In Proceedings of SAMPE International Symposium & Exhibition: Advancing Materials in the Global Economy

Applications, Emerging Markets and Evolving Technologies, May 11-15, 2003, Long Beach, California, p. 2004.

X-RAY CURING EPOXY ADHESIVE SYSTEMS

Andrea E. Hoyt Haight, Larry A. Harrah, and Melissa R. Sprouse
Adherent Technologies, Inc.
Development Laboratories
11208 Cochiti SE
Albuquerque, NM 87123

ABSTRACT
X-ray curing adhesive systems offer significant potential for use in a variety of situations where
heating is either difficult or undesirable. Such applications include aerospace repair, where
difficult-to-reach areas and large penetration depths limit the utility of e-beam curing and
detector assembly for high energy physics, where a room temperature cure is desired to minimize
stresses. Adherent Technologies has been developing epoxy-based adhesive systems modified to
promote x-ray initiated cure at relatively low doses. These systems contain a variety of fillers to
increase the x-ray dose within the bond area. Several systems have been tested that exhibit neat
resin cure at doses as low as 20kGy. Lap shear samples on G10 composite substrates have also
been prepared and tested. Performance similar to e-beam cured cationic film adhesives was
obtained.

KEY WORDS: Adhesives, Epoxy Resin, Radiation Processing

1. INTRODUCTION
Radiation curing resin systems are desirable for many applications because, in contrast to
conventional adhesive systems, polymerization can be initiated at a point specified by the user.
This allows for sufficient time to ensure adequate assembly of parts and can allow for targeted
cure initiation. Additionally, the shelf- life of the adhesive system is often greatly improved.
UV- initiated epoxy systems are an excellent example of this type of system. However, there are
many cases in which UV- initiated polymerizations are not applicable. Applications of this
nature are frequently found in the aerospace industry, where large, complex parts are frequently
adhesively bonded.

X-ray irradiation is not widely used in materials processing despite the relatively high
penetration depth relative to UV light or electron beams. However, in cases such as bonding
operations in the aerospace industry, deep penetration would be required.

Adherent Technologies has developed an epoxy-based adhesive system that will cure under
exposure to x-ray irradiation. Epoxy resins are familiar to most people as the standard two-part
system that after mixing cure to brittle solids within anywhere from five minutes to 24 hours. In
these systems the epoxy cures by reacting with another compound (incorrectly referred to as a
catalyst, actually a hardener). These compounds are frequently polymeric amines or organic
anhydrides. Epoxy compounds may also react with themselves, however, in a process called
homopolymerization. In this reaction, epoxy groups react with themselves in a chain-step
process (propagation) proceeding via a tertiary oxonium ion, which additionally generates a
living polymer. The term living polymer indicates that once the reaction is started, it continues
until all the epoxy chemistry is consumed and complete cure of the resin has been achieved.
This is sometimes referred to as a dark cure.

This chemistry requires that a small amount (0.5 to 3%) of a catalyst be present to initiate the
reaction (generate an oxonium ion). These catalysts are blended with the epoxy to give a one-
part composition. These catalysts can be thermally activated, such as dicyandiamide or boron
trifluoride MEA complex, or may be radiation activated. Typical examples of radiation activated
catalysts, also called cationic photoinitiators, include the families of triaryl sulfonium salts and
diaryl iodonium salts. These compounds are stable at room temperature, readily soluble in most
epoxy resins, do not activate epoxy cure until exposed to UV light or other ionizing radiation
sources, are insensitive to room lighting, have indefinite storage life, and initiate a living polymer
cure so that the entirety of the epoxy hardens.

1.1 Formulation for X-Ray Curing

A test matrix of resins was prepared based on cationic epoxy resins formulated by Adherent
Technologies for UV cure applications. All resins were compounded with a non-reactive filler in
order to produce materials of sufficient viscosity to allow for easy adhesive application from a
filled syringe. Additional modifications to the resins included the addition of fine metal powders
or metal salts in order to enhance the x-ray dose applied to the adhesive during the irradiation
process. The logic of this approach is discussed below.

The base formulations tested were as follows:

v AT-924F: a fast-curing UV initiated epoxy resin that provides hard coatings with good
durability, good temperature and chemical resistance, and low shrinkage. AT-924 is a
commercial product of Adherent Technologies.

v AT-945VF: a fast-curing UV initiated epoxy resin that has good durability, good
temperature and chemical resistance, and low shrinkage coupled with high strain to
failure. AT-945VF is a commercial product of Adherent Technologies.
 v ATI-ROC-KF05F: a UV initiated epoxy resin formulated with sensitizers to promote
rapid cure at low temperatures. This material is a research resin and its mechanical
properties have not been fully characterized.

1.1.1 X-ray Absorption Enhancement

Increased absorption of x-rays in the thin adhesive layer will increase its sensitivity to this
method of cure initiation and allow full cure at substantially lower exposure dose than would be
possible using un-doped formulations. By incorporating heavy element components, the energy
absorbed in the adhesive layer can be increased by a substantial amount. However, if the
additional absorbed energy is to be productive in the initiating reaction it must be partitioned into
the organic phase. Generally, the absorption of x-rays of relatively low energy is accomplished
through ejection of an electron. Two processes for the conversion of the x-ray photons energy
can occur. The most efficient process is the photoelectric conversion of the photons energy into
kinetic and potential energy of the ejected electron. A somewhat less efficient process, Compton
scattering of the photon deposits only a fraction of its energy in the event. For species containing
only carbon, hydrogen and oxygen only (such as our adhesive resin), only about half of the
events involve the photoelectric process at an x-ray energy of 60keV; for higher x-ray energy, an
even smaller fraction is deposited by that process. Incorporating higher atomic number (Z)
elements can result in nearly complete photoelectric conversion of the x-ray photons while
substantially increasing the absorption coefficient. To be effective, the incorporation of the
higher Z elements into the formulation must allow efficient transfer of the absorbed energy into
the resin component. For particulate loading, much of the kinetic energy of the ejected electron
can be lost as lattice excitation (vibration) of the particles; the potential energy of ionization may
be recovered in part by electron transfer from the surroundings (the resin). Smaller particles give
better transfer. A much more efficient method of loading would involve dissolution of the
loading material into the resin on a molecular scale. We have examined both types of loading in
our work.

To improve x-ray energy transfer, metals or metal salts were added to the base resins discussed
above. Two metals and four metal salts were evaluated. These metals provide a substantial
increase in absorption coefficient, from 170 to 280 times that of carbon (for 60 keV and below x-
ray energy). Metal halogen salts were investigated because they have substantial covalent
character and thus exhibit better solub ility in organic systems. The metals were employed
individually; the metal salts were employed as 50/50 mixtures of the chloride and bromide salts.
Experimental results of x-ray irradiation of the modified resins as well as the unmodified base
resins are discussed below. The best candidate resins were evaluated further as adhesives using
lap shear test methods.

2. EXPERIMENTAL
2.1 Adhesive Formulations

Adhesive formulations were prepared by mixing the filled base resin (AT-924F, AT-945VF, or
ATI-ROC-KF05F) with the desired amount of the metal or metal salt dose enhancing materials.
The resins were heated slightly to reduce the viscosity and to more efficiently disperse the
additives. The formulations were then either placed into small sample vials for neat resin
experiments or were packed into syringes for application to substrates.

2.2 X-ray Irradiation

X-ray irradiation was conducted at Acsion Industries, Inc. (Pinawa, Manitoba, Canada).
Accelerator parameters for the I-10/1 linear electron beam accelerator were a peak beam current
of 225 mA with 10MeV electrons. To generate x-rays, a water-cooled tungsten target was used.
It was placed 20 cm from the exit window of the accelerator, the beam scan width at the target
was 45 cm. The sample was 10 cm from the target.

The samples were fixed for the duration of the experiments. The time to cure for lap shear
samples was 100 minutes.

2.3 Differential Scanning Calorimetry

Differential scanning calorimetry (DSC) experiments were performed on a nitrogen-purged

Perkin-Elmer Diamond DSC over the temperature range 25350C at a heating rate of 10C/min.
The cure exotherms of unexposed resins were analyzed to provide baseline data for the
calculation of percentage cure in the x-ray exposed samples. X-ray exposed samples were
analyzed in a manner identical to that used for the unexposed samples. A spreadsheet was used
to tabulate data and perform calculations of percentage cure and standard deviation in the data.

2.4 Lap Shear Testing

Adhesive shear strengths were measured using the lap shear technique following the protocols
given in ASTM D1002. G10 composite lap shear test specimens were prepared using 1 in. 4
in. 0.125 in. thick substrates with a bond area of 0.5 in2 . The substrates were prepared for
bonding by sanding the bond area with 250 grit sandpaper followed in some cases by a surface
treatment. If appropriate, treatments were followed by a short bake at 150C. Specimens
receiving no surface treatments were also prepared.

After treatment, the adhesive was spread evenly onto both the top and bottom substrates. The
bottom substrate was then placed in the jig, after which a thin strip of fiberglass screen was
placed in order to ensure the same bond thickness for all samples. The top substrate was then
placed in the jig, and the top plate placed onto the apparatus and the bolts tightened a quarter turn
to hold the substrates steady in the jig.

Control specimens bonded with FM-123 adhesive (3M Corporation) were prepared for
comparison. The FM-123 is a room temperature cure epoxy adhesive. Cure was conducted
according to manufacturer instructions.

Lap shear tests were conducted on an Instron Model 1101 mechanical testing machine fitted with
a 1000-lb capacity tension/compression load cell. The samples were mounted in self-aligning
grips with shims at either end to maintain the adhesive joint parallel to the direction of pull. Data
was collected on a computer-based data acquisition system with statistical analysis capability.
 3. RESULTS AND DISCUSSION
Several different experiments were conducted in order to identify the most promising adhesive
formulations and then to evaluate bonding of these materials to an appropriate substrate. Initial
testing was conducted on small samples of resin alone; several formulations were tested. The
best formulation was then evaluated in a bonding situation. Results of these experiments are
discussed in the following sections.

3.1 Neat Resin Screening

The adhesive formulations described above were subjected to x-ray irradiation at several
different doses. This allowed the identification of the most active resin series as well as
providing guidance as to what dose would ultimately be required to cure the adhesive when used
in fabrication processes.

Small samples of the neat resins were prepared and shipped to Acsion Industries (Pinawa,
Manitoba, Canada) for x-ray irradiation. The samples were irradiated to doses of 2kGy, 5kGy,
10kGy, and 20kGy. The degree of cure was then analyzed using differential scanning
calorimetry (DSC). The best resins (i.e., those exhibiting the highest level of cure at the lowest
possible dose) were then selected for fabrication of lap shear specimens.

Adhesives based on AT-924F were immediately eliminated from further consideration. None of
the samples prepared using this resin fully cured. Even at 20kGy with the addition of metal salts,
the samples were not completely cured.

Figure 1 shows representative thermograms for one of the AT-924F systems. The presence of
residual epoxy cure exotherms is evident even at the highest doses examined.

Figure 1. DSC
thermograms of AT-
924F with added metal
salts at various
irradiation levels. Note
that a significant
amount of uncured
material remains
(residual exotherm)
even after a 20 kGy
exposure.

In contrast, both AT-945VF and ATI-ROC-KF05F looked promising in the straight resin but
when we compared the rest of the samples it appeared that the ATI-ROC-KF05F was the better
base resin. All formulations using ATI-ROC-KF05F showed full cure at the lowest dose of
2kGy. The AT-945VF formulations cured only at the higher doses; only two of these
formulations cured at 10kGy. Thermograms of both the ATI-ROC-KF05F and AT-945VF
formulations with added metal salts are given in Figures 2 and 3, respectively.

Figure 2. DSC
thermograms of ATI-
ROC-KF05F with added
metal salts. Note the
lack of significant
residual exotherms at
all doses 2 kGy and
higher.

Figure 3. DSC thermograms of

AT-945VF with added metal
salts. Note that some residual
cure exotherm remains for all
samples irradiated at less than
20kGy.

Based on these results, ATI-ROC-KF05F was chosen as the base resin for use in preparation of
lap shear samples. All formulations except the straight resin were initially evaluated for lap
shear specimens. The straight resin formulation (without metal or metal salt additives) was not
evaluated further because a general trend of improvement in the degree of cure with addition of
metals or metal salts was noted for all resins initially tested.

3.2 Lap Shear Specimens

Lap shear specimens were prepared to ASTM D1002 specifications on G10 fiberglass composite
coupons. An initial set of twelve lap shear specimens was prepared using four of the best resins
identified from the neat resin irradiations described above. The substrates were sanded in the
bonding area to promote adhesion, but no primers were employed. A fiberglass screen was
placed between the substrates to ensure reproducible bond thicknesses for all samples. These
samples were then irradiated to a dose of 5kGy and were demolded. None of the lap shear
samples were fully cured despite the fact that the neat resin specimens had all exhibited full cure
at this dose. DSC analysis of resin taken from the substrates confirmed the partial cure. These
results are summarized in Table 1.

Table 1. Summary of Cure Data for Lap Shear Specimens Irradiated to 5kGy

Sample No./number of coupons Resin % Cure (Avg.)

MG1113-1/3 coupons TS1141-9 59.7 3.5
MG1113-2/3 coupons TS1141-10 62.2 1.9
MG1113-3/2 coupons* TS1141-11 53.5 6.9
MG1113-4/3 coupons TS1141-12 81.2 1.9
* A third coupon was made using the TS1141-11 adhesive that exhibited significantly higher cure (81.7%)
than the others in this series. The reason for this is unknown at this time.

The reason for the incomplete cures on the lap shear substrates is not known at this time and is
somewhat peculiar given that all of these resins showed complete cure at 2kGy when evaluated
as neat resins. It may be possible that the substrate is interfering with the cure of the cationic
epoxy resin. The G10 composite substrate consists of glass fiber in a standard (amine-cured)
epoxy resin. There is a possibility that the lone pairs of electrons on the nitrogen atoms in this
cured structure could be interacting with the reactive cationic species in the curing adhesive in
such a way as to remove these cations from the polymerization mechanism. Similar behavior has
been reported for electron beam- initiated epoxy adhesives (also cationic systems) used with
graphite/epoxy composites [1,2]. Inhibition of cure due to the interaction between nitrogen lone
pairs and the curing cationic resin was speculated at that time.

A second set of lap shear specimens was prepared using the best resin, TS1141-12, and a silane-
based substrate treatment was used in addition to surface sanding in an attempt to improve
adhesion to the substrate. Half the samples were irradiated to 20kGy total dose and the other half
were irradiated to 50kGy total dose. DSC analysis indicated complete cure for all samples; lap
shear testing was inconclusive. However, visual observation of the failure surfaces indicated that
only the 50kGy dose was sufficient to leave cured resin on both sides of the broken lap shear
specimen. When a 20kGy dose was employed, the cured resin remained on only one of the
substrate surfaces.

A final set of lap shear specimens was prepared, again using the best resin. This time a variety
of surface treatments were evaluated. The substrates were again sanded in the bonding area and
the fiberglass screen used to control bond thicknesses. The list of primers examined is given in
Table 2.
 Table 2. Second Series of Lap Shear Specimens on G10 Substrates.

Sample No./number of test specimens Surface Treatment

MG1120-1 / 3 Plasma (oxidizing)

MG1120-2 / 3 Trimethylchlorosilane
MG1120-3 / 3 None
MG1120-4 / 3 AT-9304

These samples were irradiated to a dose of 50kGy. Full cure of the adhesives was noted for all
specimens. Results from lap shear testing of these samples were quite positive. A shear strength
of 7.23 MPa (1048 psi) was observed for one sample series. This is comparable to results
obtained by Byrne and coworkers for electron-beam cured cationic film adhesives [3]. These
results are summarized in Table 3, below. A room temperature curing two-part epoxy system,
FM-123 (3M Corporation) is included for comparison.

Table 3. Results of Lap Shear Testing for X-ray Cured Adhesive TS1141-12 versus a Standard
Room Temperature Curing System

Adhesive Primer Shear Strength, MPa (psi) Coefficient of Variation, %

TS1141-12 None 6.17 (895) 11.0
TS1141-12 Plasma 6.65 (964) 3.7
TS1141-12 Silane 3.81 (552) 21.1
TS1141-12 AT-9304 7.23 (1048) 7.3
FM-123 None 14.18 (2056) 9.3

The data in Table 3 clearly indicate that the surface of the substrate has a significant effect on the
shear strength. Samples treated with either plasma or AT-9304 showed the best shear strengths
and the most consistent sample behavior, with the untreated substrate surfaces following closely
behind. The silane treated surfaces exhibited very poor adhesion relative to all other samples
tested.

The shear strengths of the best x-ray cured samples were about half of that measured for the
standard room temperature curing system. This should not, however, be viewed as a
shortcoming of the x-ray cured adhesives since the systems are not yet optimized. The use of
aromatic epoxy monomers instead of the current aliphatic epoxy monomers in the base resin will
result in improved shear strengths.

The failure curves of the x-ray cured systems are somewhat different from those of the standard
two-part system as shown in Figure 4. Most of the radiation cured systems showed failure
steps, seen in Figure 4(b), which are commonly referred to as stick-slip failures. This type
of failure is not common in adhesives, but is well known in composites. The breaks in the data
were accompanied by audible cracking in the sample, indicating that real failures did occur.

(a) (b)

Figure 4. Failure curve for (a) FM-123 room temperature cure epoxy adhesive used as a
control sample and (b) for x-ray cured adhesive on AT-9304 treated G10 composite substrate.

Closer examination of the failure surfaces of the samples confirmed cracking into the substrate in
some cases; this is shown in Figure 5. This is a very promising result, indicating excellent
bonding between the x-ray cured adhesive and the substrate.

Figure 5. Photomicrograph of
broken lap shear sample. The
adhesive layer is pulled away from
the substrate and the substrate
(bottom of photo) shows some
cracking. Additionally, there is
clear evidence of adhesive resin
remaining on the substrate after
having been pulled apart.

In some cases, the curves for the x-ray cured samples show more abrupt, brittle failures whereas
the curves for the two-part system show a more standard elastic-plastic failure. This is to be
expected since the two-part system includes a hardener that may impart some toughening to
the overall system. The x-ray cured systems have not yet been optimized for toughness and
therefore exhibit brittle failure. The incorporation of comonomers or other tougheners into the
one-part formulation will alleviate this problem and yield results more typical of that observed
for the two-part system.

Additionally, samples prepared with either AT-9304 or plasma surface preparations all showed
significant resin remaining on both substrate surfaces after the lap shear samples were broken.
This was observed for some of the samples with untreated substrates as well. Clear evidence of
resin remaining on the substrate can be seen in Figure 5, above.

4. CONCLUSIONS
This program successfully demonstrated adhesive systems that could be cured with x-ray
irradiation at room temperature. Lap shear testing showed that shear strengths on the order of
6.9 MPa (1000 psi) were easily attained when appropriate substrate surface treatments were
employed. This result is comparable to results obtained for a cationic electron-beam cured film
adhesive. When an appropriate substrate pretreatment is employed, there is significant resin
remaining on the substrate surfaces of the broken lap shear samples. Failure curves showed
stick-slip behavior, which corresponded to audible cracking during the testing procedure.
Photomicrographs of the broken samples supported the failure curves, showing cracking of the
substrate in some cases.

The shear strength of the adhesive developed here was somewhat lower than that of the standard
room temperature curing epoxy (two-part) used as a control. The two-part system exhibited
shear strengths of approximately 15 MPa (2200 psi). It is believed that this difference arises
from the chemical nature of the resins used in formulation. The x-ray cured adhesives were
formulated using aliphatic epoxy resins rather than the aromatic epoxy resins commonly used in
most two-part systems. Future work will incorporate these aromatic epoxy resins into the x-ray
curing formulations along with comonomers and tougheners. These changes should result in
significant improvements to the shear strength of these adhesive systems.

Further exploration of the x-ray enhancing additives is also warranted to decrease the dose
required to cure. For this program, an x-ray dose of 50kGy was required to fully cure the
adhesives in the lap shear specimens. By optimizing the x-ray enhancing fillers, we hope to
lower the dose required to cure.

5. REFERENCES

1. Lopata, V.J.; Chung, M.; McDougall, T.E.; and Weinberg, V.A. Electron-Curable
Adhesives for High Performance Structures, 38th Int. SAMPE Symp. Exhib. 1994, 39, 514.

2. Farmer, J.D.; Janke, C.J.; and Lopata, V.J. The Electron Beam Cure of Epoxy Paste
Adhesives, 43rd Int. SAMPE Symp. Exhib. 1998, 43, 1639; Farmer, J.D.; Janke, C.J.; and
Lopata, V.J. Electron-Beam Cure of Composite T-38 Windshield Frame/Arch, 43 rd Int.
SAMPE Symp. Exhib. 1998, 43, 1647.

3. Byrne, C.A.; Goodman, D.L.; Palmese, G.R.; Sands, J.M.; and McKnight, S.H. Electron
Beam Curable Adhesives for Out-of-Autoclave Bonding of Large Composite Structures,
45th Int. SAMPE Symp. Exhib. 2000, 45, 2187.
 6. ACKNOWLEDGEMENTS
The authors would like to thank Mr. Vince Lopata of Acsion Industries (Pinawa, Manitoba,
Canada) for his assistance with x-ray irradiation of neat resin candidates and lap shear
specimens. We are also indebted to Ms. Tetyana Shkindel and Mr. Mark Gallegos of Adherent
Technologies, Inc. for sample preparation, and to Dr. Sheldon Wesson and Mr. Tyler Wesley,
also of Adherent Technologies, Inc., for mechanical testing.

This work was funded by the United States Department of Energy under the Small Business
Innovation Research (SBIR) program, contract number DE-FG03-01ER83178.

7. BIOGRAPHIES
Dr. Andrea E. Hoyt received her Ph.D. in Polymer Science from the University of Connecticut
Institute of Materials Science in 1993. She has a B.A. in Chemistry from the University of
Colorado and an M.S. in Polymer Science from the University of Connecticut. Dr. Hoyt is the
Manager of Polymer Projects at Adherent Technologies. In this position, she has led projects in
the development of electron beam curing resin systems, coatings for chemical sensors, liquid
crystalline thermoset adhesives, and new mo isture resistant coupling agents for glass fiber-
reinforced composites. Her current research focuses on the development of radiation curing
resin systems, on open-cell rigid polyurethane foams for self-deployable space structures, and on
fiber optic sensors for groundwater monitoring. Before joining Adherent Technologies, she was
engaged as a postdoctoral research associate in the Microsensors Research and Development
Department at Sandia National Laboratories. She has also done research in liquid crystalline
polymers and liquid crystalline thermosets at Los Alamos National Laboratory and in polyimide
synthesis and characterization at the University of Connecticut. Dr. Hoyt has over 10
publications in technical journals and conference proceedings and holds 4 patents in the areas of
liquid crystalline thermoset materials and chemical sensors.

Dr. Larry A. Harrah earned his PhD in physical chemistry at the University of Missouri,
Columbia, in 1961, where he also received his BS in chemistry in 1953. Since 1990, Dr. Harrah
has held the position of Chief Scientist at Adherent Technologies. His current research focuses
on the development of ultrafast scintillators and colorimetric dosimetry systems. Dr. Harrah
served on active duty with the US Air Force from 1957 to 1960 where he was assigned to the US
Air Force Materials Laboratory (AFML) at Wright Patterson AFB, Ohio. After receiving his
doctorate, he returned to the AFML as a civilian employee, managing the effort on radiation
damage to organic materials used in aerospace applications. In 1966, Dr. Harrah joined Sandia
National Laboratories, where he concentrated in the area of photo and radiation physics of
polymers. From 1987 through 1990, Dr. Harrah held the position of Senior Scientist at PDA
Engineering in Albuquerque, NM. He is the author of 65 technical papers and holds 7 patents in
the areas of radiation dosimetry, organic scintillators, photoresists, electric foils, and organic
getters.
Ms. Melissa R. Sprouse earned her MFS in Forensic Toxicology from George Washington
University. She has a BS in Chemistry from Gettysburg College. Before joining Adherent
Technologies, Ms. Sprouse was a Research Associate in the Analytical Department at Lovelace
Respiratory Research Institute (LRRI) an inhalation toxicology laboratory. At LRRI, she worked
on a major interdisciplinary project involving the nerve gas Sarin. Ms. Sprouse was in charge of
the analytical instrumentation on the project and was one of three HAZWOPER trained safety
officers. She was also involved in the analytical method development on the MALDI, LC/MS,
GC/MS, and the GC. Before coming to Albuquerque, New Mexico, Ms. Sprouse was employed
with Genvec, Inc., Nova Research, Inc., Fuisz Technologies, and LabCorp. In those positions,
she gained experience with unique instrumentation, forensic and research toxicology, and
analytical chemistry.