICI promised to refrain from artifi-
switched to peanuts, the nut of a sub-tropical cial "silk" fibre production. The respective tex-
plant Arachis hy'pogae 2., as the protein source. tile technology skills of these two companies
Peanuts contain the proteins arachin and con- was to be crucial in decidins the fortunes
arachin. Hydrolysis of the amino acids in these of Ardil.
proteins has shown them to be very similar to
those in wool. Peanuts or groundnuts were Under wartime production pressure, ICI and
grown widely throughout the British Empire Courtaulds joined together in 1940 to form a
as well as in America, China and Borneo. How- company named British Nylon Spinners specifi-
ever. ICI was careful to stress that Ardil had no cally set up to produce nylon for parachutes.
connection with the great groundnut scheme in Despite this, Courtaulds persisted in regarding
Africa, which collapsed in scandal in the early nylon as a competitor to its traditional interests
1950s. Nevertheless, peanuts were an important in artificial "silks," the viscose rayons based on
crop, with a yield of around eight million tons cellulose. In fact, they campaigned to have ny-
per annum during the 1930s. lon classified as a new type ofrayon. In keep-
ing with this conservative attitude, Courtaulds
Astbury, Bailey and Chibnall refined their re- refused to join ICI in a prpposed post-war joint
search and proposed a process for producing venture to develop Ardi1.6 In 1943, the earlier
a fibre from de^natured vegetable protein dis- agreement between ICI and Courtaulds lapsed.
solved in urea.3 Their patent described it as ICI felt free to move into the field of fibre pro-
"the production of artificial filaments, threads, duction, but, significantly, they lacked Cour-
films, and the like from solutions of denafured taulds's expertise in fibre technology and
or degenerate or coagulated protein material by consumer marketing.
extruding such solutions into a diluent or other
regenerating medium."' They approached ICI ICI felt sufficiently confrdent to announce the
and reached an agreement whereby the com- creation of Ardil just before the war. An Ardil
pany took over development.' The project was production plan was launched in 1938, but
based at the ICI research plant in Ardeer, Scot- suspended due to the outbreak ofhostilities.
land hence the name Ardil was registered as Sufficient fibre was produced to make a few
- mark. Of the ICI divisions, the Scot-
the trade Ardil/wool blend suits: some of these were
tish plant had the most experience of fibre spin- apparently still being worn in ,1951 when fuIl-
ning and weaving. ICI researchers developed siale production coimenced.g Post-war, ICI
practical processes for production that moved urgently needed to diversify. In particular, they
away from the original theogetical propositions needed to find an altemative production activ-
and filed their own patents.o ity for their Scottish Explosives Division. Ardil

82
was selected as an ideal new product. The pilot sales never rose above 2.6 million pounds. In
plant was established in 1946 although produc- 1955, producing Ardil cost 47 old pence a
tion was initially limited to half a ton a week. pound (approx. 20 new pence, or 40p Can.) as
In July 1941, ICI proposed a capital expendi- opposed to 24 old pence a pound (approx. 10
ture of 2.1 million pounds for Ardil with an ex- new pence, or 20( Can.) for viscose staple fibre
pected return of 70% when the main plant was despite the fact that the raw material was a com-
operating at half capacity.'" Building of the parable price. " A drop in wool prices further
manufacturing plant at Dumfries, Scotland, reduced the competitive advantage of Ardil as a
commenced in 1948 with an initial capacity of bulking fibre.
22 million pounds per year although production
was well below this amount. Caught in this vicious circle, Ardil never
gained a firm foothold in the market. However,
Simultaneously with this development, synthetic fibres overall had made a significant
Unilever, a company with various interests impact. ICI set up Group F to develop their
including margarines, was building a plant at manufactured fibres including Terylene and
Bromborough, Merseyside, to handle the raw Ardil in 1956 and were still expressing cautious
peanut meal. Unfortunately, disputes with optimism regarding the future of Ardil: "Sales
Unilever over processing charges caused de- of Ardil were disappointing ... but work is in
lays. The major problem however was an unex- hand to evaluate an Ardil fibre with imnroved
12
pected shortfall in world peanut production properties." This improvement never material-
associated with the failure of the East African ized, and in 1957 ICI took the decision to cut
groundnut scheme. Ardil manufacture com- their losses, which had been running at 3.7 mll-
menced in 195 1 again, most unfortunately, lion pounds and Ardil production ceased.
coinciding with a- depression in the textile It had an effective commercial life of only
trade. Of the 316 firms who had indicated inter- six years.
est in 1947, only 76 placed small trial orders in
195 l. The auguries were not good. Manufacturing Methods: Successes
and Problems
Technical difficulties, which will be discussed
more fully later, meant that Ardil was best used The basic principle of manufacturing was to
as a blend with wool, cotton or other manufac- extract the proteins, arachin and conarachin,
tured fibres. This resulted in marketing prob- from the peanut and treat them so that they
lems as such blends could not be marketed as formed a solution from which a fibre could
100% pure single fibre products and therefore be spun.
required special promotion to consumers.
Despite the support of the Bradford Dyer's As- The peanuts, whose fruits grow underground,
sociation and other wool and cotton manufactur- were hand pulled, shelled and the reddish skin
ers who were prepared to invest in Ardil, some removed to prevent discolouration. For maxi-
Divisions of ICI were always skeptical. The mum protein yield, the nuts needed to be ma-
Dyesfuffs Division, which had the most experi- ture but ungerminated. The decorticated nut
ence of the textile market, was never prepared consists of 43o/o to 48o/o oil,24Yo to 26Yo pro-
to declare confidence in the fibre. In the end, in- tein and 26oh to 29o/o carboltydrate. After
temal dissension and technical problems meant blanching, the nuts were ground to a meal and
that economic realities conquered. solvents used to extract the oil, which was
a useful by-product used for salad oil or
If fudil was to be attractive as abulking agent marganne.
for the more expensive natural frbres, it had to
be cheap. Such cheapness could only be Care was needed to select suitable solvents and
achieved when production levels were high to control temperature during this process. The
enough to realize economies of scale and while oil-free meal contained about 40%o to 49o/o pro-
costs were high it was hard to build up sales tein. It was dissolved in dilute alkali and acidi-
volume. Over the five years of production, fied with sulphur di-oxide to pH 4.5 to extract

83
the protein. This process was carefully control- light colours without bleaching. It was nearly
led to keep the colour as light as possible.* neutral with not more than 0.3Yo acetic acid and
a slightly higher moisture regain rating. Ardil F
The solid residue, mainly carbohydrate, was was a pale fawn colour and was used when dy-
used as a base for cattle fodder. The precipi- ing darker colours. It contained 4% sulphuric
tated protein was then washed and dried, giving acid and some formaldehyde. Ardil K was
a creamy powder known as "Ardein," which made in heavier deniers, but was otherwise
was almost pure arachin and conarachin. This similar to Ardil F. The range of deniers avail-
was then dissolved in dilute caustic soda giving able corresponded with those of wool: rang-
a viscose solution that was allowed to mature ing from fine grades to thicker deniers for
for 24 hours. This denaturation process allowed carpeting.
the folded long-chain molecules present in the
soluble protein to open out into an extended Dyeing Ardil
form so giving better mechanical properties.
Ardil dyed well both on its own and in blends.
Once this process was completed, the solution Mass-dyed Ardil was produced by introducing
was pumped through spinnerets into a coagulat- pigment into the fibre during manufacture, giv-
ing bath of 2% sulphuric acid and' l5o/o sodium ing highly light and water fast colours. This
sulphate to form the fibre. Urea or caustic soda was a cheaper dyeing process and 12 standard
solutions had been used in the earlier develop- colours were available, chiefly for use in carpet
ment stages. The tow fibre was then hardened yarns. Most wool dyes, such as acid and
by a formaldehyde treatment under acid condi- chrome dyes, as well as some direct and vat
tions to improve insolubility. The chemistry of dyes, could be used successfully. Dyes sensi-
this process of forming molecular crosslinks- . tive to formaldehyde could not be used. The dif-
was complex and not iltogether understood.14 ferential reactions of fibres to dyes meant that
Altematively, improved wet strength could be cross-dyeing of blends of Ardil with other fi-
achieved by acetylation or treatment with 0.3% bres produced interesting ingrained speckled
glyoxal polymer. A natural crimp developed if effects. Ardil could also be printed using
the fibre was wetted, stretched and released. traditional methods.
The fibre could be extruded in any diameter re-
quired. Although it could be produced as a con- Fibre Properties
tinuous filament, it was often cut into staple
lenglhs of between 1.3 cm to 20 cm depending Physical Properties: Ardil had many of the
on the application. It could then be spun using a characteristics associated with the natural pro-
woollen, worsted or cotton system. tein fibres warmth, softness, resilience and
the ability -
to absorb moisture and generate heat
Three types of Ardil were produced. Ardil B when wet. It had a remarkable wool-like tex-
was pale cream and so could be easily dyed in ture despite being less strong than wool. Al-
though not thermoplastic, its good crease
recovery and draping qualities were valuable in
*Moncrieff details one example of this process as
improving the performance of cellulose fabrics.
follows: "200 parts by weight of fat extracted peanut
Blended with synthetics, it improved wearing
meal are stirred with 3500 parts water at 20"C for
qualities through its ability to absorb moisture.
10 minutes. Then I 50 parts lo/o to 2To caustic soda
are added over a period of 20 minutes bringing the
liquid to a pH of 8.0 to 8.5. Stining is continued for Key Characteristics
one hour and the resulting solution is clarified. After
clarification the solution is pH 8.3 and is claimed to Colour: Probably as a result of some peanut
have practically no colour. Next sulphur di-oxide gas skins remaining despite skinning and blanch-
is passed through the solution until the pH value is ing, Ardil had a creamy yellow colour. Colour
4 to 5. This acidification results in a copious white varied with Ardil F and Ardil K being darker
precipitate, which allowed to settle, is centrifuged, than Ardil B. At one point, attempts were made
washed a1d dried. The yield is 84 parts air-dry to develop a peanut with a grey skin to reduce
protein."'' such colouration, but this was a long-term

84
project and such peanuts were not used for those of scoured wool. Comparative informa-
production. tion is given in Table I.

Elongation and Abrasion Resistance: Ardil


had good elongation at 50Yo with good elastic Table I
recovery so it was hard to break the fibre. How- Percentage Moisture Regain of Ardil in
ever, it had poor abrasion resistance unless Comparison to Other Fibres at 6594 RH
blended with wool. A 50/50 ArdiVwool blend and 20"C
had improved resistance possibly due to lubrica-
tion of the wool by the Ardil fibre. Fibre Moisture Regain (o/d

Wool 15.0
Behaviour with High Temperatures: Like Ardrl B, neutralrzed/bleached 14 0 to 15.0
wool, Ardil had low flammability and resisted Casernfrbre 13 0
mass-dyed,
Ardrl F, Ardil K, 12 0 to 13.0
high temperatures without softening or melting. unneutralized
It began to char at250'C. Excessive dryness Vrscose rayon 1 1.0
if the fibre was re- Vrcara protern fibre 10.0
caused embrittlement, but
Cotton LC
tumed to standard humidity and temperature, it Cellulose acelate 60
regained flexibility. Contemporary technical in- Nylon 4.5
Terylene polyester o.4
formation puts great stress on the low fire risk
of Ardil, presumably reflecting problems en-
countered when working other highly flamma- Source: lCl Technical Bulletin GL. 1
ble regenerated fibres. Extra fre cover was not "The Properties of Ardil Protein Fibre".
required by insurance companies for factories
processing Ardil. Felting and Shrinkage Behaviour Patterns:
Due to the lack of any surface scales, Ardil did
Electric Properties: Ardil had a low dielectric not felt, but felting could be induced. When
rating and gave few static electric problems blended with high grade wool, Ardil could ac-
when processed at normal humidities. celerate felting shrinkage, possibly as a result
of lubricating the wool fibres.
Tensile Strength: This remained the great
weakness of the fibre. The tensile strength of Behaviour under Ultraviolet Light: Ardil had
Ardil was ^calculated at about 8 kg/mm' to high UV resistance. An ArdiVwool blended
l0 kg/mml, which is lowler than that of wool at yam exposed to UV light for 150 hours in a
l2kg/mm' to2}kglmm". When wetted the fi- carbon arc fadeometer showed no significant
bre extends about l5% in length and about 5% strength reduction.l6
in diameter, but on drying, retums to its origi-
nal dimensions. Tensile strength was greatly Tolerance of Pests/Fungi: Unlike wool, Ardil
reduced in the wet state, dropping ffom I g/den- was mothproof (a point that was made much of
ier when dry to 0.3 g/denier when wet. Pro- in advertisements) and was more resistant to
posals for chemical treatrnents to increase mildew.
crosslinking and overcome this significant
strength loss included stretching in a bath of Comparison with Wool: Technical details,
mercuric acid and acetic acid, which had obvi- comparing the qualities of Ardil with wool, are
ous safety problems.l4 Traill and Simpson given in Table IL
showed that an increase of 50% in wet strength
and25%o in dry strength was possible if the fi- Chemical Properties
bre was treated with basic chromium sulphate
and formaldehyde after dyeing with acid or Ardil fibre is essentially made up of two pro-
chrome dves. '' teins, arachin and conarachin. Chemical reac-
tion with formaldehyde causes these to become
Moisture Absorption: The moisture content insoluble in, and resistant to, the aqueous liq-
was l4o/o. The moisture absorption and regain uors used in textile processing. Ardil F and
rates at standard conditions were similar to Ardil K and the mass-dyed fibres are acidic in

85
Behaviour with Solvents: Ardil is insoluble in
Table II the standard organic solvents. It may therefore
Technical Properti es of Ardi I be solvent cleaned, solvent selection depending
on the behaviour ofany blended fibre.
Properties Ardil Wool

Morsture regarn at Ardil and Other Fibres


65% RH & 20'C
Types F & K 12% 16%
Type B 14"k 16./" As aresult of the poor tensile and wet strength,
Ardil was almost always used in blends either
Tensrle strength (kg/mm2) 8to 10 '12lo 20
as an economic bulking agent with such natural
Tenacrty (g/denrer) fibres as wool, or to improve the characteristics
condfioned 0.7to09 ofcellulose and regenerated cellulose fibres.
wet 0.4 to 0.6
ICI recommended that it should not be used
Elongaton at break (%) alone. The blended fabrics were designed to be
dry 40 to 60 30 suitable for a wide range of end uses from cloth-
weI 50 to 70
ing to domestic fumishing fabrics. A ratio of
Heat ot wettrng (Cal./9.) 26 6 269 50% ludil/sUolo merino wool was used for
sweaters while 40Yo Ardlll60% viscose was
Torsonal ngrdrty (dynes/cm2) 1.2 x 1o1o 1.3 x 1o1o
used in carpeting. Ardil, viscose and nylon
Specrfrc gravrty 13r 'l 33 blends were used for hard-wearing, lightweight
Young's modulus
and cheaper suiting cloths. The Whitworth Art
(g/denrer % measured al 014 0.20 Gallery, Manchester holds a group of fabric
100 extenson per mrnute) samples from ICI in a wide range of blends and
fabric types. These include mass-dyed grey
Source: Anonymous untitled typescript, presumably Ardil/viscose/woolsuiting; Ardil/wool tweed;
lOl document, Whitworth Art Gallery, Manchester. Ardil/viscose/nylon shirting; printed Ardil,/Peru-
vian cotton dress fabric; ArdiVwool jersey
knit; Ardiliviscose velour coating and
reaction and contain about49/o sulphuric acid. Ardil/wooVviscose blanketing. Technical de-
Ardil B differs in being nearly neutral with ap- tails of the yams and weaves are given on the
proximately 0.3% acetic acid. swatch cards.

Acid and Alkali Resistance: Ardil was highly An Ardil blended yam may be used either as
resistant to acids and acid tendering. This en- warp or weft. Amongst the Whitworth samples,
abled it to withstand many of the standard shirting fabric Pattern 31 uses 30% Ardil B,
processes for textile production, which often in-
600Z viscose and, lUYo nylon in both warp and
volved acids. However, it had low resistance to weft. Other examples combine an Ardil blend
alkalis. The fibre would swell making it sensi- with a pure fibre as the other yam. Pattem 341,
tive to mechanical damage. Unlike wool, how- a lighfweight printed fabric, combines a warp
ever, Ardil does not require sulphur crosslinks of 100% cotton with a weft composed of 33%
for its structural stabilizing and therefore can Ardil and 670/oPeru Tanguis cotton. A few
resist chemical damage by alkalis. Normal vat swatches have differing blended yams in the
dyeing could be carried out on Ardil/cellulose warp and weft. The velour coating sample, pat-
blends. This was an important factor as many tem No. l8l, has a warp of 25% Ardil B with
of these fabrics were destined for use in 757o viscose and a weft of 50Yo Ardil B with
women's and children's clothins and were 500/o viscose.
often dyed.
Due to the differential strengths of the various
Behaviour with Bleaches: Like wool, sodium fibres, there were technical problems in prepar-
hypochlorite and sodium chlorite could cause ing the fibres. It was important that a uniform
degradation in Ardil and hydrogen peroxide blending was achieved to avoid problems and
bleaching was preferred. special techniques were required. In general,

86
Ardil and the complementary fibre(s) were Fabric Care
blended together before carding. ICI published
a range of technical information for the trade ICI recommended that Ardil could be washed
(see Appendix l). Conventional processing ma- using normal domestic or commercial methods,
chinery could be used avoiding the need to in- the governing factor being the needs of the com-
vest in an expensive new plant. plementary fibre. They rather cautiously stated
that, although ArdiVcotton blends could be
When blending with wool, Ardil needed washed as normal cotton. it was preferable th,qt
particular care to prevent overstretching and such blends should be "washed as for wool." ''
breaking. Relaxation shrinkage was possible in This presumably relates to possible increased
Ardil/wool blends when wet. Altering the per- loss of wet strength in a hotter washing proc-
centage of the blend allowed for the creation ess. Ironing temperatures were again govemed
of fabrics with different qualities. Alone, Ardil by the requirements of the other fibres used
would not permanent-pleat. However, when in the blend.
blended with a sufficiently high proportion of
a natural fibre, the resulting fabric could be in- As Ardil was stable in most organic solvents it
duced to carry a pleat. A 50/50 ArdiUwool could safely be commercially dry-cleaned.
blend gave a fabric that took a sharp crease
but hung out well whlle 25o/o to 35W65o/o to Marketing Ardil
75oh Ardlllwool was softer and did not carry
a crease well. Such blends were ideal in knits. ICI launched Ardil with extensive marketing.
Ardil was also used in hat felts. The British Hat The trade magazine International Textiles, The
and Allied Feltrnakers Research Association British Export Journal ofTextiles and Fash-
recommended blends of Ardil and rabbit fur or ions gave it wide coverage, reflecting the gen-
Ardil and wool. eral sense of amazement at this new fibre:
"Starting metaphorically with a bag of monkey
Lancashire cotton weavers were particularly nuts and an idea, British chemists of ICI have
interested in exploring blends to improve evolved a new synthetic fibre which, they
cotton's handle, crease resistance and drape. claim. is wool-like. does not shrink and is not
When working with cotton, it was recom- attacked by moths.;'20
mendet' that Ardil, the weaker fibre. should be
oiled.'* Methods of dealing with ArdiVcottg4 ln 1944, post-war shortages and technical prob-
mixes were extensively discussed by Dyke.'' lems were still holding up production and ICI
Ardil/cellulosic warps were sized either with was only able to promise that samples would
sago or Cellofas B. ln cotton blends, fudil fi- shortly be ready for the trade. However, ICI
bres showed a tendency to migrate to the sur- held a major exhibition in London in 1946.
face during washing. This could be controlled As shown in Figure l, the showcase factory
by using a tightly spun yanl gnd problems were in Dumfries was featured in an advertisement
reduced with plain weaves.'o These fabrics in the 195 I Scottish Festival of Britain
were used for nightwear, shirts, lightweight publication.2l
suits and dress fabrics.
Manufacturers were courted. ICI staff lectured
Tertiary blends exploited the wool-like handle and published articles on the new fibre. For ex-
of Ardil whilst adding strength from the other ample, n 1952 F.M. Dyke from ICI addressed
fibres. Lightweight blends of ArdiVviscose and the Oldham branch of the Textile Institute and
wool were used for velours, suitings and dress his lectures were subsequently published." ICI
fabrics. Nylon was used to improve strength went to considerable lengths to provide techni-
and abrasion resistance. Blankets were pro- cal information to textile spinners and weavers.
duced using a cotton or cottor/viscose warp (Titles are listed in Appendix 1.) ICI proposed
with an Ardil/wooVviscose weft while carpet- a wide range of end uses for Ardil, shessing the
ing could be produced using blends of Ardil, economic advantages ofcheapness and price
wool and viscose. stability in comparison with pure natural fibres.

87
contained a useful section on thgpew synthetic

rt| l}t[ PI|0TI]ll l'llfllli fibres that aimed to do just this." The new fi-
bres are described as "Among the wonders of
the post-war world." Ardil is included and
praised as being wool-like whilst being cheaper
and for improving the behaviour of cotton and
viscose. The need for blending due to lack of
strength is clearly stressed. Care advice in-
cludes a warning against shrinkage when wash-
ing wooVArdil blends.

Examples of Ardil in U.K. Museum


Collections
This new ICI factory at Durnfries is now So far, few examples of Ardil have been
producing 'Ardil' protein fibre, and full producrion located in U.K. museum collections.
- at the rate of 22 million pounds per year--will
ruchicrcd b_r the cnd of 1951.
The York Castle Museum Wallis archive con-
sists of a donation of costume, spanning four
}It,CRI,{I" CHEMICAI, INDUSTRIOS LI\,TITID generations of one family. It includes a night-
\OBEL DIVISION dress belonging to Amy Wallis dating from
the 1950s. Old-fashioned in style and similar
to Viyella, a wooVcotton blend in handle and
appearance, it was made by Potters (Museum
number 431 .78). As shown in Figure 2,thela-
Figure I ICI advertisement for Ardil featuring the bel indicates that the nightdress contains Ardil
showcase factory in Dumfries, Scotland. and that it is styled by Unique from Potter's
Ardingle. The nightdress is in reasonable condi-
Ardil was also promoted directly to the public. tion and has clearly undergone regular washing.
In the 1950s, advertisements appeared in Microscopic examination of the fibres showed
women's magtvines, such as Good Housekeep- cotton plus Ardil.
ing. Ardll is described as 'rthe man-made pro-

s-
tein fibre soft as cashmere. smooth as silk
warm and-absorbent as wool. (It's moth resis-
tant, too.) Blended with other fibres, it gives
clothes the unmistakable touch of luxury at
prices you can afford." A housewife of the
1950s is shown selecting clothes for her family
pyjamas for her son, dresses for her daugh-
-ter and shirts for her husband. The copy encour-
ages her purchases; "Happy families wear
clothes that contain Ardil." Interestingly, she
never seems to wear it herself." A scarf made
with Ardil, which is now in the Nottingham
Museum of Costume and Textiles, may have
been a promotional gift from ICI to troops in Figure 2 Labelfrom nightdress of the 1950s, Wallis
the Korean War.
family archive, York Castle Museum (Museum num-
ber 43 1.7 8).
There was considerable concern to make sure
that consumers understood the new fibres and As previously mentioned, the Nottingham
knew how to care for them. In 1956 the News Museum of Costume and Textiles holds a scarf
of the World, Household Guide and Almanac from the 1950s that is labelled "Nobel Division.

88
An Ardil Blend, ICI Ltd." (Museum number
1979,609). The other fibre in the blend is not
known. The scarf is woven with a crimson and
green cross check and blue stripes with fringed
ends. According to the donor, the scarf was sup-
plied directly from ICI, but was not hard-wear-
ing as it had an unfortunate tendency to
dissolve in saliva.*

The largest collection of fibre samples, woven


fabric swatches and technical literature from
ICI, together with a bag of peanuts, is held by
the Whitworth Art Gallery, Manchester. There
are also tow samples, apparently of all three
Figure 3 Ardil mounted in XAM, magnified 400 times
types of Ardil. With the exception of one
swatch made by T. & J. Tinker Ltd. of Holm- Chemical and Physical Tests: When analys-
firth. Huddersfield. all the fabrics are from ing blended fibres, it is necessary to use a
ICI's Ardil Fibre Factory in Dumfries. The Sci- scheme. such as that outlined by the Textile
ence Museum, London, also holds some Ardil Institute to identifli the various elements.'- In
samples. this programme, the sample is tested using
normal sodium hypochlorite and 0.5 normal
Identification of Ardil sodium hydroxide to identify the presence ofre-
generated protein, wool or tussah silk. The pres-
Ardil is not easy to identify visually, particu- ence ofother fibres can also be established and
larly asit usually occurs as a blended yam and suitable tests carried out. Other techniques are
may vary widely in appearance. This may help required to identify which protein fibre is
to explain why so few examples have been iden- presented.
tified. It is important to sample both warp and
weft in blends thought to contain Ardil as they Burn Test: This test will establish the presence
may contain different fibres in different propor- of a protein frbre without specific identifica-
tions. It is also necessarv to identifu the other tion. Ardil bums without forming a bead with a
fibres present. smell of buming feathers as do wool and other
regenerated protein fibres.
Microscopic Analysis: On a microscopic level,
the longitudinal appearance ofArdil is straight, Wet Strength: This test can be used to estab-
uniform along the length and almost structure- lish the presence ofa regenerated protein fibre.
less although with some longitudinal striations. A marked loss of shength when the fibre is wet-
The edges are smooth. Liquid paraff,rn has been ted and then tested using a Strength Test ma-
suggested as a suitable mounting medium. The chine indicates the presence of Ardil or another
cross-section is circular with some slight pit- regenerated protein fibre.
ting, as shown in Figure 3.
Measurement of Fibre Density: This test will
* identify Ardil specifically. Moncrieff describes
Scarf mid 1950s, 1105 mmx216 mm. Label woven the method of establishing the specific gravif
yellow on black, "Nobel Division, An Ardil Blend,
of textile fibres. A glass tube is filled with two
ICI Ltd." A note from the donor, R.C.G. Williams,
liquids of contrasting density, which are al-
reads "supplied direct from ICI Ltd. who developed
the fibre type Ardil made from the protein part of pea-
lowed to diffuse into each other. The tube is
nuts. The fibre was made to augment the supply of then calibrated using fibres of known density
wool and was of importance during the Korean War before testing unknown fibres. Using this
( 1950 to 1953) when wool was in short supply. The method, it is possible to distinguish Ardil with
fibre has a "wool-ile" (sic) handle but is not very hard a specific gravity of 1.30 fi'om wool (specific
wearins and so failed when wool became plentiful." gravity 1.32) and casein (specific gravity

89
1.2r.24 An alternative method using a Tecam Unfortunately, at the moment, there is insuffi-
Density Gradient Column apparatus is set out cient evidence to establish a degradation pat-
by the Textile Institute. tem for Ardil. Once more textiles containins
Ardil have been identified it will be possiblJ to
Infra-red Spectroscopy: Once a known stand- assess the behaviour of naturally aged exam-
ard has been established, Ardil can be recog- ples. Tentative conservation recommendations
nized by the chemical constitution shown on can be developed based on documentary evi-
the infra-red spectrum in comparison with the dence. Experience in dealing with degraded
spectra of other known fibre samples. At pre- wool fibres is also relevant. Full testing for fi-
sent no such standard has been established. In- bre stability and dye fastness should precede
fra-red spectroscopy showed a protein fibre, any interventive treatment. The conservation re-
similar but not identical to silk. quirements of other fibres and materials in the
garment should also be considered.
Solubitity Tests: Solubilify tests can be used
to distinguish Ardil from other protein fibres. Wet-cleaning should be approached with cau-
In 80% sulphuric acid, Ardil is unchanged tion as Ardil never achieved good wet strength.
whereas wool and casein dissolve. Caustic soda It is possible that degradation could intensify
5% at boiling point can be used to distinguish this weakness. Relaxation shrinkage could oc-
between protein and regenerated protein fibres. cur, particularly in ArdiVwool blends. Little is
Ardil will stay stable whereas wool dissolves known about the possible reaction ofdegraded
and casein shrivels up. Ardil will dissolve when Ardil with detergents or other wash bath addi-
heated at97"C in 18% solution of sodium hy- tives. High alkalinity would be undesirable as
droxide whereas ot]r_er regenerated protein fi- the fibre becomes lulnerable to mechanical
bres remain stable.2s The ability of Ardil to damage at a higher pH, and a non-ionic deter-
dissolve in a cold aqueous solutio I of sodium gent with sodium carboxy methyl cellulose as
hypochlorite (3.5 g available chlorineilihe) is an anti-soil redeposition agent would be prefer-
used to analyse.percenjgges ofcomponent fi- able. The evidence from the donor of the Not-
bres present in blends.' tingham scarf that it dissolved in saliva would
suggest that the fibre is not stable when ex-
Stain Tests: Moncrieff reports that the regener- posed to certrain enzymes.
ated protein fibres casein, Ardil, soybean and
zen all give a yellorv-orange colour with cold Bleaching protein fibres is highly problematic
Shirlastain A stain.' Shirlastain A is used for and not recommended. Ardil was never stable
the identification of non-thermoplastic fibres, in sodium hypochlorite or sodium chlorite.
such as cotton, wool and regenerated fibres.
The presence of formaldehyde may affect the When new, Ardil was stable in most orsanic
results of staining tests. solvents. Following testing, solvent cleining
may be acceptable. Suitable extraction facilities
Tests for Formaldehyde: Ardil is sfrengthened will be required to meet health and safety stand-
with a crosslinking treatment using formalde- ards. It is not clear how the use of formalde-
hyde or glyoxal to modifu the fibres so as to hyde as an after-treatment will affect the
improve stability and to allow such standard long-term stability of Ardil. Neither is it clear
textile processes as dyeing to be carried out. whether textiles containing Ardil might be a
The presence of formaldehyde canbe identified potential risk to other vulnerable objects in a
using the Chromotropic Acid Test.'" mixed media collection, particularly metals,
pigments and paper, on account oforganic acid
Implications for Conservation vapoq; resulting from the formaldehyde treat-
ment."' Isolating identified textiles and careful
Once the presence of Ardil has been estab- monitoring is recommended.
lished, the conservator should be able to make
better informed decisions regarding the treat-
ment of the textile.

90
Conclusion GL. I The Properties of Ardil Protein Fibre
GL.2 - The Construction and Finishing of
Until more textiles containing Ardil are discov- -
Dimensionally Stable Ardil Fibre/Cellulosic
ered, it is difficult to assess the long-term stabil- Fabrics
ity and behaviour of the fibre. The author WD. 1, 2,3 & 4 The Processing of
would appreciate receiving any information 3l/zDenier and 5 Denier Ardil Protein Fibre
about such items. The speed with which both on the Worsted System
known examples and contemporary informa- WN. I The Processing of Ardil Fibre on the
tion about a recent commercially produced re- Woolen- System
generated fibre have disappeared should alert CN. l. 2,3 &4
both curators and conservators to the impor- - The Processing of 2,3t/z
and 5 Denier Ardil Protein Fibre on the Cotton
tance of acquiring suitable examples and record- Spinning System
ing information on modern fibres as fully as WG. I Sizing of Ardil Fibre/Cellulosic
possible. Warps -
FF. tudil Fibre in Fur Felts
Acknowledgement FW.- I Ardil Fibre in Wool Hat Felts
FG. I - The Scouring and Bleaching of Yams
The author thanks The Conservation Unit, -
and Fabrics Containing Ardil Protein Fibre
Museums & Galleries Commission for gener- FG.2 The Application of Crease Resistant
ous support and grant aid; Fiona Strodder, Cas- Resins-to Blends Containing Ardil Fibre and
tle Museum, Norwich; Michael Robertson, ICI Viscose
Chemicals and Polymers Ltd.;Jeremy Farrell, FG. 3 The Finishing of ArdiVCotton
Museum of Costume and Textiles, Nottingham; - Fabrics
Blended
David Howell and Alain Colombini, Textile AN. I The Quantitative Analysis of Ardil
Conservation Sfudios, Hampton Court Palace - and
FibreArlylon Ardil Fibre/Cotton Blends
for infra-red spectroscopy analysis; Jennifer AN. 2 The Quantitative Analysis of Ardil
Harris and Ann Tullo, Whitworth Art Gallery, - and Ardil FibreA/iscoseA.,lylon
FibreA{ylon
Manchester; Sonia O'Connor, York Archae- Blends
ological Trust; Helen Durrant and Josie Shep- AN. 3 The Analysis of Ardil
pard, York Castle Museum; and Mark Suggitt, - andQuantitative
Fibre/Wool Ardil Fibre/Viscose/Wool
Yorkshire and Humberside Museums Council. Blends
Photographs by Richard Stansfield, York Cas- Forthcoming
tle Museum, copyright City of York Leisure The Dyeing of Ardil Protein Fibre "B",
Services. - "F" and "K" and Ardil Fibre Unions
The Printing of Ardil Fibre Unions
Source of Materials - The Application of Caledon Dyestuffs
- to Ardil Fibre "B"/Cellulosic Fibre
Shirlastain A Stain is available from Shirley Blends
Developments Ltd., PO Box 61, 856 Wilmslow
Road, Didsbury, Manchester M20 8SA Eng-
R6sum6
land.
L'Ardil, la Jibre qui disparatt?
Appendix 1 Fibre obtenue par rdgindration de prottines,
I'Ardil, qui a ,!tt produit commercialement par
Technical Bulletins I'Impeial Chemical Industies (ICI) entre 195 I
et 1957. a servi d confectionner des v€tements et
ICI published the following list of available dfabriquer des pidces d'ameublement. La prd-
and forthcoming Technical Bulletins on sente communication traite du dtveloppement de
"Ardil A New Staple Fibre for the Textile I'Ardil et des caracttristiques de ce produit, de
- (ICI Fibres Division, Harrogate, no
Industry," m€me que de ses techniques defabrication et de
date, possibly 1949): sa commercialisation. Et puisqu'il n'existe que
peu d'exemples connus de cettefibre, elle rend

9l
compte des travaux qui s'effectuent actuellement 7. Moncrieff, R.W., Man-made Fibres,6th edn.
pour retrocer les tissus contenant de l'Ardil, et (London: Newnes-Butterworths, 1 975).
foit un certain nombre de recommandations au
sujet de la conservation de ce mattriau. L'Ardil 8. Lutyens, W.F., "Icvcourtauld' s Relations, "
constitue donc un ddfi pour les spdcialistes tant Central Registry 2 lCl, 2 F ebruary 1945,
1 00 I 21
de la conservation que de la restauration, et il in: Reader, Imperial Chemica l Industries, 197 0.
illustre toute I'importance que rev€tent I'etude,
I' identification et I' enregistrement des
9. "From Ground Nuts to Ardil," ICI Magazine.
fbres synthttiques. April 1951.

References 10. "Ardil Supporting Memo," Central Registry


3l I I l, 6146, 8 July 1947,in: Reader, Imperial
l. Ministry of Economic Warfare, London, Let- Chemical Industries, 1970.
ter to NBM Legation in Berne and Stockholm,
2l October 1942. Public Record Offrce I l. "The Cost of Production of Ardil Fibre,"
L40-9n.
Central Registry 3115,9 August 1955, in:
Reader, Imperial Chemical Industries, 1970.
2. Chibnall, A.C., K. Bailey and W.T. Astbury,
"Improvements Relating to the Production of 12. ICI Review', 1956, p.21.
Artifi cial Filaments, " British P atent 467,7 04,
Provisional Specification No. 29161, 1935.
13. Moncrieff, R.W ., Man-made Fibres
(London: National Trade Press Ltd., 1957).
3. "A Perspective on the Preparation ofProtein
Fibres," Central Registry Papers (CR), 29 June
14. Traill, D., "Some Trials by Ingenious
1939. ICI Head Offrces Records Centre in:
Inquisitive Persons: Regenerated Protein
W.J. Reader, Imperial Chemical Industries. A
Fibres," Journal Society of Dyers and Color-
History (Oxford: Oxford University Press,
rsrs, July 1951, vol. 67 ,pp.257-270.
1970).
15. Traill, D., G.K. Simpson and ICI, "Improve-
4. Chibnall, A.C., K. Bailey and W.T. Astbury,
ments in or Relating to a Method for Improving
"Improvements in or Relating to the Production
the Strength of Artificial Insolubilized Protein
of Artificial Filaments, Threads, Films, and the
Filaments," British Patent 639,342 (1947).
Like," British Patent 467 ,704, Provisional
Specification No. 20927, 1936.
16. Untitled, undated typescript, presumably
from ICI, now held at the Whitworth Art
5. "Ardil Development," Notes of a Meeting
Gallery, Manchester.
on l4 February 1939. ICI Development Ex.
Committee Papers in: Reader, Imperial
17. Dyke, F.M., "Ardil Protein Fibre Blends,"
Chemical Industries, 1970.
The Textile Weekly, February 1952,p.491.

6. Various ICI Patents:


18. "The Production of Shirting Cloths," 1C1
"Hardening of Protein Filaments," British Pat-
Technical Bulletin FG.l, undated.
ent 492,67 7 (1932) and 492,89 5 ( 193 8).
"Insolubilizing Protein Fibres," British Patent 19. Ardil Protein Fibre. A New Staple Fibre
766,360 (r9s4).
"Insolubilizing Protein Fibres," British Patent for the Textile Industry (Harrogate, U.K.: ICI
Fibre Division, undated).
787,s88 (less).
"Protein Filaments," British Patent 758,445
20. "ICI Launch Monkey Nut Fibre,"
(19s6) and 7 57 ,2r5 ( 1953).
I nternation al Textil es, v ol. 12, 1944.
"Method of Insolubilizing Protein Filaments, "
British Patent 763,501 (1954).
"Insolubilizing Protein Fibres," British Patent
758,560 (1954).

92
21. Scott-Moncrieff, G., Living Traditions of 25. Canoll-Porczynski, C.2., Manual of
Scotland (London: HMSO for The Council of Man-made Fibres (Guildford, U.K.: Astex,
Industrial Design Scottish Committee on the 1960).
occasion of the Festival of Britain, l95l).
26. "Qualitative Analysis of Ardil Fibre,Atrylon
22. Advertisements: Good Housekeeping, and Ardil Fibre/ViscoseA.{ylon Blends," 1C1
August 1955, p. 16, November 1955,p.7, Technical Bulletin AN.2.lCI. 1956.
December 1958, p. 5:Woman's Journal,
April,May 1956, p.4. 27.Hatchfield, p.R. and J. CarpenteE
"Formaldehyde. How Great is the Danger to
23. News of the World, Household Guide and Museum Collections?" Centre for ConJervation
Almanac (London: News of the World, 1956). and Technical Studies. Harvard Universitv Art
Museums, 1987.
24. The Textile Institute ldentification of
Textile Materials,Tth edn. (Manchester: The
Textile Institute, 1975).

93
Plastics Found in Archives

Alan Calmes

National Archives and


Records Adminis trat io n
l[/ashington, D.C.
U.S.A.

Abstract example, have gone through a number of


formats with different plastic materials, from
Ifith the rise ofplastics in the 1950s and 1960s, physically embossed grooves on dictaphone
there emerged novel ways ofrecording informa- belts, to magnetically charged particles on
tion. Dictaphone belts, Thermofax copies, and tape and, more recently, to laminated laser-pro-
all sorts of magnetic tapes joined vin1.,l discs and duced, digitally encoded disc recordings. Each
c el luloid film as base materials for permanentll'
involves a complex combination of plastics.
holding valuable information, such as speeches,
pictures, music, and data. Such mateials t.vpi-
cally reach an archive two or three decades after The desire to have quick copies ofpaper docu-
their creation, presenting a technically compli- ments led to the development of a variety of
cated preservation challenge. The cycle ofdevel- wet and dry and sometimes heat-processed
opment, usage, support, and obsolescence of coated papers during the 1940s and 1950s.
polymers in recording media is examined. Aging Some of these processes involved plastics. Be-
c haracteri s ti cs, c o ns erv a tion m easu re s, an d con- ginning in the early 1960s, the plain paper elec-
version options are examined in the context of trostatic photocopier produced an image using
arc hiv a I pres ervation adm i ni stratio n. copolymers mixed with carbon black and fused
to the paper surface. Early conservation meth-
Overview ods used plastics for the lamination of fragile
documents. The document, tissue paper, and
Throughout the l9th century, paper was practi- sheets of cellulose acetate were heat-pressed
cally the only material used for record-keeping, into a melt. Plastic enclosures and boxes have
except for some special applications ofparch- been used in archives, and plastic parts are
ment. Often, papers were bound together and found in information recording equipment. Ad-
stored in leather volumes, or tri-folded together ditionally, paints containing plastics have been
and pressed into wooden file boxes. Cellulose used on surfaces ofshelves and on containers,
nitrate, as a filler for pyroxylin bindings and as and plastic adhesives have been used to hold
a transparent substrate for photographic filrn, boxes and envelopes together.
marked the advent of manufactured plastics in
archives. During the first half of the 20th cen- Records are usually 20 to 30 years old by the
fury, vinyl, in the form of phonograph records, time they reach the care of an archivist. Before
replaced wax cylinders and joined film as a records are transferred to an archive, they may
plastic material for holding information. Since have been stored in harsh environmental condi-
1950, more and more plastics have been used in tions and subjected to rough handling. Archival
record-keeping practices. Sound recordings, for materials made of plastic often arrive in need of

95
special attention. For example, cellulose nitrate Many product names are used by non-special-
film may arrive in an advanced state of deterio- ists as generic names of plastics, such as Du
ration and be highly flammable and in need of Pont's Mylar for poly(ethylene terephthalate),
special handling, packaging and storage; dicta- Rohm & Haas's Plexiglas for poly(methyl
phone belts may arrive cracked and broken. methacrylate), and General Electric's Lexan for
An archivist needs to know the aging charac- polycarbonate.
teristics of plastics, conservation measures suit-
able for plastics, and copying techniques. Additives
Especially vulnerable are non-paper records, Plastics are seldom used in a pure state. Polyes-
such as motion pictures, video recordings, pho- ter used for encapsulation is generally de-
tographs, sound recordings, and computer data. scribed as a simple polyester, but even it
Some of the most important information of the contains some by-products left over from the
second half of the 20th century will require spe- manufacturing process, such as lubricants and
cial conservation and duplication efforts to pre- some silica compounds to prevent blocking.
serve the history of nations. A partial list of Polypropylene and polyethylene have antioxi-
important events recorded on plastics would in- dants added so that they can be melted and
clude: political debates, presidential addresses formed into sheets or poured into molds with-
and news conferences on radio and television; out undergoing oxidation. Additives add to the
satellite mapping and environmental observa- complex nature of plastic materials. Lubricants
tions; and motion pictures of historic events. in magnetic tape, for example, can ooze out of
An example of valuable information on a plas- the tape onto its surface. The reading-head ofa
tic medium is the sound recording of the tape-drive will collect the oozed-out lubricants
Nuremberg Trials, which was recorded on a and this can cause the reading-head to either
long loop of cellulose acetate film. The em- fail to read the data or gouge the surface of
bossed grooves and bumps have nearly disap- the magnetic tape, destroying the recorded
peared as the plastic material has gradually information.
retumed to its pre-embossed smooth state. A
special machine had to be built with a special Since plastic products may melt, burn and
tracking stylus to play the loop. spread a fre rapidly, flame retardants are often
added. Some local fire codes require the addi-
The majority of plastics found in archives tion of flame retardants in plastic materials
are thermoplastic materials, such as cellulose found in household and office fumiture. and.
nitrate; cellulose acetate; polycarbonate; therefore, flame retardants should be included
poly(methyl methacrylate) (PMMA or acrylic); in plastics found in abundance inside any build-
nylon; poly(vinyl chloride) (PVC); polystyrene; ing. Some flame retardants, however, can
poly(ethylene terephthalate) (PETP or polyes- evaporate in small quantities and affect materi-
ter); polyurethane; and polyolefins, such as als in contact with them. Some halogens, for
polypropylene. Some thermosetting resins are example, may produce oxidants that can react
also found, such as polyurethane, epoxy resin, with the silver halides of photographic film.
melamine formaldehyde resin, and phenolic
resins, such as phenol-formaldehyde resin, Changes in the Composition of Plastics
and Bakelite.
Plastics have evolved rapidly since 1950.
The following are examples of plastics found Chemical formulations have been replaced or
in archives: cellulose nitrate film; various cellu- modified to achieve desired results. New appli-
lose acetate films; vinyl (PVC) phonograph cations or improved materials brought about
discs; polyester encapsulation; polyolefin the obsolescence ofone form ofplastic in favor
shrink wrapping; polyethylene boxes and enve- of another. As a result, it is difficult to identifu
lopes; polyurethane binders on magnetic tape; old plastics found in archives without conduct-
acrylic sheets and blocks used in exhibits; ep- ing laboratory tests.
oxy resin adhesives and coatings; and nylon
gears.

96
Manufacturing processes have changed. Most plastics adhere to the ink and toner ofpaper
plastics originally served more immediate documents or to the imaging layer of photo-
needs and were not designed for long life. graphic prints. PVC plastic sheets have this
There was almost an assumption in our society quality because ofthe plasticizer, such as di-
that plastic products were disposable, and if octyl phthalate, which acts like a solvent in
continued use of the product was desired, it dissolving and athacting the copolymer ingredi-
would have to be replaced by a new and better ents oftoners and dyes used in electrostatic
product. This philosophy is changing, as manu- copies and photographic images. Furthermore,
facturers are beginning to produce engineering PVC should not be used for albums or filed
plastics with substantial durability and environ- with papers because it is an unstable plastic
mental resistance. Paradoxically, there are now that decomposes to produce hydrogen chloride
environmental concems that plastics are here (HCl) and, with a little water, this forms hydro-
forever; this fear has led to the development of chloric acid.
biodegradable plastic s for throw-away produc ts.
Plasticizers are solutes placed in hard plastics
The need to carry out periodic duplication on to reduce the glass transition temperature, and
plastic media extends to all types and formats: consequently to improve the flexibility of the
sound recordings, video recordings, motion pic- material. Poly(vinyl chloride) is an example of
ture film, and computer tapes. How often this is such a material that is normally glass-like. Plas-
done depends on how the medium is stored and ticizers are not chemically bound to the poly-
handled, and on the characteristics ofthe mer, and, over time, may come out of solution
plastics used. and be found on the surface ofthe plastic from
which they may evaporate if heated, or rub off.
Physical Characteristics Eventually, plastics that are plasticized will
"dry out," shrink, and crack. Other additives,
Plastics can be rigid or flexible, soft or hard; such as oils, lubricants, antioxidants, and cy-
they can be molded to almost any shape. Dur- anamides (also known as carbodiimides) may
ing manufacturing processes, plastics are malle- ooze out onto the surface of the plastic material
able and can be stretched as well as molded. as it ages. With the subsequent reduction in the
These attributes, convenient for making any solubility of the plasticizer in the plastic, a
shaped item, can cause problems later. With white powder is often found on the surfaces of
sufficient heat, plastics can retum to the malle- old films and tapes consisting of a number of
able state and change shape. When stretched additives that have come out of solution in the
during a forming process, plastics can be made plastic support and,/or recording layer.
into thin films; however, the material will con-
tinue to have a memory of an earlier state, Aging Characteristics
and try to revert back to a previous condition.
Many plastic films will relax back to their pre- Cellulose Nitrate
stretched dimensions if they are heated even Originally, almost all black-and-white 35-mm
briefly above a transition temperature that var- motion picture film was on cellulose nitrate
ies from one plastic to another. film. It was used exclusively for studio work
from the 1920s to the 1950s. Cellulose nitrate
Plastics can be made with such a smooth sur- was not used for color film, or for l6-mm and
face that when placed in contact with another 8-mm home movie film.
smooth surface, such as a photograph, pressed,
and then removed, the photograph will be left When ignited, cellulose nitrate bums very rap-
with a shiny surface. This process is sometimes idly. Nitrate film decomposes with the emis-
called "ferrotyping" (a term borrowed from pho- sion ofoxides ofnihogen. The reactions are
tography, referring to the process oftransfer- highly exothermic and are responsible for the
ring an image directly onto the smooth surface spontaneous ignition of cellulose nitrate. De-
ofa specially prepared iron plate). Such plastic spite the hazards, commercial film makers
sheets should not be used in albums. Another preferred cellulose nitrate film over cellulose
deleterious process, "offsetting," is when

91
acetate safety film because it was easy to han- Cellulose Acetate
dle and produced a very clear, sharp image The same general mechanism of deterioration
when projected. By 1950, however, after many of cellulose nitrate occurs in a similar way in
theater fires, fire codes mandated the use of other cellulose esters. The same acid hydrolysis
safety film. During the past 10 years, after sev- occurs in all cellulose materials, but the other
eral devastating and dangerous cellulose nitrate esters produce only an acid that catalyzes fur-
film vault explosions and fires, archives and li- ther acid hydrolysis, not an oxidizer like nitro-
braries have copied most cellulose nitrate film gen dioxide in the case of cellulose nitrate. The
images onto safety film (a cellulose acetate or degradation process ofcellulose acetate film
polyester film) and disposed of the cellulose takes longer than that of cellulose nitrate film,
nitrate film. There remains, however, some but once started, the autocatalytic chemical
spliced-in cellulose nitrate film within reels reaction cannot be stopped. The result ofthe
of safety film. Nitrate-based film stock can be self-destruction of cellulose acetate film is
identified by feel; it is softer and more supple somewhat different from cellulose nitrate film,
than cellulose acetate or polyester film. When in that an intermediate stage of deterioration
degrading, its appearance may be deceiving. It between a good condition and a powdery condi-
is therefore safer to have a laboratory test film tion can be seen. Acetic acid, a product of cellu-
to confirm its nitrate content. lose acetate degradation, can be detected by its
vinegar odor. The presence ofacidic gases and
The aging of cellulose nitrate is characterized particles found in polluted air will initiate the
by rapid change once the deterioration process degradation process of cellulose acetate film.
begins. Prior to the onset of deterioration, there Unlike nitrate film, however, the image layer is
is no serious shrinkage of the material, image not chemically affected by the by-products of
quality is good, and the images can be copied the decomposition of the acetate substrate.
easily. The kinetics ofthe reaction are such that
there appears to be virtually no intermediate The first safety-based films were cellulose
stage between the time when the film is in good mono-acetate and cellulose diacetate. The term
condition and the time when it is obviously de- "safety-base" was used because acetate films do
teriorated. Archivists have seen reels of cellu- not bum easily. By the 1970s, triacetate began
lose nitrate film change from excellent to to replace diacetate as the favored substrate
extremely poor condition in two months. When for film.
it deteriorates, cellulose nitrate film produces
sticky-brownish, powdery-fibrous globs. Manufacturing experience found that the
diacetate substrate took diazo salts more readily
Gases emanating from deteriorating cellulose than the triacetate base and thus the diacetate
nitrate film can initiate the process of deteriora- base was used in diazo films until polyester
tion in neighboring films of all types. Chemi- began to be used in the 1980s.
cally, cellulose nitrate, upon decomposition,
produces its own oxidizer, and, therefore, once Researchers at the Image Permanence Institute
the chemical bonds begin to break down, a in Rochester, N.Y., are frnding little difference
rapid autocatalytic reaction sets in. The reaction between the degradation processes of cellulose
produces its own heat, which accelerates the diacetate film and cellulose triacetate film. One
initial breakdown. The process can be fast is not necessarily more stable than the other.
enough to produce fire and, if film is tightly The process, however, might take longer with a
compacted, an explosion can result. Once a de- triacetate than with a diacetate. Within one cate-
grading cellulose nitrate film has been identi- gory there are likely to be variations from one
fied, the archivist must move quickly to copy batch to another, depending on formulas. and
the images onto safety film and to dispose of additives, and manufacturing processes.'
the cellulose nitrate film. The continued useful-
ness of a reel of cellulose nitrate film depends The long, intermediate stage of deterioration of
upon good environmental conditions, such as cellulose acetate film is characterized by shrink-
clean, cool, and dry air. ase. When the cellulose acetate film base

98
shrinks, the emulsion layer on top, which does microfilm in the 1980s, but it co-existed with
not shrink, is deformed into a mass of wrinkles. cellulose triacetate-based microfilm.
Since the image is within the emulsion layer,
the image becomes illegible, unless there is a With the right combination of conditions, acid
way to copy or transfer the emulsion layer be- hydrolysis can break the bonds between mono-
fore the wrinkling obscures the image. The mers of polyester. During this process acids are
choices are to copy the image before this oc- created that in tum break more bonds. Once
curs or to laboriously remove and reapply the started, degradation is autocatalytic. Hydrolysis
image layer after it occurs. The greater the can be initiated by acids present in polluted air
shrinkage, the more difficult it is to recover the or left over from the manufacturing process.
information. When motion picture film shrinks, Oxides of nitrogen from automobile exhaust
the sprocket holes are no longer in the right form acids in the atmosphere that can acceler-
place, making the copying process, necessary ate the degradation process ofplastics. Exclu-
to save the images, difficult and expensive. sion of water from the air, that is, maintaining
The shrinkage will be uneven; consequently, en- a low relative humidity, is an effective strategy
gineered sprockets with a different spacing may to prevent hydrolysis. Using scrubbing systems
not provide a solution. or treated charcoal filters to remove pollutant
gases from the air are other strategies. The
Until the late 1950s, cellulose acetate films lower the temperature the slower the rate of
were used as a very thin base material for chemical reactions. To slow down degradation,
sound recording tape and for some early video carbodiimides are added to react with acids and
and computer tape. With age, and accelemted antioxidants are added to prevent reaction with
by elevated temperatures and high relative hu- oxygen; however, the additives will eventually
midities, thin cellulose acetate tape becomes be cons^umed, ooze out or evaporate from the
brittle and will break easily. Brittleness is an in- plastic.'''
evitable condition. Plasticizers used during the
manufacfuring process will ooze out onto the Polyester film is bi-axially oriented or "bal-
surface of the tape in the form of white droplets anced." Cellulose acetate is often uni-axially
that look like powder. To complicate matters, a stretched. Due to its limberness, polyester tape
recording layer, such as one composed of iron is more difficult to handle than cellulose acetate
oxides dispersed in polyurethane, will have tape. Polyester tends to respond to tension. For
characteri sti cs of degradation different from example, it will curl under tension. These char-
that of the base material. acteristics vary with the thickness of the tape.
The much thicker photographic film will dem-
A break in cellulose acetate magnetic tape is onstrate different properties, such as springi-
usually clean. It can be spliced back together ness rather than limbemess. Polyester has what
again with little loss of information for an ana- is called "plastic memory," and tries to go back
log sound or video recording. This is different to its pre-stretched state.
from polyester-based magnetic tapes that
stretch considerably before breaking. In order to reduce volume, polyester-based
magnetic tape for computer use has become ex-
Polyester tremely thin, but there is, concomitantly, a
Poly(ethylene terephthalate) entered the scene higher risk of loss of information, as a small
in the late 1950s. Because of its strength it was amount of stretching or some other dimensional
used, even in a very thin film, as the substrate change can cause the loss ofdata.
for all computer and video tapes. Soon thereaf-
ter, polyester replaced cellulose acetate as the Engineering Plastics
base material for sound recording tape. The Engineering plastics, such as nylon, polycarbon-
switch from cellulose acetate to polyester-based ate, and phenolic resins, have replaced metal
photographic film has been very slow. The de- for machine parts in modem information record-
mand for a strong, long lasting microfilm ing and retrieving devices. For example, video
brought about the use ofpolyester for players now have more plastic machine parts

99
than ever before. Plastic machine parts have adhesives can damage not only the surfaces
the advantage of being lightweight, tough, and where they are applied but also can initiate the
wear resistant, they do not need lubrication, degradation of nearby areas of papers, films,
and their gear trains operate quietly. Contrary and tapes.
to popular opinion that plastics are cheap substi-
tutes for metal components, plastic machine Polyolefi ns and Polystyrene
parts are often more expensive than the metal Plastic containers have significant advantages
parts they replace. Quality is a factor with plas- over metal cans and cardboard boxes. They do
tic parts as well as with metal parts. Quality of not corrode, are lightweight, and are unaffected
the product depends upon quality of the manu- by high humidity or water. Polypropylene mo-
facturing process for the plastic part and the tion picture containers are being used in some
quality of assembly into a machine. archives. There are no solvents or plasticizers
in polyolefins. However, colorants may come
There is a need for guidelines for the long-term to the surface and cause a problem and flame
maintenance of plastic machine parts. They retardants required by fne codes may slowly,
should not be lubricated. There is a problem even at normal temperatures, emit small quanti-
when part of a gear train is plastic and another ties of reactive materials, which might affect
part metal, because the lubricant needed for the the contents ofthe container.
metal part may cause the plastic part to deterio-
rate. Spare parts should be obtained before the Plastic cartridges can warp and cause magnetic
machine becomes obsolete. tape to mis-track when read and to mis-align
when rewound. In the latter case the edges of
Acrylics, Polycarbonates and Epoxies the tape may rub against the sides of the con-
Acrylic resins. polycarbonates. epoxy resins. tainer. Plastic pressure pads in magnetic tape
and various mixtures of these are used as cartridges have been known to crumble after a
shields, supports, adhesives, coatings, and few years. The pieces ofpad can damage the
toners for information recording systems. We reading machine and get between the layers of
have, however, very little experience on the use tape, causing the tape to be wound unevenly.
of these plastics in archives. Acrylics are better
known by their product niunes, such as Plexi- Because magnetic tape is thin and will sag
glas and Lucite, but acrylic products come in a against the flange when stored horizontally,
variety of forms from solid materials to liquids. magnetic tape should be stored vertically,
Some acrylic products turn yellow and become suspended on a hub; otherwise, an unevenly
brittle upon exposure to ultraviolet light or un- wound tape will result in having its outlying
frltered sunlight. Recent materials use ultravio- tape-edges folded under by the weight of the
let blockers to reduce the damage caused rest of the tape. Motion picture film is thicker
by light. and stiffer than magnetic tape and is stored hori-
zontally without flanges with the film resting
Poly(methyl methacrylate) (PMMA) has been on the surface of the container. The fihn must
used for compact discs (CDs) and CD-ROM be wound evenly and snugly. Instead of metal,
discs. Often, such discs were also coated with hubs and flanges are sometimes made of poly-
epoxy. Consumers began to note failures in styrene. In special cases glass has been used
these products in the late 1980s. Beginning in for flanges. Plastic is substantially cheaper and
1990, CDs and CD-ROMs were beginning to lighter than metal or glass; thus, we can expect
be made from polycarbonate, which is tougher to see more plastic.
and more resistant to change than PMMA.
Combinations of Plastic Materials
Much work remains to be done on plastics used
in adhesives. For example, pressrue-sensitive la- Combinations of materials abound in archives.
bels may fall offarchival boxes, tape splices on Some of these are borlnd together into lami-
motion picture film or magnetic tape eventually nated "sandwiches." For example, bound vol-
may fail and need to be replaced. Plastic umes consist of many layers of materials glued

100
together. An old phonograph record was con- produce the recording, or research will be
structed of a layer of shellac painted onto glass necessary to determine the appropriate size of
or metal. Magnetic tape base is made up of a stylus and use an appropriate transducer and
layer of pollurethane laminated to a substrate amplifier and play the recording at the
ofpolyester. Each layer has a different coeffi- correct speed.
cient of expansion. Changes in environmental
conditions can cause the separation oflayers. The pace oftechnological change has
The chemical products of decomposition of one quickened during the 20th century. Vinyl
layer can affect the other layer. One layer may records have been in use for over 50 years,
become brittle while the other remains flexible. analog sound recordings on magnetic tape for
50 years, digital recording for 30 years, and for-
Static Electricity mat changes are now occurring nearly every
decade. Plastics have played a role in the quick-
Plastics can hold a static electric charge and can ening of change by providing an infinite variety
release sparks and electromagnetic radiation. of new materials for adaptation. For example,
Static electricity is a problem for the microelec- computer information systems are constantly
tronics of some machines. Some manufacturers being upgraded with increased information den-
try to mitigate the problem of static electricity sity and reading speed, such as dye-polymers
by adding an anti-static layer to the back of used in some optical-magnetic recording sys-
tape or placing an anti-static additive directly tems to provide very high density storage and
into the plastic. Tapes coming out of long-term fast random access.
storage should be equilibrated at greater than
30% relative humidity to di-sipate the electrical The preservation of machine-readable informa-
charge and metal reels should be grounded. tion depends on periodic copying. Ifcopying is
Static electricity also attracts dust and grit that carried out properly and in a timely manner,
must be kept away from magnetic tape or disc there is little loss of information. However,
surfaces to prevent a reading-head from bump- there must be an assumption of continued cost
ing into them. Note that polystyrene, the plastic to pay for periodic duplication and new hard-
material used for tape flanges, can also collect a ware/software updates.
static charge. Dust and grit are also undesirable
on film since they scatter light, cast shadows Conclusion
and scratch the emulsion. Also, it is undesirable
to attract dust and grit at the time of polyester People in our society have ambivalent ideas
encapsulation of documents. about plastics. The memory of how plastic toys
break and cannot be fixed gives the impression
Obsolescence of plastics as cheap, ephemeral, and disposable.
Certainly, most people would not give a plastic
Rapid changes in technology during the object for a keepsake. From another point of
20th century have compounded the problems view, seeing plastics floating around in the mid-
conceming the maintenance of information in Atlantic Ocean stimulates environmental con-
other than readable form. For example, from cems that plastics pollute and will never go
the 1930s through the 1950s, sound-recorded away. The replacement of metal parts with plas-
dictation was kept on a cellulose acetate belt, tic parts is accepted by some as progress and is
sometimes referred to as a dictabelt from a dic- seen as an improvement; others would argue
taphone. At first, the sound frequencies were that the switch represents a cheapening of the
embossed mechanically into the surface of the product, even though it may look better.
cellulose acetate belt; later, the sound was re-
corded magnetically onto the surface of the Since plastics can be manufactured and molded
belt. A simple stylus can translate embossed into almost any shape at a fraction of the cost
vibrations into sounds, like that used for vinyl of other materials, there are economic forces
disc players. To play the magnetically recorded driving the use of plastics. With an acceptable
dictation on a belt, it will be necessary to obtain short life expectancy ofmost products today,
the same tlpe of machine as that used to perhaps attributable to rapid obsolescence,

l0l
frequent changes in style, and the desire to re- R6sum6
duce manufacturing costs. manufacturers are
using the less expensive plastics in products Les plastiques prisents dans les archives
that are virrually disposable after a few years.
Some plastics, however, such as polyester and L'utilisation de plus en plus rtpandue des
melamine are expected to remain durable and plastiques dan.s les annies 50 et 60 a foit naitre
to last for cenfuries. The various plastics have de n o ttve I I es faq on s d' e n regi strer I' i nfo mt ati o n.
their various environmental requirements, and Les bandes de Dictaphone, les copies Thermofax
as long as they are met, the plastic materials et toutes sortes de rubans nngnttiques ont ainsi
rejoint les disques en vinvle et les.films de cellu-
will serve their intended uses well. Maintaining
loid au rang des matiires de base qui seruent de
a benign environmental storage condition is
support petmanent aux enregistrements de dis-
key to extending the life of any material. Con- cours, de photographies, de pidces de musique,
stant low relative humidity between 30o/o and de donndes. Or, en ghtiral, ces documents ne
50% should benefit all plastics. Temperatures sont placds dans les archiyes que deux ou trois
should be kept as low as practical, between dic'ennies aprds leur creation, de sorte que leur
5oC and 20oC. consenation n'est pas sans poser d'tnormes
dfficultds techniques.
Plastics are manufactured materials subject to
endless changes in proprietary formulations. Nous examinerons, dans le cadre de la
After a formula has been used and the plastic prdsente comnunication, le cycle de developpe-
product is replaced by a new model, no one ment des poll'mires qui entrent dans la composi-
will know what formula was used or what addi- tion de tels supports et leur utilisation, tout en
tives were placed in the old plastic product; in traitant des mesures de soutien qui peuvent leur
which case, it is almost impossible to predict €tre appliqudes et de leur obsolescence. Nous
the life expectancy of the medium. Only when aborderons dgalement, du point de vue de I'ad-
we know the history of the material and its ministration d'un service d'archives, leurs carac-
chemical composition can we reasonably ex-
tiistiques de vierllissenrent, de m€me que les
diverses mestffes et options de conversion aux-
pect a certain performance and life expectancy
quelles on peut at,oir recours pour assurer la
of the plastic medium. We need the cooperation consentation de ces supports et des informations
of manufacturers to reveal the complete for- qu'ils contiennent.
mula of each plastic. With advance warning in
hand, archivists, therefore, can program replace-
References
ment costs to allow for a periodic migration of
information from one plastic medium to l. Adelstein, P.2., J.M. Reilly, D.W.
the next. Nishimura, and C.J. Erbland, "Stability of
Cellulose-Ester Based Photographic Film,"
Acknowledgement SMPTE Journal. May 1992, pp. 336-353.

The foregoing information on plastics found in 2. Smith, L.E. et al., "Prediction of the Long-
archives was derived primarily from interviews Term Stability of Polyester-Based Recording
with archivists and technical staff in the Na- Media," (Gaithersburg, Md.: National Bureau
tional Archives and Records Administration of Standards, 1986) Report No. NBSIR-
(NARA) and with polymer chemists and photo- 86t3474.
graphic film and magnetic tape standards ex-
perts at the National Institute for Standards 3. Smith, L.E., "Factors Goveming the
and Technology (NIST). Susan Lee-Bechtold, Long-term Stability of Polyester-Based
Chief Chemist, and Charles W. Mayn, sound Recording Media," Restaurator, The Interna-
and video recording engineer, both of NARA, tional Journal for the Preservation ofLibrary
and Leslie E. Smith, polymer chemist, and and Archival Materials, vol. 12, 1991,
Thomas Bagg, both at NIST, were particularly pp.20l-218.
helptul.

r02
occupe dtsormais dans notre civilisation. Le les bact,lries et m€me, la < chimie D, ce mot re-
terme ( caoutchouc > designe tout pob,a)vs q11i doutable), qui ont souvent un elfet qtnergique.
possdde ou semble poss'lder de l'ilasticit2, et il Aussi tenterons-nous, dans le cadre de la pr6-
existe aujourd'hui nombre de polym'ires d'ori- sente communication, de nous frat'er un chemin d
gine naturelle ou synthdtique qui r,lpondent d travers ce vtritable < champ de mines >, et
cette d|finition. Or, l'2lastom'ire peut tr'is bien d'tlaborer une mdthode qui permettra de choisir,
constituer moins de 50 2(' de la masse totale d'un pour chacun des divers genres d'dlastom,ires ex-
produit dit < de caoutchouc >, puisque plusieurs posis, la techniEte de restauration qui convient
substances chimiques, choisies parmi des cen- le mieux.
taines, peuvent avoir ttt ajoutdes pour lui donner
une forme utilisable. Bibliography
La prdsente communication ne vise nullement History of the Rubber Industty, eds. P.
d fournir une analyse scientifique detuiilee de la Schidrowitz and T.R. Dawson (Cambridge,
prdparation des produits de caoutchouc. Il n'en U.K.: W. Heffer and sons, 1952).
demeure toutefois pas moins que, avant d'amor-
cer tout travail de restauration d'un objet de Hancock, T., The Origin and Progress of the
muste, les sptcialistes de ce domaine devront
CAOUTCHOUC, or India-rubber Manufacture
absolument connaitre les matiriaux qui le com-
in England, eds. Longman, Brown, Green
posent et €tre bien conscients dufait qu'ils ris-
(London: Longmans and Roberts, 1857).
quent, s'ils n'appliquent pas le traitetnent
approprii, de luifaire plus de mal que de bien.
Aussi tenterons-nous de pr,!senter un historique Gottlob's Technolog, of Rubber I 92 5, Eng.
des dlastomdt'es qui met tout particuli'irement Trans., J.L. Rosenbaum (London: Maclaren and
l'accent sur certaines ddcouvertes du Xf si?cle sons,1927).
qui permettront aux spicialistes de la conserva-
tion de mieux classifier, voire de dater, les pro- Woll H. and R. Wolf, Rubber, a Story of Glory
duits dlastomdres. Au-deld de cet objectif, nous and Greed (New York: Covici and Friede,
dicrirons bi,ivement certaines des m,lthodes 1936).
simples qui peuvent ,?tre utilisdes pour analvser
ces produits. Chemistry and Technologv of Rubber, ed. C.C.
Davis (New York: Reinhold Publishing Corp.,
Il arive souvent que la sutface d'un produit t937).
ilastomdre en vienne d prtsenter des change-
ments qui, quoique ddplaisants sur le plan visuel, Buist, J.M., Ageing and Weathering of Rubber
ne sont pas toujours le prtsage d'une ddgrada-
(Cambridge, U.K.; W. Heffer and sons, 1956).
tion de la sudace; certains de ces changements
sont m€me voulus, pour permettre au produit de
Cook, J.G., Rubber (London: Frederick Muller,
durer plus longtemps. Nous passerons en revae
l 963).
les causes et les effets iventuels de tels change-
ments de surface, tout enfournissant des tech-
niques qui serviront d les distinguer entre eux. Wake, W.C., B.K. Tidd and M.J.R. Loadman,
Analysis of Rubber and Rubber-like Materials
Si I'on connait bien les ntatdiaux qui composent (London: Applied Science, 1983).
un objet de musde et que I'on peut, d'apr,is I'dtat
de sa surface, ddterminer ce qui lui est ariv,!, il Natural Rubber Science and Technologlt, ed.
sera d'autant plw Jacile de dejinir les mesures d A.D. Roberts (Oxford: Oxford University
prendre pour optimaliser sa vie utile d'exposi- Press,1988).
tion. Il n'est pas de solution qui puisse s'ap-
pliquer d tous les |lastomdres et d leurs produits, Toxicity and Safe Handling of Rubber
mais on retiendra qu'ils se digradent le plus sou- Chemicals (Birmingham: British Rubber
vent sous I'action de toute une sbrie de facteurs Manufacturers Association. I 990).
courants (l'oxvgdne, I'ozone, la chaleur, la lu-
midre , le travail ntbcanique, les mitaux oxydants,

74
Processes of Deterioration
Processus de d6gradation
Changes in Polymeric Materials with Time

David M. Wiles

Plastichem Consulting
Victoria, B.C.
Canada

Abstract Less well known is the phenomenon of physical


aging, whereby polymeric materials continue to
Virtually all plastics, fibres, rubbers, paints and alter for w,eeks, months or years after they have
protective coatings, as well crs paper, wood, skin solidified or otherwise attained the physical form
and hides, owe thetr useful characteristics to the in which they are used. Originally obsented in
relatively high molecular weights of their mole- thermoplastics belov, their glass transition tem-
cules. Values in molecular weights range from a peratures (Tg), physical aging arises because all
temperature-dependent propefties, which change
feu' t ho u s a n d t o s ev er al m i I I i on. Unfo nu n at e ly,
fabrication, handling, use or misuse of these abruptly at Tg, continue to change below that
materials result in deleteious changes (usually temperature, albeit verv slowly. Physical agtng is
decreases) in the molecular weights of the con- observed in amorphous, glassy polymers, in the
stituent molecules with concomitant reducti on amorphous phase of semi-crystalline polymers
in the materials' useful properties. The detailed and in the rubbery matrix offilled rubbers. As a
science ofthe changes at the molecular level is result, over time, many mateials become stifer
highllt complex and, frequently, not well under- and more brittle as "rates of relaxation" de-
stood. Nevefiheless, somefeatures common to the crease. Because physical aging affects polymer
degradation of many polymeric materials with segmental mobility, it should also affect chemical
time have been elucidated and can indicate the degradation, photo-oxidation, swe lling and de-
wa-v to improved preservation practices. swelling, and crosslinking reactions in a w'ide
v ari ety of macromo I ecul ar sys tems.

C hemical aging includes oxidative deterioration


as a result ofexposure in air to light, heat or ion- Introduction
izing radiation, and also includes hydrolysis and
auack fui acids or bases. Biodegradation can be The degradation of materials is identified by
considered a special case ofchemical breakdown the user/observer as an unacceptable change in
caused by microbial enzymes. The reactions in- characteristics, be they mechanical, chemical,
volved in chemical aging are numerous and some- optical, or electrical. In the case of polymeric
what material-specific, but the more important materials, that is, those comprising very large
ones canfrequently be related to the oxidation of molecules, it is usually an alteration in the mo-
liquid hltdrocarbons. The discolouration, embit- lecular weight of these molecules that results in
tlement and reduction in various phltsical proper-
an undesirable change in properties leading,
ties that accompany oxidative deterioration can
for example, to mechanical failure or discol-
be minimized by the use of appropriate combina-
tions of stabilizing additives. Microbial suscepti-
ouration. Occasionally, with some materials,
bilit)) is best dealt wilh bv "good housekeeping" crosslinking occurs (the formation of chemical
practices. bonds between molecules) with a resultant

105
increase in molecular weight, but more com- Hydrolysis, as well as deterioration from expo-
monly chemical reactions cause a reduction in sure to acids or bases, can be a problem in hu-
the molecular weight of the large molecules, mid or "hostile" environments for polymers
which collectively are the origin of the useful having specific structural features that are inher-
properties of polymers. Not infrequently, how- ently susceptible. Ester linkages are hydrolyz-
ever, unacceptable property changes can occur able, for example, and molecules that can react
even though very little change in molecular with acids or bases will usrrally be degraded by
weight can be measured. In short, polymer deg- them. Microbial attack may be a problem for
radation is widespread, complex and sometimes natural polymers since, in warm, moist air, fun-
difficult to evaluate; it can also be very difficult gal enzymes can oxidatively degrade many
to prevent. such macromolecules. Biodegradation is fortu-
nately not a factor in the deterioration of many
Much of the progress in elucidating the degra- synthetic polymers.
dation of macromolecules is summarized in
the articles and monographs listed in the refer- It is worth remembering that the molecular
ences.'-'- It is iustihable to look for simolifica- weights of polymeric materials are really very
tions and geneialities that assist in under- high and that their desirable properties will
standing and dealing with polymer degradation. be lost when only a very small fraction of the
Tuming first to chemical aging, there are nu- bonds per molecule are broken. The molecular
merous features common to the oxidative dete- weight of the molecules in cotton averages over
rioration of many kinds of macromolecules as a 2 million, in wood cellulose over 1.5 million,
result ofexposure in air to light, heat, ionizing and in paper several hundred thousand. Among
radiation or mechanical action. Some of these the synthetic polymers, nylons and polyesters
overall features are summarized in the figure having superior physical properlies may have
below. It has proven useful to relate these kinds molecular weights of only 30 thousand or so,
of chemistry to the oxidation of liquid hydrocar- but polyolefins are characterized by values
bons. Such model compound studies have iden- from a few hundred thousand to more than a
tified mechanisms that apply (at least in part) to million. The breaking of one bond in, say,
the degradation of solid polymers; kinetics, the l0 thousand in polymer chains will have a dev-
rates of the critical reactions, are rather more astating effect on the properties of the material
composition- and state-specifi c. Nevertheless, (unless the scission occurs near the ends ofthe
considerable progress has been made in devis- molecules). Such sensitivity is rarely observed
ing methods (largely chemical) of postponing in small-molecule science.
for prolonged periods the inevitable
degradation. Physical aging is an entirely different phenome-
non in that it is thermodynamically driven and
cannot be stopped although it is readily revers-
ible. Moreover, it is by no means invariably un-
desirable. All temperature-dependent properties
of materials that change abruptly at the glass-
transition temperature continue to change be-
low Tg. In effect, materials become stiffer and
more brittle; rates of relaxation decrease so the
response of a material to deformation or impact
changes, usually for the worse. Physical aging
in amorphous polymers and related materials
persists for a very long time and affects a vari-
Ro. +.OH ety of properties so that repair and replacement
,ro"r,F*t is tricky and, of course, longevity may be com-
prom ised. Additionql^information can be found
in the references.''-"'

106
Oxidative Degradation The identification ofthose processes that cause
the loss of the essential properties of polymers
A very high proportion of the common poly- has consumed a very great deal of research time
mers, both natural and synthetic, have mole- during the past 40 years and it is by no means
cules that consist of chains of carbon atoms complete. One singularly useful approach has
bonded together or chains of carbon atoms in- been to identify the more important products of
terspersed with other atoms, such as oxygen polymer degradation and, using principles es-
(e.g., in polyesters, polycarbonates, polyethers) tablished with low molecular weight analogues
or nitrogen (e.g., in polyamides, polyure- for the polymers, "reconstruct" the chemistry
thanes). Invariably, hydrogen is bonded to that gave rise to the products. Thus, under ther-
most of the backbone carbon atoms but, occa- mal stress, the weakest bonds in the system are
sionally, so is oxygen or chlorine, for example. most likely to break; these could be tertiary
Aromatic rings may be part of the polymer carbon-hydrogen bonds, for instance, or peroxi-
chains (e.g., PET, aramids) or pendant from dic linkages. In the case of light-induced dam-
them (e.g., polystyrene). Energy in one form or age, it is the shortest wavelengths (highest
another is introduced into macromolecules energy) in the incident radiation that do most of
either during fabrication (or other handling) or the damage and there must be appropriate chro-
during use and, inevitably, chemical bonds of mophores (light-absorbing groups) present for
the types referred to above will be broken ho- any damage to occur. For example, the more
molytically, that is, to produce highly reactive cotton or other cellulosic is purified, the whiter
free radicals, in pairs. Sometimes, it is these pri- it becomes and the less susceptible it will be to
mary reactions by which molecular weights are actinic deterioration because the relevant chro-
reduced below acceptable levels, but it is fre- mophores are present in extractable impurities
quently the case that the damage is done in sec- rather than in the polysaccharide chains them-
ondary or even tertiary reactions. The varied selves. Another example is that the purer a poly-
response of many common polyners to degra- olefin, the less rapidly it will photodegrade,
dative environments is illustrated in Tables I since it is the chromophoric impurities in these
and Il,.which are taken from Carlsson and thermoplastic molecules that give rise to their
Wiles.' photosusceptibility.

Table I
Susceptibility of Unstabilized Polymers to Degradation
Resistance to degradative process'

Thermal Photeoxidation
Polymer Oxidation Weathering Ozone Hydrolysis Oxidation
polyethylene p e e f
polypropylene vp vp e p
polystyrcne I Y
poly(m ethyl methacrylate) s I
poly(tetraf luoroethylene) e e vp
polyamrde (Nylon-6 and -6,6) f f f t
polyacMonrtrile p e f
s
poly(vinyl chloride) vp p p
poly(ethylene terepthalate) s g
s
polyorymethylene p p I f vp
polycarlconate I p
poly(phenylene oxrde) vp t g s
poly(ester urethane) f t I
s s
poly(ether urethane) p p f t
s
poly(/.n-phenylene
rsophthalamtde) e
poly(p-phenylene
terephthalamtde) e

' Key: e = excellent, g = good, t = farr, p = poor, and W = very poor

t07
Given that some carbon-carbon bonds are bro- apply whether bonds were cleaved initially
ken in due course, a general series ofreactions by heat, light, ionizing radiation or mechanical
may be written: stress. The behaviour of a reactive chemical
species, such as a polymer radical, is deter-
Polymer heat, lrght Carbon{entred mined by environment (both macro and micro)
Molecules RH 2R' radicals, rn
more than by origin; the radical has no memory
ronrztng parrs
radration (see Figure 1). This feature has been helpful in
the elucidation of polymer degradation mecha-
RH
R'+02-t -> ROz' TROOH*R' nisms as well as in the development of many
highly effective stabilization systems.
heat RH
ROOH 'OH + RO' ROH + R'
lrght / \\ There are many synthetic polymers that have
-> PRH -> chromophores as part ofthe repeat units, that
HzO + B' R'+ ketones
is, they are built into the polymer chains. Car-
2RO2' Stable Products bonyl groups in polyesters, nylons (including
-> aramids), polycarbonates and polyurethanes are
examples of chemical structures that absorb spe-
Two features ofthis sequence should be obvi- cific wavelengths in terrestrial sunlight. Nor-
ous. Since each initial carbon-centred radical rish-type photodegradation reactions ensue,
(R') gives rise to several other radicals, this is a leading to the formation of radicals that un-
branching chain reaction overall and, ifnothing dergo the same kinds of chemistry illustrated in
intemrpts the sequence, oxidative degradation the scheme shown earlier. There are other kinds
ofa polymer can proceed rapidly. A second fea- ofphotochemistry and other types ofdegrada-
ture is that there is likely to be some difficulty tion reaction in the case ofcer|ain specific poly-
in sorting out all the degradation products and mers, but the scheme is valid for a wide variety
where they came from, especially since some of macromolecules.
are themselves heat- and light-sensitive. A
third, less obvious feature ofthis reaction In addition to reducing the tensile or flexural
scheme is that it should apply (at least in part) performance of polymeric materials, oxidative
to a wide variety of polymer types and it should degradation llequently results in surface

Table II
Thermal Deterioration of Polymers Maximum lJse Temperaturea

Polymer Generic Type Film Thickness, mm Maximum Use Temperature,'C

poly(vrnyl chloride) 2.O 50


polyethylene f,U
polyoxymethylene 0.7 50
poly(methyl methacrylate) 1.5 50
polyslyrene 31 50
poly(phenylene oxrde) 15 50
polyamrde o.7 65
polycarbonate" o.7
epoxy resrns 3.0 90
$ilcones 0.7 105
poly(ethylene terephthalate) 0.2 105
alkyd resn 1.5 130
polytetraf I uoroethylene 0.9 150
phenohc resin 2.0 150
polyrmrde 0.1 200
aramtd" o2 220

a Recommended use temperature at whrch 50% of the ongrnal drelectric strength, tensle, and rmpact properttes are
retatned for 1 1 ,000 h under low contrnuous stress In some cases, stablhzers may be present. Underwnters
Laboratones data.
b Bisphenol A
c Aromatrc polyamrde.

r08
cracking, discolouration and enhanced surface fungi that operate by excreting water-soluble
wettability, for example. In all cases, the trick enzymes onto a biosusceptible substrate.
is to try to identify those chemical reactions Subsequent breakdown of the molecules of the
that cause the loss of useful properties since substrate material into water-soluble fragments,
these are the reactions that need to be orevented for example, two carbon-atom chunks, is fol-
as much as possible. Commonly. materials are lowed by transfer back inside the mycelial
subjected to heat and light stress simultane- cells. Fungi like to be warm and moist and they
ously. Conventional wisdom has it that photo- require oxygen because, like mammals, they
chemical reactions have no activation energy. derive energy from the oxidation ofcarbon to
that is, there is no significant effect on rate of carbon dioxide. Over tens of millions of years,
increases or decreases in temperature. Even fungal systems have evolved that can degrade a
though this is true for most primary photoproc- wide variety of materials, some of them rather
esses, a number ofthe subsequent reactions toxic. There is, however, as yet no fungus that
will, in fact, be govemed by Arrhenius princi- has the en4rmes to break down most synthetic
ples. No one has yet determined how to de- polymers, which, by and large, do not occur in
scribe this situation quantitatively, for example, nafure and have been invented within the past
in terms of lifetime predictions, but it should be 60 years. Nevertheless, microbial growth can
kept in mind that the effects of light and heat to- readily occur on surface dirt of bioinert materi-
gether on polymers are more severe, sometimes als like plastics, causing, at the very least, unac-
synergistically so, than either separately. ceptable aesthetic consequences.

Moisture-Based Deterioration Fungi proliferate by sporulation and firngal


spores are literally everywhere, waiting for con-
It is self-evident that polymers that are highly ditions favorable for germination. Commercial
hydrophobic are unlikely to be susceptible to fungicides are available, but these are relatively
hydrolysis; there are no hydrolyzable groups toxic compounds unsuitable for many materials
present. Using polyolefins as an example, these and situations. Whether there is a need to pre-
plastics cannot hydrolyze, are highly resistant vent growth on a susceptible substrate or on the
to acids and bases, and are microbially inert. contaminated surface of an inert material, the
These characteristics give rise to applications in best way is to maintain "good housekeeping"
geotextiles, food packaging and automobile practices. In other words, it is usually sufficient
parts, for example. Some polyesters are more to keep things cool, clean and dry, in air condi-
hydrolyzable than others. PET, being relatively tioned premises if possible, in order to mini-
resistant, is highly useful as textile fibres, soft mize the undesirable effects of funsi and other
drink bottles and prosthetic devices, whereas a micro-organisms.
simpler polyester, such as poly(glycolic acid),
is an effective "biodegradable" suture material. Physical Aging
Chemical structure at the molecular level is the
critical factor and a very reasonable basis for As the temperature of a solid, amorphous
material selection. It is not surprising, there- polymer is raised, the kinetic energy of the
fore, that many naturally occurring polymers molecules will increase, but the resultant
are not soluble in water (or many other sol- vibrations and rotations will be significantly
vents) at ambient temperatures. Likewise, sus- restricted as long as the material retains its
ceptibility to reaction with (and destruction by) glass-like structure. At a specific temperature,
acids and bases is predictable on the basis of characteristic of each polymer type, there is a
polymer structure compared to that of analo- measurable change in the system where glass-
gous small molecules. like properties give way to rubber-like behav-
iour, owing to the onset of greater rotational
To simplify somewhat, biodegradation is com- freedom and more segmental motion (20 to
monly molecular fragmentation resulting from 50 chain atoms) of the polymer chains. This
the chemical effects of microbial enzymes. A temperature is called the glass transition tem-
very cornmon manifestation of this would in- perature or Tg. Since the properties of a rigid
volve one of the more than 80.000 kinds of glass and a rubbery plastic are very different,

109
the Tg of a polymer is one of its most important almost impossible to quantify on the basis of
characteristics. Values range from -102"C for any free-volume models that may be applied to
cis- 1 ,4-polybutadiene, -67oC for natural rubber, the aging ofplastics.
-20"C for polyethylene, 57"C for nylon 6,6,
69oC for PET, 8l'C for PVC, 100'C for poly- Polymer Stabilization
styrene, and up to 149"C for bisphenol-A poly-
carbonate. Thus, some common polymers are A very high proportion ofthe reactions that de-
below their Tgs at room temperature. More- grade polymers involve free radicals: reactive,
over, this is the case also for the amorphous neutral species that tend to be chemically self-
phase of semi-crystalline polymers and the rub- perpetuating in a hydrocarbon matrix. Two
bery matrix of filled rubbers and other compos- ways of approaching the requirement to stabi-
ites. Amorphous glassy solids are not in lize polymers are (a) reducing the rates of for-
thermodynamic equilibrium after solidification mation of these radicals, and (b) preventing
and thus, over a wide temperature range and their destructive reactions by deactivating them
for a very long time, will undergo characteristic first. It is self-evident that in selecting a mate-
property changes. This is called physical aging rial, recognition of the need for longevity in the
and is quite different than, and distinct from, intended use environment will be factored in
chemical aging, such as thermal degradation with mechanical, aesthetic, and cost criteria. In
and photo-oxidation. connection with approach (a), it is advisable to
avoid the formation of relatively labile chemi-
The temperature range over which physical cal groups or relatively weak bonds in a poly-
aging occurs can be quite broad and frequently mer during storage, handling, application or
includes the use-temperatures of common plas- fabrication. If the material is heat sensitive, for
tics. Physical aging can be explained qualita- example, the time that it is exposed to air at
tively on the basis of the free-volume concept high temperatures should be minimized; like-
so that, for example, the time dependence of wise, light-sensitive materials should be pro-
mechanical properties is found to be inde- tected from exposure to sunlight, light from arc
pendent of chemical structure. lndeed, creep lamps or from some fluorescent lamps. Longer
(stress relaxation) curves of numerous types of wavelength light can be a problem for coloured
polymers measured at various temperatures and materials. In the case of polymers that are to be
aging times can all be superimposed to form a exposed to near ultraviolet (UV) wavelengths
single master curve. (the erythemal, or sunburn region, 290 nm to
315 nm), UV-absorbing stabilizers are com-
The consequences ofphysical aging for the ma- monly incorporated as lowlevel additives.
terials specialist are twofold. On the one hand it Some pigments are IJV protective although
is a major phenomenon that determines the be- the anatase form of TiOz can act as a
haviour of a material to a large extent by chang- photosensitizer.
ing its relaxation times. The ability of rigid
plastics to withstand prolonged stresses, that is, Hydroperoxide groups fastened to polymer
the fact that they can be used as load-bearing chains represent a particularly insidious type of
materials for long periods, is not an inherent thermal and photochemical instability. Present
property but one that is developed by physical just at or below detection limits, hydroperox-
aging during the deformation period. On the ides can initiate the degradation of polymers
other hand, because it persists for a very long containing aliphatic and even aromatic carbon-
time, physical aging affects segmental mobility hydrogen bonds. Sulphur- and phosphorus-
of polymer molecules so it probably also af- containing compounds, as well as hindered
fects chemical degradation, photo-oxidation, amines and nickel chelates, are among the
swelling and de-swelling, in addition to me- highly efficient stabilizers that are added to
chanical relaxation phenomena. A knowledge polymers to protect them by decomposing ad-
of the aging behaviour of a plastic is indispensa- ventitious hydroperoxides to form more stable
ble in the prediction of its long-term properties compounds, before those hydroperoxides can
from short-term tests. Effects on crazing and en- initiate polymer degradation.
vironmental stress-cracking are important but

ll0
It is inevitable that covalent bonds will be bro- R6sum6
ken and radicals will be formed during the fabri-
cation, application and use of macromolecular La transformation des matiriaux polymiriques
materials. With regard to approach (b) men- avec le temps
tioned earlier, stabilization can be achieved by
trapping radicals. The most corrmon kinds of ) I'instar du papier, clu bois et des peaux, pra-
radicals that can be trapped (deactivated) are tiquement tous les plastiques, tolttes les.fibres,
alkyl and alkylperoxide, with the former being tous les caoutchoucs, toutes les peintures et
much more reactive and, in the presence of air, tous les rev€tements protecteurs doivent leurs
caractiristiques utiles d la masse, relativement
converting in what is usually a very fast reac-
tlevde (variant de quelques milliers d plusieurs
tion into the latter. It is very common to use sta-
mi I I i ons ), de I eur s mol t cu I e s. M a I heu reus emen t,
bilizing additives that deactivate alkyl radicals; la fabrication, la manutention et la bonne ou mau-
hindered phenols can protect thermoplastics vaise utilisation de ces matdriaux peuvent modi-
during fabrication (in the melt) as well as dur-
.fi", - voire habituellement abatsser - la masse
ing long-term exposwe to warm temperatures. de leurs moldcules constituantes, et attdnuer
Hindered amines (HALS) react with alkyl and de fagon concomitante leurs propri ttds utiles.
peroxide radicals catalytically, as well as de- L'ttude dans le ddtail des modifications qui se
composing hydroperoxides, and impart un- produisent d l'hchelle moltculaire demeure un
usually good stability to many polymers, secteur scientifique trds complexe, qui est souvent
including paints. dfficile d saisir. Il est toutefois certains aspects,
communs au vieillissement de plusieurs mat6-
A variety of other stabilizers is included in riaux poltmtiques, qui ont ttd dlucidds. et qui
polymer formulations to cope with specific peuvent ainsi out,ir la voie d une amdlioration
instability problems, for example, metal soaps des techniques utilisees pour leur conservation.
to neutralize the HCI generated in the dehy-
drochlorination of poly(vinyl chloride); anti- La digradation chimique de ces mattriaw s'ex-
plique notamment par la ddtbrioration o4tdative
ozonants to protect polymers containing
qui se produit lorsque, exposbs d I'air, ils entrent
carbon-carbon unsaturation, such as diene rub-
en contact avec la lumiire, la chaleur ou des
bers, ffom attack by ambient ozone. lndeed, it
rayonnements ionisants, mais aussi par leur
is noteworthy that synthetic polymers are never ddcomposition sous I'elfet de I'hydrolyse ou de
pure and may include additives of many kinds I'action d'acides ou de bases. La bioddgradation
for the modification of mechanical, swface, peut, par ailleurs, €tre considirbe comme un cos
aesthetic and chemical properties. Few of these particulier de ddcomposition chinique causte
will be significant in contributing to stability. par des enzymes microbiens. Les reactions qui
interviennent dans la dtgradation chimique sont
nombreuses et, dans une certaine mesure, particu-
Conclusion lidres d chaque matidre; ndanmoins, les plus im-
portantes peuvent Ji'2quemmeft ete rattachies
Energy, oxygen and time combine to change d l'oxydation d'hydrocarbures liquides. Pour
the characteristics of polymers. The rates of un- rdduire au minimum les phdnomdnes de decolo-
desirable changes can be reduced by selecting ration, defragilisation et de perte de propietes
the most durable materials initially, combining physiques qui accompagnent la ddtdrioration
o4vdative de ces matdiaux, il sffit d'utiliser un
these where possible with stabilizing com-
mdlange appropri4 d'additifs stabilisateurs. Et la
pounds, reducing the exposure to degradative
meilleure fagon d'attdnuer leur sensibilitd aux mi-
influences, and ensuring the least harmful use
crobes demeure la mise en pratique de bonnes
conditions and service environments. The effec- mithodes d'entretien.
tiveness of such approaches is maximized by
developing a comprehensive understanding of Le vieillissement de ces nrateriaux pollmrbriques
macromolecules and the chemistrv of them. qui, solidifids ou avant atteint autrement leur
fome d2finitive d'utilisation, continueront d se
dbgrader durant des semaines, des mois, voire
des annies, demeure ndanmoins un phdnomine
beaucoup moins bien connu. D'abord observd

lll
dans les themtoplastiques, au-dessotrs de la tem- 8. DurabiIi\, o.f Mao'omolecular Materials,
ptrature de tt'ansition vitreuse (Tv), le vieillisse- American Chemical Society Symposium Series
ment de ces mattriaur se produit parce que No.95. 1979.
toutes les proptibtds lides d la temperatw'e, qui
changent bntsquement d Tv, continuent ndan- 9. Long-term Properties of Polymers and
moins d changer au-dessous de cette temptra- Polym eric Materials, Applied Polymer
ture, mCme si ce n'est que trds lentement. Le Symposium No. 35, 1979.
vieillissement s'observe che: les poll'nr'ires amor-
phes, vitreux, chez les polvmdres semi-cristallins 10. McKellar, J.F. and N.S. Allen, Photochem-
en phase amorphe, de m,?me que dans la matrice istry of Man-made Polymers (London: Applied
caoutchouteuse des caoutchoucs chargtls. Avec le Science Publishers Ltd., 197 9).
tentps, nombre de matdriaux detiennent ainsi plus
rigides et plusfragiles, aufur et d mesure que di- I l. Allara, D.L. and W.L. Hawkins, eds.
ntinue le < taux de relaxation >. Comme le vieillis- Stabilization and Degradation of Polymers,
sement influe sur la mobilite segntentale des Advances in Chemistry Series No. 169,1976.
polymires, son action devrait aussi sefairc sentir
sur la dtgradation chimique, la photo-o4vdation, 12. Ultraviolet Light-induced Reactions in
le gonJlement et le d'lgonflement et les rtactions
ers, American Chemical Society
P ol ym

de rdticulation de toute une gamme de q,s727n"t Symposium Series No. 25,1976.


ntacromoltctilaires.
13. Wiles, D.M. "The Photodegradation of
Fibre-Forming Polymers," in Degradation
References and Stabilization of Polyvners, Chapter 7 ,
G. Geuskens, ed. (London: Applied Science
l. Carlsson D.J. and D.M. Wiles, "Degrada- Publishers, 1975).
tion," in: Encyclopedia of Polymer Science
and Engineering, Volume 4, 2nd edition (New 14. Rinby, B. and J.P. Rabek, Photodegrada-
York: John Wiley & Sons Inc., 1986) tion, Pho to-oxi dation and Ph otostabi lization of
pp.630-696. Polymers; Principles and Applications (New
York: Wiley-Interscience, I 975).
2. Grassie, N., ed. Developments in Polymer
Degradation, volume series, nos. I to 5 15. Road,8.E., P.E. Tomlins and G.D. Dean,
(London: Applied Science Publishers Ltd., "Physical Ageing and Short-term Creep in
1917, 1979, 1981, 1982, 1984). Amorphous and Semi-crystalline Polymers,"
Polymer, vol.7, July 1990, pp. 1204-1215.
3. Jellinek, H.H.G., ed. Degradation and
Stabiliz ation of P olymers (Amsterdam: 16. Bouda, V., "The Nature of Glassy State In-
Elsevier, 1983). stability," Po lymer Degradation and Stability,
vol.24,1989, p. 3 19.
4. Allen, N.S., ed. Degradation and Stqbiliza-
tion of Polltolefns (London: Applied Science 17.ChaL C.X. and N.G. McCrum, "Mechanism
Publishers Ltd., 1983). of Physical Aging in Crystalline Polymers,"
Polymer, vol.21, 1980, p.706.
5. Davis, A. and D. Sims, IAeafuering of
Polwers (London: Applied Science 18. Aklonis, J.J. and W.J. MacKnight,lntroduc-
Publishers, Ltd., 1983). tion to Polymer Viscoelastrciry (New York:
John Wiley and Sons, 1983).
6. Moiseev, V.V. and G.E. Zakov, Chemical
Resistance of Polymers in Aggressive Media 19. Ferry, J.D.,Viscoelastic Properties of
(New York: Plenum Press, 1982). Polymers, 3rd edition (New York: John Wiley
and Sons, 1980).
7. Pappas, S.P. and F.H. Winslow, eds.
P ho t od egr ada ti o n and Ph oto s t abi liz ati on of 20. Stuik, L.C.E. Physical Aging in Amorphous
C o at ings, American Chemical Society Polymers and Other Materials (New York:
Symposium Series No. I5 l, 198 I .
Elsevier.1978).

tt2
The Physical Aging of Polymeric Materials

Chris topher W. Mc Gl inc hey

P aintings Conservation
TheMetropolitan Museum of Art
New York, N.Y.
U.S.A.

Abstract this phenomenon, and make superior products to


the earlier ones comprised of identical materials.
Ph1'sical aging of polymers can be defined as the Distinguishing between dffirences in chemical
changes in spatial arrangement ofmacromole- and physical aging w,ill enable conservators to
cules and side chains w,ith respect to one another. think more clearllt about theirfundamental ap-
This molecular reorganization can cause several proach to the complex aging process of
observable changes. Thermal properties, such as polltmeric materials.
melting point, glass transition and crvstallization,
can become modified, possibly transforming the Introduction
solubilin and optical characteristics in the proc-
ess. More tactile qualities, such asflexibilit-v, em- The elusive property that made macromole-
brittlernent and drape, are also changes brought cules a controversial subject fiom the time of
on bv the process ofphysical aging. their discovery until as recently as the 1920s is
their "polymeric" nature. It was precisely this
Polymer morpholog,,, the study of polymer order, property that prevented macromolecules from
depends upon the chemistry of the polymer, its
being identified using existing methods of frac-
structure and the thermal history to n'hich it has
tionation, purifi cation and crystallization. Re-
been exposed. Though the tetms amorphous and
sults flom these methods appeared to be in
crvstalline are often applied to polvnters, it is
more accurate to think of these terms as extremes direct conflict with results based on diffusion
on a spectrum and to consider that polvmers, de- and viscosity measurements that gave inordi
pending upon their stntcture, have the potential nately high molecular weights. The methods
of being crystallized (achieving a higher state of that failed to properly detect polymers were the
order) or quenched (achieving greater disorder traditional methods that had been successfully
or becoming more amorphous). Physical aging applied for centuries in the characterization of
can have signifcant effects on the aforesaid prop- inorganic and lower molecular weight organic
erties, but unlike chemical degradation, phvsical materials. Early distillations of macromolecular
aging would be completely reversible were it not materials resulted in the decomposition of the
fbr the complication that occurs when chemical parent material into fractions of low molecular
degradation alters the relationship between tem- weight. Neglecting the possibility of degrada-
peratures of degradation and melting point. tion, one attempt to solve this paradigm was
through the concept ofcolloidal forces. These
In the early stages of commercial polymers, colloidal forces, though undetectable, were
physical agingwas poorlv understood and not
thought to reside in the residue of the distilla-
kn ow i ng l1t .fac t o re d i nt o
for mu I a ti o ns. M o re r e - tion process. Two of the prominent scientists
cent polvmer scientists now carefully consider

113
that fi nally settled the "polymer""controversy dependent upon the primary bond forces. Al-
were Staudinger' and Carothers.' They showed though secondary bonds are relatively weak
through slmthesis that macromolecules consist compared to covalent bonds, they can have a
of many (poly) smaller repeat groups (mers) significant cumulative effect on the macro-
joined by conventional covalent bonds. scopic properties of molecules.

These and other historic proofs that led to It is possible that changes brought on by the
the acceptance of polymers soon gave way to altered physical properties of the polymer can
their full scale development and specialization. be reversed in-situ, while those derived from
Polymer materials rapidly began to be reliably chemical degradation cannot. Therefore it can
processed into products and finishes with a be advantageous to determine the source ofob-
wide variety of uses (e.g., automobile tires, servable changes. For example, the yellowing
plexiglass, traffic stripes and vinyl upholstery). of organic materials is obviously due to the
This was largely due to the homogeneity of cer- "irreversible" chemical production of chromo-
tain properties, which assure reproducible acti- phores. But graying, embrittlement and de-
vation conditions, melting point and solubility creased solubility are due to alterations in a
for different batches of the same material. The material's physical properties. Modified physi-
most significant properties of the polymer in- cal properties that result fiom chemical degrada-
clude chemical composition, molecular weight, tion cannot be erased; they merely become the
molecular weight distribution and branching. In new physical properties ofthe particular sys-
addition, slmthetic polymers are more likely to tem. On the other hand, the components of
be free of unstable impurities comlnon to many physical changes not initiated by chemical deg-
nafural polymers, which can initiate degrada- radation are theoretically reversible. It is evi-
tive chemical reactions. However, while these dent that the glass transition temperature (Tg)
new products were likely to have extended life- increases as oxidative degradation proceeds due
times, improperly selected processing condi- to an increase in polar secondary forces. How-
tions may have occasionally shortened their ever, it is also possible for a material to remain
life spans. Polymers do not necessarily possess chemically unchanged while showing an in-
intrinsic properties whose qualities are immedi- crease in Tg through physical aging.
ately apparent. Selection of processing condi-
tions can either bring out desired properties The purpose of this paper is to discuss and illus-
-
transparency, strength and dimensional stability trate the reversible and non-reversible changes
or result in a product that exhibits hastened that occur for thermoplastic polymers after
-embrittlement, warpage or cloudiness. processing and the passage of time. It is hoped
that this information will enlighten conserva-
In order to study the changes in synthetic poly- tors' understanding of these materials, whether
mers upon aging it is necessary to define the they are for their own use or as the medium of
differences between chemical and physical the artist or a previous conservator. This will
properties. Chemical degradation processes are enable conservators to not only choose their
related to changes in the covalent bond struc- treatments more judiciously but also assess the
ture. A covalent bond is defined as the strong aging mechanisms occurring within arfwork
attractive force derived from the sharing of containing such polymers.
two elecffons between two atoms. They are
graphically represented as the lines between Polymer Physical Properties
monomeric units, as well as between the atoms
within each monomer. Physical properties, in Thermoplastic polymers are by definition
addition to being dependent on molecular materials that can be cycled from the solid state
weight, molecular weight distribution and to the fluid state a number of times without a
branching of polymers, also depend upon secon- change in their molecular composition; this
dary bond forces. Secondary forces are attrac- trait is characteristic oftheir physical proper-
tive and repulsive forces that arise from the ties. The term thermoplastic is intended to dis-
local electron densities that surround the tinguish these materials from those that require
atoms in molecules, and are therefore in part chemical reactivity or radiation to cure them

tt4
into an ineversible polymerized state. Thermo- scanning calorimeter (DSC) measurement for
plastic polymers, while stable, are not necessar- the Tg, corresponds to the differences betw.een
ily exempt from thermo- and photo-chemically heat capacity of the liquid and glass states.'Of
induced degradation. Given this finite chemical crystalline polyrners, those that are quenched
stability, the physical changes derived from will show a more intense Tg than those an-
these chemical changes are therefore non- nealed because the phenomenon is derived
reversible. from non-cry.stalline conversely amorphous
domains. ' Below-the Tg the polymer ma-
Thermoplastic polymers have to somehow be -trix is sufficiently restricted in movement to
transferred from their raw material form into prevent further crystallization. The physical
a final shape or finish. This transformation re- processes that take place beneath the Tg in
quires energy that is supplied in the form of amorphous domains are termed physical aging.
heat, pressure and/or solvent action. Whether Physical aging occurs between the glass transi-
polymers in this transient state are in the form tion, Tg, and the next lower secondary transi-
of polymer solutions or are heated above their tion defined as Tg .o Secondary transitions are
melting point (Tm) they are by definition amor- denoted T", Tp, Ty, etc.; T6, corresponds to the
phous. This is because polymers are not locked Tg, while lower transitions maintain their
within an ordered mafrix; neighboring mole- Greek symbols and are due to weaker molecu-
cules randomly exchange their position. Rheol- lar motions at lower temperatures.
ogy, the study of the flow and deformation of
matter, is frequently used to study the response Thermal Processing
of polymer melts and solutions to stresses im- It is important for thermally processed poly-
posed by processing forces. Instrumentation for mers to be stable at the necessary processing
these measurements include viscometers, dy- temperature. If thermo-oxidative degradation is
namic mechanical analyzers, calorimeters, melt expected to occur, antioxidants in small quanti-
rheometers and melt indexers. ties can be introduced to extend the effective
processing time. Common methods of process-
The crystallization of ther^moplastics occurs ing include injection molding, film and fiber
only between Tg and Tm.r The conformation extrusion, calendaring and casting. These opera-
changes required for the transition to a crystal- tions are discussed in greater detail elsewhere.'
line state do not occur instantaneously. Time is Solvents are sometimes added to aid process-
required for polymer chains, side chains and ing. Afterwards, they are usually intended to
functional groups to organize into a three-di- evaporate out but sometimes small amounts re-
mensional order. Annealing (heating a material main as plasticizers (e.g., water in certain nylon
between its Tg and Tm) combined with a slow processes). In processing, polymer choice de-
rate of cooling will optimize the amount of or- pends upon both the intended use ofthe prod-
ganization for a given polymer. Conversely, uct and the actual processing method. There is
quenching (rapidly cooling llom above the no polymer that is ideally suited for all process-
Tm to below the Tg) of a polymer will mini- ing methods. Some polymers (e.g., polyethyl-
mize the crystallinity of its molecular structure. ene) can be easily adapted for different
Thus, it is possible to achieve a range of confor- processing methods, but usually at some com-
mations or morphologies for a polymer by promise, for example, mechanical properties at
controlling the dwation of the cooling period. the expense of optical properties or vice versa.
Annealing can also be govemed by processing
variables, such as pressure, temperature and In thermal processing, crystalline polymers
solvent evaporation. tend to be tumed into fibers and films while
amorphous materials lend themselves well as
Below the Tg, molecular motion is drastically molded objects that need good clarity and iso-
reduced. (Note: Tg is not as precise as melting tropic strength. Synthetic fibers are processed
point; the transition occurs over a temperature from crystalline polymers so that when they are
range.) This reduced molecular motion corre- drawn, their crystalline domains are oriented
sponds to a reduction in heat capacity the inten- along the hber axis (machine direction)
sity, or shift in baseline of the differential

ll5
resulting in their higher strength in that direc- cracking and crazingwithin coatings' derived
tion. For example, nylon fiber after drawing stresses imparted from drying processes, a laser
(extending its length by pulling at a rate faster interferometry technique has been dev,e^loped to
than that at which it comes out of the die exit) make these stress patlems observable.''
has a greater strength in the machine direction.
Additionally, polyethylene pellets are opaque, The preparation of polymer thermoplastic coat-
but after processing, films and fibers can be- ings can include a combination of heat, solvent
come more transparent due to orientation of and pressure. Usually, the less soluble these ma-
crystalline domains previously randomly dis- terials are, the greater their crystallinity. Sol-
tributed. Stress whitening is a result of orient- vent cast films high in crystalline content tend
ing in one direction to such an extent that the to be dwable and impervious to attack by or-
polymer develops hne voids dging processing ganic solvents when dry. (This is the benefit
that cause extensive scattering." gained from materials that require additional en-
ergy to formulate.) Thermoplastic amorphous
Processing forces, in conjunction with the resins with a low Tg are often used as compo-
cooling effect of the mold, transfer significant nents to pressure-sensitive adhesive coatings
stresses into the polymer. Stresses that are because their soft, amorphous domains reduce
formed in pattems are obseryable by the tech- the surface tension of adhesive formulations to
nique of flow bireftingence." If these pattems improve their flow, or tackiness. Amorphous
are not allowed to relax sufficiently they will coatings of higher Tg are commonly found in
become frozen within the material. These fro- over-the-counter applications because of their
zen stress patterns may relieve themselves shelf stability, solubility in mild solvents and
through the process ofphysical aging, or be ease of application. Hard amorphous materials
coaxed along by heating. A significant effort are also those most often used in conservation
has been spent in developing processing opera- practices because they are most likely to remain
tions that ease these stresses prior to being soluble in solvents similar to the ones in which
frozen into materials. they were originally dissolved.

Physical Considerations of Amorphous polymers are easily dissolved be-


Solvent-Borne Polymers cause there is no three-dimensional order that
Those formulating coatings have developed a must be broken up prior to entering the amor-
wide range of solvents and additives to accom- phous liquid state. The lack of order in amor-
modate the many specific end-use requirements phous substances permits more voids into
in the coatings industry. A host of additives which solvent molecules can penetrate, leading
have been developed to prevent almost every to the switching of polymer-polymer interac-
defect known to coatings; some examples are tions to solvent-polymer interactions. These
additives to avert floating, flooding, flattening, voids, or unoccupied spaces between atoms and
fouling, skinning and settling. Rheological molecules, are often referred to as the free vol-
modifiers are also added to improve a coating's ume of the system. Crystalline materials, on the
handling properties during application. While a other hand, require additional energy to break
solvent must have the appropriate solubility up the ordered structure that results from rela-
characteristics to make a proper paint, the rate tively strong polymer-polymer interactions and
at which it evaporates plays a significant role in generally a lower free volume compared to
how the surface sets up to dry. The evaporation amorphous substances.
t'
rate ofa solvent from a paint is different from
the evaporation rate ofthe pure solvent and var- Free volume permits solvent molecules to
ies from polymer to polymer. The surface ten- penetrate and to replace polymer-polymer
sion of the solution (which depends partlqlly interactions with polymer-solvent interactions.
upon the solvent) controls film leveling.'" Un- Since free volume increases with temperature,
suitable surface tension is responsible for such solubility of both amorphous and crystalline
flaws in pppearance as "orange peel" and "cra- materials is increased by heating.
tering."" In order to help study the checking,

l16
While solvent evaporation cannot be strictly lik-
ened to the cooling of polymer melt, some par-
allels do exist; these depend upon the rate of 116
LDPE
\
solvent release and the ability of the polymer
tE \
resin to crystallize. It has been shown that mo- 7a
I

lecular relaxation processes have occurred for I


annealed organic glasses that are the result of =

solvent removal; fu rthermore, these relaxation 375

processes are solvent dependent.'' For coatings


with crystalloid resins the rate of solvent loss
may have a great bearing on the final properties 0
60@ 70@ 80@
TemFatu6 (c)
of the surface coating. This is because the time
required for crystallinity to develop will be
Figure 1 DSC cooling curves for LDPE and
based upon the free volume available per rurit
h-PDCPD
of time; a fast evaporator will cause the free 3

volume to decrease at a fast rate, thereby re-


stricting motions that would be allowable in a
l

slow er evaporating solution. The quickly dried


sample, with its lower crystallinity, will be sub- E
I
I

sequently more soluble. If the Tg of the amor- a


E I
I

h.PDCPD
phous component is low enough, physical
aging processes mentioned above may also re- ;
sult. Latent crystallization can only occur for
materials exposed to temperafures between
their Tg and Tm for significant periods of time.
80@ r@@ 1&00
Results and Discussion
Degree of Crystallinity for Hydrocarbons Figure 2 DSCheating #)i', roru rro
h-PDCPD.
The range of crystallinity can be illustrated
by comparing the DSC cooling curves of two
The effect of processing forces on the thermal
aliphatic hydrocarbon solids, such as low den-
properties of LDPE is determined by analyzing
sity polyethylene (LDPE) and hydrogenated
the DSC curve from the first heating cycle of
poly(dicyclo pentadiene) (h-PDCPD), as shown
the injection molded resin. In this instance,
in Figure l LDPE is a long linear polymer
frozen stresses are derived fiom the imposed
with some branching and h-PDCPD is a cyclic
quench cooling ofthe high-pressure injecting
bridged structure. LDPE crystallizes at 89'C
of molten polymer into a mold much cooler
while the h-PDCPD does not crystallize. This
than the plastic. As a result it does not have the
is to be expected because the highly branched
opportunity to achieve the optimum degree of
structure of the h-PDCPD prevents crystal-
crystallinity. The DSC scan of the injection-
lization. Secondly, DSC heating curves show
molded sample shows a Tg of 4l'C and a Tm
that DCPD has a Tg at 49"C while LDPE has a
of l05oC, which suggests the presence ofboth
Tm of 106'C (Figure 2). Notice that the pro-
amorphous and crystalline components respec-
gression for thermal transition for LDPE is
tively @igure 3). After the sample is annealed
Tg<Tc <Tm, where Tc is the crystallization
at 80oC, prior to the DSC heat, the diminished
temperature. It is clear that the morphology of
Tg is the result of enhanced crystallization.
LDPE may depend upon thermal history where-
as h-PDCPD remains amorphous under the
Thin sections viewed in polarized light
same conditions. The crystalline nature of
show a radial orientation orthogonal to the
LDPE makes it more difficult to dissolve com-
picture plane (Figure 4). Annealed radial and
pared to h-PDCPD, which is more soluble and
tangential samples cut from this circular cross-
hence appropriate as a solvent cast vamish.
section show the stress-release mechanism:

Itl
cooling results (not shown) are similar to the
h-PDCPD sample. However, first heat DSCs of
a 34-year-old PVAC film, which was mechani-
11
1/l
cally removed fiom a panel painting, indicate
tg il either the release of mechanical stresses in
'il
ltl
amorphous domains or the melting of crystal-
9
I
line domains both in the range of 37'C to 44oC
I
1L (Figure 6). Normally the first heat is discarded
2
L\
I
because aging history makes it complicated and
l
rarely reproducible. This is certainly true here
l because the peak in Figure 6 has taken as long
as 34 years to develop. The fact that this is not
reproducible does not make this result any
Figure 3 DSC heating curves for injection-molded less valuable. It is simply demonstrative of the
and annealed LDPE. aging history of this particular sample. This
phenomenon must take a long time to form be-
cause, as can be seen fiom the annealed sample,
it was not possible to regain this apparent Tm
within the time frame of the DSC experiment.

After heating this sample to 45oC for five min-


utes and then reducing the temperature to 25oC,
visible light transmitklnce. measured spectro-
scopically, increased permanently by 7o/o.lt is
therefore most likely that the opacification after

Figure 4 Thin section cut from injection-molded


LDPE rod, viewed in crossed polars. 350C
unsymmetrical stress distributions in the tangen-
tial sample cause out-of-plane warpage while
the radial cut sample does not (Figure 5). This
warpage is similar to planks of wood cut with 450C
the same orientation, although the derivation
of these stresses is radically different. For com-
plex shapes of molded plastics frozen stresses
can be equally complicated. For early plastics 55nC
(whose color was a mixture of two different
pigmented resins, such as red and white poly-
styrene) the incomplete mixing of these colors
leads to a grain pattern in the product. It is pos-
sible that frozen stresses can follow this srain.
650C
This pattem, while visible. is not n"."ttuiily d.-
monstrative of fiozen stresses; the molding con-
ditions may have eased these stresses by using
a mild cooling temperature for the mold and,/or
750C
a lower injection pressure.

Effect of Thermal History on Varnishes


The amorphicity of poly(vinyl acetate) (PVAC) Figure 5 Heat-treated sample tangentially cutfrom
is illustrated by DSC cooling curves. The DSC LDPE cross-section.

118
resin and in the remaining two pairs the appear-
ance in effective solubility was about the same
for both samples. These results suggest the
main change in solubility is that the rate at
6
€ which solvent is taken up varies significantly.
B
-5 ..'-4sxeat Given enough time, however, eventual solubil-
I: ity is approximately the same.

This is not to suggest that all the changes in


this film are derived from changes in purely
physical properties. By comparing the IR spec-
tra of this 34-year-old film to fresh PVAC, it is
evident that some degree of chemical degrada-
Figure 6 DSC traces 34-year-old PVACfilm.
tion has taken place (Figure 7). The increase in
polarity from oxidation products chemically
34 years is from the formation of crystalline do-
formed during aging is likely to require a sol-
mains that are large enough to scatter light and
vent ofenhanced polarity to re-dissolve the
that the increase in transparency upon heating
is due to the conversion of crystalline domains
film. It is therefore apparent thatthese materials
change as a result of both chemical and physi-
back into amorphous ones. PVACs, though not
cal properties. Additionally, the change in
normally crystalloid, apparently are to a small
physical properties, the result from the chemi-
degree when given enough time, in this case
cal changes illustrated in the IR spectra, impart
34 years.
their own effect on the physical properties that
are not reversible. This is opposed to the por-
While long-term induced crystallization is cer-
tion ofphysical changes represented by the
tainly responsible for a portion ofthe increased
DSC results that are reversible.
opacification or grayness of these films it is
likely not to be the only reason. The denser
structure associated with the increase in crystal-
linity has (by definition) contracted during the
aging process. Ifsufficient, this contraction
may contribute to the lack of adhesion at the
paint/varnish interface and furthermore may
permit diffirsing air to accumulate in this re-
gion, thereby increasing scattering even more.
The fact that these films are so easily peeled off
of paintings lends credence in some instances
to this hypothesis. It is also possible that soft
resins allow more airbome particles to stick to
their surfaces. Figure 7 Infrared spectra ofPVACs cast on salt
plate.
Once the proportion of crystallinity has been
reduced within the material through heating, Conclusion
solubility properties change. After a five min-
ute exposure to xylene, the post treatment sam- The methods by which thermoplastic polymers
ples imbibe more solvent and coalesce more are made into their final product is as varied as
extensively. After measwing the solvent uptake the class of thermoplastics themselves. While
for five pairs of heated and unheated samples, polymers have intrinsic properties that cannot
on average the preheated sample imbibed be hidden, such observable properties as
0.51 pl and the unffeated film took 0.1I pl in strength, clarity and solubility are contingent
a five minute period. After two hours of expo- upon processing forces. It is understood that
sure the results were less reproducible: in three solvents, temperature and pressure can affect
cases the heated sample cleared away more these observable properties. Stress-release

t19
mechanisms result from frozen stresses of im- stick with wax, were suspended perpendicular
properly molded objects. The process under to the microscope's stage. This assembly was
which these stresses are relieved depend upon attached to a plate of adjustable height on either
polymer structure and ambient exposure condi- side of the stage. The vamish films were placed
tions. After sufficient aging PVAC develops beneath these pipettes and weighed down with
crystalline domains that are large enough to small washers. Prior to filling with solvent, the
scatter light and modify solubility. Addition- pipettes were adjusted to ensr.ue simultaneous
ally, the risks involved in the removal of old contact with each surface. Samples were then
synthetic restorations can possibly be induced rinsed with an aliphatic solvent and allowed to
by warming them to moderate temperatures dry ovemight and then photographed.
(e.g., 40'C) in order to reduce the time required
for their solubility. Visible light transmittance measurements were
calculated at 550 nm using a Perkin-Elmer
Finally, with the ushering in of a class of lambda anay UV/VIS spectrophotometer.
chemically stable synthetic products, yellowing
and other phenomena associated with oxidative Acknowledgement
degradation are no longer the only degradation
processes that divert artistic works from their The author thanks John M. Brealey and Hubert
original meaning. Opacifi cation, scattering and von Sonnenbwg for creating and maintaining
(when appropriate) warpage are changes that the interdisciplinary environment of the Sher-
may or may not be easily reversible. It is our re- man Fairchild Department of Paintings Conser-
sponsibility to leam more about these changes, vation; Jeflrey Jennings and Marie L. O'Shea
if not for the conservator, for the sake of future for providing valuable comments regarding the
generations so they don't romanticize about the subject matter of this paper; and Marycolette
"gorgeous grays" of the 20th century. Hruskocy for providing assistance during the
investigative stages of this paper.
Experimental
R6sum6
The Tg, Tc and Tm of polymers were deter-
mined using a liquid nitrogen cooled differen- Le vieillis sement des matiriaux polyvniriques
tial scanning calorimeter, model DSC 7
(Perkin-Elmer). Temperatures and specific Le vieillissement d'un polymdre peut se ddjinir
heats of transition were calibrated usins indium comme le rdarrangement spatial des macro-
and zinc standards. moldcules et chaines lattrales qui le composent.
Il s'accompagne pafois de plusieurs change-
ments observables. Ses propidtds thermiques
point defusion, sa transitionvitreuse et sa-
Low density polyethylene (Rexene 143,
son
Dupont) was injection molded (at 400'C under
a pressure of 9.65 MPa using a water cooled
cristallinild, notamment
- pourront ainsi se
modifier, tandis que sa solubilitd et ses cqract4ris-
Van Dorn injection molding machine) into the tiques optiques varieront d'autant. Le vieillisse-
form of a rod with a diameter of 5 mm. Plates ment suscite en outre une transformation des
I mm thick were cut from the rod, tangential proprittds plus tactiles du polymdre, telles sa
and radial samples were cut from this plate (Fig-
Jlexibilit4, safriabiliti et son drapt.
ure 1). These samples were placed on a heat-
able stage, Metratherm model 1200d (BBC), La morphologie du polymdre l'6tude de la dis-
and observed through a stereoscope (Zeiss). position des macromolicules -les unes par rapport
Approximate temperatures were determined by aux autres est lide d sa chimie, d sa structure
correcting them against the observed boiling
-
et aux tempdratures auxquelles il a dte expost. Si
point of water. Stress-release mechanisms were I'on qualifie soavent les polymires de cristallins
photographed using a Polaroid c€rmera attach- ou d'amorphes, il serait pourtant plusjuste de
ment on the stereoscope. considirer qu'il ne s'agit ld que de deux ex-
ffAmes, que les polym'ires peuvent, selon leur
Solubility tests were as follows. A pair of structure, devenir plus cristallins lorsque leurs
-
macromoldcules prdsentent, les unes par rapport
2 microliter pipettes, mormted on a wooden

120
aux autres, une disposition rdgulidre ou plus 5. Ferry, J.D., Viscoelastic Properties of Pollt-
amorphes lorsqu'elles offrent une disposition mers,3rd edn. (New York Wiley, 1980) p. 280.
-
moins rtguliire. Le ieillissement peut avoir des
effets importants sur les propi2tds hnumdrdes 6. Struik, L.C.E., "Volume Relaxation in
plus haut mais contrairement d ce qui se pro-
-
duit dans le cas d'une degradation chimique
Polymers," Rheologica Acta, vol. 5, 1966,
-
c es effets demeureraient entidrement reversibles
pp. 303-3 1 1.
si ce n'6tait de cette complication qui survient
quand une degradation chimique altdre le rap-
7.Han,C.D., Rheology in Polwer Processing
(New York: Academic Press, 1976).
port entre les temptratures de dtgradation et le
point de finion.
8. Woodward, A.E., Atlas of Polyner Morphol-
A t'Apoque des premiers polymdres commercianx, ogr (Munich: Hanser Pub., 1989) p.363.
on comprmait mal le vieillissement, et on n'en
tenait guire compte dans I'ilaboration des prdpa- 9. Meeten, G.H., ed., Optical Properties of
rations. Aujourd'hui, les sptcialistes des po- Polwers (London: Elsevier, 1986) p. 101.
lymires prennent soigneusement en compte ce
facteur, si bien que I'on obtient, d partir des 10. Keunigs, R. and D.W. Bousfield, "Analysis
m€mes matidres, des produits suptieurs aux pre- of Surface Tension Driven Leveling in
miers. Forts d'une telle distinction entre la ddgra- Viscoelastic Films," J. Non-Newtonian Fluid
dation chimique et le vieillissement, les Mechanics, vol. 22, I 987, pp. 219-233.
spdcialistes de la restauration seront en mesure
de dtfnir plus clairement leur approche fonda- I l. Quach, A., "Polymer Coatings: Physics
mentaleface d ce processus complexe que consti-
and Mechanics of Leveling," Industrial and
tue le vieillissement des matdriaux polymtiques.
Engineering Chemistry Product Research and
Development,vol. 12,no.2,1973, pp. I l0-l16.
References
12. O'Brien, R.N. and W. Michalik, "Laser
l.Staudinger, H., "The Constitution of
Interferometry for Internal Stress Measurement
Polyoxymethylenes and Other High Molecular
in Coatings," Journal of Coatings Technologt,
Compounds," Helvetia Chimica Acta, 1925,
vol. 58, no. 735, 1986, pp. 25-30.
pp.67-70.
I 3. Morawetz, H., Macromolecules in Solu-
2. Carothers, W.H., "An Introduction to the
tions,2nd edn. (New York: Wiley, 1974).
General Theory of Condensation Polymers,"
Journal of the American Chemical Society,
14. Petrie. S.E.B.. "Thermal Behavior of
vol. 5, no. 1, 1929, pp. 2548-2559.
Annealed Organic Glasses," Journal of
Polwer Science Part A-2, vol. 10,1972,
3. Hall, C., Polymer Materials,2nd edn. (New
pp.1255-1272.
York: Wiley Interscience, 1989) p. 47 .

4. Twi, 8., Thermal Characterization of


P olymericMaterials (Orlando: Academic
Press, 1981) p.167.

t2r
La pr6vision du comportement ir long terme de mat6riaux polymires
synth6tiques doapris des exp6riences de vieillissement artificiel

Jacques Lemaire

Laboratoire de photochimie
Centre national d'|valuation de photoprotection
Univers it6 B lais e-Pas cal
Clermont-Ferrand
France

R6sum6 macroscopiques (propridtds mtcaniques, opacitd,


aspect en surface, etc.). Le vieillissement acctldr!
Nous ddcrirons le photovieillissement des se confond ainsi, enfait, avec une oxydation ac-
p o \,rn'ir e s sy th dti qu es s e re tro uvan t coura m- cdl6rde, et lefacteur d'accilhration peut €tre
ment dans des sculptures modernes qui sont ex- etubh d partir de la vitesse de laformation, dans
posdes dans les musbes extirieurs. Les polyesters des conditions artificielles et naturelles, des pro-
insaturds rdticults ainsi que divers polyacrylates duits d'oxydation critiques. C'est donc dire qu'il
et polymtthacrylates seryiront d'exemples de ma- est possible de prdvoir la duree de vie d'un po-
trices non stabilisdes dont les propi2tds peuvent lymire en sefondant sur les mesures de durde de
€tre modifiees au cours de la polymdrisation, par vie obtenues dans des conditions artifcielles et
sdchage d partir de la solution ou de I'tmulsion en prenant alors en compte lefacteur d'accdl6ra-
et au moyen d'addilifs. Nous ddcrirons dgalement tion. Cette approche mdcanistique se compare
la photo-ox.vdation et I'oxydation thermique bien aux techniques empiriques fonddes sur la
d' dlas tomdres diiniques (butadidne, styrdne- simulation des conditions physiques et chimiques
butadidne et c aoutchouc de ni trile-butadidne), rdelles et sur les variations de proprietes phy-
car on ne saltrait Ahborer une strattgie de siques macroscopiques qui en rtsultent.
stabilisation sans d'abord comprendre le
mdcanis me d' oxvdation. Pour la plupart des polymdres synthttiques, la
degradation oxydative, qui tient soit d la struc-
On peut avoir recours, pour tenter de prd,,toir ture normale du monomdre polymdrise, soit d des
la vitesse du vieillissement des polymdres dtfauts chimiques, s'obserue d un tris foible de-
synthdtiques qui ont 6tt utilists, d I'approche grd d'oxydation. Ilfaut d'is lors avoir recours d
mdcanistique. Aux termes d'une telle approche, des techniques spectrophotomdtriques in situ trds
les polymires solides sont considirds comme des sensibles et informatives pour ddterminer la na-
< rdacteurs > chimiques, c'est-d-dire comme des ture des divers intermediaires et produits finals.
milieux qui sont le site de certaines riactions
chimiques. Elle suppose l'identification des Introduction
mdcanismes qui se produisent, d I'echelle mol6-
culaire, dans des conditions de vieillissement Les mat6riaux polymdres, naturels ou syn-
acc d I drt, artifi ci e I les, et d' uti I isation normale, th6tiques, que I'on utilise soit pour la conserya-
naturelles. Les effets les plus ndfastes proviennent tion et la restauration des cuwes d'art, soit
de I'oxydation causde par une tropforte exposi- pour leur 6laboration, pr6sentent I'inconv6nient
tion d la lumidre, d la chaleur ou d d'autres
majeur d'6voluer dans le temps, avec une perte
contraintes statiques et dvnamiques. Et une
progressive de leurs propri6t6s physiques et des
analyse chimique poussde permettra d' expl iquer
modifications d'aspect. Les ph6nomdnes de
les modifications de propri6tds physiques

123
vieillissement, qui sont reconnus comme com- chimiques (oxyd6s, par exemple) et cette
plexes depuis fort longtemps, ne font en fait 6volution chimique est responsable de la
I'objet d'analyses que depuis vingt ans, leur d6gradation des propri6t6s physiques. Le
6tude ayant 6t6 largement diff6r6e d la suite des vieillissement physique des polymdres n'in-
premiers succds obtenus gr6ce ir des m6thodes tervient que dans le cas des mat6riaux qui
trds empiriques. sont soumis d des contraintes physiques trds
importantes, et il n'a donc aucune importance
Les premiers probldmes de fiabilit6 des dans le cas des auwes d'art.
mat6riaux polymdres en cours d'usage se
sont en effet pos6s dds les ann6es 1945-1950 o L'6volution chimique ne d6pend pas
aux m6caniciens qui utilisaient des pidces en
des contraintes m6caniques qui sont appli-
plastique. Ces m6caniciens n'ont alors con-
qu6es; ces contraintes ne modifient que
sid6r6 que les aspects macroscopiques du vieil-
les cons6quences physiques de l'6volu-
lissement des systdmes, et ils n'ont 6tudi6 que
tion chimique.
les variations des propri6t6s physiques utiles en
appliquant d ces systdmes, en laboratoire, des
contraintes physiques et chimiques repr6senta-
. L'acc6l6ration de l'6volution chimique ne
tives de celles auxquelles ils sont soumis en d6pend donc pas des contraintes m6caniques
cours d'usage. Par exemple, le vieillissement externes et intemes. L'analyse de l'6volution
en conditions climatiques 6tait examin6 dans chimique permet de convertir la dur6e d'ex-
des enceintes de < simulation > qui reprodui- p6riences en laboratoire en dur6e d'usage
saient les contraintes naturelles de I'environne- dans des conditions naturelles (l'6volution
ment (la lumidre, la chalew, 1'Oz, l'HzO, etc.). chimique constituant dds lors la base de tout
De telles m6thodes ne permettaient de pr6voir transfert de donn6es).
le comportement d long terme des polymdres
qu'en 6tablissant une < corr6lation > avec les o L'6volution chimique est une caract6ristique
situations r6elles de vieillissement. Ces de chaque m6canisme d'6volution d'un
m6thodes de contr6le plutdt que de pr6vi- mat6riau donn6. Une formulation pr6ci-
sion - gudre 6volu6 depuis 1950; leur
n'ont se (polymdre + charges + pigments ou colo-
- s'est par contre beaucoup d6velopp6, et
emploi rants + additifs) doit 6tre caract6ris6e par un
elles ont foumi de nombreux resultats que I'on facteur d'acc6l6ration qui lui est propre. On
n'arrive pas d ( rationaliser >. ne peut transferer directement un classement
de divers mat6riaux ou de diverses formula-
A partir des ann6es 70, une approche plus tions obtenu en laboratoire aux conditions du
cognitive du comportement ?r long terme des
terrain sans tenir compte de facteurs d'ac-
mat6riaux polymdres s'est d6velopp6e, qui se
c6l6ration n6cessairement diff6rents.
fondait sur des analyses, au niveau mol6culaire,
des modifications chimiques des chaines macro-
o L'acc6l6ration des 6volutions chimiques est
mol6culaires qui apparaissent en cours d'usage.
Cette approche a surtout 6t6 utilis6e pour les non seulement autoris6e, mais elle est fonda-
polymdres synth6tiques qui, conhairement mentalement indispensable car :

aux mat6riaux organiques naturels, pr6sentent


un ordre au niveau mol6culaire, avec une a) on ne sait pas extrapoler les donn6es
r6p6tition plus ou moins r6gulidre de I'unit6 recueillies dans les phases pr6coces d'6vo-
monomdre. Acfuellement, on peut consid6rer lution des mat6riaux (les m6thodes de la
cornme acquis les principes suivants : cin6tique homogdne ne sont pas
acceptables);
o Un mat6riau polymdre 6volue ir la fagon d'un
r6acteur chimique ou photochimique; sa b) on doit amener le mat6riau d un niveau
d6gradation suppose I'apparition de concen- d'6volution chimique qui enhaine une
trations g6n6ralement faibles de groupements d6gradation m6canique.

124
o Il est, par contre, indispensable de provoquer - La description bas6e sur les produits obser-
une acc6l6ration des 6v6nements chimi- vables en spectroscopie vibrationnelle (in-
ques en maintenant leur repr6sentativit6. frarouge ou Raman) doit Otre associ6e aux
Cette repr6sentativit6 doit d'abord 6tre la variations de propri6t6s m6caniques. Ces pro-
cons6quence des contraintes physiques et duits correspondent en effet aux voies princi
chimiques qui sont appliqu6es, lesquelles pales d'oxydation: ils sont en concentrations
doivent elles-m6mes €tre repr6sentatives faibles mais leur apparition correspond ir une
des contraintes en cours d'usage. La repr6- vraie d6t6rioration de la matrice.
sentativit6 doit 6tre surtout v6rifi6e en com-
parant les m6canismes r6actionnels en - La description bas6e sur les produits
conditions acc6l6r6es, d'une part, et en condi- observables en spectroscopie 6lec-
tions d'usage, d'autre part. Dans ce domaine, tronique (ultraviolet, lumidre visible, colo-
un m6canisme r6actionnel peut Cfre d6crit d rim6trie ou 6mission de fluorescence) doit
l'aide de s6quences r6actionnelles (et non de Otre associ6e aux variations d'aspect. Ces pro-
processus 6l6mentaires), et chaque s6quence duits les produits de jaunissement, par
r6actionnelle doit 6tre reconnue dans les 6vo- exemple- sont d6tect6s ir des concentrations
lutions artificielles et naturelles. Il s'agit ld de
-
trds inf6rieures d celles des produits princi-
la base de I'approche << m6canistique > qui, de paux d'oxydation observ6s en infrarouge ou
I 975 d nos jours, a pu 6tre d6velopp6e dans en Raman, et ils ne correspondent g6n6rale-
notre laboratoire pour la plupart des po- ment pas d une perte des propri6t6s
lymdres couramment utilisds. m6caniques de la matrice.

o Tout vieillissement acc6l6r6 correspond Il est donc possible de pr6voir la dur6e de vie
en fait d une acc6l6ration de l'6volution d'un mat6riau polymdre soumis ir des contrain-
chimique. Quand un m6canisme chimique tes lumineuses ou thermiques en pr6sence d'Oz
suppose plusieurs chemins r6actionnels d'im- et d'eau :
portance 6quivalente, on ne peut esp6rer ac-
c6l6rer tous ces chemins r6actionnels avec les
r en d6terminant, en conditions acc6l6r6es de
m€mes facteurs d'acc6l6ration. L'exp6rience laboratoire, la cin6tique de I'apparition d'un
en laboratoire d6forme alors la r6alit6 des < produit critique >>, c'est-d-dire un produit
conditions d'usage. De m€me, si, d l'6volu- qui s'accumule, selon une loi lin6aire, avec la
tion chimique, se supelposent des ph6no- dur6e d'exposition et dont I'apparition traduit
mdnes de transfert physique (oxygdne ou une coupure de chaine macromol6culaire;
stabilisant, par exemple), la pr6sence de
o en d6terminant, toujours en conditions
plusieurs processus dynamiques interdit le
transfert des conditions acc6l6r6es cr66es en acc6l6r6es de laboratoire, la corr6lation exis-
laboratoire aux conditions d'usage. On ne tant entre les variations de propri6t6s macro-
peut op6rer ce transfert que pour les systdmes scopiques d'usage (propri6t6s m6caniques ou
dont l'6volution chimique est contr6l6e par aspect) et les variations de la concentration de
un m6canisme chimique pr6pond6rant. ce produit critique; on pr6cise ainsi le seuil
Cette situation se pr6sente assez souvent tol6rable d'6volution chimique;
dans le cas du photovieillissement, of le
m6canisme photo-oxydatif est alors contr6- . en comparant la cindtique de I'apparition
lant et les ph6nomdnes de diffusion de l'oxy- du produit critique en conditions acc6l6r6es,
gdne, suffisamment rapides pour ne pas d'une part, et dans une phase pr6coce de
limiter I'oxydation. vieillissement en cours d'usage, d'autre part,
on obtient le facteur d'acc6l6ration qui permet
o La description de l'6volution chimique doit, de convertir la dur6e d'exposition en labora-
enfin, Otre associ6e au critdre de d6gradation toire en duree d'usage.
choisi :

t25
Dans le pr6sent expos6, trois classes exem- Ces polymdres se retrouvent souvent dans des
plaires de polymdres seront 6voqu6es, d savoir : sculptures contemporaines (de T. Grand, de
Niki de Saint Phale ou de J. P. Raynaud, par
l. la classe des polyesters insatur6s, comme exemple). L'insolubilit6 de ces polymdres
exemples de mat6riaux polymdres dont le r6ticul6s rend l'analyse de l'6volution chimique
comportement d long terme peut etre particulidrement difficile. Il a 6t6 possible nean-
modul6 par la structure m6me du polymdre moins de d6montrer les faits suivants :
et par f introduction d'additifs stabilisants;
l. Les groupements responsables de I'ab-
2. polyacrylates et des
la classe des sorption de la lumidre naturelle ext6rieure
polym6thacrylates, comme exemples de (1, > 300 nm) ou int6rieure (1, > 340 nm)
polymdres relativement stables dont le sont les groupements de mal6ates-fumarates
comportement d long terme dEpend pourtant r6siduels et les groupements de phtalates, les
dans une large mesure du compos6 6tranger oligomdres de styrdne n'6tant pas suscep-
qui a 6t6 introduit; tibles d'absorber cette lumidre. Ces
groupements absorbants provoquent, par
3. la classe des 6lastomdres di6niques, comme dissociation en radicaux, I'amorgage de la
exemples de polymdres dont le comporte- photo-oxydation.
ment d long terme a 6t6 pr6l'u, d tort, d partir
d' essais non repr6sentatifs. 2. Ces mat6riaux sont le sidge de ph6nomdnes
de < jaunissement )) photolyique (lapho-
La photolyse et la photo-oxydation tolyse d6signant des 6v6nements photo-
des polyesters insatur6s chimiques non modifiables par l'oxygdne).
Ces produits de la photolyse qui absorbent
Les polyesters insatw6s sont des polymdres de les rayons de la lumidre visible (),:: 400 nm)
condensation r6sultant de l'action de diacides peuvent €ne d6truits par photo-oxydation.
sur des dialcools (glycols). L'un au moins des
diacides est insatur6, ce qui permet la r6ticula- 3. Ces mat6riaux sont simultan6ment le sidge
tion par oligom6risation in situ d'un agent de de ph6nomdnes photo-oxydatifs. Cette
r6ticulation. On obtient alors un r6seau tridi photo-oxydation qui d6pend de la nature
mensionnel donnant au produit final son carac- -
des diacides et du glycol utilis6s, du r6ticu-
tdre d'irr6versibilit6 thermique le classant dans lant (styrdne ou acrylates) et des modes de
les thermodurcissables. Si I'on utilise le styrdne provoque 6galement des
r6ticulation
comme agent de r6ticulation, on obtient des -
jaunissements, mais elle est surtout respon-
polyesters insatur6s de structure :
sable de I'apparition de groupements oxy-
d6s (groupements hydroxyl6s et carbonyl6s),
qui s'accumulent et qui sont associ6s aux
variations de propri6t6s physiques (des
o o microfi ssurations superficielles,
rr I IcH-c-
rl
o-R-o- par exemple).
-c-cH- I
T" 4. Les jaunissements qui apparaissent de
H-
+o fagon plus importante en I'absence ou en
d6ficit d'oxygdne qu'en pr6sence d'oxygdne
sont dus soit aux seuls groupements de
styrdne (dans les polyesters insatur6s
6labor6s uniquement d partir d'anhydride
oi R peut Otre un noyau aromatique di- mal6ique), soit ir l'action conjugu6e de grou-
substitu6 en ortho ou en m6ta selon que le pements de phtalates et de styrdne (dans les
diacide utilis6 est un acide orthophtalique polyesters 6labor6s d partir de m6langes
ou isophtalique. d'anhydride mal6ique et d'acide phtalique).

126
5. L'oxydabilit6 des polyesters insatur6s photo-amorceurs, les unit6s r6actives et la na-
est surtout attribuable aux insaturations de ture des produits observ6s. Le tableau I porte
mal6ates ou de fumarates qui n'ont pas r6agi sur les polyesters obtenus d partir d'un melange
dans la r6ticulation. Une r6ticulation obte- d'acide phtalique et d'anhydride maldique et le
nue avec du styrdne r6duira ainsi I'oxydabi- tableau II, sur les polyesters obtenus d partir
1it6, mais elle accentuera le pouvoir jaunis- d'anhydride mal6ique.
sant. Une r6ticulation obtenue avec un acry-
late et un m6thacrylate diminuera le pouvoir Puisque la structure mol6culaire d'un polyester
jaunissant, mais le taux d'insaturation r6si- insatur6 r6ticul6 peut varier trds largement
selon la nature des monomdres, des r6ticulants
duel sera plus important et l'oxydabilit6,
et du mode de r6ticulation, la photor6activit6
plus grande. Dans ce dernier cas, les cou-
sera tout aussi variable. Elle pourra ainsi notam-
ches superficielles oxyddes donneront lieu
ment Otre adapt6e pour obtenir, par exemple,
ir des microfissurations, et il en r6sultera
une microfissuration r6duite, si I'on accepte
un blanchissement.
qu'un jaunissement important puisse se pro-
duire. En fait, la durabilit6 de ce mat6riau peut
6. La nature du dialcool est 6galement impor-
6tre largement am6lior6e grdce d l'emploi de
tante I'usage de n6opentylglycol permet
- I'oxydabilit6, par exemple.
de reduire
photostabilisants. Lorsqu'un polyester est si-
multan6ment le sidge d'une photolyse et d'une
photo-oxydation, ou d'un autre double proces-
Les mecanismes d'6volution chimique des sus photochimique du genre, il convient d'utili-
polyesters insatur6s r6ticul6s d l'aide de styrdne ser un m6lange form6 des substances suivantes :
sont sch6matis6s dans les tableaux I et II,
qui indiquent les groupes chromophores r un < anti-U. Vx. )) qui inhibera partiellement
la photolyse, ainsi que la photo-oxidation

Tableau I
Les micanismes d'bvolution chimique des polyesters obtenus d partir d'un mtlange
d'acide phtalique et d'anhydride mal4ique

Groupes chrornophores
photo-amorceurs Unlt6s r6actlves Prodults

-OCOCH=CHC@-
prodults
\ absorbant
\ [-n .l R. *o -ococH-cHcoo- en IR
I
l--- R:'
2
I
li (hydroxyles
I
et
" '-
n"./'lR' :I -cH- CH
2
carbonyl6s )
I
ge
cage hors gHt
----l
lMal6ates- l' '/ I caSe
Funarates
I | | n
I
-cH 2 -c-
t-
f
^-l
\7

t27
Tableau II
Les mtcanismes d'4tolution chimique des polyesters obtenus d partir d'anh1'dride mal2ique

Groupes chronophores Unlt6s r6actlves Prodults


photo-anorceurs
-OCOCH=CHCOO-
produl ts
\ absorbant
Fnl R.*o -ococH-cH@o- en lR
luat6ates-l
l--____-;
hu l'l+ I_-_) + __2
l_---)l I (hydroxyl6s
lrunaratesl I I

In'l
L". J R'
-cH-cH
et
carbonyl6s )
cag€
l2
hora CH
cage

,0
H
I

tant de la matrice que des produits de la circonstance se rencontre, en fait, de fagon clas-
photolyse (par comp6tition d' absorption); sique, dans tous les mat6riaux polymdres qui
n'absorbent pas la lumiere lorsque )" > 300 nm;
. un ( antioxydant redox > qui n'inhibera que l'amorgage photochimique de leur oxydation
la photo-oxydation de la matrice (sans inhi- est 6galement dt d des compos6s 6trangers
bition de la photo-oxydation des produits chromophores et photo-inducteurs. Mais, dans
de la photolyse). la plupart des cas, cet amorEage ne correspond
qu'd une phase initiale, les groupements oxyd6s
L'emploi d'un seul anti-U. V. risque d'accen- form6s sur la chaine principale du polymere
tuer le photojaunissement. Il faut l'assister d'un jouant trds vite le r6le de chromophores
antioxydant redox qui ne modifiera que les photo-amorceurs. Les poly(m6thacrylates
oxydations radicalaires sans perturber les de m6thyle), ou PMMA, se caract6risent, au
photo-oxydations directes. contraire, par le fait que leur durde de vie sem-
ble compldtement tributaire des contaminants.
La photo-oxydation
La photo-oxydation du PMMA se traduit par
des d6riv6s acryliques
I'apparition de produits hydroxyl6s observa-
Les polyacrylates et les polym6thacrylates bles en spectroscopie infrarouge ii transform6e
de Fourier (IRTF) e 3 580 et3 320 cm-'. Ces
sont des polymdres satur6s qui ne pr6sentent
pas de sites intrinsdques de fragilit6 photo- produits, qui atteignent une concentration
lytique et photo-oxydative. Les processus de constante au terme d'une phase pr6coce de for-
Nonish de type II qui interviennent dans les mation, ont 6te identifi6s d des groupements
esters satur6s ont une importance r6duite. Par alcooliques associ6s, par liaisons < hydro-
gdne >. soit aux groupements esters (ir .
contre, ils peuvent 6tre le sidge d'une oxydation
photo-induite par tout compos6 6tranger suscep- 3 580 cm-') soit entre eux (d 3 320 cm-'). Les
tible d'absorber la lumidre et de donner nais- produits hydroxyl6s sont les seuls photo-pro-
sance d des radicaux. Ces polymdres sont donc duits ais6ment observables; il se forme 6gale-
trds sensibles d toute contamination (introduite, ment des groupements responsables de l'6lar-
au moment de l'6laboration ou avec le temps, gissement de la bande d'absorption des groupe-
par migration dans les systdmes solides). Cette ments esters, mais il est impossible d'attribuer
des caractdres quantitatifs ir cette observation.

128
A grandes longueurs d'onde (1, > 300 nm), la absorption au-deld de 300 nm dans les spectres
photo-oxydation est essentiellement le fait de ultraviolets de plaques 6paisses (de 3 d 6 mm).
compos6s extrinsdques chromophores et photo-
inducteurs. Aucun photoproduit chromophore Le m6canisme de I'oxydation photo-induite
n'est form6 au cours de la photo-oxydation du PMMA peut 0tre repr6sent6 ainsi :
du PMMA et n'assure le relais des photo-
inducteurs initiaux. La photor6action s'arr€te
donc quand le photo-inducteur est entidrement
X (photo-lnducteur)
consomm6; ceci intervient 6galement ir trds
faible taux d'oxydation de la matrice de I r',
PMMA expos6e sous forme de film.
'' '*'o"' '

Quand, par conhe, une plaque 6paisse de


PMMA est expos6e en conditions accel6r6es
ou en conditions naturelles, la photo-oxydation cnz-c
l*'
est localis6e dans les quelque I 000 pm de la
- -
couche superfrcielle; elle r6sulte tant de la diffu- *flt'
sion d'oxygdne dans la matrice que d'une diffu-
I
sion qui se produit en direction oppos6e aux
photo-inducteurs. Cette deuxidme diffusion a
6t6 mise en 6vidence lors d'une exp6rience,
. -c
cH
i",
au cours de laquelle deux photo-oxydations - -
sont intervenues successivement, et of les
400 premiers pm les plus photo-oxyd6s ont 6t6
pr6lev6s au terme de la premidre photo-oxyda-
'f" I

tion. La deuxidme photo-oxydation s'est pro-


l02'PH
J
duite d des vitesses nettement inferieures ir
celles qui avaient 6t6 observ6es lors de la pre-
midre. Cette remarque justifie d'ailleurs les
techniques de r6novation du PMMA qui se fon-
dent sur une 6limination des couches oxyd6es
les plus superfi cielles.

Les insaturations r6siduelles sont des sites r6ac-


tifs qui disparaissent au cours de l'oxydation de
la mafrice de PMMA, bien que, n'absorbant pas _C_C_ +.OH
les photons de l" >300 nm, elles ne soient pas
responsables de I'amorgage de la photo-
oxydation du PMMA.

La fonction d'un photo-inducteur


-
c'est-2r- flr
dire d'un compos6 susceptible d'absorber la 9Hc
I

lumidre et de donner naissance d des radicaux I -cH-c-


tl oc
olll'
libres r6actifs, qui ne se recombineront donc -C-C-+H-O
pas entre estpeu sp6cifique. La nature coocH3 \HO'ffir
l\
des contaminants chromophores du PMMA
peut etre trds vari6e et d6pendre des conditions
qui ont pr6sid6 d l'6laboration, ir la mise en (non obeervable) groupcnenta hydroxyles
cuvre et d la formulation du PMMA qui fait assocl6e cntre eux
I'objet de I'analyse, et il n'est pas utile de
pr6ciser la nature mol6culaire du photo-induc-
teur, dont la pr6sence se manifestera par une PH: Hydrogdne du polltmdre

t29
Les polyacrylates d'alkyle pr6sentent des Depuis 1985, nous analysons le m6canisme
m6canismes de photo-oxydation analogues ?r de photo-oxydation de diff6rents types de poly-
ceux du PMMA. Un ph6nomdne nouveau peut butadidnes (BR), de polyisoprdnes (IR) et de
apparaitre, par contre, quand on polym6rise copolymdres styrdne-butadidne (SBR) ou acry-
in situ des monomdres ou des oligomdres lonitrile-butadidne (NBR). Nous avons proc6d6
d'acrylates pour constituer des couches adh6- d une etude d6taill6e des evolutions photo-
sives d'assez forte 6paisseur. Cette polym6risa- chimiques en utilisant diverses techniques
tion initiale est souvent inhib6e par I'oxygdne spectroscopiques (infrarouge, IRTF, micro-
et elle ne s'effectue que trds partiellement; les IRTF, ultraviolet ou Raman). En outre, les per-
insaturations r6siduelles restent donc en concen- turbations apport6es par des noirs de carbone
trations importantes au sein de la matrice au et des oxydes photo-actifs ont 6t6 analys6es.
terme de la polym6risation. Ces insaturations L'emploi de nouvelles techniques analyiques
r6siduelles sont susceptibles de donner lieu d bien adapt6es aux milieux trds opaques la
une polym6risation d long terme et il se cr6e -
microspectrophotom6trie IRTF et la spectro-
alors un r6seau polym6rique dans la matrice photom6trie IRTF avec d6tecteur opto-
initiale, avec une r6duction du volume occup6. acoustique, par exemple ont permis de
Il apparait alors des zones d'h6t6rog6n6it6 visi- reconnaitre, lors de l'6volution de mat6riaux fi-
bles d I'ail nu. Ce ph6nomdne est courant dans nis (rulcanis6s et pigment6s), les m6canismes
le cas d'cuvres oi de fortes 6paisseurs d'ad- qui interviennent dans le cas d'6lastomdres
h6sifs acryliques ont 6t6 utilis6es. non transform6s.

La photo-oxydation Comme pour tout pollmdre < non


d'6lastomires di6niques absorbant >, la lumidre est en fait absorb6e par
des chromophores ext6rieurs aux 6lastomdres
Jusqu'2r ces dernidres ann6es, les 6lastomdres examin6s (probablement des impuret6s ou des
di6niques n'ont pu, essentiellement d cause de d6fauts que l'on ne peut contr6ler). En outre,
facteurs d'ordre analytique, 6tre examin6s dans la vitesse initiale de photo-oxydation, g6n6rale-
Ie cadre de I'approche < m6canistique >. Dans ment trds faible, est une donn6e qui d6pend es-
leur forme finie, ces materiarx sont fort com- sentiellement des antioxydants r6siduels, et la
plexes et profond6ment modifi6s par la lulcani- p6riode d'induction ne peut 6tre consid6r6e que
sation ou la r6ticulation, et ils contiennent comme une caract6ristique extrinsdque. Cette
g6n6ralement des pourcentages 6lev6s de p6riode d'induction n'excede pas une heure
charges absorbantes en ultraviolet, en lumidre dans les conditions de photo-oxydation ex-
visible et en infrarouge. Les m6thodes spectro- ploit6es dans la cadre de la pr6sente 6tude.
scopiques qui sont habituellement utilis6es Au terme de cette p6riode, les contaminants
pour observer les quelques modifications inhibiteurs sont consomm6s et des chromopho-
chimiques des chaines polym6riques (d un de- res intrinsdques se forment (hydroperoxydes,
oh) ne c6tones a-B insatur6es, c6tones satur6es). La
916 d'avancement souvent inferieur d 1
peuvent plus 6tre mises en auwe ais6ment. photo-oxydation vraie de 1'6lastomdre inter-
La durabilit6 des 6lastomdres synth6tiques vient alors, selon le sch6ma de la page suivante.
6labor6s d partir de didnes n'avait donc 6t6
examin6e que sur le plan macroscopique, le On sait que I'oxydation radicalaire des alcdnes
vieillissement de ces mat6riaux expos6s en suppose g6n6ralement une extraction d'hy-
conditions naturelles ou simul6es n'6tant carac- drogdne sur le carbone situ6 en a de la double
t6ris6 que par des variations de propri6t6s liaison. Ce m6canisme d'amorgage par extrac-
mecaniques ou d'aspect (des microfissurations tion entre en comp6tition avec I'addition du
superficielles, par exemple). Une telle approche radical avec la double liaison. Une r6action
n'a pu foumir, d 1'6vidence, aucune indication en chaine d'hydroperoxydation permet d'ex-
sur la nature exacte de l'6volution chimique, les pliquer la formation d'hydroperoxydes o-B
seules connaissances de cet ordre n'ayant pu insatur6s. Aux hydroperoxydes associes qui
6tre acquises que sur des compos6s moddles de apparaissent d des concentrations maximales
faible poids moleculaire. de 60 mmol.ks-' doit 6tre attribuee la bande

130
d'absorption d 3 400 cm-l
environ, qui n'est pas
observ6e en oxydation I
o,
thermique d 60"C et qui J*'
disparait par photolyse r. (rO^.
a\ )
sous vide d 35"C.
+m \-
Il faut admettre que I'hy- | \ addition g.r
J aonule liaiaona
droperoxydation en cr de
{llr{11:O{ll-
la double liaison intervient
+l rH(ro2tl)
avant toute saturation de
cette double liaison pour
expliquer la formation de
l.'
-{Hfcrbcr+-cx- -oEcr- ) -cx.-qf,}FcH2-
c6tones a-p insatur6es par
photolyse (ou thermolyse) o
de ces hydroperoxydes. Si
t.
rO
les c6tones cr-B insatur6es OH
ttl Jtt
absorbant d 1 696 cm-' se -cH2-cH-cH-cH- -cHfcH{t-cH-
forment dans une r6action
en cage, habituellement I o
I
rencontr6e dans les po- lr6rctio da
OH
satuation
lymdres, les groupements I
nrl
alcooliques r6sultent de I
t
radicaux alcenoxydes qui -CH"-CH:CI|-CH- q4 -q1^-611:CK-C1t- +
ne sont pas recombin6s ' I t l.
OH'

dans la cage. oH
1o.u*.i*
cag€
La bande d'absorption i Jen
| 726 cm'' a ete attribuee -CH2-CH:CH-C- + H^O
aux groupements c6toni- o
ll'
ques saturds que I'on peut
Fatofi-l \--,- r\
I gaturdes I -l r6rctidl ds I l. \r
observer ais6ment lors ,rl ,"i*"u"" I lhu'oz \ i"*'etiEation cis-trans
d'oxydations thermiques. l-;F;6
I Eatu€E I
I r--.-\ +
acides cr, p inaatur6g
En photo-oxydation, les
c6tones ne peuvent ap-
paraitre qu'en faibles con- PH: Hydrogdne du polvnftre
centrations stationnaires car les processus
photochimiques de Norrish de types I et II pro- montrent que la photo-oxydation des polybu-
voquent leur conversion. En particulier, la for- . tadidnes ne depend guere du genre de micro-
mation d'acides satur6s (absorbant d | 717 cm-' structure, puisque toute insaturation, quelle
environ) intervient dans toute matrice poly- qu'elle soit, entraine une forte photo-oxyda-
m6rique oi sont form6es interm6diairement bilit6 de ces 6lastomdres ind6pendamment
des c6tones satur6es; cette conversion de c6to- de la structure.
nes satur6es en acides suppose un processus
de Norrish de type I. La photo-oxydation d'un 6lastomdre se traduit
6galement par une modification importante du
La disparition des insaturations r6siduelles est r6seau. Aux scissions de chaines intervenant
directement observ6e d partir des spectres infra- lors de la formation des groupements oxyd6s,
rouges et elle est i l'origine de la formation de s'opposent des r6ticulations. Les photopassiva-
tous les produits d'oxydation. En outre, cette tions tres remarquables, observ6es m6me en
disparition permet d' interpr6ter I a r6ticulation films minces (de I 50 d 2 l0 pm), s'interprdtent
observ6e en R.M.N. Les donn6es cin6tiques par une augmentation de l'imperm6abilit6 d

l3l
I'oxygdne des couches superficielles oxyd6es 2. Les radicaux alc6noxydes correspondants
et r6ticul6es. Il convient de noter que, en donnent naissance i des processus plus
photovieillissement artificiel d'6lastomdres vari6s :
non transform6s, les films deviennent, sous CH:
exposition, des photordacteurs trds h6t6rogdnes I

qui ne s'oxydent que superficiellement. Le


caur du film n'est alors pratiquement
pas modifie. PH
/I
-cH-cH-c-cH2-

/u'
CH: / -
CH"-CO-GH,- +'cH-cH-
Les copolymdres SBR et NBR 6tudi6s se sont
V (i | 722 cn-r)
comport6s comme les homopolymdres de poly- -cH:cH-c-cH"-
butadidne; les unit6s de styrdne et d'acryloni-
l'
I
\
trile sont demeur6es pratiquement inertes au --ctt-cn-co-cH2-
cours de la photo-oxydation, ce qui signifie \ +'cH3
(a 1 6e3 cm-l)
que tous les radicaux produits dans la matrice
\
r6agissent en fait sur les sites insatur6s.
cH-cH-co-cH3 + 'cHr-
Les polyisoprdnes pr6sentent des photo- (A I 693 crn-1)
oxydations qui interviennent selon le m6me
m6canisme g6n6ral. Mais la structure m6me Dans le cas du polyisoprdne, les processus
du polyisoprdne entraine les differences de B-scission des radicaux alc6noxydes provo-
suivantes : quent donc des coupures de chaines. Il s'agit ld
d'un facteur qui diff6rencie nettement le poly-
l. La dissym6trie de I'unit6 isopr6nique laisse isoprdne des autres 6lastomdres pour lesquels
pr6voir la formation de deux radicaux diff6- les coupures de chaines interviennent par d'au-
rents par r6action d'extraction d'hydrogdne : tres mecanismes moins primaires et apparem-
cHs cHs
ment moins fr6quents.
I
Il est maintenant possible de d6crire l'6volu-
-CH-CH-C-CH2- ou -CH-C-CH-CH2-
tion chimique d'6lastomdres di6niques non
(a) (b) transform6s d I'aide des diff6rents produits
interm6diaires et frnals qui se forment au cours
Les hydroperoxydes apparaissent en plus
d'oxydations photothermiques (d basse tem-
p6rature) ou thermiques. Il est donc possible
fortes concenfrations dans les polyisoprdnes
de comparer cette 6volution en photovieillisse-
que dans les autres 6lastomdres, ce qui est com-
ment ou en thermovieillissement artificiel et
patible avec une structwe tertiaire de ces hy-
celle qui se produira en vieillissement clima-
droperoxydes. Les radicaux (a) semblent donc
tique. Il apparait alors clairement que le vieillis-
6tre essentiellement form6s et leurs formes
sement climatique d'6lastomdres di6niques non
m6somdres :
transform6s se r6duit essentiellement d une oxy-
CH. cHe dation photochimique et qu'il ne correspond, en
t-
-cH-cH:c-cH2 <-> -cH:cH-9-.",
I
aucun cas, ir une ozonisation (ce demier genre
d' oxydation pr6sentant une stcchiom6trie tout
d fait differente de celle d'une oxydation photo-
chi mique ou thermique).
permettent de rendre compte de cette structure
tertiaire : Par ailleurs, il convient de signaler que la
pr6sente 6tude a 6t6 prolong6e pour permettre
CHs
I'analyse de m6langes form6s d'6lastomdres
I
-cH:cH-c-cH2- vulcanis6s et contenant de forts pourcentages
I
de noir de carbone et de pigments photo-
ooH actifs (ZnO et TiOz). Les r6sultats obtenus

r32
jusqu'ir maintenant montrent, d l'6vidence et convertir les dur6es d'essai en temps r6el
-
malgr6 les difficult6s d'ordre analytique rencon- d'usage dans des conditions moyennes
tr6es que le vieillissement climatique de ces d'utilisation.
-, expos6s d la lumidre s'explique
polymdres
essentiellement par une oxydation photo- Note
chimique, et non par une ozonisation.
*Un anti-U. V. est un compos6 mol6culaire
Conclusions qui absorbe fortement les rayons ulfraviolets
et dont les 6tats excit6s sont susceptibles de se
La pr6vision du comportement d long terme d6sactiver, de fagon non radiative, avec une trds
des mat6riaux polymdres synthdtiques en condi- grande efficacit6.
tions d'usage reste diffrcile. Mais I'avancement
des connaissances sur les m6canismes d'6volu- Abstract
tion rend cette pr6vision de moins en moins
al6atoire. En fait, les progrds r6alis6s r6cem- Prediction ofthe Long-Term Behaviour of
ment r6sultent de la conionction de deux Synthetic Po lymeric Mate rials from Artilicial
facteurs, d savoir : Ageing Experiments

o La mise au point de dispositifs exp6rimentaux The photo-ageing of synthetic polymers com-


d'6tude en laboratoire qui permettent d'exa- monly used in modem sculptures found in out-
miner les seuls ph6nomdnes de photovieillis- door museums will be described. CrossJinked
unsaturated po lyesters and various polyacrylates
sement en conditions anhydres (avec applica-
and polymethacrylates will be presented as exam-
tion de contraintes de lumidr:, de chaleur ples of unstabilized matrices whose properties
et d'Oz) ou en conditions de concentration can be modified during the polvmeisation, on
d'eau maintenue (avec application de dryingfrom solution or emulsion and by introduc-
contraintes de lumidre, de chaleur, d'Oz tion of additives. Photo-oxidation and thermal
et d'eau). Il devient alors possible de oxidation of dienic elastomers (Butadiene, 54,-
hi6rarchiser I' importance des diff6rents rene-butadiene and Nitrile-butadiene rubber)
m6canismes de photovieillissement, de and oxidation ofpolyacetals will be described.
thermovieillissement et de vieillissement A basic understanding ofthe oxidation mecha-
nism is pre-requisite to any stabilisation strateg/.
hydrolytique, et de n'acc6l6rer en condi-
tions artificielles que le m6canisme le
The prediction ofthe rate ofageingfor synthetic
plus important. polymers in use can be based on the "mechanis-
tic" approach. This approach entails the recogni-
o Le d6veloppement de m6thodes microana- tion ofsolid po$rners as chemical "reactors",
lytiques qui permettent de suiwe in situ l'6vo- that is, environments where certain chemical re-
lution chimique de microzones des systdmes actions are facilitated. The mechanistic approach
polym6riques. Ces techniques de microspec- involves the identification of mechanisms at the
trophotom6trie vibrationnelle la microspec- molecular scale, in artificial accelerated condi-
-
trophotom6trie IRTF, par exemple rendent
tions and in use. The most detimental effects on
-
possible I'analyse des couches 6l6mentaires
pofi,rners are caused by oxidation which is in-
duced by light, heat, static and dynamic stress.
de 5 pm et l'6laboration des profils des pro- Vaiations in macroscopic physical properties
duits d'oxydation et additifs qui se trouvent (mechanical propenies, opacity, surface aspect,
dans les parties les plus superficielles du etc.) can be explained by analyzing the chemistry
mat6riau. Elles peuvent de plus €tre utilis6es, fully. Thus, accelerated ageing is actually acceler-
grdce ir des d6tecteurs opto-acoustiques, pour ated oxidation and the accelerationfactor can be
analyser des milieux trds opaques. Enfin, ces determined from the rates at which critical oxida-
tion products form in artificial and natural condi-
reconnaissances analytiques des m6canismes
tions. Lifetimes of polymers can therefore be
d'6volution permettent, tout en garantissant la
predicted from the measured lifetimes in artifcial
repr6sentativit6 des essais de laboratoire, de

IJJ
c ondi t i otrs, t aking ac ce I erati on fact o rs int o ac - In ntost $,nthetic pobmers, the oxidative degrada-
cottnt. This mechanistic approach contpares fa' tion, u,hich is controlled either by'the nonnal
vourablv with entpiical techniEtes based on structure of the polltnerized monomer or b1'
simulation of actual physical and chemical condi' chemical dekcts, is obsented at a very lou' degree
tions and the resulting vanations in macroscopic of oxidation. Ven' sensitive and infornative in-
physical properties. situ spectrophotometric techniques are required
to detetmine the nature of the various intermedi-
ate and final products.

t34
Composition Implications of Plastic Artifacts:
A Survey of Additives and Their Effects on the Longevity of Plastics

R. Scott Williams

an C ons erv ation In s titute


Can adi
Communications Canada
Ottawa. Canada

Abstract Introduction
Plastic objects are usually described by A plastic is not a pure chemical product but
reference to their main polymer component, rather a formulation or composition made by
i.e., polyetfuilene sheet, vinyl or poly(vinyl chlo- proper mixture of a base polymer with a combi-
ride) upholstery, acrylic sculpture, polystyrene nationof additlves. Some cornmon base poly-
or styrene box, etc. Infact, all manufactured mers are listed in Table I. Polymers can be
plastic items are complexformulations of poly- produced as pure materials but these are never
mers with the additives that are required to adequate for serviceable products. To produce a
give the base pol"vmer suitable end-use proper- useful plastic, the inherent chemical, physical,
ties. Additives such as plasticizers, stabilizers, mechanical, optical, electrical, and other proper-
colorants, processing aids, etc. are as impor- ties of the polymer must be modified by the in-
tant as the base polymer in determining how corporation of additives, such as those listed in
long an object serves its intended purpose, or Table II. Plastics also are subjected to a variety
how long it survives in a museum. of physical treatrnents to modify such proper-
ties as appearance, printability, and shape.
The main polymer groups that comprise (Table III lists some of these.) The nature of
plastics found in typical museum objects are the base polyrner, additives, and fabrication all
brieflv introduced, and the additives in the plas- affect the service life and museum-longevity
tics madefrom these polymers are discussed in ofobjects. This paper discusses the mode of
detail. Results of recent chemical analyses of action of some additives and the deterioration
deteriorated plastic objects are used to illus- ofplastic objects due to their change or loss.
trate the effect of additives and base polymers
on the longevity of the plastics. Deterioration General Characteristics of Additives
and damage such as accretiow, blooming,
cracking, crazing, discoloration, embrittle- Additives should be efficient. that is. effective
ment, oozing, softening, etc. are related to at low concentration (most are used at concen-
changes in, or loss of, the additives. These trations of less than lo/o,with the notable excep-
changes are Wuenced by the museum environ- tions of plasticizers, pigments, fillers, and
ment. Generql guidelines for storage of plastic reinforcing fibres typically used at concentra-
objects, both to increase their longevity and to tions of l0o/o to 35olo), low-cost, convenient,
prevent damage to neighbouring objects, safe to use and handle, and should not impart
are given. undesirable characteristics such as colour. taste.

135
Compatibility and Permanence
Table I Polymers and additives are compatible if they
Common Base Polvmers in Plastics can be intimately blended with each other to
form a homogeneous composition. Additives
(Note: The abbreviaions are those grven In ASTM that are highly compatible with a given resin do
Standard D 1600-83: Standard Abbrevratrons of
Terms Relatrng to Plastrcs by the American Society for not exude to form droplets or liquid surface
Testino and Materials.) films, nor do they bloom as a crystalline sur-
face crust. Incompatibility is usually indicated
by migration and exudation of substances on
Acrylonrtnle-butadrene-styrene ABS the plastic surface (also called spewing or
Cellulose acetate CA
Cellulose acetate butyrate CAB blooming), or by poor physical properties. This
Cellulose acelate propronate CAP is not always evident after short storage periods
Cellulose plastrcs, general CE but may take months or years to appear. Perma-
Cresol formaldehyde UT
Cellulose nitrate CN nence rcfers to the ability of an additive to re-
Cellulose propionale CP main unchanged within various environments
Casern UJ
so that the plastic retains its desired properties
Cellulose tnacetate CTA
Epoxy, epoxide EP during use, as opposed to simple incompatibil-
Ethyl cellulose EC ity,which becomes evident by time alone and
Ethylene-vinyl acetate EVA
lmpact polystyrene IPS
is not related to the exposure environment. The
MelamrneJormaldehyde MF permanence of an additive is usually deter-
Polyamrde (nylon) PA mined by its volatility, migration, extractabil-
PolycaIbonate rw
Polyethylene PE ity, and stability to heat and light.
Poly(ethylene terephthalate) PET
Phenol{ormaldehyde PF
Compatibility can be thought of as mutual
Polyisobutylene PIB
Poly(methyl methacrylate) PMMA solubility, although phenomena other than solu-
Polypropylene PP bility can create compatibility (e.g., molecular
Polystyrene t'b
Polytehaf luoroethylene PTFE
entanglement). Additives are often relatively
Polyurethane PUR small molecules that can diffuse through the
Poly(vrnyl acetate) PVAC polymer matrix. The larger the additive mole-
Poly(vrnyl alcohol) PVAL
Poly(vnyl butyral) PVB cule, or the more crosslinked or crystalline the
Poly(vnyl chlonde) PVC polymer, the slower is the diffusion. As a result
Poly(vnyhdene chlonde) PVDC
Styrene-acrylonitnle SAN
Styrene-butadrene SB
Srlcone plastrcs
Urea{ormaldehyde
sl
UF
Table II
Unsaturated polyester UP Additives in Plastics

Plasticrzers (50% rn some PVC)

Stabrlizers
or odour. They should also maintain their Antioxdants
properties over time, be resistant to extraction Heat stabllizers
UV absorbers
during service, be thermally stable at processing
temperatures, be stable to light, and be unaf- Processng Ards
fected by pollutants. Additives should be Intemal lubncants
Mould release agents
compatible with the polymer and other addi- Slip agents
tives, and be capable of blending mutually. Blowrng agents
They are intended to remain effective through-
End-use Modifiers
out the serttice lifetime of the plastic. Note that Colourants
the designed service lifetime of a plastic object - Organrc dyes and prgments
- Inorganlc prgments
is always much shorter than the desired mu-
Reinforcing frbres
seum lifetime. It would be nice if additives Fillers and extenders
remained effective throughout their museum Anltstattc agents
Antrblock agents
lifetimes, but this is seldom the case. Barrier coatngs
Lamrnatrng process

136
of diffusion driven by concenhation gradients, Hedley 1980, or Horie 1987 for discussions
additives can migrate to surfaces where, if they of solubility parameters). Table IV lists the
are not volatile, they collect as a discrete ex- Hildebrand solubility parameters of some
uded layer (this property is sometimes used common polymers and compounds used as
intentionally as in the case of slip agents and plasticizers. Compounds most commonly used
antistats). This migration continues until the as plasticizers for poly(vinyl chloride) (PVC)
concentration gradient disappears, at which have Hildpbrand solubility parameters between
time the rate of diftrsion from the interior to 17 .2 MPai'z and 23.3 MPa''2 bracketing
the surface equals the rate of reverse diffirsion 19.6MPa'2, the solubility parameter of PVC.
from the surface to the interior. Ifthe exudate The solubility parameters of other polymers
is removed, a new concentration gradient is set (e.g., cellulose nitrate and polyethylene) are
up and more additive migrates to the surface to outside the compatibility range of plasticizers
replace the lost material. Volatile additives that are suitable for PVC. Thus different types
evaporate at the surface, to be replaced by more ofplasticizers are required for different poly-
additive from the bulk of the plastic. In this situ- mers, as is verified by formulation practice.
ation there is a constant loss ofadditive to the
atmosphere and a constant diffusion of additive Compatibility and Dielectric Constant
from the interior to the surface. Eventually, all Solubility parameter is only one of many
the additive will be lost (and possibly redepo- properties that can be evaluated to predict plas-
sited elsewhere to cause harm). ticizerlpolymer compatibility. Dielectric con-
stant is also an important properly of a solvent.
Although plasticizers will be discussed in detail The dielectric constant of a compound is a
later, additive/polymer compatibility using plas- measure of its polarity and its polarizability.
ticizers as an example is discussed here to Polarizability is a measure of the response of
clarify discussions of all additives. a molecule to an electric field and thus of the
intermolecular forces between a solvent and
Compatibility and the a solute, that is, between a plasticizer and a
Hildebrand Sotubitity Parameter polymer. The interaction of these forces affects
Plasticization is similar to dissolution of the compatibility. Observation of compatible plasti-
polymer by the plasticizer compound. Thus the cizers shows that those with dielectric constants
Hildebrand solubility parameter, which is a between 4 and 8 are compatible with PVC,
measure of solubility or solvent power, may be which has a dielectric constant of 3.2
a predictor of compatibility (see Barton 1983, (Table IV).

Exudation of Additives due to


Table III Incompatibility
Pl as ti c F abric ation Pro c ess es Figure 1 shows the relationship between solu-
bility parameter and dielectric constant for
compounds commonly used as plasticizers (for
Shaping Processes all plastics, not just PVC). The zone of compati-
Molding, e}trusron, casting bility for PVC is indicated by the dotted lines.
Machining (cuttrng, gnndrng, dnlIng,
stamping, etc.) As can be seen there is a restricted group of
Foamrng compatible plasticizers for PVC. Similar zones
of compatibility could be plotted for all other
Surface Treatments
Surfa@ textunzrng, embosstng plastics and would include a different group
Surface activation (to permit printrng) of plasticizers.
- Flame
- Corona drscharge
Coating This Figure helps to explain the occurrence
- Painting, pnntrng of exudates and blooms. Some additives have
- Metallrzrng, electroplatrng
- Lamrnaton solubility parameters and dielectric constants
Joining outside the zone of compatibility. We should
Adhesive bonding expect these to exude because they are incom-
Welding
Mechanrcal fastening (screws, nvets, etc.) patible. Exudations may also occur if the

t37
Table IV
Pl asti cizers : Solubilit"v Parameters
and Dielectric Constants

Solubility
Paralpeter Dielectric
Plasticizers (Mpa") Constant

Phthalate esters:
Di (2-ethylhexyl phthal ate), 14.9-18.0 5.2
(dioctyl phthalate), DOP
Dibutylphthalate, DBP 17.0-19.2 6.4
Butyl benzylphthalate 15.5-18.2 6.4

Phosphate esters:
Triphenyl phosphate, TPP 1e.6 (20.7)
Tricresyl phosphate, TCP 17.2-20.2 7.2

Diacid esters:
Di (2-ethyl hexyl) adipate,
(dioctyladipate), DOA ('t7.4) 4.1
Di(2-ethylhexyl) azelate,
(dioctylazelate), DOZ 19.0 4.0
Diethylene glycol dibenzoate 7.1

Miscellaneous:
Epoxidized soya oil, ESO (18.2) 5.5
Adipic acid polyester 19.0 6.0
Tri(2-ethylhexyl)
trimellitate, (trioctyl
trimellitate). TOTM 18.4 4.7
Camphor 18.2 11.4
Castor oil 8.9

Polymers

Polytetrafl uoroethylene 12.7 2.0


Polyethylene 16.2-16.8 2.25-2.35
Polypropylene 18.8 2.2
Polystyrene 17.6-19.8 2.4-2.6
Poly(vinyl acetate) 18.0-22.5 3.5
Poly(methyl methacrylate) 18.4-19.4 3.3-3.6
Poly(vinyl chloride) 19.0-20.2 3.3-3.5
Polycarbonate 19.0-20.2 3.0-3.2
Poly(ethylene terephthalate) 21.9 3.0-3.6
Cellulose nitrate 21.5-23.5 7.0-7.5
Cellulose diacetate 22.3-23.3
Cellulose acetate 27.2-27.8 3.5-7.5
Nylon 66 27.8 4.O

r38
solubility parameters and dielectric constants of mutual solubilities and may lead to separation
polymers and additives shift out of the compati- of components (incompatibility). As aging pro-
bility zone as chemical reactions take place in ceeds, these changes make the plastic composi-
the plastic as it ages. For instance, it is possible tion more and more different from the initial
that dibutyl phthalate, a very compatible PVC design formulation. Since the initial formula-
plasticizer, may hydrolyze to form butyl alco- tion was designed by delicately balancing many
hol and phthalic acid. The solubility parameter properties of different polymers and additives,
and dielec,tric constant of Qutyl alcohol are the likelihood of incompatibilities increases.
23.3 MPa''2 and 36.4 MP a''2 respectively, val- The plastic should be designed so that these
ues well outside the zone of compatibility. We developing incompatibilities do not affect the
would expect this hydrolyic degradation prod- properties of the plastic during its design life-
uct (of an additive, not the polymer) to exude. time. Unforfunately, over the extended duration
This could happen for any additive that changes in a museum, problems of incompatibility start
upon aging. Similarly, if the polymer changes to show up.
enough to shift its solubility parameter and di-
electric constant, then the changed polymer Deterioration of Plastics
would have a new compatibility zone. Un-
changed additives might now lie outside the Plastics deteriorate by l) chemical degra-
new compatibility zone of the aged (changed) dation of the base polymer and/or the additives,
plastic. They might now be incompatible, 2) physical processes like bending and breaking
and therefore exude. or exudation of components (perhaps resulting
from chemical changes), and 3) biological
Incompatibility due to Aging of Base agents (molds, fimgi, rodents, people, etc.),
Polymer and Addifives just as all other organic and biological museum
There is a particular relationship befween the objects deteriorate (Table V). Physical proc-
base polymer and its additives in an unaged esses include diffusion and molecular redistri-
plastic. lnitially, everything is designed to be bution where there is no change in the chemical
mutually compatible. Upon aging, compatibili- bonds in the polymers or additives. These
ties change. Light-, heat-,
and oxygen-induced reac-
tions of the various com-
20 tldninal
ponents occur. Stabilizers
cornpa6lity
prevent reactions of the lmrts basod
on€
polymer macromolecules,
but in doing so the stabi-
lizers are changed into r- 14
@
new products, with differ- 6
ent compatibilities with !12
the base polymers. And" Eto
t Nminal
@mpa&a.ty
in spite of the stabilizers, o lmil8 baaocl
mA
changes do occur in the 88
polymer. Slow macro-
molecular rearangements
increase crystallinity. A
few crosslinks pull poly-
mer molecules together
and decrease molecular
o 5 r0 15 20 25 Xt 35 40 45 50 55 60
free volume, squeezing
U6locticdr6tan, t
out additives. Oxidation
of macromolecules and Figure I Compatibility zone defned by solubility parameter and dielectric
additives tend to increase constants ofplasticizers. Curyes: l, carboxtlic acids; 2, hltdrocarbons:
their polarities. All this 3, chlorinated hydrocarbons; 4, esters: 5, ethers; 6, alcohols; 7, all;yl nitiles;
leads to changes in 8, aldehydes;9, ketones; 10, nitro compounds. (from Darbv, Touchette and
Sears 1967.)

t39
Degradation Profile
Table V To help visualize degradation processes,
Agents of Degradation degradation profiles can be created by plotting
changes in such properties as colour, brittle-
Heat during Processing and use
ness, tack, surface gloss or crazing or chalking,
Light during use
Atmosphere oxygen and Pollutlon release of volatile substances, and many other
Morsture hydrolysts chemical, mechanical, and physical properties.
Solvents dtssolutton and solvolysls
Feller (1977) described four stages ofdegrada-
Brologrcal Agents bactena, fungi, insects,
rodents, PeoPle tion inception, induction, increase to maxi-
Physrcal Changes crystallizalon, migratlon mum- rate or steady state, and decreasing rate
Mechanical Wo|* fatigue, creep
stages and introduced a generalized profile
- degradation as measured by oxy-
of oxidative
are usually slow and can be accommodated by gen uptake. Examples of degradation profiles
compatibility considerations during design of are shown in Figure 2 for yellowing, for weight
the plastic (See also McGlinchey, this publica- gain due to oxygen uptake during oxidation, for
tion). This sort of degradation takes a long time solubility, and for oxidation (measured by IR
to show. Often it is not detected, or not signifi- absorbance due to hydroperoxide formation).
cant, during the service lifetime of a plastic, but These degradation profiles show the state of
becomes a problem after long periods in a mu- degradation ofa plastic object at any stage in
seum. Chemical degradation alters the arrange- its lifetime and permit some prediction of the
ment and type of chemical bonds present in the course of its degradation. This is valuable in the
plastic. The primary chemical changes are due museum/conservation context. Chemical degra-
to oxidation, although hydrolysis is a factor in dation follows the same general pattem for all
some polymers. Oxidation can be initiated or plastics and organic materials although the
promoted by heat (thermal oxidation) and light
breadth and height of the curve may change.
(photo oxidation). Light in the absence of It is worth examining degradation profiles in
oxygen can also cause chemical degradation. more detail to see how plastic objects degrade,
to identify the stage ofdegradation achieved by
The inherent properties of plastics affect their a particular object, to predict what future can be
degradation. For example, amorphous polymers expected for a particular object, and to deter-
allow easier diffusion of oxygen and water than mine how the course ofdegradation ofobjects
do crystalline polymers, therefore oxidation can be altered so as to prolong their useful life.
and hydrolysis are likely to be more pro-
nounced in the former. Plastics with glass tran- Consider a plastic with a degradation profile
sition temperatures below room temperature are shown in curve A in Figure 3. If the duration of
in a rubbery state, which allows greater molecu- the inceptiorVinduction stage can be increased,
lar mobility and greater oppornrnity for reactive the amount of degradation that will have oc-
degradative species or degradation products to curred at a future observation date will be
move through the plastic thereby increasing the
decreased, even ifthe shape (slope) ofthe later
rate ofdegradation reactions. The presence or portions of the curve is unaltered. This situation
absence ofcertain functional groups in the poly-
is shown on curve B. Note that the amount of
mer macromolecules affects the susceptibility
degradation at the future observation date on
to certain degradation reactions. For instance, curve B (labelled Ds) is less than that on the
esters and amides are susceptible to hydrolysis,
original curve (Da). If the inception/induction
aromatic rings and carbonyl groups absorb ul- period cannot be increased, but the rate of deg-
traviolet (UV) light and are sensitive to pho- radation can be reduced (i.e., reduced slope),
todegradation, and pure hydrocarbons like then the total amount of degradation at the fu-
polyethylene, polypropylene, and rubber con- ture observation date can be reduced. This is
tain no chemically bound oxygen and so will shown in curve C. Note that Dc is less than De.
oxidize only if exposed to oxygen. Unforhr- As another example of possible intervention,
nately these latter hydrocarbons contain tertiary
consider that the durations ofthe various stages
hydrogens, which makes them very susceptible
cannot be changed but the magnitude of each
to degradation by free-radical processes. can be decreased. This is shown in curve D.

140
O Sample I
aSampte 2

al Slmpl6 on Gls
oA Samptes in Alumlnrum Panr

Hours in Own * IOOoC

*,1'J.',,1:';:u^11"T.0r!i,ij0,tror: res,r
leo
loo ."f
3ao
dza ."I
d cxi rol I
4?O 'ool
801
{ro €oi
Eeo .ol
ito 201
100 ol

Figure 2 Degradation profiles produced by measuring diferent properties of plastics Top teft yellowing
o/'
ru-bber cement (from Feller and Encke 1982); Top right: ireghigain of rubier cement (from
Feller and'Encke
I 982 ): Botrom lefl:, Hydloperoxide concentration in photo-oxldiid poiyprop,tene
as measured bv IR absor-
bance at J400 cm ' with no inhibitor (A), with nickel chelate inhibitor (B) and with piperidine
H)LS inhibitor
(C) (from Grattan 1978); Bottom right: Change in solubility of varnish r'esins (from Felter
1975).

Again' the total amount of degradation at the grofile. They function either by preventing the
future observation date @o) is less than in the initiation reactions caused by light and h# ex-
case without intervention. posure, or by interfering witir th"e free-radical
chain reactions, thereby preventing propagation
In general, to decrease the degradation ofaplas- reactions. They also scavenge non-free-ra"dicat
tic at some future date the objectives should be products of degradation that-cause or catalyze
to 1) increase the duration ofeach ofthe early further degradition. Ifstabilizers are effeciive
stages ofthe deterioration_profile (i.e., delay they alterihe degradation profile by lengthen-
the onset of later stages), 2) decrease the rate of ingthe time scai6 and reducing the exteit of
change from each earlier stage to the next, and degradation. When stabiliz"r.iuil (by loss or
3) decrease the magnitude of degradation for change) the degradation profile becomes much
each stage. steepir and the plastic degrades more rapidly.

Stabilizers The stabilizers to be discussed here include


Stabilizers (inhibiton) are additives that are antioxidants, heat stabil izers, and ultraviolet
used to make changes in the degradation

t4l
cleavage to produce two new oxygenated radi-
De4radation Trofile cals (hydroxyl and alkoxyl). Thus the number
of radicals multiplies and oxidative degradation
o of the polymer via free-radical reactions is
E accelerated.
\E
o In the initial stages of this chain reaction, radi-
o cals react with non-radical polymer macromole-
cules. Eventually, as the radicals multiply, their
concentration becomes large enough that they
react with each other as well as with the non-
'6
radical macromolecules. The reaction of one
free-radical with another produces a non-radi-
cal. Thus radical-radical reactions terminate
t'(r)
Jl
6'a) free-radical chain reactions. Eventually the
lncreaainq Aqe of }bJect, I
rate of termination of radicals equals the rate
ofproduction ofradicals and a steady state is
Figure 3 Altering the degradation profile of a plas- reached. This corresponds to the steady state
tic. Curtes: A, unaltered profile; B, induction period
region ofthe degradation profile degrada-
increased but shape ofprofile unchanged (B is "par- -
tion continues, but at a constant, non-accelerat-
allel"to A); C, induction period unchanged but rate
ofdegradation decreased (slope decreased); D, in- ing rate.
duction period and rate ofdegradanon unchanged
but magnitude ofdegradation decreased. D* Da, Prevention of Oxidation
Dg, and DL, are amounts of degradation at some ar- Key stages in the oxidation process are the
bitrary future observation time, t, -for each scenario. initial formation of alkyl free-radicals by ener-
The end of the induction period is indicated by getic processes and the subsequent production
ttme, tt. ofhydroperoxides by reaction ofthe alkyl
free-radicals with oxygen. Oxidation can be
light stabilizers (including ultraviolet light
prevented if formation of either of these species
absorbers).
is prevented or if these are converted to some-
thing innocuous before they have a chance to
Antioxidants react with polymer macromolecules. Thus to
prevent oxidation the aims should be the
Oxidation Reactions
following:
To understand how antioxidants work, the
chemical reactions that take place in the base
e prevent access ofoxygen to the plastic by
polymer during oxidation must first be consid-
removing oxygen from its surroundings
ered. This subject is vast and has been dis-
(anaerobic conditions) or by using oxygen
cussed in the conservation context by Grattan
(1978) and de la Rie (1988) and in much barrier layers on the plastic
greater technical detail by Carlsson and Wiles
(1986). For oxidation to occur, oxygen must be r prevent initial formation of alkyl radicals,
absorbed and dissolved in the polymer. It then or scavenge those radicals formed before they
diffuses to reactive sites on the polymer macro- have a chance to react with oxygen in the
molecule where oxidation reactions can take plastic (radical scavenger antioxidants)
place. Such reactive sites are usually free-radi-
cals produced when energy in the form ofheat, o destroy peroxides that do form before they
radiation (commonly UV and visible light but cleave to produce new radicals (peroxide
also gamma- and X-rays, or electron beams), decomposer antioxidants)
and mechanical work, etc. is absorbed by the
plastic and breaks chemical bonds in polymer r prevent the development of, or remove,
macromolecules. These free-radicals react with active sites from the polymer macromolecules
oxygen to produce hydroperoxides. Hydroper-
oxides readily undergo thermal and photo

142
Anaerobic (oxygen-free) atmospheres can be greater detail, with emphasis on conservation
produced using bell jars with inert gas purging applications. Free-radical scavengers are typi-
or by enveloping objects in oxygen barrier cally hindered phenols (e.9., butylated hydroxy
membranes (including bell jars) with enclosed toluene, BHT) or secondary aryl amines (e.g.,
oxygen scavenger chemicals such as Ageless diphenylamine derivatives). These react with
(Grattan l99l). This method may be suitable peroxy and alkoxy free-radicals to convert
for a few special objects in a museum, but it is them to non-radicals. Organophosphites and
useless for plastics in normal use. various organic sulfur compounds, such as
thiodipropionic acid, are effective peroxide
Oxygen-impermeable layers can be applied, decomposers. Organophosphites react with hy-
or allowed to develop by controlled exudation droperoxides to form alcohols and phosphates.
from the body of the plastic. Some wax "anti- Organic sulfides react with hydroperoxide to
oxidants" function in this manner during fab- form alcohols and organic sulfoxides. These
-
rication of the plastic a wax is incorporated that antioxidants are usually used in combinations
migrates to the plastic surface where it forms a with each other and with other stabilizers. A
thin oxygen-impermeable barrier layer that pre- typical stabilizer package may consist of pheno-
vents oxygen dissolution in the plastic. This is lic or amine antioxidant, a thiodipropionate
an example of an intentional exudation of an ad- ester. a phosphite, a metallic stearate (for lubri-
ditive. It is commonly used for rubber. Disrup- cation), and a metal deactivator (for chelating
tion ofthis barrier layer gives oxygen access to any traces of detrimental metal ions, usually
the plastic and oxidative degradation can occur. from polymerizatron catalyst residues).
Thus conservation treatments and handling pro-
cedures should be designed so that this surface
oxygen-barrier layer is not disrupted.
Table VI
Protection by Antioxidants Antioxidants
Antioxidants are compounds that inhibit or re-
Charn Propagatton Radtcal Termtnators
tard oxidative degradation of polymers by at-
mospheric oxygen during fabrication, storage, Alkylated phenols and polyphenols
and use. They are added during formulation - Butylated hydroxy toluene (BHT)
Secondary aryl amrnes
and become intimately mixed with the polymer.
They function at the molecular level by react- Hydroperoxrde Decomposers
ing with a particular functional group of a
Organophosphttes
nearby polymer macromolecule that is reacting, Thtoesters
or has just reacted, with oxygen dissolved in
Metal Deactrvators
the polymer. There are two classes of antioxi-
Hydrazdes
dants. Free-radical scavengers (radical or Tnazoles
chain terminators) react with chain-propagating - Benzotnazole

free-radicals before these free-radicals have a


chance to react with oxygen or polymer macro-
molecules, thereby preventing the formation of Polymers with hydrogens on tertiary carbon
more radicals and propagation of the chain reac- atoms, such as polypropylene and its copoly-
tions of oxidative degradation. Peroxide decom- mers, are particularly susceptible to oxidation
posers (sometimes referred to as secondary initiated by free-radicals produced by hydrogen
antioxidants or synergists) convert peroxides abstraction. During manufacture and process-
and hydroperoxides into non-radical and stable ing, unstabilized polypropylene yellows and
products before they become reactive free- changes in melt viscosity (a function of molecu-
radicals that propagate degradation reactions. lar weight). These polymers require antioxidant
stabilizers. Since they are exposed to high proc-
es sing temperatures, I ow volati lity c ompounds
Many compounds are used as antioxidants and
only the more common can be described (Table should be used. Typical antioxidants for
VI). A review by de la Rie (1988) describes polypropylene are high molecular weight,
polymer stabilizers and their reactions in low volatility alkylated phenols and polyphe-
nols with esters of thiodipropionic acid

143
(a sulfur-containing compound) and phosphites in polyrner chain breakage or chain-scission
at total concentration of 0.25o/o to l%o. and the formation of various oxygen-containing
species, such as peroxides, hydroperoxides,
Metal ions from catalyst residues, such as alcohols, aldehydes, and ketones. Physical
titanium and metal containing pigments (zinc, changes accompany these chemical changes,
iron, chromium, cadmium, etc.), induce hydro- leading to increased stiffness, insolubility, and
peroxide decomposition and cause undesirable exudation ofplasticizers (in the case ofplasti-
discolouration. Such metal deactivators as cized products). The degradation and stabiliza-
hydrazides and triazoles often are added in tion of PVC is described in great detail by Nass
conjunction with antioxidants to protect against /1976).
adverse effects of metals.
The main degradation reactions involve certain
Most hydrocarbon polymers (e.g., natural and labile chlorine atoms that are easily abstracted
synthetic rubbers based on polyisoprene, poly- from PVC molecules. The aims of stabilization
butadiene and its copolymers or blends like are to prevent the abstraction of chlorine atoms,
acrylonitrile-butadiene-styrene, polyisobuty- and/or prevent the reaction between chlorine
lene, etc., and such synthetic plastics as atoms and the polymer. This can be done by the
high-impact polystyrene that have a rubber following methods:
component) have unsaturated carbon-carbon
double bonds that are subject to oxidation that . exchanging chlorine atoms with stable
is accelerated by heat and UV light. This oxida- substituents that become more firmly bonded
tive aging is often characterized by yellowing to the PVC molecule
and embrittlement. These are commonly pro-
tected by 0.lo/o to 2.5%o of an antioxidant sys- r preferentially reacting the abstracted chlorine
tem generally consisting of organic phosphites with additives instead of PVC
and low-volatility hindered alkylidene bisphe-
nols often in conjunction with UV stabilizers. . scavenglng hydrogen chloride to prevent it
from catalyzing further reactions
Heat Stabilizers
o using antioxidants to decrease oxidation reac-
Heat stabilizers are added to polymers to retard tions in the PVC or the polyene sequences to
their decomposition when exposed to heat en- prevent chain-scission, crosslinking, and
ergy and mechanical stress during mixing, pro- other free-radical reactions
cessing, reworking of scrap, outdoor exposure,
and/or other storage or use conditions through- Table VII lists some of the more common types
out their service lives. Since they are most of heat stabilizers for PVC. Predominant in use
commonly used in PVC plastics, heat stabi-
lizers for this plastic will be discussed in detail.
Table VII
PVC degrades by a mechanism referred to as Heqt Stabilizers for PI/C
dehydrochlorination, in which hydrogen chlo-
ride is produced when hydrogen and chlorine Primary Stabrhzerc
atoms are eliminated from adjacent carbons on Banum/cadmium soaps
a PVC molecule. The acidic hydrogen chloride Calcium/zrnc soaps
catalyzes further degradation of the PVC and Lead soaps
Antrmony compounds
corrodes production equipment. The hydrogen Organotins
chloride elimination creates polyene sequences - Trn caboxylates
- Tin mercaptrdes and mercaptoesters
consisting of conjugated carbon-carbon double
bonds. This causes yellowing, which deepens Secondary (auxrlary) Stabr lrzers
through red and brown to black as degradation
Epoxy
increases. Oxidation reactions of non-reacted - Epoxrdrzed soya orl (ESO)
polymer molecules and of the polyene se- Organophosphrtes
quences of degraded polymer molecules result

144
among these is the group with heavy metal salts Thus, PVC stabilized as above, has a period
(zinc, cadmium, stannous tin, etc.) paired with with no deterioration while the barium (or
alkaline or alkaline earth metal salts (sodium, calcium) soap is present, followed by very
potassium, magnesium, calcium, barium, etc.). rapid deterioration after the barium (or calcium)
The synergism between the two types of salt soap is consumed leaving the cadmium (or
make the combinations much more cost-effec- zinc) chloride to catalyze degradation. This sort
tive than either single salt. of behaviour is exemplified by the deterioration
of vinyl rooftops on cars, which takes place
Primaty stabilizers are a group of compounds after about 5 to 10 years. On the degradation
that can be used as additives, by themselves, to profile this manifests itself as a lengthened in-
impart significant heat stabilization to a PVC duction period followed by a very rapid rate of
polymer. Secondary or auxiliary stabilizers deterioration in the second stage (e.g., Figure 3,
cannot by themselves impart stability, but can Curve B). It is the duration of the induction pe-
dramatically increase the efficiency of other riod that is of critical imoortance in the mu-
stabilizers by synergistic effects. seum context.

Primary Heat Stabilizers Among the earliest effective heat stabilizers


Barium/cadmium carboxylate soap systems for PVC were inorganic lead compounds (e.g.,
are the most widely used PVC heat stabilizers. basic lead sulfates, dibasic lead phosphite, ba-
Cadmium exchanges its carboxylate anion for sic lead carbonates, and complexes like basic
labile chlorine atoms of PVC creating a car- lead silicosulfate). Now there is a larger vari-
boxylate-containing PVC molecule that is ety ofboth inorganic and organic lead salts
stabilized against loss of other chlorine atoms (mainly soaps like basic lead phthalates, basic
in the form ofhydrogen chloride. This ex- lead maleates, and dibasic and normal lead
change forms cadmium chloride, which, stearates, which are also used for their lubricat-
unfortunately, catalyzes PVC degradation. Deg- ing properties). Lead functions in the same way
radation is prevented by reacting the cadmium as cadmium andzinc, stabilizing the PVC by
chloride with the barium soap also present in exchange of labile chlorine atoms with other
the heat stabilizer to make barium chloride and anions (e.g., stearates). Lead chloride is pro-
regenerate the cadmium carboxylate soap. Bar- duced, but, unlike cadmium andzinc chlorides,
ium chloride does not cause PVC to degrade. this does not catalyze PVC degradation. Eventu-
The regenerated cadmium soap is available for ally these stabilizers are consumed and become
more stabilization by labile chlorine exchange. ineffective. Lead stabilizers tended to be used
Eventually the barium soap is consumed by in opaque, rigid, and flexible PVC extrusion
reaction with the cadmium chloride. Then, no and injection moldings such as phonograph re-
barium soap remains to scavenge cadmium cords, conduit, pipe, and wire and cable. Lead
chloride, which then catalyzes PVC degrada- stabilizers are cornmon in wire and cable
tion resulting in catastrophic deterioration. insulation because of their electrical properties
Because these barium/cadmium stabilizers are (low volume resistivity) and because lead
toxic they are not likely to be found in objects chloride is water-insoluble and non-
that contact food and drug products. conducting.

Calcium/zinc soap stabilizers function in the Antimony stabilizers have been used for
same way as barium/cadmium soaps, but they about 25 years, especially in phonograph rec-
are non-toxic and have been sanctioned for ords. These have drawbacks because they are
food-packaging film, blister packaging, blow- unstable in UV light and, like lead stabilizers,
molded bottles, beverage tubing, blood bags, also suffer from sulfide staining (producing an
etc. Like the barium/cadmium system, the cal- orange stain) when in contact with any sulfides.
cium/zrnc system is effective as long as there is They are cost-effective because they have a syn-
calcium soap available to scavenge zinc chlo- ergistic response with metal soaps and improve
rides and regenerate zinc soaps. Eventually performance at lower concentrations.
the calcium soap is exhausted and degradation
proceeds.

t45
Recently commercialized organotin stabilizers Advantages and Disadvantages
are most efficient for rigid PVC, because, al- of Different Stabilizers lvlussurn
though more expensive, they excel in the most Consequences -
demanding high temperafure and high pressure Lead stabilizers are relatively inexpensive but
(high shear) fabricating processes that must be they have several drawbacks. They are opaque
used for non-plasticized PVC. These are based salts and cannot be used in clear plastics. Lead
on tetravalent tin (they are usually dialkyl tin compounds are toxic and do not have FDA ap-
esters) and are true organometallic compounds proval. They are also susceptible to sulfur stain-
where the metal is covalently bonded to carbon. ing, which may be of some consequence in
Organotin carboxylates like dibutyltin dilaurate museums. Early "plastic" objects are often
and dibutyltin maleate were among the earliest made of sulfur-vulcanized rubber. Storage of
organotin stabilizers. These have low toxicity, lead-stabilized PVC with sulfur-vulcanized rub-
low volatility, and good UV stability. Organo- ber may lead to sulfur staining of the PVC, that
tin maleate provides UV stability and sparkling is, the production ofbrown or black lead sul-
clarity to flexible and rigid PVC. Some alkyl fide by reaction of lead from the stabilizer with
tins have low toxicity and are sanctioned by the volatile sulfur-lulcanizing agents from the rub-
U.S. Food and Drug Administration (FDA) ber. Gaseous sulfur pollution may also cause
(e.g., di-n-octyltinbis-(iso-octylthioglycolate) lead staining. Additional sources of sulfur are
and di-n-octyltin maleate). Organotin mercap- other (more modem) PVC objects that contain
tides and mercaptoacid esters are liquids that organotin sulfide based heat stabilizers. Cad-
are used in the most difficult-to-process rigid mium and antimony stabilizers undergo similar
PVC applications, such as pipe and profile ex- reactions with sulfur to produce yellow and or-
trusion and injection and blow molding. These ange stains. Thus heavy metal stabilizers and
have poor UV stability relative to organotin sulfur-containing stabilizers cannot be used in
maleates and strong characteristic skunky the same product because of the formation of
odours. lead sulfide stains. Similarly PVC with heary
metal stabilizers cannot be stored with
Secondary Heat Stabilizers sulfur-containing plastics and rubbers, direct
Epoxy compounds with oxirane and ester contact especially should be avoided.
functional groups improve the performance of
heavy metal carboxylate soaps. These are pre- Ultraviolet Stabilizers
pared by epoxidizing natural unsaturated vege-
table oils like soya bean and linseed oil, or by When exposed to light, especially UV, some
epoxidation of synthetic esters of unsaturated plastics are degraded. To have an effect, the
animal or vegetable fatty acids (e.g., butyl-9,10- light must be absorbed by the plastic and must
epoxy-stearate and octyl-epoxy-tallate). Their have sufficient energy to rupture chemical
ability to accept chloride ions preferentially bonds. In general, UV light has sufftcient en-
(thereby preventing the formation of heary ergy and, for cerlain bonds, so does blue light.
metal chlorides that catalyze degradation) is the Certain functional groups, called chromo-
source of their enhanced performance. Thus the phores, absorb light and transform into
more the oxirane content, the more effective is energetically excited and highly reactive states.
the epoxy. Epoxidized soya and linseed oils are Because the functional groups have absorbed
also plasticizers that add low temperature flexi- UV light energy they contain more energy than
bility. These are also FDA sanctioned. in their unexcited normal or "ground" states.
This excess energy must be dissipated and the
Organophosphites (di- and trialkyl, di- and extent to which the plastic will, or will not, be
triaryl, and mixed alkyl aryl phosphites) are degraded depends on how this energy
most often used in conjunction with barium/ dissipation occurs.
cadmium and calcium/zinc stabilizers or some-
times with organotin or antimony stabilizers. Absorbed energy in the excited chromophore
They provide heat and UV light stability and can be dissipated by 1) non-damaging conver-
improve early colour and long-term UV light sion to heat (i.e., vibration of molecules),
stability. 2) non-damaging energy transfer to nearby

r46
molecules which in turn dissipate the energy o be compatible with the polymer
as non-damaging heat, or 3) degradative rup-
ture of weak chemical bonds in the chromo- o be stable at processing temperatures
phore molecule or in nearby molecules to
which the energy is transferred. . have low volatility so that it is not lost during
processlng
Different plastics vary greatly in their resis-
tance to light damage (Carlsson and Wiles o not contribute colour
1986). Some plastics, such as polycarbonates,
polyesters, and aromatic polyurethanes, have Choice also is influenced by other additives pre-
strongly absorbing chromophores as parts of sent, such as antioxidants, heat stabilizers, fill-
their molecular structures. Other plastics, such ers, and pigments.
as polyethylene and polypropylene, have no
chromophores in the pure polymer, but unfortu- Three types of uV stabilizers are used UV
nately they usually contain small amounts of absorbers (UV screens), energl transfer - agents
catalyst residues from polymerization, or ke- (quenchers), and fr e e -r' a d i c a I s c ave n gers. Each
tones and hydroperoxides from oxidation, of these goups contains many compounds that
which are chromophores and absorb energy, are effective UV stabilizers, some of which are
leading indirectly to bond cleavage and forma- listed in Table VIII. Synergistic enhancement
tion offfee-radicals in the plastic. In air, of stabilizing activity is often achieved by
free-radicals higger chain reactions leading to simultaneous use of stabilizers that act by dif-
more bond cleavage and destruction. Once initi- ferent mechanisms. Some compounds act by
ated by light, these oxidative chain reactions several different mechanisms. In his review of
require no light to continue, so it is important polymer stabilizers de la Rie (1988) describes
to prevent or inhibit the initial photochemical UV stabilizers and their reaction mechanisms
events that produce these radicals. Since photo- in greater detail.
chemical production of radicals leads to sub-
sequent oxidation reactions, antioxidants
(preferably UV-stable ones) are usually added
to retard thermal oxidative degradation by
Table VIII
Ultravi olet St abi lizers
scavenging free-radicals and chemically
inactivating peroxides. UV Absorbers

Carbon black
Thus there are two main mechanisms for
Zrnc oxrde
UV degradation: I ) direct rupture of chemical Substrtuted benzophenones
bonds by absorbed UV energy followed by (yellow drscolouratron)
Substrtuted benzotnazoles
rapid oxidation of the radical fragments, and Substrtuted acrylonrtnles
2) rupture in nearby molecules caused by en- Aryl esters
ergy transferred from UV-excited impurities, - Salicylates and benzoates

such as catalyst residues or oxidation products Energy Transter Agents


in the plastic. (Quenchers)

Cobalt and nrckel chelates


Choice of UV absorber stabilizers is governed (green/blue drscolouratron)
by the spectral region in which light absorption
Free-radical Scavengers
leads to degradation. This spectral region,
namely the "activation spectrum," is that por- Hrndered amrnes (HALS)
tion of the UV (and/or visible) spectrum, pri-
marily responsible for the degradation of a
particular polymer. A UV stabilizer should thus IIV Absorbers
do the following:
UV absorbers compete with the polymer for
absorption of the incident UV light. The UV ab-
. absorb strongly at the wavelengths sorber absorbs the UV light before or more effi-
of maxi-
mum sensitivity of the polymer ciently than the polymer molecule. The excess
energy in the UV absorber is converted into

t47
harmless heat (vibrational energy of the mole- Nickel chelate complexes are commonly used
cule). By absorbing the UV light at the surface, quenchers that function by absorbing LfV light
less UV light reaches the interior of the plastic. energy into their highly conjugated structures
It is important to note that this is essentially a and dissipating the energy as harmless IR radia-
surface phenomenon. Although the entire bulk tion by a resonance stabilization mechanism.
of the plastic may contain UV absorber, only They also have some hydroperoxide decompos-
that absorber in a surface layer measuring a few ing abilities. Unfornrnately they are inherently
tens of microns thick is the effective stabilizer. greenish or tan coloured. Some benzophenone
derivatives function both as direct IJV absorb-
Some inorganic pigments are efficient UV ab- ers and as energy-transfer agents, absorbing in
sorbers. Carbon black is most common and ef- the range of 230 nm to 350 nm. As noted, these
fective, but titanium dioxide andzinc oxide are have a tendency to yellow under processing or
also used. When high levels of inorganic pig- exposure to light.
ments can be used, additional UV stabilizer is
not required. The main disadvantage with these Free-radical Scavengers
pigments is that they produce opaque plastics. Free-radical scavengers do not absorb or block
UV light. They scavenge the free-radicals pro-
Colourless and transparent UV absorbers, duced by photolytic reactions. They prevent
such as benzophenone derivatives (e.g., ortho- degradation by reacting so quickly that the free-
hydroxybenzophenone) absorbing at 230 nm to radicals have no time to undergo subsequent
350 nm, andbenzotriazole derivatives absorb- photo-oxidative degradation reactions with the
ing at 280 nm to 390 nm, can be used in clear polymer.
plastics. Although these UV absorbers are not
consumed during their interaction with the UV Compounds typically used for this include
light, some are consumed by reaction with free- substituted tetramethyl piperidines, a class of
radicals (this is especially true for o-hydroxy- compounds referred to as hindered amine light
benzophenone, which is as commonly added stabilizers (HALS). These HALS have excep-
for its free-radical scavenger antioxidant tional activity that is attributed to mechanisms
properties as for its UV absorber properties). in which the active species are regenerated and
An unfortunate property ofbenzophenones and recycled. They are not consumed by their stabi-
benzotriazoles is that the products oftheir reac- lizing reactions and therefore have long life-
tions with free-radicals are usually highly col- times in the plastic. HALS function as energy
oured, causing yellowing of the plastic. Aryl quenchers, peroxide decomposers, and/or alkyl
esters including aryl salicylate, resorcinol radical terminators.
monobenzoate, and aryl esters of terephthalic
and isophthalic acid undergo light-induced Plasticizers
reaffangement to give photochemically stable
products, which are derivatives of2-hydroxy- A polymer is said to be plasticized when it
benzophenone which absorb in the range of is made more flexible by the addition of com-
230 nm to 350 nm and which are effective pounds to the base polymer. Typical plasti-
IJV absorbers. Unfortunately these also cause cizers are high-boiling organic liquids or
discolouration by formation of quinoid by- low-melting solids that are added to hard or
products, like the benzophenone derivatives. tough resins to impart softness or flexibility, re-
duce stiffness, increase impact resistance, and
Quenchers (Energy Transfer Agents) ease processing by lowering melt viscosity and
Quenchers interact with energetic excited-state fabricating temperatures. Camphor, the first
chromophores. Excess energy is transferred plasticizer, inhoduced about I 870, was used at
from the excited chromophore to the quenching concentrations of about 50o/o in cellulose nitrate
stabilizer, allowing the excited chromophore to to make celluloid. This formulation was the
return to its stable ground state. The excess en- principal thermoplastic until camphor supplies
ergy in the excited quencher is dissipated by from Japan were threatened during WW II.
non-damaging conversion to heat (vibrational Increasing scarcity combined with camphor's
energy) in the quencher molecule. objectionable odour led to its gradual

148
replacement with phosphate e sters, parti cularly softening action (plasticization) usually is attrib-
triphenyl phosphate. Phthalates were developed uted to their ability to reduce the intermolecular
in the 1920s. Di-2-ethylhexyl phthalate, com- attractive forces between the polymer macro-
monly referred to as dioctyl phthalate or DOP, molecule chains. The plasticizer molecules
currently the predominant plasticizer for all insinuate themselves between those of the
plastics, was patented in 1933. Modem plasti- polymer, thereby preventing the polymer-to-
cizers are used at concentrations of 10% to polymer interactions that create rigidity (i.e.,
25o/obut sometimes as high as 50o/o (Figure 4). increased tensile strength and decreased elonga-
Eighty-five per cent of plasticizers are used in tion). The plasticizer acts as a lubricant to facili-
PVC and these are most commonly esters of tate movement of the polymer macromolecules
carboxylic acids or phosphoric acid. Plasticiza- over each other and provides internal lubricity.
tion of PVC is reviewed in detail by Darby and Plasticizers have varying degrees of solvating
Sears (1976). action on resins. A plasticizer is a solvent that
is involatile.
Plasticizers are not chemically bound to the
polymer but merely dissolved in it. Plasticizer Different plasticizers have different efhcien-
cies different concentrations are required to
- the same level of plasticization. In gen-
achieve
eral, monomeric plasticizers are the most effi-
cient. Plasticizer efficiency may be measured
by a number ofphysical properties, such as ten-
E
-t''.\'.\ sile strength, modulus of elasticity, elongation
"'.i".
t'F at break, and hardness. Figure 4 shows the vari-
o
z ation of the mechanical properties of tensile
strength and elongation at break with plasti-
cizer concentration and plasticizer type. Hori-
J
6 zontal lines indicate the onset of plasticization.
zU
At higher concentrations plasticization is evi-
dent, whereas at lower concentrations the plas-
tic remains stiff and relatively unaffected by the
*

.-t;;ffi
2? -'
i plasticizer content. Figure 4 also shows that
; 200 different plastic izers have different effi ciencies,
0
z that is, at a particular concentration different
ro0
ot
,,'/4.4i2( compounds produce different levels of
flexibility.
lo 20 30
o/o
PLASTTCTZER CONC.
Plasticizers that exhibit good compatibility
are designated primary plasticizers. Those that
Figure 4 Efect of concentration ofvarious plasticiz- exhibit partial compatibility or that exude or
ers on tensile strength and elongation at break of bloom on standing are usually called secondary
PVC. Plasticizers: I, dicycloheryl phthalate; 2, plasticizers and cannot be used alone. The
phthalate polyester; 3, adipic polltester; 4, ticresyl
distinction between primary and secondary
phosphate ; 5, I iquid butadiene-acrylonitrile copoly-
plasticizers is vague, depending partly on poly-
mer; 6, di-2-ethylhexyl phthalate; 7, di-2-ethylhexyl
adipate (from Darby and Sears 1976). The concen- mer, concentration desired, and environment
tration at v'hich the curves cross the line marking the and conditions ofend use (see previous discus-
initial value of tensile strength or elongation indi- sion of plasticizers in section on additive com-
cates the onset ofplasticization. At any plasticizer patibility). These are external plasticizers,
concentration greater than this, the plastic will be compounds added to the base polymer. Internal
sofier and more flexible than in the non-plasticized plasticization can be accomplished by copoly-
state. For these plasticizers the concentration at on- merization of a flexible polymer with a rigid
set ofplasticization ranges from about 694 to 22% polymer. Vinyl chloride-vinyl acetate copoly-
plasticizer content. This is also a measure of the
mer is an example where the vinyl acetate im-
effciency of the plasticizer-the lower the onset
parts flexibility to the PVC. In this type of
concentrafion, the more efticient the plasticizer.

149
plasticization, no "external" compounds are of acids and alcohols. The acids may be cyclic
added to the base polymer. Instead a new poly- (e.g., phthalic, benzoic, trimellitic, terephthalic)
mer with greater flexibility is chemically pro- or linear (e.g., adipic, azelaic, sebacic, ethylhex-
duced by modifying the polymer backbone anoic, isodecanoic). The alcohols may be
with flexible units. Such "intemally plasticized" monohydric (e.9., 2-ethylhexanol, isodecanol,
plastics avoid plasticizer migration problems. butanol) or polyhydric (e.g., propylene glycol,
Internally plasticized plastics will not be dis- ethylene glycol, pentaerythritol). Furthermore
cussed fi.uther. the alcohol may be linear or branched. Aryl (cy-
clic) and alkyl (linear) esters ofphosphoric acid
Pressure exerted on plastics can cause are also cofiImon.
migration of plasticizers. This can lead to exu-
dation and blooming of plasticizers or to loss of Secondary plasticizers for PVC include various
plasticizer from specific areas in compression aromatic and mixed aromatic-aliphatic oils,
(Figure 5). This might cause excessive exuda- chlorinated paraffrns, some po ly(alpha -methyl
tion at the weight-bearing points of objects rest- sfyrene) derivatives, and inexpensive esters of
ing in storage. Also where there are snap high molecular weight alcohols and organic
fittings under compression, plasticizer canbe acids that have marginal compatibility with
lost. This leads to embrittlement. Examples of PVC.
this are the location where a doll's arm snaps
over a protruding socket at the shoulder, or The most common general purpose plasticizers
where a hose fitting over a tube is often held by are cyclic dialkyl phthalates. As the alkyl side
a hose clamp that adds more pressure and in- chain of the plasticizer increases in molecular
creases migration and subsequent embrittle- weight volatility decreases, water extractability
ment due to plasticizer loss. Plasticizer loss is decreases, oil extractability increases, effi -
usually accompanied by shrinkage resulting ciency decreases, processing becomes more
from loss of volume (i.e., that normally occu- difficult, and specific gravity decreases. Trimel-
pied by the plasticizer in the plastic). This leads litate esters (i.e., derived from trimellitic acid)
to cracking, again exacerbated in areas often- with branched or linear alkyl side chains having
sion or compression. 7 to 10 carbons are cyclic plasticizers ofhigher
molecular weight, which have low volatility
Composition of PVC Plasticizers and low water extractability. Major applica-
All primary plasticizers for PVC are esters, tions are in high temperature wire and non-fog-
functional groups that are reaction products ging automotive compounds for upholstery and
panels in car interiors. (Note: Fogging plasticiz-
ers are those that evaporate from plastic compo-
nents inside automobiles then condense as the
bluish oily film on the inside of the car
a<
windows.)
U;
6
o
) Linear plasticizers have improved low tempera-
N
ture performance. Their linear structure (no
9 rings) make them less compatible with PVC
F
and they tend to exude. Substitution ofalkyl
groups by aryl groups (rings) improves per-
formance. Alkyl benzyl phthalates are high
solvating plasticizers with increased
PRESSURE, kg,/cm2 compatibility with PVC. For example, butyl
benzyl phthalate is used in vinyl asbestos floor
Figure 5 Loss ofplasticizer due to pressqre on
PVC. Note that the pressure of 0.1 kg/cm' (which n tiles because it has less tendency to migrate
equivalent to that exerted bv a stack of 100 dimes or from the tile into the flooring adhesive, thus
quarters) can be easily achieved on objects in stor- preventing adhesive softening caused by the
age, d the entire weight of the objects rests on small plasticizer. Citrate esters are approved for food
area bearing points. (from Darb)t and Sears I 976) products.

150
Phosphate plasticizers impart flame resistance. phthalate, for plasticization. Triacetin may also
Triaryl (tricresyl, cresyl diphenyl, triphenyl) be used. Such materials have greater volatility
phosphates dissolve easily and are low in vola- than dioctyl phthalate (which is typically used
tility. As aryl groups are substituted by alkyl in PVC) and are more subject to loss by evapo-
groups, the properties ofthe plasticized plastic ration leading to embrittlement and cracking
tend to become less useful. For example, tri-2- due to shrinkage.
ethylhexyl phosphate is only good at low tem-
peratures, has high oil extractability, high Plasticizers with high solvating power for
volatility, and poor heat stability (needs PVC are suitable for poly(vinyl acetate),
stabilizer). PVAC. Dibutyl and butyl benzyl phthalates,
and glycol dibenzoates are widely used in
Polymeric plasticizers (polyesters) are usually PVAC. Phthalates and dibenzoates have been
based on linear dibasic acids plus polyols (espe- detected in several PVAC adhesives (white
cially glycols). Their peffnanence improves glues) analyzed in the Adhesive Testing Project
with increased molecular weight. They gener- at CCI (Down and Williams 1988).
ally have low volatility, good to very good ex-
traction resistance to aqueous and hydrocarbon Toluene ethyl sulfonamides are used to soften
solvents, but have poor low temperature per- some grades of nylon. Poly(vinyl butyrals) are
formance and are inefficient and difficult to most often plasticized with glycol derivatives,
process. Epoxy plasticizers are generally epoxi some adipates, and some phosphates. Alkyl
dized oils (linseed and soya bean) and alkyl ep- benzyl phthalates and phosphates are used in
oxy esters (tallates, oleates, linseed oil acids) acrylics. Polprethanes can be plasticized with
with performance properties like polyesters. many of the same compounds that are used
The presence of the oxirane goup allows these for PVC.
to also serve as auxiliary heat and light stabi-
lizers. Some soya and linseed oil epoxies are Extraction of Plasticizers
also FDA approved. Extraction of plasticizers from PVC by coat-
ings, such as nitrocellulose, can mar finishes
Aliphatic and aromatic hydrocarbons and and cause discolouration and tackiness. This
chlorinated paraffins are generally not compat- problem can be greatly reduced by the use of
ible with PVC. These are used as minor compo- plasticizers with little or no affinity for the
nents of a plasticizer system as extenders to coating and by the selection of plasticizers with
reduce cost in those few applications where resistance to migration. Isophthalates (DIOP)
they provide advantageous end-use properties. and terephthalates (DOTP) have very low
migration to nitrocellulose coatings and offer
Plasficizers for Plastics Other Than PVC a marked improvement over all orthophthalate
The amorphous character of PVC allows esters, but are very expensive.
pasticization over a wide range of plasticizer
concentrations, whereas the higher crystallinity As a consequence of these mutual compatibili-
of other polymers means that the region in ties, it is essential that objects made of different
which plasticizing is possible is smaller. This polymers (such as these) be kept out ofcontact
means that plasticizer compatibility must be with each other. Plastic2er migration and plasti-
critically evaluated. Plasticizers must be chosen cizer transfer will occur, usually with softening
with care. and increased tackiness ofobject surfaces. This
also applies to the choice of storage materials.
Generally the plasticizers used for poly(vinyl Storage materials must be carefully chosen to
chloride) work well for cellulose nitrate, resist damage by plastic objects and their
ce llulose acetate propionate, cellulose acetate additives.
butyrate, and ethyl cellulose. Cellulosics are
best plasticized with dibutyl phthalate, alkyl- Conclusion
oxy phthalates, triaryl phosphates, and sul-
fonamides. Cellulose acetate requires lower When collecting, displaying, and storing plastic
alkyl phthalates, such as dimethyl and diethyl objects it must be remembered that plastics are

l5l
complex mixtures of polymers and additives. Nous decrirons donc brievernent, dans le cadre
Changes in either the polymer or any of its addi- de la prdsente communication, les principaux
tives will result in changes in the plastic, usu- groupes de polymires auxquels appartiennent
ally in a way that is described as degradation. les plastiques qui entrent dans la composition
This paper describes the wide variety of addi- des objets de musie les plus typiques, pour en-
tives that are used and the delicate balance of suite traiter dans le detail des additifs qui sont
properties that are required to ensure that all prdsents dans les plastiquesfabriques d partir
the ingredients are, and remain, compatible in a de ces polymdres. Les rtsultats d'analyses
well formulated plastic. The effects of antioxi- chimiques ricentes, portant sur des objets en
dants, heat and light stabilizers, and plasticizers plastique dtgradis, seront utilisbs pour mon-
on the longevity ofplastics has been discussed trer I'incidence des additifs et des polymdres
in some detail. Other additives not discussed de base sur la longivite des plastiques. Leur
here, such as colourants (pigments and dyes), digradation, de m€me que d'autres dommages
mineral additives like fillers, extenders and rein- qui leur sont causis accrition, cristallisa-
forcing agents, and processing aids like anti- -
tion, craquellement, fssuration, ddcoloration,
static agents, intemal lubricants, mold release fragilisation, suintement, ramolliss ement,
agents, and slip agents, also affect the longevity etc. est lide d une modification ou d une
of a plastic, although probably not to the same perte-,d'additifs. Et les conditions ambiantes ne
degree as those discussed. Degradation profiles sont en outre pas ttrangdres d de tels change-
were introduced as an aid to understanding the ments. Nous fournirons enfin des lignes direc-
states and stages ofdegradation ofplastics as trices gmtrales sur la mise en rdserve des
determined by the rates and extents ofdegrada- objets en plastique, qui permettront, d'une
tion processes occurring in aging plastics. The part, d'accroitre leur longtvite et, d'autre part,
ways that additives and changes in additives d'empdcher qu'ils ne causent des dommages
can influence the shape of the degradation pro- aux m ati eres envi ro n na n tes
file and thus the longevity of the plastic object
were discussed. References

R6sum6 Barton, A.F.M. CRC Handbook of Solubility


Parameters and Other Cohesion Parameters
La composition des objas en plastique (Boca Raton, Florida: CRC Press, Inc., 1983).
et leur longevitd
Carlsson, D.J. and D.M. Wiles, "Degradation,"
Les objets en plastique sont gendralement in: Encyclopedia of Polymer Science and Engi-
ddcrits d'apris leur principal constituant po- neering, Volume 4 ed. H.F. Mark et al. (New
lymdrique unefeuille de polyithyldne, un York John Wiley & Sons, 1986) pp. 630-696.
-
rev€tement en vinyle ou en polychlorure de
vinyle, une sculpture en acrylique ou une bofte Darby, J.R. and J.K. Sears, "Theory of Solva-
en polystyrdte ou en styrdne, par exemple. Or, tion and Plasticization," in: Encyclopedia of
tous les articlesfabriquis en plastique sont PVC, Volume l, Chapter I0 ed.LJ. Nass (New
constituts de prtparations complexes de York Marcel Dekker, lnc.,1976) pp. 385-503.
polym,ires, auxquels viennent s'ajouter les ad-
ditifs necessaires pour conferer au polymdre de Darby, J.R., N.W. Touchette and K. Sears,
base les propridtds qu'exige son utilisation fi- "Dielectric Constants of Plasticizers as Predic-
nale. Et les additifs notqmment les plasti- tors of Compatibility with Polyvinyl Chloride,"
fiants, les stabilisants, les colorants, les agents Polymer Engineering and Science, October
favorisant le traitement, etc. sont tout aussi 1967 , pp. 295-309.
importants que le polyrndre de - base quand il
s'agit de determiner la durte de vie utile de de la Rie, R., "Polymer Stabilizers. A Survey
tels objets ou la ptriode durant laquelle ils with Reference to Possible Applications in the
pourront €tre conservds dans un muste. Conservation Field," Studies in Conservation,
vol. 33, no. 1, 1988, pp.9-22.

t52
Down, J.L. and R.S. Williams, "Report on Ad- Grattan, D.W., "The Oxidative Degradation of
hesive Testing at the Canadian Conservation In- Organic Materials and Its Importance in the
stitute," Symposium '88 Proceedings (Ottawa: Deterioration of Artifacts," Journal of the Inter'
Canadian Conservation Institute, in press). national Institute for Conservation - Canadian
Group,vol. 4, no. 1, 1978, pp. 17-26.
Feller, R.L., "Studies on the Photochemical Sta-
bility of Thermoplastic Resins," Preprints of Grattan, D.W., "Degradation Rates for Some
the 4th Triennial Meeting of the International Historic Polymers and the Potential of Various
Council of Museums Committeefor Conserva' Conservation Measures for Minimizing Oxida-
tion Committee. Venice. 7 5 12214, 197 5, tive Degradation," this volume.
pp. l-11.
Hedley, G., "solubility Parameters and Varnish
Felleg R.L., "Stages in the Deterioration of Or- Removal: A Survey," The Conservator,vol.4,
ganic Materials," in Preservation of Paper and 1980, pp. l2-18.
Textiles of Historic and Artistic Value. Ad-
vences in Chemistry Series, No. 164 ed. J.C. Horie, C.V., Materials for Conservation :
Williams (Washington, D.C.: American Chemi- Organic Consolidants, Adhesives and Coatings
cal Society, 1977) pp. 314-335. (London: Butterworths, 1987).

Feller, R.L. and D.B. Encke, "Stages in Deterio- Nass, L.I., "Theory of Degradation and Stabili-
ration: The Examples of Rubber Cement and zation Mechanisms," in: Encyclopedia of PVC.
Transparent Mending Tape," tn'. Science and Volume I, Chapter 8 ed. L.I. Nass (New York:
Technologlt in the Sertice of Conservation. Pre- Marcel Dekker, Inc., 1976) pp.27l-293.
prints of the Contributions to the Washington
Congress, 3-9 September 1982 ed. N.S. Brom- Nass, L.I., "Actions and Characteristics of Sta-
melle and G. Thomson (London: International bilizers," in: Encyclopedia of PVC. Volume I ,
Institute for Conservation of Historic and Artis- Chapter 9 ed.Ll. Nass (New York: Marcel
tic Works, 1982) pp. 19-23. Dekker, Inc., 1976) pp.295-384.

153
The Nature and Origin of Surface Veneer
Checking in Plywood

Mark D. Minor

Department of Obj ects


C onserv ation
TheMetropolitain Museum of Art
New York, N.Y.
U.S.A.

Abstract By the third week, checks extendingfrom the


original lathe checl<s had extended to approxi-
Pllnuood is a material occarring with greater matelv 7596 to 8096 of the veneer thickness. By
frequency in both decorative andfine arts collec- the ninth week, several cracks had reached the
t i o n s. LThi I e fund am ent a I ly w o o d, i t s s t ru ct ure sudace of the panels. At week 20, a majority of
and characteristics fall into a completelv different the cracks had extended to the surface, regardless
realm than solid timber. One degradation prob- of the severity of the moisture change.
lem evident in many pieces offurniture and paint-
ing supports are longitudinal surface veneer A modelforformation of checks is suggested,
checks. These cracks are disfiguing to the dis- based on compressive damage to the wood cells
play surface, they break up the continuity ofthe caused by moisture driven expansion, followed
surface coatings, and they expose the underlying by tensile stress in subsequent shrinkage. The con-
layers and adhesives to air, moisture and grime. cept that plywood is not an inherently stable mate-
As a precursor to responsible treatment of these rial but afragile one is stressed. Implicationsfor
objects, the nature and mechanism of these conservation are discussed, in particular the po-
checks need to be understood by conservators. tential dangers ofsurface stabilization and
reintegration.
The study to be discussed attempted to show a
tangible connection between an inherent vice in Introduction and Objective
rotarv-peeled plvwood veneer (lathe checks) and
the formation of surface checking in pfi,wood Plywood is a material showing up with greater
panels. It consists oftwo sections: a survev ofre- frequency in decorative arts collections. While
search literature on the mechanics ofthe veneer plywood is fundamentally wood, its structure
cutting process combined with a microscopical
and characteristics make it completely different
examination of plywoods qtcled to equilibrium
from solid timber. One problem evident in
at different moisture contents. A sample group of
many pieces of l9th- and 2Oth-century objects
p[u:oods was cycled through two difering mois-
ture extremes, representing an acceptable and an that incorporate plywood is longitudinal sur-
extreme climatic change, for 20 weeks. Periodic face veneer checking. These checks are disfigur-
samples were taken and examined microscopi- ing to the surface, they break up the continuity
callt in cross-section to assess any changes to ofthe surface coatings, and they expose under-
the wood structure. lying layers and adhesives to air, moisture
and grime.
Damage to the wood structure occurred much
more quickly than expected, even in the accept- To devise optimal treatment methods and pre-
able 3594 to 5994 relative humidity (RH) cycle. ventive conservation measures, the nature and

r55
origin of veneer checks need to be fully under- severe situation can be seen in a close-up ofthe
stood. Specifically, it needs to be ascertained chair by Charles Eames (Figures 3 and 4)
whether (and to what extent) the checks are a which he termed "DCW" @ining Chair Wood).
result of relative humidity (RH) fluctuation. Also, the effects of surface checking can be
(RH fluctuation could cause mechanical dam- seen on many painted plywood panels, both
age on the cellular level, which would be ob- new and old. Obviously, where ground and
served as checks and these checks could be paint layers are not flexible enough to remain
augmented by processing methods, such as intact over a developing check, they will crack,
rotary lathe checking.) in some cases much to the detriment of the
pictorial effect of the painting.
The purpose of this study is to provide empiri-
cal information on the nature and origin of sur-
face checks in plywood. The study will seek a
connection between rotary-peel lathe checks in
tangentially cutveneer, and surface check for-
mation in plywood panels composed of such ve-
neer. This will be shown by exposrng plywood
to numerous high/low extremes of RH, and sub-
sequently examining it in cross-section. The
study will underscore the misconception that
plywood is an inherently stable material. It
will also help in the construction of a plausible
model that explains the propagation of surface
checks in plywood, in relation to the veneer cut- Figure 2 Detail ofsurface veneer checking, Breuer
tabl e. Note the accompanying finish film cleavage.
ting process, and the hygromechanical proper-
ties of wood. It is hoped that the study will
create interest in plywood degradation among
other conservators, and provide some prelimi-
nary recommendations on approaches to the
conservation of ply.wood.

Consider some examples of surface checks in


plywood objects (Figures I and2). The frst is
a 1936 table by Marcel Breuer. The surface
checks are evident on this piece, and have exac-
erbated a condition of brittle delamination of
the cellulose nitrate hnish. A similar but less
iiil:t,ii:uii:iJlt::iiJJ\i]jjJ:ii\

Figure I NestingTable, Mqrcel Breuer, 1936. The


Metropolitan Museum of Art, Gtft of Daniel and Figure 3 DCW Chair, Charles Eames, 1946.
Suzanne Geller in memory of Bert and Phyllis Geller Collection of the Fine Arts Museum of
(1e8e.s2). San Francisco (I 987.62. I ).

156
grain direction would render such designs
useless.

The third advantage, ofgreatest interest here, is


the dimensional stability of plywood. Plywood
is widely considered by non-specialists to be ul-
timately stable, that is, absolutely immune from
dimensional movement, since its alternating
layers effectively lock each other into place. ln
truth, it is not dimensionally stable. In going
from the green state, or with moisture content
at fiber saturation point, to the oven-dry condi-
Figure 4 Detail ofsurface veneer checks, tion, plyr,vood experiences a dimensional
DCW Chair. change that is but a fraction ofthat ofsolid
wood.^The shrinkage is 0.2o/o compared to
Plywood as an Art Material 8.0o ,2 smaller by a factor of 40. Because of its
significantly reduced potential for hygroscopic
Plyrvood has a diverse history of usage, which dimensional changes, plywood panels can be ef-
can be seen as a response to its unique charac- fectively glued to perpendicular or cross-grain
teristics. Plywood embodies at least three key frame elements. Such construction is either dan-
advantages. The flrst advantage lies in its gerous or impossible in solid wood, as can be
strength properties, which show remarkable seen in numerous fumiture case constructions,
advantages over solid wood of equivalent di- where drawer rails and leg posts have been fas-
mension.' Since plywood, by definition, con- tened tight across the width of the sides. Inevita-
sists ofan odd number ofveneer plies bound bly, such constructions lead to compression set
at 90o angles to one another, the weakness of the panel as relative swelling, caused by
inherent in a thin piece of wood that of stiff- moisture absorption, exceeds the proportional
ness or modulus of elasticity perpendicular to limit in compression of the wood. (Concise de-
the grain direction is mitigated. Plywood has scriptions of the nature of compression set are
virtually isotropic (i.e., directionless) strength given by Hoadley'and Stamm.") Since longitu-
properties in its "sheet" or planar dimension. dinal compressive and tensile strengths exceed
Obviously, a thin sheet of plywood, when incor- those perpendicular to the grain direction, sub-
porated into a seat, will support a much greater sequent shrinking results in tensile cracking
load than a piece of solid wood of similar size; across the grain.
a distinct advantage, elegantly exploited by
such designers as Alvaar Aalto in the 1930s. Plywood gains in strength and stability relative
to solid wood through its cross-banded struc-
A second advantage is the ability of pll.wood ture. Forced stability in plywood is not without
to be formed. Again, it is the fundamental pre- its drawbacks, however. The restricted dimen-
condition of being laminated that allowed users sional change (i.e., compared to solid wood) in
to modify plywood from flat sheets to fantastic- relation to moisture content change is the cause
ally curved forms like the Arabesque table of of many of the problems apparent in art and
Carlo Mollino. artifacts incorporating pl1.wood.

Such curved forms rely on the adhesive bond Lathe Checking and the Veneer
between laminae to "set" the shape upon cur- Cutting Process
ing. Solid wood could be processed to equiva-
lent shapes, but at great effoft, and again, Surface checking can be found on weathered
would maintain its unfortunate perpendicular- solid wood ut *E[ as plywood.s The mecha-
to-grain weakness. In addition, solid wood nisms involved in either case are not obvious.
shaped into tight curves would incorporate For plywood, it is also not obvious from which
areas of "short grain," wherein the far weaker direction (i.e., surface downward, or flom the
shear strength of solid wood petpendicular to

r57
undemeath working upwards) the checks upward and forward as a result of the cutting
evolve. The obvious culprits are lathe checks. wedge entering the tissue. Since the resulting
chip is effectively levered upward, there is a
Since the I 890s, rotary lathe cutting has been noteworthy compression on the upper surface
the preferred manner to obtain the large sec- of the wood blank.
tions of venee,r needed to form sheets for mak-
ing plywood.o A log or "blank" of wood, first The combined stresses form a shear force,
boiled or steamed for softening, is spun against resulting from the very complex mixture of ten-
alarge fixed knife, cutting a carefully control- sion and compression. The shear plane is ori-
led thickness ofcontinuous veneer. Such ve- ented at approximately 60o to the direction of
neer will obviously be sliced as a tangential cut. At some point in the cutting process, the
spiral from the blank. Unforfunate by-products shear stress will exceed the tensile strength of
of this cutting action are lathe checks, which the wood cells, and failure (i.e., a lathe check)
form on the downward or knife side of the ve- will occur. At this moment, the stresses are re-
neer being formed. Lathe checks are periodic, lieved by the rapid increase in strain (deflection
radially oriented cracks in veneer, resulting upward of the veneer). The stresses will be cy-
from a complex interaction of compressive _ clic, since both the tension and compression
and tensile stresses during rotary lathe cutting/ will increase as a chip is forced up the knife
(Figure 5). From early on, lathe operators face, and be released somewhat after failure,
learned to optimize the settings of the lathe and only to increase again as the cut continues.
the pre-conditioning of the wood to minimize
lathe checking. Refinements such as the nose- The result is a sheet of veneer with periodic
bar, a pre-compression apparatus, fupher re- checking on its under (or loose) side. Even with
duced the formation of lathe checks.d modem refinements, production of rotary-cut
veneer without knife checks is a theoretical pos-
sibility, and in practice, all such veneer will
contain checks to some extent.

The Hygromechanical Properties


of Plywood
ri
How might the hygromechanical properties of
plywood affect its degradation? As can be seen
trlI iri r" in the following data, there is a notable differ-
rtr'
Ir-t+\ ence in the equilibrium moisture content
(EMC) of plywood in comparison to that of,
solid wood up to the fiber saturation point.'"

Figure 5 Stress distribution duringveneer Plywood EMC RHo/o Solid Wood EMC
cutting. "C" denotes compression, "T" tension,
and "M" moment. (After Lenev 1960). 6% 40 -8%
10% 70 -13%
28k 100 -28"k
Theoretical work begun in the 1940s continues
In simple terms, plywood is less hygroscopic
to shed a great deal oflight on the origin of
than solid wood. It is prudent to consider what
lathe checks. There are a number of studies on
might cause this phenomenon. A ready explana-
this topic that are noteworthy.g-l3 Thir research
tion is the volumetric presence of a great deal
proved the phenomenon to be extremely com-
of adhesive. For a given volume of plywood,
plex. Based on these findings, a very idealized
compared to an equal volume of solid wood,
and simplified veneer cut can be summarized
more water-impervious material, in the form
here. As a knife enters a block of wood (Figure
ofadhesives, is present. But this is only part of
5), various compressive and tensile stresses
the answer.
become apparent. Wood cells are forced both

158
A more interesting explanation is that of demonstration was devised to enable a cross-
reduced hygroscopicity: wood under high com- sectional viewing of wood cells in plywood
pressive stress experiences a small but signifi- through numerous cycles ofRH changes. Con-
cant reduction in hygroscopicity, known as the ditions of extreme RH that an artifact might
Barkas effect.'' Obviously. individual plies, be- actually be exposed to were established. (The
ing crossJaminated, are restrained ffom swel- soaking, freezing and boiling associated with
ling in the tangential direction by the adhesive industrial standards were avoided. For an inter-
bond to the adjacent ply (or plies). esting critique ofindustrial test standards in
plyw6od. see Canoll.l8) Two relative humidity
We can deduce a situation in which hyg.o- regimes were chosen: a moderate one of 33%
scopicity reduces as RH and, therefore, mois- to 59%io RH, and a more drastic yet still realistic
fure content (MC) increases. It follows from one of l2o/o to 85% RH. Small chambers were
general principles that as long as there is any constructed, using saturated salt solutions to
degree of hygroscopicity in wood, any rise in create the two regimes. (The following satu-
RH must be accompanied by a certain degree rated salt solufions were used: lithium chloride
of swelling. It is easy to see the somewhat can- l2Yo,magnesium chloride 33%, sodium bro-
celing effects that further swelling would have mid,e 59o/o and ammonium sulfate 85%.)
on further hygroscopicity. But the reduction in
hygroscopicity will not be enough to account By monitoring the equilibrium moisture con-
for the reduction in swelling.seen in plywood tents (EMCs) of the samples, it was decided
as compared to solid wood.'o that a 3.5 day changeover time (i.e., a complete
low-high cycle took one week) was adequate.
Where does the swelling go? This is not a naive Weights were recorded at both the low and
question. On a cellular level, one theory hy- high EMCs to determine whether a significant
pothesizes "inward swelling, " where restrained gain or loss in weight occurred during the test.
wood is forced to expand "inside" into the cell At the end of the exposure period to low humid-
lumen, or the empty space. gt the center of ves- ity (EMC at minimum, shrinkage at maximum)
sels, fibers, and fracheids.' 'Theoretically. on a several circular samples (6 mm in diameter)
molecular level, enough compression might in- were removed from each test piece.
duce a physical/chemical change in the wood
tissue, with a redistribution of bonds between Two types of plywood samples were used in
polymeric cellulose chains within the lignin this demonstration. The first, representing a
matrix. Both models imply permanent set (this "new" piece of plywood, was commercial
phenomenon is known in the world of wood Baltic birch plywood. This consisted of nine
technology as compression set). As compressed plies of rotary-cut birch veneer, glued with a
wood shrinks in a decreasing RH, it will do so hot-press urea-formaldehyde adhesive. The sec-
from its compressed dimension, effectively try- ond, representing an "old" or historic plywood,
ing to further reduce its size. But plywood is was taken from an early 2Oth-cenflry object. It
bound, one layer to another. Eventually, the consisted ofthree plies oftangentially cut birch
compression of the swelling surface veneer veneer, with a casein or soya adhesive.
will become a tension that exceeds the tensile
strength of the shrinking veneer. In this man- All the samples were embedded, and examined
ner, through numerous MC cycles, we can hy- microscopically, in both reflected visible light
pothesize that tensile cellular failure might and long-wave ultraviolet illumination (UV ex-
occur in the surface veneer. This failure may citer wavelength 365 nm, splitter wavelength
perhaps eventually become a full-fledged check. 395 nm, barrier wavelength 420 nm). The con-
dition(s) of the surface veneers were noted and
An Empirical Demonstration photographed. Long-wave UV has been pre-
of Check Formation viously used in wood research to examine the
wood-adhesive interface and the onset of fun-
As a means of visually gauging the onset and gal degradation.l9'20 It has the advantage
propagation ofsurface checks in plywood, a of showing the lignin-rich areas of wood

r59
structure, and in this work aided the detection
ofcheck proliferation and cell wall breakage.

Results of Demonstration
l. The direction/evolution/resolution of lathe
checks.

The lathe checks in the new plywood, while


quite variable in nature, extended approxi-
mately one-third of the veneer thickness
(Figure 6). The dark lumens seen in the photo-
micrographs are those completely infused Figure 7 Typical lathe check in the uncycled "old"
sample. Note the nonlangential orientation of the
with embedding medium, and tend to indicate
upper veneer, as well as the compression visible in
increased cellular permeability due to damage. vessel elements.
Lathe checks occur with a spacing that is
roughly equal to the thickness ofthe veneer,
and extend upward at an angle ofapproxi- the samples was the extension of existing lathe
mately 45o. There appears to be a great ten- checks (Figure 8). After this first cycle, the
dency for lathe checks to resolve or end at lathe checks had extended a great distance into
some barrier point, such as a ray, a row of late- the veneer, in some cases to within 10 cells of
wood cells, or a vessel element. Figure 7 shows the surface. But most noteworthy was that the
a lathe check zone in the old plywood sample, degree of damage was largely independent of
the lower check resolvins at a ravlvessel the amplitude of RH variation. Note the dam-
intersection. age to the sample in the "extreme" cycle seen
in Figure 8, compared to the sample from the
"moderate" or regime more likely to be seen in
a museum, in Figure 9. There is an alarmingly
similar degree of damage. This very pointedly
underscores the sensitivity of plywood to
swings in RH in the early stages of its life. The
sample shown in Figure 10 again demonstrates
the possibility of extreme damage in one cycle.
Note the crushed layer of cells at the top of this
sample they are a result of compression and
-

Figure 6 Typical lathe check in the uncycled "new"


plywood, extending to the left at an angle, then
changing direction towards vertical. Note extent oJ'
adhesive layer penetration. (N.B.: all photomicro-
graphs are at 150 x.)

2. Checking evolves from lathe checks, much


more quickly than would be expected.

An original goal of this study was to visually


determine from which direction, up or down, Figure 8 Cycle #1, extreme RH range ("new" ply).
veneer checks originated and evolved. By the Note extension of crack system, and middle lamella
completion of the first humidity cycle, it was (inter-cellular) failure in vertical section at lower ar-
evident that the primary change occurring with row. The check terminates at a rqv. near the suiace.

160
shown in Figure 14, vtrtually no damage that
can be directly ascribed to this humidity cy-
cling occurred, since extant surface coating ma-
terial can be seen penetrating the check. This
indicates that it is an old check. The great de-
gree of variation (Figure l5) in damage to the
older samples makes it difficult to connect any
damages noted with this demonstration with
particular causes. Since lathe/surface checks of
many different lengths and widths were noted
at the start, it became impossible to attribute
any widening or lengthening to further RH cy-
Figure 9 Cycle#1, moderate RH range ("new" ply) cling. It was apparent, however, that the fre-
Photo shot4)s relative damage after one cycle in the quency of checking did not increase, that is, it
moderate RH range. Top margin oJ photo cone- did not appear that any new cracks developed
sponds to surface of the veneer. during the demonstration in either type of sam-
ple. This strongly suggests that checks are only
associated with original lathe checks, since
no new checks developed, either from above
or below.

Figure l0 Cycle #1, extreme RH range ("new" ply).


Extension o;f crack system to within I0 cells o/
surface. Nctte crushed cells at surface infused with
embedding medium.
Figure I I Cycle #2, moderate kH range ("new"
ply). This sample shows no more damage than that
machine surfacing during manufacturing, and ofthe first cvcle.
are not directly related to MC cycling.

3. Very little change occurs after the initial


damage.

Aside from a slight "widening" of the area of


damage within a crack system, and a gradual in-
crease in the length and definition ofthe cracks,
very little change was noted in 18 humidity cy-
cles for both plywoods. Figures I I to 16 show
a series of cross-sections of plywood samples
exposed to an increasing number of RH cyclic
changes. In cycle 9 (Figure 12) the crack has
reached the surface, while another sample (Fig- Figure l2 Cycle #9, moderate RH range ("new"
ure 13) exposed to 18 cycles shows a check re- ply). Typical damage after nine cycles. Check is
solved at aray somewhat further down, albeit effectively to surJace. Note wider "footprint" oJ'
near the surface. In the aged pl1'wood sample crack system.

161
Figure I 3 Cycle #18, extreme RH range ("new" Figure l5 Cycle #18, moderate RH range ("old"
ply). Note long, vertical, middle lamella failure, as ply). No extended lathe checks are visible in this
well as severe and abrupt changes of direction. sample; only surface damage in conjunction with
vessel elements can be seen.

Figure 14 Cycle#6, moderate RH range ("old" ply).


Note the sur/hce coating moterial penetrating the Figure l6 Cycle #18, extreme RH range ("new"
check at the su4face. ply). Full check extension.

Two or three moderate humidity cycles pro- plausible explanation of the decline in crack
vided enough compressive stress to extend a propagation with increasing RH cycles is that
lathe check nearly to the surface, yet another l6 the lengthened crack becomes its own expan-
cycles caused, at most, the crack to be barely sion "release valve." Cracks provide a space to
complete and visible only under high magnifica- absorb expansion before it causes compressive
tion (Figure l6). Since the percentage of weight stress leading to fuither failure. It follows that
gain did not change noticeably with ftrther cy- as the crack lengthens or widens, the relative
cling, the amount of expansion due to adsorp- degree ofexpansion needed to induce compres-
tion must remain virnrally constant as shown in sive damage would increase. Hence the slow
the data below. progress ofthe crack after its initial extension.

Weiqht Gain Due to Moisture UDtake bv Plvwood 4. Checks do not typically evolve at ray
Samole Group Weight Gain % (Mean) interfaces on the surface and work downward.
new plv, extreme cvcle 2.56
old plv, extreme cvcle 3.58 One theory ofcheck propagation predicts that
new ply, mgderate cycle o.74 differing cellular orientation would make the
old ply, moderate cycle 0.84 interface of rays with vessels and fibers the
primary location of check formation.'' The
(Note the difference in moisture uptake be- reasoning is that any tangential or radial shrink-
tween the "new" and "old" sample groups.) One age in vessel or fiber tissue would be restrained

162
by the radially orientated rays, causing a local- degree ofRH change will be enough to propa-
ized stress zone. No downward extending gate checks that extend throughout the thick-
checks were ever noted at a raylfi,ber or raylves- ness ofplywood veneers, and in a short period.
sel interface. After l8 "extreme" cycles, the On the bright side, it also has been shown that
new plywood showed no surface damage (i.e., while damage may develop quickly, it occurs
damage other than the crushed outer layer of on a relatively minor scale, and the propagation
cells, an occurence that is independent ofhu- of damage is slow. While checks were gener-
midity cycling). In the old sample only, how- ated during the test, at no time were any
ever, a small number of surface checks of gaping checks produced, such as those on the
varying sizes were noted from the start. These Breuer table.
were always associated with a vessel element
(Figure 15). The fact that these checks often It is hoped that the importance of paintings in
extended fairly deep implied that they were not most collections will help to underscore this
merely a phenomenon caused by mechanical point, for we only have to look at a small sam-
surfacing. The frequency ofthese checks did pling of the many paintings executed on ply-
not change throughout the demonstration; in no wood grounds to see the effects ofsurface
instance was a surface check seen to connect veneer checking. For paintings on plywood, it
with a lathe check to become a full-thickness is not only the wood surface that is disfigured,
check. While surface checkingproper onply- but the decorative paint layers as well, quite
wood does exist, its frequency ofoccurrence is possibly to the great disfigurement of the pic-
far less than that of surface damage resulting ture. Just as furniture finishes tend to give way
from the extension ofa lathe check. This latter with the evolving check, so do the ground and
damage seems more in line with the explana- color layers in paintings.
tion of weathering offered by Feist and Hon for
solid wood.' No other explanation of the origin This leads to an important point with respect
ofthese checks can be offered. to recommendations. What do we do when we
are "too late"? As long as the RH variability
Conclusion and Conservation that causes checks continues, any filling of
Recommendations these checks will exacerbate the crack. The
fill provides a wedge for further compression,
It should be evident that plywood is far from which negates the "release valve" potential. It is
a perfectly stable material. In view of its poten- unknown at this time how great an effect filling
tial for checking, we must conclude that ply- will have. It is also unknown whether or not the
wood is not as stable as solid wood {iom a filling ofan applied decorative layer (such as
conservation point of view. Plywood was re- paint) with differing mechanical properties will
garded as a panacea on its inception, and as have a subsequent effect on the underlying ply-
a material that allowed artistic freedom for wood veneer. The object's environment, and
designers in the 1920s and 1930s. We as conser- the mechanical properties of the fill material,
vators, however, must now consider plywood will play an imporiant role in determining the
as a product fraught with inherent problems. extent of further damage. These are important
This insight demands a deparhre from the intui- areas for further study of plywood
tive understanding held by most in favor of a conservation.
new understanding: plywood is not stable, not
inert, and is not able to survive the most de- Checks also provide a receptacle into which
manding environmental conditions. dust and airbome grime may settle. Covering
plywood objects, whenever possible, is there-
Ifplywood surface checks are considered to be fore a more important preventive conservation
disfiguring to objects, then pieces in decorative measure than it is for solid wood objects. Not
arts collections composed of plywood should only do years of accumulated grime begin to
and must be considered to be as fragile or pose the same, inadvertent potential harm as a
perhaps more fragile than those constructed of fill, but dirt-filled cracks have the added effect
solid wood. It has been shown that even a mild of becoming "beautiful black lines" across

163
blonde wood, a problem known and loved by Les dommages touchant la structure du bois
all who have tried to deal with it. sont appants beaucoup pltts rapidement que
privu, m€me dans le cas du cycle reprdsentant
Further investigations will perhaps yield en- un changenent de conditions ambiantes accep-
table (de 35 d 59'% d'humiditd relative). D'is la
couraging treatment possibilities. For now there
troisiime semaine, les craquelures qui partaient
are no easy answers. Surface veneer checking is
de la craquelure de ddroulage originale attei-
in large part a classic "inherent vice" problem.
gnaient un,e profondeur de quelque 75 d 80 % du
Although latent, it is a difficulty we must unfor- placage. A la neuviime semaine, nombre d'entre
tunately live with. elles avaient atteint la surface du panneau. Et, d
la vingtidme semaine, la maioritd s'dtaient dten-
R6sum6 dues d la sudace, inddpendamment de I'amplett
du changement d' humiditt.
Une etude priliminaire de l'origine et de Ia
nature des craquelures qui se produisent dans Ie Un moddle est proposi pour expliquer la.forma-
placage des contreplaquis tion des craquelures. Il sefonde sur les dom-
mages risultant de la compression que subissent
Le contreplaqut est un matiriau qui se retrouve les cellules de bois lorsque I'humiditd, qui afa-
de plus en plus souvent dans les collections d'art vorist wte expansion du mattriau, finit par se
ddcoratifet de beaux-arts. Bien que ce produit retirer, lui intposant ainsi un effurt de tension.
soitfondatnentalementfait de bois, sa structure et On insiste sur lefait que le contreplaqud est un
ses caracftristiques le placent dans une cattgorie matiriau non pas stable maisfragile. Et I'on
totalement dilftrente de celle des pi'ices de bois traite de l'incidence de ces dtcouvertes sur la
ntassif. Les craquelures longitudinales qui se pro- consen,ation de ce matdriau, notamment en
duisent dans le placage de nombreux meubles ou ce qui a trait aux risques tventuels que peuvent
supports de peinture constituent un exemple des prdsenter Ia stabilisation et la r2paration
probldmes de dbgradation que peut prdsenter ce de la sudace.
matdriau. De telles craquelures d,!figurent la sur-
face exposie, elles brisent la continuitd des rev€te- References
ments de surface et ellesfont que les couches
sous-jacentes et les adhdsifs entrent en contact
l. Forest Products Laboratory, Forest Service,
avec I'air, I'humidtrd et Ia saletd. Pour qu'un
U.S. Dept. of Agriculture, Handbook of lilood
traitement tclaird puisse €tre appliqut d ces
and Wood-Bas ed Materia ls.for Engineers,
objets, ilfaut d'abord que les specialistes de
la restauration comprennent la nature de ces Architects, and Builders (New York: Hemi-
craquelures et les mdcanismes qui prdsident d sphere Publishing Corp., 1989) pp. 1l-17.
leur formation.
2. Schniewind, A.P., ed., Concise Encyclopedia
La prtsente communicationfail ttat d'une ttude of llood and llood-Based Materials (Oxford:
qui visait d etablir I'existence d'un lien riel entre Pergamon Press, 1989) p.224.
un vice prdsent dans le placage (des craquelures
de ddroulage) et la formation de craquelures d la 3. Hoadley, Bruce, Understanding Wood
surface du contreplaqui. Elle comportait deux vo- (Newtown, CT: The Taunton Press, 1980).
lets : I'etubhssement d'une liste de documents de
recherche portant sur la mtcanique du dtroulage 4. Stamm, A.J.. Wood and Cellulose Science
des placages el un examen au nticroscope de (New York: Ronald Press, 1964).
contreplaquds qui avaient traversd un cycle, en
iquilibre stable, d diverses teneurs en humidit2. 5. Feist, W.C. and D. Hon, "Chemistry of
Un 1channllon de contreplaqud a ainsi ttd sou'
Weathering and Protection," in: The Chemistry
mis, pendant vingt semaines, d deux cltcles de
of Solid l(ood, ACS, Advances in Chemistry
t a ux d' h umi di t t diffh'e n t s, rep rt s e n t a n t re sp e c -
Series, Roger Rowell, ed. (Washington D.C.:
tivement un changentent de conditions ambiantes
acceptable et extr€me. Des prdl'ivements oil eft American Chemical Society, 1984).
faits pdriodiquement, qui ont ttt examints, en
coupe, au microscope pour dvaluer toute
modification de la structure du bois.

164
6. Susan Klim, "Composite Wood Materials in 1 4. Schniewind. Concise Encl,clopedia, 1989,
Twentieth Century Fumiture," Preprints of the p.224.
Wooden Artifucts Group, American Institute
for Conservation Annuql Meeting, 1990. 15. Barkas, W.W., "The Swelling of Wood
Under Stress," Forest Products Research
7 . Leney, L., Mechanism of Veneer Formation Bullerin (G.8.),1946.
at the Cellular Level, Research Bulletin
No. 744, University of Missouri Agricultural 16. Kollman, F.P., E.W. Kuenzi and A.J.
Experiment Station, 1960. Stamm, Principles of lTood Science and
Technologt Vol. II: l4/ood Based Materiqls
8. McMillan, C.W., "The Relation of (New York: Springer Verlag, 1975) p. 116.
Mechanical Properties of Wood and Nosebar
Pressure in Production of Veneer," Forest 17. Skaar, C., "Wood-Water Relationships," in:
Products Journal, vol. 8, no. 1, 1958, The Chemistry of Solid Wood, ACS, Advances
pp.23-32. in Chemistry Series, Roger Rowell, ed.
(Washington, D.C.: American Chemical
9. Fleischer, H.O., "Experiments in Rotary Society, 1984) p. 143.
Veneer Cutting," Forest Products Research
SocieQ Proceedings, vol. 3, no. 20,1949. 18. Canoll. M.N.. "We Still Don't Boil
Houses: Examination of the Boil-Dry-Boil
10. Kivimaa, E., "Investigating Rotary Veneer Test in Plywood Standards," Forest Products
Cutting with the Aid of a Tension Test," Journal, vol. 28, no. 5, 1978, pp.23-27.
Forest Products Journal, vol. 6, no. 7, 1957,
pp.251-255. 19. Murmanis, L. and B. River, "Microscopy of
Abrasive-Planed and Knife-Planed Surfaces in
1 l. Feihl, O., "Peeling Defects in Veneer: Wood-Adhesive Bonds," Wood and Fiber
Their Causes and Control," Canadian Forest Science, vol. 15, no. 2, pp. 102-l 15.
Products Research Bureau Technical Note
No.25. 20. Krahmer, R., "Detecting Incipient Brown
Rot with Fluorescence Microscopy," llood
12. Cumming, J.D. and B.M. Collett, Science, vol. 15, no. 2, pp. 78-80.
"Determining Lathe Settings for Optimum
Veneer Quality," Forest Products Journal, 21.Schniewind. A.P.. "Mechanism of Check
vol.20, no. ll, 1970,pp.20-27. Formation," Forest Products Journal, vol. 13,
no. I 1, 1963, pp. 47 5-480.
13. Chih, L.C., Fundamental Analysis of the
Ven eer C u tti ng Pro ces s, unpublished
dissertation (University of Califomia at
Berkeley, 1970).

165
Case Studies and Specific Problems with Materials
Etudes de cas et problimes particuliers pos6s par les mat6riaux
Deterioration of Cellulose Nitrate Sculptures
Made by Gabo and Pevsner

Michele Derrick, Dusan Stulik Eugena Ordonez

Getty C onserv ation Institute The Museum of Modern Art


Marina del Rey, California New York, New York
U.S.A. U.S.A.

Abstract together to create open structures in which the


interior space of the sculptures actively inter-
Three sculptures by Naum Gabo and Antoine acted with the exterior space. That is, Gabo
Pevsner in The Museum of Modern Art's collec- started constructing sculpture in terms of space
tion exhibit varying degrees of deteioration instead of volume or mass. A few years later,
and rangefrom very good to poor in overall Gabo's older brother, Antoine Pevsner, a
condition. The sculptures are made of thermally painter, also began to create three-dimensional
shaped pieces ofcellulose ester (cellulose nitrate) constructions. Actually, Gabo described their
attached together with an adhesive also based on works as four-dimensional because time as ex-
cellulose ester. The translucent cellulose nitrate
pressed through rhythm was also an element
pieces, which originally may have been uncol-
(Lassaw 1957, 160). The works of these two
ored, are now shades ofbrown. The opaci/ied
brothers highlight the formal vocabulary of the
pieces, which were colored white or black, have
most likel, also changed in tone. Some of the constructivist tradition for which conventional
pieces exhi bit finefi, spaced cracks whi le others methods and materials were no longer applica-
have small drops of a clear to light-brown liquid ble. Although Gabo and Pevsner did not ascribe
present on the surface. Conosion products arc to the utilitarian, propagandistic agenda that the
frequently found where metallic components are Russian constructivists, such as Vladimir Tatlin
near the deteriorated cellulose nitrate. The proc- and Alexander Rodchenko, pursued in their art,
ess of deteioration of these works as well as there are formal similarities in these artists'
works from other collections were studied using work, including their unconventional use of
F o uri e r t ra n sfo rm i nfr ar ed nt i c ro s c opy (F TI R ), modern materials. (See Lodder 1983 for an
X-ray dffiaction (XRD), scanning electron mi- in-depth study of Russian constructivism.)
croscopy coupled with X-ray spectrometry and
backscatter electron imaging (SEM-EDS) and The manufacture of new materials, such as plas-
light microscopy. Information obtained about tics, allowed artists to explore new territories
the possible deterioration mechanisms and the
technically. The modem plastics of the early
role thatfllers and plasticizers may plalt is pre-
1900s provided sculptors with a medium that
sented and proposed conservation treatments
was lightweight, easily cut and molded into a
are discussed.
variety of shape-maintaining forms. The plastic
could be made either transparent or opaque and
Introduction in many colors. The versatility of this new ma-
Naum Gabo moved to Oslo in l9l4 where in terial of the 20th century appears to have been
perfect for their sculptures of the future except
1915 he created his first constructions. These
consisted of flat and curved planes joined
for one problem, the plastics in the early works

r69
of Gabo and Pevsner were cellulose esters. The which is usually responsible for the fast deterio-
early cellulose-derived plastics are inherently ration ofa work, is dependent on the presence
unstable and undergo slow, spontaneous degra- ofoxygen and water because the nitrogen ox-
dation under normal room conditions. ides themselves do not react with the cellulose
nitrate polymer.
This study incorporates the results of the exami-
nation and analysis ofthree sculptures in The Nearly every step in the production of cellulose
Museum of Modem Art's (MoMA) collection nitrate has an effect on the final stability of the
in New York as well as the general findings on material. Cellulose nitrate is prepared by treat-
several works by Gabo and Pevsner in other ing purified cellulose with nitric acid in the
museums. In all six works sfudied, covering a presence of sulfuric acid. The starting cellulose
period ofabout l0 years (1917 to 1926), cellu- needs to be lignin-free since lignin is unstable
lose-based plastics were an integral parl of the and can photodegrade to produce a complex
construction. The range and degree ofdeteriora- mixture of acidic materials that will promote
tion that these plastics have undergone reflect degradation of the cellulose nitrate (Yarsley
numerous factors, including the composition 1964). However, because pure cellulose was ex-
of the starting material for the plastic and the pensive at the tum of the century, many compa-
environmental history of the work of art. nies bought cheaper material, then cleaned and
bleached it. This processing usually left resid-
Cellulose Nitrate Preparation ual acids and chlorine in the cellulose, which in
and Deterioration turn could speed up the deterioration ofthe cel-
lulose nihate. Thus, to regain some stability, ad-
Pullen and Heuman (1988) found that a number ditional materials, such as sodium bisulfite and
of Gabo's sculptures from the 1920s in the Tate triphenyl phosphate, were put in as scavengers
Gallery were made fiom cellulose nitrate and ofthe acids and chlorine.
that "most were in good condition." Their study
showed that the problematic plastic sculptures In the nitration step, sulfuric acid was incorpo-
were those made from cellulose acetate. Inter- rated to safely control the rate and extent of
estingly though, all four works by Pevsner in the reaction. However, this resulted in small
the present study were found to be made from amounts of sulfate esters in the polymer, which
cellulose nitrate and these works showed seri- later produced sulfuric acid and initiated rapid
ous overall or local deterioration. Prior to dis- deterioration. The early remedy for this was to
cussing six works by Gabo and Pevsner, a brief add alkaline additives to the polymer. But when
background on cellulose nitrate, its manufac- added in small amounts, the alkalis were used
ture and degradation processes, will be given. up rapidly while large amounts of added alkali
attacked the cellulose nitrate itself.
Thermal, hydrolytic and photochemical proc-
esses are the primary decomposition mecha- The nitrated cellulose (pyroxylin) was com-
nisms for the degradation of cellulose nitrate. bined with necessary solvents and plasticizers
(For a thorough description ofthese processes to form a clear, plastic material. A variety of
see Selwitz 1988.) In addition, secondary plasticizers were tried for the production of cel-
mechanisms due to reactions of additives and lulose nitrate, such as phosphates, phthalates,
impurities in the cellulose nitrate also influence oils, resins, gums and camphor. Camphor was
the degradation rates. Even when prepared un- considered the best and was the most com-
der the best conditions, cellulose nitrate under- monly used since it not only acts as a plasti-
goes slow, spontaneous degradation in which cizer but also as a stabilizer. It forms hydrogen
the NOz groups split off to form nitrogen oxide bonds with the unsubstituted hydroxyl group
gases. This thermal degradation is accelerated producing a non-crystalline, solid solution of
by ultraviolet light. Upon exposure to water cellulose nitrate that is strong and flexible. Cel-
and oxygen, the nitrogen oxides form nitric and lulose nitrate with camphor as its plasticizer is
nitrous acids, which then catalyze hydrolyic de- referred to as celluloid.
composition reactions. The autocatalyic cycle,

170
Although camphor was the best plasticizer were used to identify and quantify additives in
available it caused some problems. Impurities the polymer as well as identify the principal
in the camphor are one of the causes of the yel- component of the plastic.
lowing of cellulose nitrate objects. Also, be-
cause it forms hydrogen bonds with the Sculptures
cellulose nitrate, the proportion of camphor in
the final product is important. Too little cam- The earliest sculpture included in the present
phor results in a brittle material subject to craz- study is Head of a Woman by Naum Gabo,
ing. On the other hand, excess camphor slowly shown in Figure 1 and in the collection of The
sublimes from the object, leaving a porous or Museum of Modem Art. It is generally be-
cracked material that is more susceptible to oxi- lieved to have been made in Russia after the art-
dation and hydrolysis and thus to accelerated ist's retum there in 1917, although the exact
deterioration. Even in materials that have the date of execution is not known. Nash bnd Merk-
optimum proportion of camphor (20%to 40%), ert (1985) give arguments for possible dates of
the camphor will slowly sublime from the ex- 1917 to 1920 as well as circa 1923.
posed surfaces promoting deterioration, and
when the object degrades, more surface is usu-
ally exposed and camphor is lost more rapidly.
(For further information on the use of camphor
in celluloid see Reilly 1991.)

Stabilizers were often added to scavenge free


radicals and unwanted acids. Urea and diphenyl
amine were found to be effective initially, but
later crystallized out and became more harmful
than beneficial. Zinc oxide was commonly used
as a filler and opacifier. It also acts as a stabi-
lizer by rapidly consuming nitric acid to form
zinc nitrates.

Sample Analyses

The decomposition of cellulose esters has been


documented and studied extensively and the in-
stability of these materials is generally acknow-
ledged (Sirkis 1982). What is intriguing in the
case of these sculptures, however, is the type
and extent ofdegradation that occurs from
work to work and flom location to location
Figure I Naum Gabo, Head of a Woman (ca. l9l7
within a sculpture. To obtain a better under-
to 1920, after awork oJ 1916). Construction in
standing of the relation between various celluloid and metal, 62.2 cm x 48.9 cm x 35.4 cm.
deterioration mechanisms and the material Collection, The Museum oJ Modem Art, New York,
composition, chemical analyses were per- Purchase.397. 38.
formed on samples from several sculpfures
and from several sites within each sculpture
The work was formed ffom thermally shaped,
whenever possible.
cream-colored pieces of opacified cellulose ni-
trate. Frequently Gabo would make models of a
The samples, acquired as broken fragments or
work in materials, such as cardboard or wood,
removed as scrapings. cross-sections and
prior to executing the sculpture in metal. In a
exudates, were analyzed by FTIR microscopy
letter from 1948, Gabo described MoMA's
at the Getty Conservation Institute and by light
Head of a lloman as being "finished and com-
microscopy, SEM-EDS and X-ray diffraction at
plete" but "in fact, the original model which I
The Museum of Modem Ar1. These methods

t71
always make before executing a work in a In Alexi Pevsner's biography of his two broth-
proper and solid material. In the case of this ers (1964), Naum and Antoine, he mentions
head...the material is celluloid and therefore un- that Gabo taught Pevsner "how to handle metal
stable..." (Nash and Merkert 1985,202).In and plastic materials." Head of a Woman (Fig-
1951 and 1964 Gabo borrowed the head to use ure 2) made by Antoine Pevsner is dated 1923
as a model for a bronze version now in the col- (inscribed under the artist's signature in the
lection of Gabo's family. lower left corner) and is in the collection of the
Hirshhom Museum, Washington, D.C. The
Since its acquisition in 1938, the sculpture has work is in extremely poor condition and is no
been frequently exhibited at the Museum. longer exhibited. The sculpture's pieces were
When on display, it has been hung above eye made from translucent cellulose nitrate except
level in a comer with no vitrine in a similar for the creamy-white base that was opacified
fashion to the way it was presented in the Van with zinc oxide. As shown in Figure 3 the trans-
Diemen Gallery in 1922. (See Nash and Merk- lucent pieces have yellowed and cracked se-
ert 1985, 2l for aphotograph ofthe 1922 exhi- verely, joined pieces have separated and forms
bition showing this work. It should be noted have fallen off. The pieces have also distorted
that Nash believes the object shown in this pho- out-of-plane by forming bubbled areas.
tograph is actually an earlier metallic version
and not this celluloid version as other scholars
believe.) In contrast to some other works done
about this time by Gabo and Pevsner, to be dis-
cussed, this work is in very good condition.
Nonetheless, by 1947, Gabo objected to the
work being exhibited in its "present condition"
(Richards 1947).lt appears that the damage in-
cluded at least two pieces of plastic that had
fallen off. The work was not repaired by the afi-
ist in time for the Gabo-Pevsner Exhibition in
1948 and therefore was not included. It may a ::,1:'aa.:a/:l

have been this sculpture that Gabo is indirectly Figure 2Antoine Pevsner, Head of a Woman
quoted as describing as "an early celluloid (1923) Hirshhorn Museum and Sculpture Garden
(Photo taken in l99l).
model, now in the collection of Alffed Barr,
[which] has cracked and discolored severely,
but was not intended to be permanent" (Siegl When the work was acquired in the early
and d'Hamoncourt 1968). 1970s, it was placed in a wall-mounted vitrine
case. By 1982 the sculpture had darkened,
Currently, there are long, fine cracks especially cracks had formed along the join lines and
in the forehead area, some broken or missing "beads of clear oily liquid" were present on
pieces and separation ofa few pieces from the the surface of the sculpfure (Aks 1991). These
main body, for example, the right proper shoul- drops were analyzed in 1983 and found to be
der has separated along the join line. Most of an aqueous basic zinc nitrate solution (Erhardt
the damage appears to have resulted from me- et al. 1983). The 1983 analytical report fuither
chanically induced stresses caused by handling states that "zinc nitrate is a result ofattack of
the work or environmental factors. The pieces nitric acid on the zinc oxide pigment of the cel-
are a creamy-white except for three that are dis- luloid on which the drops form....It was only
tinctly yellow, namely the proper right shoulder when the object was sealed up that the gener-
and two pieces near the chin area. FTIR analy- ated acid was confined so that reaction with
sis of the discolored pieces showed them to be the available alkali (zinc oxide) became inevita-
cellulose nitrate (as were the other pieces). X- ble." After cleaning the sculpture, it was placed
ray diffraction analysis indicated that zinc ox- in storage in its vitrine and an efforl was made
ide was the principal opacifying pigment both to minimize exposure to light by covering the
in the yellow and cream-colored pieces. The case with brown paper. By 1988 the sculpture
cause ofdiscoloration has not been determined. was in a "severe state of deterioration....A clear

112
Figure 4 Antoine Pevsner, Bust (1923 to
1924). Construction in metal and celluloid,
53 cm x 59.4 cm. Collection, The Museum of
Modern Art, New York. Purchase. 396. 38.

color, as shown in Figure 5), as well as the


formation of light-brown drops of liquid over
several ofthe pieces. There is also localized cor-
rosion of the metal in areas where the deterio-
Figure 3 Antoine Pevsner, Head of a Woman, rated cellulose nitrate contacts metallic pieces,
detail of bottom portion. such as the copper background or iron bolts.

residue has reformed over the entire sculpfural


surface mostly in the form of oily beads" (Aks
1988). It was also noted that the cracking had
worsened significantly and that "several plastic
panels had collapsed and separated from the
sculpture" (Aks 1988). For a while the work
was kept in a fume hood where it was well ven-
tilated and currently it is kept in an uncovered
ply.wood box in storage. When the work was
examined recently for this study, only a small
amount of the oily drops was observed.

The presence ofa liquid exudate has been no- Figure 5 Antoine Pevsner, Bust, detail ofupper
ticed on other works such as Pevsner's -Basl portion showing extensive cracking of a translucent
(shown in Figure 4) completed between 1923 piece and spotting on an opaque piece (Photo taken
199r).
and 1924. This relief sculpture, acquired by
MoMA in 1938, is in relatively good condition
even though it also has problems. It is made out Until 1980 this work was frequently exhibited
of translucent and opacified cellulose nitrate at MoMA. Exhibition photographs show that it
pieces, the latter primarily opacified with zinc was hrurg without a vitrine on a wall. In 1989,
oxide. The cellulose nitrate has deteriorated in a for as yet undetermined reasons, the sculpture
varief of ways, including cracking, discolora- started producing a liquid exudate especially
tion of translucent pieces, spotting (the pres- on the opacified cellulose nitrate pieces,
ence of small, roundish areas slightly whiter in such as the topmost curved piece. A small
portion of the liquid was analyzed with FTIR

t73
works at MoMA showed that potassium, chlo-
rine, silicon, calcium and sulfur are usually pre-
sent in both the opacified and the translucent
cellulose nitrates.

Figure 6 Calcium sulfate dihydrate crystals


precipitated out of liquid degradation product,
5 00 times magniJication.

spectroscopy and determined to be primarily Figure 7 Higher atomic weight elements appear
composed of inorganic nitrate(s). Over several brighter. Top surface ofthe cross-section is diago-
months, while the liquid was stored in a cov- nally across the lower left corner, 500 times
ered well-slide, crystal formation took place. magniJication.
With slight heat more crystals formed. Several
chemical species appear to be present in the A third piece at MoMA, Zorso (illustrated in
exudate and have only been partly identified at Figure 8) dated 1924 to 1926, was also made
this time. X-ray diffraction of a large cluster of by Pevsner. This work consists primarily of
crystals showed these to be calcium sulfate di- translucent cellulose nitrate pieces that are
hydrate. Figure 6 is a secondary electron image often attached with metal hardware to similarly
of these crystals magnified approximately 500 shaped copper pieces. (Figure 9 is a detail of
times. Energy-dispersive spectrometric analysis the proper right hand showing the copper and
(EDS) of the same crystals showed that along cellulose nitrate held together with a screw.)
with calcium and sulfur, zinc was also present Torso was exhibited for many years on a pedes-
(though apparently not as a crystalline com- tal and without a vitrine. In December 1954,
pound since it was not distinguished by XRD). the sculpture was accidently knocked over
SEM-EDS examination of a cross-section from while on view, resulting in several broken
an area where excessive liquid had accumulated pieces. The treatment report states that the
indicates that there is about a 20 micron deple- physical effects of the fall were exacerbated by
tion of higher atomic weight elements {iom the the deteriorated condition of the cellulose ni-
surface ofthe piece as indicated in Figure 7. trate (which was incorrectly identified as cellu-
(Because of the sample size required for cross-
lose acetate in the report). The treatment
sections and the delicate nature ofthe sculp- consisted of replacing a number of the pieces,
ture, confirmation of these preliminary results mostly around the lower head area. After being
will have to be carried out on cellulose nitrate repaired, the work was heavily exhibited until
artifacts exhibiting similar symptoms.) Macro- 1968 when it was removed after the black cellu-
scopically, this effect was observed by an in- lose nitrate base (which had been progressively
creased translucency at the surface, albeit deteriorating since at least 1964) was too buck-
exhibiting a slightly browner tonality. It thus led and fragmented. For most of the time be-
appears that zinc salts as well as other com- tween 1968 and 1978 the sculpture was kept in
pounds (such as calcium salts) are leached out storage. Part ofthe reason for storing this ob-
as part ofthe degradation process. EDS analy- ject, as well as other cellulose nitrate objects,
sis on a number of small scrapings of cellulose was the fear that they would spontaneously
nitrate from this work as well as the two other

174
exhibition and to lightly drape them with paper
to keep dust off if they are not on view.

Torso was once described as being able to


"exploit light through the various densities
and directions of the planes" (Read 1948). The
transparency ofthe plastic was also used to set
up a "rhyhmic dialogue" with the curved cop-
perbacking (Rankin 1988). The photograph of
the sculpture presented in this paper was taken
in 1933 and indeed shows how light could have
penetrated the cellulose nitrate forms and re-
flected offthe copper pieces. It no longer is
able to do this as effectively because the cellu-
lose nitrate pieces are all now discolored and
some have cracked badly enough to form essen-
tially opaque layers. A detail ofthe head area
(illustrated in Figure l0) shows what is left of a
cellulose nitrate piece and how corroded some
ofthe copper pieces have become.

Figure 8 Antoine Pevsner, Torso (1924 to


1926). Construction in plastic and copper,
74.9 cm x 29.4 cm x 38.7 cm. Collection, The
Museum of Modern Art, New York.
Ktttherine S. Dreier Beauest. ,/85. 53.

Figure I0 Pevsner,Torso, detailJi'om head area


(Photo taken in 1991).

At present there is only a slight amount of


liquid formation on the cellulose nitrate sur-
face. FTIR analysis of the liquid, which in this
case crystallized out instantly on contact with a
Figure 9 Pevsner, Torso, detail ofhand (Photo microscope slide, showed it to be inorganic ni-
taken in I99I). trates (Figure I I is an FTIR spectrum of these
crystals). The XRD pattem indicated the pres-
combust. Although it was unlikely that the ob- ence of ammonium nitrate as one of the compo-
jects would combust since the percent nitrate in nents (JCPDS 8452). The copper pieces have
polymers (10%) is lower than that in explosives also corroded to various degrees forming a blue-
(13% to l4o/o) or frlm (l2o/o), suggestions were green, very fine-grained powder. Figure l2 is a
made to put them in airtight vitrines filled with backscatter electron image of the powder show-
nitrogen or to store them in explosion-proof ing the general size distribution ofthe particles.
steel cabinets. It appears instead that Torso was Analyses so far indicate that it is a form of
kept in storage with a plastic film draped over copper nitrate, possibly gerhardite. The same
it. The present policy at MoMA is to keep air corrosion product was found on the copper
freely circulating around these objects while on backins in Pevsner's Busf.

115
Figure I I FTIR spectrum of crystallized liquid
exudate from Pevsner Torso.
very yellow and brittle

3998 3332 2666 2000 1741 182 1223 9U 705


wavenumber (CM_r)

Figure I 3 Infrared spectrafor cellulose nitrate at


va ri ous s ta ges of degrada t ion.

Figure I 2 Conosion product from the Pevsner


Torso, 2,000 times magnifcation.

FTIR analysis ofcondensed fractions collected


from heating cellulose nitrate pieces show that
camphor was used as a plasticizing agent along
with an additional oil. There is a direct correla-
tion between the absence of camphor and se-
vere degradation of the sample as shown in
Figure 13. The top infrared spectrum is ofcellu-
lose nitrate in good condition and the bottom
one is ofthe nitrate salt degradation product.
Two stages of the deterioration are shown in
between. Noticeable in the spectra is the in-
crease in the intensity of a nitrate salt band
at 1340 cm-'. There is a corresponding de-
crease in the intensity of the carbonyl band at
1735 cm'' , which is due to decreasing amounts
of camphor, the plasticizer. (See Figure 14 for
reference spectra ofcellulose nitrate and
wavenumber (cM_l)
camphor.)
Figure 14 Infrared spectrafor cellulose nitrate,
Cross-sections taken from broken fragments camphor and liquid obtained from the Pevsner
of the sculpture were examined to assess the Torso.

r76
extent of deterioration into the pieces of cellu-
lose nitrate (i.e., depth-wise). Figure l5 shows
an inffared spectrum produced by analyzingthe
surface ofthe cellulose nitrate piece. The other
spectra were collected further into the sample,
specifically at I mm and 2 mm. The spectra
show that while the surface of the sculpture con-
tains cellulose nitrate with a diminished amount
of plasticizer, the celluloid 2 mm into the sam-
ple is in good condition.

2 mm inside

1 mm inside

Outside surlace

F i gu re t 5 ;:;":;; ";:;i:o * *
"
co r e c te d a t
dffirent ", ",,in a cross-section sample
depths from the
PevsnerTorso. The spectra obtainedfrom the outer
Figure I6 Antoine Pevsner, Portralt of Marcel
area ofthe cross-section show that there is a
Duchamp, Yale University Art Gallery, Gift of
depletion of camphor near the suryface.
Collection Sociiti Anonyme.

Another work of this period by Pevsner is the The sculpture is undergoing degradation proc-
Portrait of Marcel Duchamp in the Yale Uni- esses, such as excessive liquid production on
versity Art Gallery collection, which was com- the surface (illustrated in Figure l7), cracking
missioned in the spring of 1926 and completed and buckling of the cellulose nitrate pieces and
in the summer of the same year. This work, like extensive corrosion of metallic parts. These me-
the Bust mentioned above, is a relief attached tallic parts include the copper backing, which is
to a metallic backing. The arlist used both clear now covered with a blue-green corrosion prod-
and opaque plastic for this sculpture including uct, an iron-based strip of metal, which is now
pieces colored purple and black. By 1941, a pile of rust and the copper- and iron-based
when the Yale University Art Gallery acquired hardware, which have also corroded. The Yale
the work, conservation records indicated that University Art Gallery files indicate that at
pieces were already warped and discolored. It least since 1984 until now the work has been
was restored in 1957 by the artist after it was kept in storage in an essentially airtight vitrine.
smashed en route to an exhibition. At that time Andrew Petryn, former conseryator at Yale,
the artist replaced broken pieces and changed thought that the work might have been kept
the backing (originally a zinc plate) to copper in a vitrine for a much longer period of time,
because the latter harmonized better with the although an exact date was not known
discolored pieces (Herbert 1984). Figure l6 (personal communication, I 99 l).
shows the work after the 1957 repair. Within a
few years the copper backing started corroding A detailed photograph from late 1984 shows
as the plastic pieces kept deteriorating. By an active corrosion problem but for the most
7979, a replica in plexiglass had been made for part the pieces ofcellulose nitrate do not appear
the gallery by a former assistant of Pevsner badly cracked or out of plane. As mentioned
(Herbert 1984). earlier though, the pieces are now badly

177
Figure 1 7 Pevsner, Portrait of Marcel Duchamp,
detail ofbottom edge oJ'sculpture (Photo taken
in 199 I ).

cracked, many of them exhibiting widely


spaced cracks with the plastic in between the
cracks badly buckled, as shown in Figure 18.
It thus appears that a significant amount of the
cracking and buckling has occurred in the last
few years. As will be discussed fuither in the
following paragraphs, this type of damage ap-
pears to occur when the objects are kept in a
sealed vitrine and then suddenly exposed to a
new environment when the vitrine is opened.
This change in environment allows the immedi-
ate loss ofvolatile degradation byproducts.
This in furn appears to create intemal stresses Figure I9 Antoine Pevsner, Portrait of Marcel
subsequently released by crack formation. It Duchamp, Yale University Art Gallery, Gilt of
C ol le c tion S oc i et e Anorytme.
is known that the vitrine of the Portrait of

Marcel Duchampwas opened in 1986 to re-


move some fragments that had fallen offf.
Rapid deterioration appears to have followed as
shown by comparing the photograph in Figure
19 (taken around the mid-1980s) with the re-
cently photographed details pictured in
Figures 17 and 18.

When Gabo was asked how the invention of


plastics affected the course of his works, he re-
plied, "I did a great deal of work in plastics for
only one reason: to accentuate the transparent
character ofspace" (Kuh 1962, 102). The cones
in Construction in Space: Two Cones (1921)
in the Philadelphia Museum of Art exemplify
Gabo's use of plastic for its transparent quali-
ties. The artist also used opaque plastics
though, in white, black and red for some of
the pieces in this work, thus emphasizing their

*These fragments were analyzed for this study and

Figure 18 Pevsner, Portrait of Marcel Duchamp, found to be cellulose nitrate. It is not known what
detail ofplastic in upper right portion oJ'sculpture piece(s) they are from or if they are fiom the original
(Photo taken in I99I). 1926 pieces or the last 1957 restoration by the artist.

178
planes and curves. In its present condition, as
the detail shown in Figure 20 indicates, both
transparency and form have been seriously
compromised with the deterioration of the plas-
tics. (For an earlier photograph, see Nash and
Merkert 1985.215.)

Figure 2l Fragments from Naum Gabo's


*ff
Construction in Space: Two Cones (Photo taken
in 1991).

the 1927 work), which he had in his studio and


which were still in good condition (Siegl and
d'Hamoncourt 1968). These were analyzed and
all were found to be cellulose acetate (Miklas
Figure 20 Naum Gabo, Construction in Space: 1968). As described by Pullen and Heuman
Two Cones, detail (Photo tqken in 1991). (1988) though, when Gabo used cellulose ace-
tate in 1968 to make a replica of this work for
For the most part, since the work was acquired the Tate Gallery, rapid deterioration again oc-
in 1952 it has been kept in a sealed vitrine. Seri- curred. Unfortr-rnately, as had been the case
ous physical damage occurred after the sculp- for the Philadelphia Museum of Art work, the
ture was removed for study from its airtight Tate Gallery version had also been stored and
vitrine in 1960 when condensation was noticed exhibited in a closed virrine.
on the glass (Siegl 1966). Upon opening the
case "a strong odor escaped from the case, then During an interview in 1953, Gabo was asked
cracks developed in some parts ofthe construc- if his constructions, which seemed so delicate,<