UNI VERS ITE DE SHERBROOKE

Faculte de genie
Departement de genie chimique et de genie biotechnologique

ETUDE DE PRODUCTION ET DE CARACTERISATION DE

BIOCHARBONS DE PANIC ERIGE (Panicum virgatum L.)
OBTENUS PAR PYROLYSE

These de doctorat
Speciality : genie chimique

Guillaume PILON

Jury : Jean-Michel LAVOIE (directeur)

Nicolas ABATZOGLOU
Denis GROLEAU
Valerie ORSAT

Sherbrooke (Quebec) Canada Mai 2013

 Library and Archives
1+1 Canada
Published Heritage
Branch
395 Wellington Street
Ottawa ON K1A0N4
Canada
NOTICE:
The author has granted a non­
exclusive license allowing Library and
Archives Canada to reproduce,
publish, archive, preserve, conserve,
communicate to the public by
telecommunication or on the Internet,
loan, distrbute and sell theses
worldwide, for commercial or non­
commercial purposes, in microform,
paper, electronic and/or any other
formats.
The author retains copyright
ownership and moral rights in this
thesis. Neither the thesis nor
substantial extracts from it may be
printed or otherwise reproduced
without the author's permission.
In compliance with the Canadian
Privacy Act some supporting forms
may have been removed from this
thesis.
W hile these forms may be included
in the document page count, their
removal does not represent any loss
of content from the thesis.
Canada
Bibliotheque et
Archives Canada
Direction du
Patrimoine de I'edition
395, rue Wellington
Ottawa ON K1A 0N4
Canada
Your file Votre reference
ISBN: 978-0-494-96324-1
Our file Notre reference
ISBN: 978-0-494-96324-1
AVIS:
L'auteur a accorde une licence non exclusive
permettant a la Bibliotheque et Archives
Canada de reproduire, publier, archiver,
sauvegarder, conserver, transmettre au public
par telecommunication ou par I'lnternet, preter,
distribuer et vendre des theses partout dans le
monde, a des fins commerciales ou autres, sur
support microforme, papier, electronique et/ou
autres formats.
L'auteur conserve la propriete du droit d'auteur
et des droits moraux qui protege cette these. Ni
la these ni des extraits substantiels de celle-ci
ne doivent etre imprimes ou autrement
reproduits sans son autorisation.
Conform em ent a la loi canadienne sur la
protection de la vie privee, quelques
formulaires secondaires ont ete enleves de
cette these.
Bien que ces formulaires aient inclus dans
la pagination, il n'y aura aucun contenu
manquant.
RESUME
Dans le cadre de cette recherche, la production de biocharbon par pyrolyse est etudiee a des
conditions visant sa valorisation comme biochar (amendement pour le sol), tout en considerant
son potentiel comme charbon vert (pour bioenergie ou transformations subsequentes). La
production du charbon de biomasse s’est effectuee a l’aide de deux reacteurs a lits fixes de
types batch, d’une capacite de 1 et 25 gb.h./batch, respectivement. Le panic erige (Panicum
virgatum) est la biomasse lignocellulosique qui a ete utilisee dans le cadre des tests. Les
facteurs de production etudies sont principalement la temperature (300, 400 et 500 °C) pour un
court temps de residence en reacteur (2,5 et 5 min) et l’effet de l’usage du CO 2 plutot qu’un
environnement plus conventionnel d’azote. L ’effet de ces facteurs est etudie par rapport aux
caracteristiques physico-chimiques des biocharbons obtenus. Un suivi des produits
pyrolytiques complementaires (bio-huile et gaz) a aussi ete effectue. Des extractions de
biocharbon par Soxhlet (a l’aide de dichloromethane) ont ete analysees par GC-MS et ont
permis 1’identification d ’une multitude de produits presents dans les biocharbons. Les
conditions specifiques utilisees pour la pyrolyse, entre autres la convection forcee avec taux de
chauffe rapide a 300 °C - N 2 a l’aide du reacteur 1 g/batch, ont presente des rendements et
proprietes de biocharbon avantageux pour l’amelioration de la productivity du procede de
torrefaction (en comparaison avec des travaux rapportes, notamment ceux de Gilbert et al.
[2009]). Les analyses des extractions de charbons de biomasse ainsi que des bio-huiles (par
GC-MS), produites a l’aide du reacteur 25g/batch, ont permis d’observer des differences
significatives dans les composes obtenus lors de l’usage de CO2 vs N2. Plusieurs composes
observes dans les extraits de biocharbons, produits en atmosphere de N 2 , se retrouvent en
quantites moindres dans les extraits de biocharbons produits en atmosphere de CO 2 pour des
temperatures communes. A titre d ’exemple, le furfural, un compose aromatique commun
provenant de la degradation des glucides, s’est retrouve uniquement dans les extraits de
biocharbons en presence de N 2 vs CO 2 , a 400 °C. Parmi l’ensemble des conditions etudiees
(pour les 2 reacteurs), uniquement le naphtalene et des derives du naphtalene sont observes
comme hydrocarbures aromatiques polycycliques, et ce, uniquement suivant les traitements a
500 °C. L ’etude de l’usage du CO 2 comme gaz d ’entree en reacteur mena a des differences
significatives pour l’ensemble des temperatures etudiees, et ce, tant pour les biocharbons que
pour les produits liquides et gazeux. A 300 °C, en environnement de CO2 compare a N2, il est
possible d’observer une production de bio-huile significativement plus faible (18,0 vs 24,6 %;
C 0 2 vs N2 pour P<0,002), ce qui represente un resultat coherent avec l’obtention de
biocharbon au contenu en composes volatils significativement plus eleve obtenu aux memes
Conditions (0,29 V S 0,35 g com poses v olatils - biocharbon^§ biom asse originate! CO2 V S N 2 p O U r P=0,1). De
plus, a 500 °C, un contenu en cendres de biocharbon significativement plus faible a ete
observe en environnement de CO 2 vs N 2 (P<0,06).

M ots-cles: Pyrolyse, torrefaction, biocharbon, biochar, charbon vert, CO 2 , extractions

Soxhlet, caracterisation.
ABSTRACT
This research aimed at the production of biomass char under pyrolytic conditions, targeting
biochar as soil amendment, while also considering its application as biocoal, either for
bioenergy or subsequent upgrading. The production o f biomass char was performed using two
bench-scale, batch-type, fixed-bed reactors, each with an operating capacity of 1 and 25
gw.b./batch, respectively. Switchgrass (Panicum virgatum) has been used for the tests.
Production conditions studied implied temperatures of 300, 400 and 500 °C with short
residence times (2.5 and 5 min). As well, the effect of using CO 2 as vector gas has been
compared to a common inert environment of N 2 . The effects o f the previously mentioned
parameters were correlated with some important physicochemical characteristics o f biomass
char. Analyses were also performed on complementary pyrolytic products (bio-oil and gas).
The biomass char extraction was performed using a Soxhlet and dichlromethane was used as
extracting solvent. The extracts were then characterized by GC-MS thus allowing the
identification o f several compounds. Specific pyrolysis conditions used at 300 °C - N 2 with
the 1 g/batch reactor, such as high heating rates as well as high convection conditions,
presented advantegeous biomass char yields and properties, and, possible torrefaction process
productivity improvement (in comparison to reported literature, such as Gilbert et al. [2009]).
The char extracts as well as the bio-oils analysis (also performed using GC-MS), all generated
from the 25 g/batch reactor, showed major differences among the compounds obtained from
the CO2 and N 2 environments, respectively. Several compounds observed in the char extracts
appeared less concentrated in the CO2 environment vs N 2 , for the same reaction temperatures.
As an example, at 400 °C, furfural was found only in char extracts from N 2 environment as
compared to the CO2 environment. Among all studied conditions (for both reactors), only
naphthalene and naphthalene derivatives constituted the PAHs content, which was only
detected for the chars produced at 500 °C. The use of CO 2 as pyrolysis vector gas led to a
significant difference for every temperature conditions studied for the biomass char as well as
for the liquid and gas products. At 300 °C, in CO2 environment, it is possible to observe a bio­
oil production significantly lower than within a N 2 environment (18.0 vs 24.6 %; CO 2 vs N 2
for P<0.002), a result consistent with the biomass char volatile content that was shown to be
significantly higher under the same conditions (0.29 vs 0.35 g char volatile* c o m e m /g raw b io m a ss;
P=0.1). In addition, at 500 °C, the char ash content was observed to be significantly lower in
C 0 2 than in N2 (P<0.06).

Keywords: Pyrolysis, torrefaction, char, biochar, biocoal, CO2 , Soxhlet extractions,

characterization.
REMERCIEMENTS
Un projet de la sorte fiat rendu possible seulement grace a l’apport considerable de personnes
cles et de ressources essentielles tout au long de sa realisation.

Je tiens ainsi a remercier sincerement:

• Prof. Jean-Michel Lavoie pour m’avoir permis d’effectuer ce projet et pour tout le
support qu’il y a apporte et l’interet qu’il y a demontre.
• M. Henri Gauvin pour son attitude optimiste, son aide technique et sa disponibilite.
• M. Serge Gagnon pour son support technique et sa disponibilite.
• M. Gaston Boulay (aide GC-MS), M. Michel Trottier (aide GC-MS), Mme. Eva Capet
(aide analyses Panic erige), M. Marc G. Couture (aide machinage & soudures), M. Carl
St-Louis (aide BET), Mme. Sonia Blais (aide CHONS), Mme Irene Kelsey (aide
MEB), M. Serge Berube et M. Stephane Gutierrez pour leur support technique et leur
disponibilite pour repondre aux questions.
• Le Fonds Quebecois de Recherche sur les Nouvelles Technologies.
• Les Fonds de recherche de demarrage de Prof. Jean-Michel Lavoie par le Departement
de genie chimique et genie biotechnologique de l’Universite de Sherbrooke.
• Les membres du jury pour 1’evaluation de la these : Prof. Nicolas Abatzoglou, Prof.
Denis Groleau, et Prof. Valerie Orsat.
• Les membres du jury pour l’examen predoctoral: Prof. Marc Olivier, Prof. Nicolas
Abatzoglou et Prof. Francois Gitzhofer.
• Prof. Esteban Chomet pour le temps qu’il m’a accorde et son support pendant la
periode preliminaire a ce projet.
• Prof. Jean Lessard du Departement de chimie de l’Universite de Sherbrooke pour
l’acces a son laboratoire.
• Les stagiaires pour leur aide technique lors de manipulations laboratoires : M. Khalid
Oukarame (aide pour valeurs calorifiques), Mme. Marie-Pier Gravel (aide analyses
Panic erige et Karl Fisher) et M. Osvaldo Serrano (Karl Fisher).
• Des collegues-etudiants de travail de l’Universite de Sherbrooke pour leur aide
technique et discussions : Mme Clemence Fauteux-Lefebvre, M. Jasmin Blanchard
(analyse des gaz), Dr Guillaume Leonard (aide programmation Labview), M. Alexis
Cadoret, M. Mohammad Zakaria et Mme Julie Roy.
• Tout le personnel du secretariat et des laboratoires du Departement de genie chimique
et genie biotechnologique de l’Universite de Sherbrooke.
• Ma compagne, Veronique Gagnon, pour son support moral, son encouragement, son
ecoute et sa comprehension.
• Ma famille et particulierement mes parents, Adrien Pilon et Louise Potvin, pour leur
inspiration, ecoute, conseils et encouragements.
• Prof Vijaya G.S. Raghavan, qui m ’a fortement encourage et influence a poursuivre
mes etudes jusqu’au 3e cycle.
• Anciens collegues, amis et colocs pour d ’importantes discussions et leurs
encouragements : Dr Venkatesh Sosle, Thomas Fortin-Chevalier, Michele et co.,
Philippe-David et sans oublier les amis de « jams » de musique.
TABLE DES MATIERES
CHAPITRE 1 : INTRODUCTION........................................................................................ 1
1.1 Mise en contexte et problematique.................................................................................... 1
1.2 Definition du projet de recherche...................................................................................... 4
1.3 Objectifs du projet de recherche........................................................................................ 7
1.4 Contributions originates...................................................................................................... 7
1.5 Plan du document.................................................................................................................9
CHAPITRE 2 : CADRE DE REFERENCE ET ETAT DE U ART.................................. 10
2.1 La biom asse....................................................................................................................... 10
2.1.1 Types de biomasses.........................................................................................................10
2.1.2 Composantes de la biomasse........................................................................................... 11
2.2 Charbon de biomasse vegetale..........................................................................................16
2.2.1 Biochar........................................................................................................................... 17
2.2.2 Charbon de biomasse torrefiee et charbon vert................................................................ 19
2.2.3 Autres produits carbones................................................................................................21
2.3 Conversion de la biomasse en biocharbon par pyrolyse............................................... 24
2.3.1 Pyrolyse.......................................................................................................................... 24
2.3.2 Effet de la temperature et de la composition de la biomasse sur la pyrolyse...................28
2.3.3 Effet de l’usage du C 02 lors de la pyrolyse.................................................................... 39
2.3.4 Effet du temps de residence du biocharbon en reacteur sur la production de biocharbon 52
2.3.5 Effet du taux de chauffe sur les caracteristiques du biocharbon....................................... 53
2.3.6 Effet du transfert de chaleur et de masse sur les caracteristiques du biocharbon.............. 56
2.3.7 Effet de la pression sur les caracteristiques du biocharbon.............................................. 57
2.3.8 Composition chimique des extraits de biocharbons.........................................................58
CHAPITRE 3 : ARTICLE 1 - PRODUCTION DE BIOCHARBON EN
REACTEUR A LIT FIXE DE CAPACITE 1g/batch - EFFET DE LA TEMPERATURE
A DE COURTS TEMPS DE RESIDENCE.................................................................................. 61
3.1 Introduction........................................................................................................................63
3.2 Experimental.....................................................................................................................66
3.2.1 Biomass..........................................................................................................................66
3.2.2 Experimental setup and biomass char production............................................................66
3.2.3 Feedstock and char characterization................................................................................ 67
3.3 Results and discussion !...............................................................................................69
3.3.1 Balance on char production............................................................................................. 69
3.3.2 Heating value.................................................................................................................. 72
3.3.3 SEM figures and BET results.......................................................................................... 74
3.3.4 Char extraction and composition analyses...................................................................... 76
3.4 Conclusions........................................................................................................................ 82
CHAPITRE 4 : ARTICLE 2 - PRODUCTION DE BIOCHARBON EN REACTEUR
A LIT FIXE DE CAPACITE lg!batch EN ENVIRONNEMENT DE C 02 vs N2 - ETUDE
PRELIM INAIRE............................................................................................................................ 83
4.1 Introduction........................................................................................................................ 85
4.2 Materials and methods...................................................................................................... 89
4.2.1 Biomass.......................................................................................................................... 89
4.2.2 Experimental setup..........................................................................................................89

iv
 4.2.3 Biomass char production and product mass balance...................................................... 91
4.2.4 Feedstock characterization.............................................................................................92
4.3 Results and discussion...................................................................................................... 93
4.3.1 Heating rate effect.........................................................................................................93
4.3.2 Gas environment effect..................................................................................................94
4.3.3 Temperature effect.........................................................................................................95
4.4 Conclusions.........................................................................................................................97
CHAPITRE 5 : ARTICLE 3 - ETUDE DES BILANS DE PRODUITS DE
PYROLYSE OBTENUS EN ENVIRONNEMENTS DE N2 vs C 0 2 EN REACTEUR A
LIT FIXE D’UNE CAPACITE DE 25g/batch..............................................................................99
5.1 Introduction.......................................................................................................................102
5.2 Materials and methods.....................................................................................................104
5.2.1 Feedstock..................................................................................................................... 104
5.2.2 Experimental setup description and operation.............................................................. 104
5.2.3 Feedstock and char characterization..............................................................................106
5.2.4 Condensable products preparation and analyses.......................................................... 106
5.2.5 Gas analyses................................................................................................................ 106
5.3 Results and discussion................................................................................................... 107
5.3.1 Reactor operation......................................................................................................... 107
5.3.2 Products yields analysis................................................................................................108
5.3.3 Specific surface area development...............................................................................119
5.4 Conclusions..................................................................................................................... 121
CHAPITRE 6 : ARTICLE 4 - ETUDE DE LA COMPOSITION CHIMIQUE DES
EXTRAITS DE BIOCHARBONS ET DES BIO-HUILES OBTENUS DE LA
PYROLYSE DU PANIC ERIGE EN ATMOSPHERES DE C 0 2 vs N2...................................123
6.1 Introduction..................................................................................................................... 126
6.2 Materials and methods................................................... 128
6.2.1 Experimental setup...................................................................................................... 128
6.2.2 Solid material extraction and bio-oil preparation..........................................................129
6.2.3 Product characterization and GC-MS analysis............................................................. 129
6.3 Results and discussion.....................................................................................................130
6.3.1 Char extracts and bio-oils chemical composition..........................................................130
6.3.2 Calorific value............................................................................................................. 144
6.4 Conclusions...................................................................................................................... 145
CHAPITRE 7 : CONCLUSION......................................................................................... 148
7.1 Sommaire : ........................................................................................................................148
7.2 Contributions:................................................................................................................. 151
7.3 Travaux futurs :................................................................................................................ 153
ANNEXE A PHOTOGRAPHIES DU MONTAGE ET REACTEUR 1 gbi0masJb a tc h .... 156
ANNEXE B SCHEMA ET PHOTOGRAPHIES DU MONTAGE ET REACTEUR
25 gbiomassJbotch ...........................................................................................................................158
ANNEXE C EXEMPLE DE CALCUL POUR LES VALEURS DU TABLEAU 3.3....... 163
ANNEXE D COMPOSITION ELEMENTAIRE METALLIQUE DE PANIC ERIGE..... 164
REFERENCES ...........................................................................................................................165

v
LISTE DES FIGURES
Figure 1.1: Projections de l’accroissement de la population mondiale.................................. 2
Figure 1.2 : Projection de la consommation energetique mondiale..........................................2
Figure 2.1 : Representation schematique de la paroi cellulaire des plantes.......................... 13
Figure 2.2 : Le cellobiose, dimere de D-glucose est 1’unite de repetition du polymere
de cellulose.............................................................................................................14
Figure 2.3 : Principaux composes de l’hemicellulose..............................................................15
Figure 2.4 : Structure des unites de phenylpropane C6 C 3 presentes dans la lignine............. 15
Figure 2.5 : Decomposition dans le sol d ’une biomasse non carbonisee vs du biochar........18
Figure 2.6 : Emissions de CO 2 generees par l’application reguliere de biomasse vs du
biochar en sols.......................................................................................................19
Figure 2.7 : Analyses thermogravimetriques des composantes principales de la
biomasse.................................................................................................................29
Figure 2.8 : Diagramme Van Krevelen de diverses matieres solides..................................... 31
Figure 2.9 : Analyses thermiques differentielles de Festuca mairei lavee vs non lavee....35
Figure 2.10 : Temperature de degradation de la cellulose selon le contenu en sodium et
potassium............................................................................................................... 36
Figure 2.11: Developpement de la surface specifique du biocharbon de bambou en
fonction de la temperature....................................................................................37
Figure 2.12 : Evolution de la structure du charbon de biomasse en fonction de la
temperature............................................................................................................ 38
Figure 2.13 : Modification de la structure du biocharbon en fonction de la temperature......39
Figure 2.14 : Bilans des produits de pyrolyse rapide de rafles de mai's obtenus dans
divers environnements gazeux.............................................................................41
Figure 2.15 : Analyses thermogravimetriques differentielles de pyrolyse de charbon
mineral en atmospheres de N 2 ou de CO2 ........................................................... 42
Figure 2.16 : Etude thermogravimetrique de pyrolyse de charbon mineral en
atmospheres de N2 ou de CO 2 ..............................................................................42
Figure 2.17 : Effet du taux de chauffe sur la pyrolyse de charbon mineral en atmosphere
de C 0 2.................................................................................................................... 43
Figure 2.18: Monoxyde de carbone produit lors de la conversion de charbon mineral en
atmospheres de CO2 ou de N 2 .............................................................................. 43
Figure 2.19 : Bilans des produits de pyrolyse obtenus en atmospheres de C 0 2 ou de N 2 . ...44
Figure 2.20 : Analyses thermogravimetriques et analyses thermodifferentielles de boues
municipales en atmospheres comparatives de N 2 ( • ) ou de CO 2 ( ° ) ...............45
Figure 2.21 : Production de CO et CO 2 lors de la pyrolyse de boues municipales en
atmosphere de CO2 ou de N2................................................................................46
Figure 2.22: Concentration de groupes de composes chimiques obtenus des bio-huiles
de la pyrolyse rapide de rafles de mai's sous environnements de N2, CO 2 ,
CO, CH4 o u H2...................................................................................................... 48
Figure 2.23 : Classes de composes de produits condensables obtenus en atmospheres de
N2 (a) ou de C 0 2 (b)..............................................................................................49
Figure 2.24 : Spectres FTIR de biocharbons de boues municipales produits par pyrolyse
en atmospheres de N 2 (a) ou de CO2 (b)............................................................. 50
Figure 2.25 : Resultats ATG de lignine en atmosphere de CO 2 . A : de 22 a 860 °C a
1 °C/min, B : de 22 a 930 °C a 1 °C/min, C : de 22 a 1000 °C a
100 °C/min.............................................................................................................55
Figure 2.26 : Effet de la pression et du debit gazeux (inerte) sur le rendement en
biocharbon lors de la pyrolyse de cellulose........................................................ 58
Figure 2.27 : Contenu en HAP de divers types de charbons de biomasse.............................. 60
Figure 3.1: Schematic representation o f the experimental setup........................................... 67
Figure 3.2: Side view of biomass holder..................................................................................67
Figure 3.3: Bottom view of biomass holder............................................................................ 67
Figure 3.4: Example o f a representative temperature profile within biomass holder
during a 1g/batch test. (500 °C, 5 min)............................................................... 68
Figure 3.5: Char yield obtained from pyrolysis of switchgrass at 300, 400 and 500 °C,
for 2.5 or 5 minutes under N 2 atmosphere.......................................................... 71
Figure 3.6: Scanning electron microscopy pictures o f raw switchgrass and switchgrass
chars. Enlargement = 45X.................................................................................... 74
Figure 3.7: Scanning electron microscopy pictures o f raw switchgrass and switchgrass
chars. Enlargement = 500X.................................................................................. 76
Figure 3.8: FTIR spectra of chars obtained using attenuated total reflectance
equipment (ATR) for 300-500 °C and 2.5 min pyrolysis conditions............... 80
Figure 4.1: Schematic representation o f the experimental setup........................................... 91
Figure 5.1: Schematic representation o f the experimental setup......................................... 107
Figure 5.2: Condensable products amounts collected along sections of the condensing
train.......................................................................................................................108
Figure 5.3: Pyrolysis products yields..................................................................................... 115
Figure 5.4: Char elemental organic content expressed on original feedstock basis...........116
Figure 5.5: Liquid elemental organic content expressed on original feedstock basis........117
Figure 5.6: Example o f reactor temperature profile at 500 °C - N 2 obtained during the
special air treatment following normal test.......................................................121
Figure 6.1: Mass o f extracts generated by Soxhlet extractions of chars using
dichloromethane.................................................................................................. 131
Figure 6.2: Relative content of various categories of identified compounds in the char
extracts................................................................................................................. 143
Figure 6.3: Relative content of various categories of identified compounds in the bio­
oils fractions.........................................................................................................143
Figure 6.4: Higher heating value of chars and bio-oils produced under various
temperature and vector gas conditions.............................................................. 145
LISTE DES TABLEAUX
Tableau 2.1 : Analyse elementaire de sources de biomasse...................................................... 11
Tableau 2.2 : Fractions des macromolecules constituant divers types de biomasses
lignocellulosiques....................................................................................................12
Tableau 2.3 : Sommaire des precedes thermochimiques de production de charbon de
biomasse.................................................................................................................. 27
Tableau 2.4 : Rendements en biocharbons de lignine et de cellulose obtenus par ATG en
atmospheres de CO2 ou de N2................................................................................ 52
Table 3.1: Production mode, yield, composition, heating value and surfacearea of
representative switchgrass chars............................................................................ 73
Table 3.2: Switchgrass characterization.................................................................................. 73
Table 3.3: Losses occurring from original switchgrass for char production........................ 73
Table 3.4: Identified compounds from dichloromethane extractions of switchgrass
chars..........................................................................................................................81
Table 3.5: Amounts of extracted material from switchgrass chars........................................ 81
Table 4.1: Torrefaction and pyrolysis product analyses.........................................................96
Table 4.2: Original composition of Panicum virgatum straw ............................................... 96
Table 5.1: Switchgrass pyrolysis products characterization................................................. 118
Table 5.2: Switchgrass characterization.................................................................................118
Table 5.3: Reactions possibly occurring under pyrolysis conditions.................................. 119
Table 5.4: Char surface area after specific treatments.......................................................... 120
Table 6.1: Physicochemical characteristics of raw switchgrass, chars and bio-oils
derived from switchgrass................................................................................... :. 132
Table 6.2: GC-MS compound identification of DCM char extracts and bio-oils............... 140
LISTE DES ACRONYMES
ATG: Appareil thermogravimetrique
B EFSC I: Bioenergy and Food Security Criteria and Indicators (Criteres et indicateurs sur
la bioenergie et la securite alimentaire)
BET: Methode Brunauer-Emmett-Teller
b.h. : Base humide
CHP : Combined Heat and Power (production combinee de chaleur et d'electricite ou
cogeneration)
DCM : Dichloromethane (dichloromethane)
DTG : Analyse thermogravimetrique differentielle
E IA : U.S. Energy Information Administration
EPA : U.S. Environmental Protection Agency
FAO : Food and Agriculture Organization of the United Nations (Organisation des
Nations Unies pour l’alimentation et Tagriculture)
FID : Flame ionization detector (detecteur a ionisation de flamme)
FTIR-ATR: Fourier transform infrared spectroscopy-Attenuated total reflectance
(Spectrometre infrarouge a transformee de Fourier-reflectance totale attenuee)
GC-MS : Gas chromatography-mass spectrometry (Chromatographie en phase gazeuse-
Spectrometrie de masse)
GES : Gaz a effet de serre
GFC : Groupement Fran 9 ais de Combustion
HAP : Hydrocarbure aromatique polycyclique
IC PE S: Inductively-coupled plasma emission spectrometer
LCA : Life cycle analysis (Analyse de cycle de vie)
MEB : Microscope electronique a balayage (Scanning electron microscopy)
MEB-EDX : Microscopie electronique a balayage avec spectrometrie par rayons X a
dispersion d'energie
NIST: National Institute o f Standards and Technology
ONU: Organisation des Nations Unies
PAH : Polycyclic aromatic hydrocarbon
RetD: Recherche et developpement
SEM : Scanning electron microscopy (Microscope electronique a balayage)
TCD : Thermal conductivity detector (Detecteur a conductivity thermique)
U NCTA D: United Nations Conference on Trade and Development (Conference des
Nations Unies sur le commerce et le developpement; CNUCED)
w.b. : Wet basis
CHAPITRE 1 : INTRODUCTION

1.1 Mise en contexte et problematique

La hausse de la population mondiale et la gestion des ressources necessaires pour sa
subsistance represented des enjeux majeurs pour nos societes actuelles et futures. II est estime
que la population mondiale pourrait passer de 7 a 9.25 milliards d’habitants d’ici 2050 (Figure
1.1) [ONU, 2011]. Cette hausse de la population s ’accompagnerait certes par une
augmentation de la nourriture requise, de l’energie (Figure 1.2), de biens materiels consommes
ainsi que par la necessite de gerer les impacts lies a cette consommation. II est estime que la
pression exercee par ces changements sur l’usage des terres arables se solderait par un besoin
de doubler la production alimentaire actuelle [Foley et ah, 2011], La consommation mondiale
d ’energie devrait, quant a elle, augmenter d ’environ 35 % par rapport a la consommation
actuelle, et ce, d ’ici les 20 prochaines annees (Figure 1.2). Ainsi, ce besoin d’energie et de
biens de consommation, dans une realite ou les ressources fossiles sont limitees, peut signifier,
la necessite d’une gestion durable des ressources renouvelables dont la biomasse fait partie.
Cependant, la population doit avant, toutes autres activites, assurer sa securite alimentaire
[FAO, 2012; Foley et al., 2011], bien qu’elle necessite a la fois energie et produits de
consommation afin de subsister.

Selon une etude recente de l'Organisation des Nations Unies pour l'alimentation et l'agriculture
(FAO), les criteres et indicateurs sur la bioenergie et securite alimentaire (BEFSCI) suggerent
que les besoins alimentaires de la population grandissante et les espaces necessaires pour la
valorisation non alimentaire de la biomasse peuvent cependant cohabiter [FAO, 2012]. Tel
qu’indique par le Sous-Directeur general de la FAO, Alexander Mueller, « La production de
bioenergie, si elle est entreprise de maniere responsable et opportune, peut permettre aux
agriculteurs et aux populations rurales de participer a la construction d’une nouvelle economie
verte, et contribuer a contrer les effets de decennies de sous-investissements dans l’agriculture
et les zones rurales du monde en developpement». Ce demier precise cependant que « le
developpement des bioenergies ne doit en aucun cas compromettre la securite alimentaire » et
que son developpement « doit etre administre judicieusement, et doit etre regi par des

1
principes directeurs d ’ordre social, tels que le developpement rural durable, la reduction de la
pauvrete et la securite alimentaire » [FAO, 2012].

Population mondiale [x 109 habitants]

- i t - Constant fertility variant

High fertility variant

Medium fertility variant
- o — Instant replacement fertility variant
Low fertility variant

2100-
Figure 1.1: Projections de 1’accroissement de la population mondiale.
Source : ONU [2011]

250.0

200.0

<u 150.0

5 loo.o
c
v

1990 2000 2008 2015 2020 2025 2030 2035

Annees

Figure 1.2 : Projection de la consommation energetique mondiale.

Source : EIA [2011]

2
En general, le constat est clair : un usage plus intensif des terres arables requiert plus d’eau,
plus de fertilisants et genere une plus grande pression sur les sols (dont sur les macros et
microelements des sols et la matiere organique). L’agriculture represente 70 % de la
consommation hydrique mondiale, tous domaines confondus [FAO, 2011], De plus, pour
assurer une agriculture viable, les besoins en matiere organique du sol sont essentiels [Brady
et Weil, 2002; Olivier, 2007], La matiere organique est generalement renouvelee en partie par
les residus agricoles, ces residus etant grandement convoites par d ’eventuelles bioraffmeries.

Un autre enjeu majeur auquel la population mondiale fait face est l’accroissement incessant de
la quantite de gaz a effet de serre (GES) de source anthropique (CO 2 ou autres GES regroupes
sous l’appellation CO2 e q u iv a le n t), issus majoritairement de la combustion [Olivier, 2007], Ainsi,
la population mondiale est confrontee a une realite oil les besoins des individus et les
ressources disponibles deviennent en concurrence et les impacts doivent etre minimises.

Ce projet de recherche vise l’etude d’un bioproduit pouvant constituer un element de solution
pour l’ensemble de ces enjeux. Ce bioproduit, le charbon de biomasse (biocharbon) est parmi
les plus anciens produits synthetiques fabriques par l’humain. II est utilise et etudie depuis
plus de 38 000 ans, et ce, principalement a des fins energetiques (charbon de bois) [Antal et
Gronli, 2003],

Depuis les demieres annees, l’engouement pour le charbon de biomasse a titre d ’amendement
pour le sol (c.-a-d. : biochar) a nettement retenu l’attention de la communaute scientifique
suite a la parution des deux articles de Johannes Lehmann et Emma Marris dans la revue
Nature [Lehmann, 2007; Marris, 2006]. II fut observe que des charbons de biomasse utilises
comme amendement dans certains types de sols permettraient: une augmentation de la
production des recoltes (jusqu’a 2 fois), une reduction des besoins en fertilisants (grace au
pouvoir adsorbant du biochar) et la sequestration du carbone par la resistance du biocharbon a
la degradation. De plus, produit a plus basses temperatures (<300 °C), un charbon de
torrefaction (pyrolyse douce) se presente comme une pretransformation de la biomasse
pouvant faire beneficier le reste de la chaine de valorisation de la biomasse incluant le passage
menant aux bioenergies et bioproduits de chimie « verte ». Ce benefice est rendu possible

3
grace aux avantages que le procede peut conferer a la matiere premiere : par 1’augmentation de
la densite energetique du produit qui se traduit en des economies lors du transport, la
resistance a la biodegradation pouvant ameliorer l’entreposage de la matiere et la reduction
des proprietes mecaniques de la biomasse facilitant le dechiquetage de la matiere premiere, qui
se solde en des economies energetiques [Felfli et al., 2005; Phanphanich et Mani, 2011; Uslu
et al., 2008; Zwart et al., 2006],

La recherche et le developpement entourant le charbon de biomasse ont jusqu’a maintenant ete

concentres davantage sur le charbon de bois pour valorisation energetique, sur le charbon de
biomasse en tant que residu de pyrolyse rapide qui vise a priori la production de bio-huile et
finalement sur la production de charbon active. Par ailleurs, 1’etude du biocharbon ciblant les
applications de biochar (amendement pour le sol) et charbon vert (valorisation en energie ou
chimie « verte ») constitue une approche encore tres peu exploree, mais dont le potentiel
justifie l’avancement de son developpement. Le biocharbon est un terme au sens large
representant un produit dont les caracteristiques peuvent varier immensement selon une
panoplie de facteurs. Dans le domaine agronomique, il a ete observe que le biochar pouvait
etre tres benefique pour le rendement des cultures. Toutefois, les sols et les types de cultures
sont aussi d ’une grande variability et la comprehension de l’interaction biochar-sol-plante est
encore tres jeune. Utilise pour des essais agronomiques, le biochar a souvent ete le residu de
divers procedes thermochimiques, obtenu de conditions n ’ayant pas ete ciblees pour une telle
application. De plus, certaines analyses de caracteristiques du biochar, telle l’analyse de ses
extraits, ont ete tres peu repertoriees mais sont reconnues par la communaute scientifique
comme pouvant ameliorer la comprehension de ce bioproduit. Ce projet de recherche explore
certaines de ces conditions de production par pyrolyse a basses temperatures et analyse les
caracteristiques physico-chimiques jugees comme pouvant contribuer a l’avancement du
domaine du biocharbon et a la communaute scientifique en general.

1.2 Definition du projet de recherche

Telles que presentees dans la section precedente, deux options de valorisation sont visees par
la presente etude, le biochar (amendement pour le sol) et le charbon vert (pour energie et
procedes subsequents), avec l’accent porte sur le biochar. Le potentiel du charbon vert est
4
aussi aborde etant donne ses conditions de production avoisinant celles du biochar et la
complementarite des deux produits dans la chaine de valorisation de la biomasse.

Dans le cadre de ce projet, la production du charbon de biomasse est etudiee a 3 temperatures

distinctes; 300, 400 et 500 °C. Cette plage de temperatures, relativement basses, a ete choisie
pour l’obtention d ’un rendement interessant en charbon de biomasse, car ce dernier diminue
en fonction d’une augmentation de la temperature et parce que des caracteristiques d ’interet
pourraient etre obtenues a ces temperatures selon les rapports actuels. Bien que la temperature
soit le parametre considere comme ayant le plus d’effet sur les caracteristiques des
biocharbons, plusieurs autres facteurs sont aussi tres significatifs. Parmi ces parametres
d’operation additionnels, les parametres suivants ont egalement ete selectionnes : de courts
temps de residence du biocharbon en reacteur, un taux de chauffe modere-eleve, une pression
quasi atmospherique, un echange thermique et un transfert de masse favorise en reacteur. Le
choix de ces conditions de production est base sur l’analyse de connaissances rapportees par la
communaute scientifique comme pouvant favoriser certaines caracteristiques souhaitees pour
le biochar et le charbon vert, ainsi que sur 1’estimation des conditions industriellement
applicables qui pourraient mener a la conception d’unites mobiles de pyrolyse. Le
developpement d ’unites mobiles de pyrolyse pourrait faire en sorte d’eviter le deplacement de
la biomasse brute, souvent tres humide, et ainsi cela permettrait un traitement in situ de celle-
ci, ce qui representerait un benefice potentiel pour l’economique du procede et/ou de la chaine
d’approvisionnement en biomasse. Afin de convertir la biomasse par torrefaction ou pyrolyse
a un rythme s’approchant de la vitesse de recolte, soumettre la biomasse a de courts temps de
residence en reacteur serait souhaitable pour restreindre le volume des reacteurs. De plus,
l’usage de l’azote comme milieu gazeux inerte peut representer un defi dans le cas des unites
portables de pyrolyse. Dans cette optique de conception de l’unite mobile, et pour la
valorisation d’un produit gazeux residuel, l’utilisation du CO2 comme environnement reactif
est etudiee dans le cadre des presents travaux. Bien qu’un melange gazeux facilement
accessible pour des unites portables serait l’utilisation des gaz d ’echappement des moteurs, le
C 0 2 etant un constituant majeur de ce melange gazeux justifie done l’interet d ’explorer
l’utilisation de ce gaz lors des reactions de pyrolyse. De plus, la valorisation du CO2 en
general est nettement d ’interet d’un point de vue de reduction des gaz a effet de serre. L ’etude

5
de l’usage du CO 2 lors de la pyrolyse s’est done effectuee dans les conditions prealablement
mentionnees, mais tout en comparant la conversion et les produits obtenus a un environnement
gazeux inerte conventionnel d’azote (N2 ). Ces conditions ont tres peu ete rapportees jusqu’a
maintenant, mais sont ciblees comme un besoin par la communaute scientifique [Duan et al.,
2009; Jindarom et al., 2007a; Naredi et Pisupati, 2011]. Parmi les autres conditions
d ’operations du present projet, l’usage de biomasse a dimension brute (n’ayant pas ete
prealablement broyee) est vise pour l’applicabilite de la technologie sans l’etape energivore de
la pulverisation de la biomasse. Le panic erige (Panicum virgatum) est la plante graminee
utilisee pour les tests, etant donne son potentiel deja demontre comme culture en climats et
terrains hostiles de l’Amerique du Nord, ainsi que pour son usage comme plante energetique
dediee. Le panic erige est une plante vivace, native de l’Amerique du Nord et qui necessite
peu d ’intrants (fertilisants ou pesticides) [David et Ragauskas, 2010; Mullen et Boateng,
2008],

En termes de caracterisation des charbons de biomasse, les analyses de routine sont les bilans
et les analyses immediates, elementaires et calorifiques. Toutefois, d’autres caracterisations
ont ete tres peu effectuees et consisteraient en un apport considerable pour la communaute
scientifique. Pour les conditions de pyrolyse a basses temperatures (300 a 500 °C), tres peu
d’analyses des extraits de charbons de biomasse sont presentees en litterature. Certaines visent
seulement quelques composes specifiques tels des HAP et dioxines [Hale et al., 2012], Ce
type d’analyse a recemment ete soul eve par la communaute scientifique comme etant un
bassin de connaissances qui pourrait contribuer significativement au domaine du biochar
[Krull et al., 2009]. De plus, dans le domaine de la pyrolyse en general, l’analyse des
composes chimiques s’est plutot concentree sur celle des bio-huiles. Par consequent et par la
meme occasion, ces tests peuvent contribuer a mieux comprendre les mecanismes de
conversion de la matiere premiere et de la formation du charbon de biomasse.

6
1.3 Objectifs du projet de recherche
• Concevoir les reacteurs (bancs d ’essais) necessaires permettant l’etude de la
production et des caracteristiques du biocharbon
• Etudier la production de biocharbon a basse severite (temperature), a des conditions
applicables pour des unites de pyrolyse portables industrielles
• Etudier l’effet du CO 2 lors du procede de pyrolyse et l’impact de ce dernier sur les co-
produits obtenus, avec I’accent porte sur le charbon de biomasse
• Caracteriser le charbon de biomasse et particulierement en extraire une partie de son
contenu chimique et 1’analyser

1.4 Contributions originales

Telle qu’abordee dans les sections precedentes, cette etude contribue a la communaute
scientifique a plusieurs egards. Premierement, l’approche suivie lors du projet vise l’etude
d’une production dediee de biochar (tout en considerant le charbon vert) et non la valorisation
d’un residu de pyrolyse visant l’optimisation d’un autre co-produit (ex. : pyrolyse rapide et
bio-huile). Cette forme de contribution devient originale par les conditions d ’operation
specifiques qui sont explorees (1’usage de taux de chauffe moderes a rapides, temps de
residence du biocharbon en reacteur courts, convection forcee elevee et temperatures de faible
intensite (300 a 500 °C)) et par les resultats de caracterisations des biocharbons et co-produits
qui y sont obtenus. De plus, le fait que les tests visent des conditions favorables pour une
application industrielle, pouvant mener au developpement d ’unites portables de pyrolyse,
represente egalement une forme de contribution. Celle-ci est d ’autant plus interessante etant
donne qu’elle genere des resultats uniques et d ’interet pour la continuation de la R et D sur la
valorisation du biocharbon (biochar et charbon vert) et la comprehension de la pyrolyse en
general. La presentation et l’usage de deux nouveaux bancs d’essais laboratoires avec
reacteurs originaux parametrables, qui permettent l’etude de la production du charbon de
biomasse, represented aussi un apport. Certaines conditions explorees et caracteristiques
d’interet observees peuvent contribuer a l’avancement de la valorisation du biocharbon
comme amendement pour le sol, alors que d ’autres, particulierement celles a 300 °C, cement
des conditions qui apparaissent comme nettement avantageuses pour le charbon vert.

7
L’identification de composes chimiques des extraits de biocharbons compte parmi les
contributions importantes apportees. Ce domaine avait recemment ete souleve par la
communaute scientifique comme necessitant une recherche approfondie, specialement pour
1’application du biochar, mais aussi, pour le domaine de la pyrolyse en general, dans lequel ce
type d’analyse est tres peu retrouve. Dans le domaine du biochar en particulier, la publication
de composes identifies contribue a ce que les chercheurs(ses) du domaine agronomique
puissent etablir des liens entre les applications de biochar et la reponse par le sol et les plantes
cultivees. De plus, les conditions d ’operation du procede, etant mentionnees, elles pourront
contribuer, suite a une comparaison avec d’autres etudes similaires, a la comprehension entre
les caracteristiques chimiques obtenues selon diverses conditions de production de biocharbon
et la variete de facteurs pouvant influencer la composition chimique de ces biocharbons. La
variete de facteurs en question se rapporte principalement aux types et aux caracteristiques de
biomasse ainsi qu’aux conditions de pyrolyse. Ces analyses ont mene, par la meme occasion, a
l’observation de composes particulierement d ’interet pour l’usage du biocharbon en sol
(biochar), mais egalement, pour le charbon vert, le biocharbon en general et le procede de
pyrolyse en tant que tel. Pour le procede de pyrolyse, le fait d ’explorer les extractions de
biocharbons a plusieurs temperatures, de concert avec les bio-huiles (uniquement dans le
chapitre 6), permet de faire la lumiere sur des mecanismes de pyrolyse et la production de
biocharbon, et ce, autant pour les tests plus conventionnels en atmosphere d’azote, que pour
les tests effectues en atmosphere de CO 2 (les tests en presence de CO2 ont aussi ete soumis
aux memes analyses de caracterisation).

Finalement, l ’etude de l’ajout du CO2 lors de la pyrolyse a basse severite represente une
contribution majeure en soit. La pyrolyse en atmosphere de CO2 avait recemment ete reconnue
par la communaute scientifique comme necessitant davantage d ’investigations du au potentiel
et au besoin environnemental de valoriser ce gaz residuel [Duan et a l, 2009; Jindarom et a l,
2007a; Naredi et Pisupati, 2011]. Tres peu d ’etudes semblent avoir ete effectuees sur la
pyrolyse de matiere lignocellulosique en atmosphere de CO 2 . Ainsi, tous les resultats et les
observations notes dans ces conditions contribuent a l’avancement de la valorisation de ce gaz
pour le procede de pyrolyse en general, et pour la valorisation des co-produits obtenus.
L ’exploration de l’usage du CO2 a ces conditions a permis d ’observer des differences

8
significatives en comparaison avec 1’environnement d ’azote gazeux. Parmi les differences
significatives notees, certaines d’entre elles ouvrent des pistes d’explorations futures comme
plusieurs autres observations mentionnees tout au long du projet. La presentation de ces pistes
de travaux futurs represente aussi, en quelque sorte, une forme de contributions.

1.5 Plan du document

Le present document est separe en 6 sections principales. Le cadre theorique couvre tout
d’abord les notions essentielles necessaires a l’exploration du sujet et qui ont mene a la
conception des deux reacteurs, aux conditions d ’operation des reacteurs, ainsi qu’au plan
experimental. L ’approfondissement de l’etat de l’art est effectue sur les conditions de
production et des caracteristiques plus specifiques du biocharbon et applicables a ce projet.
Ces informations sont presentees tout au long du chapitre 2 et approfondies lors de l’analyse
des resultats. Les resultats du projet sont presentes dans les chapitres 3 a 6 sous forme
d ’articles scientifiques ayant ete publies dans des joumaux ou chapitre de livre evalues par les
pairs. Les articles (chapitres) sont presentes par ordre chronologique, suivant l’ordre dans
lequel les etudes ont ete effectuees. L ’article 1 (chapitre 3) presente une etude de la production
de biocharbon en atmosphere de N 2 a 300, 400 et 500 °C, a l ’aide du reacteur de 1 g/batch,
comparant 2 courts temps de residence en reacteur, soit 2 et 5 minutes. L ’article 2 (chapitre 4)
presente une etude preliminaire de l’usage du CO 2 , effectuee avec le reacteur de 1 g/batch,
comparant les atmospheres de N 2 et CO 2 ainsi que l’impact de ces demiers sur les
biocharbons. L’article 3 (chapitre 5) approfondit l’etude de l’usage du CO2 grace a l’utilisation
d’un reacteur de plus grande capacite (25 g/batch), permettant ainsi un suivi accru sur
1’ensemble des produits de pyrolyse (biocharbons, bio-huiles et gaz non condensables). Cette
etude porte d ’ailleurs specialement sur le bilan des co-produits pyrolytiques obtenus. L ’article
4 (chapitre 6), quant a lui, presente une etude specifique a la composition des extraits de
biocharbons et des bio-huiles obtenus avec le reacteur de 25 g/batch et comparant la pyrolyse
en atmospheres de CO 2 et N 2 . Le chapitre 7 conclut la these par un retour sur l’ensemble des
travaux effectues dans le cadre du projet; les contributions au domaine scientifique y sont
presentees plus en details et des pistes de travaux futurs y sont proposees.

9
CHAPITRE 2 CADRE DE REFERENCE ET
ETAT DE L’ART

2.1 La biomasse
Le biocharbon etant un produit issu de la biomasse, bien qu’il soit influence par les conditions
en reacteur (Section 2.3), ses caracteristiques proviennent, en partie, directement de la matiere
premiere a partir de laquelle il est obtenu. Dans cette section, les types de biomasses et les
caracteristiques generates liees a la biomasse lignocellulosique y sont presentes.

2.1.1 Types de biomasses

Plusieurs types de matieres premieres peuvent entrer sous la definition de « biomasse ». La

biomasse est essentiellement une substance issue de matiere biologique (vegetate ou animale).
Cependant, lorsqu’il est question de biomasse energetique, cette matiere premiere implique les
cultures agricoles dediees, les produits forestiers et marins, les boues municipales ou
d’industries, mais aussi les dechets municipaux dont la composition renferme des produits de
source « petroliere ». La biomasse fait cependant habituellement reference a une matiere
premiere de source non fossilisee [Demirbas, 2000], La « biomasse » contient principalement
les elements organiques suivants : le carbone, l’hydrogene et l’oxygene (C, H ,0) ainsi qu’une
quantite moindre d’azote et des quantites minimes de souffe (Tableau 2.1). La composition de
la biomasse en souffre est habituellement de <0,5 % en comparaison a environ 0,5 a 4 % pour
plusieurs types de charbons « mineraux » [Demirbas, 2004; The Engineering Toolbox, 2012],
La biomasse peut etre categorisee aussi en fonction de son degre d’homogeneite chimique.
Une biomasse tres homogene peut correspondre a une partie precise d ’une plante (ex. : bran de
scie provenant en majeure partie de l’aubier et du duramen d ’un tronc d’arbre). Une biomasse
dite non homogene peut correspondre a un melange de sections de la plante (e x .: ecorce,
aubier et rameaux), a un melange d ’especes de plantes et meme jusqu’au bois residuel
comprenant divers degres de contamination. Une biomasse dite « chimiquement heterogene »
correspond a diverses sources de biomasses melangees dont le degre de contamination avec

10
des matieres autres que biologiques peut survenir (ex. : des plastiques dans une biomasse
« urbaine ») [Chomet, 2005],

Tableau 2.1 : Analyse elementaire de sources de biomasse

Type de biomasse C H N S O (diff.)

Paille de ble 41,8 5,5 0,7 - 35,5
Bois d'epinette 51,9 6,1 0,3 - 40,9
Rafle de mai's 49,0 5,4 0,5 0,2 44,5
Tige de mai's 42,5 5,0 0,8 0,2 42,6
Panic erige 46,7 5,9 0,8 0,2 37,4
Source : Demirbas [2004]

Dans le cadre de la presente etude, il est question de biomasse dite lignocellulosique. La

biomasse lignocellulosique comprend les cultures energetiques dediees dont des plantes
herbacees (ex. : le panic erige, Panicum virgatum L.) ou des essences de bois, tels les taillis en
courte rotation (e x .: essences de saule, Salix). La biomasse lignocellulosique comprend aussi
la biomasse residuelle (ex. : residus de culture de mais). Les composantes de ces types de
biomasse sont decrites plus en details dans les sections suivantes. Le panic erige, le saule et les
residus de mai's sont des exemples de biomasses non homogenes etant donne qu’il existe une
disparite entre les differentes parties d ’une meme plante (ex. : composition elementaire des
rafles et des tiges de mai's; Tableau 2.1).

2.1.2 Composantes de la biomasse

La biomasse vegetale est constitute principalement de matiere organique ainsi que de matiere
inorganique en plus faible quantite. La matiere organique a l’echelle macromoleculaire
consiste en la cellulose, l’hemicellulose, la lignine et de matieres extractibles. Les polymeres
de cellulose et d’hemicellulose sont formes de glucides et referent ensemble a l’holocellulose.
Avec la lignine, ces trois types de macromolecules composent les parois cellulaires des plantes
(Figure 2.1). C’est la lignine qui donne cependant la rigidite a la fibre vegetale. Les matieres
extractibles, quant a elles, represented principalement le materiel non structural de la plante.
Elies sont extractibles a l’aide de solvants (tels l’eau et/ou l’ethanol) et sont constitutes de

11
cires, graisses, resines, tannins, sucres non structuraux, amidons, pigments et autres [Guo et
al., 2010; Sluiter et al., 2008], Une fraction des plantes est aussi constitute de matiere
inorganique, cette demiere ayant ete assimilee lors de la croissance de la plante a partir du sol
ou recueillie lors de manipulations de la biomasse (ex. : paille en contact avec le sol et
particules de sol restant collees a sa surface). Un des defis concemant la valorisation de la
biomasse est essentiellement la variabilite dans la composition d ’une meme plante ainsi
qu’entre les types de plantes (Tableau 2.2). Les proportions de chacun des constituants de la
biomasse lignocellulosique varient selon les facteurs suivants [Rousset et al., 2011;Vassilev et
al., 2010; Yang et al., 2006] :
• le type de biomasse, le type d’espece ou la partie de la plante;
• les conditions climatiques et geographiques pouvant affecter la croissance (ex. :
periode d ’ensoleillement, sol, eau);
• l’age de la plante;
• les fertilisants et pesticides utilises (affectant principalement les elements
inorganiques);
• les saisons de recolte, les techniques de recolte, de transport et d’entreposage.

Tableau 2.2: Fractions des macromolecules constituant divers types de biomasses

lignocellulosiques

Type de biomasse Cellulose Hemicellulose Lignine References

[%] [%] [%]
Coniferes 35-50 20-32 25-35 [Prins et al., 2006b]
Feuillus 40-50 15-35 18-25 [Prins et al., 2006b]
Tige de mai's 32-36 19-22** 11-15 [Weiss et al., 2010]
Paille de ble 38-41 29-31 8-14 [Butterman et Castaldi, 2009]
et [Bridgeman et al., 2008]
Saule 49* 14* 20* [Bridgeman et al., 2008]
Ecorce d’erable 18 33 19 [Butterman et Castaldi, 2008]
*Base seche sans cendre, **xylane seulement

12
 Bioenergy Crop
■ Y Plant Cells

Plant Cell Wall

V ,^'~-Ugnin
Cellulose H em icellulose
Microfibril
C ellu lose

Sugar
Molecules
Glucose

Figure 2.1 : Representation schematique de la paroi cellulaire des plantes.

Source : Sannigrahi et al. [2010]

Cellulose
La cellulose est un polysaccharide, forme de monomeres de glucides, mais plus
particulierement de D-glucose ((C6 Hio0 5 )n), dont l’unite de repetition est le cellobiose, un
dimere de glucose (Figure 2.2) [Demirbas, 2000; Lee et al., 2010], La cellulose est constitute
de longs polymeres non reticules qui sont principalement organises sous une forme cristalline
ou pouvant aussi se presenter sous forme entrelacee, correspondant ainsi a la cellulose
amorphe. L ’organisation des polymeres de cellulose de forme cristalline rend ce constituant de
la biomasse tres resistant a l’hydrolyse (reaction de degradation a partir des ions de l’eau; H+
et OH ). La cellulose cristalline est aussi plus resistante a la pyrolyse que la cellulose amorphe
et l’hemicellulose [Rousset et al., 2011]. La degradation de la cellulose et celles des autres

13
composantes de la biomasse lors de reactions thermochimiques sont presentees plus en detail
dans la section 2.3.2.

OH OH

.o h I p i .OH
HO HO
HO' 'O H
HO
OH HO

OH OH

Unite de cellobiose

Figure 2.2 : Le cellobiose, dimere de D-glucose est l’unite de repetition du

polymere de cellulose. Source : Mohan et al. [2006]

Hemicellulose
L ’hemicellulose est composee de glucides polymerises enchevetres avec la cellulose et qui
entrent dans la formation des membranes cellulaires (Figure 2.1) [Reed, 1981], La structure
des polymeres d’hemicellulose est similaire a celle de la cellulose excepte que l’hemicellulose
est reticulee alors que les polymeres de cellulose sont lineaires. La structure des polymeres
d ’hemicellulose, avec la variete des sucres la constituant, confere a 1’hemicellulose une
structure amorphe. Cette structure formee de liaisons faibles la rend moins resistante a la
degradation thermique que la cellulose cristalline [Almeida et al., 2010].

L ’hemicellulose est formee habituellement de 2 a 6 glucides. Le D-xylose et le D-mannose

sont les sucres les plus courants. D’autres sucres, tels que le D-glucose, le D-galactose et le L-
arabinose peuvent egalement s’y retrouver (Figure 2.3). Le xylane, polymere de D-xylose, est
generalement la fraction la plus abondante et la plus reactive des hemicelluloses [Prins et al.,
2006b; Rousset et al., 2011]. Le xylane representerait environ 80-90% de l’hemicellulose des
bois de feuillus et jusqu’au tiers de tous les hydrates de carbone pour l’ensemble des types de
biomasses lignocellulosiques [Reed, 1981; Sun et Cheng, 2002], L’hemicellulose des
coniferes differe de celle des feuillus et est constitute a 60-70% de glucomannanes [Prins et
al., 2006b; Rousset et al., 2011]. Le panic erige (Panicum virgatum), espece employee dans la
presente recherche, a deja ete rapporte comme comportant des glucanes (41,6 %), xylanes
(17,3 %), galactanes (2,0 %), arabinanes (3,3 %) et mannanes (0,8 %) [He et al., 2009],

14
 .OH OH .OH
HO

Glucose Galactose Manose

COOH

CH2OH
Xylose Arabinose Acide glucuronique

Figure 2.3 : Principaux composes de l’hemicellulose.

Source : Mohan et al. [2006]

Lignine
La lignine constitue la portion aromatique et en consequence non-glucidique de la paroi
cellulaire [Reed, 1981; Rousset et al., 2009]. Elle est composee de polymeres phenoliques
dont l’unite principale est le phenylpropane (Figure 2.4) [Reed, 1981; Sharma et al., 2002]. La
lignine comporte une structure complexe, variable, et entrelacee avec les autres polymeres de
cellulose et d ’hemicellulose (Figure 2.1) [Butterman et Castaldi, 2010; Rousset et al., 2009].
Les noyaux aromatiques entoures de groupes fonctionnels varies sont omnipresents dans la
lignine. Le lien principal a l’interieur meme de l’unite de phenylpropane de la lignine est le
lien P-O-4 (Figure 2.4).

HC ( ^ ) V — Lignin

CHOH

OR'

Figure 2.4 : Structure des unites de phenylpropane C 6 C 3 presentes dans la lignine.

(Ri et R 2 variant selon la structure). Source : Rousset et al. [2009]

15
II est important de noter que la lignine extraite de biomasse et souvent utilisee pour des
analyses de conversion thermochimiques (ex. : analyses par ATG) peut varier selon le type de
biomasse et selon le mode d ’extraction utilise pour son isolation de la matrice
lignocellulosique. Ainsi, les etudes qui presentent des resultats de decomposition thermique de
la lignine ne correspondent pas necessairement a sa decomposition lorsque celle-ci est
presente dans la biomasse originale [Mohan et al., 2006], Parmi les precedes menant a
l’isolement de la lignine se trouvent, entre autres, le precede de l’industrie des pates et papier
qui genere la lignine kraft ou lignine alcaline, l’explosion a la vapeur ainsi que divers precedes
de production d’ethanol a partir de biomasse lignocellulosique ou la lignine se retrouve
comme sous-produit [Pandey et Kim, 2011],

Matieres extractibles
Les matieres extractibles constituent la matiere non structurale de la biomasse et peuvent etre
extraites de la biomasse a l’aide de solvants aussi bien polaires que apolaires [Sluiter et al.,
2008]. Les bois mous abondent en resines et terpenes, 2 types de matieres extractibles.
Cependant, en general, la plupart des essences de bois comportent des cires, des acides gras,
des pigments et des glucides (ex. : sucres simples, amidon), tous etant definis comme matieres
extractibles [Bourgois et al., 1989; Guo et al., 2010; Reed, 1981]. Le panic erige (Panicum
virgatum L.) extrait avec 95 % d’ethanol et analyse par GC-MS a demontre la presence de
glucides, d'acides gras, d ’alcools gras, de glycerol, d ’alcanes et de sterols [Yan et al., 2010],
En general, le contenu total en matieres extractibles varie selon le type de biomasse, se situant
toutefois sous 10 % de 1’ensemble de la biomasse [Guo et al., 2010].

2.2 Charbon de biomasse vegetale

Le charbon de biomasse vegetale (biocharbon) fait reference a un produit obtenu lors de
transformations thermochimiques de la biomasse, au contenu en carbone eleve. Lors de
transformations thermochimiques d ’une matiere organique et carbonee (c.-a-d.: contenant du
carbone), Paction thermique vaporise Phumidite et la portion volatile de la biomasse, ce qui a
pour effet de concentrer le residu solide en carbone. La concentration du contenu de la matiere
solide en carbone se nomme la carbonisation. II est a noter que la plupart des matieres

16
inorganiques initialement presentes dans la biomasse se retrouvent egalement dans le
biocharbon. Ces matieres inorganiques peuvent etre modifiees selon leur composition et
l’environnement reactif dans lequel elles se trouvent et peuvent a leur tour avoir influence
(catalyse) les reactions de pyrolyse de la biomasse [Mohan et al., 2006], Les variations de la
composition de la biomasse initiale peuvent affecter directement les caracteristiques du
biocharbon [Downie et al., 2009]. Les processus thermochimiques par lesquels la
carbonisation peut s’effectuer so n t: la torrefaction, la pyrolyse, la gazeification, une
combustion incomplete, la carbonisation hydrothermale, ou toutes variantes de ces precedes
(Tableau 2.3) [Amonette et Joseph, 2009; GFC, 2004; Prins et al., 2006b]. Le precede utilise
et etudie dans la presente recherche pour la production de biocharbon est la pyrolyse ainsi
qu’une sous-division de la pyrolyse, la torrefaction (une pyrolyse douce). Les caracteristiques
du biocharbon sont nettement influencees par ces precedes et specialement par les conditions
en reacteur (ex. : temperature, taux de chauffe, transfert de masse, temps de residence,
conditions oxydatives, pression, environnement gazeux et usage de catalyseurs). D ’autres
facteurs tels que les pretraitements de la biomasse (sechage, broyage) et les post-traitements
(activation, broyage) viennent aussi affecter les caracteristiques du biocharbon [Downie et al.,
2009], Ainsi, la variability des caracteristiques du biocharbon est immense. L ’etude de
facteurs controlables en reacteur permet done de cibler des conditions pour des
caracteristiques et applications specifiques (ex. : biochar produit en fonction du type de sol).
Certains de ces facteurs, dont la composition de la biomasse et les conditions en reacteur, sont
presentes en section 2.3.

2.2.1 Biochar

Le biochar est un charbon de biomasse dont l ’utilisation est specialement vouee a

Pamendement des sols. Un biochar dont les caracteristiques agronomiques sont optimales peut
contribuer a augmenter la productivity des cultures, sequestrer du carbone et adsorber des
fertilisants et certains contaminants, permettant ainsi de reduire leur lessivage [Boateng, 2007;
Keiluweit et Kleber, 2009; Lehmann, 2007; Major et al., 2009; Prins et al., 2006b]. Le
biochar, avec ses caracteristiques variables, peut ainsi influencer les proprietes physiques et
chimiques du sol. Sa porosite en fait un milieu favorisant la croissance des microorganismes
[Laird et al., 2009; Thies et Rillig, 2009], qui a leur tour favorisent la formation d ’agglomerats

17
pouvant contribuer a une meilleure aeration du sol et absorption de l’eau [Brady et Weil,
2002], La consistance du sol se retrouve ainsi affectee tant au niveau de sa texture, sa
structure, sa profondeur, sa porosite, sa surface specifique et sa consistance [Downie et al,,
2009], De plus, l’amendement de biochar dans le sol consiste en une methode de sequestration
de carbone [Gaunt et Lehmann, 2008; Lehmann, 2007; Lehmann et al., 2006]. Cette
particularity du biocharbon a ete observee par les historiens et scientifiques alors que du
charbon de biomasse incorpore dans le sol y serait reste pendant des millenaires, proposant
ainsi l’usage du biochar comme methode de sequestration de carbone. La modification de la
structure de la plante mere en biocharbon genere une structure beaucoup plus resistante a la
biodegradation. Cette structure est obtenue suite a la concentration de la lignine (< 625 °C)
(polymere de la biomasse le plus resistant a la biodegradation) et/ou par la modification meme
de la structure carbonee qui genere une structure plus resistante a la biodegradation. Ainsi, la
degradation du biochar peut s’effectuer sur plus de 100 a 1000 ans, selon les conditions de sa
formation et de sa composition. En comparaison, une biomasse laissee sur le sol se biodegrade
naturellement par oxidation avec l’oxygene de l’air et par la presence des organismes du sol
(organismes heterotrophes) qui utilisent la matiere organique comme source d’energie et de
carbone en l’oxydant, du carbone qui est ultimement relache sous forme de CO2 [Brady et
Weil, 2002], II prend generalement de 5 a 10 ans a la biomasse pour se biodegrader
naturellement (Figure 2.5). C’est done cette difference de vitesse de degradation entre le
biochar et la biomasse brute, jumelee a une application repetee du biochar, qui consisterait en
une methode de sequestration du carbone (Figure 2.6).

100

o>

Bio-char

U n-charred organic m atter

10 ■

1 2 3 4 5
Y ears

Figure 2.5 : Decomposition dans le sol d’une biomasse non carbonisee vs du

biochar. Source : Lehmann et al. [2006]

18
 Reduction
ou agricoles
CO,

Figure 2.6 : Emissions de CO 2 generees par l’application reguliere de biomasse

vs du biochar en sols

Certains elements inorganiques presents dans le charbon de biomasse peuvent quant a eux
contribuer a augmenter le pH du sol. Le compose d ’oxyde de calcium pouvant se retrouver
dans le biochar peut mener a la liberation d ’ions OH' et, par consequent, augmenter le pH du
sol (Reaction : CaO + H20 -> Ca(OH)2 -► Ca2+ + 20H ) [Brady et Weil, 2002], De plus,
certains elements inorganiques presents dans le biochar constituent une source d’echanges
cationiques dans le sol qui peuvent adsorber des nutriments tels les fertilisants azotes qui sont
dans bien des cas lixivies dans le sol. Ces pertes en elements azotes occasionnent ainsi des
couts en pertes d ’engrais et des dommages environnementaux [Major et al., 2009], La capacite
du charbon de biomasse a adsorber des substances toxiques dans le sol pourrait ameliorer la
culture des plantes, affectant leur sante, leur survie et, consequemment, leur productivity
[Antal et Gronli, 2003; Keiluweit et Kleber, 2009; Major et al., 2009], Certains types de
biochar ont la capacite d’adsorber des composes organiques dont des HAP [Keiluweit et
Kleber, 2009; Smemik, 2009].

2.2.2 Charbon de biomasse torrefiee et charbon vert

La torrefaction de la biomasse lignocellulosique s’effectue a des temperatures sous 300 °C. Le

charbon de biomasse issu de la torrefaction correspond a un type de charbon vert {green
charcoal ou green coal ou bio-coal) qui est voue a un usage energetique. Par definition, le
charbon vert peut avoir ete obtenu a des temperatures au-dela de 300 °C, mais il exclurait le
bois, en se limitant ainsi a des residus de biomasses [L'Encyclopedie de_TAgora, 2008;

19
UNCTAD, 2012], II est a noter que cette exclusion serait plutot contradictoire car le bois est
une plante et aussi une biomasse. Malgre cette possible contradiction de definitions,
contrairement au charbon de bois, le charbon de biomasse torrefiee inclut divers types de
biomasses lignocellulosiques, tels les residus agricoles ou cultures energetiques dediees dont
le bois fait partie. Contrairement a la torrefaction (pyrolyse douce) oil le procede se limite par
definition a 300 °C, la production de charbon de bois peut s ’operer a des temperatures
beaucoup plus elevees [Prins et al., 2006b], Les charbons de biomasse torrefiee et charbons
verts sont similaires a un charbon de bois classique en ce qui conceme l’obtention d ’un
produit dont la valeur calorifique est augmentee par rapport a la biomasse originale. Cette
augmentation de la valeur calorifique resulte de la perte de volatiles; augmentation au cours de
laquelle le contenu en oxygene est diminue et la concentration en carbone augmentee. Par
contre, dans une approche different de celle de la torrefaction, la production de charbon de
bois classique vise a confiner les volatiles en presence du charbon pour favoriser la reaction
entre carbone volatil et carbone fixe [Antal et Gronli, 2003]. Par cette approche, il en resulte
une restructuration de la biomasse initiale ainsi que des rendements eleves en charbons et
concentres en carbone. Tous ces types de charbons menent a une reduction du ratio O/C de la
biomasse originale. II fut demontre qu’une efficacite plus elevee pouvait etre atteinte par la
gazeification de matieres premieres dont le ratio O/C etait minimise. D’un point de vue
exergetique, un biocarburant ideal pour gazogene doit etre faiblement oxygene [Prins et al.,
2006a]. II fut aussi reconnu que le bois torrefie constituait un carburant solide plus propre pour
la combustion, produisant une fumee plus faible en composes acides. Cependant, ce dernier se
consumerait plus rapidement que le charbon de bois [Prins et al., 2006a; Prins et al., 2006c].
En comparaison avec la biomasse, le charbon de biomasse torrefiee a une valeur calorifique
plus elevee, un contenu moindre en composes volatils et en eau, ce qui en fait un bon candidat
pour la combustion et la gazeification. La gazeification de charbon de biomasse torrefiee tend
aussi a generer moins de goudrons. Cela est du au faible contenu en composes volatils du
biocharbon comparativement a sa biomasse originale. Lors de la torrefaction, malgre le fait
qu’une partie de l’energie de la biomasse soit perdue lors de la liberation des composes
volatils (10-20% de la biomasse originale), ces composes chimiques peuvent etre captures et
valorises [Phanphanich et Mani, 2011],

20
Un avantage additionnel que le procede de torrefaction (pyrolyse douce) confere au charbon
de biomasse, constitue en sa perte de proprietes mecaniques. Ceci serait du au bris des liens
glucosidiques entre les monomeres de glucose et les liens hydrogene inter et intramoleculaires
[Rousset et al., 2011]. Plus la temperature est elevee lors de la torrefaction, plus les proprietes
mecaniques de la biomasse sont reduites. Ainsi, il est estime que l’energie necessaire a reduire
en particules et pour granuler la biomasse est diminuee de moitie en comparaison avec les
memes etapes pour la biomasse brute. Le charbon de torrefaction est done plus facilement
pulverisable, pouvant etre benefique pour la production de granules ou pour des procedes
(ex. : gazeification) ou une particule de petite taille peut etre souhaitee [Boateng, 2007;
Rousset et al., 2011], De plus, le procede liberant l’eau libre et structurale permet au charbon
de biomasse de se conserver plus longtemps sans se biodegrader. Le procede de torrefaction
reorganise les parois cellulaires de la biomasse et diminue ainsi le nombre de liens OH-H, ce
qui lui confere un caractere hydrophobe, l’empechant de regagner l’humidite tel que le ferait
une biomasse brute [Kocaefe et al., 2007; Repellin et al., 2010], Le caractere non
hygroscopique de la biomasse torrefiee est aussi un avantage au niveau de la stabilite du
materiel [Kocaefe et al., 2007]. Des granules de biomasse torrefiee evitent de regagner
l’humidite et de gonfler telles le feraient des granules de biomasse brute ayant ete sechees.
Felfli et al. [2005] effectuerent des travaux sur la torrefaction de briquettes de bois et leurs
resultats montrent une reduction de 73 % du degre d’humidite d ’equilibre. Le caractere non
hygroscopique de la biomasse torrefiee se repercute aussi au niveau de la valeur calorifique et
de l’usage de la biomasse torrefiee qui demeure stable malgre les variations d’humidite
relative. Une meilleure stabilite de la granule peut etre benefique tant pour les systemes
d ’alimentation des reacteurs que pour la reaction en reacteur con 9 u pour une particule de
dimension specifique.

2.2.3 Autres produits carbones

Plusieurs facteurs tels que le type de matiere organique, le traitement thermochimique et les
conditions dans lesquelles ces traitements thermochimiques operent conferent au biocharbon
une variability au niveau de ses proprietes [Downie et al., 2009]. Les proprietes affectees par
la transformation thermochimique sont principalement : la surface specifique, la porosite, la
structure du pore (de taille micro a macrometrique), la composition chimique du biocharbon,

21
Fanalyse elementaire et le contenu en cendres [Ioannidou et Zabaniotou, 2007]. Certaines de
ces caracteristiques et les facteurs les influen 9 ant sont traites dans la section 2.3. D ’autres
types de produits carbones peuvent aussi etre obtenus lors de procedes thermochimiques, mais
dont les caracteristiques, les applications et la nomenclature different du biochar et du charbon
vert. Certains de ces produits carbones sont definis dans la section suivante afin de mieux les
distinguer.

Carbonisat
Le carbonisat est un produit de carbonisation obtenu a partir de matiere organique. La
particularity du carbonisat comparativement au biocharbon est qu’il n ’est pas passe par un etat
fluide. Ainsi, le carbonisat ne serait pas le resultat de la cokefaction ou du cokage ou de
condensation de composes volatils [GFC, 2004], Dans le cas de la presente recherche, le
charbon de biomasse peut passer par un etat fluide, ainsi, l’utilisation du terme charbon de
biomasse ou biocharbon plutot que carbonisat a ete maintenue.

Char
Terme anglophone englobant divers types de residus carbones (charbon, carbonisat et noir de
carbone) [Bureau de la traduction, 2012].

Noir de carbone
Correspondant au carbon black en langue anglaise, le noir de carbone est un materiau carbone
obtenu des industries thermiques a des conditions de production controlees. Ce produit est
valorise notamment comme pigment dans la production de pneus automobiles ou de tableau
noir. Ce materiau est generalement de forme spherique ou d ’agregat de spheres, de dimension
de 10 a 1000 nm [Druide, 2011; GFC, 2004],

Suie
Les particules de suie consistent en des feuillets de cycles de benzene resultants de
combinaisons radicalaires formees en milieu quasi ou totalement anaerobique [Olivier, 2007].
Suie et noir de carbone sont souvent pris l’un pour l’autre. La suie serait obtenue dans des
conditions non controlees, contrairement au noir de carbone obtenu en conditions controlees.

22
De plus, la suie comporte un taux de goudron, de cendres et d’impuretes qui differe du noir de
carbone [GFC, 2004],

Carbone noir
Le carbone noir (anglais : black carbon) represente habituellement un eventail de produits
issus de transformations thermochimiques tels que des matieres partiellement carbonisees, des
particules de suie et du graphite. Ce terme est generalement utilise en climatologie en
reference a toutes ces structures carbonees similaires qui represented des particules absorbant
l’intensite lumineuse ou foumissant de l’information sur l’histoire et des evenements s’etant
deroules au fil du temps [GRID-Arendal, 2001; Krull et al., 2009; Schmidt et Noack, 2000].
La structure du carbone noir est generalement aromatique et formee de couches graphitiques
empilees aleatoirement [Krull et al., 2009].

Graphite
Le graphite est une structure de carbone dont 1’organisation des elements a mene a une
structure cristalline consistant uniquement en des cycles aromatiques de carbone organises en
feuillets [Larousse, 2005]. Dans les calculs thermodynamiques faisant intervenir le carbone
avec d’autres elements, le carbone refere habituellement au graphite.

Charbon de bois
Traditionnellement, le charbon de bois (charcoal en anglais) refere au residu de carbone
obtenu a la suite d'une conversion du bois en conditions de pyrolyse [Borman et Ragland,
1998]. Le terme charbon correspond a un produit riche en carbone obtenu a la suite d'une
action thermique sur la matiere organique ayant vaporise la partie volatile. Les charbons
produits specialement pour fins energetiques sont produits a haute temperature visant a reduire
le contenu en oxygene. La pression favorise aussi la formation du charbon de bois. La pression
peut etre augmentee en for9 ant les composes volatils a rester dans l’environnement reactif du
charbon, ce qui permet a ces composes de reagir et de se transformer avec le contenu meme du
charbon de bois et ainsi augmenter le rendement de ce dernier [Antal et Gronli, 2003]. Des
etudes d ’extraction des composes aromatiques polycycliques de ces charbons (ex. : briquettes

23
de charbons de bois) ont demontre la presence d’une grande quantite d’HAP dans ces types de
biocharbons [Kushwaha et al., 1985].

Charbon mineral
Le terme « charbon » refere habituellement au secteur du charbon mineral (coal en anglais)
qui est obtenu par extraction miniere. II est important de noter que « coal ch a r» et
« charcoal» (termes anglophones) sont d ’origine minerale et de bois, respectivement. Le
charbon mineral regroupe 3 categories (la tourbe, le lignite et la houille), mais il fait
habituellement davantage reference a la houille [Foucault et Raoult, 1995]. Plusieurs
subdivisions se retrouvent dans la houille (ex. : l’anthracite). La houille est aussi utilisee pour
la fabrication du coke. Des initiatives sont prises afin de melanger les charbons mineraux et le
charbon vert afin de reduire les emissions de GES et de soufre [Laird et al., 2009].

Coke
Le coke consiste en une matiere carbonee habituellement obtenue par la cokefaction de la
houille [Larousse, 2010]. Le coke peut aussi correspondre a des depots carbones sur les
surfaces de l’environnement dans lequel il se produit. Cette formation de depots correspond
cependant au cokage a partir de diverses matieres carbonees. Bien que cokage et cokefaction
resultent du meme processus chimique, les deux termes sont appliques selon leur matiere
premiere respective (c.-a-d. : cokefaction pour la transformation specifique a la houille)
[Comite d'etude des termes techniques franfais, 2001],

2.3 Conversion de la biomasse en biocharbon par pyrolyse

2.3.1 Pyrolyse

La pyrolyse consiste au bris de liens chimiques d’une matiere premiere sous l’effet de la
chaleur. Le procede de pyrolyse de biomasse consiste a chauffer la matiere premiere sans
oxygene, la transformant ainsi en 3 categories de produits : un biocharbon, des vapeurs
condensables et des gaz non condensables. Les vapeurs condensables forment une solution
aqueuse (bio-huile), une microemulsion formee de produits de decomposition d’holocellulose

24
et de lignine [Mohan et al., 2006]. Le procede de pyrolyse est habituellement considere
comme endothermique et s’effectue en continu grace a un apport d ’energie exteme; un flux de
chaleur de 50 W/cm2 serait necessaire pour la pyrolyse rapide [Mohan et al., 2006].
Cependant, il fut observe qu’au-dela de 300 °C, Toxygene libere avec les composes volatils
tend a oxyder certains des constituants, pouvant mener a des plages de temperatures ou le
bilan des reactions serait exothermique [Bourgois et al., 1989; Demirbas, 2000; Mohan et al.,
2006; Rousset et al., 2011; White et al., 2011]. La technologie de pyrolyse s’opere selon deux
grands regimes, la pyrolyse lente et la pyrolyse rapide, selon le taux de chauffe de la matiere
premiere et le temps de residence en reacteur (Tableau 2.3). En variant les parametres
d’operation du reacteur de pyrolyse tels que la temperature, la pression, le taux de chauffe et le
temps de residence de la biomasse en reacteur, differentes fractions et compositions de ces
produits sont alors obtenues. A basse temperature (200 a 300 °C), a une pression de 101.3 kPa
(generalement) et dans une atmosphere inerte ou faible en O 2 , le procede refere a la
torrefaction (pyrolyse douce; Tableau 2.3) [Prins et al., 2006c; Repellin et al., 2010; Rousset
et al., 2011].

Lors de la pyrolyse, les changements s’effectuant dans la biomasse sont les suivants [Mohan et
al., 2006] :
1- La chaleur est transmise a la biomasse par une source de chaleur exteme,
augmentant ainsi sa temperature.
2- Les premieres reactions de pyrolyse se deroulent a la suite de la hausse de
temperature. Lors de ces reactions, des composes volatils (fractions de mono et
polysaccharides de biomasse [Prins et al., 2006c]) sont relaches et les biocharbons
sont formes.
3- Les composes volatils peuvent entrer en contact avec des zones du charbon de
biomasse non pyrolysees dont la temperature est plus basse que celle a laquelle ces
composes volatils sont produits. Un transfert de chaleur peut s’y effectuer,
provoquant ainsi la condensation des composes volatils a Tinterieur meme de la
particule au lieu qu'ils soient expulses de la structure carbonee. Les composes
volatils produits a partir de la decomposition des constituants de la biomasse

25
 peuvent subir des reactions secondaires entre eux et avec le biocharbon. Les
reactions des etapes 2 et 3 se produisent simultanement.
4- D ’autres modifications telles des decompositions thermiques, du reformage, des
recombinaisons de radicaux et de multiples reactions peuvent aussi survenir selon
les conditions d’operation du procede.

Le charbon de bois voue a une application thermique, ainsi que l’usage de charbon de
biomasse comme charbon active, ont tous deux grandement contribue a l’avancement de la
caracterisation du produit solide de pyrolyse [Antal et Gronli, 2003; Ioannidou et Zabaniotou,
2007]. L ’essor majeur qu’a connu la pyrolyse de la biomasse depuis les demieres decennies
est grandement lie a la maitrise du procede favorisant la fraction liquide, la bio-huile. Cette
bio-huile contient une multitude de produits, estimee a plus de 300 composes [Czemik et
Bridgwater, 2004], Ces composes peuvent etre valorises en biocarburants (necessitant des
etapes subsequentes) ou autres produits a valeur ajoutee tels des extraits dedies au domaine
alimentaire (ex. : saveurs barbecue). Les bio-huiles sont tres oxygenees. Si elles ne sont pas
traitees, il en resulte des biocarburants au contenu energetique moindre, une acidite elevee
comparativement aux carburants petroliers, en plus d’etre immiscibles avec ces demiers
[Imam et Capareda, 2012]. Depuis les demieres annees, le domaine de la pyrolyse se penche
sur les transformations subsequentes de la bio-huile et, entre autres, sur la desoxygenation
pour la production de biocarburants, tels les biocarburants d ’aviation [Goodfellow, 2009], Les
conditions specifiques d’operation des reacteurs de pyrolyse visant l’obtention de biocharbons
aux caracteristiques souhaitables pour la valorisation comme biochar et\ou comme charbon
vert constituent des conditions d ’operation pouvant differer de celles deja raportee dans le
domaine scientifique. Plus de details quant aux conditions d ’operation menant a des
caracteristiques specifiques sont abordes dans les sections suivantes.

26
Tableau 2.3 : Sommaire des precedes thermochimiques de production de charbon de biomasse
Temps de Bilan de
Proclde Taux de
Environnement Pression Temperature residence en biocharbon References
thermochimique chauffe
reacteur obtenu
°C °C min'1 %
Carbonisation aqueux 0 ,5 - 1 8 0 -3 5 0 min a 1 jour variable >30 [Amonette et Joseph, 2009;
hydrothermale 2 MPa Heilmann et al., 2011;
Hoekman et al., 2011 ]
Torrefaction anoxique ou 1 atm 2 0 0 -3 0 0 min a <50 >40 [Amonette et Joseph, 2009;
faible en 0 2 heures Prins et al., 2006b]
Pyrolyse lente anoxique ou 1 atm -5 0 0 min a -5 -3 0 -3 5 [Brown, 2009; Downie et al.,
faible en 0 2 jours 2009]
Pyrolyse rapide anoxique variable -5 0 0 sec-min +++ 20-40 [Boateng, 2007; Brown, 2009]
Gazeification Appauvri en 0 2 variable 550-750 sec-min 10 12 [Brown, 2009]

Combustion Riche en 0 2 1 atm >1000 variable ++ minimal

27
2.3.2 Effet de la temperature et de la composition de la biomasse
sur la pyrolyse

La temperature maximale atteinte lors du procede de pyrolyse est le facteur ayant le plus
d’effet sur les proprietes du charbon de biomasse. Plus la temperature est elevee, plus la
productivity en biocharbon diminue [Ioannidou et Zabaniotou, 2007; Sharma et al., 2002], La
cinetique de la reaction globale de pyrolyse suit la fonction d ’Arrhenius, indiquant l’intime
relation de la temperature sur le procede de pyrolyse et, consequemment, sur les proprietes des
biocharbons obtenus [Antal et Gronli, 2003]. La diminution du rendement en biocharbon en
fonction de la temperature resulte d ’une decomposition dite « primaire », formant et liberant
des composes volatils. Sous 400-450 °C, la majorite des produits volatils sont formes et
liberes; conditions ou la production de charbon de biomasse peut decroitre drastiquement
[Butterman et Castaldi, 2010], II s ’en suit une decomposition dite « secondaire»,
correspondant aux transformations subsequentes des composes volatils et du biocharbon. Les
reactions secondaires menent habituellement a une production de gaz non condensables.

Etudiees independamment, la pyrolyse des composantes de la biomasse (hemicellulose,

cellulose, lignine, matieres extractibles) resulte en des produits pouvant differer de ceux
obtenus lorsque ces composantes sont pyrolysees ensemble, a l’interieur d ’une meme
biomasse [Butterman et Castaldi, 2010]. Chaque composante de la biomasse a un
comportement par rapport a la temperature qui lui est propre (Figure 2.7). Les composantes
des biomasses pouvant varier selon plusieurs facteurs (section 2.1.2), leur influence sur les
proprietes et les rendements en biocharbons est majeure [Antal et Gronli, 2003; Krull et al.,
2009]. Dans les sections qui suivent, 1’effet de la temperature est premierement aborde par
rapport a la conversion de l’ensemble des composantes, suivi par son effet sur la conversion
des composantes analyse individuellement.

28
 2.5
100 H em icellulose
C e ll u lo s e
LiQnin
80

60

|
ft
40

20

0 1
---------t'.-i 0 Q

200 400 600 800

T em p eratu re(°C )

Figure 2.7 : Analyses thermogravimetriques des composantes principales de la

biomasse. Source : Yang et al. [2006]

Effet de la temperature sur la pyrolyse de I’ensemble des composantes de la

biomasse lignocellulosique
Lors de la periode de prechauffage de la biomasse (pour pyrolyse a pression atmospherique),
la biomasse passe tout d’abord par une phase d ’extraction de l’eau a des temperatures
avoisinant 100 °C. Cette eau est presente sous 2 formes dans la biomasse : l’eau libre et l’eau
liee. Une deuxieme phase a lieu generalement entre 100 et 250 °C et modifie les constituants
contenant des groupes hydroxyles accessibles; cette phase cause des degradations irreversibles
dans la biomasse [Almeida et al., 2010]. A ces temperatures, des groupes carboxyles et
carbonyles sont generes et menent a la formation de CO et CO2 . En effet, lors de la
torrefaction (jusqu’a 300 °C), la decarboxylation compte parmi les reactions importantes se
produisant avec la deshydratation. Ces deux reactions reduisent la masse de la biomasse
originale et s’accompagnent d’une augmentation de la valeur calorifique suite a une
carbonisation et une reduction du contenu en eau. Lors de la decarboxylation, la formation
d ’acides est favorisee. Ces acides, a leur tour, catalyseraient la deshydratation. Les produits de
pyrolyse douce (torrefaction) obtenus so n t: un biocharbon, un liquide acide jaunatre
(constitue d’eau, d ’acide formique, d ’acide acetique, de furfural, etc.) et des gaz non
condensables (tels que le monoxyde de carbone et le dioxyde de carbone) [Prins et al., 2006b;
Prins et al., 2006c]. Le CO ne decoulerait pas de la decarboxylation ou de la deshydratation. II
fiat observe que la formation de CO etait plus elevee en presence de certains types de mineraux
[Prins et al., 2006c], Selon Prins et al. [2006b], la cinetique de reaction de la torrefaction pour
une plage de 230 a 300 °C pourrait se modeliser par un mecanisme en deux etapes. Une
premiere etape consiste en la decomposition rapide de l’hemicellulose et, une deuxieme, en la
decomposition de la cellulose. De plus, le bris des liens, entre 120 et 300 °C, favorise la
formation de radicaux libres et ces demiers favoriseraient la degradation de la biomasse a ces
temperatures. La formation des radicaux libres serait favorisee par : la chaleur, la presence de
matieres inorganiques, l’oxygene present dans la structure de la biomasse originale, ainsi que
par l’oxygene present en faibles quantites lors de l’insertion de la biomasse dans le reacteur.
Certains de ces radicaux seraient recombines et contribueraient ainsi en la formation de la
structure du biocharbon en tant que tel [Amonette et Joseph, 2009].

La hausse de la temperature au-dela de 375 °C favorise la production d’aromatiques,

d ’alcanes, de derives de benzene et de composes oxygenes [Butterman et Castaldi, 2010].
L’observation de composes aromatiques a ces temperatures est en accord avec la composition
de la lignine, sa decomposition s’effectuant davantage a ces temperatures. Generalement, a des
temperatures sous 400 °C, le taux auquel la carbonisation s’effectue est tres lent [Antal et
Gronli, 2003]. De 300 a 600 °C, la formation de liquides prend de l ’ampleur, reduisant ainsi la
production de biocharbon. Les vapeurs condensables produites a partir de la cellulose sont
composees principalement de levoglucosane. A des temperatures sous 500 °C, une partie des
vapeurs liberees condensent alors dans les pores du biocharbon et il en resulte une substance a
multiphases (phase solide et phase liquide-goudronnee) [Amonette et Joseph, 2009], Au-dela
de 550 °C, la formation de sucres anhydres peut s’effectuer a partir desquels peuvent se former
les cycles de furanes [Zhang et al., 2011]. Au-dela de 600 °C, la formation de gaz non
condensables est davantage favorisee, minimisant ainsi la formation de goudron et de
biocharbon. D ’autre part, a ces temperatures, la carbonisation du charbon de biomasse s’en
trouve favorisee, alors que le contenu en O, H, N et S diminue. II fut observe que la
concentration en radicaux libres augmentait avec l’augmentation de la temperature, jusqu’a
environ 500 °C. Ces radicaux libres sont tres reactifs avec l’Ch et c’est ce qui expliquerait la
tendance a l’ignition spontanee du biocharbon. Le contenu en radicaux tendrait cependant a
diminuer a des temperatures superieures [Amonette et Joseph, 2009; Bourke et al., 2007],

30
En general, l’augmentation de la temperature de pyrolyse mene a un biocharbon dont le
contenu en carbone aromatique est plus eleve [Baldock et Smemik, 2002; Sharma et al.,
2002], La biomasse brute contient un ratio H/C (xlO) et O/C d ’environ 1,5 et 0,7,
respectivement. Suivant une augmentation de la temperature, les ratios H/C (xlO) et O/C
passent generalement de 0,45-0,85 (H/C) et 0,15-0,45 (O/C) pour 300 °C a 0,3-0,55 (H/C) et
0,07-0,4 (O/C) pour 500 °C. Les ratios de H/C et O/C peuvent etre utilises afin d’estimer le
degre d ’aromaticite ou une diminution du ratio correspond a une augmentation de
l’aromaticite [Krull et al., 2009]. La diminution des ratios O/C et H/C a partir de 250 °C peut
aussi etre associee a la formation de structures carbonees non saturees [Baldock et Smemik,
2002], De plus, les ratios O/C et H/C peuvent etre utilises pour estimer la valeur calorifique du
produit, ou la valeur calorifique augmente au fur et a mesure que le ratio O/C de la biomasse
diminue. Cette correlation est generalement exprimee selon le diagramme de Van Krevelen
(Figure 2.8). Lors de la pyrolyse, la liberation des composes volatils permet d ’expulser
l’oxygene. II en resulte une reduction du ratio O/C et une augmentation de la valeur
calorifique du charbon de biomasse.

1.8

1.6

1.4

1.0
0.8
0.6
0.4
Increased Heating Value
0.2

0.0
0.0 0.2 0.4 0.6 0.8
Atomic OiC Ratio
Biomass
Peat
3 Lignite
tSSfiSS Coal
KSSftSSfl Anthracite
■ Wood
A Lignin

Figure 2.8 : Diagramme Van Krevelen de diverses matieres solides.

Source : McKendry [2002]

31
Effet de la temperature sur la pyrolyse des hemicelluloses
Les hemicelluloses forment la composante la plus reactive de la biomasse [Prins et al., 2006c;
Rousset et al., 2011]. Ainsi, les hemicelluloses sont facilement depolymerisees entre 200 et
230 °C, mais leur decomposition se complete entre 230 et 300 °C [Amonette et Joseph, 2009;
Prins et al., 2006b; Repellin et al., 2010; Rousset et al., 2011], Lors de la premiere etape de la
degradation des hemicelluloses, les composes volatils se forment (portions de mono et
polysaccharides). Ces demiers peuvent subir une degradation catalytique par les mineraux de
la biomasse, resultant ainsi en la formation de CO et CO2 [Prins et al., 2006c], C’est la
degradation des hemicelluloses qui reduirait le caractere hygroscopique du biocharbon de
torrefaction en limitant la readsorption de l’eau [Kocaefe et al., 2007]. Parmi les polymeres
formant les hemicelluloses, les xylanes sont les polymeres les plus reactifs et sont tres
susceptibles de se decomposer a des temperatures entre 200 et 260 °C [Prins et al., 2006c;
Repellin et al., 2010]. Les xylanes sont la source principale de la portion volatile de la
biomasse (principalement les furanes et les acetaldehydes). Les xylanes contribuent aussi a
amorcer et propager les reactions de pyrolyse favorisant la liberation de substances volatiles
condensables et non condensables (CO 2 , CO et des traces de H2 et CH4) [Prins et al., 2006c;
Rousset et al., 2011]. Parmi les essences de bois, les coniferes contiennent generalement
moins de xylanes et davantage de glucomannanes. Ainsi, comparativement aux feuillus, le
rendement en biocharbon se retrouve done souvent plus eleve en conditions de torrefaction
chez les coniferes. Prins et al. [2006c] observerent que les feuillus (hetre et saule) et la paille,
contenant davantage de xylanes et de mineraux, produisaient plus de methanol et d’acide
acetique que les coniferes (meleze). Tout comme la cellulose, l’hemicellulose produirait aussi
des composes contenant le cycle de furane, des chaines droites avec groupements fonctionnels
C =0 et des sucres [Zhang et al., 2011]. La decomposition des hemicelluloses forme une
portion importante de l’acide acetique produit, et ce, a la suite d ’une reaction de deacetylation
[Boateng et al., 2007; Mohan et al., 2006].

Effet de la temperature sur la pyrolyse de la cellulose

La cellulose resiste davantage a la degradation thermique en comparaison aux hemicelluloses.
Un comportement qui est du principalement a la cristallinite de la cellulose alors que les
regions amorphes de la cellulose seraient plus susceptibles a la degradation thermique

32
[Rousset et al., 2011]. La cellulose peut etre depolymerisee et deshydratee lors de la
torrefaction a plus haute temperature [Repellin et a l, 2010], Prins et al. [2006b] rapportent
que la depolymerisation de la cellulose a moins de 250 °C peut se produire, mais que la
volatilisation s’effectue lentement. Des liens glucosidiques entre les monomeres de glucides et
les liens hydrogene inter et intramoleculaires seraient susceptibles de se rompre lors de la
torrefaction et meneraient a une diminution des proprietes mecaniques de la biomasse
originale [Rousset et al., 2011]. Entre 250 et 350 °C, la depolymerisation complete de la
cellulose serait atteinte, liberant ainsi beaucoup de substances volatiles et laissant une matrice
de carbone amorphe [Amonette et Joseph, 2009], Butterman et Castaldi [2010], quant a eux,
avancent que la decomposition du polysaccharide de cellulose s’effectue entre 325 et 425 °C.
Certains auteurs rapportent que la cellulose pourrait se transformer en des polymeres de
furanes et de phenols a partir de 270 °C. La pyrolyse de la cellulose a basses temperatures
(<500 °C) produirait aussi des chaines aliphatiques (avec ou sans groupements fonctionnels
C=0) et des sucres [Sharma et al., 2002; Zhang et al., 2011].

En fonction de la temperature, la cellulose se degrade en 2 etapes reactives. A basse

temperature, la degradation graduelle de la cellulose favorise une production elevee en
biocharbon. Cependant, a des temperatures plus elevees, la formation de composes volatils est
favorisee, les reactions s’effectuant beaucoup plus rapidement et menant a la formation de
levoglucosane. Lors de la pyrolyse de la cellulose, le levoglucosane est obtenu en grandes
quantites, tous deux ayant la meme formule elementaire C6Hi0O 5 [Demirbas, 2000]. Dans ces
conditions, les chaines de cellulose sont brisees en monomeres de glucose et alors le glucosane
est obtenu a la suite de l’enlevement d ’une molecule d ’eau. Les premieres reactions de
degradation de la cellulose comportent la depolymerisation, l’hydrolyse, l’oxydation, la
deshydratation et la decarboxylation [Demirbas, 2000], Lors de la pyrolyse, la cellulose
produit principalement du dioxyde de carbone, de l’eau et du methane. Ces gaz diminuent en
fonction de la temperature au profit, entre autres, du monoxyde de carbone [Antal et Gronli,
2003]. Lors de la pyrolyse de la cellulose, des radicaux »OH sont aussi relargues de la
structure et peuvent ainsi reagir avec le CO, le methane ou autres hydrocarbures ou radicaux
libres. La reaction de «OH avec le CO produirait ainsi du CO 2 et »H; ce dernier qui peut reagir
avec des radicaux libres, dont le radical »CH3, et former du CH4 [Butterman et Castaldi, 2010].

33
Effet de la temperature sur la pyrolyse de la lignine
La lignine, comparativement a l’hemicellulose et a la cellulose, est plus resistante a la
degradation lors d’un traitement thermique [Rousset et al., 2011], C’est la lignine qui forme
generalement la quantite la plus considerable de biocharbon parmi les composantes de la
biomasse [Butterman et Castaldi, 2010], La degradation de la lignine s’effectue en 2 etapes, la
premiere a partir de 250 °C, et la deuxieme a des temperatures jusqu’a 625 °C [Butterman et
Castaldi, 2010; Prins et al., 2006c; Sharma et al., 2004]. A 300 °C, seulement 20-25 % de la
lignine serait convertie en produits volatils (Figure 2.7) [Sharma et al., 2004; Yang et al.,
2006], Lors de sa degradation, la lignine perd plusieurs liaisons P-O-4 (Figure 2.4), mais de
nouveaux liens plus resistants seraient reformes [Almeida et al., 2010]. En general, la
comprehension des mecanismes expliquant la reactivite de la lignine en fonction des
traitements thermiques serait moindre que celle de la cellulose et de l’hemicellulose, celles-ci
ayant ete etudiees davantage [Rousset et al., 2009], La pyrolyse de la lignine est reconnue
comme favorisant la formation de composes phenoliques a partir de son unite de
phenylpropane (Figure 2.4), tels que le guaiacol et le syringol [Amen-Chen et al., 2001;
Butterman et Castaldi, 2010; Zhang et al., 2011], Tel que rapporte par Butterman et Castaldi
[2010], entre 250 et 400 °C, la decomposition de la lignine formerait et libererait des radicaux
hydroxyles, methyles et methoxy. A partir de 400 °C, sa decomposition favoriserait la
formation de monoxyde de carbone lors de reactions de carbonisation [Butterman et Castaldi,
2010],

Effet combine de la temperature et des elements inorganiques sur la pyrolyse

de la biomasse
Les types et la quantite d’elements inorganiques dans la biomasse varient selon le type de
biomasse et peuvent aussi varier a l’interieur d’une meme espece de biomasse (voir section
2.1.2). Certains de ces elements ont un effet catalytique sur les reactions de pyrolyse des
constituants de la biomasse [Butterman et Castaldi, 2010; Mohan et al., 2006]. Les biomasses
contenant davantage d’elements inorganiques ont d ’ailleurs tendance a se transformer a des
temperatures plus basses (Figure 2.9 et Figure 2.10) [Butterman et Castaldi, 2010; Fahmi et
al., 2007], Une plus grande conversion de la biomasse originale ainsi qu’une diminution du
rendement en biocharbon sont generalement observees en presence de mineraux [Prins et a!.,
2006b]. Dans certains cas, 1’effet est plutot contraire. L’enlevement partiel de sodium et de

34
potassium par lessivage de la biomasse, bien qu’il augmente la temperature de transformation
(Figure 2.9 et Figure 2.10), aurait toutefois reduit le rendement en biocharbon de biomasse
ainsi que la formation d ’HAP [Sharma et Hajaligol, 2003; Fahmi et al., 2007]. II serait
cependant important de verifier si le pretraitement de lessivage, qui consiste aussi en une
technique d ’extraction de produits de la biomasse [Sluiter et a l, 2008], ne serait pas en cause
dans les variations observees plutot que la diminution en elements inorganiques suite au
lessivage.

Butterman et Castaldi [2010] ont calcule des valeurs d’energie d ’activation suite a des tests
effectues en atmospheres de H20/N 2, C 0 2, C 0 2/N2/H20 , et 0 2/N2. Pour tous les tests, les
valeurs d’energie d’activation etaient toujours plus basses dans le cas des plantes herbacees
que pour le bois. Un resultat qu’ils attribuent a la presence de mineraux plus elevee dans les
plantes herbacees. Le MgO et le CaO seraient deux mineraux participant a la reactivite et a la
fragilite du biocharbon [Butterman et Castaldi, 2010]. L ’effet marque des elements
inorganiques lors des transformations thermochimiques de la biomasse pourrait en quelque
sorte mener a une reaction telle celle presentee par Jenkins et al. [1998] :
CxiHx2Ox3NX4SX5ClX6Six7KX8CaX9MgxioNaxiiPxi2Fexi3Alxi4Tixi5+niH20+n2(l+e)(02+3,76N2)=
n 3 C 0 2+ n 4 H 20 + n 5 0 2+ n 6 N 2+ n 7 C O + n 8 C H 4 + n 9 N O + m o N 0 2+ n n S 0 2+ n i 2H C l+ n i3 K C l+ n i4 K 2S 0 4 +

ni5C+... .

' ' W ashed F estuca malrei-4x

400 450 500 550 600 660 TOO 750 800

Temperature (K)

Figure 2.9 : Analyses thermiques differentielles de Festuca mairei lavee vs non

lavee. Les echantillons non laves (unwashed) contiennent plus
d’elements inorganiques. Source : Fahmi et al. [2007]

35
 650
640 - ! * Unwashed
|
= ■£ 630 ■ Washed

C. 610

580

570
0 5000 10000 15000 20000 25000 30000 35000 40000
Na+K Content (ppm)
Figure 2.10: Temperature de degradation de la cellulose selon le contenu en
sodium et potassium. Source : Fahmi et al. [2007]

Effet de la temperature sur la structure du biocharbon

La structure du biocharbon peut varier grandement. Une structure poreuse ou de surface
specifique elevee est souhaitable, et ce, autant pour le biochar que le charbon vert. Pour le
biochar, elle signifie une structure pour Fetablissement de microorganismes ou des sites
d ’adsorption potentiels. Dans le cas du charbon vert, une surface specifique elevee favorise
une plus grande reactivite pour les procedes subsequents. Une augmentation de la surface
specifique est generalement observee a partir de temperatures de pyrolyse relativement
elevees, entre 500 a 800 °C (Figure 2.11) [Biagini et al., 2009; McLaughlin et al., 2009],
Generalement, la porosite des charbons de biomasse augmente a la suite de la liberation de la
portion des composes volatils, pouvant correspondre a des goudrons s’etant loges a l’interieur
des pores par condensation [Jindarom et al., 2007a; Lehmann et Joseph, 2009]. La liberation
des composes volatils (du goudron) reduit cependant la viscoelasticite et la solidification du
biocharbon. La carbonisation complete qui s’en suit favorise la production de gaz non
condensables et resulte en un biocharbon plus fragile a l’effritement [Sharma et a l, 2002],

Lors de la pyrolyse, selon la temperature maximale atteinte, plus la temperature est elevee,
plus les structures de carbone varient et tendent a s’organiser en feuillets. A des temperatures
en dessous de 250 °C, les premiers changements de la biomasse s’en tiennent a une legere

36
depolymerisation de la cellulose et une deshydratation. Entre 250 et 350 °C, la
depolymerisation complete de la cellulose se produit et la cellulose perd sa cristallinite [Paris
et al., 2005], Cela se traduit par une perte significative en composes volatils, y laissant une
matrice de carbone amorphe. La presence de carbone aromatique apparait a 330 °C et, au-dela
de 350 °C, les feuillets individuels de graphene se developpent. A ces temperatures (300-
350 °C), le charbon de biomasse a des allures plus semblables a la biomasse initiale. Le
passage a des temperatures au-dela de 625 °C provoque la formation de structures presque
totalement aromatiques [Sharma et Hajaligol, 2003]. De plus, le processus de carbonisation est
defini comme debutant a des temperatures de 600 °C. A ces temperatures, les atomes, autres
que le carbone, seraient liberes sous forme de substances volatiles et les feuilles de graphene
auraient tendance a se former et croitre lateralement [Amonette et Joseph, 2009]. Par la suite
(T ~ 800 °C), le biocharbon tend a former des feuillets de structures carbonees aromatiques
conjuguees, organisees de maniere dite turbostratique (organisees en plan, mais des plans
n ’ayant pas d ’ordre entres eux) (Figure 2.12 et Figure 2.13). A une temperature plus elevee,
l’organisation s’ameliore au niveau de l’ordre entre les plans (3e dimension) menant a la
formation du graphite (T > 2000 °C; Figure 2.13) [Kumar et al., 1993],

■150 -
* «r

'a *- * A' WP ’ *1 r - ' -J

-j 350 -
■ >->■i - —i fa s - »
V jW ■. »y * ^ *'» -iyr* \ ■J »?■ t ■p.*'* -v J
‘ - I- ^ '■ - . - 1 ' . . \ ^ ■

I 200 •
•/
r 150 •

£100 ■
- V 1' I V .

0 ■
300 100 500 600 700 SOO 900 1000
T e r mi n a l p y r o l y s i s t e m p e r a t u r e ( O

Figure 2.11 : Developpement de la surface specifique du biocharbon de bambou

en fonction de la temperature. Source : McLaughlin et al. [2009]

37
Recemment, Keiluweit et al. [2010] etudierent la dynamique de la production de biocharbon et
de ses caracteristiques en fonction de la temperature (100 a 700 °C), et ce, a partir d ’une
essence de bois et d’une plante graminee (Pin, Pinus ponderosa et fetuque elevee, Festuca
arundinacea, respectivement). Pour les plages de temperatures allant de 100 a 700 °C, ils ont
defmi 4 categories de biocharbons (Figure 2.12) : (i) les biocharbons de transition, ou la
cristallinite de la cellulose diminue avec la temperature; (ii) les biocharbons amorphes; (iii) le
biocharbon composite constitue de graphenes desordonnes et dont la formation de pores est
favorisee; (iv) les biocharbons turbostratiques domines par des cristallites desordonnes et dont
les pores tendent a diminuer [Keiluweit et al., 2010],

unaltered plant
material

amorphous
lignin
pyrogenlc
amorphous carbon

crystalline
cellulose turfaostratic
crystallites

amorphous
hemlcelubse

■ash
a volatile matter
char yield ■non-volatile matter (fixed C)

Charring Intensity

Figure 2.12 : Evolution de la structure du charbon de biomasse en fonction de la

temperature. Source : Keiluweit et al. [2010]

38
 Figure 2.13: Modification de la structure du biocharbon en fonction de la
temperature (a) augmentation du carbone aromatique, desordonne,
dans une structure amorphe; (b) developpement de feuillets de
carbone aromatique, organisation turbostratique; (c) structure
graphitique. Source : Downie et al. [2009]

2.3.3 Effet de I’usage du C 0 2 lors de la pyrolyse

Plusieurs systemes de pyrolyse industriels fonctionnent a partir d ’un gaz inerte d ’azote ou
d ’argon, alors que d’autres n’utilisent pas de gaz (ex. : reacteurs a lits fluidises vs reacteurs par
ablation ou a vis, respectivement) [Bridgwater et Peacocke, 2000], Les gaz inertes sont
generalement utilises pour favoriser l’echange de chaleur (chauffer les particules de
biomasse), transporter les gaz produits et favoriser le transfert de masse (expulser et
transporter les vapeurs et gaz produits). A temperature ambiante, le CO2 est considere comme
etant non reactif avec le carbone, mais le devient a la suite de 1’augmentation de la
temperature (aG=0 a 704 °C, 1 atm) par la reaction inverse de Boudouard (C + CO 2 «-* 2 CO).
La reaction de Boudouard est reconnue comme pouvant etre catalysee en presence d ’elements
metalliques des groupes VIII (tel le nickel et le fer) ainsi que des sels alkalins et alkalino-
terreux (Jin et al. [2012]). Cette reaction est favorisee dans le procede de gazeification pour la

39
conversion du carbone en monoxyde de carbone (un constituant des gaz de synthese; CO et
H 2 ). La litterature aborde tres peu l’utilisation du CO 2 en conditions de pyrolyse. Or, c’est un
domaine qui beneficierait d’etre etudie davantage [Duan et a i, 2009; Jindarom et al., 2007a;
Naredi et Pisupati, 2011]. Malgre le fait que le CO2 soit habituellement inerte a basses
temperatures, les composantes lignocellulosiques de la biomasse ont des mecanismes reactifs
avec le CO2 qui leur sont propres et specialement lorsque la presence de mineraux entre en
consideration [Butterman et Castaldi, 2008], Des reactions avec la matiere premiere pourraient
se produire tant avec la partie solide qu’avec les composes volatils radicalaires, ou encore,
suivant une modification des proprietes physiques du CO2 [Naredi et Pisupati, 2011], Dans les
prochaines sections, des experiences ayant ete effectuees avec l’usage de CO 2 en pyrolyse et
des effets sur l’ensemble des produits de pyrolyse (dont le biocharbon) y sont presentes.

Effet du CO2 sur le rendement en biocharbon et autres co-produits de pyrolyse

Zhang et al. [2011] ont etudie l’effet du type de gaz lors de la pyrolyse rapide, en utilisant N 2 ,
CO 2 , CO, CH4 et H2 . Lors de tests en lit fluidise (lit de sable de silice, reacteur laboratoire de
30mm diametre X 400 mm hauteur), avec temps de residence de 10 minutes, a 550 °C et en
utilisant des particules seches (l- 2 mm) de rafles de mats, l’environnement de CO 2 en
comparaison avec l’azote a mene a plusieurs differences parmi les produits de pyrolyse. Une
baisse des gaz condensables et du biocharbon, accompagnee d ’une hausse des gaz non
condensables, ont ete observees en environnement de CO2 vs N 2 (Figure 2.14). En presence de
CO 2 , la composition de leurs gaz non condensables produits presente une baisse de la quantite
de CO2 produit, mais une stabilite au niveau du CO (Figure 2.14c). L ’utilisation du vecteur
gazeux de CO 2 , tout comme les autres environnements carbones (CO ou CH4 ), a resulte en un
lit de teinte plus foncee; une situation ne se presentant pas avec l’usage de N 2 . Les chercheurs
associent cette teinte plus foncee a un possible cokage, sur la base de l’analyse (par MEB-
EDX) de la surface de leur lit de silice qui presentait des teneurs elevees en carbone. Les
analyses GC-MS presentent des produits condensables contenant moins de composes methoxy
et davantage de phenol (monofonctionnel). L’environnement de CO 2 est decrit comme
oxydatif doux [Jindarom et al., 2007b; Zhang et a i, 2011],

40
 B T o tal liquid B O il fraction D W ater D Nofwondensahle ■ Char U Coke

15.0

Atm osphere — * n Atmosphere

W. 10

Atm osphere —► »\’7 )2

CC CO CH« H,

Figure 2.14 Bilans des produits de pyrolyse rapide de rafles de mai's obtenus
dans divers environnements gazeux. (a) produits condensables,
fraction bio-huile et fraction eau; (b) gaz non condensables,
biocharbon et coke; (c) composantes des gaz generes. Source :
Zhang et al. [2011]

Dans le cadre de recherches faites avec la technologie de combustion Oxy-fuel, qui est
consideree comme l’une des technologies les plus prometteuses pour la sequestration de
carbone, Duan et al. [2009] ont effectue des tests avec du charbon mineral en presence de
CO2 . Parmi leurs resultats, il est possible d ’observer la liberation plus rapide de composes
volatils entre 425 et 550 °C en presence de N 2 comparativement a une atmosphere de CO 2
(Figure 2.15). Les auteurs de l’etudes rapportent que cette difference s’est repetee tout au long
de leurs experiences et ils attribuent cette difference a un craquage thermique et a la
gazeification au CO2 (sans toutefois preciser la reaction de gazeification). De plus, ils notent,
qu’en atmosphere de CO 2 , le rendement en composes volatils augmente avec l’augmentation
de la temperature (davantage a partir de 550 °C - Figure 2.16). L ’augmentation du taux de
chauffe a aussi ete notee comme diminuant la production de composes volatils en presence de

41
CO 2 ; une tendance qui n ’a pas ete verifiee en presence de N 2 (Figure 2.17) mais qui
correspond a des resultats deja observes par Butterman et Castaldi [2010]. II est interessant de
noter, qu’entre 500 et 700 °C, le CO produit est favorise en presence de CO 2 comparativement
a l’environnement de N 2 (Figure 2.18).

N. atmosphere
C 0 2 atmosphere

K -0.

O -0.08

8 ' ° 10
- 0.12

200 400 600 800 1000

Temperature [ °C ]

Figure 2.15: Analyses thermogravimetriques differentielles de pyrolyse de

charbon mineral en atmospheres de N 2 ou de CO2 .
Source : Duan et al. [2009]

100
N2 atmosphere
CO, atmosphere

to
<N

200 400 600 800 1000

Temperature [ °C ]

Figure 2.16: Etude thermogravimetrique de pyrolyse de charbon mineral en

atmospheres de N 2 ou de CO2 . Source : Duan et al. [2009]

42
 77 o.oo
^ -0.02
-0 04

c3 -0.06

« -0.08

« -0.10
- 0 .1 2

200 400 600 800 1000

Temperature [ °C ]

Figure 2.17 : Effet du taux de chauffe sur la pyrolyse de charbon mineral en

atmosphere de CO2 . Source : Duan et al. [2009]

0.04
N2 atmosphere
0.03
C 0 2 atmosphere
O
J 0.02
u.
O
tfi

< 0.01

0 .0 0

300 400 500 600 700 800 900 1000

Temperature [ °C ]

Figure 2.18 : Monoxyde de carbone produit lors de la conversion de charbon

mineral en atmospheres de CO 2 ou de N 2 . Source : Duan et al.
[2009]

Naredi et Pisupati [2011] ont effectue la pyrolyse d ’un charbon bitumineux a faible contenu en
composes volatils en comparant l’atmosphere de CO 2 a celle de l’argon. En presence de CO 2 ,
a environ 400 °C, la perte massique du charbon mineral etait d’environ 6 %, comparativement
a 12 % en presence d’argon. Ce comportement a cependant demontre un effet inverse pour les
temperatures repertoriees au-dela de 900 °C, causant une plus grande conversion en composes
volatils en presence de CO2 . Ainsi, pour Naredi et Pisupati [2011] et Duan et al. [2009], la

43
presence de CO 2 versus N 2 semble reduire la formation de composes volatils entre 400 et 550
°C lors de la pyrolyse de charbon mineral.

Jindarom et al. [2007b] observerent dans leurs travaux visant a comparer la pyrolyse de boues
municipales en environnements de CO 2 ou de N 2 (utilisant un reacteur a lit fixe) que la
presence de CO2 reduisait legerement la production de biocharbon et augmentait la production
des composes volatils condensables ainsi que celle des non condensables (Figure 2.19). Ces
resultats ont ete corrobores par des analyses thermogravimetriques (ATG) qui presentent une
baisse de la production de charbon a partir de 300 °C et qui se maintient jusqu’a environ
550 °C (Figure 2.20 (a)). Leur resultats ATG sont aussi supportes par des resultats d’analyse
thermogravimetrique differentielle (DTG) ou le taux de degradation (variation de la masse par
unite de temps) en fonction de la temperature variait en accord avec leurs resultats ATG. Les
resultats DTG presentent une variation de la premiere reaction de decomposition thermique de
boues municipales significativement acceleree en environnement de C 0 2. Des resultats qui
sont appuyes par d’autres auteurs auxquels Jindarom et al., [2007b] font reference (Figure
2 .2 0 (b et c)).

70 ■

60

350 550 650

T e m p e ra tu re (°C)

Figure 2.19 : Bilans des produits de pyrolyse obtenus en atmospheres de CO 2 ou

de N2. Legende : solides (•), liquides (A), gaz (■), CO 2 (vide) et
N 2 (plein). Source : Jindarom et al. [2007b]

44
 20

•o
10

£ 05 ■
X.
X. 0.00
O
-e
C4j 40

- . 0 2.
$ .s
E
£
p
*
200 300 400 500 600 ^
Temperature (°C)

Figure 2.20 : Analyses thermogravimetriques et analyses thermodifferentielles

de boues municipales en atmospheres comparatives de N 2 (• ) ou
de CO 2 (o). Source: a : Jindarom et al. [2007b], b & c :
Thipkhunthod et al. [2006] (tire de Jindarom et al., [2007b]).

Une difference a aussi ete notee au niveau de la composition des gaz non condensables a partir
de 550 °C oil le CO augmente en presence de CO2 alors que le CO2 produit diminue (Figure
2.21). De plus, l’hydrogene et le methane presentent aussi des concentrations plus elevees en
environnement de CO2, et ce, a partir de 450-550 °C. Les auteurs de l’etude attribuent la
difference de ces compositions de gaz possiblement a la presence de metaux dans les boues
municipales qui agiraient comme catalyseurs aux temperatures notees pour la reaction inverse
de Boudouard (C + CO2 = 2CO; AH298 = + 164,9 kJ/mol), pour la reaction inverse de
conversion a la vapeur d ’eau (CO2 + H2 = CO + H20 ; (AH298 = + 42,3 kJ/mol)) et de
reformage sec du methane (CH4 + C 0 2 = 2CO + 2H2; (AH298 = + 261,0 kJ/mol)). Leur
explication par rapport a l’augmentation du CO en presence de CO2 a 550 °C, faisant reference
a la reaction inverse de Boudouard, serait en accord avec les travaux d’Osaki et Mori [2006]
ainsi qu’avec les resultats de Duan et al. [2009]. De plus, etant donne que les tests de Jindarom
et al. [2007b] ont ete effectues dans un reacteur batch ou 1’environnement gazeux est
constamment renouvele par un apport nouveau en C0 2 (vecteur gazeux), sans apport en
biomasse (boues municipales dans ce cas), les gaz produits sont done continuellement dilues

45
dans 1’apport C 0 2 (vecteur gazeux), tout en restant dans le reacteur pour un temps de reaction
tres court. Le C 0 2 (vecteur gazeux) ne semble apparemment pas avoir ete considere dans le
bilan des gaz produits. Ainsi les reactions directes (vers la droite) des reactions suggerees
(Boudouard inverse, conversion a la vapeur d’eau inverse et reformage sec du methane)
seraient favorisees plutot que l’atteinte de l’equilibre chimique, ceci pouvant done ainsi
constituer une explication pour l’augmentation du CO observee en presence de
1’environnement gazeux de C 0 2.

5.00 COj (N, atmosphere)

COj (COj atmosphere)
4.00 CO (N2 atmosphere)
CO (COj atmosphere)
3.00 o-

2.00

1.00

0.0
350 450 550 650
Temperature < ’C)

Figure 2.21 : Production de CO et C 0 2 lors de la pyrolyse de boues municipales

en atmosphere de C 0 2 ou de N 2. Source : Jindarom et al. [2007b]

Effet du CO2 sur la composition des produits condensables et des biocharbons

Jusqu’a maintenant, les quelques etudes portant sur la pyrolyse de biomasse en atmosphere de
C 0 2 ont effectue un suivi principalement sur la composition des bio-huiles plutot que sur celle
du biocharbon obtenu. Des similarites doivent tout de meme exister entre la composition des
composes contenus dans le biocharbon et celle des bio-huiles etant donne que certains
composes du biocharbon resultent de la condensation des composes volatils lors de la
devolatilization. Ainsi, des resultats d’etudes portant sur les produits condensables obtenus en
pyrolyse a basse severite, en atmosphere de C 0 2, et retrouves dans la litterature sont presentes
dans les sections suivantes.

Zhang et al. [2011] ont etudie l’influence de l’environnement gazeux lors de la pyrolyse
rapide de rafles de ma'fs sous environnement de N2, C 0 2, CO, CH4 ou H2. Ils ont note des

46
variations entre la composition des produits condensables de pyrolyse selon un environnement
de N 2 ou de CO2 . En presence de CO 2 et comparativement a l'atmosphere de N 2 , les contenus
relatifs (supposes dans ce cas comme etant bases sur l’aire sous la courbe du chromatogramme
GC; non specifie dans l’etude en question) des composes de bio-huiles sont passes de 9 a 16 %
pour les composes acides, de 15 a 17.5% pour les cetones, de 33 a 27 % pour les phenols, de
7,5 a 6 % pour les sucres alors que les esters, aldehydes, alcools et ethers sont restes stables
(Figure 2.22). De plus, l’atmosphere de CO 2 a produit moins de composes contenant des
fonctions methoxy (precurseur de la polymerisation des bio-huiles). Parmi les composes
principaux de ces categories, en atmosphere de C 0 2, le contenu relatif en acide acetique etait
plus eleve (11 % en CO2 vs 7,25% en N 2 ). Jindarom et al. [2006] firent aussi cette
observation par rapport a 1’augmentation du contenu en acide acetique. Ces auteurs expliquent
leurs resultats par des reactions entre les composes volatils et le CO 2 et les biocharbons avec le
CO 2 [Jindarom et al., 2006; Zhang et al., 2011], Ces reactions pourraient en fait s’expliquer
par des reactions de carboxylation qui se produiraient en presence de CO2 . L ’equipe de Zhang
et al. [2 0 1 1 ] a aussi observe une difference (en % relatif) pour les composes suivant: acide
glycerique (4,9% en CO 2 vs1 1,3% en N 2 ), furfural (8 , 2 en CO 2 vs 7,5% en N 2 ), 1,1-
dimethoxy-propane (3,0 % en CO 2 vs 1,3 % en N 2 ), phenol (4,1 % en CO2 vs 3,0 % en N 2 ), 4-
methoxybenzene-l,2-diol (1,2 % en CO 2 vs 2,8% en N 2 ), 2-methoxy-4-vinylphenol (3,0 % en
CO 2 vs1 4,6% en N 2 ), l,6-anhydro-(3-D-glucopyranose (5,1 % en CO2 vs 3,4% en N 2 ), 2,3-
dihydro-benzofurane (5,5 % en CO2 vs 7,0 % en N 2 ). De plus, la variation de l’atmosphere
gazeuse s'est soldee par un pouvoir calorifique superieur des liquides en presence de CO2 vs
N 2 (20,2 MJ/kg vs 17,8 MJ/kg, respectivement).

47
 40

35 A tm osphere

B N,
30
■ C 02

£ 25
O CO
c ■ ch 4
0

20 □ h2

a 15
10

Acids E sters A ldehydes A lcohols K etones Phenols Ethers O thers

containing

Figure 2.22: Concentration de groupes de composes chimiques obtenus des bio-

huiles de la pyrolyse rapide de rafles de mat's sous environnements
de N2, CO 2 , CO, CH4 ou H2. Source : Zhang et al. [2011]

Jindarom et al. [2007b] effectuerent la pyrolyse de boues municipales a des temperatures de

350, 450 et 550 °C, comparant l’atmosphere de CO2 a celle de N 2 et observerent que l’usage
de CO2 eut certaines repercussions sur la composition des composes volatils condensables.
Selon leurs observations, les composes aliphatiques ont diminue et la formation de composes
carboxyliques et de cetones a ete favorisee (Figure 2.23). Des resultats qui supportent, une fois
de plus, une possible carboxylation en presence du CO 2 .

48
 60 -

50 -

40 -

30 -

20 -

™ M onoarom atics
Aliphatics
— Oxygenates
1----- 1 Nitrogenates
SlcriiKis
50 - ■ ■ ■ PACs

40 -

30 -

20 -

10 -

a l

350 450 550 650

Temperature (°C)
Figure 2.23 : Classes de composes de produits condensables obtenus en
atmospheres de N 2 (a) ou de CO 2 (b). PA C s: Composes
polycycliques aromatiques. Source : Jindarom et al. [2007b]

Jindarom et al. [2006] firent aussi la pyrolyse de boues municipales a des temperatures de 105
a 500 °C, en atmosphere de N 2 ou de CO2 et ils observerent, qu’a partir de 500 °C, en
atmosphere de CO 2 , le H 2 , le propylene et le butane ont ete favorises alors que moins de CO2 a
ete produit qu’en atmosphere de N 2 . L’acide acetique a aussi ete favorise en environnement de
CO2 , mais seulement pour les temperatures au-dela de 500 °C. L’analyse par FTIR des
charbons de boues municipales obtenus par pyrolyse en atmospheres comparees de CO 2 vs N 2
(de 350 °C a 750 °C) n ’a toutefois pas presente de differences notables (Figure 2.24)
[Jindarom et al., 2007a].

49
 N350
C350
M 50

C450

N550

N650
J C550
C650

N750
C750

i
4000 3400 2800 2200 1600 1000 400 4000 3400 2800 2200 1600 1000 400
W avenum ber (cm !) W avcnum ber ( c m 1)

Figure 2.24 : Spectres FTIR de biocharbons de boues municipales produits par

pyrolyse en atmospheres de N 2 (a) ou de C 0 2 (b). Source :
Jindarom et al. [2007a]

Effet d’une atmosphere gazeuse de C02sur la structure du biocharbon

Lors de la production de charbon active, 1’activation connue sous l’appelation « activativation
physique» (activation thermique avec agent gazeifiant) se produit en conditions de
gazeification enrichies au C 0 2 dans le but de retirer les goudrons et maximiser la porosite du
charbon [Smemik, 2009], Cependant, l’activation physique s’effectue generalement a des
temperatures de 900 a 1200 °C, favorisant ainsi la reaction inverse de Boudouard (CO 2 + C *-*
2CO). Ceci laisse tout de meme presager un potentiel pour les conditions a basses
temperatures, mais plus specialement aux temperatures avoisinant la temperature d ’equilibre a
pression atmospherique (704 °C; pour AG=0, sans catalyseur).

Encinar et al. [1998] firent des tests de pyrolyse en presence de CO2 avec des bagasses
d’olives et ont obtenu des aires de surface de 126, 207 et290 m2/g pour 300, 400, 500 °C,
respectivement. Ces resultats ont cependant ete obtenus pour un temps de residence de 30 min,
et en presence d’un agent utilise pour l’activation chimique des biocharbons, le ZnCl2. A ces

50
temperatures, l’etude ne rapporte toutefois pas la surface specifique etudiee sans ZnCh ou en
presence d’un vecteur gazeux inerte (ex. : N 2 ou Ar).

Jindarom et al. [2007a], ayant pyrolyse des boues municipales dans un environnement de CO 2
ou de N 2 ont observe une surface specifique de leurs biocharbons plus importante en presence
de CO2 , mais seulement a partir de 650 et 750 °C (jusqu’a 60,7 m 2 /g). Pour les temperatures
moindres, 350, 450 et 550 °C, le type d ’environnement gazeux a eu peu d ’effet sur la surface
specifique (BET), celle-ci variant de 15 a 20 m /g pour chaque condition. Les auteurs
expliquent que le changement en fonction de la temperature resulte de la reaction inverse de
Boudouard agissant sur les atomes de carbone a 1’interieur de la particule et reagissant avec le
CO2 pour ainsi favoriser la formation de pores.

Butterman et Castaldi [2010] ont effectue des tests de conversion de biomasses en

environnement de CO2 en comparaison avec un environnement de vapeur d ’eau et d ’azote
melanges. Leurs etudes de conversions de divers sources et constituants de biomasses dans ces
environnements gazeux ont ete effectuees par analyses ATG. Les observations par rapport aux
biocharbons obtenus lors des tests ATG presentent des biocharbons plus poreux (macropores)
en atmosphere 100% CO 2 . Cependant, il semble que la porosite des biocharbons ait ete
verifiee seulement a partir de 860 °C. La presence de CO2 , comparativement a
Tenvironnement du melange de vapeur d’eau et N 2 , a permis de transformer completement le
biocharbon (au-dela de 930 °C), contrairement au biocharbon en environnement H2 O-N 2 ,
lequel est reste en partie sous forme solide. Un comportement qu’ils ont attribue a la reaction
inverse de Boudouard et au developpement des macropores par Tusage de CO 2 favorisant une
conversion plus complete aux temperatures superieures, specifiquement pour un taux de
chauffe tres bas (1 °C/min). Ces auteurs ont, entre autres, presente des resultats d ’analyses
ATG de constituants de la biomasse. Bien que leur etude ne visait pas la comparaison avec un
environnement a 100 % N 2 , la comparaison avec des donnees de la litterature permet
d ’observer que, pour des conditions similaires, une difference apparait dans la conversion de
la lignine a 500 °C (Tableau 2.4). La conversion de la lignine en composes volatils a 500 °C
est plus importante en condition de CO 2 que de N 2 (rendements en biocharbons moindres),
cependant, il est a noter que les lignines comparees proviennent de methodes d ’extraction

51
differentes et que ces methodes peuvent affecter la lignine et les analyses de ces demieres
[Mohan et al., 2006],

Tableau 2.4 : Rendements en biocharbons de lignine et de cellulose obtenus par ATG en

atmospheres de CO2 ou de N 2

Taux de Debit m. Rendements en biocharbons

Temp. Env. gaz Reference
chauffe gazeux echantillons Cellulose Lignine

°C *C/min ml/min mg %' %'

300 100%CO2 10 75 20-25 92 83,5* [BuUennan et
Castaldi, 2010]
400 100%CO2 10 75 20-25 10 61* [Butterman et
Castaldi, 2010]
500 100%CO2 10 75 20-25 8 47* [Butterman et
Castaldi, 2010]
300 100 % n 2 10 120 20 94 80** [Yang etal., 2006]
400 100 % n 2 10 120 20 7 68** [Yang et at., 2006]
500 100 % n 2 10 120 20 7 62** [Yang etal., 2006]
Legende : Temp. : Temperature, Env.: Environnement, *Lignine “Organosolv”, **Lignine de Sigma-Aldrich
Chemie GmbH,1 : par rapport au constituant initial.

Teng et al. [1997] ont compare la preparation de charbon active a partir de charbon mineral
avec et sans pretraitement. Le pretraitement consistait a soumettre le charbon a 200 °C dans un
environnement d’oxygene avant un traitement commun de chauffage jusqu’a environ 900 °C
en atmosphere de CO 2 . Les auteurs ont observe que le charbon pretraite avec O 2 montrait une
surface specifique beaucoup plus grande que celle du charbon non pretraite. Des resultats qui
seraient compatibles avec de T information rapportee par Downie et al. [2009] ou, lors de tests
de gazeifications (comportant CO2 et faibles quantites d’C>2 ), Teffet oxydatif de l’oxygene sur
le carbone fixe aurait eu un effet significatif sur l’augmentation de la surface specifique.

2.3.4 Effet du temps de residence du biocharbon en reacteur sur la

production de biocharbon

L ’augmentation du temps de residence du biocharbon en reacteur tend a sa devolatilization

complete, done a reduire la valeur des ratios H/C & O/C [Krull et al., 2009], Ceci est
cependant favorise pour des vapeurs et des gaz restant de courtes periodes en contact avec le
biocharbon en reacteur [Ioannidou et Zabaniotou, 2007], II est done important de distinguer
temps de residence des biocharbons en reacteur vs temps de residence en reacteur des

52
composes volatils produits. Les composes volatils produits peuvent etre forces a rester et a
reagir avec le biocharbon (et potentiellement augmenter la production de biocharbon), comme
dans le cas du charbon de bois, ou etre forces a sortir rapidement du reacteur, comme dans le
cas de la pyrolyse rapide pour production de bio-huile. Les principaux parametres operatoires
en reacteur juges comme favorisant la surface specifique et les micropores des biocharbons
so n t: la temperature, le taux de chauffe et le temps de residence du biocharbon [Downie et al.,
2009], L ’economie du procede peut cependant etre grandement affectee par le temps de
residence du produit solide en reacteur. Pour une production en continu, plus le temps de
residence des biocharbons en reacteur est long (jusqu’a plusieurs heures pour une pyrolyse
lente), plus volumineux doivent etre ces reacteurs. Cette etape de conversion peut alors
devenir tres limitante pour la chaine de valorisation de la biomasse. C’est pour cette raison que
des tests de production de biochar et charbon vert doivent favoriser 1 ’etude de temps de
residence courts d’un point de vue industriel et, specialement, pour le developpement d ’unites
portables (appuye par Downie et al. [2009]). La litterature actuelle couvre davantage la
pyrolyse rapide et la pyrolyse lente de biomasse lignocellulosique; ainsi, des etudes sur une
pyrolyse moderee dont le temps de residence se situerait entre ces plages d’operation (temps
de residence de quelques minutes; Tableau 2.3) pourrait etre fort interessantes pour la
communaute scientifique. De plus, un survol de la litterature et des etudes portant sur la
pyrolyse douce a 300 °C (torrefaction) permet de constater que la tres grande majorite des
travaux disponibles visent des temps de residence tres longs et des environnements ou la
convection est generalement tres faible.

2.3.5 Effet du taux de chauffe sur les caracteristiques du

biocharbon

Generalement, plus le taux de chauffe est eleve et plus la formation et l’expulsion de

composes volatils de la biomasse sont favorisees, reduisant ainsi le rendement en biocharbon.
A des taux de chauffe eleves, de grandes quantites de composes volatils sont forcees a
Texpulsion alors que la structure solide passe, quant a elle, par des etapes d ’affaiblissement,
de gonflement et de fragmentation qui peuvent aussi influencer la liberation des composes
volatils [Biagini et al., 2009; Zanzi et al., 1996]. Un taux de chauffe eleve aurait tendance a
briser des liens et former des polymeres fragmentes representant des sites reactifs pour des

53
reactions heterogenes de gaz et matieres solides [Biagini et al., 2009], L’augmentation du taux
de chauffe peut favoriser la formation de fissures et/ou de porosite dans le biocharbon. Ces
fissures et cette porosite aurait aussi comme effet d’augmenter la surface specifique et done de
favoriser la reactivite du charbon vert pour une valorisation energetique. Ces fissures et sites
reactifs seraient aussi potentiellement favorables pour l ’etablissement de microorganismes et
pour les proprietes adsorbantes dans le cas du biochar. A des taux de chauffe eleves, la
dimension des particules a un impact car, pour de petites particules de bois, la structure de bois
serait maintenue, evitant ainsi la formation de fissures dans le biocharbon [Paris et al., 2005].
Pour les grosses particules, la force d’expulsion des composes volatils favorise la formation de
fissures. Ceci est en accord avec la pyrolyse rapide, ou les particules sont generalement de
dimension tres fine (< 6 mm [Bridgwater et Peacocke, 2000]) afin de favoriser le transfert de
chaleur et atteindre des taux de chauffe tres eleves. Les particules liberent leurs composes
volatils et, par consequent, augmentent la production de bio-huile. Les charbons de biomasse
de pyrolyse rapide sont, en effet, observes comme ayant des surfaces specifiques faibles
[Brewer et al., 2009].

A l’inverse, une diminution du taux de chauffe favorise un rendement plus eleve en

biocharbon, une tendance qui est cependant asymptotique a partir de taux de chauffe sous
1 °C/min [Antal et Gronli, 2003; Biagini et al., 2009; Demirbas, 2000; Sensoz et Kaynar,
2006], La diminution du taux de chauffe favorise aussi la formation des micropores. Les
micropores sont davantage associees a une surface specifique elevee que peuvent l'etre les
macropores formees a des taux de chauffe plus eleves [Downie et al., 2009]. Une diminution
du taux de chauffe a des niveaux tres faibles menerait cependant a des pratiques non
economiquement viables d ’un point de vue industriel [Antal et Gronli, 2003].

Dans le cas specifique de l’utilisation du CO 2 comme vecteur gazeux, l’augmentation du taux

de chauffe sur la pyrolyse de la cellulose a ete observee (lors de tests ATG) comme
augmentant la temperature de la reaction primaire de devolatilization [Butterman et Castaldi,
2010]. De plus, toujours concemant l’utilisation du CO2 comme vecteur gazeux, lors de la
pyrolyse de la lignine, l’augmentation du taux de chauffe, qui a favorise la liberation des
composes volatils, a cependant ete observee comme ayant des limites. Butterman et Castaldi

54
[2 0 1 0 ] ont observe que, lorsque convertie en presence de CO 2 , la lignine se comportait
differemment a 1 °C/min qu’a 100 °C/min (Figure 2.25, B & C). L ’usage du CO 2 (a faible
taux de chauffe) a favorise la formation de macropores (mesure suivant l’atteinte de 860 °C),
facilitant Faeces du CO 2 a l’interieur du biocharbon et favorisant la conversion complete du
biocharbon lorsque soumis a des temperatures plus elevees (sous CO 2 , 930 °C, 1 atm). Cette
tendance a cependant change lorsque le taux de chauffe a ete augmente a 100 °C/min, cas ou
la conversion complete n’a pas ete possible (Figure 2.25). L ’utilisation de taux de chauffe plus
eleves (100 °C/min) a mene a la formation d’une couche exteme sur le pourtour des charbons
de biomasses plutot qu’en la formation de macropores et a restreint F evacuation des composes
volatils restants, resultant ainsi en un gonflement du biocharbon (section 2.3.3; Figure 2.25c)
[Butterman et Castaldi, 2010]. A des taux de chauffe eleves, la structure solide de la biomasse
passe par des etapes d’affaiblissement, de gonflement, et de fragmentation qui peuvent aussi
influencer la liberation des composes volatils [Biagini et al., 2009], La lignine etant reconnue
comme formant des composes oligomeres de masse moleculaire elevee [Boateng et al., 2007],
celle-ci et ses composes ont possiblement subi davantage un affaiblissement plutot qu’une
fragmentation comme la chaine de cellulose pourrait subir par depolymerisation.

Dans le projet actuel, l’usage de biomasse a l’etat brute (non broyee), a des taux de chauffe
moderes-eleves, serait done justifie, tant pour favoriser les fissures dans le biochar que pour
1 ’application industrielle beneficiant d’une particule non broyee et d ’un court temps de
reaction.

Figure 2.25 : Resultats ATG de lignine en atmosphere de CO 2 . A : de 22 a 860 °C

a 1 °C/min, B : de 22 a 930 °C a 1 °C/min, C : de 22 a 1000 °C a
100 °C/min. Source : Butterman et Castaldi [2010]

55
2.3.6 Effet du transfert de chaleur et de masse sur les
caracteristiques du biocharbon

Les facteurs d ’operation en reacteur de pyrolyse observes comme ayant le plus d ’effet sont la
temperature, le taux de chauffe et le temps de residence. Cependant, le transfert de masse et le
transfert de chaleur peuvent aussi compter parmi les facteurs d’importance ayant des
repercussions sur les caracteristiques du biocharbon [Downie et al., 2009; Ioannidou et
Zabaniotou, 2007]. Ainsi, le debit d’un gaz porteur (ex. : gaz inerte et/ou gaz vecteur de
chaleur) ou la dimension de la particule, tous deux affectant le transfert de masse et de chaleur,
peut ainsi avoir un impact sur les caracteristiques du charbon de biomasse. Selon Willner et
Brunner [2005], lors de la modelisation de la pyrolyse de particules de bois, il importe de
considerer ces deux facteurs. Generalement, les particules de plus grandes dimensions ont un
plus faible transfert de masse, ce qui a pour effet de contrer la liberation des composes volatils,
causant des rendements en biocharbon plus eleves. Pour de tres fines particules (ex. : mechantiiion
entre 20 et 300 mg obtenues de poudres finement pulverises), ce comportement n ’est
normalement pas observe [Sharma et al., 2002]. Plusieurs analyses thermogravimetriques
utilisent frequemment des masses de cet ordre. Les resultats des analyses de pyrolyse (ATG vs
en reacteur) retrouves en litterature doivent ainsi etre analyses en prenant en consideration ce
facteur. Comme la reduction de la taille des particules de biomasse necessite des couts eleves,
la presente recherche vise done une taille de particule du panic erige s’approchant de l’etat
brut. De plus, tel que presente dans la section precedente (Effet du taux de chauffe sur les
caracteristiques du biocharbon : 2.3.5), l’utilisation d ’une particule de plus grosse taille
soumise a un taux de chauffe plus eleve peut favoriser la . formation de fissures lors de
1’expulsion des composes volatils. Ainsi, l’utilisation de taux de chauffe moderes a
relativement eleves est visee dans le cadre de cette etude. De plus, afin de compenser les
transferts de masse et de chaleur qui pourraient etre reduits par des particules de tailles non
reduites et des temps de residence relativement courts, l’usage d’un gaz porteur (azote ou
CO2 ) sera fait a des vitesses superficielles favorisant une importante convection, mais restant
sous le seuil de fluidisation des particules. Le transfert de masse et de chaleur jouant un role
important dans le procede de pyrolyse de biomasse, bien que des taux de chauffe tres eleves
puissent etre commandes au reacteur, la temperature reelle a l’interieur de la particule de
biomasse (et ultimement du biocharbon) ne suit pas necessairement ce taux de chauffe. C ’est

56
pourquoi les reacteurs con^us dans le cadre de cette etude doivent faire un suivi de la
temperature par 1 ’insertion de thermocouples tres pres du biocharbon (entre les biocharbons).

2.3.7 Effet de la pression sur les caracteristiques du biocharbon

Une augmentation de la pression tend a augmenter le rendement en biocharbon lors de la

pyrolyse (Figure 2.26). Tel que rapporte dans Antal et Gronli [2003], Taugmentation de la
pression pouvant etre obtenue par la reduction du taux de purge des gaz force ainsi les
composes volatils a rester en contact avec le biocharbon, leur permettant de reagir ensemble,
ce qui a ete observe comme augmentant le rendement de charbon de biomasse. A l’oppose,
des conditions de pression sous vide diminuent la formation de biocharbon [Antal et Gronli,
2003], Pour une application industrielle avec des unites de pyrolyse portables, l’usage d ’un
reacteur sous pression pourrait s’averer non justifie etant donne la complexity d ’operation et
les couts potentiels supplementaires lies a ce type de reacteur. De plus, des extractions sur des
charbons de bois avaient demontre une concentration elevee en divers types de HAP
[Kushwaha et al., 1985], Par consequent, bien que l’etude en question ne specifiait pas le
mode de production, il est probable que ces charbons de bois aient ete produits dans un
reacteur limitant la liberation des composes volatils afin de favoriser le rendement en charbons
de bois (souvent obtenus a pression au-dela de 1 atm). Le biochar, devant minimiser ce type de
composition, doit done potentiellement favoriser la liberation des substances volatiles. Pour
ces raisons, l’emphase a ete portee sur T operation a une pression atmospherique ou les
composes volatils sont facilement expulses du reacteur.

57
 24
23
22
21 low flow
20
19
medium flow
18
17
16
15 high flow -
14
13
12
11
10
9
8
0.0 0.5 1.0 1.5 2.0 2.5
P re s su re [MPa]

Figure 2.26 : Effet de la pression et du debit gazeux (inerte) sur le rendement en

biocharbon lors de la pyrolyse de cellulose. Source : Antal et Gronli
[2003]

2.3.8 Composition chimique des extraits de biocharbons

La litterature actuelle portant sur les composes formes lors de la pyrolyse de la biomasse
couvre principalement la composition des bio-huiles. Tres peu d ’etudes portent sur les
composes chimiques dans la matrice des divers biocharbons. En general, l’heterogeneite
chimique a Tinterieur d ’un meme biocharbon peut etre tres grande etant donne les multiples
reactions chimiques s’y produisant [Amonette et Joseph, 2009]. Bien que cette information
soit tres peu disponible, il est probable que certains composes du charbon de biomasse
s’apparentent aux composes presents dans les bio-huiles. Le suivi des composes produits lors
de la pyrolyse (bio-huile) provient surtout de l’interet entourant la pyrolyse rapide ayant connu
un essor au cours des demieres decennies tel que rapporte par Mohan et al. [2006], Selon
Krull et al. [2009], l’etude des composes extractibles des biocharbons par des solvants
contribuerait significativement a l’avancement du biochar et de sa comprehension comme
amendement pour le sol.

Parmi les quelques etudes repertoriees sur le sujet, des briquettes de charbon de bois ont ete
soumises a une extraction afm d ’analyser leur contenu en HAP. Une grande variete et des

58
concentrations importantes de HAP se retrouvaient dans ces charbons [Kushwaha et al.,
1985]. Aucun detail quant aux conditions de production n ’apparait dans l’etude en question.
Cependant, il est possible de supposer que les conditions de production de ces charbons de
bois limitaient l’expulsion des composes volatils, et ce, pour augmenter le rendement en
charbons. Plus recemment, des extractions de charbons obtenus d ’un melange de residus de
plastiques, de bois et de pneus ont ete effectuees dans le but de reduire la composition des
charbons en contaminants organiques [Bernardo et al., 2010; Bernardo et al., 2009], Une etude
recente a investigue specifiquement la presence d ’HAP et de dioxines dans les extraits de
differents types de biocharbons (Figure 2.27) [Hale et al., 2012], Selon ces auteurs, tres peu
d ’etudes ont paru sur l’extraction du charbon de biomasse et aucune ne s’est penchee
specifiquement sur la composition en HAP et en dioxines. Les auteurs concluent de leur etude
que les contenus obtenus en HAP et dioxines (total et biodisponible), de biocharbons apres
pyrolyse lente, etaient sous les niveaux critiques et pouvaient done etre appliques au sol.
Cependant, ces resultats devaient se limiter aux biocharbons de pyrolyse lente etudies.
Amonette et Joseph [2009] rapportent la composition chimique de composes preleves de
biocharbons de fumier de poule. Ces demiers comportaient 6 types de composes : des N-
heterocycliques, des derives de furane, des derives de phenol, des derives de benzene, des
carbocycliques et des aliphatiques.

La presence de tres peu d ’etudes pour ce type de caracterisation du biochar dans la litterature,
et la pertinence de ce type de caracterisation des biocharbons pour la comprehension de
l’interaction biochar-sol-plante, justifient d’integrer ce type d’analyse a l ’etude du biocharbon.
La presente etude vise done, entre autres, a contribuer a l’avancement du biochar grace a la
caracterisation chimique de ses extraits. De plus, dans le cas de l’usage du CO2 , cette
caracterisation n ’ayant pas ete effectuee, elle permettrait d ’explorer 1’effet potentiel du CO 2
aux conditions visees par l’etude. II est a noter que les conditions concemees par la presente
etude visent une pyrolyse moderee. Ainsi, les composes identifiables peuvent done differer des
produits condensables generalement etudies lors de la pyrolyse rapide de biomasse et
constituer des donnees d’interet pour la communaute scientifique. De plus, la biomasse
torrefiee, pouvant servir de matiere premiere tant pour la bioenergie que pour la chimie
« verte », pourrait done beneficier d ’une connaissance accrue de ces extraits.

59
 : H P-H eartbnd Ptoe(gasHfcation)
HW-Maedhardwood {fan pyrolysis)
4.0- U P -L o d g ep ctep n e (modem stow pyrolysis)
DOM • Qpestod dairy menure (modem stow pyrolysis)
FW- Food w i f i {modem stow pymtyeis)
PM W -Paper mAwato (modem stow pyrolysis]
3.5- CS - Com stove* (modem slow pyrolysis)
EFB VESTO - empty fmk bunches (VESTO method)
l

EFB NTS - empty fruit bunches (Matoysan tredtoenal stove)

EFB TUX) - empty fruit tomch e s (top tit up ifraft)
Total MH oofleenSotfton M

3.0 - CCS TLUO - eooonut A iel (top a i p draft}

CCS 3 stone-ooconut shaft (traditional 3 stone <re)
WS - Wheat straw (micro— t t )
RWSO - Rubber wood sawdust (modem slow pyrolysis}
0 - Oak (modem stow pyrolysis, m ode famace)
2.5-
P - Pme (modem stow pyrolysis. m u«e fcmaee]
6 - Grass (modem slow pyrolysis, muffle furnace)
OA - O at aped (modem Aow pyrolysis todowed by leaching with disttBed wator)
Wk - Ptoe aped (modem stow pyrolysis faiowed by toaohmp with dabbed wetor)
2.0 G A -G raw aged (modem stow pymfysts todowed by ieaohinp edh dhaiedw W r)
PW >Pine wood (modem stow pyrofysto)
SO • Swdeh Grass (modem stow pyrofysts)
Zambia chamaofdest- m in d savannah vegetation ftncMonalIds. stowpyrolysis}
1.5 - 2amtod stover • oem (tnM onat U K stow pymyAs)
todonesia mixed - forest (tradrtonati kiln, stow pyvetysw)
indOMfiattMfc-hM (drouble dnan, slow pyrolysis)
Kenya corncob, comstover and sawdust (vaddkmal 3 stone toe)
1,0

as-

0,0 iii
H$88888Sg8g§l8g?8gg§§fS8i88§IHgaSg88g8S88g 888888888888881!
vVVi l i § £^^^iIII8S8| s i | i | | 0 0 0 “ ‘ 0 0 0 »88SS«83 8 I H i H i 8 8 8 8 8 8 *

30mm

P y ro ly s is ttm ®

Figure 2.27 : Contenu en HAP de divers types de charbons de biomasse. Source : Hale et al. [2012]

60
CHAPITRE 3 ARTICLE 1 - PRODUCTION
DE BIOCHARBON EN REACTEUR A LIT FIXE
DE CAPACITE 1glbatch - EFFET DE LA
TEMPERATURE A DE COURTS TEMPS DE
RESIDENCE
Avant-propos

Auteurs et affiliation :
G. Pilon : etudiant au doctorat, Universite de Sherbrooke, Faculte de genie, Departement
de genie chimique et de genie biotechnologique.
J.-M. Lavoie : professeur, Universite de Sherbrooke, Faculte de genie, Departement de
genie chimique et de genie biotechnologique.
Date d’acception : 4 octobre 2011
Date de publication : 6 octobre 2011
R evue: BioResources
Reference : [Pilon et Lavoie, 201 lb]
Titre fran^ais : Caracterisation du biocharbon de panic erige produit en conditions de
torrefaction et de pyrolyse
Contribution de l’article a la these :
Cet article contribue a la these en investiguant des conditions de production de
biocharbon qui simulent l’entree soudaine d’une biomasse (panic erige) en reacteur
industriel. La dimension des particules n’a pas ete reduite par broyage afin de simuler un
procede qui eviterait l’energie su p p lem en tal liee a ce pretraitement. Le reacteur, con?u
et assemble par G. Pilon, d ’une capacite de 1 g/batch, est opere a 300, 400 et 500 °C, a de
courts temps de residence (2,5 et 5 min), a un taux de chauffe eleve, ainsi qu’a des
conditions ou la convection est favorisee. L ’etude du biocharbon dans ces conditions,
regroupant l’usage d’un court temps de residence avec les autres conditions d’operation

61
 du reacteur reunies, n ’a pas ete couvert dans la litterature et specialement a 300 °C. De
plus, l’extraction des composes des biocharbons par Sohxlet et les analyses de ces extraits
par GC-MS ont ete effectuees pour en identifier le contenu. Ces identifications sont tres
rares dans le domaine de la pyrolyse de biomasse lignocellulosique en general.
Resume :
Dans le cadre de cette recherche, le biocharbon fut prepare a partir de l’espece « Cave-in-
R ock» du panic erige (Panicum virgatum). Les biocharbons furent produits et
caracterises dans une perspective de valeur ajoutee, a la fois comme amendement pour le
sol ou comme matiere premiere dans la production energetique. Le biocharbon de
torrefaction/pyrolyse fut produit dans un reacteur « banc d’essai» a lit fixe, dans une
atmosphere d’azote, a trois temperatures (300, 400, et 500 °C) et pour deux temps de
residence (2,5 et 5 minutes). Les analyses immediates, elementaires et de la valeur
calorifique furent effectuees sur chacun des biocharbons obtenus. Des photos par
microscope electronique a balayage furent prises. La composition organique des extraits
de biocharbons obtenus par Soxhlet (avec dichloromethane) fut identifiee a l’aide d ’un
chromotographe en phase gazeuse muni d’un spectrometre de masse (GC-MS). Le
biocharbon fut aussi analyse par spectrometre infrarouge a transformee de Fourier muni
du dispositif de reflexion totale attenuee. La majorite des caracteristiques des biocharbons
analyses ne demontrerent pas de difference significative selon le temps de residence (2,5
ou 5 min), et ce, pour chaque temperature etudiee. Parmi les composes identifies, deux
pourraient etre d ’un interet particulier par rapport au domaine du biochar. Le 1-
undecanol, un compose utilise comme regulateur de croissance des plantes en agriculture,
a ete observe dans les biocharbons de 300 °C - 5 min et de 400 °C - 2,5 min de temps de
residence respectivement. Le naphtalene, couramment utilise comme un ingredient de
pesticides, fut le seul hydrocarbure aromatique polycyclique (HAP) observe dans le
biocharbon. Ce dernier fut retrouve seulement dans les extraits de biocharbon produit a
500 °C (pour les deux temps de residence).

Mots cles : biocharbon, pyrolyse, torre/action, biochar, panic erige, extractions Soxhlet.

Note : A la suite des corrections demandees par les membres du jury, le contenu de cet article
differe de celui qui a ete accepte.

62
Characterization of switchgrass char produced in torrefaction and pyrolysis
conditions

Guillaume Pilon and Jean-Michel Lavoie

Abstract
In this research, char was prepared from Switchgrass (Panicum virgatum) Cave-in-Rock
cultivar and characterized from the perspective of value addition either as a soil amendment or
as a feedstock for energy production. Char from torrefaction/pyrolysis was produced in a fixed
bed lab scale reactor, using a N 2 flow at three temperatures (300, 400, and 500 °C) and for two
residence times (2.5 and 5 minutes). Proximate and ultimate analyses as well as calorific value
tests were done on each of the chars obtained. Chars scanning electron microscopy (SEM)
pictures were taken. Organic composition of char leachates was characterized using gas
chromatography-mass spectrometry (GC-MS) following a Soxhlet extraction with
dichloromethane. Char was analyzed using FTIR-ATR. No significant difference was
observed in most char characteristics between 2.5 and 5 min residency times, for each
temperature studied. Among the compounds identified, two may be of particular interest with
respect to the biochar domain. 1 -undecanol, a compound commonly used as a growth
regulator in agriculture, was observed in 300 and 400 °C chars for 5 and 2.5 min residence
time, respectively. Naphthalene, commonly used as an ingredient in pesticide products, was
the only polycyclic aromatic hydrocarbon (PAH) observed in chars and was found solely in
the extracts obtained from the char produced at 500 °C (for both residency times).

Keywords: Char; Pyrolysis; Torrefaction; Biochar; Switchgrass; Soxhlet extractions.

3.1 Introduction
Biomass char, for a long time considered as a residue, may play a major role in biomass value
addition chain. Char from torrefied biomass (< 300 °C) can contribute to the economic
feasibility o f biomass supply chain as “solid fuel” by reducing the energy requirements during
grinding, as well as transportation via densification, while improving storage stability [Felfli et

63
al., 2005; Phanphanich and Mani, 2011; Uslu et al., 2008; Zwart et al., 2006], For energy
production, char may be advantageous for the overall process, either for gasification (CO 2
environment) or for combustion [Boateng, 2007; Prins et al., 2006a]. As reported by
Phanphanich and Mani [2011], torrefaction char, in comparison to raw biomass, may also
produce less tar during gasification, while its energy content is superior to original feedstock.
Nevertheless, char “solid fuel” feasibility still depends on its final price for the gasification or
combustion user. Char obtained at higher severity conditions may also serve as a soil
amendment, which can increase crop yield, store carbon (CO 2 reduction) and reduce nutrients
leaching due to surface adsorption [Lehmann, 2007].

Torrefaction techniques have been studied for a long time, however, its investigations have
mostly emphasized the use o f wood as the original feedstock. In the past 5 years, published
results using other lignocellulosic biomass such as agricultural wastes and grasses have been
reported to some extent. Nevertheless, reported results usually have involved long reaction
times, while convection within the reactor was rarely high. Sadaka and Negi [2009] studied
how torrefaction of wheat straw, rice straw, and cotton gin waste produced at residence time
up to 60 min (260 °C) improved physical and thermochemical properties o f the feedstocks for
further utilization. Bridgeman et al. [2008] studied torrefaction of reed canary grass and wheat
straw using thermogravimetric analysis (TGA) with the specific reacting conditions provided
by TGA apparatus (low convection and few mg of powdery form material). However,
torrefaction at 300 °C, at high heating rates, with high convection, and short residence time
(< 1 0 min) of non woody lignocellulosic biomass, investigating physicochemical
characteristics o f char has been, to our knowledge, sparsely reported in literature.

As for pyrolysis char (>300 °C), especially for soil amendment applications (biochar),
multitudes of factors influence its properties [Joseph et al., 2009]. Strictly looking at physical
properties, the factors considered to have a possible major effect on biochar properties are the
original biomass (chemical composition, ash composition, size), the operating conditions
(temperature, heating rate, mass transfer rate, residence time, oxidative conditions, pressure,
type of input (gas - vapour), catalyst), the pre-treatment (drying, crushing), and the post­
treatment (activation, crushing) [Downie et al., 2009]. Among these factors, the highest

64
reacting temperature was reported as the major factor influencing the physical properties,
while heating rate and pressure came in second. The overall biochar characterization domain
with respect to production conditions is in development. According to Krull et al. [2009],
some characterization methods such as solvent-extractable components o f biochar have rarely
been attempted and could provide valuable insight on characterization and eventually
application o f char as a soil amendment. Extraction o f chars, followed by compound analysis,
conducted on different wood charcoal barbecue briquettes have previously been reported in
literature. Authors found presence of various PAHs although no details on the production of
the wood briquette were specified [Kushwaha et al., 1985]. More recently, extraction o f chars
obtained from pyrolysis o f plastic, wood and tire wastes mixture (at 0.41 MPa and 420 °C for
15 min) was conducted in order to reduce their organic contaminants composition [Bernardo
et al., 2010; Bernardo et al., 2009].

In this research, char was prepared from switchgrass {Panicum virgatum), a species that is
being considered as an energy crop in many geographical location in North America,
including Quebec (Canada). Switchgrass was selected due to its potential as a dedicated crop
for energy production as well as for its agronomic attributes [Samson et al., 2005]. Chars’
physicochemical characteristics in function of temperature (300, 400, 500 °C) and residence
time (2.5 and 5 min), at high heating rate and moderately high convection were analyzed. Such
residence times were chosen in order to verify conditions that might be beneficial for smaller
reactor sizing as well as for the biomass supply and transformation chain rate. Chars were
characterized for proximate and ultimate analyses as well as for their calorific value. Scanning
electron microscopy (SEM) pictures were taken to follow the changes in biomass physical
structure. BET analysis was carried out to study effects related to surface area. FTIR using
attenuated total reflectance equipment (ATR) was also utilized to follow char chemical
transformation as function of temperature. Organic composition of char leachates were
qualified using gas chromatography-mass spectrometry (GC-MS) following Soxhlet extraction
with dichloromethane.

65
3.2 Experimental
3.2.1 Biomass

Feedstock used for the experiments was Switchgrass {Panicum virgatum), Cave-in-Rock
cultivar. The feedstock was grown in the Eastern Township region of Quebec, Canada. Crops
were cultivated in a clay soil; the material selected had a 1.5 year root system and was
harvested in Fall (November 2009). The biomass was harvested and baled in 500 kg bales with
moisture content below 10 % and stored for about 6 months under open structures protected
from rain before samples were taken in laboratory. Switchgrass samples used for this research
were maintained in unsealed plastic bags at room temperature for 1 to 4 months (20 to 25 °C).

3.2.2 Experimental setup and biomass char production

Char production was conducted into a lab-scale fixed bed reactor. The reactor consists in a
stainless steel tube installed within a furnace and mounted vertically (Figure 3.1). Switchgrass
was cut into pieces shorter than 10 cm. Biomass humidity levels, which varied all along
experiments, typically ranged from 4 to 10% (dry basis). Before each test, a biomass holder
(Figure 3.2 and Figure 3.3) was preheated at reactor’s temperature outside the reactor, using a
preheater installed underneath the reactor, open to ambient air environment. This step was
used to simulate the feeding o f biomass into a reactor; such biomass would suddenly enter the
reactor, therefore achieving high heating rates. The short contact with ambient air while
passing from preheater to reactor simulates air that can enter in minor quantity within a non
purged continuous feeding system. For each batch, biomass was inserted in the preheated
biomass holder, at lg/batch and quickly inserted in the reacting chamber using a thermocouple
as a displacing rod. This technique resulted in an almost instantaneous biomass heat up to
desired reacting temperature (Figure 3.4). The thermocouple (displacing rod) allowed for the
monitoring o f temperature variations within the biomass holder (Figure 3.1 to Figure 3.3). A
quick connection around the thermocouple allows a sealed connection with the reactor when
the biomass holder and thermocouple were inserted. High convection around particles was
obtained for a N 2 flow rate of 0.115 L/s (STP: 101.3kPa - and 25 °C) using an inline
flowmeter. The N 2 flow resulted in a superficial velocity estimated at 0.33 m/s in the holder

66
(STP). These conditions were estimated to be slightly below minimal fluidization levels o f the
biomass particles. The inert gas was preheated at reactor’s temperature using an inline
preheater and passed within the furnace before entering the reactor (Figure 3.1). Chars were
produced at 300, 400 and 500 °C (± 5 °C) for two distinct residence times in the reactor (2.5
and 5 min.) calculated from the instant biomass was inserted in the reactor (Figure 3.4). Once
the time o f reaction was successfully reached, char was quickly taken out o f the reactor,
transferred in a pan and maintained into a desiccator. Chars produced in several batches from
same specific conditions were homogenized as a unique sample prior to analysis. Further
details about the reactor, the experimental setup, and the operating conditions were presented
in a previous report (see Pilon and Lavoie [201 la])1.

3.2.3 Feedstock and char characterization

Proximate and organic elemental analysis, calorific value and specific surface
Analytical methods for proximate analysis, organic elemental analysis, calorific value
determination, BET surface area as well as cellulose, hemicelluloses, and lignin content in the
original biomass were all reported previously in Pilon and Lavoie [201 la]1.

B iom ass F urnace

holder's
preh eater R eactor

B iom ass
holder

Figure 3.1: Schematic representation o f the experimental setup.

Figure 3.2: Side view of biomass holder.
Figure 3.3: Bottom view o f biomass holder.

1 Bien que cet article soit le premier de la presente these, sa publication dans un ouvrage de la litterature a suivi
l’article du chapitre 4. Certaines sections du present chapitre font done reference a Particle du chapitre 4.

67
 600
Biomass insertion in biomass holder
550
followed by entrance into reactor
,u 500

n 450
CL
400 Biomass holder Char holder removed
preheating from reactor
350

300
0 100 200 300 400 500
Residence time [s]
Figure 3.4: Example o f a representative temperature profile within biomass
holder during a lg/batch test. (500 °C, 5 min).

Scanning electron microscopy pictures

Scanning electron microscopy was performed using a Hitachi model VPSEM S-3000N device.
Prior to SEM analyses, char was first applied on an autosticking carbon and its surface
metalized, thus allowing conduction. Metallisation was done using an Anatech Hummer VI
sputtering system using a 60%Au-40%Pd electrode. Pictures presented were taken at a
magnification of 45X and 500X. Representative pictures were selected from pictures taken of
few samples.

Organic components extraction and GC-MS analysis

For each extraction, one gram of ground char (-90% 10-200 mesh particles size) was initially
placed in 25 mm ID X 90 mm cellulose cartridges and extracted with 150 mL of
dichloromethane. All extractions were done in triplicates. Extractions were conducted for 8

hours at about 10 cycles/hr. After extraction, the solvent and extracts were filtered to remove
any char particles and transferred into a round flask for concentration using a vacuum rotary
evaporator. Each concentration was done at 40 °C under a vacuum of 20 kPa relative to
ambient pressure. Reduction was done until complete solvent evaporation. The concentrated
product was collected using about 1-2 mL of dichloromethane, and the vials were kept at 4 °C
prior to injection in the GC-MS. The gas chromatograph utilized was a HP 5890 Series II
equipped with a HP 5971A Mass Selective detector and a HP 7673 controller. The column

68
was an HP-lms GC column (Agilent Technologies). Each sample was analyzed in a splitless
mode. Gas chromatography and mass spectroscopy were performed using temperatures o f 325
°C (inlet) and of 280 °C (detector). The oven’s temperature profile consisted in an initial
temperature of 90 °C for 1 min, followed by a temperature increase at 10 °C/min up to 180 °C.
The heating rate was then changed to 5 °C/min up to 270 °C where it was maintained for a
total run time of 58 min. Compounds were identified using the Wiley7 library.

Char FTIR analysis

Solid char was analyzed by infrared spectroscopy using attenuated total reflectance (FTIR-
ATR). The FTIR used in this experiment was a model 640 (Varian), while the ATR was
purchased from Pike Technologies. Char pieces obtained for every studied condition were
analyzed in four replicates. For each sample, 300 scans were run between 600 and 4000 cm '1.

3.3 Results and discussion

3.3.1 Balance on char production

Figure 3.5 depicts char yield reduction behaviour with increasing temperature. Despite a slight
reduction at 300 and 400 °C, ANOVA statistical method showed no significant difference
among each residence time for all temperatures investigated (a = 0.05). Char yields and
analytical results (on a char basis) are presented in Table 3.1. Based on values obtained in
Table 3.1, biomass losses (on an original biomass basis), in function of temperature, are shown
in Table 3.3 in order to follow the composition of feedstock losses. For the char obtained at
300 °C (2.5 or 5 min.), volatile content seemed to be the main loss (Table 3.3). Volatiles
losses were mostly related to oxygen losses, which were found to be a twofold value in
comparison to the carbon losses (Table 3.3). The 31.1-32.3% mass loss for the char produced
at 300 °C (Table 3.3; expressed as CHON or volatile losses) could correspond mostly to
hemicelluloses decomposition; hemicelluloses are known to decompose at temperatures
ranging from 225 to 300 °C [Prins et al., 2006c], Hemicellulose content in switchgrass Cave-
in-Rock cultivar was evaluated at 36.3% (determined by difference between holocellulose and
a-cellulose; see Pilon and Lavoie [2011a]) (Table 3.2), a value close to the one obtained by

69
Lemus et al. [2002] who obtained a value o f 32.0% for the same cultivar. On the other hand,
char losses cannot be related solely to hemicelluloses degradation. Despite their conversion,
hemicelluloses volatiles may also condense to form tar or coke, partly constituting char
[Sharma et al., 2002]. Cellulose and lignin are known to decompose between 305 and 375 °C
and between 250 to 500 °C, respectively; therefore these two biomass constituents, especially
lignin, must also be linked to a certain extent, for the losses observed at 300 °C.

Results obtained at 300 °C are comparable to those of Gilbert et al. [2009], who torrefied
switchgrass at 250 °C for 2h and obtained char yields of 72%db. with elemental and proximate
analyses fairly similar to the results observed in this study (Table 3.1). This suggests that using
much shorter residence time (2.5 min), much higher heating rates and superficial velocity
(about 100X more), and a higher temperature (300 °C vs 250 °C), similar torrefied biomass
properties can be obtained. As a result, this could represent a huge potential improvement in
torrefaction conditions in terms of energy requirements, system’s production rate, and reactor
sizing due to much lower residence time. Further research should, therefore, be conducted on
these operating conditions in fonction of torrefied feedstock properties obtained.

For char production at temperatures of 400 °C and 500 °C, the char yield was drastically
reduced (27.3-25.3 and 12.4-12.3%, respectively) (Table 3.1). Hemicelluloses, cellulose and
lignin are known to decompose at temperatures of 225 to 300 °C, 305 to 375 °C and 250 to
500 °C, respectively [Prins et al., 2006c]. Cellulose was reported by Lemus et al. [2002] to
represent 37.8% o f the bone-dry mass of switchgrass, which is comparable to the 33.7% value
found in the biomass used in this study (Table 3.2); therefore the 72.6-73.7% mass reduction
(Table 3.3) at 400 °C may very well correspond to the total holocellulose (cellulose and
hemicellulose) content of switchgrass (see Table 3.2) and part o f lignin. Lignin must have
continued its decomposition further when pyrolysis temperature was increased to 500 °C,
resulting in the 86.2-87.6% mass loss from original biomass for the char produced at 500 °C.
Boateng et al. [2007] worked on the same feedstock, using a bench-scale fluidized bed fast
pyrolysis reactor. They ran their tests at 500 °C and obtained 12.9% o f char yields. Their
results are comparable to the values (12.4 and 12.3%) that were obtained in the present
research for 2.5 and 5 min retention time respectively (Table 3.1). Nonetheless, at 500 °C, the

70
contact of air with biomass while entering the reactor resulted in momentary ignition while on
exiting the reactor it presented signs o f char glowing red (oxidative pyrolysis). These
momentary conditions necessarily contributed to mass losses. This may also account for the
char ash losses observed at 500 °C (see Table 3.3). No ignition or red glowing char was noted
for 400 °C; however short oxidation may have occasioned char yield reductions. Sharma et al.
[2 0 0 2 ] observed char yield reduction during oxidative pyrolysis of tobacco under similar
conditions. Brewer et al. [2009] also produced switchgrass char at 500 °C using a fluidized
bed reactor and obtained slightly higher char yields (15-20 %).

Char moisture content increased with process temperature (see Table 3.1). This behaviour is
consistent with hydrophobic character of torrefaction chars and water sorption properties o f
pyrolysis biochar [Felfli et al., 2005; Rutherford et al., 2010],

12.5 min
15 min
£
2
at
'>•
ro
JC

300 400 500

T em p e ra tu re [°C]

Figure 3.5: Char yield obtained from pyrolysis of switchgrass at 300, 400 and
500 °C, for 2.5 or 5 minutes under N 2 atmosphere, (d.b.: dry basis).
Error bars correspond to standard deviation.

71
3.3.2 Heating value

The maximal higher heating value (HHV) was obtained for the chars produced at 400 °C
(Table 3.1). This difference between 400 °C chars compared to 300 and 500 °C chars was
significant for 2.5 min chars only (both for P<0.005). For the 400 °C chars, the 2.5 and 5 min
processes resulted in significant increases of 36.6% (P=0.01) and 23.1% (P=0.002),
respectively, with reference to the original feedstock. On the other hand, ANOVA statistical
analysis showed that there was no significant difference between chars produced for the two
residence times of 2.5 and 5 min for each specific production temperature (a = 0.05). The
decrease in heating value at 500 °C could be resulting from the decreasing volatile content and
the increasing ash content (Table 3.1). The HHV obtained for the 500 °C char (19.6 and 22.4
MJ/kg for 2.5 and 5 min, respectively) are comparable to values obtained by Boateng et al.
[2007], who obtained 19.4 MJ/kg using a bench-scale fluidized bed reactor (see Table 3.1).

72
Table 3.1: Production mode, yield, composition, heating value and surface area of representative switchgrass chars

Mat. R.Bed J O
RT HR Conv Yield C H O N H/C O/C MC Ash VC FC HHV S.A. Reference

I°C| [min] |°C/min] * [%]' |%]‘ [%]' 1%]' ** ** [%] |%]' |%]‘ [%]' |MJ/kg] [mJ/g|
1%]'
RSG na na na na na na 44.5 5.8 45.7 0.5 1.55 0.77 6.2 3.7 81.0 15.3 19.5 < 1 Pilon and Lavoie2
SGC Fixed 300 2.5 Fast ++ 70.9 51.3 5.5 38.8 0.5 1.30 0.57 0.4 6.7 72.8 20.5 20.3 < 1 Pilon and Lavoie2
SGC Fixed 300 5 Fast ++ 65.6 51.7 5.5 37.7 0.5 1.28 0.55 0.2 5.3 69.9 24.7 22.7 < 1 Pilon and Lavoie
SGC Fixed 400 2.5 Fast ++ 27.3 65.4 3.8 20.9 0.7 0.70 0.24 1 .8 1 1 .8 34.2 53.9 26.6 3.6 Pilon and Lavoie
SGC Fixed 400 5 Fast ++ 25.3 67.2 3.7 19.6 0.6 0.66 0.22 1 .6 14.6 33.2 52.2 24.0 4.5 Pilon and Lavoie
SGC Fixed 500 2.5 Fast ++ 12.4 63.9 2.6 15.0 0.6 0.48 0.18 1.3 26.6 20.8 52.6 19.6 66.6 Pilon and Lavoie2
SGC Fixed 500 5 Fast ++ 12.3 67.5 2.8 13.8 0.6 0.49 0.15 2.4 23.8 21.7 54.6 22.4 84.6 Pilon and Lavoie
[Gilbert et al..
SGC Fixed 250 120 10 + 72.0 50.9 5.6 37.3 0.8 1.33 0.55 3.7 4.8 70.7 20.8 20.6 n/a 2009]
[Boatcng et al
SGC Fluid. 480 quick Fast +++ 12.9 60.7 4.0 8.7 0.8 0.79 0.11 3.8 25.9 28.4 42.0 19.4 7.8 2007]
[Brewer et a l,,
SGC Fluid. 500 0.025 Fast +++ 15-20 39.4 1.3 6 .1 0.7 0.4 n/a 0.9 52.5 7.1 39.5 15.37 50.2 20091
n.b. : Percentages are expressed on a dry basis. : Part o f these results is presented in Pilon and Lavoie [201 la] along a complementary study
Legend: Mat.: starting material R.Bed: reactor bed RSG: raw switchgrass SGC: switchgrass char T°: temperature na: not applicable
RT: residence time HR: heating rate Conv: convection MC: moisture content VC: volatile content FC : fixed carbon
HHV: higher heating value S.A.: surface area Fluid.: fluidized n/a: not available *: low “+” to high “+++” molar basis

Table 3.3: Losses occurring from original switchgrass for char production 1 Table 3.2: Switchgrass characterization
Comparison CHON VC + FC Identification Mass Units (±S.D.)*
C H O N Ash VC FC
Conditions losses losses
Ash 3.7 (± 0.2)
[%] [%1 [%] [%] [%] [%] [%] [%] [%] a-cellulose 33.1 (± 0.7)
RSG to SGC 300 °C 9.3 2.0 19.6 0.2 31.1 0 32.3 0 32.3 Hemicellulose 36.3 (± 1)
Lignin 17.3 (± 0.6)
RSG to SGC 400 °C 27.1 4.8 40.4 0.4 72.6 0.2 72.1 1.3 73.5
* Expressed in terms of 100 mass units of oven dry
RSG to SGC 500 °C 36.4 5.4 43.9 0.5 86.2 0.6 78.4 8.7 87.1 switchgrass straw.
n.b.: Percentages are expressed on a dry biomass basis & RT 2.5 and 5 min results were averaged for every
temperature condition. 'Calculation example presented in Annexe C.

73
3.3.3 SEM figures and BET results

Scanning electron microscopy pictures (see Figure 3.6 and Figure 3.7) present the degree of
decomposition as function o f temperature at two enlargement levels, 45X and 500X,
respectively. The 45X enlargement (Figure 3.6) shows the structural breakage as temperature
increased, which is consistent with observations from previous research aimed on structural
breakage during devolatization [Downie et al., 2009], These results are consistent with the
observations made during grinding o f material, where grinding of the original biomass was a
tedious process, while 300 °C char crumbled easily, and a powdered material was obtained
much more easily at 400 and 500 °C.

Raw Switchgrass 300 °C - 5 min

400 °C - 5 min 500 °C - 5 min

Figure 3.6: Scanning electron microscopy pictures of raw switchgrass and
switchgrass chars. ^Ellipses show examples o f regions where cavities
development occurs. Enlargement = 45X.
74
At 500X enlargement (Figure 3.7), magnification showed that, as temperature increased, some
biomass sublayers and structures appeared more pronounced, which might contribute to
increased surface area (Table 3.1). As reported by Downie et al. [2009], temperature increase
may result in enhancement o f char pore development to some extent. However, it is not
possible to discern macropores development from the pictures taken (Figure 3.7). It has to be
noted that the surfaces presented (Figure 3.7) are the side o f an original plant structure and not
the tip of a cut plant structure, where cavities resulting from inner plant cells could appear
more pronounced. Such types o f cavities (side view) may be observed in Figure 3.6, resulting
from structural breakage at higher temperatures. Despite the fact that pore development could
not be observed using the SEM (see Figure 3.7), BET surface area increased at 500 °C,
reaching values ranging from 6 6 .6 to 89.0 m2/g for 500 °C at 2.5 min and 5 min, respectively.
Increasing BET surface area at 500 °C may result from this structural breakage, allowing for
access to inner plant cells’ cavities, enhanced by devolatilization. High heating rates such as
the one obtained along these experiments, combined with the higher temperature levels, are
reported to enhance devolatilization, resulting in structure breakage [Biagini et al., 2009], BET
surface area results obtained for 500 °C are similar to those reported by Brewer et al. [2009]
who obtained 50 m /g using a fluidized bed reactor. On the other hand, Boateng [2007] got a
BET surface area of 7.8 m2/g for their char produced in a fluidized bed reactor. Despite the
increasing BET surface area, values obtained in this study for the 2.5 and 5 min chars (500 °C)
have shown no significant statistical difference.

75
 Original 300 °C - 5 min

400 °C - 5 min 500 °C - 5 min

Figure 3.7: Scanning electron microscopy pictures of raw switchgrass and
switchgrass chars. Enlargement = 500X.

3.3.4 Char extraction and composition analyses

The compounds identified in the dichloromethane char extracts are reported in Table 3.4. In
addition, the amounts of extracted compounds are presented in Table 3.5. Among the
identified compounds, aliphatic compounds were observed only at 300 and 400 °C. This
observation is consistent with the work of Sharma et al. [2002], who observed from FTIR and
NMR char analyses a rapid loss in aliphatic groups and an increase in aromatics for
temperatures above 450 °C. Additionally, the same authors reported that chars formed from
cellulose below 400 °C had a higher concentration in aliphatics than chars obtained at higher
temperature. This assumption could be supported by the FTIR-ATR analysis on char, where

76
bands around 800 cm ' 1 were observed as increasing, as well as one around 1600 cm '1, with
increasing temperatures (Figure 3.8). In both cases, the band appeared at temperatures starting
from 400 °C. Wavenumbers o f 700 to 900 cm ' 1 can be associated with aromatics, as well as
the peak at 1600 c m 1, which can correspond to a conjugated C=C linkage, also common in
aromatic compounds. The band decreasing for an increasing temperature between 2800 and
3000 cm ’ 1 would also be consistent with this assumption, since this band may correspond to
two aliphatic bands (aliphatic CH3 (2946-2987 cm '1) and aliphatic CH2 (2888-2946 cm '1)).
The observation is also supported by the H/C and O/C ratios, decreasing with increasing
temperature (Table 3.1) which may correspond to an increase in aromatic structure
concentration [Krull et al., 2009]. The band growing at wavenumber of 885 cm '1, possibly
indicating the appearance o f aromatic C-H at temperatures higher than 500 °C, would be
consistent with observations by Uchimiya et al. [2011] as well as with the increasing char
aromatic content.

The appearance of naphthalene, a polycyclic aromatic hydrocarbon (PAH), observed at

operating conditions of 500 °C, also supports the previous assumptions (Table 3.4). PAH
formation usually increases with operating temperature, where aromatic rings may lose their
functional groups, eventually combining into PAH [McGrath et al., 2001; Nakajima et a l,
2007], Additionally, this temperature corresponds to the level at which lignin thermal
decomposition is enhanced; therefore aromatisation of dimeric fractions o f lignin may have
occurred, leading to naphthalene formation. Sharma et al. [2004] concluded that char aromatic
content and inorganics may act as catalysts in PAH formation. As reported by McGrath et al.
[2 0 0 1 ], naphthalene appearance might be critical for further thermochemical transformations
since heavier PAHs can be created from aliphatic substrates through naphthalene. Naphthalene
has been classified as a group 2B chemical, a category regrouping compounds possibly
carcinogenic to humans [Nakajima et al., 2007]. The same product is also commonly sold as
mothballs, and naphthalene containing products are already used in soil as a fumigant
pesticide under risk mitigation measures [EPA, 2008], Extraction yields at 500 °C correspond
to 1.3 and l.lm g of extracted material per g o f char (2.5 and 5 min, respectively) (Table 3.5).
Linked with the relative content of naphthalene obtained, these values could provide a rough
estimate on how much char produced under these specific conditions could be applied on soil

77
with respect to environmental regulations. However, because several other compounds remain
to be identified and quantified as well as their effect and interaction with soil and plants
remain to be clarified, these values provide a first estimate for further research in the domain,
which should rely on an exhaustive qualification and quantification of all the compounds
extracted from the char.

The 1-undecanol was identified at 300 °C - 5 min and 400 °C - 2.5 min and it is a compound
that could be o f interest for biochar as soil amendment, since this compound is already sold as
a growth regulator [SciFinder, 2011], Occurrence o f such compounds highlights the necessity
to investigate more intensively the chemical composition o f char with respect to char testing
conducted in soil in order to possibly establish some relationships between char composition
and agronomic responses.

It may be observed that vanillin, a benzene ring compound, was found only at 300 and 400 °C
(CAS # 121-33-5). Since vanillin was shown to decompose only at temperatures above 550 °C
[Shin et al., 2001], it is most likely that this compound disappeared simply due to its phase
change, its boiling point being estimated at 285 °C [Lide, 2008]. These results are consistent
with softwood lignin volatiles analysis reported by Sharma et al. [2004], which contained
mostly vanillin and guaiacol. Guaiacol was also observed at a temperature of 300 °C - 5 min
residence time (CAS #90-05-1, Table 3.4).

As reported by Uchimiya et al. [2011], the decrease in bond stretching between 3200-3500
cm '1, with increasing temperature from 300 to 500 °C, represents a decrease in O-H related to
hydrogen bonded hydroxyl groups. These findings may correspond to water formation,
possibly linked to the aqueous content found in pyrolysis oils. The 1700-1740 cm' 1
wavenumbers showed an increase from 300 to 400 °C and a decrease from 400 to 500 °C
which may be associated with a carboxylic acid group (Figure 3.8). However, the only
compound identified containing a carboxyl group is hexadecanoic acid observed at 300 °C, 2.5
min (Table 3.4). A decrease in the IR absorbance from 1030-1100 cm' 1 for 300 to 400 °C
chars may be due to the decreasing amount of C -0 bonds, which would be consistent with the
high reduction in volatiles and oxygen content observed (Table 3.3), an observation that can

78
be correlated with IR results presented by Uchimiya et al. [2011]. The negative peak observed
near 2350 cm ' 1 corresponds to CO2 ; the latter may become negative during background
subtraction due to laboratory environment gas variations.

As shown in Table 3.5, the amount of extracted material from char behaved as could be
expected; it decreased with increasing temperature as well as with increasing reactor residence
time. It is known that compounds present on or in char may result from volatile products
condensation [Sharma et al., 2002] and that increasing temperature and residence time
enhances further conversion of these condensed volatiles. In terms o f specific compounds
identified, Table 3.4 shows their relative content obtained along GC-MS analysis, therefore,
further analyses coupled with compound quantification would benefit the results. In addition,
char characterization studies could gain from having higher char reactor productivity; char
content for extraction could also be increased in order to increase abundance response. In
terms of value addition, especially for biochar as a soil amendment, char could be produced
with a specific content of desired or undesired compounds, which will necessarily have to be
considered along with technology and product development. These results give some initial
insights. However, follow-up work is needed for a better understanding and for sustainable
development o f this domain. For instance, work is needed to clarify why biochar may
contribute to soil and plant productivity in some conditions and less in others, as well as in
defining conditions under which the desired chemical and physical characteristics are
obtained.

79
 0 08j

ooasI M
A* AC
$
Absorbance (arbitrary units)
Char500C .2 S r m 1 I
n§L
_
- T

ft
0 02
& I 'T ~
I

1
I
000| I
-0 0?g I ri-
-004*: Char 4000 - 2.5 rr4n jX
i
__________
"I
i I
ooe-:; i I
i I
J
-010& Char 300C l
I
-012 *: I ,/
I i^
I
i
$
■0 16 f C
A h ': _i
IKw 1 >» . JSkrtf? iT. I/• ;V' ■».' k l’ii'*'-* .V-'.l.
3800 3600 3400 3200 3000* * 2800* 2600 2400 2200 2000 1800 1600 1400 1200 1000 800

Wavenumber (cnr1)

Figure 3.8: FTIR spectra of chars obtained using attenuated total reflectance equipment
(ATR) for 300-500 °C and 2.5 min pyrolysis conditions.

80
Table 3.4: Identified compounds from dichloromethane extractions of switchgrass chars

RT 300 °C 300 °C 400 °C 400 °C 500 °C 500 °C

Identified Compound* CAS#
(min) 2.5 min 5 min 2.5 min 5 min 2.5 min 5 min
Aliphatic
1- D e c e n e 8 7 2 -0 5 -9 1 0 .4 4 3.1

1- u n d e c a n o l 1 1 2 -4 2 -5 1 0 .4 7 3 .2 17.9

H e p ta d e c a n e 6 2 9 -7 8 -7 1 2 .4 4 7.7 2 .3

1 - T e tr a d e c a n a m in e , N ,N - d im e th y l- 1 1 2 -7 5 - 4 1 4 .0 8 12.5 6 .6

H e x a d e c a n o ic a c id 5 7 -1 0 - 3 1 8 .0 2 2.1

Aromatic
P h e n o l, 2 - m e th o x y - ( g u a ia c o l) 9 0 -0 5 -1 5 .1 2 4 .0

P h e n o l, 4 - e th y l- 2 - m e th o x y 2 7 8 5 -8 9 -9 7 .7 0 2 .0 5.3

P h e n o l, 2 ,6 - d im e th o x y - ( v a n illin ) 9 1 -1 0 -1 8 .5 3 2.1 12.1

B e n z a ld e h y d e , 4 - h y d r o x y - 3 -m e th o x y - 1 2 1 -3 3 -5 9 .1 5 13.1 1 9 .4 10.0 2.3

B e n z a ld e h y d e , 3 ,4 - d im e th o x y - 1 2 0 -1 4 - 9 1 0 .1 2 6.3

B e n z a ld e h y d e , 4 -h y d r o x y - 3 ,5 - d im e th o x y - 1 3 4 -9 6 -3 1 2 .6 4 9 .5 9.5

Polycyelic aromatic hydrocarbon

N a p h th a le n e , d e c a h y d r o - , tra n s - 4 9 3 -0 2 -7 5 .0 4 4 1 .5 2 5 .4

N a p h th a le n e , d e c a h y d r o - , c is - 4 9 3 -0 1 -6 5 .6 1 16.8 1 8 .0

RT: retention time * Presence o f compound is indicated by its relative content.

Table 3.5: Amounts o f extracted material from switchgrass chars

yo °C 300 300 400 400 500 500
RT min 2.5 5 2.5 5 2.5 5
Extract m g/g char 2.8 2.5 1.9 1.5 1.3 1.1
Legend : T° : Temperature, RT : residence time

81
3.4 Conclusions
1. Among the identified compounds found in char extracts, two compounds are already used
in agriculture under risk mitigation measures: 1 -undecanol, a growth regulator, and
naphthalene, a compound used in soil as fumigant. Naphthalene was the only polycyclic
aromatic hydrocarbon found in the produced chars.
2. Char production at 400 °C gave the maximal HHV increase over the original biomass.
Nonetheless, mass reduction with respect to original biomass was much greater at 400 °C
(72%) than at 300 °C (31%). At 300 °C, it was also shown that char yields and properties
were fairly similar to results obtained by other researchers operating at 250 °C for 2
hours, using different process conditions.
3. Residence time increasing from 2.5 to 5 min did not lead to significant difference between
most char characteristics.
4. Scanning electron microscopy pictures showed that charred switchgrass surface did not
appear more porous (in terms o f macropores) under the studied conditions. The increased
BET surface area at 500 °C could mostly be the result o f structural breakage at these
temperature levels, allowing an access to inner plant cells’ cavities, enhanced by
devolatilization.

Acknowledgments
Authors would like to thank Le Fonds Quebecois de la Recherche sur la Nature et les
Technologies (FQRNT) for the financial support of Mr. G. Pilon, the University of
Sherbrooke, the Faculty of Engineering, and the Department of Chemical Engineering and
Biotechnological Engineering for supporting this research via the starting funds offered to the
corresponding author as well as the Industrial Chair in Ethanol Cellulosic o f the Universite de
Sherbrooke in lending research equipment. The authors would also like to acknowledge
participation by Ms. Eva Capek, Mr. Henri Gauvin, Mr. Michel Trottier, Mr. Serge Gagnon,
Mr. Marc G. Gagnon, Ms. Sonia Blais, Mr. Gaston Boulay, Mr. Khalid Ouakarame, and Ms.
Marie-Pier Gravel of the Universite de Sherbrooke for their technical support. Finally, Mr.
Daniel Clement, switchgrass producer from the Eastern Township (Quebec, Canada), is
acknowledged for providing the feedstock used in this research.

82
CHAPITRE 4 ARTICLE 2 - PRODUCTION
DE BIOCHARBON EN REACTEUR A LIT FIXE
DE CAPACITE iglbatch EN ENVIRONNEMENT
DE C 02 vs N2 - ETUDE PRELIMINAIRE

Avant-propos

Auteurs et affiliation:
G. P ilo n : etudiant au doctorat, Universite de Sherbrooke, Faculte de genie, Departement
de genie chimique et de genie biotechnologique.
J.-M. Lavoie : professeur, Universite de Sherbrooke, Faculte de genie, Departement de
genie chimique et de genie biotechnologique.
Date d ’acception : janvier 2011.
E tat de 1’acceptation : version finale publiee
Travaux de conference : Energy and Sustainability III, WIT Press
Reference : [Pilon et Lavoie, 201 la]
Titre fran 9 ais : Production de charbon de biomasse a conditions de faible severite en
environnements de C 0 2 ou de N 2
Contribution de Particle a la these :
Cet article contribue a la these en explorant, avec une approche preliminaire, la production
de biocharbon par pyrolyse dans un environnement de C 0 2 compare a l’azote. La
caracterisation des biocharbons produits a partir de biomasse lignocellulosique (panic
erige), dans un environnement de C 0 2, aux temperatures visees, ne semble pas avoir ete
repertoriee dans la litterature. Les conditions d ’operation du procede, qui font suite et qui
comparent certains resultats avec Particle precedent, utilisent le reacteur de 1 g/batch, un
court temps de residence en reacteur (2,5 min uniquement), une convection elevee en
reacteur et l’usage de taux de chauffe allant de modere a eleve. L’effet de l’usage du C 0 2
sur les caracteristiques du biocharbon est introduit dans le cadre de cet article.

83
R esum e:
Dans un contexte de valorisation de la biomasse, le biocharbon, longtemps considere
comme un residu, se presente maintenant comme un possible vecteur de developpement
durable dans le secteur de la biomasse. Certains types de biocharbons (charbons verts) ont
demontre un grand potentiel comme carburant solide ou comme precurseur pour des
transformations subsequentes tout en montrant des avantages au niveau des etapes
d’entreposage et de transport de la matiere premiere. D ’autres types de biocharbons
(biochar) ont demontre un potentiel comme methode de sequestration du carbone et
comme amendement pour le sol, pouvant meme favoriser le rendement des cultures
lignocellulosiques. Selon plusieurs facteurs, mais plus particulierement en fonction de la
biomasse et des conditions de production, les caracteristiques physico-chimiques du
biocharbon peuvent varier enormement. Ainsi, afin d’utiliser le biocharbon pour des
applications precises (ex. : biochar en fonction du type de sol), ces caracteristiques doivent
etre precisement comprises et controlees. Dans cette etude, des biocharbons furent produits
a l’aide d’un montage laboratoire utilisant un reacteur a lit fixe. Tout au long des
experiences, une variete de charbons de biomasse furent produits sous CO2 ou N 2 pour des
temperatures de 300 et 500 °C. Le biocharbon fut produit a partir de panic erige {Panicum
virgatum L.), une plante energetique cultivee au Canada. Le biocharbon fut caracterise par
analyses elementaires et immediates ainsi que pour sa valeur calorifique. De plus, la
surface specifique fut determinee par la technique Brunauer-Emmett-Teller (BET).

Mots cles : BET, biochar, biomasse, charbon vert, CO2 , pyrolyse, torrefaction

Note : A la suite des corrections demandees par les membres du jury, le contenu de cet article
differe de celui qui a ete accepte.

84
 Biomass char production at low severity conditions under C 0 2 and N2
environments

Guillaume Pilon and Jean-Michel Lavoie

Abstract
In a perspective of biomass value addition, biomass char, a thermochemical product, long time
considered as a residue, is now getting attention and may represent a vector in the
sustainability of the whole biomass sector. Some char types have been shown to have great
potential as a solid fuel or precursor for further transformations as well as having attributes for
storage and transportation. Other types showed potential as a soil carbon sequestration
technique and soil amendment enhancing biomass yields. Depending on several factors, but
mostly on biomass and production conditions, biomass char’s physico-chemical characteristics
may vary tremendously. Therefore, in order to be used in accordance with specific
applications (e.g.: biochar as function of soil type), its characteristics must be carefully
understood and controlled. In this study, chars with varying properties were produced in a
custom-made lab-scale fixed bed reactor. Along these experiments, various biomass chars
were produced under C 0 2 or N 2 at temperatures of 300 and 500 °C. Char was produced from
switchgrass (Panicum virgatum L.), an energy crop grown in Canada. It was then
characterized for ultimate and proximate analysis as well as for calorific value. In addition,
specific surface was characterized by the Brunauer-Emmett-Teller (BET) technique.

Keywords: BET, biochar, biomass, bio-coal, CO2, pyrolysis, torrefaction.

4.1 Introduction
Biomass char refers to biochar with respect to soil amendment, and to charcoal when referring
to the charred organic matter used as a source of energy [Lehmann and Joseph, 2009;
McLaughlin et al., 2009], Biochar has been shown to be a potential soil amendment that
would enhance biomass production/hectare. It was also shown to have positive effects on soil
health as well as being able to store carbon for reduction o f C 0 2 emission [Gaunt and
Lehmann, 2008; Lehmann, 2007], Another application, potentially complementary to soil

85
amendment [Laird, 2008], relates to the potential use o f biomass char to provide the required
heat for pyrolysis’ endothermic reaction during bio-oil production. Furthermore, it was shown
to be advantageous as feedstock for combustion or for syngas production through gasification
[Boateng, 2007; Uslu et al., 2008].

Biomass char can be obtained through different thermochemical processes operating under
low oxygen content using varying degrees o f severity. In increasing degrees o f severity order,
torrefaction, slow and fast pyrolysis and gasification are processes under which biomass char
can generally be obtained. Torrefaction can be defined as the conversion o f lignocellulosic
material occurring without oxygen, within a temperature range of 200 to 300 °C, at
atmospheric pressure [Prins et al., 2006b; Uslu et al., 2008], It consists mostly in
hemicellulose degradation resulting in a solid material with low moisture content, hydrophobic
properties and high calorific value if compared to original biomass. Most likely, biomass char
obtained under torrefaction may not meet biochar stable soil characteristics since the original
biomass will need to be treated at a sufficient temperature (>300 °C) in order to yield
carbonized material with modified chemical bonds, which is less likely biodegradable
[McLaughlin et al., 2009]. Slow pyrolysis operates usually at temperature between 300 and
500 °C, at heating rates lower than 20 °C/min and with long residence time (min to days),
conditions that were shown to enhance char production over non condensable gases and oil.
Fast pyrolysis operates at higher heating rates (up to 10 000 °C/s), shorter residence times (< ls
to 30s) and favors bio-oil over char and non condensable gas production [Brewer et al., 2009;
Mohan et al., 2006]. Gasification conditions operate with a low-mid level of oxygen, yielding
generally about 5-10% biomass char. Despite their lower char production yield compared to
slow pyrolysis, fast pyrolysis and gasification chars were recently considered as having soil
amendment potential [Brewer et al., 2009].

Biochar’s physico-chemical properties vary to a large extent depending on a multitude of

factors, especially the reaction conditions used [Joseph et al., 2009]. The operating parameters
that have been reported to influence char properties are heating rate, the highest operating
temperature, pressure, reaction residence time, reactor design, biomass pretreatments, input
flow rate and post-treatments [Downie et al., 2009]. Among these numerous factors,

86
temperature is considered one of the most important factors affecting char properties as it
provides the necessary activation energy for the chemical reactions involved; furthermore it
influences char physical changes resulting from volatilization and structural melts. Studies on
wood showed that the three main biomass components decompose at temperatures ranging
from 200 - 325 °C for hemicelluloses, from 240 to 375 °C for cellulose and from 280 to 500
°C for lignin [Downie et al., 2009; Prins et al., 2006b]. Demirbas [2001] studied carbonization
(medium rate pyrolysis at 10 °C/s) o f several agricultural and forest residues, for temperatures
ranging from 300 to 900 °C. He reported that for every feedstock, solid products were found
to decrease, gases increased and liquid fractions decreased with increasing temperatures;
furthermore, higher heating values (HHV) were also found to increase with increasing
temperature.

Pyrolysis’ gaseous environment also plays a major role. Conventionally, pyrolysis operates
within an environment free o f oxygen, therefore using inert gases such as nitrogen. CO2 is also
an inert gas at room temperature, however, with temperature increasing, CO 2 becomes reactive
and tends to react with carbon char through the reverse Boudouard reaction (C + C 0 2 —> 2
CO).

There are numerous thermochemical processes that could take advantage o f using CO2 . In the
combustion field, especially in the coal industry, the Oxyfuel Technology which consists of
using flue gases with pure oxygen (CO 2 , H 2 O and O 2 ) instead of air (N 2 , O 2 , CO 2 and H2 O)
appears as a promising technology to recycle CO2 especially from the output CO 2 purity
(without N 2 ) and potential CO2 sequestration point o f view [Buhre et al., 2005]. In addition,
flue gases as a gas source contain residual heat which is beneficial for energy flows within the
overall process [Butterman and Castaldi, 2010]. In gasification, use of CO 2 to enhance the
reversed Boudouard reaction and CO formation for syngas production is also a common
practice. During physical activation o f char for activated carbon production, CO 2 is the
medium used for surface area and pore enhancement. Physical activation is usually done at a
temperature around 900 to 1200 °C [Jindarom et al., 2007a]. In presence o f CO 2 , the char
yield is usually reduced towards formation of volatiles [Butterman and Castaldi, 2010;
Minkova et al., 2000], Jindarom et al. [2007a] observed that, for temperatures from 350 to 750

87
°C, the use of CO2 in comparison to N 2 resulted in chars with higher char specific surface and
alkalinity levels.

High specific surfaces as well as high porosity are two desired characteristics for soil
amendment applications of biochar since they increase the number of favourable sites for
microorganisms’ proliferation as well as act as an exchange surface for nutrient retention or
water absorption attributes [Lehmann and Joseph, 2009]. Under pyrolysis conditions, specific
surface was observed to reach peak values at temperatures varying from 350 to 700 °C
[McLaughlin et al., 2009; Sharma et al., 2004], Decrease in specific surface area at higher
temperature was attributed, by some researchers, to melts in char structure. Increasing
retention time was also shown to increase porosity since it allowed most volatilization
reactions to reach completion. As for the heating rate, lower heating rates would allow for
volatile release with less morphological changes, therefore, maintaining natural porosity in
comparison with higher rates which would even lead to morphological transformation of the
cell structure [Downie et al., 2009],

In this research, biomass char was prepared from switchgrass (Panicum virgatum L.).
Switchgrass was chosen due to its beneficial characteristics as an energy crop such as its
hardiness in harsh climate conditions and poorer soils [He et al., 2009; Martel and Perron,
2008]. In a first step, a lab-scale fixed bed reactor was tested using specific conditions
simulating a high heating rate reactor operating under a N 2 atmosphere. In a second step, the
same reactor was operated at a lower heating rate, but adding new atmospheric conditions,
first, under N2, and, secondly, under CO 2 . Temperature, heating rate and gas atmosphere were
the three main parameters studied and their effect on the char’s physico-chemical
characteristics was analysed. The chars obtained were characterized after proximate and
ultimate analyses, for BET as well as for their calorific value. Bio-oil yield was also verified
and the ones derived from the slow heating rate experiments were analyzed by GC-MS.

88
4.2 Materials and methods
4.2.1 Biomass

Switchgrass (Panicum virgatum L.) Cave-in-Rock cultivar utilized for the experiments was
grown in the Eastern Township region of Quebec, Canada. The crop was grown in a clay soil
and had a 1.5 year root system. The feedstock was harvested in Fall (November 2009) at
moisture content below 10% in 500 kg bales and stored in open structures protected from rain.
Switchgrass samples used for experiments were maintained in unsealed plastic bags at room
temperature (20-25 °C).

4.2.2 Experimental setup

For experimentation, a lab scale fixed bed pyrolysis reactor was used. It consists in a stainless
steel (SS) cylindrical reactor of 26.6 mm ID and 0.44 m length installed in a 2400W electric
tube furnace mounted vertically. Biomass is added to the reactor using a biomass carrier
(Figure 4.1). The latter is 13 cm long, 25 mm OD in diameter and is made of stainless steel.
Carrier’s top, bottom and along the length is partially opened in order to allow gas flow
through its structure. A wire mesh is installed within the carrier to retain the biomass from
falling down. Reactor’s entrance is controlled via a gate valve equipped with a motor for quick
control.

During the first experimental tests, operating at a higher heating rate, a preheater installed
underneath the reactor was used in order to preheat the biomass carrier at the reactor’s
temperature. The carrier has a top opening in order to fill it with biomass. The action o f filling
the reactor and taking it into the reactor was done for every test in less than 10 s. During the
second step o f experiments, at lower heating rates, the biomass carrier preheater was not used.
The biomass carrier is attached to a thermocouple (T/C), which serves as a rod to move it into
the reactor. The T/C also provides an indication on temperature evolution within the biomass
carrier.

89
Nitrogen or carbon dioxide were the two gas vectors utilized while also serving to enhance
heat and mass transfer between gases and feedstock. Gas flow input was maintained at 0.115
L/s (STP: 101.3kPa - and 25 °C) for every test and was controlled by a flowmeter before
entering the reactor. The flowmeter was calibrated with respect to each gas. Nitrogen gas and
CO 2 were, respectively, preheated before entering the reactor using an inline gas heater and
while passing in the annulus region between, the reactor and the tube furnace. Gases and
volatiles exiting the reactor passed through a heated line maintained at around 200 °C +/- 25
°C before being diverted to an exhaust system.

For pyrolysis vapour mass balance, vapours were collected within two consecutive glass
condensers; both were kept within an isothermal container filled with dry ice. Pyrolysis gases
were calculated from the difference from char and condensable liquid yields.

Real time temperature monitoring was done on the overall system using the National
Instrument Data Acquisition System NI cDAQ-9172 equipped with Labview software.
Thermocouples were positioned all along the system as shown in Figure 4.1. In Figure 4.1, T
corresponds to thermocouples, where T1 refers to biomass carrier, T2 to gas outlet, T3, T4 and
T5 to the annulus region between the tube furnace and the reactor, T6-T9 to outlet gas lines,
inside and at surface temperature and T10 at gas inlet to monitor gas inlet temperature.
Pressure (PI, Figure 4.1) was verified on the reactor exit using a mechanical pressure gauge in
order to maintain an absolute pressure close to 101.3 kPa in the reactor.

90
 VT-2AF ■J-6 ,
R e a c to r
— L S
Tube ____— 1 |
F u rn a c e i <

J4 . JS, Bio-oil
A n a ly se s

•JH vT-a
G as
E x h au st
J-5 ,

Bio-oil
C o n d e n s e rs
B io m a ss
c arrier
p re h e a te r

J -1
B io m a ss
C har
c arrier
A n a ly se s

Figure 4.1: Schematic representation o f the experimental setup.

4.2.3 Biomass char production and product mass balance

Biomass was first cut into pieces shorter than 10 cm and the humidity level, which ranged
between 4 and 10%, was monitored all along the char production periods. Typically, one gram
of switchgrass was inserted in the biomass carrier per batch. For each batch operating at higher
heating rates, the biomass carrier was first preheated until reaching reactor’s temperature,
before being rapidly filled with biomass and inserted into the core of the reactor. This resulted
in almost instantaneous heating up o f biomass. Nitrogen (N 2 ) provided the gaseous
environment at the higher heating rate. At lower heating rate, a comparison between N 2 and
CO 2 was done. In these latter conditions, the biomass carrier was not preheated; biomass was
inserted into the carrier, approached to the reactor’s opening to flush air and rapidly inserted
into the reactor. Biomass was kept in reactor for the desired residence time o f 2.5 minutes,

91
after which it was removed from the reactor, transferred and cooled down at room temperature
in a desiccator and later on transferred to hermetic containers until further analysis. During
experiments at the high heating rate, time was monitored from the moment the biomass was
inserted into the preheated carrier. For moderate heating rates, time was counted once
temperature reading indicated the desired temperature. It took about 7.5 and 9.5 min. to reach
500 and 300 °C, respectively, for slow heating rates conditions. For all batches, the tube
furnace was maintained at a fixed temperature. Chars were produced at 300 and 500 °C +/- 10
°C. Chars produced under specific conditions were mixed together before analysis. For a
moderate heating rate only, a mass balance on products was performed. Bio-oil was collected
using condensers and non condensable gases were determined by mass balance difference.

4.2.4 Feedstock characterization

Proximate analysis and calorific value

Char as well as original feedstock were both characterized for ash content and volatile content
following ASTM E1755-01 and ASTM E 872 methods. Fixed carbon was determined by
difference from the ash and volatile content. Higher heating value (HHV) was determined
using an oxygen bomb calorimeter, Model 1341 from Parr Co.

Organic elemental analysis

For CHON organic elemental content analysis, finely powdered feedstock was analyzed using
a Leco TruSpec Micro CHNS equipped with the TruSpec Micro Oxygen Add-On Module for
oxygen analysis.

BET analyses
Biomass samples were first sieved and the 40-60 mesh material was used for BET analyses.
Surface area characterization was performed using a Micromeritic physical adsorption
apparatus, ASAP 2020, using nitrogen adsorption based on BET theory. Degassing was done
under vacuum from an initial heating rate o f 10 °C/min up to a temperature o f 110 °C kept for
a minimum of 16 hours.

92
Organic components
Bio-oils samples were collected from the slow heating rate conditions only and were analyzed.
Bio-oils collected into condensers were diluted into acetone before being directly injected into
GC-MS. The gas chromatogram used is a HP 5890 Series II equipped with a HP 5971A Mass
Selective Detector and a HP 7673 Controller.

4.3 Results and discussion

4.3.1 Heating rate effect

At 300 °C (N2 atmosphere), change in heating rate (from fast to slow) did not seem to have a
major effect on char’s physicochemical properties (organic elemental composition, proximate
analysis and specific surface), despite the fact the char yield slightly decreased while higher
heating value (HHV) slightly increased (Table 4.1). For 300 °C (N2 atmosphere), char yield
and char volatile content could have been expected to be lower for higher heating rates, since
generally higher heating rates favour bio-oil production, which was not the case at 300 °C,
but was the case at 500 °C. Perhaps that in this case (at 300 °C - N2 ), the longer residence
time in the reactor during temperature progression period had a more pronounced effect on
char yield than the quenching effect of high heating rates usually enhancing bio-oil yield. Char
yield being lower at 300 °C - N 2 (66.9 v.v 70.9% for slow and fast heating rate, respectively) is
at least consistent with the higher fixed carbon content (24.4 vs 20.5% for slow and fast
heating rate, respectively) that may have resulted from longer residence time (during heating
stage), releasing more volatiles at temperatures lower than 300 °C. This latter would also
explain the increase in heating value content. The higher heating values obtained for 500 °C
(under N 2 atmosphere) in this research are really close to what was previously reported in
literature using a bench scale fluidized bed reactor design (19.6 and 23.1, for fast and slow
heating rates, respectively [this study] vs 20.1 MJ/kg [Boateng et al., 2007]). At 500 °C,
decreasing heating rate seems to have a pronounced effect on surface area passing from 67 to
82 m 2 /g. It has to be noted that contact with oxygen during the preheating at high heating rate
may have had impact on char properties due to oxydation o f char, which would be consistent
with the lower char yield obtained under 500 °C high heating rates. Bio-oil content was not

93
collected at high heating rates since biomass carrier preheating step, achieved outside the
reactor, resulted in volatiles losses.

4.3.2 Gas environment effect

At 300 °C, the gas environment did not seem to have a major effect on physicochemical
properties. All parameters studied did not vary, which is consistent with theory keeping in
mind that CO2 does not react with carbon at lower temperatures [Jindarom et al., 2007a]
(Table 4.1). On the other hand, despite the fact it would require further analyses, the presence
o f compounds such as syringol, methoxyeugenol and palmitic acid was noted in 300 °C - CO 2
oils, but not at 300 °C N 2 . A further investigation on oil is however required in order to prove
that observation.

At 500 °C though, the gas environment seemed to have a more pronounced effect. First o f all,
decrease in char’s yield (15.9 to 13.6%, N 2 vs CO2 , for n=20&22, respectively) and char’s
carbon content (65.4 to 61.8%, N2 vs CO 2 , respectively; significantly different for P=0.02) in
presence of CO2 suggests that the Boudouard reversed reaction could have taken place,
resulting in higher carbon conversion (as shown previously) and gas yield (73.6 vs 79.5%, N 2
vs CO2 , respectively). Since 500 °C is a relatively low temperature for the reversed Boudouard
reaction to take place, it seems that at least an element could have acted as a catalyst. It this
case it could be the nickel and the iron present in the stainless steel of the reactor, the
thermocouple and the biomass carrier due to the close contact with biomass. As reported by
Osaki and Mori [2006], in presence o f Ni, the reversed Boudouard reaction during gasification
may be enhanced at 500 °C. Boudouard reaction is known to be catalyzed by VIII metals as
well as by alkali and alkali-earth salts (Jin et al. [2012]). It has to be noted also that the ash
content expressed per mass unit of original biomass presented a diminution at 500 °C - CO 2 vs
N2; 0.016 vs 0.025 g ash m Char/g biomass d.b.> respectively; typical metal inorganic content of
switchgrass is shown in Annexe D. Butterman and Castaldi [2010] observed that the activation
energy for pyrolysis and gasification in a CO 2 atmosphere was lowered in presence of
herbaceous feedstock in comparison to wood. This difference was attributed mostly to the
higher ash content in herbaceous feedstocks.

94
Specific surface is also another parameter which was influenced by carbon dioxide. At 300 °C,
there was no change noticed, however at 500 °C, the specific surface passed from 82 to 189
m2/g (Table 4.1), which is consistent with results from Jindarom et al. [2007a] who observed a
drastic change in specific surface at 550 °C.

4.3.3 Temperature effect

Temperature may have a pronounced effect on surface area. Increasing temperature from 300
to 500 °C increased surface area from less than lm 2/g, for all 300 °C char production
strategies studied, to 67 m2/g using a high heating rate and to 87 m2/g at a low heating rate
(Table 4.1). Despite the fact that the high increase in surface area at 500 °C might be
beneficial for a soil amendment application, the decrease in char yield, passing from about
68 % to 14% (for 300 vs 500 °C, respectively; heating rate and gas environment conditions
confounded), must be seriously considered while estimating the economics o f the process,
depending on the desired application. Char yields ranging from 66 to 71 % at 300 °C could
result mostly from hemicellulose decomposition as reported by Prins et al. [2006b]. This latter
assumption would be consistent with the switchgrass hemicelluloses content which was
estimated at 36.3% (Table 4.2). In each case, char yield greatly decreased with increasing
temperature. This mass balance is consistent with observations made by other authors showing
that cellulose decomposition occurs at temperatures from 305 to 375 °C and lignin at
temperatures from 250 to 500 °C [Prins et al., 2006b], Switchgrass cellulose content
represents 33.7% of the biomass composition (Table 4.2). Lignin being more reluctant to
decomposition, would balance the remaining losses and leaving about 14% in char yield.
Boateng et al. [2007] worked on the same plant species using a bench-scale fluidized bed fast
pyrolysis reactor. They ran their tests at a temperature o f 500 °C and the char produced
corresponded to 12.9% o f the original biomass, which is very close to the values obtained in
our study.

It has to be noted that char was immediately removed from the reactor once the residence time
was completed. As a result, for the 500 °C assays, the biomass and biomass carrier remained
at a temperature high enough to allow char to react with the ambient oxygen, resulting in char

95
I

glowing red. This definitively may have had an effect on char yields’ decrease as well as having an impact on char’s characteristics.

Table 4.1: Torrefaction and pyrolysis product analyses

Production conditions Mass balance Char analyses

Feedstock Heating 'po Gas Char Oil Gas C H O N Ash Volatiles FC HHV S.A.
rate Pci Env. [%1 [%] (%1 f%l f%l [%] [%1 [%) [%1 [%1 MJ/kg m2/g
Raw sg. n.a. n.a. n.a. n.a. n.a. n.a. 44.5 5.8 45.7 <1 3.7 81.0 15.3 19.5 --
Char FAST 300 n2 70.9 -- - 51.3 5.5 38.8 <1 6.7 72.8 20.5 20.3 <1
Char FAST 500 n2 12.4 - - 63.9 2.6 15.0 <1 26.6 20.8 52.6 19.6 67
Char SLOW 300 n2 66.9 9.0 24.1 53.3 5.5 38.9 <1 4.7 70.9 24.4 23.5 <1
Char SLOW 300 co2 66.6 10.6 22.8 54.9 5.5 38.4 <1 5.4 68.9 25.7 23.1 <1
Char SLOW 500 n2 15.9 10.6 73.6 65.4 2.6 17.4 1 15.1 19.7 65.2 23.5 82
Char SLOW 500 co2 13.6 6.9 79.5 61.8 2.1 18.6 1 14.7 21.5 63.8 25.1 189
n.b.: Percentages are expressed on a d ty basis.
Legend: sg. : Switchgrass T °: Temperature FC : Fixed carbon
n.a.: Not applicable Env.: Environment S .A .: Surface area

Table 4.2: Original composition o f Panicum virgatum straw

Identification Standard Test Mass Units (±S.D)*
Ash ASTM E1755-01 3.7 (± 0.2)
a-cellulose ASTM D ll03-60 33.7 (± 0.7)
Hemicellulose By difference 36.3 (± 1)
Lignin ASTM D1106-56 17.3 (±0.6)
* Expressed in terms o f 100 mass units o f oven dry switchgrass straw

96
4.4 Conclusions
Preliminary investigations on char formation from switchgrass under N 2 environment show
results similar to those reported so far in the literature. Lowering the heating rate greatly
enhanced the specific char surface area, but only at 500 °C, as shown in this study. Carbon
dioxide showed to have a great impact on specific surface at 500 °C, leading to a specific
surface of 189 m2 /g, which more than doubled the N 2 results. For a biochar-soil amendment
application, where optimized yields are desired as well as a high specific surface, C 0 2 might
then represent a strong advantage since the char yield was comparable to the yield using 500
°C - N2. Leaching of ashes prior to pyrolysis or using a reactor free o f Ni could be
recommended in further experiments in order to target whether an element or another was
responsible for the enhancement of C 0 2 effect. Despite the fact C 0 2 seems to be inert at 300
°C based on physical characteristics of chars, preliminary results for bio-oil content analysis
revealed some differences which must be further investigated. It was also reported that final
reacting conditions at open atmosphere resulted in decreased char yield. Despite the fact this
obvious reaction with oxygen could easily be reduced by waiting for char to cool down, the
effect o f such post treatment on char properties should also be analyzed further with respect to
char value additions and the practical applications of the technology.

Acknowledgements
Authors would like to thank Le Fonds quebecois de la recherche sur la nature et les
technologies (FQRNT) as well as the Universite de Sherbrooke, the Faculty o f engineering
and the Department o f chemical engineering for the financial support offered to Mr.
Guillaume Pilon’s PhD thesis. Thanks also to the Universite de Sherbrooke, to the Faculty of
Engineering and to the Department o f chemical engineering and biotechnological engineering
for supporting this research (through starting funds for Prof. Jean-Michel Lavoie and lending
research equipment) as well as to the Industrial Chair in Ethanol Cellulosic in lending research
equipment. The authors would also like to acknowledge Ms. Eva Capek, Mr. Henri Gauvin,
Mr. Serge Gagnon, Mr. Marc G. Couture and Mr. Michel Trottier o f the Universite de
Sherbrooke for their technical support. Finally, we wish to thank Mr. Daniel Clement,

97
switchgrass producer from the Eastern Township, for providing us the feedstock for the
experiments.

98
CHAPITRE 5 ARTICLE 3 - ETUDE DES
BILANS DE PRODUITS DE PYROLYSE
OBTENUS EN ENVIRONNEMENTS DE N2 us
C 0 2 EN REACTEUR A LIT FIXE D’UNE
CAPACITE DE 25glbatch

Avant-propos

Auteurs et affiliation :
G. Pilon : etudiant au doctorat, Universite de Sherbrooke, Faculte de genie, Departement
de genie chimique et de genie biotechnologique.
J.-M. Lavoie : professeur, Universite de Sherbrooke, Faculte de genie, Departement de
genie chimique et de genie biotechnologique.
Date de soumission : 19 septembre 2012.
Date de publication : 30 novembre 2012.
Revue : ACS - Sustainable Chemistry and Engineering
Reference : [Pilon et Lavoie, 2013a]
Titre fran^ais : Pyrolyse de panic erige (Panicum virgatum L.) a basses temperatures dans
des environnements de N 2 ou de CO 2 ; etude du bilan des produits
Contribution de Particle a la these :
Cet article contribue a la these en explorant l’effet du CO2 lors de la pyrolyse et son
impact sur le biocharbon et sur les co-produits pyrolytiques. Dans cette etude, un
nouveau reacteur et banc experimental d ’une capacite superieure (25 g/batch) a ete con?u
et assemble par G. Pilon. Chaque test (replicat) de pyrolyse permet l’obtention de la
quantite necessaire de biocharbon pour les tests de caracterisation vises. De plus, l’usage
d’un reacteur de plus grande capacite permet un suivi accru du bilan de l’ensemble des
co-produits de la pyrolyse (charbon de biomasse, bio-huile, gaz non condensables), tout

99
 en permettant l’echantillonnage et l’analyse de chacune de ces fractions. Le taux de
chauffe du reacteur peut etre controle a des taux moderes. L’effet du CO 2 est etudie par
comparaison a l’usage de N 2 , aux temperatures de 300, 400 et 500 °C et pour un taux de
chauffe de 55 °C/min. Le suivi accru de l’ensemble des co-produits de pyrolyse permet
d ’investiguer l’effet du CO 2 sur le biocharbon produit a ces conditions d ’operation, et par
le fait meme, de lier de possibles observations a l’ensemble des reactions de pyrolyse
pouvant s’effectuer dans le reacteur.
Resume :
Le panic erige {Panicum virgatum L.) fut pyrolyse dans des environnements de N 2 ou de
CO 2 afin de verifier si le CO2 pouvait avoir un effet sur les fractions de pyrolyse
produites a faibles temperatures. Le panic erige fut pyrolyse a 300, 400 et 500 °C pour
2.5 min dans un reacteur a lit fixe d ’une capacite de 25 g/batch. Les biocharbons, les
liquides et les gaz non condensables ont ete suivis et caracterises, et les rendements furent
compares. Un rendement significativement plus bas en liquides fut observe a 300 °C en
atmosphere de CO2 en comparaison a N 2 (18,0 vs 24,6 % en C 0 2, P<0,002), une
difference qui fut egalement refletee dans le contenu en composes volatils plus eleve dans
les biocharbons a 300 °C - CO2 et dans la composition elementaire (C,H et O; exprime
sur une base de panic non transforme) plus elevee chez les produits solides et plus basse
chez les liquides (CO2 v.v N 2 ). Le CO 2 a aussi demontre un effet a 500 °C, conditions sous
lesquelles le contenu en liquides baissa et le contenu en gaz s’est accru de maniere
significative en comparaison avec l’environnement d’azote. A 400 °C -C 0 2 , les
concentrations en CO furent plus basses et plus elevees en CO2 comparativement aux
niveaux a 400 °C-N 2 , alors qu’a 500 °C -C 0 2 , les concentrations en CO furent plus
elevees et plus basses en CO 2 et CH4 comparativement aux niveaux a 500 °C -N 2 . A 500
°C, une reduction significative du contenu en cendres des biocharbons (par rapport a la
biomasse initiate, P<0,06) fut observee, et ce, seulement dans l’environnement de C 0 2.
Des tests additionnels furent effectues suivant les tests reguliers a 500 °C, avec l’ajout
d’un traitement a fa ir avant la sortie du reacteur. Le traitement mena en une surface
specifique des biocharbons augmentee d’environ 20 fois; tant pour l’environnement
initial de CO 2 que pour celui de N 2 .
Mots-cles : Pyrolyse, torrefaction, CO2 , bilans, proprietes physico-chimiques

100
 Pyrolysis of switchgrass (Panicum virgatum L.) at low temperatures within
N2 and C 0 2 environments; products yields study

Guillaume Pilon and Jean-Michel Lavoie

Abstract
Bioproducts and biofuels production from pyrolysis might benefit the recycling of C 0 2 within
the process instead o f using a conventional inert gas such as N2. In this study, switchgrass
(Panicum virgatum L.) was pyrolyzed within C 0 2 and N 2 environments at 300, 400 and 500
°C for 2.5 min in a lab-scale fixed-bed reactor of 25 g/batch capacity. Chars, liquids and non
condensable gases were monitored, characterized and yields were compared. A significant
lower liquids yield was observed in 300 °C - C 0 2 environments in comparison to 300 °C - N 2
environments (18.0 vs 24.6 % in C 0 2, P<0.002), which was also reflected by a higher char
volatile content under 300 °C - C 0 2 and within elemental composition (C,H and O; expressed
on an original feedstock basis) higher for the solids and lower for the liquid products. Carbon
dioxide also showed an effect at 500 °C, where liquids and gases yields significantly
decreased and increased, respectively. At 400 °C -C 0 2, CO concentrations were lower and
C 0 2 was higher compared to the levels at 400 °C -N 2, while at 500 ° C -C 0 2, CO
concentrations were higher, C 0 2 was lower, and CH4 was lower compared to levels at 500
°C -N 2. At 500 °C, a significant ash content reduction compared to the levels found in the
original biomass was observed only in the C 0 2 environment (P<0.06). Additional tests were
conducted by following regular 500 °C tests with a char air quench. This resulted in a char
specific surface increase of about 20 times; C 0 2 and N 2 initial environments confounded.

Keywords: Pyrolysis, torrefaction, carbon dioxide, yields, physicochemical properties

Note : A la suite des corrections demandees par les membres du jury, le contenu de cet article
differe de celui qui a ete accepte.

101
5.1 Introduction
Plant biomass represents an important source for sustainable energy as well as a neat feedstock
for bio-products and green chemistry development. Biomass use also represents the potential
to reduce CO2 emissions in comparison to petroleum-based products since the balance
between CO2 emissions during combustion or biodegradation o f bio-based products and CO 2
capture during photosynthesis for biomass growth is theoretically neutral. On the other hand,
in order to reach this neutral CO2 balance, the biomass value addition chain must be overall
well implemented. Cultivation (for agro-energy crops), harvesting and transportation may
themselves represent a major cost, energy needs and CO2 emissions sources along the biomass
value addition chain. Cherubini and Jungmeier [2010] did a life cycle analysis (LCA) on a
biorefinery based on switchgrass for bioethanol, bioenergy and biochemicals production and
showed that, in comparison to petroleum products, the biorefmery products could greatly
reduce greenhouse gases emissions (GHG) and non-renewable energy needs. Along their
study, they pointed out that their results were mostly influenced by switchgrass production
issues, especially those related to pelletizing, drying, transportation and manufacturing of N-
based fertilizers. The use of torrefied biomass and biochar at the beginning o f the chain shows
great potential in terms of energy use efficiency and GHG reduction. Compared to raw
biomass, torrefied biomass (char) may benefit the energy balance and GHG emissions
reduction during transportation, pelletization, and grinding [Bergman and Kiel, 2005; Uslu et
al., 2008]. Biochar for soil amendment has the potential to increase plant yield, being
responsible for N-retention through adsorption (then potentially reducing N-based fertilizer
needs), while consisting in a carbon sequestration method due to its very slow decomposition
rate in soils [Lehmann, 2007; Major et a l, 2009]. In addition to these benefits, a rising hope is
whether it would be possible to recycle residual CO2 within the torrefaction or pyrolysis
process instead of using N 2 or other inert (Ch-free) vector gases. Some drying systems as well
as the Oxy-fuel combustion technology already make use of residual CO2 from flue gases and,
consequently, reduce their energy needs by benefitting from gases sensible heat content
[Buhre et al., 2005; Roos, 2008], Recycling CO2 from residual gases into the pyrolysis process
is something that might be economically beneficial [Jindarom et al., 2007b]. However, it can
be expected that modifying the inert environment (N 2 or Ar) to CO2 may have an effect on

102
product composition as previous observations have shown [Duan et al., 2009; Jindarom et al.,
2007b; Zhang et al., 2011],

Use o f CO 2 along gasification has already been studied in order to enhance char conversion in
CO (a targeted syngas constituent) or to enhance char surface area along activated char
production; however, the use of CO2 within torrefaction or pyrolysis conditions (especially
from 300 to 500 °C) has been sparsely reported [Duan et al., 2009; Jindarom et al., 2007a;
Naredi and Pisupati, 2011]. Recently it was reported that pyrolysis studies in a CO 2
atmosphere were lacking [Naredi and Pisupati, 2011], To the best knowledge of the authors,
literature covering pyrolysis in CO 2 environment at lower temperatures (300 to 500 °C) is
really sparse and requires further investigation.

Among the few studies known to be available on the subject, very few o f them used
lignocellulosic biomass at lower temperatures. Zhang et al. [2011] compared CO 2 and N 2
pyrolysis environments within a fluidized bed reactor at 550 °C using corncob. They observed
a slight decrease in condensable gases with slightly higher water content and lower bio-oil
fraction. Char decreased in presence of CO 2 whilst non condensable gases increased. Coking
was observed in presence of CO2 , in comparison to N 2 , on their silica sand bed material. The
same authors observed a decreased production in methoxy-containing compounds and an
increased production in monofunctional phenols in the condensable product fraction in
presence of CO 2 . As for Jindarom et al. [2006], who studied sewage sludge pyrolysis in a CO 2
environment, Zhang et al. [2011] also observed an increase in acetic acid production in
presence o f CO2 , which may be attributed to carboxylation reactions. Butterman and Castaldi
[2010] conducted tests with woody and herbaceous biomass in a CO 2 environment compared
to an environment made of steam mixed with N 2 . Within the 110-450 °C range, their TGA
analyses showed that these two particular gaseous environments (CO2 vs N 2 +steam) lead to
fairly similar results. Encinar et al. [1998] pyrolyzed agricultural residues for activated char
production under a CO2 environment using ZnC^ as chemical additive. For 300, 400 and 500
°C, they obtained specific surface values o f 126.3m2/g, 206.8 m2/g and 290.1 m 2 /g,
respectively. Jindarom et al. [2007a], who conducted tests with sewage sludges in CO 2 vs N 2
environments, observed a higher char specific surface at 550 °C - CO2 conditions (20 vs 15

103
m 2 /g, in CO2 vs N 2 respectively). In a complementary study, Jindarom et al. [2007b] continued
their investigations using TGA, DTG and a batch-type thermal process. They observed from
DTG analysis that under C 0 2 (vs N2), the first pyrolysis reactions were significantly
accelerated (between 275 and 325 °C, results confirmed by TGA weight loss differences) and
the secondary reactions were shifted to lower temperatures (Figure 2.20).

In this study, switchgrass {Panicum virgatum L.) was pyrolyzed within 2 respective gas
environments (C 0 2 and N2) using a fixed bed batch type reactor that was operated at three
temperatures (300, 400 and 500 °C) for a 2.5 min residence time. Two additional tests were
also performed at 500 °C (C 0 2 and N 2 environment, respectively) using a special air quench
following the above process at 500 °C for 2.5 min in order to compare the actual work with a
preliminary study presented in the work of Pilon and Lavoie [201 la].

5.2 Materials and methods

5.2.1 Feedstock

Switchgrass, Cave-in-Rock cultivar {Panicum virgatum L.) was the feedstock utilized for the
experiments. Further details about this switchgrass’s origin and its composition are presented
in the work o f Pilon and Lavoie [201 lb].

5.2.2 Experimental setup description and operation

Thermochemical conversions were conducted in a lab-scale fixed bed reactor. The reactor is
composed of a thin stainless steel tube with a wall thickness of 1 mm, an inner diameter of
83.5 mm and a length of 370 mm. The reactor’s surface is surrounded by electrical band
heaters, which provide an equally distributed heat flux. The gas inflow used for experiments
was either C 0 2 or N2. Gas was first preheated at reactor’s average temperature using an inline
gas heater and, then, it entered the vertically positioned reactor from its bottom (Figure 5.1).
The band heaters were surrounded by insulated shell covers to minimize thermal losses. The
latter were removed once the desired reacting time had been reached (allowing for a faster
temperature drop). A pressure gauge was installed near the gas outlet. The gases produced

104
passed through a heated line maintained at around 250 °C before passing through a series of
four empty glass condensing tubes (38mm o.d. by 23 cm long) immersed into ice baths (water
and ice), all connected by stainless steel tubing equipped with a filter (tube filled with cotton
wool pieces) (Figure 5.1). A portion o f the non condensable gases passed through a molecular
sieve (Aldrich 3A; 1.6 mm pellets) before being collected in 25L Tedlar® gas sampling bags
(Supelco). Gas sampling was done throughout the experiment at a constant flow rate starting
from a reference reactor temperature o f 100 °C. Temperature was monitored using K-type
thermocouples installed at various locations: after the gas inlet (T-0), inside the reactor (T-l to
T-4) at the reactors surface, at its outlet (T-5) and before the condensers (T-6 ) (Figure 5.1).
Real time temperature monitoring was done on the overall system using the National
Instrument Data Acquisition System NI cDAQ-9172 equipped with Labview software.

Before each test, switchgrass was cut into <10 cm pieces and about 25 g o f feedstock was
utilized per test. Feedstock humidity levels were evaluated prior to each test; levels that
typically ranged from 4 to 9%. The gas inflow rate was set to 0.5 L/s (STP: 101.3kPa - and 25
°C) and the heating rate was set to 55 °C/min by controlling the inline preheater as well as the
reactor’s band heaters. The reactor’s operating absolute pressure corresponded to about 115
kPa, which is slightly higher than the atmospheric pressure resulting from friction along the
gases outlet system (Figure 5.1). Tests were conducted at 300, 400 and 500 °C for two distinct
gas environments (CO2 or N 2 ). Every tested condition was conducted in triplicates. Statistical
analyses of result averages comparison was expressed with the P value. The P value
corresponds to the smallest level o f significance that lead to the rejection o f the null
hypothesis (where averages compared are equal).

For two additional tests, one at 500 °C - N2 and the other one at 500 °C - CO2 , the tests were
first conducted normally (as described above) except that, after completing the normal test
period o f 2.5 min residence time, air replaced N 2 or CO 2 (from the moment the reactor reached
400 °C during the cooling phase), at the same flowrate (0.5 L/s), for 2 minutes. After air
treatment, the reactor was left to cool as for the previous tests. These 2 tests were carried in
order to compare the new results with those derived from a previous experiment where the
products became in contact with ambient air after pyrolysis [Pilon and Lavoie, 2011a]. Since

105
these tests were conducted especially to support BET measurements, only char yields and BET
values were determined (see section 5.3.3).

5.2.3 Feedstock and char characterization

Analytical methods for proximate analysis, organic elemental analysis, BET surface area as
well as for cellulose, hemicellulose, and lignin content, on the feedstock and, to a lower extent,
on chars (Table 5.1 and Table 5.2) were all reported in Pilon and Lavoie [201 lb]. For further
analyses, the elemental analysis results were transferred on an original biomass basis using,
for each replicate, its respective pyrolysis fraction yield.

5.2.4 Condensable products preparation and analyses

Condensable products were first analyzed by gravimetry using the content collected within the
condensing tubes, the connecting tubings and the filter cotton wool. After each experiment,
each condenser and their lines were thoroughly washed and rinsed with water followed by
acetone and the cotton wool in the filter was replaced. Only the condensable products from the
condensing tubes (not from the cotton wool) were collected for further analyses. Their content
collection was completed using acetone after which the solution was filtered through glass
frits to remove any suspension particles. Before analyses, acetone was removed from the
condensable product samples in a rotary evaporator at a temperature of 65 °C and a vacuum
pressure of 15 mmHg (50 kPa).

The condensable products’ organic elemental composition was analysed using the same
methods used for char and feedstock and described in the previous section (section 5.2.3).
Water content of the bio-oils was analyzed using Mettler Toledo DL31 Karl-Fisher Titrator
and anhydrous ethanol reagent.

5.2.5 Gas analyses

Non condensable gas analyses were conducted with a gas chromatograph (GC; Varian
C P3800) using a thermal conductivity detector (TCD ) for CO2, N 2 , CO and a flame ionization
detector (FID ) for CH4, C2H4 and C2H6. For G C -T C D analyses, columns used were Hayesep T

106
(CP81072), followed by Hayesep Q (CP81073) and Molecular Sieve 13X (CP81071), using
He as carrier gas. The FID was installed in series with the TCD. Gas analyses were conducted
in duplicates. For gas result analyses, CO2 from the vector gas was considered as inert,
therefore, the inflow volume was removed within the calculations. Hydrogen could not be
analyzed due to its low abundance, overall interfering with the injection valve peak.

R eactor Bio-oil
A nalyses

^ 4 n .:n ;n :n
% j ■ Filter
< N C1 C2 C3 C4

Bio-oil G as
C o n d e n se rs Volume
M eter

G as
E xhaust
M oisture -
A bsorber

Flow M eter
G as ^ G as
P re h e a te r
OO Sam pling
OO
Flow m eter Char

Figure 5.1: Schematic representation o f the experimental setup (not at scale).

5.3 Results and discussion

5.3.1 Reactor operation

The reactor was operated under conditions permitting to reach an average o f the set point
temperature within its core. Results for the oil condensing train as well as for bio-oil collection
efficiency are presented in Figure 5.2. Interestingly, despite the fact each condensing tube had
its own cooling bath (Annexe B, Figure B.4), about third o f the liquid obtained was from the
cotton wool placed following the condensing train. These results are similar to those of
Boateng et al. [2007] who collected their pyrolysis condensable products using a condensing
train of 4 tubes followed by an electrostatic precipitator (ESP), which collected an additional
107
42.7% o f liquids. In their case, it appeared that most o f the water present in the oil was
condensed within the four condensers and the higher molecular compounds resulting from
lignin decomposition were collected by the ESP [Boateng et al., 2007]. Since only the liquids
from the condensing tubes were considered for further analysis during the present study, it is
possible, therefore, that higher molecular weight compounds were not in the sample.

4.0

□ 300*C - N j

□ 300'C - C O 2

■ 400"C - N j

Q 4 0 0 “C - C O ,

D SOO'C - N ;

■ 500"C - CO?

T ubel T ube 2 T ube 3 T ube4 Tubing 1 T ubing 2 Tubing 3 T u b in g 4 C o tto n

Filter
Liquid products collecting sections

Figure 5.2: Condensable products amounts collected along sections of the

condensing train. Error bars correspond to standard deviation.

5.3.2 Products yields analysis

Among the products recovered (Figure 5.3), the first most significant result observed was the
bio-oil yield at 300 °C, which was significantly lower (P<0.002) in presence o f CO 2 (18.0 vs
24.6 % in CO2 vs N 2 respectively). From Figure 5.3, for 300 °C, it is also possible to observe
chars reaching higher yields in presence of CO2 than with N2 (59.2 vs 54.4% respectively),
therefore, less products possibly converted into liquid products. No major non condensable gas
yield variation with respect to the gas environment was observed although yields under CO 2
always appeared slightly higher (Figure 5.3).

108
These primary observations may provide an overview of the reactions that were influenced by
CO2 as vector gas during pyrolysis at 300 °C. Liquid condensable products are the result of
depolymerisation and fragmentation reactions o f the original feedstock’s polymers
(holocelluloses and lignin) leading to products that may tend to recombine and be condensed
as liquids [Butterman and Castaldi, 2010; Mohan et al., 2006], Therefore, a first explanation is
that CO 2 could affect the decomposition reactions or volatile products recombination.
Analyses on the 300 °C chars showed that char volatile content was higher in presence o f CO 2
than in presence of N 2 (61.0 vs 57.1%, corresponding to 0.35± 0.01 vs 0.29±0.05 g/g biomw.b.
for P=0.1, respectively) (Table 5.1). These results are consistent with char elemental
composition where carbon, hydrogen and oxygen contents were higher in presence o f CO 2
compared to N 2 (Figure 5.4 a and b). In the case o f liquid products at 300 °C, C, H and O were
found in lower concentrations in samples produced under CO2 and the results were statistically
significant (P< 0.1, 0.002 and 0.02, respectively) (Figure 5.5 a and b). Since the conversion of
char material was not as complete in CO 2 as in N 2 (for 300 °C), these results are consistent
with the assumption that a CO 2 environment affects reactions such as decomposition. During
torrefaction (up to 300 °C), important reactions taking place are dehydration and
decarboxylation [Prins et al., 2006b], On the other hand, because the gas yield (non
condensable gases) was slightly higher in presence o f CO2 (Figure 5.3), this means the simple
compounds of low molecular weight that may result from these decomposition reactions are
not much affected or, on the contrary, are slightly enhanced. In addition, there was no
difference in the gas composition at 300 °C (C 0 2 vs N2) while CO 2 gas composition content
was quite high (85-87%; Table 5.1), which is consistent with decarboxylation known to occur
at 300 °C [Prins et al., 2006b]. As a result, the fact that only the bio-oil, but not the non-
condensable gases were affected at 300 °C, could mean that either the cleavage o f the original
feedstock polymers into heavier molecular weight fragments (forming bio-oil) or the
recombination o f small molecular weight fragments into heavier compounds could be
impaired by a CO2 environment. This behaviour appears to occur especially at 300 °C since
bio-oil yield difference between the two gas environments decreased with increasing
operating temperature (Figure 5.3).

109
The feedstock polymeric constituents that are known to decompose around 300 °C are mostly
related to hemicellulose and lignin [Prins et al., 2006a; Rousset et al., 2011; Sharma et al.,
2004], In addition, cellulose, which is usually known to decompose starting from about 325
°C, was shown to be more vulnerable to temperature in presence of minerals [Yang et al.,
2006]. Since xylans are abundant polymers within hemicelluloses and since they especially
tend to decompose at torrefaction temperatures [Repellin et al., 2010], it could be interesting
to verily whether or not CO2 has an effect on these specific polymers. In addition, Prins et al.
[2006b] observed that deciduous trees and straw containing more xylans and minerals
produced more acetic acid; moreover, acetic acid was reported to be favoured during CO 2
pyrolysis [Jindarom et al., 2006; Zhang et al., 2011]. As a result, this would all be coherent.
As reported by Amonette and Joseph [2009], lignocellulosic degradation in the vicinity o f 300
°C is believed to involve free radical processes initiated by the thermal action of inorganic
impurities present in the feedstock as well as the heat effect on structural O. As a result, CO2
could perhaps affect the inorganic impurities, which would subsequently affect the
condensable products formation and content. Some observations on the effect of CO2 on the
inorganic content were made under 500 °C conditions and are presented in the following
sections.

At a temperature of 300 °C, very few studies are available on the impact o f a C 0 2
environment. Jindarom et al. [2007b] had observed a difference at 300 °C based on the gas
environment, through DTG analysis; however, their results indicated an enhanced mass loss
rate in a CO 2 environment, which appears contradictory to the results of the present study.
However, this contradictory result could be due to the fact that their feedstock was sewage
sludge, which is fairly different from switchgrass in terms of mineral content, moisture
content, and hemicellulose content and composition; to note only these three differences in
feedstock characteristics [Jouraiphy et al., 2005; Pettersson et al., 2008],

Increasing temperature from 300 to 400 °C led to lower char yields and favoured an increase
in liquids as well as non condensable gases (Figure 5.3). Gas composition, with respect to
temperature only, showed a decrease in CO 2 concentrations at 400 °C in comparison to 300 °C
while the CO content increased (both in N 2 and CO 2 environments) (Table 5.1). This may

110
result from oxygen content (trapped in biomass) reacting with carbon. However, oxygen being
limited within the biomass, incomplete combustion reaction (reaction #3; Table 5.3) may be
favoured over complete carbon oxidation (C + O 2 —>CO 2 ). As O radicals are formed, carboxyl
and carbonyl groups are cleaved to yield CO2 and CO [Amonette and Joseph, 2009]. Another
explanation comes from Butterman and Castaldi [2010] who suggested that OH, which usually
reacts with CO to form CO 2 (reaction #7; Table 5.3), would not be produced from lignin at
400 °C (requiring higher temperatures for its decomposition) and this would then lead to
increase CO concentrations.

Regarding again results obtained at 400 °C, comparing N 2 and CO 2 atmospheres, no

statistically significant difference was noted on the conversion yields of the original feedstock
into the three main fractions: char, liquid and gas fractions (Figure 5.3). The fraction where a
difference (with respect to the gas atmosphere used) may be noted is in the non condensable
gases one. The gas composition in CO 2 environments showed higher CO 2 concentrations
while CO, C2H4 and C2 H6 all decreased compared to a N2 atmosphere (Table 5.1). In order to
explain these gas composition differences, it is first possible to observe that the oxygen
elemental content, with respect to the char and liquid fractions, was not significantly affected
by the gas environment (Figure 5.4a and Figure 5.5a). As a result, CO lower concentration in
CO 2 environment could be enhanced by the Boudouard reaction or reversed steam-reforming
reaction (reactions #1&2; Table 5.3). Zhang et al. [2011] observed that, under a CO 2
atmosphere (at 550 °C), coking was occurring on their silica sand bed which was not observed
under N 2 . Since this reaction (Boudouard, reaction #1; Table 5.3) is thermodynamically
favoured at lower temperature, a CO2 environment could affect indirectly the reaction thus
enhancing carbon deposition and increasing carbon dioxide formation. As for the C2H4 and
C2 H6 concentrations obtained in CO2 environments versus N 2 (0.12 and 0.12 vs 0.26 and
0.24% for C2H4 and C2H6, respectively; Table 5.1), this might be consistent with the previous
observations at 300 °C where CO2 could have affected some decomposition reactions. The
reaction that could have been impaired in this case is demethylation leading to subsequent
formation of these compounds (e.g.: C2 H4 and C2 H6) through combination o f radicals.
Demethylation is known to decompose biomass constituents between 350 °C to 450 °C and
especially phenol methoxy groups from lignin [Butterman and Castaldi, 2010; Mohan et al.,

Ill
2006]. This could be consistent with the previous assumption where CO2 would have induced
a reduction o f the depolymerisation process, whereas lignin usually leads to high molecular
weight compounds during pyrolysis and where CH3 radicals could be exposed following
depolymerisation. A lower depolymerisation of lignin would have reduced the amount of
methyl groups being exposed thus, lowering ethylene and ethane formation.

At 500 °C, feedstock conversion was further enhanced in comparison to 400 °C conditions;
char yields passing from about 36 to 28% (400 vs 500 °C respectively, N 2 and CO 2
environments confounded; Table 5.1). Glancing at the 500 °C experiments, with respect to gas
environment, it appeared that char yield was not affected, but the condensable gases yields
were significantly lower (P<0.05) and non condensable gases higher (P<0.1). The condensable
gases yield results, showing a diminution in presence of CO 2 vs N 2 are consistent with results
obtained by Zhang et al. [2011] with com cob. As for gas composition, the CO 2 environment
appears to have enhanced the formation of CO. Therefore, the reversed Boudouard reaction
(reaction #1 inversed; Table 5.3) might have taken place, enhancing formation o f CO instead
of bio-oil. Despite the fact that thermodynamics does not favour the reversed Boudouard
reaction below 704 °C (C + CO2 —> 2 CO; reversed reaction #1, Table 5.3), but mostly the
forward reaction; previous observations showed that in presence o f Ni or Fe (e.g.: from the
biomass in close contact with the stainless steel material), the reversed Boudouard reaction
may be catalyzed, thus, allowing the reaction at lower temperatures [Osaki and Mori, 2006
and Jin et al. 2012]. Despite the fact that char yield did not vary with respect to gas
environment, 500 °C - C 0 2 char’s volatile content was reduced and this translated into an
almost significant higher fixed carbon content (Table 5.1). As a result, this could mean that the
presence o f CO 2 enhances the formation of gases from the volatile fraction o f the feedstock,
but not from the most stable carbon structure. A behaviour that would be consistent with the
use of CO2 for activated char through the removal o f tar within its pores. In fact, at 500 °C, in
a CO2 atmosphere compared to a N 2 atmosphere, bio-oil showed a lower elemental carbon
content while char carbon content remained essentially stable (Figure 5.5a and Figure 5.4a
respectively). The carbon dioxide content was also lower in the C 0 2 environment than in N2,
which would be consistent with the possible enhancement o f the reversed Boudouard reaction.
Based on a mechanism reported in Butterman and Castaldi [2010], another explanation for the

112
higher CO concentration (at 500 °C under CO 2 ) could be the decrease in OH radical branching
off from lignin and cellulose which usually tends to combine with CO (reaction #7; Table 5.3)
producing CO2 and H radicals. Since it was hypothesized from our observations in the course
o f this study that a C 0 2 environment could affect production of heavier compounds through
depolymerisation, the “OH branching o ff’ could then have been reduced, consequently
avoiding CO reduction through reaction 7 (Table 5.3).

It may be observed from the 500 °C reaction that the formation o f CH4 was enhanced
compared to 300 and 400 °C (Table 5.1). CH 4 may be formed from hydrogenation-
methanation reactions (reactions #4, 5, 6 ; Table 5.3) as well as from H and CH3 radicals
formed above 250 °C and combining into CH4 [Butterman and Castaldi, 2010], This is
consistent with the fact that the hydrogen content was greatly reduced in chars when
comparing with the reactions performed at 300 and 400 °C respectively (Figure 5.4b). Results
showing a higher CH4 yield around 500 °C as well as a lower CH4 yield in presence of CO2
than in N 2 (500 °C) are consistent with the ones from Butterman and Castaldi [2010]. They are
also consistent with the explanation given for the 400 °C results concerning demethylation
reactions which could have decreased due to lower depolymerisation o f biomass constituents
such as lignin (at these temperatures) under a CO2 environment. Another explanation for the
lower CH4 content in a CO2 environment at 500 °C would be the enhancement of CO2
reforming of CH4 reaction (CH4 + CO 2 —►2CO + 2 H 2 ) which could occur to some extent at
500 °C [Nikoo and Amin, 2011]. This latter reaction could also have contributed to the CO
enhancement observed under the CO 2 atmosphere (Table 5.1).

The hydrogen content in bio-oils appeared lower in presence o f CO 2 than with N 2 ; mostly at
300 °C (Figure 5.5b). H 2 was not specifically analyzed, however it is known to appear
essentially starting at 600 °C [Butterman and Castaldi 2010]. It may also be observed that the
water content in the bio-oils was significantly higher at 500 °C than under most conditions at
300 and 400 °C. This may result from 500 °C temperature range enhancing combustion
reaction of H2 with O2 (from biomass).

113
Still at 500 °C, the char ash content was significantly lower under a CO 2 environment
(11.5±0,7 vs 13.5±0.6%, P<0.02, corresponding to 31.2±2.6 vs 36.4±2.6 mg/g biomass d.b.,
P<0.06; CO2 vs N 2 respectively; Table 5.1). Such an ash content difference was not observed
with the other temperatures and conditions studied here. Since a CO2 effect was noted with all
conditions studied, investigations on mineral interactions with respect to a CO2 atmosphere
and with pyrolysis reactions mechanisms would be of great interest for further studies. As
reported in Fryda et al. [2010], the fate of minerals (e.g.: lowering ash melting temperature
and vaporization) may be affected by reaction conditions. Therefore, since CO 2 was observed
to have an effect on ash content at 500 °C, the variations noted with respect to the gas
environment (in CO 2 ) could, then, be the result of interactions between CO 2 and minerals
present in the biomass. Frequently it was observed and reported that the inherent biomass
inorganic content had an effect on pyrolysis mechanisms and often a catalytic effect during
biomass pyrolysis [Butterman and Castaldi, 2010; Jindarom et al., 2007b; Mohan et al., 2006;
Prins et al., 2006b]. Overall, that could mean that CO2 simply inhibits the catalytic action of
some biomass inorganic material (especially the metallics; see typical switchgrass composition
in Annexe D). Further analyses of char inorganic content such as XRD might be used to verify
reaction o f CO2 with inorganics and corroborate the assumption. Noteworthy during the
pyrolysis experiments, CO2 seemed to affect the stainless steel material exposed to the
reaction by modifying the colour (multiple-colors appearing) as well as having a different
smell under CO2 compared to N 2 when opening the reactor. This may be in agreement with
CO2 affecting (possibly reacting with) some inorganic components such as the alloying metals
in stainless steel. As a result, it could be o f great interest to investigate more specifically the
effect o f Fe, Ni or other stainless steel constituents on biomass pyrolysis under CO2 .

114
 70

Products yields [%]

C3 C4 C5 L3 L4 L5 G3 G4 G5

Figure 5.3: Pyrolysis products yields. Error bars correspond to standard

deviation. Legend: {(C, L or G), (3, 4 or 5), (N or C)} standing for
{(Char, liquid or non condensable gases), (300,400 or 500 °C), and
(N2 or CO2 )} respectively; applicable to figures 5.4a to 5.5b included.

115
 Content [g/g biom.] (a)

■ C

■0

RSG C3N C3C C4N C4C C5N C5C

Samples

0 .0 6

(b )
0 .0 5
Content [g/g biom.J

0 .0 4

1H

0.02 ■ N

0.01

0
RSG C3N C3C C4N C4C C5N C5C
Samples

Figure 5.4: Char elemental organic content expressed on original feedstock basis
(wet basis): (a) carbon and oxygen content; (b) hydrogen and
nitrogen content. Error bars correspond to standard deviation.
Legend: RSG: Raw switchgrass

116
 0.3
(a)
U.25

0.2

§ 0 .1 5
■c

r
<u
c 0.1 ■o

0 .0 5

L3N L3C L4N L4C L5N L5C

S a m p les

0 .0 1 8

0 .0 1 6 (b)
0 .0 1 4

| 0.012
oo 0.01

t i 0 .0 0 8 ■ H

0 .0 0 6 ■ N

0 .0 0 4

0.002

L3N L3C L4N L4C L5N L5C

Samples
Figure 5.5: Liquid elemental organic content expressed on original feedstock
basis (wet basis): (a) carbon and oxygen content; (b) hydrogen and
nitrogen content. Error bars correspond to standard deviation.

117
Table 5.1: Switchgrass pyrolysis products characterization

Conditions Solids Liquids Gases

yo Gas Yields Ash Volatiles FC S.A. Yields Water Yields CO; CO ch4 C;H, C;H6
|°C]
300
Env.
N,
!%]'
54.4
i%r
7.3±1.0
i%r
57.140.6
i% r
35.641.2
m2/g
1.3
i% i
24.6
[%l
26.645.3
l%l
21.0
|% |2 1%I2 |% |2 |% |2
84.8 14.8 0.3 0.06
|% ]2
0.04
300 CO; 59.2 6.340.5 61.043.9 32.843.5 1.4 18.0 23.5412.9 22.7 86.8 12.8 0.3 0.05 0.05
400 N; 36.7 10.8±0.7 35.842.8 53.543.4 1.1 35.6 25.545.3 27.7 64.1 34.6 0.9 0.26 0.24
400 CO; 35.9 11.5±0.3 36.542.1 52.041.9 1.1 33.7 27.643.6 30.4 74.5 24.5 0.7 0.12 0.12
500 N, 28.2 13.540.6 28.341.9 58.341.7 9.3 37.7 39.042.2 34.1 58.7 35.4 4.1 0.97 0.86
500 CO; 28.1 _ 11.540.7
_ _ _ 24.843.4 63.743.1
....
9.2 36.0 36.742.6 35.9 52.9 42.8 2.2 0.78 1.39
“± ” corresponds to standard deviation.1: Dry char basis. Volume basis. Legend: T°: Temperature, Env.: environment, FC: Fixed carbon, S.A.: surface area.

Table 5.2: Switchgrass characterization

Characteristics Units Content
C % 44.5
H % 5.8
O % 45.7
N % 0.5

Ash %' 3.7

Volatiles %' 81.0
Fixed carbon %' 15.3

Moisture content %‘ 4-9

a-cellulose % 33.7
hemicellulose % 36.7
Lignin % 17.3

Surface area m2/g <1

HHV MJ/kg 19.5

Dry basis

118
 Table 5.3: Reactions possibly occurring under pyrolysis conditions
AH AH T
# Reaction (300 °C) (500 °C) (AG < 0)*
kJ/m ol kJ/mol °C
1 2CO — C + C 0 2 -174 -172 <704
2 CO + h 2 -> c + h 2o -134 -135 <672
3 C + V2O2 - CO -110 -111
4 2 C 0 + 2H2 -» C H 4+ C 0 2 -256 -259 <640
5 CO + 3H2 ►CH4 + H20 -217 -222 <612
6 C + 2H2 — CH4 -83 -87 <532
7 CO + OH - C 0 2 + H -103 -101 **
♦Obtained from FactSage software (P = 1 atm) ♦♦For values from 0 to 1200 °C

5.3.3 SPECIFIC SURFACE AREA DEVELOPMENT

It may be observed that the char’s surface area slightly increased at 500 °C (Table 5.1). It went
from <1 m2/g for the original feedstock to about 1 m2/g for 300 and 400 °C chars and to 9
m2/g for 500 °C chars (for both gaseous environments) (Table 5.1). These results for surface
area with respect to temperature are consistent with other studies on switchgrass pyrolysis,
such as those of Boateng et al. [2007], who obtained 7.8 m2/g with a fluidized bed pyrolysis
reactor (at 480 °C). Nonetheless, Brewer et al. [2009] and Pilon and Lavoie [201 lb] obtained
50.2 and 6 6 .6 m 2 /g, respectively, for similar reacting conditions (at 500 °C). With respect to
the gas environment, especially for higher temperatures, a higher surface area could have been
expected due to the possible reversed Boudouard reaction (possibly catalyzed by feedstock
minerals) enhancing carbon conversion and emptying pores. Butterman and Castaldi [2010]
have observed higher pore formation behaviour in presence of C 0 2, although the chars tested
were derived from higher temperature treatments (860 °C), which is known to enhance surface
area [McLaughlin et al., 2009], Since a lower volatile content was obtained at 500 °C - C 0 2
(Table 5.1), increased emptying of the pores would be expected, thus increasing specific
surface, which did not occur in the current study.

As reported in a preliminary study on the subject [Pilon and Lavoie, 2011a], in presence o f
C 0 2, a higher specific surface had been observed (189 vs 82 m2/g in a C 0 2 or N 2 environment,
respectively; Table 5.4). The possible reasons for this important difference may be related to

119
the differences between the two reactors design and operations which implies: a lower
superficial gas velocity in this study (0.24 vs 0.54 m/s; evaluated at 500 °C, at 1 atm), the
amount of material used (25 g vs lg) and there was contact with air after treatment during the
preliminary study o f Pilon and Lavoie [201 la]. In order to verify whether the gas environment
(N 2 or CO 2 ), followed by contact with air had an effect on char porosity, two additional tests
were conducted specifically for the 500 °C conditions, at which a high surface area had been
observed in the preliminary study. The experiments gave chars with similar surface areas of
183 and 198 m2/g for N 2 + air and CO2 + air environments, respectively (Table 5.4). In both
cases, the maximal temperature reached was 650 °C (Figure 5.6). As a result, the CO 2
environment used during the heating period and the 500 °C treatment, previous to the oxygen
quenching treatment, did not show an impact on surface area compared to N 2 . Noteworthy that
char yields for these two additional treatments were 15.8 and 15.1% for N 2 -air and C0 2 -air,
respectively (Table 5.4). Despite the fact that CCh-air in comparison to N 2 -air did not show
any major difference, the contact with air after pyrolysis treatment greatly increased the
surface area. This simple technique o f air insertion within the reactor (at low temperature char
production) may have important implications, especially for char plus value when a high
surface area is desired; such as biochar (soil amendment applications). On the other hand, the
technique would gain at being optimized since char yield decreased drastically under the
conditions used (decreased from about 28 to 15% for both gaseous environments; Table 5.4).
In addition, clarifications whether it is the oxidation, the temperature or a correlation between
both factors which had this effect on specific surface enhancement should be undertaken.

Table 5.4: Char surface area after specific treatments

Char yield Surface Area
Conditions Ref.
[%]* m2/g
500 ° C -N 2 28.2 9 1
500 °C - C 02 28.1 9 1
500 °C - N2* 15.8 183 1
500 °C - C 0 2* 15.1 198 1
500 °C - N2* 15.9 82 2
500 °C - C 02* 13.6 189 2
References (ref.): 1: This study; 2: Pilon and Lavoie [2011a] *Chars put in contact
with air after the 0 2-free test.

120
 700

■T 1
GOO

■T 2
500

u" - T3
%
.- /
/ ' /■
400
A '
E
41 T4
f 300 :rT
.4

■if
jif
200 .*'/.....
:>!
100 •>/...
/*

500 1000 1500 2000 2500

Time [s]

Figure 5.6: Example o f reactor temperature profile at 500 °C - N 2 obtained

during the special air treatment following normal test. T1 to T4
represent temperature signals from thermocouples positioned as
indicated in the schematic representation o f the reactor (Figure 5.1).

5.4 Conclusions
Pyrolysis of Panicum virgatum L. at 300 °C using CO2 as a gas vector resulted in significant
reduction of the bio-oil yield, which was consistent with the higher char volatile content under
the same conditions. In addition, in the 500 °C - CO2 environment only, a significant
reduction in ash content was observed. Given the importance o f some inorganic elements
(especially the metallic ones) as reaction catalysts, how CO2 affect them and influences,
consequently, subsequent biomass conversions, represent important areas for further research.

121
Acknowledgements
Authors would like to thank the Faculty o f Engineering and the Department o f chemical
engineering and biotechnological engineering (DCEBE) of the Universite de Sherbrooke
(UdeS) for the financial support to Mr. Guillaume Pilon’s PhD thesis as well as for supporting
this research through starting funds of the corresponding author. In addition, authors would
like to thank the Industrial Chair in Ethanol Cellulosic of the UdeS in lending research
equipment. The authors would also like to acknowledge: Mr. Jasmin Blanchard from the
Groupe de recherches sur les technologies et precedes de conversion (GRTPC) o f UdeS for
gas analyses, Ms. Sonia Blais from the Centre de caracterisation de materiaux of the UdeS,
Mr. Serge Gagnon, Mr. Henri Gauvin and Mr. Osvaldo Serrano from DCEBE for their
technical support and, finally, Mr. Daniel Clement, switchgrass producer (Qc, Canada), for
providing the feedstock used in this research.

122
CHAPITRE 6 ARTICLE 4 - ETUDE DE LA
COMPOSITION CHIMIQUE DES EXTRAITS DE
BIOCHARBONS ET DES BIO-HUILES OBTENUS
DE LA PYROLYSE DU PANIC ERIGE EN
ATMOSPHERES DE C 02 vs N2.

Avant-propos

Auteur et affiliation :
G. Pilon : etudiant au doctorat, Universite de Sherbrooke, Faculte de genie, Departement
de genie chimique et de genie biotechnologique.
J.-M. Lavoie: professeur, Universite de Sherbrooke, Faculte de genie, Departement de
genie chimique et de genie biotechnologique.
Date de soumission : septembre 2012.
Date de publication : 11 fevrier 2013.
Revue : Journal of Analytical and Applied Pyrolysis
Reference : [Pilon and Lavoie 2013b]
Titre fran^ais : Pyrolyse de panic erige (Panicum virgatum L.) a basses temperatures dans
des environnements de N 2 et C 0 2; etude de la composition chimique des lixiviats de
biocharbons et des bio-huiles
Contribution de Particle a la th&se :
Cet article contribue a la these en elaborant davantage sur P extraction des biocharbons
(par Soxhlet et dichloromethane) et continue, par la meme occasion, 1’investigation de
l’effet de l’usage du C 0 2 comme environnement gazeux sur les biocharbons et le procede
en general. L ’experience acquise pendant l’etude du chapitre 3 (article 1) a permis de
noter que l’extraction d’une plus grande quantite de biocharbon permettrait
1’augmentation du signal GC-MS, et, par consequent, permettrait d’augmenter le nombre
de composes observes et identifiables. Dans la presente etude, le reacteur d’une capacite

123
 de 25 g/batch a done ete utilise pour generer une quantite importante de co-produits de
pyrolyse. De plus, dans le cadre de cette etude, 1’analyse des extraits de biocharbon a ete
effectuee de concert avec celle de la bio-huile. Ces deux analyses permettent de clarifier
et d’appuyer des observations faites sur l’usage du CO 2 dans le chapitre precedent et de
decrire davantage les comportements reactionnels du procede.
R esum e:
Le developpement de produits a valeur ajoutee a partir du biocharbon de pyrolyse, tel un
amendement pour le sol ou matiere pour conversion thermochimiques subsequentes,
necessite la connaissance de la composition chimique de ses extraits. Dans le cadre de
cette etude, le panic erige (Panicum virgatum L.) fut pyrolyse en utilisant deux vecteurs
gazeux distincts, soit l’azote et le dioxyde de carbone, a 300, 400 et 500 °C, pour 2.5 min
de temps de residence, dans un reacteur a lit fixe de 25 g/batch et en utilisant un
prechauffage a un taux de 55 °C/min. Les biocharbons furent extraits avec du
dichloromethane a l’aide d ’appareils Soxhlet. Ces extraits de biocharbons ainsi que les
liquides pyrolytiques furent analyses par GC-MS. Par rapport a la temperature, le procede
effectue a 500 °C resulta en des quantites extraites des biocharbons significativement plus
faibles que celles obtenues des biocharbons produits a 300 et 400 °C. Des differences ont
ete observees en fonction de l’environnement gazeux, aux 3 temperatures d’operation.
Parmi ces differences, la pyrolyse effectuee a 300 °C en presence de CO2 (vs N2) mena a
une quantite plus elevee d’extraits de biocharbon (9 vs 6 mg/g Char)- La valeur calorifique
fut aussi evaluee pour les biocharbons et les bio-huiles. II fut possible d ’observer que les
biocharbons produits a 300 °C avaient une valeur calorifique significativement plus faible
en environnement de CO2 plutot qu’en presence de N 2 (22,2 vs 24,5 MJ/kg
respectivement; P<0,05). A 400 °C, le furfural a ete seulement retrouve dans les extraits
de biocharbons obtenus sous 1 ’environnement de N 2 , alors que dans les bio-huiles, son
contenu relatif fut significativement plus eleve sous N 2 vs CO 2 . Les extraits de
biocharbons ainsi que les bio-huiles ont demontre des contenus marques en phenols,
cyclopentanones et composes derives du furane.

Mots-cle : Pyrolyse, torrefaction, biocharbon, extractions Soxhlet, dichloromethane, bio-

huile, GC-MS

124
 Pyrolysis of switchgrass (Panicum virgatum L.) at low temperatures in N2
and C 0 2 environments: a study on chemical composition of chars extracts
and bio-oils

Guillaume Pilon and Jean-Michel Lavoie

Abstract
With the increasing value addition applications of biomass pyrolysis char such as a soil
amendment or feedstock for subsequent thermochemical conversions, chemical composition
of char extracts becomes a necessity. Along this study, switchgrass (Panicum virgatum L.)
was pyrolyzed using two distinct gas vectors, N2 vs C 0 2, at 300, 400 and 500 °C for 2.5 min in
a 25 g/batch fixed bed reactor using a 55 °C/min preheating stage temperature ramp. Chars
were extracted with dichloromethane using Soxhlet, and these mixtures, in common with
pyrolytic liquids were analyzed by GC-MS. With respect to temperature, the process
performed at 500 °C resulted in significantly lesser amounts o f char extracts as compared to
the ones produced at 300 and 400 °C. Differences were observed with respect to gas
environments at the 3 operating temperatures. Among the differences, pyrolysis performed at
300 °C under C 0 2 (vs N2) led to a higher char extracts content (9 vs 6 mg/g Char)- Calorific
value was also determined for chars and bio-oils. Chars obtained at 300 °C had a calorific
value significantly lower when produced in a C 0 2 environment as compared to a N2
environment (22.2 vs 24.5 MJ/kg respectively; P<0.05). At 400 °C, furfural was only found in
char extracts from N2 gas inflow environment, while in bio-oil, it was also found at
significantly higher relative content under N2 vs C 0 2. The char extracts as well as the
condensable chemical fractions have demonstrated predominant levels o f phenols,
cyclopentanones and furan-derived compounds.

Keywords: Pyrolysis, char, Soxhlet extractions, dichloromethane, bio-oil, GC-MS

Note : A la suite des corrections demandees par les membres du jury, le contenu de cet article
differe de celui qui a ete accepte.

125
6.1 Introduction
Pyrolysis o f lignocellulosic biomass is a process by which the feedstock’s macromolecular
compounds are fragmented to smaller molecules by the effect o f heat within an anoxic or
almost anoxic environment. From the three main pyrolysis fractions, char, gases and liquids,
the latter is the fraction that has been characterized more thoroughly due to its multiple
possible applications, for the variety o f compounds it contains as well as for its energy density.
The solid residue known as the char, generated after a certain loss of volatiles, is also getting
increased consideration due to the numerous potential applications such as its use as an
improved feedstock for combustion or gasification. It also generates increasing attention as a
precursor for activated carbon and as a soil amendment, either for CO2 sequestration, or as a
crop growth enhancer or for adsorption of organic and inorganic compounds for
environmental purposes (e.g.: N-fertilizer leaching control) [Lehmann, 2007; Major et al.,
2009]. Char, depending on the reacting conditions and the biomass used, as well as on pre-
and post- treatments, may vary chemically and physically [Downie et al., 2009]. Very few
analyses have explored the chemical composition o f char extracts synergistically with the
condensable gases portion. In accordance to Krull et al. [2009], the solvent-extractable
components from char have rarely been investigated and they could provide valuable insight
for char application as a soil amendment, especially in terms of char-soil and/or char-plant
interactions, without making abstraction of the environmental point o f view regarding soil
amendment regulations. Char chemical composition may also be o f interest for further
downstream processes along the biorefinery concept where char would serve as feedstock for
various streams (e.g.: energy, chemical extracts or chemical synthesis).

Hale et al. [2012] recently published one of the few studies about the presence o f PAHs and
dioxins in char extracts. They analyzed mostly slow pyrolysis chars from several biomass
sources, reactor types and production methods (temperatures from 250 to 900 °C) and they
quantified total PAH in those chars. They concluded that the total as well as the bioavailable
contents of PAHs and dioxins found in their slow pyrolysis chars were below environmental
levels (threatening levels) for soil application. However, the authors specified that their results
should not be extended any more than for slow pyrolysis chars and limited to their studied
conditions.

126
As well, the use of CO2 as an inflow gas environment in low temperatures pyrolysis conditions
has not been thoroughly reported in the open literature. Pyrolysis as well as thermochemical
conversion processes in general are attracting genuine interest for recycling flue gases and
residual CO 2 , although, most of the reported experiments have used carbon dioxide at higher
temperatures (above 600 °C). Despite the fact that CO 2 is generally considered to be inert at
low temperatures [Duan et al., 2009], interesting phenomenon may be observed at low
pyrolysis temperatures of 300-400 and 500 °C [Pilon and Lavoie, 2013a]. Potential interaction
of CO2 with chemical functions found in biomass are possible to exist based on differences
noted in CO 2 vs N 2 atmospheres in the yields and characteristics o f the 3 main fractions (chars,
liquids and non condensable gases). The reaction o f carbon with CO 2 , usually unfavourable at
such low temperatures, the mineral content o f the biomass might have been affected by the
CO2 environment (ash content difference noted at 500 °C) and this could have promoted
interactions such as fixing CO 2 (COO + H) to yield a carboxylic acid function.

A similar possible reaction was observed based on the results o f Zhang et al. [2011] and
Jindarom et al. [2007b] who pyrolyzed corncobs at 550 °C and sewage sludge at 350 to 650
°C, respectively, both within a CO 2 environment. Both groups analyzed their bio-oils and
observed an increase in compounds containing carboxylic acid functions while operating in
CO 2 in comparison to a N 2 environment. In addition, Zhang et al. [2011] observed an increase
in acid content and overall lower amounts of phenol derivatives and sugars. On the other hand,
the phenol concentration (C6H60 ) of their bio-oils reached a higher value following pyrolysis
in a CO2 environment. In general, there was also a correlation with the lower methoxy
function when using CO 2 , which could be consistent with the overall higher concentrations of
acetic acid, which may lead to the assumption that CO 2 has been fixed as a carboxylic acid
function. The authors also compared calorific values for condensable products and they
observed an increase from 17.8 MJ/kg with N 2 to 20.2 MJ/kg with CO 2 . Jindarom et al.
[2007b] generally observed more oxygenated compounds in a CO 2 atmosphere, a reduction in
aliphatics content and an enhancement in carboxyl groups and ketones formation.

127
Along this study, switchgrass (Panicum virgatum) was used due to its potential as an energy
crop, a native crop from North America which represents an interesting feedstock for
pyrolysis. Work reported on this feedstock pyrolysis presented bio-oils composition in
standard inert atmosphere only (i.e., N2). Generally, the bio-oils resulting from these studies
showed contents rich in acetic acid, furfural, acetol, levoglusosan, guaiacol and phenol,
contents that approach those of typical wood bio-oils values [Boateng et al., 2007; Bridgeman
et al., 2007; He et al., 2009; Mullen and Boateng, 2008].

This work presents the chemical analysis of the char extractives as well as of the condensable
gases (bio-oils) obtained from pyrolysis of switchgrass (Panicum virgatum) at 300, 400 and
500 °C within N2 and C 0 2 environments. Char extracts presented in common with bio-oils
compositions along the same study in addition of using C 0 2 as an inflow gas environment in
low temperatures pyrolysis conditions are rather scarce in literature for those specific
conditions as well as for pyrolysis in general.

6.2 Materials and methods

6.2.1 Experimental setup

The experimental setup has been previously described in details [Pilon and Lavoie, 2013 a] and
no modifications were made on the process for this study. About 25 g o f switchgrass at an
initial moisture content level o f 4-9% were pyrolyzed per test in a process by which the
feedstock is first inserted into the sealed reactor and heated using nitrogen or carbon dioxide
as heat vector gas (0.5 L/s). Heating is started at 55 °C/min, up to the reactor’s desired
operating temperature. Operating temperatures (300, 400 and 500 °C) were maintained for 2.5
min after which the vector gas flow was stopped and the reactor was cooled down to room
temperature. Each studied pyrolysis reaction set o f conditions was conducted in triplicate.
Statistical comparisons among the results were done with one factorial ANOVA; mostly
looking at the gas environment effect per specific operating temperature. Condensable
products were collected along a series o f 4 chilled condensers (glass tubes in water and ice)
installed right after the reactor. In order to alleviate the text, samples abbreviations will be

128
given from this point as follows: {(C or L)(3, 4 or 5)(N or C)} standing for {(Char or Liquid
bio-oil),(300, 400 or 500 °C), and (N2 or CO 2 )} (ex.: C3N corresponds to char produced at 300
°C, in a N 2 environment).

6.2.2 Solid material extraction and bio-oil preparation

For each Soxhlet extraction, three grams o f biomass (-90% 10-200 mesh particle size) were
placed in a cellulose cartridge (25 mm ID X 90 mm) and extracted with 150 mL o f methylene
chloride. Extractions were performed for 8 hours, after which the solvent and extracts were
filtered through grade 2 Whatman© filter papers to remove any char particles, before
transferring the liquid to a round-bottom flask for concentration using a rotary evaporator.
Each concentration was done at 40-45 °C under a vacuum of 50 kPa relative to ambient
pressure. Concentration was performed until complete solvent evaporation. The remaining
extracts were collected using 1-2 mL of dichloromethane in order to have a sufficiently
concentrated extract for GC-MS analysis.

Condensable products were collected from each o f the condensing tubes and connecting lines.
The contents were collected using acetone after which each solution was filtered through a
glass frit funnel to remove particles. Prior to analysis, acetone was completely removed in a
rotary evaporator (at 65 °C and under a vacuum of 50 kPa relative to ambient pressure). After
concentration, the bio-oil samples (homogeneous) were mixed with dichloromethane for
further GC-MS analysis. Product identification was performed using the NIST mass spectra
database (for >85% correlation spectrum identification). The gas effect comparison per
specific temperature was carried out from relative compounds content (area based), calculated
from the peak area of the total ion chromatogram.

6.2.3 Product characterization and GC-MS analysis

The gas chromatograh utilized was an Agilent 6890N network GC system equipped with 5973
network mass selective detector. The column was a DB-5ms MSD (30m x 0.25mm) (#122-
4532). Each sample was analyzed in a splitless mode, no pulse, at 2 pl/injection and with 2
min solvent delay. Gas chromatography and mass spectroscopy were performed using

129
temperatures o f 250 °C (inlet) and of 250 °C (detector) (MS: 200V, El: 70eV, Mass range: 10-
400 g/mol). The oven’s temperature program started at 60 °C for 3 min, followed by a
temperature increase (4 °C/min) up to 310 °C which was maintained for 1 min. Methods used
for elemental composition, calorific value, Karl Fisher water content, and proximate analysis
have been described in Pilon and Lavoie [2013a].

6.3 Results and discussion

6.3.1 Char extracts and bio-oils chemical composition

Looking in general terms at the results for char extracts, the quantity o f extracted material
varied with respect to temperature and gas environment (Figure 6.1). With respect to
temperature, the 500 °C conditions resulted in significantly less char extracts compared to the
ones produced at 300 and 400 °C. This result is in agreement with the higher feedstock
conversion (lower char and higher gases yields) resulting from a higher temperature and with
the lower volatile content in chars produced at 500 °C (Table 6.1). With respect to the gas
environment, as presented in Table 6.1 and discussed in further detail in Pilon and Lavoie
[2013a], at 300 °C - CO2 , char had a higher volatiles content (P=0.1) and the bio-oil yield was
significantly lower (P<0.002) compared to the 300 °C - N 2 conditions. The char extracts
results obtained for the 300 °C chars are in agreement with these results since the extracted
amount (on a char basis) was higher for the CO2 atmosphere. The char’s extracted material
may correspond to the volatiles that condensed on the char matrix during devolatilization.

130
 12

Figure 6.1: Mass of extracts generated by Soxhlet extractions o f chars using

dichloromethane.

131
Table 6.1: Physicochemical characteristics of raw switchgrass, chars and bio-oils derived from switchgrass
HHT Gas Char Oil Gas C H O N Ash Volatiles FC HHV HHV Water
H/C* O/C"
1°C] Env. i% r i% r |%J*b l%l<l l%]4 I%|d 1%]" [%r- i%r |% fd | MJ/kg jd IMJ/kg] i% r
Switchgrass n.a. n.a. n.a. 47.1*0.5 5.4±0.0 42.7*0.8 0.6±0.0 1.4 0.7 3.7*0.2 81.0*0.2 15.3*0.2 20.7*1.1 n.a. n.a.
300 N, 54.4±2.2 24.6*0.6 21.0*2.4 54.7*0.7 4.7*0.1 34.2*2.4 0.7*0.1 1.0 0.5 7.3*1.0 57.1*0.6 35.6*1.2 24.5*1.1 15.4*2.0 26.6*5.3
300 co2 59.2*4.1 18.0*1.5 22.7*4.7 53.7*0.8 4.9±0.2 34.6*1.2 0.7*0.1 1.1 0.5 6.3±0.5 61.0*3.9 32.8*3.5 22.2*0.7 15.7*1.1 23.5*12.9
400 N, 36.7±2.9 35.6*3.3 27.7*3.4 61.1*2.5 3.6*0.4 22.8*2.8 0.8±0.0 0.7 0.3 10.8*0.7 35.8*2.8 53.5*3.4 26.6*2.0 15.8*1.6 25.5*5.3
400 CO: 35.9*1.6 33.7*1.4 30.4*2.2 61.5*0.1 3.7±0.3 21.1*1.9 0.9*0.1 0.7 0.3 11.5*0.3 36.5*2.1 52.0*1.9 25.8*1.6 15.4*1.1 27.6*3.6
500 n2 28.2*0.7 37.7*0.7 34.1*0.7 63.8*1.3 2.5±0.1 20.0*0.5 0.8*0.1 0.5 0.2 13.5*0.6 28.3*1.9 58.3*1.7 26.7*1.2 14.0*1.1 39.0*2.2

500 co2 28.1*0.8 36.0*0.6 35.9*1.1 64.4*2.0 2.5±0.0 19.0*0.5 0.7*0.1 0.5 0.2 11.5*0.7 24.8*3.4 63.7*3.1 28.3*1.6 15.6*1.4 36.7*2.6
“Expressed on a raw biomass (dry basis). bObtained by difference. 'Source: Pilon and Lavoie [2013a], dDry basis. “±” corresponds to std. dev. (for n=3).
Legend: HHT: Higher Heating Temperature; Env.: Environment; FC: Fixed carbon; HHV: Higher Heating Value.

132
The char extracts contained a wide variety o f compounds with several types o f functionalities.
These compounds may have been formed during decomposition and recombination reactions
and possibly condensed back in char along the devolatilization process [Mohan et al., 2006],
Due to the relatively short residence time o f 2.5 min used in this study (compared to slow
pyrolysis), this might have favoured a higher condensed volatiles content (tar) on the char
itself instead of enhancing complete devolatilization [Hale et al., 2012], The conditions
actually used are most representative o f medium pyrolysis based on the heating rate level
employed and char’s residence time in reactor [Boateng, 2007; Brown, 2009; Downie et al.,
2009],

Among the compounds identified, only a few different types of alkanes were observed in C3N
(all with a low relative content) and none o f them were present within C3C (Table 6.2). A
reason explaining the different types o f alkanes found at 300 °C may be in part related to the
decomposition o f the extractives, starting at around 200 °C up to 600 °C [Guo et al., 2010],
Extractives (and most specifically long chain fatty acids found in switchgrass extractives [Yan
et al., 2010]) are known to decarboxylate at such temperatures and possibly to pass through
cracking reactions, which could explain the occurrence of long alkane chains [Bourke et al.,
2007; Jindarom et al., 2007b; Maher et al., 2008]. The fact that the C3C samples did not
contain any of these long alkane chains shows that CO 2 had a strong effect on the extracts at
that specific temperature. This phenomenon might be explained by extraction o f the alkanes in
this condition (C3C) that would be inhibited by a new link resulting from CO 2 environment.
Biomass is known to decarboxylate at torrefaction temperature [Prins et al., 2006c], however,
a CO2 environment may reduce decarboxylation and even enhance carboxylation with CO 2
reacting with the alkane functional groups leading to more acidic compounds. These new
acidic compounds, expected to be more polar, may become less prone to extraction by the
solvent. It has to be noted that, at 500 °C, despite the fact a high relative content of alkanes
was found in the char extracts (Figure 6.2), the overall amounts o f extracted material (Figure
6.1) showed that the quantities of produced alkanes, on a char basis, were similar to the ones
produced at lower temperatures. The relative contents would, however, be higher under a CO2
atmosphere (vs N 2 ) for 500 °C as well as for 400 °C, which is the inverse as the behavior
observed at 300 °C. Differences observed among the identified alkane components with

133
respect to the gas environment for 400 and 500 °C, respectively, were however not statistically
significant.

The compounds identified in the char extracts were mostly phenol derivatives (Table 6.2).
Phenols may be found in raw switchgrass, tightly bound within the lignin structure or as part
o f other extractable phenolics [Sarath et al., 2007], A higher relative content o f phenol
derivatives was found in the 300 °C - N 2 char extracts compared to the 300 °C - CO 2
environment (52% vs 32%), however since more char extracts were obtained under CO 2
(Figure 6.1), the content (on a char basis) could show that both gas vectors generated
comparable amounts o f phenolics. On the other hand, results differed at 400 and 500°C, where
phenol derivatives production appeared to be enhanced by a CO2 environment, especially for
char extracts (relative content of 36 and 50% at 400 °C and 12 and 23% at 500 °C for N2 and
CO2 environments, respectively; Table 6.2). Following that trend, phenol, 4-methyl-phenol,
2,4-dimethylphenol showed a significant difference with respect to gas environment; all were
in char extracts, from 400 °C (P=0.09), 500 °C (P=0.01) and 500 °C (P<0.04) respectively.
Results show also that phenol derivatives (relative content) from chars decreased with
increasing temperature, mostly at 500 °C, which seems to be consistent with the increasing
phenol derivatives concentrations in the bio-oils fractions with increasing temperature (Table
6.2, Figure 6.2 and Figure 6.3). The behaviour of phenol derivatives with respect to
temperature is supported by the observations showing that char and char extracts yields
decreased and bio-oils yields increased when temperature was increased from 300 to 500 °C
(Table 6.1). This is also consistent with the fact that thermal degradation o f the phenyl units
composing lignin leads to phenolics production [Amen-Chen et al., 2001] and that lignin is
known to degrade at temperatures ranging from 250 up to 625 °C [Butterman and Castaldi,
2010]. It may be noted that phenol, 4-methylphenol and 2-methoxy-4-methylphenol are
among the phenol derivatives found in bio-oils that show an increase in relative content with
respect to temperature. These results are consistent with the literature where phenols have
been reported to increase in bio-oils with temperature (up to about 600 °C) and where mostly
alkyl phenolics may be found [Mohan et al., 2006], In bio-oils, with respect to the gas
environment, the phenol derivatives relative content at 300 °C differs from 13.9 to 15.2% (N 2
and CO2 respectively); a difference which could be inversed if these phenol yields were

134
calculated on an original biomass basis since the bio-oil yield was significantly lower at 300
°C under CO 2 environment (Table 6.1). Still in the char extracts, most of the phenol
derivatives containing methoxy group(s) were present only at 300 & 400 °C (Table 6.2).
Among the phenol derivatives, example of products following that behavior are 2-methoxy-
phenol (guaiacol), 2-methoxy-4-methyl-phenol (4-methylguaiacol), 2,6-dimethoxyphenol
(syringol) and 4-ethyl-2-methoxy-phenol, which were all present in higher concentrations.
Loss o f methoxy functions from the aryl lignin structure with increasing level of severity
conditions has been reported earlier [Beauchet et al., 2012], Although this reaction has been
also observed at a lower temperature and in an aqueous NaOH solution, to a certain extent,
this observation could be extrapolated to what was observed in our case and for other
pyrolysis reactions at 450-500 °C [Amen-Chen et al., 2001]. These results are consistent with
the increasing phenol (CAS# 108-95-02) content observed in increasing bio-oils yields at 500
°C (Table 6.1 and Table 6.2) [Amen-Chen et al., 2001].

Among the ketones (including cyclopentanone derivatives), the results showed that none could
be observed within chars produced at 500 °C which, in this case, could be related to the
complete degradation of the carbohydrates structures in hemicelluloses or cellulose (Table
6.2). Ketones are well known to be related to hemicelluloses and cellulose degradation [Lv et
al., 2010; Wang et al., 2011], The loss o f carbonyl groups is in agreement with increases in
gas yields at that temperature (Table 6.1). Cyclopentanone derivatives are among the major
compounds found in bio-oil for every temperature (Table 6.2 and Figure 6.3). Cyclopentanone
derivatives in the char extracts seemed to be affected by the gas environment since an overall
lower relative content was obtained with the use o f CO 2 compared to N 2 for 300 °C chars
(1.80 vs 3.08% respectively) (Table 6.2). This difference might be attenuated by the higher
extraction yields obtained for the 300 °C-C02 chars (i.e.: on a char basis). On the other hand,
some specific compounds such as 2-methyl-2-cyclopenten-l-one, 3-methyl-2-cyclopentan-l-
one and 2,3-dimethyl-2-cyclopenten-l-one were solely found in char extracts when using N 2
(at 300 °C). The lower percentages o f ketones resulting from the CO 2 environment may once
again, result from the C 0 2 reacting with some o f the compounds by carboxylation, increasing
their polarity, therefore becoming less prone to being solubilised by dichloromethane. Chars
produced at 400 °C also showed a difference in relative total ketone content with respect to the

135
gas environment: 7.8% vs 11.5%, for C4N and C4C, respectively (Table 2). This behavior was
consistent with the observation of Jindarom et al. [9] who explained the increasing ketone
content under CO2 by an insertion of C 0 2 through unsaturated compounds. Among the bio-oil
compounds, in that sense, the 2-hydroxy-2-cyclopenten-l-one content was higher under CO 2
vs N2 (P<0.07) at 400 °C.

Chars produced at 500 °C also seemed to enhance releasing compounds such as furan
derivatives, alcohols, as well as methoxybenzene, ethylbenzene and other benzene derivatives
(Table 6.2). Much of these compounds, found at lower temperatures (300 and 400 °C), were
not present in the 500 °C char extracts. For all these benzene derivatives that are mostly
present at 400 °C, the increment of temperature towards 500 °C might have favoured the
formation of naphthalene derivatives within char which were mostly found at 500 °C (Table
6.2). This would be consistent with a known mechanism, where a temperature increase would
result in aromatic rings losing their functional groups and eventually recombining into PAHs
[Hajaligol et al., 2001; McGrath et al., 2003]. It is also consistent with an observation
previously made by Pilon and Lavoie [2011b] despite the important fact that the reactor’s
operation was different (higher heating rate, higher convection, and partial oxidation). The
reduction in chars’ H/C and O/C ratios observed with increasing production temperature is in
agreement with an increase in aromatic compounds concentration among char extract
components (Table 6.1 and Table 6.2). On the other hand, PAH levels in char were observed
by Hale et al. [2012] to be at higher levels following low pyrolysis temperatures, while
Keiluweit et al. [2012] similarly observed maximal char PAH concentrations at 400-500 °C,
both for slow pyrolysis, at low heating rates and possibly under low convection conditions. In
both cases, their studies resulted in several types of higher molecular weight PAHs for
temperatures from 300 to 500 °C [Hale et al., 2012 and Keiluweit et al., 2012] compared to
the present study where only low molecular weight naphthalene derivatives were observed.
As a result, heating rate, char residence time, gas flow and gas residence time could account
for the PAHs formation, which would be an interesting subject for further investigation. In
accordance to Hale et al. [2012], PAHs have strong sorption on char due to the 71-71 interaction
between the char and the coplanar structure of PAHs themselves, while some o f the PAHs
might also be trapped within char structure as a result o f mechanical occlusion o f the char’s

136
pores. As a result, more PAHs could possibly be expected due to inadequacy of
dichloromethane to extract all o f the compounds, which could be more easily extracted by an
aromatic solvent such as toluene. From a soil amendment point o f view, despite the presence
of naphthalene derivatives, the quantification of bioavailable PAHs would be o f importance
since it may differ from the total PAHs available [Hale et al., 2012]. As for the effect o f the
gas environment, CO2 seemed to have enhanced naphthalene derivative formation since 2-
methyl-naphthalene and 1,7-dimethyl-naphthalene showed more than a twofold increase for
about the same char extracts content, both for C5N and C5C (Figure 6.1); with the PAHs’ total
relative content also increasing and passing from 6.7 and 10.5% (N2 vs CO 2 , respectively)
(Table 6.2). No PAH was detected in the liquids.

Among the carbohydrate derivatives found in char extracts, l,4:3,6-dianhydro-a-D-

glucopyranose was found at an higher concentration under the 500 °C - N 2 conditions with
regards to all other studied conditions (Table 6.2). This can simply result from the much lower
char extract yields at 500 °C (Figure 6.1), which would lead, on a char basis, to concentrations
similar to those measured for 300 and 400 °C chars. Noteworthy that l,4:3,6-dianhydro-a-D-
glucopyranose saccharide was observed only in C5N, but not in C5C conditions (Table 6.2).
In addition, lower levels o f l,4:3,6-dianhydro-a-D-glucopyranose were also observed for C3C
and C4C compared to C3N and C4N. This observation may signify that, once again,
carboxylation could have occurred, CO2 reacting with l,4:3,6-dianhydro-a-D-glucopyranose
or its precursor, thus adding acidic function(s) (COO + H) and, as a result, the new compounds
would have been less prone to solubilisation with dichloromethane. Nonetheless, the lower
content o f l,4:3,6-dianhydro-a-D-glucopyranose resulting from the C 0 2 environment could
also be coherent with the lower furfural content found for C4C conditions compared to C4N
(Table 6.2). The l,4:3,6-dianhydro-a-D-glucopyranose has been proposed to precede furfural
formation [Branca et al., 2010], Furfural was also identified in bio-oils, with a slightly higher
relative content under N 2 for each temperature studied (8.4 vs 7.8; 6.6 vs 6.1 and 8.0 vs 7.3 for
L3N vs L3C; L4N vs L4C and L5N vs L5C respectively; significant at 400 °C for P<0.065).
Despite the fact furfural was not found at C4C, C5N and C5C, these conditions (and especially
400 °C - CO2 ) does not seem to impair furfural formation since it is present in bio-oils. The
low l,4:3,6-dianhydro-a-D-glucopyranose content in bio-oils, could be due to a better affinity

137
o f this later compound with aqueous bio-oil fraction than with CH2 CI2 . Furfural is also
considered to be a degradation product from hemicellulose where cellulose presence or
cellulose degradation products would enhance furfural formation [Branca et al., 2010; Wang
et al., 2011]. The presence of furfural in chars extracts at 300 and 400 °C (Table 6.2) would
then be consistent with hemicelluloses and cellulose transformation known to occur around
225-300 °C and 305-375 °C, respectively [Prins et al., 2006b].

It is also possible to notice, among the GC-MS identifications (Table 6.2), that compounds
such as alkanes, naphthalene derivatives and benzene derivatives were found in char extracts,
but not in the bio-oils although they could be expected in this fraction [Fabbri et al., 2010;
Onay, 2007]. A reason explaining their absence from bio-oils could be related to the high
solubility of some of them with the aqueous fraction o f bio-oil that is not miscible with
CH2 CI2 . Water, acids, alcohols, aldehydes, ketones and sugar constituents are known to be
water soluble [Boateng et al., 2007]. As a result, in order to improve compounds
identifications (for char extracts and bio-oils), other types o f solvent, with different polarity
levels, could be used, including methanol and water. The use of solvent with a higher polarity
could also corroborate the explanations about CO2 involved in carboxylation, which would
have led to the resistance of compounds to dichloromethane solubilisation and extraction.

Based on the grand total content values (Table 6.2), it may be observed that the total amount
o f compounds identified in the char extracts for each specific temperature shows a difference
with respect to the gas environment. At 300 °C, more compounds were identified under a N 2
gas inflow environment than under CO2 ; the inverse is however observed for 400 and 500 °C.
This observation is similar to those obtained with some specific compound categories such as
phenol derivatives amd alkanes, and with cyclopentanone derivatives to some extent (Table
6.2, Figure 6.2 and Figure 6.3). This difference was mostly observed with char extracts while
the contents of bio-oils varied much less with the process conditions tested with respect to gas
environment per specific temperature. These results show that CO 2 environment has an effect
on chemical products formed at low temperatures within the specific conditions of this study.
Since CO 2 atmosphere was already observed to enhance carboxylic acid such as acetic acid
[Zhang et al., 2011 and Jindarom et al., 2006] and that several observations made along this

138
study could be explained by CO2 reacting by carboxylation with the biomass constituents and
pyrolysis products, the CO 2 environment could have reduced the decarboxylation reaction
(prone at 300 °C) and even enhanced the reversed reaction (carboxylation) due to high CO 2
concentrations. Under a CO2 inflow environment at 400 or 500 °C, it was possible to note that
several methyl- and methoxy-containing compounds showed higher relative contents (Table
6.2; CO2 v.v N 2 per specific pyrolysis temperature). Further quantitative analyses of CO 2 -
pyrolysis products should be conducted on the most important compounds in order to clarify
the mechanisms that could have led to the differences observed.

139
Table 6.2: GC-MS compound identification of PCM char extracts and bio-oils (Contents from GC peak area, n=3 and for *, n=2)
Char Extracts by DCM Condensable Products
Compounds CAS#
C3N C3C* C4N C4C C5N C5C L3N L3C L4N L4C* L5N L5C
Alkanes and cycloalkanes
Undecane 001120-21-4 0.81
Dodecane 000112-40-3 0.32 4.35
Tridecane 000629-50-5 0.06 0.23 0.23 3.46 9.51
Tetradecane 000629-59-4 0.22 0.11 0.24 2.19 1.74
Pentadecane 000629-62-9 0.49 0.17 0.34 2.53 4.80
Hexadecane 000544-76-3 0.12 0.14 0.18 1.43 1.13
Heptadecane 000629-78-7 0.17 0.10 0.21 2.43 0.18
Octadecane 000593-45-3 0.12 0.13 0.13 0.56
Nonadecane 000629-92-5 0.13 0.33 0.46
Gicosane 000112-95-8 0.09 0.04 0.08
Heneicosane 000629-94-7 0.06 0.13
Cyclododecane 000294-62-2 0.13 0.08
Total (relative content): % 1.27 0.00 1.24 1.94 12.92 22.97 0.00 0.00 0.00 0.00 0.00 0.00
Phenol derivatives
Phenol 000108-95-2 4.82 3.82 4.09 5.78 1.42 1.66 1.87 2.27 2.16 3.64 3.60
Phenol, 2-methyl- 000095-48-7 2.19 0.99 1.95 2.22 4.89 1.12 0.58 1.29 1.45
Phenol, 3-methyl- 000108-39-4 0.27 3.09 1.76 0.09
Phenol, 4-methyl- 000106-44-5 3.70 0.41 2.29 4.93 0.39 5.93 0.11 0.17 0.82 0.63 2.14 2.48
Phenol, 2-ethyl- 000090-00-6 0.63 1.08 0.23 0.43
Phenol, 4-ethyl- 000123-07-9 4.28 1.89 3.40 3.66 0.49 0.56 0.70 0.75 0.76 0.71 0.81
Phenol, 2,3-dimethyl- 000526-75-0 0.10 0.70 0.64
Phenol, 2,4-dimethyl- 000105-67-9 0.18 0.79 0.64 2.51 0.08 0.11
Phenol, 2,5-dimethyl- 000095-87-4 0.42 2.32
Phenol, 3,4-dimethyl- 000095-65-8 0.34 0.34 0.61
Phenol, 3,5-dimethyl- 000108-68-9 1.89 0.08 0.17 4.08 3.64 0.09
Phenol, 2-methoxy- 000090-05-1 8.43 8.50 5.89 11.83 4.05 4.32 5.27 4.48 5.41 5.56
2-Methoxy-6-methylphenol 002896-67-5 0.69 1.32
Phenol, 2-methoxy-4-methyl- 000093-51-6 5.36 3.43 6.44 8.33 1.43 1.30 2.50 2.18 2.59 2.64
Phenol, 2-ethyl-6-methyl- 001687-64-5 0.08 0.05
Phenol, 3-ethyl-5-methyl- 000698-71-5 0.25 0.27
Phenol, 4-ethyl-2-methoxy- 002785-89-9 7.71 5.63 6.21 5.79 1.37 1.49 1.87 1.74 1.82 1.92
Phenol, 2,6-dimethoxy- 000091-10-1 1.91 2.55 3.05 1.71 0.64 0.71 1.01 1.03 0.94 1.02
2-Methoxy-4-vinylphenol 007786-61-0 1.63 2.05 0.59 3.29 3.84 2.82 3.51 2.70 2.84
Phenol, 2-methoxy-3-(2-propenyl)- 001941-12-4 0.09 0.09 0.13
Phenol, 2-methoxy-4-propyl- 002785-87-7 0.18 0.34 1.02 0.68 0.02 0.18 0.08
3-Allyl-6-methoxyphenol 000501-19-9 0.06 0.08 0.14
Eugenol (or 2-methoxy-4-(2-propeny!)phenol) 000097-53-0 0.64 0.05
Phenol, 2-methoxy-4-( 1-propenyl)-,(E)- 005932-68-3 0.61 0.75 0.70 0.16 0.58 0.63 0.70 0.73 0.57 0.56
Phenol, 2-methoxy-3-methyl- 018102-31-3 0.44 0.81 0.06
Phenol, 2,6-dimethoxy-4-(2-propenyl)- 006627-88-9 0.03 0.09 0.13

140
Vanillin (or4-hydroxy-3-methoxybenzaldehyde) 000121-33-5 _________________________________________ 0.02 0.04 0.02 0.10______________
Total (relative content):_______________________ %_________ 44.9 31.1 35.6 49.5 12.4 23.1 13.9 15.2 19.6 18.3 21.9 23.1
Ketones
Acetophenone 000098-86-2 0.09 0.05
2-Pentadecanone, 6,10,14-trimethy1 000502-69-2 0.01 0.03 0.03
2-Cyclohexen-1-one 000930-68-7 0.06 0.14
Total (relative content): % 0.00 0.00 0.15 0.20 0.00 0.00 0.01 0.03 0.00 0.03 0.00 0.00
Cyclopentanone derivatives
Cyclopentanone 000120-92-3 0.31 1.71
2-Cyclopenten-1-one 000930-30-3 0.17
2-Cyclopenten-l-one, 2-methyl- 001120-73-6 0.34 1.44 1.78 0.42 0.37 0.51 0.45
2-Cyclopenten-l-one, 3-methyl- 002758-18-1 0.61 2.50 3.59
2-Cyclopenten-1-one, 2,3-dimethyl- 001121-05-7 0.84 1.12 2.17 0.11 0.53 0.52 0.52
2-Cyclopenten-1-one, 2-hydroxy-3-methyl- 000080-71-7 1.29 1.62 1.85 1.63 1.32 1.40 2.16 2.12 1.86 1.99
2-Cyclopenten-1-one, 3,4-dimethyl- 030434-64-1 0.19 0.18
2-Cyclopenten-l-one, 2-hydroxy- 010493-98-8 3.94 3.94 4.96 5.97 3.92 2.77
2-Cyclopenten-1-one, 3-ethy1-2-hydroxy- 021835-01-8 0.10 0.16 0.10 0.25 0.48 0.51
2-Cyclopenten-l-one, 3-ethy1- 005682-69-9 0.08 0.26 0.27 0.02 0.14
Total (relative content): % 3.08 1.80 7.84 11.50 0.00 0.00 5.77 6.07 8.15 8.51 6.80 5.73
Furan derivatives
2-Furanmethanol 000098-00-0 3.21 6.84 1.91 0.52 4.35 4.98 4.79 5.06 4.92 4.89
2-Furancarboxaldehyde, 5-methyl- 000620-02-0 1.60 1.26 0.59 0.30 0.58 0.51 0.43
Ethanone, l-(2-furanyl)- 001192-62-7 1.11 1.06
2(5H)-Furanone, 5-methyl- 000591-11-7 0.11 0.38
Butyrolactone (or 2(3H)-Furanone, dihydro-) 000096-48-0 3.19 6.08 4.17 2.60 0.54 0.40
Furfural 000098-01-1 3.33 3.93 3.86 8.43 7.78 6.59 6.05 7.99 7.26
Total (relative content): % 9.74 16.84 12.65 5.44 0.00 0.00 13.49 13.30 11.79 12.07 13.42 12.98
Benzofuran derivatives
Benzofuran 000271-89-6 0.18
Benzofuran, 7-methyl- 017059-52-8 0.20
Benzofuran, 2,3-dihydro- 000496-16-2 0.56 1.51 1.28
Benzofuran, 2-methyl- 004265-25-2 0.41 0.23
Total (relative content): % 0.00 0.00 0.41 0.23 0.00 0.38 0.00 0.00 0.56 1.51 1.28 0.00
Ethylbenzene derivatives
Benzene, l-ethyl-2-methyl- 000611-14-3 0.28
Benzene, l-ethyl-4-methoxy- 001515-95-3 0.17 0.31
Benzene, 4-ethyl-1,2-dimethoxy- 005888-51-7 0.19 0.10 0.38
Methoxybenzene derivatives
Benzene, 1-methoxy-4-methyl- 000104-93-8 0.35 0.53
Benzene, 1,2-dimethoxy- 000091-16-7 0.08 0.22
3,4-Dimethoxytoluene 000494-99-5 0.22 0.25 0.43
Methylbenzene derivatives
Benzene, 1,3,5-trimethyl- 000108-67-8 0.42
Benzene, 1,2,3,4-tetramethyl- 000488-23-3 0.18

141
Other benzene derivatives
Bicyclof4.2.01octa-l,3.5-triene 000694-87-1 0.12 0.28
Total (relative content): % 0.41 0.00 1.07 2.15 0.00 0.88 0.00 0.00 0.00 0.00 0.00 0.00
lH-Indene derivatives
lH-Inden-l-one, 2,3-dihydro- 000083-33-0 0.76 0.82 0.04 0.05
1H-Inden-1 -one, 2,3-dihydro-3-methyl- 006072-57-7 0.23 0.33
Total (relative content): % 0.76 0.00 0.23 1.19 0.00 0.00 0.00 0.00 0.04 0.05 0.00 0.00
Naphthalene derivatives
Naphthalene, 1-ethyl- 001127-76-0 0.35 0.18
Naphthalene, 1,4-dimethyl- 000571-58-4 0.25
Naphthalene, 2,3-dimethyl- 000581-40-8 0.43
Naphthalene 000091-20-3 0.87
Naphthalene, 2-methyl- 000091-57-6 1.57 2.69
Naphthalene, 1,7-dimethyl- 000575-37-1 0.20 1.40
Naphthalene, 1,6,7-trimethyl- 002245-38-7 0.67 0.33
Naphthalene, 1,8-dimethyl- 000569-41-5 0.19
Naphthalene, 2,7-dimethyl- 000582-16-1 1.22 2.52
Naphthalene, 1,6-dimethyl- 000575-43-9 0.72
Naphthalene, 2,6-dimethyl- 000581-42-0 1.49 2.09
Total (relative content): % 0.00 0.00 0.00 0.00 6.65 10.52 0.00 0.00 0.00 0.00 0.00 0.00
Alkenes and cycloalkenes
5-Tetradecene, (E)- 041446-66-6 0.25
1-Tridecene 002437-56-1 0.07 0.08 0.35 0.15
Total (relative content): % 0.07 0.00 0.00 0.08 0.35 0.40 0.00 0.00 0.00 0.00 0.00 0.00
Carbohydrate derivatives
1,4; 3,6-Dianhydro-a-D-glucopyranose 4451-30-3 1.58 0.72 1.81 1.11 2.66 0.07 0.35 0.42 0.14
2,3-Anhydro-d-mannosan 1000129-98-0 0.15 0.06 0.09
Total (relative content): % 1.58 0.72 1.96 1.17 2.66 0.00 0.00 0.07 0.35 0.51 0.00 0.14
Acids
propanoic acid 000079-09-4 2.34 1.75 0.28 1.43 0.79 1.39
Hexadecanoic acid, methyl ester 000112-39-0 0.13 0.16 0.13 0.33
Pentadecanoic acid, 14-methyl-, methyl ester 005129-60-2 0.15 0.16 0.23
Total (relative content): % 0.28 0.32 0.36 0.33 0.00 0.00 2.34 1.75 0.28 1.43 0.79 1.39
Alcohol
1-Hydroxy-2-butanone 005077-67-8 1.22 3.00 3.60 7.64
Aldehydes
Cyclopropaneoctanal, 2-octyl- 056196-06-6 0.32
Ester
2H-Pyran-2-one, tetrahydro- 000542-28-9 0.22 0.45 0.46
Others
Azulene, 4,6,8-trimethyl- 000941-81-1 0.11
Grand Total (relative content): % 63.5 54.0 62.0 74.2 35.0 58.4 39.1 44.0 40.8 42.4 44.2 43.3

142
 60
□ C3N

-C3C
~
T
•
> 50
m
a
A
1 40 ■ C4C
£
£dl □ C5N
| 30
o0 *C5C
>
|01 20
a:
10

X .H _ il

o -f jf
^ Compounds categories
Figure 6.2: Relative content of various categories o f identified compounds in the
char extracts.

60
□L3N
r L3C
_ 50
•B
•
a
A
M ■L4N
I 40 ■L4C
□L5N
4c>
E 30
o «sL5C

20
60

10 ni
^
ilII1
^ y tgr & & y
, ilii I
/ / / / y s y y * y y ^r
y y y / / / / /
* o' / y */ y Compounds categories
<f

Figure 6.3: Relative content of various categories o f identified compounds in the

bio-oil fractions.

143
6.3.2 Calorific value

In addition to chemical composition, the calorific value o f the products was also determined.
A significant lower calorific value was observed for chars produced at 300 °C in C 0 2 vs a N 2
environment (22.2 vs 24.5 MJ/kg, respectively; P<0.05) (Figure 6.4). However, the char yields
were lower at 300 °C under N2 conditions vs C 0 2 (Table 6.1). For the C3N conditions, char
production resulted in a higher volatile loss and a higher bio-oil yield compared to the C3C
conditions. As a result, a higher fixed carbon content was obtained, which led to a higher
calorific value due mostly to the carbon’s complete oxidation (C + 0 2 —*■C 0 2; A H f M 25 °c ' -
393.5 kJ/mol) being more exothermic than the volatiles’ oxidation containing other elements
such as hydrogen, oxidizing through 2H + V2 O2 —*■H20 (AHfat 25 °c : -241.5 kJ/mol). At 500
°C, the high fixed carbon yield with a lower ash content may explain the slightly higher
heating value for C5C vs C5N. As for the bio-oils heating value, no significant difference was
observed with regards to temperature and vector gas. Bio-oils’ calorific value levels were
consistent with the ones found in literature (-15 vs 16-19 MJ/kg) [Zhang et al., 2007].
However, the lower calorific value level for the bio-oils might be due to some compounds that
were not possibly condensed and collected. The bio-oils condensing train, from which the
samples were taken, was shown in a previous analysis to collect only about 2/3 o f the bio-oils
while the remaining bio-oils (not used along this study) were trapped in a cotton filter [Pilon
and Lavoie, 2013a], This would be in agreement with Boateng et al. [2007]’s bio-oil
collecting system using condensers followed by an electrostatic precipitator, which lead to
bio-oils heats of combustion of 16.0 and 21.7 MJ/kg (for condenser and electrostatic
precipitator respectively).

144
 Figure 6.4: Higher heating value o f chars and bio-oils produced under various
temperature and vector gas conditions.

6.4 Conclusions
The characterization of the solid and liquid pyrolysis products obtained from medium
pyrolysis o f switchgrass showed that:
• At 300 °C, char produced under CO2 yielded higher amounts o f extracts while its
calorific value was significantly lower (P<0.05) compared to the N 2 environment (22.2
vs 24.5 MJ/kg, respectively).
• At 400 °C, in char extracts, furfural was present only under N 2 conditions and found at
significantly higher relative contentd in the bio-oils under N2 vs CO 2 environment. The
lower furfural content under CO 2 conditions is consistent with the l,4:3,6-dianhydro-a-
D-glucopyranose concentrations in the char, which was found at higher concentrations
under N 2 than CO 2 environment and is considered to be a precursor along furfural
production.
• Char extract yields were significantly lower at 500 °C which was in accordance with
the lower volatile contents found at these temperatures.
• Among the char extract compounds identified, lower grand total relative contents were
obtained at 300 °C in a CO2 gas inflow environment (vs N 2 ), while the inverse was
observed at 400 and 500 °C, a behavior consistent mostly with the phenol derivative
and alkane contents, and with cyclopentanone derivative contents to some extent.
 • Char extracts had high phenol contents, which appeared to decline with respect to
increasing temperature while increasing with the bio-oil for the same production
conditions. Phenol derivatives with methoxy functional groups, were present in char
extracts only from 300 and 400 °C operating temperature conditions, while other types
of phenol derivatives did not show that trend.
• In terms of PAHs, only naphthalene derivatives were found, and only at 500 °C. At
500 °C, CO2 appeared to enhance particularly the formation o f 1,7-dimethyl-
naphthalene and 2,7-dimethyl-naphthalene.
• Compounds containing carbonyl functional groups, such as ketones, esters, acids,
aldehydes, furan derivatives and alcohols, were found only at 300 and 400 °C in char
extracts.
• In char extracts, compounds containing carbonyl functional group, such as various
ketones, were found only at 300 and 400 °C.

Our emphasis on the analysis of the chemical compounds extracted from char together with
pyrolysis studies provided information for further compound value addition, information that
may also provide a better overview of pyrolysis conversion mechanisms, especially when
presented in concert with bio-oil composition. The use o f solvents with higher polarity levels
to extract and investigate further the acidic compounds possibly resulting from carboxylation
in a CO2 environment as well as the use o f aromatic solvents to support further studies on
PHAs and other aromatic compounds would be highly beneficial. Further quantification o f the
products of interest and especially those whose presence or concentration was influenced by
the gas environment, will eventually be needed to corroborate the observations reported here
and for the value additions of char (e.g.: biochar for soil amendment) with compounds that
may or may not be desired at certain levels. In addition, since the CO 2 might come from flue
gases in the future, further work using typical flue gas mixtures might also be o f interest for
industrial applications o f the technology.

146
Acknowledgements:
The authors would like to thank the Faculty o f Engineering and the Department o f chemical
engineering and biotechnological engineering (DCEBE) of the Universite de Sherbrooke
(UdeS) for financial support through the starting funds of Prof. J.-M. Lavoie.
Acknowledgements are also presented to the Industrial Chair in Ethanol Cellulosic of the
UdeS as well as to Professor Jean Lessard from the Chemistry Department o f UdeS in lending
research equipment. The authors would also like to acknowledge: Mr. Gaston Boulay from the
Chemistry Department o f (UdeS) for his technical support with the GC-MS analyses, Ms.
Sonia Blais from the Centre de caracterisation de materiaux of UdeS, Mr. Serge Gagnon, Mr.
Marc G. Couture, Mr. Henri Gauvin, and Mr. Osvaldo Serrano from DCEBE for their
technical support and, finally, Mr. Daniel Clement, switchgrass producer (Qc, Canada), for
providing the feedstock used in this research.

147
CHAPITRE 7 : CONCLUSION

7.1 Sommaire:
Ce projet de recherche s’est penche sur la production et la caracterisation du biocharbon par
pyrolyse de biomasse lignocellulosique. La communaute scientifique des demieres annees a
observe que l’usage du biocharbon comme amendement au sol representait un grand potentiel
pour 1’augmentation du rendement des cultures agricoles et comme methode de sequestration
du carbone. Cependant, les biochars (et biocharbons en general) comportent des
caracteristiques d’une grande variability qui peuvent a leur tour influencer la reponse
agronomique. La presente etude s’est done penchee sur l’etude de caracteristiques specifiques
du biocharbon en fonction de conditions de production pouvant influencer l’obtention de
proprietes d’interet pour le biochar. De plus, les conditions de production visaient une possible
transposition au niveau industriel, notamment pour la conception d ’unites portables de
pyrolyse. Ces conditions visaient une plage de temperatures relativement basses (entre 300 et
500 °C) et d’autres parametres d ’operation pour lesquels les rendements, la surface specifique
et la composition chimique pourraient avantager le biochar. L ’usage du CO 2 , compare a
l’environnement plus conventionnel de N 2 , a aussi ete etudie comme parametre d ’operation
lors de la pyrolyse. Le CO2 est un gaz residuel qui pourrait etre recycle et valorise dans le
precede. Etant donne que les conditions de production etudiees chevauchaient celles pour
l’obtention d’un charbon de torrefaction (un charbon vert pour valorisation energetique et
precedes subsequents), cette valorisation complementaire au biochar a aussi ete consideree
lors des analyses, mais tout en maintenant l’accent sur le biochar.

Pour cette etude, deux reacteurs a lit fixe de capacites differentes ont ete utilises (1 et 25
gbiomasse/batch, respectivement). La conception d ’un premier reacteur d ’une capacite de
lg/batch a permis la production de biocharbon dans un environnement de N 2 , a un taux de
chauffe eleve, ou la convection en reacteur etait favorisee par un important debit de gaz
entrant. A ces conditions, l’etude a tout d’abord porte sur l’effet de la temperature a de courts
temps de residence du biocharbon en reacteur (2,5 et 5 min). Lors d’une 2e etude, l’usage de

148
ce meme reacteur a permis l’obtention de resultats preliminaires visant la comparaison des
atmospheres de CO 2 et N2. Afin de poursuivre l’investigation de l’effet de l’usage du CO 2 en
pyrolyse, la necessite d ’effectuer plusieurs tests de caracterisation sur les biocharbons obtenus,
le besoin de controle sur le taux de chauffe, ainsi que le besoin d’une precision accrue au
niveau des bilans des co-produits pyrolytiques ont tous justifie la conception et la construction
d ’un reacteur de plus grande capacite. Un second reacteur d’une capacite de 25g\,\omasJ batch a
ainsi permis d ’explorer davantage l’usage de l’environnement de C 0 2 avec un suivi sur
1’ensemble des produits pyrolytiques obtenus (biocharbon, bio-huile et gaz). Ce demier a ete
opere a un taux de chauffe de 55 °C/min, un temps de residence en reacteur de 2,5 min, tout en
poursuivant l’etude de l’influence de la temperature (300, 400 et 500 °C) en atmosphere de
CO 2 ou de N2.

Lors des tests en reacteur de 1 g/batch (chapitre 3), tres peu de differences ont ete observees
dans les caracteristiques des biocharbons en fonction du temps de residence en reacteur (2,5 et
5 min), et ce, pour une meme temperature. Ainsi, uniquement le temps de 2,5 min a ete
maintenu dans les etudes suivantes (chapitres 4, 5 et 6) afin d’operer a des conditions
industriellement favorables. En terme de rendement en biocharbon, les conditions d ’operation
ont particulierement ete avantageuses a 300 °C et 2,5min, oil cette temperature combinee aux
conditions specifiques employees (taux de chauffe rapide et convection elevee en reacteur) ont
genere un biocharbon au rendement et aux proprietes physico-chimiques tres comparables aux
biocharbons presentes dans la litterature pour une temperature de 250 °C. L ’avantage repose
sur le temps de residence du biocharbon en reacteur qui s’est vu reduit d ’au moins 10X pour
une augmentation en temperature de 50 °C seulement. Ce type de resultat peut representer un
avantage majeur au point de vue applications industrielles futures dans le domaine de la
torrefaction.

Les resultats des extractions chimiques des biocharbons ont permis d’identifier une multitude
de produits. Parmi ces resultats, lors de l’usage du reacteur de 1 g/batch, le 1-undecanol a ete
observe a 300 et 400 °C en atmosphere de N 2 . Ce compose est d’interet pour le biochar
(amendement pour le sol) etant donne qu’il est deja utilise en agriculture comme agent
controlant pour la croissance des plantes. Dans le cas des 2 reacteurs utilises, la presence du

149
naphtalene et de ses derives a ete observee. Bien que le naphtalene soit un HAP ayant un
certain degre de toxicite, il est le seul HAP ayant ete observe, et ce, seulement a 500 °C.
D ’ailleurs, le naphtalene entre deja dans la composition de certains pesticides, ce qui pourrait
faire en sorte de ne pas limiter l’usage du biochar produit a ces conditions, si une
quantification de ce HAP, liee a une application en champ donnee (T b io c h a r/h a ), menait a des
concentrations conformes aux normes environnementales. De plus, le fait qu’aucun autre HAP
n ’ait ete observe lors des conditions utilisees represente un avantage, car plusieurs autres HAP
avaient recemment ete repertories dans la litterature pour du biocharbon produit par pyrolyse
lente. Avec les 2 reacteurs etudies, la quantite d’extraits de biocharbon a diminue avec une
augmentation de la temperature; particulierement avec le reacteur de 25 gbiomassJbatch, ou des
diminutions plus prononcees furent notees a 500 °C. Les extractions des biocharbons resultant
du reacteur de 25 gbiomassJbatch ont, quant a elles, permis de dresser un large eventail de
produits pouvant se retrouver dans les biocharbons, et ce, autant en atmosphere de N2 qu’en
atmosphere de C 0 2. Ces resultats pourraient, entre autres, s’averer utiles afin d ’etablir des
liens entre certains types de biocharbon et des observations agronomiques, faites a la suite de
l’utilisation de ces types de biochars en sols.

L’usage du C 0 2 lors de la pyrolyse de biomasse lignocellulosique a d ’ailleurs presente des

differences notables par rapport a l’atmosphere inerte conventionnelle d ’azote. Un effet
significatif a ete note a 300 °C sur le contenu en composes volatils du biocharbon, ainsi que
sur le bilan obtenu en bio-huiles (plus eleve et plus faible, respectivement, pour C 0 2 vs N 2).
L ’usage du C 0 2 a aussi influence la matiere inorganique a 500 °C, ou une diminution
significative (P < 0.06) du contenu en cendres fut observee en presence de C 0 2 (v.s N2). Cette
reaction des elements inorganiques avec le C 0 2 suggere une inhibition par le C 0 2 de faction
catalytique de certains elements inorganiques (particulierement les elements metalliques), une
action catalytique reconnue en atmosphere conventionnelle (de N2) lors de reactions de
pyrolyse. Ce comportement pourrait, entre autres, expliquer les resultats obtenus a 300 °C qui
s’apparente a une diminution de la depolymerization de la biomasse (mentionnes ci-dessus).
La carboxylation semble avoir ete favorisee en presence de C 0 2, ce qui pourrait expliquer
certaines des differences notees dans les bilans et dans les analyses des extraits de biocharbons
et des bio-huiles. Parmi les differences notees dans les composes identifies, plusieurs

150
composes se sont retrouves en concentrations plus faibles en atmosphere de CO 2 (vs N 2 ) dans
les extraits de biocharbons ainsi que dans les bio-huiles, dont le furfural a 400 °C. Bien que
des differences aient ete notees selon 1’atmosphere gazeuse utilisee, pour la valorisation du
biochar (en sol), c ’est davantage les reponses agronomiques qui determineront si les variations
notees dans les compositions et caracteristiques physico-chimiques seront benefiques ou non,
etant donne la multitude de produits observes et les effets que pourraient avoir ces demiers.
Les differences notees ne semblent toutefois pas freiner le potentiel que pourrait avoir la
valorisation du CO 2 dans le precede, notamment tout le CO 2 residuel provenant de precedes
thermochimiques et associe aux changements climatiques. Les observations notees ouvrent
plutot des pistes d ’explorations futures afin de poursuivre l’analyse des mecanismes ayant pu
survenir et, potentiellement, d ’en tirer diverses applications.

7.2 Contributions:
Cette etude a premierement contribue a foumir des donnees visant une production dediee de
biochar et non seulement a la valorisation d ’un residu de pyrolyse. Ainsi, le fait de cibler des
conditions de production de biocharbon applicables au developpement d ’unites portables et
visant un niveau de productivity interessant pour la chaine de valorisation de la biomasse
industrielle est une forme de contribution en soi. Comme il a ete presente au courant de
1’etude, la communaute scientifique a recemment reconnu que la production de biochar par
pyrolyse lente (regulierement utilisee pour le biochar), et de charbon vert par torrefaction,
beneficieraient d ’etre effectuees a des temps de residence plus courts afin d’atteindre de
meilleurs niveaux de productivity, tout en maintenant les caracteristiques souhaitees. En ce
sens, la conception et l’usage de 2 montages de laboratoire originaux avec reacteurs
parametrables operes a de telles conditions (2.5 min de temps de residence du biocharbon;
taux de chauffe modere a eleve; convection forcee par un vecteur gazeux; temperatures de
basse severite de 300, 400 et 500 °C; usage de CO2 vs N 2 , et usage de biomasse non-broyee)
represented egalement une contribution dans le domaine du biocharbon. Certaines
caracteristiques specifiques obtenues, telles que presentees dans la section precedente,
justifient la pertinence d ’etudier la reponse de ces types de biochar en sols, d ’autant plus

151
qu’une partie importante de leurs caracteristiques physicochimiques, dont plusieurs composes
chimiques des biocharbons, sont maintenant presentees dans cette recherche.

L ’analyse des extraits de biocharbon a genere des resultats de caracterisation qui peuvent
grandement contribuer a la recherche sur le biochar en sol et a la pyrolyse en general. La
plupart des tests de biochar effectues dans le domaine agronomique ont habituellement une
caracterisation tres partielle des biocharbons et de ses conditions de production. La
communaute scientifique avait recemment note le besoin d ’investiguer ce type de
caracterisation dans le cas du biochar. Ainsi, l’ensemble de ces identifications foumissent de
1’information pertinente pour etablir des liens entre des effets observes avec le sol et les
plantes a la suite d'une application de biochar. Ces conditions en reacteur sont d’autant plus
une forme de contribution car elles ont mene a l’identification de composes d’interet au point
de vue agronomique, tel le 1-undecanol qui s’est trouve produit dans le reacteur lg /batch, a
300 et 400 °C. De plus, en termes d’hydrocarbures aromatiques polycycliques (HAP), qui sont
habituellement obtenus en pyrolyse et qui represented une classe de composes chimiques qui
pourrait restreindre l’utilisation du biochar a cause de leur toxicite, uniquement le naphtalene
et ses derives ont ete observes (a 500 °C). Le naphtalene est un compose aromatique
polyeyclique compose de 2 cycles aromatiques qui est deja utilise comme ingredient de
pesticides agricoles. L’ensemble des identifications des substances presentes dans les extraits
de biocharbons se retrouve rarement dans le domaine de la pyrolyse en general,
particulierement pour la pyrolyse a basse severite (P ~ atm, T : 300 a 500 °C) et est quasi
inexistant dans le cas des biocharbons produits en environnement de CO2 . Ainsi, l’analyse de
ces extraits de biocharbons obtenus sous atmosphere de C 0 2 contribue non seulement a
l’etude du biochar, mais aussi a 1’investigation des mecanismes reactionnels s’effectuant lors
de la pyrolyse et particulierement, lorsque presentee de concert avec l’analyse des bio-huiles
(tel qu’effectue dans le chapitre 6).

Les analyses des extractions de biocharbons effectuees de concert avec celles des bio-huiles
ont permis de noter des differences reliees a l’environnements gazeux (C 0 2 vs N2) et meme
d ’appuyer l’hypothese ou la carboxylation semble survenir en presence de C 0 2. La
contribution de l’usage du C 0 2 ne tient pas uniquement de l’analyse des extractions des

152
biocharbons. Recemment, la communaute scientifique a reconnu le besoin d ’explorer
davantage le potentiel de valoriser le CO2 a l’interieur du procede de pyrolyse; ainsi,
l’ensemble des resultats generes dans cet environnement gazeux represente une contribution
notable. L’etude de la comparaison CO2 vs N 2 ne semble pas avoir ete rapportee pour la
pyrolyse de matiere lignocellulosique aux temperatures investiguees dans la presente etude.
Tel que presentees tout au long de cette etude, les differences au niveau des bilans d ’huiles,
des volatils du biocharbon, des contenus en cendres, des valeurs calorifiques et des composes
chimiques identifies pour des conditions d’operation precises contribuent a concretiser l’effet
que peut avoir le remplacement de gaz inertes conventionnels dans le procede de pyrolyse par
l’usage du CO 2 . Plusieurs pistes de travaux futurs ont ete notees lors de l’etude de l’usage du
CO2 ainsi que pour les autres conditions etudiees. Ces pistes de travaux futurs represented
aussi une forme de contribution pour l’avancement de la recherche dans ce domaine.

7.3 Travaux futurs :

L ’interet que peuvent soulever les resultats d’extraction des biocharbons (tel que mentionne
ci-dessus) justifie une etude approfondie de la composition chimique de ces extraits en
fonction des conditions en reacteur, et ce, autant pour l’usage du biocharbon comme biochar
que pour le charbon vert utilise comme matiere premiere pour des transformations
subsequentes. Pour la recherche fondamentale, poursuivre l ’analyse des extraits de
biocharbons de concert avec celle des bio-huiles apporterait une vue globale, telle que
presentee dans le chapitre 6, et contribuerait a l’avancement des connaissances sur les
mecanismes reactionnels s’effectuant lors de la pyrolyse. Ajouter d ’autres solvants pour viser
une gamme elargie de composes, comme les composes aromatiques, ainsi que ceux a polarite
plus elevee, serait egalement souhaitable pour elargir le spectre des identifications obtenues.
Une quantification complete (avec standards) des produits juges d ’interet parmi ceux identifies
permettrait d’effectuer un pas de plus dans la connaissance des extraits de biocharbons. La
poursuite de l’etude du contenu de ces extraits de biocharbons pourra aussi contribuer a
l’avancement d’etudes toxicologiques sur le biochar et permettra de s’assurer que les
concentrations des composes identifies rencontrent les normes environnementales en vigueur.

153
Le domaine de la torrefaction gagnerait a atteindre des temps de residence plus courts en
reacteurs, tels ceux verifies dans le cadre de ce projet. Ainsi, des etudes plus approfondies des
methodes de transfert de masse et de chaleur pour reduire le temps de reaction et ainsi obtenir
des biocharbons aux caracteristiques souhaitees permettraient de consolider les resultats de
1’etude actuelle.

Pour l’utilisation du CO 2 , les travaux a effectuer sont nombreux dans ce domaine etant donne
tout le potentiel et le besoin environnemental lies a la valorisation du CO 2 residuel. Une
continuation de ces travaux, par rapport aux observations notees, contribuerait autant a la
recherche fondamentale sur la comprehension de mecanismes pyrolytiques qu’a la recherche
appliquee. Une recherche appliquee sur ce gaz pourrait viser l’optimisation de certains
composes ou caracteristiques de produits obtenus. En atmosphere de CO 2 , comme dans tout
autre type d’atmosphere de pyrolyse, les observations faites au niveau du contenu en cendres
(baisse a 500 °C avec CO2 ; chapitre 5) et sur le materiel du reacteur (changement de couleur;
chapitre 5), justifient qu’un suivi accru sur les elements inorganiques soit effectue, et
particulierement sur les elements metalliques, sachant l’effet catalytique possible de certains
d’entre eux lors de reactions de pyrolyse.

Toujours en terme d ’environnement gazeux, explorer davantage l’effet du contact du

biocharbon avec Pair a la sortie du reacteur (il a ete demontre que la surface specifique
pouvait etre augmentee; chapitre 5), serait d’interet, et ce, particulierement pour une
application industrielle ou de grands volumes de biomasse traites seraient tres probablement
mis en contact avec Pair a la sortie des reacteurs.

Des essais agronomiques de biochar en sols a partir d’un reacteur parametrable preparant le
biochar en conditions controlees, tout en effectuant le suivi des caracteristiques obtenues,
seraient parmi les travaux futurs a effectuer. Le charbon de biomasse est un terme generique
correspondant a un materiel dont les caracteristiques peuvent varier immensement. Le sol qui
refoit le biochar peut quant a lui varier autant ou meme davantage. Pouvoir effectuer des tests
en sols avec du biocharbon qui aurait ete obtenu dans des conditions controlees, et dont les
caracteristiques pertinentes auraient ete determinees, permettrait ainsi d ’etablir certaines

154
correlations aujourd’hui encore meconnues. Viser des types de sols plus pauvres dont la
fertilite gagnerait a etre amelioree serait d ’ailleurs vivement d ’interet et particulierement par
rapport aux enjeux auxquels nos societes actuelles font face. Ces enjeux incluent le besoin
d’optimiser l’usage des terres arables prioritairement pour la production agro-alimentaire
mondiale requise, tout en considerant les besoins en bio-produits ainsi que les biocarburants de
fa9 on responsable.

Bien que plusieurs pistes restent encore a explorer pour ce produit regulierement considere
comme un residu, ce dernier semble de plus en plus se presenter comme un « Vecteur de
developpement durable dans la chaine de valorisation de la biomasse ».

155
ANNEXE A PHOTOGRAPHIES DU MONTAGE
ET REACTEUR 1 gbiomassJbatch

Figure A. 1 : Prechauffeur du support a biomasse.

Figure A.2 : Support a biomasse (et biocharbon) installe sur un thermocouple.

156
Figure A.3 : Support a biomasse insere dans le reacteur (lg /batch).

157
ANNEXE B SCHEMA ET PHOTOGRAPHIES
DU MONTAGE ET REACTEUR 25 gbiomassJbatch

a c q u is itio n 6e d o n n e e s

Figure B .l : Schema du reacteur 25 gbi, Jbatch.

158
Figure B.2 : Reacteur 25gbiomassJbatch sans isolation.

159
Figure B.3 : Reacteur 25gbiomaSse/batch avec isolation.

160
Figure B.4 : Condenseurs des bio-huiles installes pour le montage 25 gbiomass Jbatch

161
Figure B.5 : Condenseurs des bio-huiles pour le montage 25 gbiomassJbatch.

Figure B.6 : Filtre suivant les condenseurs de bio-huiles.

162
ANNEXEC EXEMPLE DE CALCUL POUR
LES VALEURS DU TABLEAU 3.3

Exemple presente pour la perte en carbone elementaire (C) dans le biocharbon obtenu a
300°C comparativement au panic erige non pyrolyse. Exemple representatif des autres
pertes presentees dans le tableau 3.3.

1) Transfert des donnees sur une base de panic erige (exemple presente pour biocharbon
300°C -2.5m in)

( - • 3 o o ° c -2 .5 m in b.s. = contenu en C char(tab. 3.1) * ren d em en t char (tab. 3.1)

^'3 o o ° c - 2 .5 min b.s. = 51.3 g C/lOOg char * 70.9 g char/lO O g panic

^ 3 o o ° c - 2 .5 min b.s. = 36.4 g C /1 0 0 g panic

2) Moyenne des contenus en carbone des biocharbons obtenus a 300 °C 2.5 min et 5 min

N.b. : tel que mentionne en section 3.3.1, cette analyse de pertes s’est effectuee
independamment du temps de residence en reacteur etant donne que les resultats
pour les analyses immediates et elementaires, en fonction du temps de residence,
n’etaient pas significatifs (a = 0.05).

n _ £300°C-2.5m m b.s. "b ^'300°C-5min b.s.

L300°C b.s. ~ 2

36.4 g C/lOOg panic + 33.9 g C/lOOg panic

Qoc)°C b.s. = 2

^ 3 oo°c b.s. = 35.2 g C/lOOg panic

3) Difference entre contenu en carbone dans le biocharbon et dans le panic (exemple pour
300°C char)

&Cpanic-3 oo°c b.s. = 44.5 g C /1 0 0 g panic erig e — 35.2 g C/lOOg panic erige

h C P a n ic -3 0 0 ° c b.s. = 9.3 g C /1 0 0 g panic erige

163
ANNEXE D COMPOSITION ELEMENTAIRE
METALLIQUE DE PANIC ERIGE

Tableau D.l : Composition en elements metalliques du panic erige*

Contenu
Element ■
ppm
Na 158
Mg 542
Al 102
K 717
Ca 6173
Mn 41
Fe 113
Cu 2
Mo 0
*Obtenu par ICPES; Source : Fahmi et al. [2007]

164
REFERENCES
Almeida, G., Brito, J. 0 . et Perre, P. (2010). Alterations in energy properties o f eucalyptus
wood and bark subjected to torrefaction: The potential o f mass loss as a synthetic
indicator. Bioresource Technology, volume 101, numero 24, p. 9778-9784.
Amen-Chen, C., Pakdel, H. et Roy, C. (2001). Production o f monomeric phenols by
thermochemical conversion of biomass: a review. Bioresource Technology, volume 79,
numero 3, p. 277-299.
Amonette, J. E. et Joseph, S. (2009). Characteristics o f biochar: microchemical properties.
Dans Lehmann, J. et Joseph, S., Biochar fo r Environmental Management; Science and
Technology, l ere edition. Earthscan, London et Washington DC, UK et USA, p. 33-52.
Antal, M. J. et Gronli, M. (2003). The art, science, and technology of charcoal production.
Industrial & Engineering Chemistry Research, volume 42, numero 8, p. 1619-1640.
Baldock, J. A. et Smemik, R. J. (2002). Chemical composition and bioavailability o f thermally
altered Pinus resinosa (Red pine) wood. Organic Geochemistry, volume 33, numero 9,
p. 1093-1109.
Beauchet, R., Monteil-Rivera, F. et Lavoie, J.M. (2012). Conversion o f lignin to aromatic-
based chemicals (L-chems) and biofuels (L-fuels). Bioresource Technology, volume
121, numero 24, p. 328-334.
Bergman, P. C. A. et Kiel, J. H. A. (2005). Torrefaction fo r biomass upgrading (ECN-RX-05-
180; http://www.ecn.nl/docs/library/report/2005/rx05180.pdf). Energy research Centre
o f the Netherlands, Petten, The Netherlands, 8 p.
Bernardo, M., Lapa, N., Gonfalves, M., Barbosa, R., Mendes, B., Pinto, F. et Gulyurtlu, I.
(2010). Toxicity of char residues produced in the co-pyrolysis o f different wastes.
Waste Management, volume 30, numero 4, p. 628-635.
Bernardo, M. S., Lapa, N., Barbosa, R., Gon9 alves, M., Mendes, B., Pinto, F. et Gulyurtlu, I.
(2009). Chemical and ecotoxicological characterization o f solid residues produced
during the co-pyrolysis of plastics and pine biomass. Journal o f Hazardous Materials,
volume 166, numero 1, p. 309-317.
Biagini, E., Simone, M. et Tognotti, L. (2009). Characterization o f high heating rate chars of
biomass fuels. Proceedings o f the Combustion Institute, volume 32, numero 2, p. 2043-
2050.
Boateng, A. A. (2007). Characterization and thermal conversion of charcoal derived from
fluidized-bed fast pyrolysis oil production of switchgrass. Industrial & Engineering
Chemistry Research, volume 46, numero 26, p. 8857-8862.
Boateng, A. A., Daugaard, D. E., Goldberg, N. M. et Hicks, K. B. (2007). Bench-scale
fluidized-bed pyrolysis o f switchgrass for bio-oil production. Industrial & Engineering
Chemistry Research, volume 46, numero 7, p. 1891-1897.
Borman, G. L. et Ragland, K. W. (1998). Combustion Engineering, l ere edition. McGraw Hill,
Boston, USA, 613 p.
Bourgois, J., Bartholin, M. C. et Guyonnet, R. (1989). Thermal treatment o f wood: analysis of
the obtained product. Wood Science and Technology, volume 23, numero 4, p. 303-
310.
Bourke, J., Manley-Harris, M., Fushimi, C., Dowaki, K., Nunoura, T. et Antal, M. J. (2007).
Do all carbonized charcoals have the same chemical structure? 2. A model o f the

165
 chemical structure o f carbonized charcoal. Industrial & Engineering Chemistry
Research, volume 46, numero 18, p. 5954-5967.
Brady, N. C. et Weil, R. R. (2002). The Nature and Properties o f Soils, 13th edition. Prentice
Hall, Upper Saddle River, New Jersey, USA, 960 p.
Branca, C., Di Blasi, C. et Galgano, A. (2010). Pyrolysis o f corncobs catalyzed by zinc
chloride for furfural production. Industrial & Engineering Chemistry Research,
volume 49, numero 20, p. 9743-9752.
Brewer, C. E., Schmidt-Rohr, K., Satrio, J. A. et Brown, R. C. (2009). Characterization of
biochar from fast pyrolysis and gasification systems. Environmental Progress &
Sustainable Energy, volume 28, numero 3, p. 386-396.
Bridgeman, T. G., Darvell, L. I., Jones, J. M., Williams, P. T., Fahmi, R., Bridgwater, A. V.,
Barraclough, T., Shield, I., Yates, N., Thain, S. C. et Donnison, I. S. (2007). Influence
of particle size on the analytical and chemical properties o f two energy crops. Fuel,
volume 86, numero 1-2, p. 60-72.
Bridgeman, T. G., Jones, J. M., Shield, I. et Williams, P. T. (2008). Torrefaction o f reed
canary grass, wheat straw and willow to enhance solid fuel qualities and combustion
properties. Fuel, volume 87, numero 6, p. 844-856.
Bridgwater, A. V. et Peacocke, G. V. C. (2000). Fast pyrolysis processes for biomass.
Renewable and Sustainable Energy Reviews, volume 4, numero 1, p. 1-73.
Brown, R. (2009). Biochar production technology. Dans Lehmann, J. et Joseph, S., Biochar
fo r Environmental Management; Science and Technology, l ere edition. Earthscan,
London et Washington DC, UK et USA, p. 127-146.
Buhre, B. J. P., Elliott, L. K., Sheng, C. D., Gupta, R. P. et Wall, T. F. (2005). Oxy-fuel
combustion technology for coal-fired power generation. Progress in Energy and
Combustion Science, volume 31, numero 4, p. 283-307.
Bureau de la traduction (2012). Termium. http://www.btb.termiumplus.gc.ca/tpv2alpha/alpha-
ffa.html?lang=fra&i=&index=ent& index=ent&srchtxt=char (page consultee le 7
avril 2012).
Butterman, H. C. et Castaldi, M. J. (2008). CO 2 enhanced steam gasification o f biomass fuels.
Dans Proceedings o f the 16th Annual North American Waste to Energy Conference,
NAWTEC16, ASME, Philadelphia, PA, USA, p. 157-172.
Butterman, H. C. et Castaldi, M. J. (2009). Syngas production via CO2 enhanced gasification
of biomass fuels. Environmental Engineering Science, volume 26, numero 4, p. 703-
713.
Butterman, H. C. et Castaldi, M. J. (2010). Biomass to fuels: impact of reaction medium and
heating rate. Environmental Engineering Science, volume 27, numero 7, p. 539-555.
Cherubini, F. et Jungmeier, G. (2010). LCA of a biorefinery concept producing bioethanol,
bioenergy, and chemicals from switchgrass. The International Journal o f Life Cycle
Assessment, volume 15, numero 1, p. 53-66.
Chomet, E. (2005). Thermochemical strategies for biofuels, green chemicals, polymeric
biomaterials and biofuels. Dans Bio-refinery workshop, Edmonton and Montreal
workshop on bio-refineries; Capturing Canada's natural advantage, Paprican, Pointe-
Claire, Qc, Canada.
Comite d'etude des termes techniques fran9 ais (2001). Termes techniques fran?ais : essai
d'orientation de la terminologie. Dans Bureau de la traduction, Termium.
http://www.btb.termiumplus.gc.ca/tpv2alpha/alpha-fra.html?lang=fra&i=l&index=frt
& index=frt&srchtxt=cokage (page consultee le 14 mai 2012).

166
Czemik, S. et Bridgwater, A. V. (2004). Overview o f applications of biomass fast pyrolysis
oil. Energy & Fuels, volume 18, numero 2, p. 590-598.
David, K. et Ragauskas, A. J. (2010). Switchgrass as an energy crop for biofuel production: A
review of its ligno-cellulosic chemical properties. Energy and Environmental Science,
volume 3, numero 9, p. 1182-1190.
Demirbas, A. (2000). Mechanisms of liquefaction and pyrolysis reactions o f biomass. Energy
Conversion and Management, volume 41, numero 6, p. 633-646.
Demirbas, A. (2001). Carbonization ranking of selected biomass for charcoal, liquid and
gaseous products. Energy Conversion and Management, volume 42, numero 10, p.
1229-1238.
Demirbas, A. (2004). Combustion characteristics o f different biomass fuels. Progress in
Energy and Combustion Science, volume 30, numero 2, p. 219-230.
Downie, A., Crosky, A. et Munroe, P. (2009). Physical properties o f biochar. Dans Lehmann,
J. et Joseph, S., Biochar fo r Environmental Management - Sience and Technology, l ere
edition. Earthscan, London et Washington DC, UK et USA, p. 13-32.
Druide (2011). Antidote HD, version 5.1. Druide Informatique inc., Montreal, Canada.
Duan, L., Zhao, C., Zhou, W., Qu, C. et Chen, X. (2009). Investigation on coal pyrolysis in
CO 2 atmosphere. Energy & Fuels, volume 23, numero 7, p. 3826-3830.
EIA (2011). World energy consumption, 1990-2035. Dans U.S. Department o f Energy,
Independent Statistics and Analysis; U.S. Energy Information Administration.
http://www.eia.gov/forecasts/ieo/index.cfm (page consultee le 2012-04-04).
Encinar, J. M., Beltran, F. J., Ramiro, A. et Gonzalez, J. F. (1998). Pyrolysis/gasification of
agricultural residues by carbon dioxide in the presence of different additives: influence
o f variables. Fuel Processing Technology, volume 55, numero 3, p. 219-233.
EPA (2008). Reregistration Eligibility Decision fo r Naphthalene (Prevention, Pesticides and
Toxic Substances, Case no. 3058, EPA 738-R-07-010) United States Environmental
Protection Agency, USA, 71 p.
Fabbri, D., Adamiano, A. et Torri, C. (2010). GC-MS determination o f polycyclic aromatic
hydrocarbons evolved from pyrolysis o f biomass. Analytical and Bioanalytical
Chemistry, volume 397, numero 1, p. 309-317.
Fahmi, R., Bridgwater, A. V., Darvell, L. I., Jones, J. M., Yates, N., Thain, S. et Donnison, I.
S. (2007). The effect o f alkali metals on combustion and pyrolysis o f Lolium and
Festuca grasses, switchgrass and willow. Fuel, volume 86, numero 10-11, p. 1560-
1569.
FAO (2011). World Water Day 2012: Water and Food Security. Dans FAO, FAO Water
Development and Management Unit, http://www.fao.org/nr/water/news/wwdl2.html
(page consultee le 22 juillet 2012).
FAO (2012). La FAO publie de nouvelles orientations pour exploiter le potentiel de la
bioenergie. Dans Centre d ’actualites de l’ONU, Depeches du service d'information des
Nations Unies.
http://www.un.org/apps/newsFr/storyF.asp?NewsID=27724&Cr=FAO&Crl=, (page
consultee le 4 avril 2012).
Felfli, F. F., Luengo, C. A., Suarez, J. A. et Beaton, P. A. (2005). Wood briquette torrefaction.
Energy fo r Sustainable Development, volume 9, numero 3, p. 19-22.
Foley, J. A., Ramankutty, N., Brauman, K. A., Cassidy, E. S., Gerber, J. S., Johnston, M.,
Mueller, N. D., O'Connell, C., Ray, D. K., West, P. C., Balzer, C., Bennett, E. M.,
Carpenter, S. R., Hill, J., Monfreda, C., Polasky, S., Rockstrom, J., Sheehan, J.,

167
 Siebert, S., Tilman, D. et Zaks, D. P. M. (2011). Solutions for a cultivated planet.
Nature, volume 478, numero 7369, p. 337-342.
Foucault, A. et Raoult, J.-F. (1995). Dictionnaire de geologie. Dans Bureau de la traduction,
Termium. http://www.btb.termiumplus.gc.ca/tpv2alpha/alpha-fra.html?lang=fra&i= 1&
index=frt& index=frt&srchtxt=Houille&comencsrch.x=13&comencsrch.y=8 (page
consultee le 19 juin 2012).
Fryda, L., Sobrino, C., Cieplik, M. et van de Kamp, W. L. (2010). Study on ash deposition
under oxyfuel combustion of coal/biomass blends. Fuel, volume 89, numero 8, p.
1889-1902.
Gaunt, J. L. et Lehmann, J. (2008). Energy balance and emissions associated with biochar
sequestration and pyrolysis bioenergy production. Environmental Science <6
Technology, volume 42, numero 11, p. 4152-4158.
GFC (2004). Les mots de la combustion, http://www.gfcombustion.asso.fr/index.php
?item=glo&page=glossaire.php&langue-fr (page consultee le 7 avril 2012).
Gilbert, P., Ryu, C., Sharifi, V. et Swithenbank, J. (2009). Effect of process parameters on
pelletisation of herbaceous crops. Fuel, volume 88, numero 8, p. 1491-1497.
Goodfellow, R. (2009). Pyrolysis: Energy and chemicals from residues. Dans, Atelier Quebec-
Ontario pour explorer les nouvelles opportunites d ’affaires dans le secteur de la
bioenergie par gazeification et methanisation, Centre quebecois de valorisation des
biotechnologies, visioconference Montreal-Samia, Qc-Ontario, Canada.
GRID-Arendal (2001). Bilan 2001 des changements climatiques : Rapport de synthese. Dans
Bureau de la traduction, Termium. http://www.btb.termiumplus.gc.ca/tpv2alpha/alpha-
fra.html?lang=fra&i=l&index=frt& index=frt&srchtxt=Carbone+noir&comencsrch.
x=12&comencsrch.y=10 (page consultee le 19 juin 2012).
Guo, X.-j., Wang, S.-r., Wang, K.-g., Liu, Q. et Luo, Z.-y. (2010). Influence of extractives on
mechanism of biomass pyrolysis. Journal o f Fuel Chemistry and Technology, volume
38, numero 1, p. 42-46.
Hajaligol, M., Waymack, B. et Kellogg, D. (2001). Low temperature formation o f aromatic
hydrocarbon from pyrolysis of cellulosic materials. Fuel, volume 80, numero 12, p.
1799-1807.
Hale, S. E., Lehmann, J., Rutherford, D., Zimmerman, A. R., Bachmann, R. T.,
Shitumbanuma, V., O ’Toole, A., Sundqvist, K. L., Arp, H. P. H., et Comelissen, G.
(2012). Quantifying the total and bioavailable polycyclic aromatic hydrocarbons and
dioxins in biochars. Environmental Science & Technology, volume 46, numero 5, p.
2830-2838.
He, R., Ye, X. P., English, B. C. et Satrio, J. A. (2009). Influence of pyrolysis condition on
switchgrass bio-oil yield and physicochemical properties. Bioresource Technology,
volume 100, numero 21, p. 5305-5311.
Heilmann, S. M., Jader, L. R., Sadowsky, M. J., Schendel, F. J., von Keitz, M. G. et Valentas,
K. J. (2011). Hydrothermal carbonization o f distiller’s grains. Biomass and Bioenergy,
volume 35, numero 7, p. 2526-2533.
Hoekman, S. K., Broch, A. et Robbins, C. (2011).Hydrothermal carbonization (HTC) of
lignocellulosic biomass. Energy and Fuels, volume 25, numero 4, p. 1802-1810.
Imam, T. et Capareda, S. (2012). Characterization o f bio-oil, syn-gas and bio-char from
switchgrass pyrolysis at various temperatures. Journal o f Analytical and Applied
Pyrolysis, volume 93, p. 170-177.

168
Ioannidou, O. et Zabaniotou, A. (2007). Agricultural residues as precursors for activated
carbon production—A review. Renewable and Sustainable Energy Reviews, volume 11,
numero 9, p. 1966-2005.
Jenkins, B. M., Baxter, L. L., Miles Jr, T. R. et Miles, T. R. (1998). Combustion properties of
biomass. Fuel Processing Technology, volume 54, numero 1-3, p. 17-46.
Jin, C., Xia, W., Chen, P. et Muhler, M. (2012). On the role of the residual iron growth
catalyst in the gasification of multi-walled carbon nanotubes with carbon dioxide.
Catalysis Today, volume 186, numero 1, p. 128-133.
Jindarom, C., Meeyoo, V., Kitiyanan, B., Rirksomboon, T. et Rangsunvigit, P. (2007a).
Surface characterization and dye adsorptive capacities of char obtained from
pyrolysis/gasification of sewage sludge. Chemical Engineering Journal, volume 133,
numero 1-3, p. 239-246.
Jindarom, C., Meeyoo, V., Rirksomboon, T., Kitiyanan, B. et Rangsunvigit, P. (2006).
Production of bio-oil by oxidative pyrolysis of sewage sludge in rotating fixed bed
reactor. Asian Journal on Energy and Environment, volume 7, numero 04, p. 401-409.
Jindarom, C., Meeyoo, V., Rirksomboon, T. et Rangsunvigit, P. (2007b). Thermochemical
decomposition o f sewage sludge in CO 2 and N 2 atmosphere. Chemosphere, volume 67,
numero 8, p. 1477-1484.
Joseph, S., Peacocke, C., Lehmann, J. et Munroe, P. (2009). Developing a biochar
classification and tests methods. Dans Lehmann, J. et Joseph, S., Biochar fo r
Environmental Management, l ere edition. Earthscan, London et Washington DC, UK et
USA, p. 107-126.
Jouraiphy, A., Amir, S., El Gharous, M., Revel, J.-C. et Hafidi, M. (2005). Chemical and
spectroscopic analysis of organic matter transformation during composting of sewage
sludge and green plant waste. International Biodeterioration & Biodegradation,
volume 56, numero 2, p. 101-108.
Keiluweit, M. et Kleber, M. (2009). Molecular-level interactions in soils and sediments: the
role of aromatic 7t-systems. Environmental Science & Technology, volume 43, numero
10, p. 3421-3429.
Keiluweit, M., Nico, P. S., Johnson, M. G. et Kleber, M. (2010). Dynamic molecular structure
o f plant biomass-derived b la ck . carbon (biochar). Environmental Science and
Technology, volume 44, numero 4, p. 1247-1253.
Keiluweit, M., Kleber, M., Sparrow, M. A., Simoneit, B. R. T. et Prahl, F. G. (2012). Solvent-
extractable polycyclic aromatic hydrocarbons in biochar: influence of pyrolysis
temperature and feedstock. Environmental Science and Technology, volume 46,
numero 17, p. 9333-9341.
Kocaefe, D., Chaudhry, B., Poncsak, S., Bouazara, M. et Pichette, A. (2007).
Thermogravimetric study o f high temperature treatment o f aspen: effect o f treatment
parameters on weight loss and mechanical properties. Journal o f Materials Science,
volume 42, numero 3, p. 854-866.
Krull, E. S., Baldock, J. A., Skjemstad, J. O. et Smemik, R. J. (2009). Characteristics of
biochar: organo-chemical properties. Dans Lehmann, J. et Joseph, S., Biochar fo r
Environmental Management - Science and Technology, l ere edition. Earthscan, London
et Washington DC, UK et USA, p. 53-65.
Kumar, M., Gupta, R. C. et Sharma, T. (1993). X-ray diffraction studies o f Acacia and
Eucalyptus wood chars. Journal o f Materials Science, volume 28, numero 3, p. 805-
810.

169
Kushwaha, S. C., Clarkson, S. G. et Mehkeri, K. A. (1985). Polycyclic aromatic hydrocarbons
in barbecue briquets. Journal o f Food Safety, volume 7, numero 3, p. 177-201.
L'Encyclopedie de l'Agora (2008). Charbon vert. Dans Bureau de la traduction, Termium.
http://www.btb.termiumplus.gc.ca/tpv2alpha/alpha-fra.html?lang=fra&i=l&index=frt
& index=frt&srchtxt=charbon+vert&comencsrch.x=l 4&comencsrch.y= 14 (page
consultee le 2 aout 2012).
Laird, D. A. (2008). The charcoal vision: a win win win scenario for simultaneously producing
bioenergy, permanently sequestering carbon, while improving soil and water quality.
Agronomy Journal, volume 100, numero 1, p. 178-181.
Laird, D. A., Brown, R. C., Amonette, J. E. et Lehmann, J. (2009). Review o f the pyrolysis
platform for coproducing bio-oil and biochar. Biofuels, Bioproducts and Biorefining,
volume 3, numero 5, p. 547-62.
Larousse (2005). Grand dictionnaire encyclopedique Larousse. Dans Bureau de la traduction,
Termium. http://www.btb.termiumplus.gc.ca/tpv2alpha/alpha-fra.html?lang=fra&i= 1&
index=frt& index=frt&srchtxt=Graphite&comencsrch.x= 11 &comencsrch.y=8 (page
consultee le 14 mai 2012).
Larousse (2010). Grand dictionnaire encyclopedique Larousse. Dans Bureau de la traduction,
Termium. http://www.btb.termiumplus.gc.ca/tpv2alpha/alpha-fra.html?lang=fra&i= 1&
index=frt& index=frt&srchtxt=Coke&comencsrch.x= 13 &comencsrch.y=9 (page
consultee le 14 mai 2012).
Lee, S.-M., Jin, L., Kim, J., Han, S., Na, H., Hyeon, T., Koo, Y.-M., Kim, J. et Lee, J.-H.
(2010). P-Glucosidase coating on polymer nanofibers for improved cellulosic ethanol
production. Bioprocess and Biosystems Engineering, volume 33, numero 1, p. 141-
147.
Lehmann, J. (2007). A handful of carbon. Nature, volume 447, numero 7141, p. 143-144.
Lehmann, J., Gaunt, J. et Rondon, M. (2006). Bio-char sequestration in terrestrial ecosystems
- a review. Mitigation and Adaptation Strategies fo r Global Change, volume 11,
numero 2, p. 395-419.
Lehmann, J. et Joseph, S. (2009). Biochar fo r Environmental Management; Science and
Technology, l ere edition. Earthscan, London et Washington, DC, UK et USA, 416 p.
Lemus, R., Brummer, E. C., Moore, K. J., Molstad, N. E., Burras, C. L. et Barker, M. F.
(2002). Biomass yield and quality o f 20 switchgrass populations in southern Iowa,
USA. Biomass and Bioenergy, volume 23, numero 6, p. 433-442.
Lide, D. R. (2008). CRC Handbook o f Chemistry and Physics, 88e edition (CD-ROM Version
2008). CRC Press/Taylor and Francis, Boca Raton, FL, USA.
Lv, G.-j., Wu, S.-b. et Lou, R. (2010). Characteristics of com stalk hemicellulose pyrolysis in
a tubular reactor. BioResources, volume 5, numero 4, p. 2051-2062.
Maher, K. D., Kirkwood, K. M., Gray, M. R. et Bressler, D. C. (2008). Pyrolytic
decarboxylation and cracking o f stearic acid. Industrial & Engineering Chemistry
Research, volume 47, numero 15, p. 5328-5336.
Major, J., Steiner, C., Downie, A. et Lehmann, J. (2009). Biochar effects on nutrient leaching.
Dans Lehmann, J. et Joseph, S., Biochar fo r Environmental Management; Science and
Technology, l ere edition. Earthscan, London et Washington, DC, UK et USA, p. 271 -
287.
Marris, E. (2006). Putting the carbon back: Black is the new green. Nature, volume 442,
numero 7103, p. 624-626.

170
Martel, H. et Perron, M.-H. (2008). Compilation des essais de Panic Erige realises au
Quebec. Centre de reference en agriculture et agroalimentaire du Quebec, Quebec,
Canada, 195 p.
McGrath, T., Sharma, R. et Hajaligol, M. (2001). An experimental investigation into the
formation of polycyclic-aromatic hydrocarbons (PAH) from pyrolysis of biomass
materials. Fuel, volume 80, numero 12, p. 1787-1797.
McGrath, T. E., Chan, W. G. et Hajaligol, M. R. (2003). Low temperature mechanism for the
formation of polycyclic aromatic hydrocarbons from the pyrolysis o f cellulose. Journal
o f Analytical and Applied Pyrolysis, volume 66, numero 1-2, p. 51-70.
McKendry, P. (2002). Energy production from biomass (part 2): conversion technologies.
Bioresource Technology, volume 83, numero 1, p. 47-54.
McLaughlin, H., Anderson, P. S., Shields, F. E. et Reed, T. B. (2009). All biochars are not
created equal, and how to tell them apart. Dans International Biochar Initiative, IBI
Publications, http://www.biochar-intemational.org/sites/default/files/All-Biochars—
V ersion2-0ct2009.pdf (page consultee le 10 decembre 2010).
Minkova, V., Marinov, S. P., Zanzi, R., Bjombom, E., Budinova, T., Stefanova, M. et Lakov,
L. (2000). Thermochemical treatment of biomass in a flow of steam or in a mixture of
steam and carbon dioxide. Fuel Processing Technology, volume 62, numero 1, p. 45-
52.
Mohan, D., Pittman, C. U. et Steele, P. H. (2006). Pyrolysis o f wood/biomass for bio-oil: A
critical review. Energy & Fuels, volume 20, numero 3, p. 848-889.
Mullen, C. A. et Boateng, A. A. (2008). Chemical composition o f bio-oils produced by fast
pyrolysis of two energy crops. Energy & Fuels, volume 22, numero 3, p. 2104-2109.
Nakajima, D., Nagame, S., Kuramochi, H., Sugita, K., Kageyama, S., Shiozaki, T., Takemura,
T., Shiraishi, F. et Goto, S. (2007). Polycyclic aromatic hydrocarbon generation
behavior in the process o f carbonization of wood. Bulletin o f Environmental
Contamination and Toxicology, volume 79, numero 2, p. 221-225.
Naredi, P. et Pisupati, S. (2011). Effect of CO 2 during coal pyrolysis and char bumout in oxy-
coal combustion. Energy & Fuels, volume 25, numero 6, p. 2452-2459.
Nikoo, M. K. et Amin, N. A. S. (2011). Thermodynamic analysis of carbon dioxide reforming
of methane in view o f solid carbon formation. Fuel Processing Technology, volume
92, numero 3, p. 678-691.
Olivier, M. J. (2007). Chimie de I'environnement, 5e edition, Les productions Jacques Bernier,
Quebec, Canada, 315 p.
Onay, O. (2007). Influence of pyrolysis temperature and heating rate on the production of bio­
oil and char from safflower seed by pyrolysis, using a well-swept flxed-bed reactor.
Fuel Processing Technology, volume 88, numero 5, p. 523-531.
ONU (2011). World population prospects: The 2010 revision. Dans ONU, United Nations,
Department o f Economic and Social Affairs, Population Division, Population
Estimates and Projections Sections. http://esa.un.org/wpp/Analytical-
Figures/htm/fig_l.htm (page consultee le 4 avril 2012).
Osaki, T. et Mori, T. (2006). Kinetics o f the reverse-Boudouard reaction over supported nickel
catalysts. Reaction Kinetics and Catalysis Letters, volume 89, numero 2, p. 333-339.
Pandey, M. P. et Kim, C. S. (2011). Lignin depolymerization and conversion: A review of
thermochemical methods. Chemical Engineering & Technology, volume 34, numero 1,
p. 29-41.

171
Paris, O., Zollfrank, C. et Zickler, G. A. (2005). Decomposition and carbonisation o f wood
biopolymers—a microstructural study of softwood pyrolysis. Carbon, volume 43,
numero 1, p. 53-66.
Pettersson, A., Amand, L.-E. et Steenari, B.-M. (2008). Leaching o f ashes from co-combustion
of sewage sludge and w ood-Part II: The mobility of metals during phosphorus
extraction. Biomass and Bioenergy, volume 32, numero 3, p. 236-244.
Phanphanich, M. et Mani, S. (2011). Impact o f torrefaction on the grindability and fuel
characteristics o f forest biomass. Bioresource Technology, volume 102, numero 2, p.
1246-1253.
Pilon, G. et Lavoie, J.-M. (201 la). Biomass char production at low severity conditions under
CO2 and N 2 environments. Dans Villacampa, Y., Mammoli, A. A. et Brebbia, C. A.,
Energy and Sustainability III, WIT Press, Ashurst, Southampton, UK, p. 109-121.
Pilon, G. et Lavoie, J.-M. (2011b). Characterization o f switchgrass char produced in
torrefaction and pyrolysis conditions. BioResources, volume 6, numero 4, p. 4824-
4839.
Pilon, G. et Lavoie, J.-M. (2013a). Pyrolysis of switchgrass (Panicum virgatum L.) at low
temperatures within N2 and C 0 2 environments: product yield study. ACS - Sustainable
Chemistry and Engineering, volume 1, numero 1, p. 198-204.
Pilon, G. et Lavoie, J.-M. (2013b). Pyrolysis of switchgrass (Panicum virgatum L.) at low
temperatures in N 2 and C 0 2 environments; a study on chemical composition o f chars
extracts and bio-oils. Journal o f Analytical and Applied Pyrolysis, volume 101, p. 122-
131.
Prfns, M. J., Ptasinski, K. J. et Janssen, F. J. J. G. (2006a). More efficient biomass gasification
via torrefaction. Energy, volume 31, numero 15, p. 3458-3470.
Prins, M. J., Ptasinski, K. J. et Janssen, F. J. J. G. (2006b). Torrefaction of wood: Part 1.
Weight loss kinetics. Journal o f Analytical and Applied Pyrolysis, volume 77, numero
l , p. 28-34.
Prins, M. J., Ptasinski, K. J. et Janssen, F. J. J. G. (2006c). Torrefaction of wood: Part 2.
Analysis of products. Journal o f Analytical and Applied Pyrolysis, volume 77, numero
l , p. 35-40.
Reed, T. B. (1981). Biomass Gasification; Principles and Technology. Solar Energy Research
Institute, Park Ridge, New Jersey, USA, 401 p.
Repellin, V., Govin, A., Rolland, M. et Guyonnet, R. (2010). Modelling anhydrous weight loss
of wood chips during torrefaction in a pilot kiln. Biomass and Bioenergy, volume 34,
numero 5, p. 602-609.
Roos, C. J. (2008). Biomass drying and dewatering fo r clean heat & power (WSUEEP08-
015). Northwest CHP Application Center, Olympia, WA, USA, 35 p.
Rousset, P., Davrieux, F., Macedo, L. et Perre, P. (2011). Characterisation of the torrefaction
o f beech wood using NIRS: Combined effects of temperature and duration. Biomass
and Bioenergy, volume 35, numero 3, p. 1219-1226.
Rousset, P., Lapierre, C., Pollet, B., Quirino, W. et Perre, P. (2009). Effect of severe thermal
treatment on spruce and beech wood lignins. Annals o f Forest Science, volume 66,
numero 1, p. 110.
Rutherford, D. W., Rostad, C. E. et Wershaw, R. L. (2010). Sorption of water by biochars.
Dans American Chemical Society, 239th ACS National Meeting.
http://envirofacs.org/symposia/239_san%20francisco/finalprogram_for_posting.pdf
(page consultee le 18 aout 2012).

172
Sadaka, S. et Negi, S. (2009). Improvements o f biomass physical and thermochemical
characteristics via torrefaction process. Environmental Progress and Sustainable
Energy, volume 28, numero 3, p. 427-434.
Samson, R., Mani, S., Boddey, R., Sokhansanj, S., Quesada, D., Urquiaga, S., Reis, V. et Ho
Lem, C. (2005). The potential o f C 4 perennial grasses for developing a global bioheat
industry. Critical Reviews in Plant Sciences, volume 24, numero 5/6, p. 461-495.
Sannigrahi, P., Pu, Y. et Ragauskas, A. (2010). Cellulosic biorefmeries-unleashing lignin
opportunities. Current Opinion in Environmental Sustainability, volume 2, numero 5-
6, p. 383-393.
Sarath, G., Baird, L. M., Vogel, K. P. et Mitchell R. B. (2007). Intemode structure and cell
wall composition in maturing tillers o f switchgrass (.Panicum virgatum L.).
Bioresource Technology, volume 98, p. 2985-2992.
SciFinder (2011). Regulatory information detail; Cas registry number: 112-42-5. Dans
SciFinder, https://scifmder-cas-org.ezproxy.usherbrooke.ca/scifinder/view/scifmder/
scifinderExplore.jsf (page consultee le 18 aout 2012).
Schmidt, M. W. I. et Noack, A. G. (2000). Black carbon in soils and sediments: Analysis,
distribution, implications, and current challenges. Global Biogeochemical Cycles,
volume 14, numero 3, p. 777-793.
Sensoz, S. et Kaynar, I. (2006). Bio-oil production from soybean (Glycine max L.); fuel
properties of bio-oil. Industrial Crops and Products, volume 23, numero 1, p. 99-105.
Sharma, R. K. et Hajaligol, M. R. (2003). Effect of pyrolysis conditions on the formation of
polycyclic aromatic hydrocarbons (PAHs) from polyphenolic compounds. Journal o f
Analytical and Applied Pyrolysis, volume 66, numero 1-2, p. 123-144.
Sharma, R. K., Wooten, J. B., Baliga, V. L., Lin, X., Geoffrey Chan, W. et Hajaligol, M. R.
(2004). Characterization of chars from pyrolysis o f lignin. Fuel, volume 83, numero
11-12, p. 1469-1482.
Sharma, R. K., Wooten, J. B., Baliga, V. L., Martoglio-Smith, P. A. et Hajaligol, M. R.
(2002). Characterization of char from the pyrolysis of tobacco. Journal o f Agricultural
and Food Chemistry, volume 50, numero 4, p. 771-783.
Shin, E.-J., Nimlos, M. R. et Evans, R. J. (2001). A study o f the mechanisms of vanillin
pyrolysis by mass spectrometry and multivariate analysis. Fuel, volume 80, numero
12, p. 1689-1696.
Sluiter, A., Ruiz, R., Scarlata, C., Sluiter, J. et Templeton, D. (2008). Determination o f
Extractives in Biomass (NREL/TP-510-42619). National Renewable Energy
Laboratory, Golden, Colorado, USA, 12 p.
Smemik, R. J. (2009). Biochar and sorption of organic compounds. Dans Lehmann, J. et
Joseph, S., Biochar fo r Environmental Management; Science and Technology, l 6rc
edition. Earthscan, London et Washington, DC, UK et USA, p. 289-300.
Sun, Y. et Cheng, J. (2002). Hydrolysis of lignocellulosic materials for ethanol production: A
review. Bioresource Technology, volume 83, numero 1, p. 1-11.
Teng, H., Ho, J. A. et Hsu, Y. F. (1997). Preparation of activated carbons from bituminous
coals with C 0 2 activation-Influence o f coal oxidation. Carbon, volume 35, numero 2,
p. 275-283.
The Engineering Toolbox (2012). Classification o f coal, http://www.engineeringtoolbox.com/
classification-coal-d_164.html (page consultee le 3 avril 2012).
Thies, J. E. et Rillig, M. C. (2009). Characteristics of biochar: biological properties. Dans
Lehmann, J. et Joseph, S., Biochar fo r Environmental Management; Science and

173
 Technology, l ere edition. Earthscan, London et Washington, DC, UK et USA, p. 85-
105.
Uchimiya, M., Wartelle, L. H., Klasson, K. T., Fortier, C. A. et Lima, I. M. (2011). Influence
o f pyrolysis temperature on biochar property and function as a heavy metal sorbent in
soil. Journal o f Agricultural and Food Chemistry, volume 59, numero 6, p. 2501-2510.
UNCTAD (2012). The Biofuels Controversy. Dans Bureau de la traduction, Termium.
http://www.unctad.org (page consultee le 2012 08 02).
Uslu, A., Faaij, A. P. C. et Bergman, P. C. A. (2008). Pre-treatment technologies, and their
effect on international bioenergy supply chain logistics. Techno-economic evaluation
o f torrefaction, fast pyrolysis and pelletisation. Energy, volume 33, numero 8, p. 1206-
1223.
Vassilev, S. V., Baxter, D., Andersen, L. K. et Vassileva, C. G. (2010). An overview o f the
chemical composition of biomass. Fuel, volume 89, numero 5, p. 913-933.
Wang, S., Guo, X., Wang, K. et Luo, Z. (2011). Influence o f the interaction of components on
the pyrolysis behavior of biomass. Journal o f Analytical and Applied Pyrolysis,
volume 91, numero 1, p. 183-189.
Weiss, N. D., Farmer, J. D. et Schell, D. J. (2010). Impact of com stover composition on
hemicellulose conversion during dilute acid pretreatment and enzymatic cellulose
digestibility of the pretreated solids. Bioresource Technology, volume 101, numero 2,
p. 674-678.
White, J. E., Catallo, W. J. et Legendre, B. L. (2011). Biomass pyrolysis kinetics: A
comparative critical review with relevant agricultural residue case studies. Journal o f
Analytical and Applied Pyrolysis, volume 91, numero 1, p. 1-33.
Willner, T. et Brunner, G. (2005). Pyrolysis kinetics o f wood and wood components.
Chemical Engineering & Technology, volume 28, numero 10, p. 1212-1225.
Yan, J., Hu, Z., Pu, Y., Charles Brummer, E. et Ragauskas, A. J. (2010). Chemical
compositions o f four switchgrass populations. Biomass and Bioenergy, volume 34,
numero 1, p. 48-53.
Yang, H., Yan, R., Chen, H., Zheng, C., Lee, D. H. et Liang, D. T. (2006). In-depth
investigation o f biomass pyrolysis based on three major components: Hemicellulose,
cellulose and lignin. Energy and Fuels, volume 20, numero 1, p. 388-393.
Zanzi, R., Sjostrom, K. et Bjombom, E. (1996). Rapid high-temperature pyrolysis o f biomass
in a free-fall reactor. Fuel, volume 75, numero 5, p. 545-550.
Zhang, H., Xiao, R., Wang, D., He, G., Shao, S., Zhang, J. et Zhong, Z. (2011). Biomass fast
pyrolysis in a fluidized bed reactor under N2, CO2, CO, CH4 and H2 atmospheres.
Bioresource Technology, volume 102, numero 5, p. 4258-4264.
Zhang, Q., Chang, J., Wang, T. et Xu, Y. (2007). Review o f biomass pyrolysis oil properties
and upgrading research. Energy Conversion and Management, volume 48, numero 1,
p. 87-92.
Zwart, R. W. R., Boerrigter, H. et van der Drift, A. (2006). The impact o f biomass
pretreatment on the feasibility of overseas biomass conversion to Fischer-Tropsch
products. Energy & Fuels, volume 20, numero 5, p. 2192-2197.

174