polymers

Article
High Temperature Epoxy Foam: Optimization of
Process Parameters
Samira El Gazzani 1,2 , Valrie Nassiet 1, *, Jean-Pierre Habas 3 , Christian Freydier 2
and Aline Hilleshein 1
1 Universit de Toulouse, INPT, Laboratoire Gnie de Production, Ecole Nationale d'Ingnieurs de Tarbes,
47 Avenue dAzereix, B.P. 1629, 65016 Tarbes cedex, France; samira.elgazzani@enit.fr (S.E.G.);
alinehilleshein@hotmail.com (A.H.)
2 Roxel France, N151, 18570 Le Subdray, France; c.freydier@roxelgroup.com
3 Institut Charles Gerhardt, Equipe Ingnierie et Architectures Macromolculaires, UMR CNRS 5253, CC 1702,
Universit de Montpellier, Place Eugne Bataillon, 34095 Montpellier, France;
jean-pierre.habas@univ-montp2.fr
* Correspondence: valerie.nassiet@enit.fr; Tel.: +33-562-44-2930

Academic Editor: Antonio Pizzi

Received: 5 May 2016; Accepted: 30 May 2016; Published: 7 June 2016

Abstract: For many years, reduction of fuel consumption has been a major aim in terms of both
costs and environmental concerns. One option is to reduce the weight of fuel consumers. For this
purpose, the use of a lightweight material based on rigid foams is a relevant choice. This paper deals
with a new high temperature epoxy expanded material as substitution of phenolic resin, classified as
potentially mutagenic by European directive Reach. The optimization of thermoset foam depends on
two major parameters, the reticulation process and the expansion of the foaming agent. Controlling
these two phenomena can lead to a fully expanded and cured material. The rheological behavior
of epoxy resin is studied and gel time is determined at various temperatures. The expansion of
foaming agent is investigated by thermomechanical analysis. Results are correlated and compared
with samples foamed in the same temperature conditions. The ideal foaming/gelation temperature
is then determined. The second part of this research concerns the optimization of curing cycle of a
high temperature trifunctional epoxy resin. A two-step curing cycle was defined by considering the
influence of different curing schedules on the glass transition temperature of the material. The final
foamed material has a glass transition temperature of 270 C.

Keywords: expandable microsphere; rigid foam; TETM trifunctional epoxy; thermomechanical analysis

1. Introduction
Polymerfiber composite materials are widely used in many applications as they combine a low
weight with satisfying mechanical properties. A decrease in weight is a recurrent requirement in
industry, particularly in the aerospace field where fuel cost has a significant impact. Producing foamed
materials is a way to achieve mass reduction.
There are different methods to produce foam: decomposition of chemical substance with gas
formation, under heating [14]; bubble creation by vaporization of a liquid [510]; gas introduction in
the resin under high-pressure follow-up with expansion when pressure drops [1113]; and mechanical
beating [14,15]. The low boiling temperature liquids used for the expansion by vaporization are usually
hydrofluorocarbon, which is problematic because they contribute to global warming. Other methods
allow lightening a structure, for example by filling a matrix with microballoons to produce syntactic
foam [1618], or with water-soluble crystals in a net immersed in water so that holes supplant the
crystals [19,20]. However, these processes do not involve expansion, so they are not relevant for
our application.

Polymers 2016, 8, 215; doi:10.3390/polym8060215 www.mdpi.com/journal/polymers

Polymers 2016, 8, 215 2 of 19

In the case of thermoset foams, the foaming and the crosslinking reaction have to take place
simultaneously. Sometimes, the blowing agent is involved in the crosslinking reaction. For example,
the reaction between isocyanate and water, which produce CO2 , is a well-known method used in
the manufacture of polyurethane foams. Li et al. [21] and Basso et al. [22,23] explored the addition
of isocyanate in the manufacture of tannin foams. Dogan and Kusefoglu [24] studied the reaction
between malonic acid and epoxidized soybean oil. Malonic acid reacts with epoxy groups on the
one hand and produces CO2 by decarboxylation on the other hand, which acts as a blowing agent
leading to in situ foaming of the polymer. Stefani et al. [25] used a polysiloxane as blowing agent to
produce epoxy foam. They postulate that the amino hardener reacts with the polysiloxane, leading
to dihydrogen production. Indeed, the amino hardener has a dual function: crosslinking agent and
gas co-producer. These two reactions follow distinct kinetics, which make the crosslinking/foaming
process difficult to control. Then, the use of a foaming agent independent of a crosslinking reaction
would be preferable. For example, sodium bicarbonate is a harmless substance that can be used as a
blowing agent since it decomposes during heating, releasing CO2 and H2 O. It was used in various
studies with epoxy resins [1,26]. Najib et al. used it as a foaming agent in natural rubber [27].
Expandable microsphere is a non-traditional physical blowing agent initially patented by
Morehouse in 1971 [28]. The expandable microspheres are prepared by a suspension-type
polymerization of droplets of a mixture of monomers and blowing agents. This leads to a microsphere
of thermoplastic copolymer encapsulating a volatile hydrocarbon liquid. Expansion occurs because
the thermoplastic shell softens, while the encapsulated hydrocarbon gasifies, increasing the internal
pressure in the particle. The copolymer is made of three different monomers assumed to have a specific
role: the one with the lowest glass transition temperature (Tg ) offers initial softening as the liquid
starts to vaporize; the one with middle Tg allows the expansion of the microspheres; and the one
with highest Tg lends a relative stiffness. The particles expand as the hydrocarbon vapor pressure
overcomes the yield strength of the polymer shell [29]. In order to contain the gas, a monomer with
gas-barrier properties such as acrylonitrile (AN) or methacrylonitrile (MAN) is generally used as
a polymerizable monomer [30]. There is a large range of products, made of different copolymers
and hydrocarbons, leading to various temperatures of expansion. They are commercially available
under the trade name of Expancel [31]. This has been used for different purposes like car protection
production [32] and for the manufacture of an expanded composite material named Roxalte.
Roxalteis developed by Roxel, a company dedicated to propulsion systems, and is of increasing
interest to the aerospace industry. This material is made from three major components: phenolic resin,
a basalt fiber mat, and an expandable microsphere [33].
The phenolic resin is a resole type, composed of a hydroxymethyl phenol, phenol, and
formaldehyde solution. The high concentration of aromatic cycles after curing provides excellent
thermal properties at high temperatures. Some tests were performed on the phenolic resin and a glass
transition temperature of 380 C was found [34].
Anyway, phenol and formaldehyde are both hazardous molecules, classified as potentially
mutagenic and carcinogenic, respectfully, by European directive Reach. The substitution of this
resin is then required. Epoxy resin has been selected as a relevant choice due to its good mechanical
properties, resistance to corrosive liquids and environments, resistance to moisture absorption, and
durability. The main criterion of selection of epoxy bi-components was to obtain a high glass transition
temperature (above 250 C). The link between the chemical structure of the material and the Tg was
studied by various authors [35,36]. The Tg has been found to increase with the degree of crosslinks
and chain rigidity of the polymer network [37]. The degree of crosslinking is inversely proportional to
the average molecular weight between crosslinks (Mc). Then molecules with a high reactive group
ratio should enhance the Tg . On the other hand, the chain rigidity is greater for an aromatic cycle
than for aliphatic ones [38]. The link between the presence of an aromatic cycle and the high Tg of the
structure is illustrated in Van Krevelens study [39]. In order to reach a high Tg , the triglycidyl ether
of tris(4-hydroxyphenyl)methane (TETM), a trifunctional epoxy with high aromatic ratio, has been
Polymers 2016, 8, 215 3 of 19

Polymers 2016, 8, 215 3 of 19

triglycidyl ether of tris(4-hydroxyphenyl)methane (TETM), a trifunctional epoxy with high aromatic
ratio, has been chosen. Indeed, the high energy chemical bonds of the aromatic cycle are responsible
for theIndeed,
chosen. properties
the highstability at high
energy chemical bonds temperature.
of the aromatic Because
cycle areitresponsible
is an aromatic amine,
for the properties
diaminodiphenylsulfone
stability at high temperature. (DDS) has it
Because been
is anselected
aromaticasamine,
curingdiaminodiphenylsulfone
agent. Moreover, TETM is known
(DDS) has been to
have superior
selected as curing thermal
agent. oxidative
Moreover,stability
TETM isover known other types
to have of epoxy
superior resins
thermal [40]. Aronhime
oxidative stability overand
Guillham
other typesestimated
of epoxy resinstheoretically the glassand
[40]. Aronhime transition
Guillham temperature of TETM/DDS
estimated theoretically formulation
the glass transition at
about 352 Cof
temperature TETM/DDS formulation at about 352 C [41].
[41].
For the
For thereasons
reasonspreviously
previouslyexposed,
exposed,the the substitution
substitution of of phenolic
phenolic resin
resin in Roxalte
in the the Roxalte
was was the
the first
first purpose
purpose of this of study.
this study. Subsequently,
Subsequently, different
different foaming
foaming processes
processes werewere considered.
considered. Starting
Starting withwith
the
the blowing
blowing agentagentused used in the
in the Roxalte,
Roxalte, i.e.i.e. expandable
expandable microspheres,seemed
microspheres, seemedlogical.
logical.Indeed,
Indeed,as as far
far as
we know,
we know, thethe association
association of of the
the preselected
preselected epoxy epoxy formulation
formulation withwith these microspheres
microspheres has has not been
studied. Further developments on other foaming processes with different foaming agents will be the
another paper.
aim of another paper.
Wang et
Wang et al.
al. [42]
[42] produced
producedepoxy/expandable
epoxy/expandable microsphere microsphere foams in two steps. The epoxy was was
bisphenol-A. A
diglycidyl-ether of bisphenol-A. A precure
precurestep stepwas wasfirst
firstdone
doneon onthe
theepoxy/curing
epoxy/curingagent/blowing
agent/blowing
agent formulation
agent formulationinin order
order to increase
to increase the viscosity
the viscosity by starting
by starting the polymerization
the polymerization process. process.
The foaming The
foamingoccurred
process process duringoccurred theduring
secondthe stepsecond
wherestep where the formulation
the formulation was put under wasa higher
put under a higher
temperature.
temperature.
This This process
process intended intendedthe
to improve to cell
improve the cell sizein
size distribution distribution
the foam. in In the
ourfoam.
study,Inwe our study,
propose
we propose manufacturing
manufacturing the foam in the onefoam
step in oneby
only, step only, the
tuning by tuning the microspheres
microspheres foaming
foaming kinetic to kinetic
the resin to
the resinkinetic.
gelation gelation Thekinetic. The crosslinking
crosslinking reaction andreaction and the
the expansion expansion oftake
of microspheres microspheres take place
place simultaneously.
simultaneously.
Obtaining an expandedObtaining an expanded
material dependsmaterial dependsphenomena,
on two distinct on two distinct phenomena,
the foaming agentthe foaming
expansion
agent
on oneexpansion
hand and the on one
resinhand andon
gelation thetheresin
othergelation on thetwo
hand. These other hand. These
phenomena two phenomena
are independent are
of each
independent of each other but both
other but both depend on time and temperature. depend on time and temperature.
propose aa complete
First, we propose complete physicochemical
physicochemical study of expandable expandable microspheres, in order to to
understand their their behavior.
behavior.InInparticular,
particular,the expansion
the expansion is studied
is studiedby thermomechanical
by thermomechanical analysis and
analysis
correlated
and to the
correlated gel gel
to the time of of
time thethe
epoxy
epoxy formulation
formulation with
withthetheaim
aimofofidentifying
identifying thethe ideal process
temperatureand
temperature andproducing
producing the optimized
the optimized foamed foamed
material.material. The optimization
The optimization of the epoxy of formulation
the epoxy
formulation
curing curing
cycle has beencycle has been
realized realized
to achieve thetomanufacture
achieve the manufacture of enhanced
of enhanced expanded expanded foam.
foam.

2. Materials
Materialsand
andMethods
Methods

2.1.
2.1. Materials
Materials
The
The triglycidyl ether ofether
triglycidyl tris(4-hydroxyphenyl)methane
of (TETM) and the diaminodiphenylsulfone
tris(4-hydroxyphenyl)methane (TETM) and the
(DDS) were kindly offered by Huntsman Advanced
diaminodiphenylsulfone (DDS) were kindly offered by Huntsman Materials (Basel, Switzerland).
Advanced Their commercial
Materials (Basel,
names are Tactix
Switzerland). 742commercial
Their and Aradurnames976, respectively.
are Tactix 742 Theand
foaming
Aradur agents from the Expancelrange
976, respectively. The foaming
were supplied by Rossow Industries (Gennevilliers, France). The Expancel950
agents from the Expancel range were supplied by Rossow Industries (Gennevilliers, DU 80 was selected
France). The
(Particule size 1824 m; Density 12 kg/m 3 ). The microsphere copolymer is made of acrylonitrile,
Expancel 950 DU 80 was selected (Particule size 1824 m; Density 12 kg/m3). The microsphere
methylmetacrylate,
copolymer is made and methacrylonitrile
of acrylonitrile, and the volatileand
methylmetacrylate, liquid inside is an isooctaneisopentane
methacrylonitrile mix.
and the volatile liquid
Its
inside is an isooctaneisopentane mix. Its technical data sheet also reveals the presence of
technical data sheet also reveals the presence of magnesium hydroxide. The chemical structures of
TETM and DDS
magnesium are presented
hydroxide. in Figure
The chemical 1.
structures of TETM and DDS are presented in Figure 1.

O O
O O
H 2C C C C C CH 2
H H2 H2 H
H H 2N NH 2
C

S
O O
O
O
C C CH 2
H2 H

(a) (b)
Figure 1. Chemical structures of (a) TETM and (b) DDS.
Figure 1. Chemical structures of (a) TETM and (b) DDS.
 Polymers 2016, 8, 215 4 of 19
Polymers 2016, 8, 215 4 of 19

2.2.Preparation
2.2. PreparationofofEpoxy
EpoxyResin
Resin
The epoxy
The epoxy resin
resin formulation
formulationwas wasprepared
prepared in in
stoichiometric
stoichiometric proportion: the stoichiometric
proportion: ratio,
the stoichiometric
denoted as r, is defined by the ratio between the number of amine and epoxy group
ratio, denoted as r, is defined by the ratio between the number of amine and epoxy group and has to and has to be equal
to 1.
be Then,tothe
equal 1. mass
Then,proportions
the mass were 71% of TETM
proportions were for71%29% of of DDS. for
TETM The 29%
melting temperature
of DDS. of DDS
The melting
is 178 C. As evidence, the TETM was first heated at 180 C and the DDS was added and mixed under
temperature of DDS is 178 C. As evidence, the TETM was first heated at 180 C and the DDS was
mechanical
added stirringunder
and mixed until apparent
mechanical homogeneity. Indeed,
stirring until we canhomogeneity.
apparent see in Figure 2Indeed,
the difference
we canbetween
see in
the TETM/DDS formulations mixed at 100 and 180 C. At 100 C, the viscosity is small enough to
Figure 2 the difference between the TETM/DDS formulations mixed at 100 and 180 C. At 100 C, the
allow mixing but the DDS is not melted, leading to a low homogeneity, compared
viscosity is small enough to allow mixing but the DDS is not melted, leading to a low homogeneity, to the mix realized
at 180 C. to the mix realized at 180 C.
compared
Themix
The mixisiscooled
cooledatatambient
ambienttemperature
temperaturebefore
beforestarting
startinganyanyanalyses.
analyses.

(a) (b)
Figure
Figure 2.
2. Picture
Pictureof
of TETM/DDS
TETM/DDSmixed
mixedatat(a)
(a)100
100C
Cand
and(b)
(b)180
180C.
C.

2.3.
2.3.Preparation
PreparationofofEpoxy
EpoxyFoam
Foam
To
To prepare
prepare the
the epoxy
epoxy foam,
foam, TETM
TETM was
was first
first mixed
mixed into
into DDS
DDS atat 180
180 C,
C, as
as above.
above. Then
Thenthe
the
temperature was lowered to 100 C
and the microspheres were added and mixed under
temperature was lowered to 100 C and the microspheres were added and mixed under mechanical mechanical
stirring
stirringagain
againuntil
untilapparent
apparentuniformity.
uniformity. First
Firstexperiments
experimentswere
wereuseful
usefultotoevaluate
evaluatethe
thenecessary
necessary
quantity
quantity of blowing agent. Finally, the mixture was poured into an aluminum pot and curedin
of blowing agent. Finally, the mixture was poured into an aluminum pot and cured inan
an
oven at constant temperature.
oven at constant temperature.

2.4.
2.4.Scientific
ScientificCharacterization
Characterization
All
Alldifferential
differentialscanning
scanningcalorimetry
calorimetry(DSC
(DSCQQ200200fromfromTA TAInstruments,
Instruments,New NewCastle,
Castle,DE,DE,USA)
USA)
analyses
analyseswerewereperformed
performedunder
underinert
inertatmosphere
atmosphere(N (N22) )using
usingan analuminum
aluminumcrucible
cruciblewith
withaacalibrated
calibrated
hole
holeand
andan anempty
emptycrucible
crucibleas asreference.
reference.
First,
First,the
thetemperature
temperaturedomaindomain of of
curing reaction
curing reactionof epoxy system
of epoxy was was
system determined
determinedby DSC. The
by DSC.
reactive formulation
The reactive was was
formulation prepared as mentioned
prepared before
as mentioned andand
before a small sample
a small sample of of
about
about5 5mg mgwaswas
removed for testing.
removed for testing.
To
To understand
understandthe theexpandable
expandablemicrospheres
microspheres behavior,
behavior, DSCDSCexperiments
experiments were performed
were performed in
dynamic mode, on usual microspheres and on compressed microspheres.
in dynamic mode, on usual microspheres and on compressed microspheres. The compressed The compressed
microspheres
microsphereswere were obtained
obtained using
using aa manual
manual hydraulic
hydraulic presspress (Specac,
(Specac, 1515 tons).
tons). AAfew fewgrams
grams of of
microspheres
microspheres were were introduced intointo the
thepress
pressand
andaa9-ton
9-tonloadloadwas
was applied.
applied. This
This allows
allows forfor throwing
throwing the
the liquid
liquid hydrocarbons
hydrocarbons out.DSC
out. The Theanalyses
DSC analyses were performed
were performed between 20between
and 40020 C,
andat a400 C, at a
temperature
temperature
ramp of 3 C/min.ramp ofTo3 be C/min. To be
free from freehydrocarbon
liquid from liquid hydrocarbon
evaporation and evaporation
to focus on and to focus on
determination
determination of the glass
of the glass transition transition temperature,
temperature, DSC analysesDSC wereanalyses
carried were
out oncarried
another outsample
on another samplea
as follows:
as
sample of compressed microspheres has been heated from 50 to 200 C, at a rate of 10 rate
follows: a sample of compressed microspheres has been heated from 50 to 200 C, at a of
C/min,
10 C/min,
then cooledthen cooled
before beforetemperature
a second a second temperature
scan at 10scan
at 10to
C/min C/min to Tg determination.
Tg determination. The firstThe first
heating
heating allows for evacuating the residual volatile liquid and the second
allows for evacuating the residual volatile liquid and the second one allows the Tg measurement. one allows the Tg
measurement. Thermogravimetric analyses (TGA-50 Shimadzu, Kyoto, Japan) were achieved in
order to get the degradation temperature. Experiments were performed between 25 and 500 C, at
temperature ramps of 1, 5, 10, and 20 C/min, under air atmosphere.
Polymers 2016, 8, 215 5 of 19

Thermogravimetric analyses (TGA-50 Shimadzu, Kyoto, Japan) were achieved in order to get the
degradation temperature. Experiments were performed between 25 and 500 C, at temperature ramps
of 1, 5, 10, and 20 C/min, under air atmosphere.
The study of expansion of expandable microspheres was performed by thermomechanical analysis
(TMA 7 from Perkin Elmer, Waltham, MA, USA). This instrument is used to measure the dimensional
variation of a sample under constant strain, as a function of time or temperature. A low strain is
applied by a probe on the sample to maintain the contact between the probe and the sample, and
the position of the probe is registered. The tests were done in isothermal mode, chosen to be in the
appropriate temperature range of the curing reaction. A temperature ramp of 30 C/min was applied
until the isothermal temperature was reached. The strain applied was 10 mN.
The determination of the gel time of the epoxy system was realized by viscosimetric experiment
using a dynamic rheometer (MCR 302 from Anton Paar, Graz, Austria) equipped with a cup-plate
geometry, in which the diameter of the upper plate is much smaller than the diameter of the cup (25
and 41 mm), in order to prevent side effects and slippage. The gel time is measured at the divergence
of viscosity and strongly depends on the temperature used in the kinetic. So the experiment was
performed in isothermal mode, chosen to be in the appropriate temperature range of the curing
reaction, in steady state with a shear rate of 20 s1 . For these experiments, the reactive formulations
were prepared at a temperature of 180 C. The mixture was then poured into the dedicated cup in the
instrument and the test started when the temperature was reached.
Scanning electronic microscopy (SEM) was used to study the morphology of the foam. SEM was
done using model ZEISS EVO HD15 (Jena, Germany). The variable pressure (VP) imaging was carried
out at 30 Pa. There was no need to coat non-conductive materials prior to analysis. The incident
electron beam was kept at 10 kV.
The study of optimization of the curing cycle of the epoxy system was performed by DSC and
thermomechanical analysis by using a dynamic rheometer equipped with rectangular torsion geometry.
First, epoxy formulations were subjected to different curing times (30, 40, 50, 60, 120, and 180 min)
at the same temperature. Then, around 6 mg of each material were used for DSC analysis. Tests
were performed at a heating rate of 5 C/min, from 25 to 400 C. Samples for thermomechanical
analysis were prepared as follows. The epoxy formulation was mixed, poured into an aluminum
mold, and cured in an oven under a determined cycle. The samples stayed in the oven for slow
cooling. The dimensions of the final samples were 2 mm 8 mm 45 mm. Tests were performed
with a temperature ramp of 3 C/min, from 25 to 350 C, with a deformation of 0.1% and an angular
frequency of 1 rad/s.
Thermomechanical analyses using a dynamic rheometer equipped with rectangular torsion
geometry were also performed on epoxy foam. The foamed material was cut to form a parallelepipedal
sample (3 mm 12 mm 45 mm). Tests were performed with a temperature ramp of 3 C/min,
with an angular frequency of 1 rad/s and a deformation of 0.05%.

3. Results and Discussion

3.1. Preliminary Study of the Epoxy Formulation and Expandable Microspheres Characterization
First, the epoxy formulation was studied by DSC in order to determine the temperature range
of the curing reaction. The DSC thermogram exposed in Figure 3 shows a first exothermic peak
corresponding to the curing reaction and a second one, noisy and repeatable, probably due to the
degradation. It is presumed that the exothermic peak of degradation is due to the condensation of
polyaromatic species by radical carbonization, as was postulated by Rose et al. [43]. According to
the DSC curve, this process starts around 330 C. It can be observed that the first exothermic peak is
included between 130 and 240 C, for a heating rate of 3 C/min.
Polymers 2016, 8, 215 6 of 19
Polymers 2016, 8, 215 6 of 19

0.25
0.15
Heat flow (W/g)

0.05
-0.05
-0.15
-0.25
-0.35
25 50 75 100 125 150 175 200 225 250 275 300 325 350 375 400 425 450
Temperature (C)

Figure 3. DSC scan of TETM/DDS formulation performed under nitrogen at a heating rate of 3 C/min.
Figure 3. DSC scan of TETM/DDS formulation performed under nitrogen at a heating rate of
3 C/min.
Since the reaction starts below 180 C, the question was asked about the mixing temperature of
the formulation. In orderstarts
Since the reaction to know whether
below 180 C,the gelation
the questionstarted duringabout
was asked the mixing, the gel
the mixing times wereof
temperature
determined at different temperatures for twothemixes: onestarted
mixed during
at 100 theC, and the second C.
at 180were
the formulation. In order to know whether gelation mixing, the gel times
Results are gathered
determined in Table
at different 1. The gel for
temperatures times
twoprove theone
mixes: difference
mixed atbetween
100 C, the
andtwo
the mixes.
secondAnyway,
at 180 C.
if Results
the mixare
is realized at 100 C, we may have some concerns about DDS precipitation as it is not well
gathered in Table 1. The gel times prove the difference between the two mixes. Anyway,
mixed into the epoxy prepolymer (Figure 2). Thus, we have chosen to perform the mixing at 180 C,
if the mix is realized at 100 C, we may have some concerns about DDS precipitation as it is not well
aware
mixed that the
into gelation
the is only slight.(Figure
epoxy prepolymer However the gelation
2). Thus, we have did not continue
chosen to performwhenthethe microspheres
mixing at 180 C,
were added and mixed at 100 C, as the curing reaction only starts at 130 C (see Figure 3).
aware that the gelation is only slight. However the gelation did not continue when the microspheres
were added and mixed at 100 C, as the curing reaction only starts at 130 C (see Figure 3).
Table 1. Comparison of gel times for two mixing temperatures.

Table 1. Comparison of gel times for two mixing temperatures.

Gel time (min)
Mixing temperature
130 C 140 C 150 Gel
C time
160(min)
C 170 C 180 C 190 C
Mixing temperature
100 C 130C
65.6 4 140C
46.3 3.5 150
33.8 C2.2 160
18.3C
1 170 C 1 180
10.3 6.1C
0.9 190
3.3C
0.8
100C
180 C 65.644
58.6 46.33.3
41 3.5 33.8 2.2
24.3 2 18.3 1 10.3 1 6.1 0.9 3.3 3.0
13.6 1 8.6 1 5.0 0.8 0.8
0.8
180 C 58.6 4 41 3.3 24.3 2 13.6 1 8.6 1 5.0 0.8 3.0 0.8
Thermochemical characterizations (DSC and TGA) were performed on compressed and
Thermochemical
uncompressed microspheres characterizations
to understand(DSC and TGA) were performed on compressed and
their behavior.
uncompressed microspheres to understand
The DSC analyses performed on compressed and their behavior.
uncompressed microspheres are presented in
The DSC analyses performed on compressed
Figure 4a. There are three major reactions. The first one, and uncompressed
measured at 160 microspheres
C, appears for are presented in
uncompressed
Figure 4a. There
microspheres are is
only. This three majordue
certainly reactions. The first one,
to the hydrocarbon measured
liquid at 160According
vaporization. C, appears for
to the
uncompressed microspheres only. This is certainly due to the hydrocarbon
technical data sheet, there are two hydrocarbons, isopentane and isooctane. Their vaporization liquid vaporization.
According to
temperatures arethe28technical
and 99 C,data sheet, there
respectively, are two hydrocarbons,
at atmospheric isopentane
pressure. However, whenand theisooctane. Their
microspheres
vaporization
are unexpanded, temperatures
the internal are 28 and
pressure in 99 C,
this respectively,
confined at higher
space is atmospheric
than inpressure.
atmospheric However, when
conditions,
leading to a higher temperature of vaporization [44]. A pressure drop is possible when the microspheresin
the microspheres are unexpanded, the internal pressure in this confined space is higher than
atmospheric
expands, leading conditions, leading toofathe
to the evaporation higher temperature
condensed blowing ofagent.
vaporization [44]. A pressure
The vaporization occurreddrop
hereis
possible when the microspheres expands, leading to the evaporation
at around 160 C, which is therefore the temperature of starting expansion. This vaporization should of the condensed blowing
agent. Thetovaporization
correspond occurredthe
isooctane. Therefore, here at aroundof160
vaporization C, which
isopentane is therefore
is not noticeable.the temperature
It was supposedof
starting
that expansion.ofThis
the proportion this vaporization
hydrocarbon should
was toocorrespond to isooctane.
low to be clearly detected Therefore, the vaporization
by DSC. Figure 4b exposesof
the analyses performed on compressed microspheres to reveal the glass transition temperature.too
isopentane is not noticeable. It was supposed that the proportion of this hydrocarbon was low to
A major
be clearly detected by DSC. Figure 4b exposes the analyses performed on
Tg appears around 107 C and a secondary one around 146 C. There are various Tg values reported in compressed microspheres
to literature
the reveal theforglass transition temperature.
the components of the copolymerA major Tg appears
capsule (PMMA, around 107 C and
PAN, PMAN) a secondary
[45,46], because theone
around 146 C. There are various T g values reported in the literature for the components of the
Tg is depending on molecular weight. We do not know the molecular weight of the copolymers so it is
copolymer
difficult capsulewhich
to conclude (PMMA, PAN, PMAN)
copolymer [45,46],
is related because
to these the Tg is depending
Tg . However, we assumed onthat
molecular
the firstweight.
glass
We do not know the molecular weight of the copolymers so it is difficult
transition (Tg 1) offers initial softening to reduce local pressure in the inner part of cell wall; the second to conclude which
copolymer is related to these T g. However, we assumed that the first glass transition (Tg1) offers
Tg (Tg 2) leads to the softening of the outer part of the cell wall, allowing the hydrocarbon vaporization
initial
due softening
to drop to reduce
in pressure. Onelocal
of thepressure in the inneris part
three copolymers of cell wall;
not associated the any
with second Tg (T
Tg ; we g2) leads to the
assume that its
softening of the outer part of the cell wall, allowing the hydrocarbon vaporization due to drop in
pressure. One of the three copolymers is not associated with any Tg; we assume that its role is to
 ensure
ensure relative
relative stiffness.
stiffness. TheThe proximity
proximity of of these
these twotwo Tg Tvalues
g valuesmademadeus us think
think of of
a Tag Tgradient,
g gradient,
in in
which
which thethe third
third copolymer
copolymer Tg T g should
should be be contained.
contained.
TheThe second
second and and third
third major
major reactions
reactions appear
appear forfor both
both compressed
compressed and and uncompressed
uncompressed
microspheres.
microspheres. It is
It is thusthus assumed
assumed thatthat they
they areare related
related to to
thethe capsule
capsule of of
thethe microspheres.
microspheres. The The
exothermic
exothermic
Polymers peak,
peak,
2016, 8, 215 between
between 220220
and and
290290C,C, is due
is due to the
to the cyclization
cyclization reaction
reaction in polyacrylonitrile
in polyacrylonitrile 7 and and
of 19
polymetacrylonitrile
polymetacrylonitrile [47,48],
[47,48], exposed
exposed in Figure
in Figure 5. The
5. The third
third reaction
reaction is an
is an endothermic
endothermic peak
peak between
between
320320
andand380380
C.C. It attributed
It is is attributed to to
thethe decomposition
decomposition of of magnesium
magnesium hydroxide
hydroxide to to
formform magnesium
magnesium
role is to ensure
oxide and relative
water, stiffness. Theinproximity ofwhere
these two Tthermogram
g values made us think of a Thydroxide
g gradient,
oxide and water, as as illustrated
illustrated Figure
in Figure 6, 6,
where thethethermogram of of magnesium
magnesium hydroxide is is
in which the third
represented. We copolymer
can see Tgthe
that should be contained.
endothermic peak is between 260 and 360 C.
represented. We can see that the endothermic peak is between 260 and 360 C.

1 1 2 2
0.80.8
0.60.6
Heat flow (W/g)
Heat flow (W/g)

0.40.4 3 3
0.20.2
0 0 1 1
-0.2-0.2
-0.4-0.4
-0.6-0.6
-20-20 80 80 180180 280280 380380
Temperature
Temperature (C)
(C)
uncompressed
uncompressed microspheres
microspheres compressed
compressed microspheres
microspheres

(a)(a)

0 0
-0.1-0.1
Heat flow (W/g)
Heat flow (W/g)

-0.2-0.2
-0.3-0.3 Tg1=
Tg1= 107107
C C
-0.4-0.4 Tg2=
Tg2= 146146
C C
-0.5-0.5
-0.6-0.6
-0.7-0.7
-50-50 0 0 50 50 100100 150150 200200
Temperature
Temperature (C)
(C)

(b)(b)
Figure
Figure
Figure 4. 4. DSC
4. DSC
DSC analyses
analyses
analyses of of compressed
of compressed
compressed andand
and uncompressed
uncompressed
uncompressed microspheres
microspheres
microspheres at 33at/min,
at 3 /min,
/min, with
with
with thethe
the three
three
three
major
major reactions
reactions encircled
encircled (a) (a)
andand
DSC DSC analysis
analysis performed
performed on on compressed
compressed microspheres
microspheres
major reactions encircled (a) and DSC analysis performed on compressed microspheres (second heating) (second
(second
forheating)
heating) forfor
thethe determination
determination
the determination of glass
of glass
of glass transitions transitions
transitions
at at (b).
at 10
10 C/min 10 C/min
C/min (b).(b).

Figure
Figure 5. Cyclization
Figure
5. reaction
5. Cyclization in polyacrylonitrile
reaction [47].
in polyacrylonitrile [47].

The second and third major reactions appear for both compressed and uncompressed
microspheres. It is thus assumed that they are related to the capsule of the microspheres.
The exothermic peak, between 220 and 290 C, is due to the cyclization reaction in polyacrylonitrile and
polymetacrylonitrile [47,48], exposed in Figure 5. The third reaction is an endothermic peak between
320 and 380 C. It is attributed to the decomposition of magnesium hydroxide to form magnesium oxide
Polymers 2016, 8, 215 8 of 19

and water, as illustrated in Figure 6, where the thermogram of magnesium hydroxide is represented.
Polymers
can 2016,
Polymers
We 8,8,215
2016,
see that the endothermic peak is between 260 and 360 C.
215 8 8ofof1919

00
-0.2
-0.2
-0.4

Haet flow (W/g)

-0.4

Haet flow (W/g)

-0.6
-0.6
-0.8
-0.8
-1-1
-1.2
-1.2
-1.4
-1.4
-1.6
-1.6
00 100
100 200
200 300
300 400
400
Temperature
Temperature(C)
(C)

Figure
Figure 6.6.
Figure
6. DSC
DSC analysis
DSC
analysis ofof
analysis
of magnesium
magnesium
magnesium hydroxide,333C/min.
hydroxide,
hydroxide, C/min.
C/min.

InInorder
In order toto
orderto confirm
confirm
confirm the
thethephenomena
phenomena
phenomena responsible
responsible
responsible for
for forthe
the the weight
weight
weight loss,
loss, the
loss,
the theTGA
TGATGA curves, performed
curves,
curves, performed
performed at
atat10
10C/min,
C/min, ofofcompressed
compressed microspheres,
microspheres, PAN,
PAN,and
and magnesium
magnesium hydroxide
10 C/min, of compressed microspheres, PAN, and magnesium hydroxide were compared (Figure 7). hydroxide were
were compared
compared (Figure
(Figure
7).7).PAN
PAN PAN analysis
analysis
analysis isisreported
in ainstudy
reported
is reported ina astudy
study bybyKorobeinyk
Korobeinyk
by Korobeinyk etet[47].
et al. al.al.[47].
It isItIt
[47]. isisobserved
observed
observed thatthat
thethe
that themass
mass lossloss
mass loss
duedue
due
to
to
the the
to PAN PAN
the PAN cyclization
cyclization
cyclization reaction
reaction
reaction matches
matchesmatches the major
the major
the major microspheres
microspheres
microspheres weight
weight weight loss.
loss.loss. A similar
A similar
A similar mechanism
mechanism
mechanism is
isisobserved
observed for
for PMAN
PMAN [48].
[48].WeWe can
canalso
also see
seethat
thatmagnesium
magnesium hydroxide
hydroxide
observed for PMAN [48]. We can also see that magnesium hydroxide degradation occurs in the same degradation
degradation occurs
occurs in
inthe
the
same
sametemperature
temperature range.range.
temperature range.

100
100
9090
8080
Residual weight (%)
Residual weight (%)

7070
6060
5050
4040 Compressed
Compressedmicrospheres
microspheres
3030
PAN
PAN
2020
1010 Magnesium
Magnesiumhydroxide
hydroxide
00
00 100
100 200
200 300
300 400
400 500
500 600
600 700
700
Temperature
Temperature(C)
(C)

Figure 7. Comparison of TGA of compressed microspheres, PAN (polyacrylonitrile) and magnesium

Figure
Figure7.7.Comparison
ComparisonofofTGA
TGAofofcompressed
compressedmicrospheres,
microspheres,PAN
PAN(polyacrylonitrile)
(polyacrylonitrile)and
andmagnesium
magnesium
hydroxide, at 10 C/min, under air. For PAN analysis, see Korobeinyk et al. [47].
hydroxide, at 10 C/min, under air. For PAN analysis, see Korobeinyk et al. [47].
hydroxide, at 10 C/min, under air. For PAN analysis, see Korobeinyk et al. [47].

Thermogravimetric
Thermogravimetric analyses were performed at various heating rates on compressed microspheres
Thermogravimetric analyses analyses were were performed
performed atat various
various heating
heating rates
rates on on compressed
compressed
in order to determine the
microspheres degradationthe temperature. Results are reported in Figure 8. The degradation
microspheresininorderordertotodetermine
determine thedegradation
degradationtemperature.
temperature.Results
Resultsarearereported
reportedininFigureFigure8.8.
temperature
The and polymer degradation depend ondegradation
the temperature ramp.on Moreover, when the heating
The degradation temperature and polymer degradation depend on the temperature ramp.
degradation temperature and polymer depend the temperature ramp.
is performed slowly,
Moreover, the degradation phenomenon the happens gradually. On the contrary, when heating
Moreover,whenwhenthe theheating
heatingisisperformed
performedslowly,slowly, thedegradation
degradationphenomenon
phenomenonhappens happensgradually.
gradually.
is
On performed quickly, the main degradation kinetics are graphically clear, which explains the shape
Onthe
thecontrary,
contrary,when
whenheating
heatingisisperformed
performedquickly,quickly,thethemain
maindegradation
degradationkinetics
kineticsare aregraphically
graphically
differences
clear, between the curves.
clear,which
whichexplains
explainsthe theshape
shapedifferences
differencesbetween
betweenthe thecurves.
curves.
For
For every curve, a small decrease of about 2%3% appears around 200 C;
C; this is attributed to
Forevery
everycurve,
curve,a asmall
smalldecrease
decreaseofofaboutabout2%3%
2%3%appears
appearsaround
around200200 C;thisthisisisattributed
attributedtoto
the
the vaporization of residual hydrocarbon liquid. Then, for every curve the degradation temperature
thevaporization
vaporizationofofresidual
residualhydrocarbon
hydrocarbonliquid. liquid.Then,
Then,for
forevery
everycurve
curvethe
thedegradation
degradationtemperature
temperature
is measured at
isismeasured at aa mass
mass loss
lossof of5%5%tototake
takeinto
intoaccount
accountthe the polymers degradations. The curve of
measured at a mass loss of 5% to take into account the polymers degradations. The curveofofthe
polymers degradations. The curve the
the degradation temperature
degradation in function of thetemperature
temperature rampisisdrawn drawn and the the degradation
degradationtemperature
temperatureininfunctionfunctionofofthe the temperatureramp ramp is drawnand and thedegradation
degradation
temperature
temperatureatata averyverylow lowheating
heatingrate rate(close
(closetoto0 0C/min)
C/min)isisdetermined
determinedby byextrapolation.
extrapolation.As Asisis
shown
shownininFigure
Figure9,9,thisthisvalue
valueisisaround
around235 235C.C.The
The8.5%
8.5%residual
residualmass
masscorresponds
correspondstotothe the
magnesium
magnesiumoxideoxideformation,
formation,whichwhichoccurs
occursbetween
between260260andand360
360CCaccording
accordingtotothe theheating
heatingrate rateand
and
resists
resiststemperatures
temperaturesup uptoto700700C,
C,all
allthe
theother
othercarbon
carbonspecies
speciesbeing
beingtransformed
transformedtotoCO CO 2. 2.
Polymers 2016, 8, 215 9 of 19

temperature at a very low heating rate (close to 0 C/min) is determined by extrapolation. As is shown
in Figure 9, this value is around 235 C. The 8.5% residual mass corresponds to the magnesium oxide
formation,
Polymers 2016,
which
8, 8,
215
occurs between 260 and 360 C according to the heating rate and resists temperatures
9 of 1919
Polymers 2016,
215 9 of
up to 700 C, all the other carbon species being transformed to CO2 .

100
100

8080
Residual weight (%)
Residual weight (%)
6060

4040 1 C/min
1 C/min
5 C/min
5 C/min
2020 1010
/min
/min
2020
C/min
C/min
00
0 0 100
100 200
200 300
300 400
400 500
500 600
600 700
700
Temperature
Temperature(C)
(C)

Figure
Figure 8.8.
8.
Figure TGA ofof
TGA compressed
compressed microspheres
microspheres
compressed atat
at
microspheres various
various temperature
temperature
various ramps,
ramps,
temperature under
under
ramps, air.
air.
under air.
Degradation temperature (C)

300
Degradation temperature (C)

300

280
280

260
260

240
240
Tdeg(v0) = 235
Tdeg(v0) CC
= 235
220
220
00 55 1010 1515 2020
Temperature
Temperatureramp (C/min)
ramp (C/min)

Figure
Figure 9.9.Influence
Figure 9.Influence
Influence ofoftemperature
of temperature
temperature ramp
ramp on ononthe
thedegradation
the degradation
ramp temperaturetemperature
degradation of compressed
temperature ofof compressed
microspheres,
compressed
microspheres,
by TGA, under by TGA,
air. under
microspheres, by TGA, under air. air.

3.2.
3.2. Determination
Determination
3.2. of
Determination the
ofof
the Optimal
Optimal
the Foaming
Foaming
Optimal Temperature
Temperature
Foaming Temperature
In our
InIn our process,
our process,
process,wewe propose
wepropose to manufacture
propose totomanufacture the foam
manufacture thethein onlyinone
foam
foam step.one
inonly
only Then
one we Then
step.
step. need
Then towe determine
we need
needtoto
the temperature
determine
determine the that leads to
thetemperature
temperature thatequilibrium
that leads between the
leadstotoequilibrium
equilibrium microspheres
between
between the foaming kinetic
themicrospheres
microspheres foaming and
foaming the resin
kinetic
kinetic and
and
gelation
the resin kinetic.
gelation First, the
kinetic.gel time
First, of
thethe epoxy
gel time system
of the is determined
epoxy
the resin gelation kinetic. First, the gel time of the epoxy system is determined at differentsystemat different
is temperatures,
determined at after
different
which the microspheres
temperatures,
temperatures, after
afterwhich foaming
which the is studied. foaming
themicrospheres
microspheres foamingisisstudied.
studied.
Gelation
Gelationtime
Gelation timeis
time isismeasured
measuredat
measured atatthethedivergence
the divergenceof
divergence ofofviscosity
viscosityand
viscosity andobviously
and obviouslydepended
obviously dependedon
depended ononthe
the
the
temperature,
temperature, as
as shown
shown in
in
temperature, as shown in Figure 10. Figure
Figure 10.
10.
Microsphere expansion study was performed by TMA at isothermal temperatures, from 160 C
Microsphere
Microsphere expansion
expansion study
study was
was performed
performed byby TMA
TMA atat
isothermal
isothermal temperatures,
temperatures, from
from 160
160CC
to 200 C. Results of analysis are given in Figure 11. The relative expansion is the ratio of the sample
toto
200 C. Results of analysis are given in Figure 11. The relative expansion
200 C. Results of analysis are given in Figure 11. The relative expansion is the ratio of the sample is the ratio of the sample
thickness
thickness
thickness totothe
tothe initial
the sample
initial
initial sample thickness.
sample thickness. For each
thickness. For temperature,
Foreach
eachtemperature, this deformation
temperature, this of expansion
thisdeformation
deformation reaches
ofofexpansion
expansion
areaches
maximum value
reachesa amaximum
maximum thenvalue
drops.
valuethen There
then are two
drops.
drops. phases:
There
There are
aretwo expansion
twophases:
phases:and shrinkage.
expansion
expansion and The
and deformation
shrinkage.
shrinkage. The
The
of expansion
deformation is higher and is happening faster when the temperature
deformation of expansion is higher and is happening faster when the temperature increases. Thisisis
of expansion is higher and is happening faster when the increases.
temperature This
increases.is because
This
of the higher
because
because ofofthe foaming
the higher temperature
higherfoaming
foamingtemperature generated
temperature by a higher
generated
generated pressure,
bybya ahigher which which
higherpressure,
pressure, stretches the cell
whichstretches
stretches thewall
thecell
cell
even
wallmore,
wall even as
evenmore, was
more, reported
asaswas by other
wasreported
reported authors
bybyother [27,42].
otherauthors
authors At 180,
[27,42].
[27,42]. 190,
AtAt180,and
180,190,200
190,and
and C200
the
200Cexpansion
Cthetheexpansionis quite
expansion isis
quite important compared to at temperatures of 160 and 170 C. The
quite important compared to at temperatures of 160 and 170 C. The shrinkage phase is even faster shrinkage phase is even faster
when
whenthe thetemperature
temperatureisishigh. high.
Polymers 2016, 8, 215 10 of 19

Polymers 2016, 8, 215 10 of 19

Polymers 2016,
important 8, 215
compared to at temperatures of 160 and 170 C. The shrinkage phase is even faster
10 of 19
when
the temperature is high.

1200
1200
160
160 C
C
1000
1000 170
170 C
C
(Pa.s)
Viscosity(Pa.s)
800
800 180
180 C
C
190
190 C
C
600
Viscosity
600
200
200 C
C
400
400
200
200 ttgel(160 C)=820 s
gel(160 C)=820 s
00
10
10 100
100 1000
1000
Time
Time (s)
(s)

Figure 10. Determination

Figure of the gel time of TETM/DDS formulation at various temperatures.
Figure 10.
10. Determination
Determination of
of the
the gel
gel time
time of
of TETM/DDS
TETM/DDS formulation
formulation at
at various
various temperatures.
temperatures.

99
Time
Time of
of expansion
expansion at 200 C
at 200 C
88 200
200 Expansion
Expansion 160
160 C
C
77 Expansion 170 C
Expansion 170 C
expansion

(C)
Relativeexpansion

66 150 Expansion
Expansion 180
180 C
Temperature(C)

150 C
55 Expansion
Expansion 190
190 C
Temperature

C
44 100 Expansion
Expansion 200
200 C
C
Relative

100
33 Temperature
Temperature 160 C
160 C
22 50 Temperature
Temperature 170
170 C
C
50
Temperature
Temperature 180
180 C
C
11
Temperature
Temperature 190
190 C
C
00 00
00 55 10 15 20 25 30 Temperature 200 C
Temperature 200 C
10 15 20 25 30
Time
Time (min)
(min)
Figure
Figure 11.
11.
Figure TMA
11.TMA of
TMAof expandable
ofexpandable microspheres
microspheres Expancel
expandable microspheres 950
950 DU
Expancel950
Expancel DU80
DU 80at
80 atvarious
at varioustemperatures.
various temperatures.
temperatures.

It
It has
It has to to
has bebe
to be noted
noted
noted thatthat
thethe
that the expansion
expansion
expansion starts
starts
starts during
during
during the
thethe temperature
temperatureramp
temperature ramp after
after444min,
rampafter min, that
thatisis
min,that istoto
tosay
say
say atat
150 C. This temperature corresponds with the T g2 (see Section 3.1) responsible for the
at 150 C.150
ThisC. This temperature
temperature corresponds corresponds
with thewith Tg 2 the (seeTSection
g2 (see 3.1)
Section 3.1) responsible
responsible for the
for the expansion.
expansion. The The time of
of expansion
expansiontocorresponds to the
the expansion phase: from t == 44 min
min to the
Theexpansion.
time of expansion timecorresponds corresponds
the expansion to phase: expansion
from t = phase:
4 min to from
the tmaximum to the of
value
maximum
maximum value
value of
of the
the curve,
curve, as
as shown
shown in
in Figure
Figure 11
11 for
for the
the temperature
temperature
of
of 200
200 C.
C.
the curve, as shown in Figure 11 for the temperature of 200 C.
Results
Resultsfrom from
fromthe the
theTMA TMA
TMA and and epoxy
and epoxy system
system gel
epoxy system gel time analysis are gathered in Figure 12.
Results gel time time analysis
analysisare aregathered
gatheredin inFigure Figure 12.12.
TETM/DDS/Microspheres
TETM/DDS/Microspheres foam foam samples
foam samples were manufactured as described in Section 2 and cured at
at aa at
TETM/DDS/Microspheres sampleswere weremanufactured
manufactured asas
described
described in Section
in Section 2 and cured
2 and cured
constant temperature
constant temperature
temperature from from 160 to 200 C, with the aim of comparing with the previous analysis. The
a constant
pictures from160 160toto 200200C, C,
withwiththe the
aimaim of comparing
of comparing with with
the previous analysis.
the previous The
analysis.
pictures are are given
given inin Figure
Figure 12, 12, associated
associated withwith each
each temperature.
temperature.
The pictures
Between are given in Figure 12, associated with each temperature.
Between 160 160 and
and 180180 C,
C, the
the gel
gel time
time isis quite
quite high
high compared
compared to to the
the time
time of of expansion.
expansion. At At 160
160 and
and
170Between 160 and 180 C, the gel time is quite high compared to the time of expansion. At 160
170 CCthe
the pictures
pictures showshow that that expansion
expansion is is very
very limited
limited and and aa phase
phase separation
separation can can bebe observed:
observed: the the
and 170 Care
bubbles thelighter
picturesthan show
the that expansion
resin and go up is the
to verysurface.
limitedThis andphenomenon
a phase separationcan canbecause
occur be observed:
the
bubbles are lighter than the resin and go up to the surface. This phenomenon can occur because the
theresin
bubbles are lighter than theon. resin and go upthe to the surface. This phenomenon can occur because the
resin gelation
gelation occurs
occurs laterlater on. At At180
180 C,C, the picture
picture shows shows aa great
great expansion,
expansion, without
without phasephase
resin gelation
separation. occurs
Indeed, later
as on.
was At 180
shown C,
in the picture
Figure 11, shows
at this a great expansion,
temperature
separation. Indeed, as was shown in Figure 11, at this temperature the expansion is relatively the without
expansion phase
is separation.
relatively
Indeed, as wasMoreover,
important.
important. shown in even
Moreover, Figure
even ifif11, theatgelation
the this temperature
gelation occurs
occurs late, thethe
late, expansion
the shrinkage
shrinkageis relatively
kinetic
kinetic is isimportant.
quite slowMoreover,
quite slow so
so the
the
bubble
even if the size decreases
gelation occurs
bubble size decreases slowly. slowly.
late, the shrinkage kinetic is quite slow so the bubble size decreases slowly.
Polymers 2016, 8, 215 11 of 19
Polymers 2016, 8, 215 11 of 19

Between 190 and

Between and C,
190 200 200theC,gel time
the gel istime
close
is to the time
close to theoftime
expansion. The expansion
of expansion. is important
The expansion is
important
and faster; thereand
is nofaster;
phase there is no phase
separation. separation.
At these At these the
temperatures temperatures
shrinkagethe shrinkage
kinetic is fast,kinetic is
as is shown
fast, 11,
in Figure as is shown
which in Figure
confirms 11,the
that which confirms
gelation has that
to bethe gelation
close to thehas to be
time close to thetotime
of expansion avoidof the
expansion to avoid the shrinkage of the foam. Furthermore, a progressive
shrinkage of the foam. Furthermore, a progressive darkening is observed inside the sample at 190 and darkening is observed
inside the sample at 190 and 200 C. This coloration is due to the exothermicity of the curing reaction
200 C. This coloration is due to the exothermicity of the curing reaction so the temperature in the sample
so the temperature in the sample is greater than the isothermal temperature in the oven. Therefore it
is greater than the isothermal temperature in the oven. Therefore it is reasonable to think that an oxidative
is reasonable to think that an oxidative reaction occurred in these samples. The cause of this
reaction occurred in these samples. The cause of this oxidation is explained later in the text.
oxidation is explained later in the text.
The density datadata
The density forforthese
thesesamples
samples were measuredbyby
were measured using
using a pycnometer
a pycnometer and
and are are gathered
gathered in
in Table 2. The
Table density
2. The densitydecreases
decreases with
with the
the temperature, whichconfirms
temperature, which confirms that
that thethe expansion
expansion is more
is more
important at 190atand
important 200 200
190 and C than
C thanat 180 C. However, a better compromise would be a temperature of
at 180 C. However, a better compromise would be a temperature

180 C, of because
180 C, because
it alliesit aallies
great a great expansion
expansion withwith a slow
a slow shrinkage
shrinkage andavoids
and avoidsthermal
thermal degradation.
degradation.

18
Time of expansion Gel time
16

14

12
Time (min)

10

0
150 160 170 180 190 200
Temperature (C)

Figure
Figure 12. Microspheres
12. Microspheres TMATMA andepoxy
and epoxysystem
system gel
gel time
timeanalysis:
analysis:correlation of results.
correlation of results.

Table 2. Density of foam samples.

Table 2. Density of foam samples.
Temperature 160 C 170 C 180 C 190 C 200 C
Temperature 160 C 170 C C
Density (kgm ) 909
3 14 797 C22 676 27 495 38 462 42 200 C
180 190
Density (kg m3 ) 909 14 797 22 676 27 495 38 462 42
A SEM (scanning electronic microscopy) micrograph of the epoxy formulation/expandable
microsphere foamed at 180 C is given in Figure 13. The cell size distribution is homogeneous, which
Ais SEM
a sign (scanning electronic microscopy)
of good synchronization between the micrograph of the
occurrence of the epoxy
gelation andformulation/expandable
the time of expansion
microsphere foamed at 180 C is given in Figure 13. The cell size distribution is homogeneous, which is
[42].
a sign of good synchronization
In order to observe the between
behaviorthe of
occurrence
different of the gelation
grades and themicrospheres,
of expandable time of expansion
three [42].
additional
In order to expandable
observe themicrospheres
behavior ofgrades were
different selected
grades of and tested at 180
expandable C (920 WUFthree
microspheres, 40, 920 DU
additional
40, 980 DUX
expandable 120). They grades
microspheres were chosen
wereaccording
selectedtoandtheirtested
expansion temperature,
at 180 C (920 WUFwhich 40,
was 920
around
DU 40,
180 C. Referring to the supplier data sheet, these four grades are made
980 DUX 120). They were chosen according to their expansion temperature, which was around of the same copolymer:
acrylonitrilemethylmetacrylatemethacrylonitrile. However, the hydrocarbons inside the shell are
180 C. Referring to the supplier data sheet, these four grades are made of the same copolymer:
not the same for all. For instance, 920 WUF 40 and 920 DU 40 contain isopentane, 980 DUX 120
acrylonitrilemethylmetacrylatemethacrylonitrile. However, the hydrocarbons inside the shell are not
contains isooctane, and 950 DU 80 contains an isopentaneisooctane mix. As was explained in
the same for 3.1,
Section all. expansion
For instance, 920 WUF
is possible only40ifand
the 920
shellDU 40 contain
is softened isopentane,
enough. Then the980 DUX 120 contains
hydrocarbons are
isooctane,
allowed to expand. Knowing the nature of the copolymer is not sufficient becauseinthe
and 950 DU 80 contains an isopentaneisooctane mix. As was explained Section
Tg is 3.1,
expansion is possible only if the shell is softened enough. Then the hydrocarbons are allowed to
expand. Knowing the nature of the copolymer is not sufficient because the Tg is dependent on the
molecular weight of each copolymer. So the nature of the gas inside the shell as well as the mass of
each copolymer is needed to predict the temperature of expansion.
 Polymers 2016, 8, 215 12 of 19

Polymers 2016,
Polymers 8, 215
2016, 8, 215 1219
12 of of 19
dependent on the molecular weight of each copolymer. So the nature of the gas inside the shell as
well as the mass of each copolymer is needed to predict the temperature of expansion.
dependent on the molecular weight of each copolymer. So the nature of the gas inside the shell as
well as the mass of each copolymer is needed to predict the temperature of expansion.

Figure13.
Figure 13.SEM
SEMmicrograph
micrograph of the foam
foam sample
sample(magnification
(magnification30x).
30x).
Figure 13. SEM micrograph of the foam sample (magnification 30x).
FourFour
epoxyepoxy formulation/expandable
formulation/expandable microspheres
microspheres mixes mixes were prepared.
were prepared. The epoxyThe epoxy
formulations
formulations
were prepared were prepared
as before, as before, in
in an aluminum crucible,an aluminum crucible,
at 180 C. The at 180
temperatureC. The temperature
was dropped was C,
Four epoxy formulation/expandable microspheres mixes were prepared. The toepoxy
100
dropped to 100 C, at which time the microspheres were added and mixed until apparent
at which time the
formulations microspheres
were prepared as were added
before, in and mixed until
an aluminum apparent
crucible, at homogeneity. The four mixtures
180 C. The temperature was
homogeneity. The four mixtures had exactly the same weight and a proportion of 5% of
had exactly the
dropped sameC,
to 100 weight and atime
at which proportion of 5% of microspheres.
the microspheres were addedThen, and the crucibles
mixed until were placed
apparent
microspheres. Then, the crucibles were placed in an oven at 180 C, for 30 min.
C, for
in homogeneity.
an oven at 180 The four30 mixtures
min. had exactly the same weight and a proportion of 5% of
The pictures of the cured mixes with the four different expandable microspheres are presented
microspheres. Then, the crucibles
The pictures of the cured mixes with were placed in andifferent
the four oven at 180 C, for 30 microspheres
expandable min. are presented
in Figure 14. Tstart corresponds to the temperature when the foaming begins and Tmax corresponds to
in Figure The14.
pictures of the cured mixes
Tstart corresponds to thewith the four different
temperature when the expandable microspheres
foaming begins and Tmax are corresponds
presented
the temperature when the microsphere volume is maximal. These values are issued from the
in
theFigure
to technical 14. Tstart corresponds
temperature to the temperature when the foaming begins and Tare
max corresponds to
data sheet. The samples were cut down the middle into two parts; we can see thefrom
when the microsphere volume is maximal. These values issued 5 cmthe
the
technicaltemperature
data sheet. when the microsphere
The samples were the volume
cut920
down is
themaximal. These values are issued from the
length profiles. The samples containing WUF 40middle
and 920intoDU 40two parts; we can
microspheres seeaverage
show the 5 cm
technical
length dataThe
profiles. sheet. The samples
samples werethe
containing cut920
downWUFthe40 middle
and into
920 DU two
40 parts; we can see
microspheres showthe average
5 cm
expansion and bubbles of heterogeneous and large dimensions. This is certainly due to coalescence.
length
expansion profiles. The
and bubbles samples containing the 920 WUF 40 and 920 DU 40 microspheres show average
As the bubbles grow, of heterogeneous
adjacent ones tend and
to large dimensions.
coalesce in order to This is certainly
reduce the totaldue toenergy
free coalescence.
by
expansion and bubbles of heterogeneous and large dimensions. This is certainly due to coalescence.
Aslowering
the bubblesthe grow,
surface adjacent
area of ones tend to coalesce
cells [4951]. in order tohere
This is happening reduce the total
because free energy bycapsule
the thermoplastic lowering
As the bubbles grow, adjacent ones tend to coalesce in order to reduce the total free energy by
thecollapsed
surface area
due ofto cells [4951].
the high This is happening
temperature here becausereaction
while the crosslinking the thermoplastic capsule
was not really collapsed
consequent.
lowering the surface area of cells [4951]. This is happening here because the thermoplastic capsule
due to theithigh
Indeed, has temperature while the
been demonstrated by crosslinking
various authors reaction
that was not really
too low consequent.
a viscosity increasesIndeed,
the
collapsed due to the high temperature while the crosslinking reaction was not really consequent.
probability
it has of coalescenceby
been demonstrated [50,52].
various authors that too low a viscosity increases the probability of
Indeed, it has been demonstrated by various authors that too low a viscosity increases the
coalescence [50,52].
probability of coalescence [50,52].

Figure 14. Expansion temperature data of microspheres and pictures of expandable microspheres/
epoxy formulation foams cured at 180 C.
Figure
Figure Expansion
14.14. Expansiontemperature
temperaturedata
data of
of microspheres and pictures
microspheres and picturesofofexpandable
expandablemicrospheres/
microspheres/
epoxy
epoxyformulation
formulationfoams
foamscured 180C.
curedatat180 C.
The sample containing the 950 DU 80 microspheres presents greater expansion and bubbles of
homogeneous and microscopic dimensions.
The
The samplecontaining
sample containing the950
950 DU 80 80 microspheres
microspheres presents greater expansion and bubbles of of
The sample containingthe the 980DU DUX 120 presents
microspheres shows greater expansion
a low expansion, onand
thebubbles
surface
homogeneous
homogeneous andmicroscopic
and microscopicdimensions.
dimensions.
only. Referring to producer data, this grade needs higher temperature to expand than previous ones.
The
The samplecontaining
sample containingthe
the 980DUXDUX 120 microspheres
microspheres shows aalow expansion, onon
thethe
surface
The temperature in the core is 980
not high 120
enough to produceshows
a wholelow expansion,
expansion compared tosurface
the
only.
only. Referring
Referring toto producer
producer data,
data, this
this grade
grade needs
needs higher
higher temperature
temperature totoexpand
expandthan previous
than previousones.
ones.
surface because of the weak thermal conductivity of epoxy resins.
The temperature in the core is not high enough to produce a whole expansion compared to the
The temperature in the core is not high enough to produce a whole expansion compared to the surface
surface because of the weak thermal conductivity of epoxy resins.
because of the weak thermal conductivity of epoxy resins.
Polymers 2016, 8, 215 13 of 19

Polymers 2016, 8, 215 13 of 19

3.3. TETM/DDS Cure Cycle Optimization

3.3. TETM/DDS Cure Cycle Optimization
Based on the previous study, a first temperature T1 = 180 C was set for the curing process of
Based on the previous study, a first temperature T1 = 180 C was set for the curing process of the
the TETM/DDS
TETM/DDS system. DSCanalyses
system. DSC analyseswerewereconducted
conducted as as described
described before.
before. Results
Results are presented
are presented in in
FigureFigure
15. 15.
Heat flow ( W/g)

30 55 80 105 130 155 180 205 230 255 280 305 330
Temperature (C)

30 min 40 min 50 min 1h 2h 3h

15. DSC scanscan

of TETM/DDS C, performed
FigureFigure 15. DSC of TETM/DDSformulations
formulations after varioustimes
after various timesofof cure
cure at at
180180
C, performed
underunder
nitrogen withwith
nitrogen a heating rate
a heating 5 5C/min.
of of
rate C/min.

An exothermic
An exothermicpeakpeak
is remained around
is remained around C C
240240 forfor
each time
each of of
time cure. The
cure. Thelonger
longerthe
thecuring
curinglasts,
lasts, the
the lower the remaining
lower the remaining peakHowever,
peak height. height. However,
even aftereven
3 hafter 3 h C,
at 180 at 180 C, the
the cure curecompleted.
is not is not
completed. The T g depends on the degree of cure reaction [53]. Then the system needs to be fully
The Tg depends on the degree of cure reaction [53]. Then the system needs to be fully cured to increase
the Tgcured
. to increase the Tg.
A post-curing temperature T2 = 240 C was then added to the cycle. To determine the optimal
A post-curing temperature T2 = 240 C was then added to the cycle. To determine the optimal
cure thermal cycle, six curing cycles were prepared and tested by thermomechanical analysis using a
cure thermal cycle, six curing cycles were prepared and tested by thermomechanical analysis using a
dynamic rheometer equipped with rectangular torsion geometry:
dynamic rheometer equipped with rectangular torsion geometry:
C1: 1 h at180 C + 1 h at 240 C
C1 : 1C : 1 h at180
h 2at180 C +C1 h+ 2ath240 C
at 240 C
C2 : C
1 h at 180 C + 2 h at 240 CC
3 : 2 h at 180 C + 1 h at 240

C3 : 2C : 2 180
h 4at h at 180
C +C1 +h2ath 240
at 240
CC
C5: 3 h at 180 C + 1 h at 240 C
C4 : 2 h at 180 C + 2 h at 240 C
C6: 3 h at 180 C + 2 h at 240 C
C5 : 3 h at 180 C + 1 h at 240 C
The temperature ramp between the two stages was 3 C/min. DSC analyses of each sample after
C6 : 3 h at 180 C + 2 h at 240 C
the two steps curing cycles were performed and no remaining peak was observed.
Results fromramp
The temperature rectangular
between torsion tests,
the two performed
stages was 3 atC/min.
an angular frequencyofofeach
DSC analyses 1 rad/s,
sample areafter
gathered in Figure 16. The samples that went through a second
the two steps curing cycles were performed and no remaining peak was observed. stage of 1 h at 240 C (C 1, C3, and C5)

are similar, likewise the samples that had a second stage of 2 h at 240 C (C2, C4, and C6). Two alpha
Results from rectangular torsion tests, performed at an angular frequency of 1 rad/s, are gathered
relaxations, a rheological signature of glass transition temperature, are observed at the peak on the
in Figure 16. The samples that went through a second stage of 1 h at 240 C (C1 , C3 , and C5 ) are similar,
G curves of C1, C3, and C5 samples, 1 at 230 C, and 2 at 330 C. Indeed, G is sensitive to the
likewise the samples that had a second stage of 2 h at 240 C (C2 , C4 , [54].
long-range molecular motion occurring during the relaxation
and C61).indicates
Two alpha relaxations,
molecular
a rheological
movements, which confirms that the curing process is not over and that more time for more energyof C1 ,
signature of glass transition temperature, are observed at the peak on the G curves
C3 , and C5 samples,
is needed aroundthis 230 C, and
1 at temperature. Evolution

2 at 330 inC.theIndeed, G is sensitive
polymerization reactiontoisthe long-range
occurring during molecular
this
motion occurring
analysis. The during the relaxation
macromolecules [54]. allows
conformation 1 indicates
for molecular
the reaction movements,
of chemical which
entities ( confirms
1 ). As the that
heating is slow during the test, the molecules have the time to react, leading to
the curing process is not over and that more time for more energy is needed around this temperature. 2 , which actually is
at the same temperature as . Then, this test is allowed to complete the curing
Evolution in the polymerization reaction is occurring during this analysis. The macromolecules reaction for samples
that went allows
conformation throughfor
a second stage ofof
the reaction only 1 h at 240
chemical C (C1(
entities , C3,).and C5). The second stage of 2 h at 240
1 As the heating is slow during the test,
C is then needed to complete the curing process. However, a first stage of 1 h at 180 C is enough
the molecules have the time to react, leading to 2 , which actually is at the same temperature as .
Then, this test is allowed to complete the curing reaction for samples that went through a second stage
of only 1 h at 240 C (C1 , C3 , and C5 ). The second stage of 2 h at 240 C is then needed to complete
the curing process. However, a first stage of 1 h at 180 C is enough since there is no difference when
Polymers 2016, 8, 215 14 of 19

Polymers
increasing
Polymers2016,
the 8, 8,
215
time
2016, of the first stage from 1 h to 3 h (C2 , C4 , and C6 ). Finally, the optimized 14
215 of of
cure
14 19cycle
19 is

1 h at 180 C and 2 h at 240 C (C2 ). The alpha () temperature, evaluated at the peak on the G curve,
since
sincethere
there is nono difference when increasing the time of the first stage from 1 h to 3 h (C2, C4, and C6).
is then equal to is330 difference
2 C. when increasing the time of the first stage from 1 h to 3 h (C2, C4, and C6).
Finally,
Finally, the optimized curecycle
the optimized cure cycleis is1 1h hatat180
180CCandand2 2h hatat240
240CC(C(C2). The alpha () temperature,
2). The alpha () temperature,
Epoxy
evaluated foam
at the was
peak realized
on the as
G described
curve, is then in Section
equal to 2. The
330 2 mixture was
C. cured in an oven, where the
evaluated at the peak on the G curve, is then equal to 330 2 C.
cycle C2Epoxywas applied. Parallelepipedal samples were cut from the foamed material andwhere rectangular
Epoxyfoam foamwas wasrealized
realizedasasdescribed
describedininSection
Section2. 2.The
Themixture
mixturewas wascured
curedininananoven,
oven, where
torsion
the analyses
thecycle
cycleC2C2was
were performed.
wasapplied.
Results are given
applied.Parallelepipedal
Parallelepipedalsamples
in Figure
sampleswere werecut
17, with
cutfrom
fromthe
the curves
thefoamed
ofmaterial
the non-foamed
foamedmaterial andand
sample previously
rectangular obtained. The moduli values of the foam are logically
rectangular torsion analyses were performed. Results are given in Figure 17, with the curves ofthe
torsion analyses were performed. Results are given in Figure lower
17, with because
the curvesof air
of presence.
the
Thenon-foamed
Tnon-foamed
of the foam sample
sample previously
is also obtained.
less important.
previously obtained. The
Indeed,
Themodulithevalues
moduli ofof
endothermic
values the
thefoam are
reaction
foam logically
are due tolower
logically the because
vaporization
lower because of
the of airair
liquid
of presence.
during
presence. The
the
TheTTof the foam
expansion of is
of the foam the also
is less important.
microspheres
also less important. Indeed,
consumes
Indeed, the endothermic
athe
part reaction
of the heat
endothermic due
released
reaction duetoduring
the
to the the
vaporization
vaporization
exothermic of the liquid
of the liquid
cure reaction during
of theduring the
TETM/DDS expansion
the expansion of
system. the microspheres consumes a part
of the microspheres consumes a part of the heat of the heat
released during
released the
during exothermic
the cure
exothermic reaction
cure ofof
reaction the TETM/DDS
the system.
TETM/DDS system.

Figure
Figure 16.16.
Figure Rheological
Rheological
16. analyses
analyses
Rheological ofof
of
analyses TETM/DDS
TETM/DDS formulations
TETM/DDS cured
formulations
formulations according
cured
cured toto
according
according various cure
to various
various cycles.
curecure cycles.
cycles.

Figure
Figure 17.17.
Figure Comparison
17. ofofrheological
Comparison
Comparison rheological
of analyses
rheological of
analyses
analyses foamed
ofof and
foamed
foamed non-foamed
and
and sample,
non-foamed
non-foamed cured
sample,
sample, according
cured
curedaccording
according to
toto
C2Ccycle.
C2 cycle. 2 cycle.
The
TheDSC
DSCanalyses
analysesininFigure
Figure1818allowallowusustotocompare
comparethe thethermograms
thermogramsofofTETM/DDS
TETM/DDS
The DSC analyses
formulation, in Figure 18ofallow us to compare the thermograms of TETM/DDS formulation,
formulation, one containing 5% of expandable microspheres and one without microspheres,with
one containing 5% expandable microspheres and one without microspheres, with
onethe
containing 5% of expandable microspheres and one without microspheres, with the thermogram
thethermogram
thermogramofofexpandable
expandablemicrospheres
microspheresonly. only.The
Thevalue
valueofofheat
heatflow
flowforformicrospheres
microsphereswas was
of expandable
reducedtotomicrospheres
reduced the
theweight
weightofofonly. The valuecontained
hydrocarbons
hydrocarbons of heat flow
contained for
ininthe microspheres
theepoxy
epoxyformulation was with
formulation reduced to the weight
withexpandable
expandable
of hydrocarbons
microspheres,ascontained
microspheres, asthey
theyare in the epoxy
areresponsible
responsible formulation
forforthe
theendothermic
endothermic with expandable
reaction,
reaction,oror1% microspheres,
1%ofofthe
theepoxy as they are
epoxyformulation
formulation
responsible for the endothermic reaction, or 1% of the epoxy formulation (20% of hydrocarbons in
5% of microspheres). There is a small interval of 5 C between the peaks on the thermograms of
 Polymers 2016, 8, 215 15 of 19
Polymers 2016, 8, 215 15 of 19
Polymers 2016, 8, 215 15 of 19
(20% of hydrocarbons in 5% of microspheres). There is a small interval of 5 C between the peaks on
thethe
(20%thermograms
of hydrocarbons
formulation of the
without formulation
inand
5% of withoutThere
microspheres).
with microspheres, andcorresponding
with microspheres,
is a small interval 15corresponding
to of 5min.
C between topeaks
the 15 min.
This temperature This is
on shift
the thermograms
temperature shiftofisthe formulation
explained by without
the and with peak
endothermic microspheres,
of corresponding
vaporization, which todelays
15 min.theThis
curing
explained by the endothermic peak of vaporization, which delays the curing reaction. As the expansion
temperature
reaction. As shift is explained by the during
the expansion endothermic peak of vaporization,
of the curingwhich delays the curing
happens during the first stagehappens
of the curing processthe first
(at stage
180 C), process
it was postulated (at 180
that more C), it was
time at this
reaction. As the expansion happens during the first stage of the curing
postulated that more time at this temperature was needed to complete the curing reaction process (at 180 C), it was of the
temperature
postulated
was needed
that more time
to complete
at 230
this C,
the
temperature
curing reaction
was needed
of the foam.
to complete
At temperatures over 230 C,
foam. At temperatures over the increased heat flow encircledthe curing
in the reaction ofofthe
thermogram epoxy
thefoam.
increased heat flow
At temperatures encircled in
over 230 is the thermogram
C,related
the increased of epoxy formulation
heat flow encircled with microspheres
of epoxy related
is
formulation with microspheres to a degradation, which is in the thermogram
discussed later.
to aformulation
degradation,
withwhich is discussed
microspheres later.
is related to a degradation, which is discussed later.

0.2
0.2
0.1
0.1
(W/g)
Flow(W/g)

0
0
HeatFlow

-0.1
-0.1
Heat

-0.2
-0.2

-0.3
-0.3
100
100 150
150 200200 250 250
Temperature (C)(C)
Temperature
without
withoutexpandable microspheres
expandable microspheres
with expandable microspheres
with expandable microspheres
uncompressed
uncompressed microspheres
microspheres

Figure DSC analyses

18.18.
Figure of compressed microspheres and TETM/DDS
TETM/DDSformulation
formulationwith
withand
andwithout
Figure 18. DSC
DSC analyses
analysesofofcompressed
compressedmicrospheres
microspheres and
and
C/min.TETM/DDS formulation with and
expandable microspheres, under air, at a heating rate of 3
without expandable microspheres, under air, at a heating rate of 3 C/min.
without expandable microspheres, under air, at a heating rate of 3 C/min.

Then
Then twoother
two other cycles were tested forfor
2 h at 180 180
C +2Ch+at2240 C,240
and 3h
C, at 180 Cat + 2180
h at C
240+ 2 h at
Then two othercyclescycleswere
were tested
tested for 22hhatat180 h 240
C + 2 h at at C, and and
3h at3h
180 C + 2 h at 240
C
240C C (C 4 and C6). Results are given in Figure 19. Applying the cure cycle C2 allows for reaching 250 2
(C(C 4 and
4 and C6).CResults
6 ). Results are given
are given in Figure
in Figure 19. Applying
19. Applying the curethe cure
cycle C2cycle
allowsC2forallows
reachingfor 250
reaching
2
C as the glass transition temperature. Increasing the time of the first stage of 1 h allows for
250C as2 C theasglass
the glass transition
transition temperature.
temperature. Increasing
Increasing the the time
time of of first
the the first stage
stage of 1ofh1 allows
h allowsforfor
increasing the glass transition temperature by about 20 C, leading to a T of 270 2 C. However, a
increasing
increasing the glass
the glass transition
transition temperature
temperature by
by about 20
about 20C, leadingtotoa aTTof
C,leading of
270270
2 C.2 However,
C. However,a
first stage of 3 h at 180 C is not needed.
a first
firststage
stageof
of 33 hh at
at 180 C
C is
is not
notneeded.
needed.

Figure 19. Rheological analyses of TETM/DDS/expandable microspheres foam cured according to

various cure cycles, with a heating rate of 3 C/min.
Figure 19. Rheological analyses of TETM/DDS/expandable microspheres foam cured according to
Figure 19. Rheological analyses of TETM/DDS/expandable microspheres foam cured according to
various cure cycles, with a heating rate of 3 C/min.
various cure cycles, with a heating rate of 3 C/min.
Polymers 2016, 8, 215 16 of 19
Polymers2016,
Polymers
Polymers 2016,8,
2016, 8,215
8, 215
215 16of
16
16 of19
of 19
19

Thefoams
The
The
The foamspresented
foams
foams presentedaaaadark
presented
presented dark
dark coloration,
coloration,
darkcoloration,
coloration,as as
as shown
shown
asshown
showninin
in Figure
Figure
inFigure 20.
20.
Figure20. We
We
20.We
We assume
assume
assume
assume that
that
that this
this
this
that coloration
coloration
coloration
this coloration is
is
is
due
due
due
is due to
to
to tothe
the degradation
degradation
thethedegradation of
of expandable
expandable
degradationofofexpandable microspheres
microspheres
expandablemicrospheres
microspheresand and
and particularly
particularly
andparticularly
particularlytoto
to the
the
to the polymerization
polymerization
the polymerization
polymerization of of
of
of
sequences
sequences
sequences of
of
of adjacent
adjacent
adjacent nitrile
nitrile
nitrile groups
groups
groups [55],
[55],
[55], in
in
in the
the
the cyclization
cyclization
cyclization reaction
reaction
reaction occurring
occurring
occurring in
in
in
sequences adjacent nitrile groups [55], in the cyclization reaction occurring in polyacrylonitrile and polyacrylonitrile
polyacrylonitrile
polyacrylonitrile
and polymethacryclonitrile
and
and polymethacryclonitrile
polymethacryclonitrile
polymethacryclonitrile
[47,48].
[47,48].
[47,48].
[47,48].
Indeed,
Indeed,
Indeed,
Indeed,
this darkening
this
this
this darkening
darkening
darkening
occurred
occurred
occurred
occurred during the
during
during during
the post-curing
the
the post-curing
post-curing
post-curing atas
at
at
at 240 C,
240
240
240
is
C,
C,
C, as
as is
is highlighted
highlighted
as is highlighted
highlighted
in
in Figure
Figure
in Figurein21.
Figure 21.21.
21.

Figure20.
Figure
Figure
Figure 20.Cross-sectional
20. Cross-sectionalpicture
Cross-sectional
Cross-sectional pictureof
picture ofTETM/DDS/expandable
of TETM/DDS/expandablemicrospheres
TETM/DDS/expandable
TETM/DDS/expandable microspheres
microspheres foam,cured
foam,
microspheresfoam,
foam, curedaccording
cured
cured accordingto
according
according to
to
to
theC
the
the
the C222cycle
C
C cycle(length
cycle
cycle (length===
(length
(length =7777cm).
cm).
cm).
cm).
2

Figure21.
Figure
Figure 21.Pictures
21. Picturesof
Pictures ofsamples
of samplescured
samples curedwithout
cured withoutaaapost-cure
without post-cure(left)
post-cure (left)and
(left) andwith
and withaaapost-cure
with post-cure(right)
post-cure
post-cure (right)at
(right)
(right) at240
at
at 240C.
240
240 C.
C.
C.

Moreover,in
Moreover,
Moreover,
Moreover, inorder
in
in orderto
order
order toconfirm
to
to confirmthat
confirm
confirm thatthe
that
that theTETM/DDS
the
the TETM/DDSformulations
TETM/DDS
TETM/DDS formulationsdo
formulations donot
do notstart
not startto
start todegrade
to degradeduring
degrade during
during
thepost-cure
the post-cureat
post-cure at240
at 240C,
240
C,
C, TGAanalyses
C, TGA
TGA analyseswere
analyses wereperformed
were performedon
performed ontwo
on twosamples
two samples(Figure
samples (Figure22).
(Figure 22).One
22). Onesample
One samplehas
sample has
has
the
been
been subjected
subjected
subjected to to
to the
the
the first
first
first stagestage
stage
at at
at
180 180
180
C C
C
only, only,
only,
and and
and
the the
the
second second
second
one hasone
one
been has
has been
been
subjected subjected
subjected
to the to
to
post-cure the
the
been subjected to the first stage at 180 C only, and the second one has been subjected to the
post-cure
at
post-cure
240 C.at at
at
We 240
240
can C.
C. We
We can
observe can
canthatobserve
observe that
that
thesethatcurves these
these curves
curves
arecurves are similar,
arewhich
similar, similar, which confirms
which confirms
confirms confirms
that that the
that the post-cure
the post-cure
the post-cure post-cure
does not
post-cure 240 C. We observe these are similar, which that
does
does not
not initiate
initiate the
the degradation
degradation
of the of
of
epoxy the
the epoxy
epoxy
resin. The resin.
resin. The
The experimental
experimental
does not initiate the degradation of the epoxy resin. The experimental result corroborates that
initiate the degradation experimental result result
result
corroboratescorroborates
corroborates
the fact the fact
the
the fact
fact
the
that
that the
the
increase increase
increase
in the in
in the
the
thermogram thermogram
thermogram
(Figure (Figure
(Figure
18) of an18)
18) of
of
epoxy an
an epoxy
epoxy formulation
formulation
formulation with with
with microspheres
microspheres
microspheres at 230 Cat
at 230
230
is C
C
only
that the increase in the thermogram (Figure 18) of an epoxy formulation with microspheres at 230 C
is only
is only
related
is related
only related
related to
to thethe degradation
the degradation
to the degradation
to degradation of the
of the microspheres.
the microspheres.
of the microspheres.
of microspheres.

Figure22.
Figure
Figure 22.Comparison
22. Comparisonof
Comparison oftwo
of twoTETM/DDS
two TETM/DDSsamples
TETM/DDS
TETM/DDS samples
samples cured
cured
samplescured underdifferent
under
curedunder
under differentcuring
different curingcycles,
curing cycles,by
cycles, byTGA,
by TGA,at
TGA, ataaa
at
heating rate
heating rate
heating of
rate of 10
of 10 C/min,
10 C/min,
C/min,
under
under
C/min,under air.
air.
underair.
air.
Polymers 2016, 8, 215 17 of 19

4. Conclusions
This paper reports on a complete study leading to the manufacture of a high temperature epoxy
foam. The analyses performed on the expandable microspheres highlighted the different steps leading
to the expansion. The copolymer shell passes through two main glass transition temperatures: the first
one offers initial softening to reduce local pressure in the inner part of cell wall, whereas the second
one leads to the hydrocarbons vaporization and the cell expansion. It was demonstrated that the
expansion kinetics performed by TMA is even more important at higher temperatures. These results
were correlated to the gel time of the epoxy formulation. Indeed, in order to maintain the bubble
size and a homogeneous size distribution, gelation and foaming times have to be close. A good
compromise temperature is 180 C, because it allies great expansion with slow shrinkage and avoids
thermal degradation.
In the second part of this paper, the TETM/DDS formulation curing process was optimized and a
T of 330 2 C was obtained after a two-step curing cycle. Finally, the epoxy foam was manufactured
and its curing process was optimized, leading to a T of 270 2 C.

Acknowledgments: The authors express their sincere thanks to Roxel for the financial support. Huntsman
Advanced Materials and Rossow are gratefully acknowledged for supplying products.
Author Contributions: Samira El Gazzani performed the experiments. Samira El Gazzani and Valrie Nassiet
analyzed the data and wrote the paper. Jean-Pierre Habas and Christian Freydier helped analysis and discussed
about the data. Aline Hilleshein contributed performing the experiments.
Conflicts of Interest: The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript:
TETM Triglycidyl ether of tris(4-hydroxyphenyl)methane
DDS Diaminodiphenylsulfone
DSC Differential scanning calorimetry
TGA Thermogravimetric analysis
TMA Thermomecanical analysis
SEM Scanning electronic microscopy

References
1. Mazzon, E.; Habas-Ulloa, A.; Habas, J.-P. Lightweight rigid foams from highly reactive epoxy resins derived
from vegetable oil for automotive applications. Eur. Polym. J. 2015, 68, 546557. [CrossRef]
2. Wang, L.; Yang, X.; Jiang, T.; Zhang, C.; He, L. Cell morphology, bubbles migration, and flexural properties
of non-uniform epoxy foams using chemical foaming agent. J. Appl. Polym. Sci. 2014, 131, 41175. [CrossRef]
3. Luo, Y.; Zhang, J.; Qi, R.; Lu, J.; Hu, X.; Jiang, P. Polylactide foams prepared by a traditional chemical
compression-molding method. J. Appl. Polym. Sci. 2013, 130, 330337. [CrossRef]
4. Huang, Q.; Klotzer, R.; Seibig, B.; Paul, D. Extrusion of microcellular polysulfone using chemical blowing
agents. J. Appl. Polym. Sci. 1998, 69, 17531760. [CrossRef]
5. Rutz, B.H.; Berg, J.C. A review of the feasibility of lightening structural polymeric composites with voids
without compromising mechanical properties. Adv. Colloid Interface Sci. 2010, 160, 5675. [CrossRef]
[PubMed]
6. Lacoste, C.; Basso, M.C.; Pizzi, A.; Laborie, M.-P.; Celzard, A.; Fierro, V. Pine tannin-based rigid foams:
Mechanical and thermal properties. Ind. Crops Prod. 2013, 43, 45250. [CrossRef]
7. Basso, M.C.; Li, X.; Fierro, V.; Pizzi, A.; Giovando, S.; Celzard, A. Green, formaldehyde-free, foams for termal
insulation. Adv. Mater. Lett. 2011, 2, 378. [CrossRef]
8. Basso, M.C.; Lagel, M.-C.; Pizzi, A.; Celzard, A.; Abdalla, S. First tools for tannin-furanic foams design.
BioResources 2015, 10, 52335241. [CrossRef]
9. Tondi, G.; Fierro, V.; Pizzi, A.; Celzard, A. Tannin-based carbon foams. Carbon 2009, 47, 14801492. [CrossRef]
10. Meikleham, N.E.; Pizzi, A. Acid- and alkali-catalyzed tannin-based rigid foams. J. Appl. Polym. Sci. 1994, 53,
15471556. [CrossRef]
Polymers 2016, 8, 215 18 of 19

11. Sun, H.; Mark, J.E.; Tan, S.C.; Venkatasubramanian, N.; Houtz, M.D.; Arnold, F.E.; Lee, C.Y.-C. Microcellular
foams from some high-performance composites. Polymer 2005, 46, 66236632. [CrossRef]
12. Liu, T.; Liu, H.; Li, L.; Wang, X.; Lu, A.; Luo, S. Microstructure and properties of microcellular poly(phenylene
sulfide) foams by mucell injection molding. Polym.-Plast. Technol. Eng. 2013, 52, 440445. [CrossRef]
13. Tomasko, D.L.; Li, H.; Liu, D.; Han, X.; Wingert, M.J.; Lee, L.J.; Koelling, K.W. A review of CO2 applications
in the processing of polymers. Ind. Eng. Chem. Res. 2003, 42, 64316456. [CrossRef]
14. Lacoste, C.; Basso, M.C.; Pizzi, A.; Celzard, A.; Laborie, M.-P. Natural albumin/tannin cellular foams.
Ind. Crops Prod. 2015, 73, 4148. [CrossRef]
15. Li, X.; Pizzi, A.; Cangemi, M.; Navarrete, P.; Segovia, C.; Fierro, V.; Celzard, A. Insulation rigid and elastic
foams based on albumin. Ind. Crops Prod. 2012, 37, 149154. [CrossRef]
16. Zhang, L.; Ma, J. Effect of carbon nanofiber reinforcement on mechanical properties of syntactic foam.
Mater. Sci. Eng. A 2013, 574, 191196. [CrossRef]
17. Kishore, N.G.; Woldesenbet, E.; Sankaran, S. Studies on compressive failure features in syntactic foam
material. J. Mater. Sci. 2001, 36, 44854491.
18. Huang, Y.-J.; Wang, C.-H.; Huang, Y.-L.; Guo, G.; Nutt, S.R. Enhancing specific strength and stiffness of
phenolic microsphere syntactic foams through carbon fiber reinforcement. Polym. Compos. 2010, 31, 256262.
[CrossRef]
19. Burns, J.; Prime, R.; Barrall, E.; Oxsen, M. Chemistry of an epoxy-phenolic magnetic disk coating. In Polymers
in Information Storage Technology; Springer: New York, NY, USA, 1988; pp. 237256.
20. Loera, A.G.; Cara, F.; Dumon, M.; Pascault, J.P. Porous epoxy thermosets obtained by a
polymerization-induced phase separation process of a degradable thermoplastic polymer. Macromolecules
2002, 35, 62916297. [CrossRef]
21. Li, X.; Basso, M.C.; Braghiroli, F.L.; Fierro, V.; Pizzi, A.; Celzard, A. Tailoring the structure of cellular vitreous
carbon foams. Carbon 2012, 50, 20262036. [CrossRef]
22. Basso, M.C.; Giovando, S.; Pizzi, A.; Celzard, A.; Fierro, V. Tannin/furanic foams without blowing agents
and formaldehyde. Ind. Crops Prod. 2013, 49, 1722. [CrossRef]
23. Basso, M.C.; Pizzi, A.; Lacoste, C.; Delmotte, L.; Al-Marzouki, F.M.; Abdalla, S.; Celzard, A. MALDI-TOF and
13 C NMR analysis of tanninfuranicpolyurethane foams adapted for industrial continuous lines application.

Polymers 2014, 6, 29853004. [CrossRef]

24. Dogan, E.; Ksefoglu, S. Synthesis and in situ foaming of biodegradable malonic acid ESO polymers. J. Appl.
Polym. Sci. 2008, 110, 11291135. [CrossRef]
25. Stefani, P.M.; Barchi, A.T.; Sabugal, J.; Vazquez, A. Characterization of epoxy foams. J. Appl. Polym. Sci. 2003,
90, 29922996. [CrossRef]
26. Wan Hamad, W.N.F.; Teh, P.L.; Yeoh, C.K. Effect of acetic acid as catalyst on the properties of epoxy foam.
Polym.-Plast. Technol. Eng. 2013, 52, 754760. [CrossRef]
27. Najib, N.N.; Ariff, Z.M.; Bakar, A.A.; Sipaut, C.S. Correlation between the acoustic and dynamic mechanical
properties of natural rubber foam: Effect of foaming temperature. Mater. Des. 2011, 32, 505511. [CrossRef]
28. Morehouse, D.S. Expansible thermoplastic polymer particles containing volatile fluid foaming agent and
method of foaming the same. US Patent 3615972 A, 26 October 1971.
29. Jonsson, M.; Nordin, O.; Malmstrm, E. Increased onset temperature of expansion in thermally expandable
microspheres through combination of crosslinking agents. J. Appl. Polym. Sci. 2011, 121, 369375. [CrossRef]
30. Kawaguchi, Y.; Itamura, Y.; Onimura, K.; Oishi, T. Effects of the chemical structure on the heat resistance of
thermoplastic expandable microspheres. J. Appl. Polym. Sci. 2005, 96, 13061312. [CrossRef]
31. Expancel ProductsAkzoNobel ExpancelThe multi-performance additive. Available online:
https://www.akzonobel.com/expancel/products/ (accessed on 19 February 2016).
32. Tomalino, M.; Bianchini, G. Heat-expandable microspheres for car protection production. Prog. Org. Coat.
1997, 32, 1724. [CrossRef]
33. Rumeau, N.; Buisson, A.; Trouillot, P. Novel composite materials and methods for manufacturing same.
U.S. Patent 20140220841 A1, 25 August 2011.
34. Samira, E.G. Influence de la structure chimique et des proprits physiques dun systme poxyde sur
la formation et la morphologie de structures poreuses. In Proceeding of the Matriaux 2014, Montpellier,
France, 2428 November 2014.
Polymers 2016, 8, 215 19 of 19

35. Pascault, J.-P.; Sautereau, H.; Verdu, J.; Williams, R.J.J. Thermosetting Polymers; CRC Press: Boca Raton, FL,
USA, 2002.
36. Nakka, J.S.; Jansen, K.M.B.; Ernst, L.J. Tailoring the viscoelasticity of epoxy thermosets. J. Appl. Polym. Sci.
2013, 128, 37943806. [CrossRef]
37. Halary, J.L. Structureproperty relationships in epoxy-amine networks of well-controlled architecture.
High. Perform. Polym. 2000, 12, 141153. [CrossRef]
38. Nakka, J.S.; Jansen, K.M.B.; Ernst, L.J. Effect of chain flexibility in the network structure on the viscoelasticity
of epoxy thermosets. J. Polym. Res. 2011, 18, 18791888. [CrossRef]
39. Pearce, E.M. Properties of polymers, their estimation and correlation with chemical structure, 2nd ed.;
Van Krevelen, D.W., Nijenhuis, K., Eds.; Elsevier: New York, NY, USA, 1976.
40. Shieh, J.-Y.; Ho, T.-H.; Wang, C.-S. Synthesis and modification of a trifunctional epoxy resin with
amino-terminated poly(dimethylsiloxane)s for semiconductor encapsulation. Angew. Makromol. Chem.
1997, 245, 125137. [CrossRef]
41. Aronhime, M.T.; Gillham, J.K. Time-temperature-transformation (TTT) cure diagram of thermosetting
polymeric systems. Adv. Polym. Sci. 1986, 78, 83113.
42. Wang, L.; Yang, X.; Zhang, J.; Zhang, C.; He, L. The compressive properties of expandable
microspheres/epoxy foams. Compos. Part. B Eng. 2014, 56, 724732. [CrossRef]
43. Rose, N.; Le Bras, M.; Delobel, R.; Costes, B.; Henry, Y. Thermal oxidative degradation of an epoxy resin.
Polym. Degrad. Stab. 1993, 42, 307316. [CrossRef]
44. Jonsson, M.; Nordin, O.; Kron, A.L.; Malmstrm, E. Thermally expandable microspheres with excellent
expansion characteristics at high temperature. J. Appl. Polym. Sci. 2010. [CrossRef]
45. Katircioglu, T.Y.; Gven, O. Glass-transition behavior of poly(methacrylonitrile). J. Appl. Polym. Sci. 2001, 82,
19361943. [CrossRef]
46. Mark, J.E. Polymer Data Handbook; Oxford University: Oxford, UK, 1999.
47. Korobeinyk, A.V.; Whitby, R.L.D.; Mikhalovsky, S.V. High temperature oxidative resistance of
polyacrylonitrile-methylmethacrylate copolymer powder converting to a carbonized monolith. Eur. Polym. J.
2012, 48, 97104. [CrossRef]
48. Bashir, Z.; Packer, E.J.; Herbert, I.R.; Price, D.M. Base-induced gelation of polymethacrylonitrile. Polymer
1992, 33, 373378. [CrossRef]
49. Park, C.B.; Behravesh, A.H.; Venter, R.D. Low density microcellular foam processing in extrusion using CO2 .
Polym. Eng. Sci. 1998, 38, 18121823. [CrossRef]
50. Klempner, D.; Sendijarevi1 c, V.; Aseeva, R.M. Handbook of Polymeric Foams and Foam Technology; Hanser Verlag:
Munich, Germany, 2004.
51. Matuana, L.M.; Park, C.B.; Balatinecz, J.J. Processing and cell morphology relationships for microcellular
foamed PVC/wood-fiber composites. Polym. Eng. Sci. 1997, 37, 11371147. [CrossRef]
52. Ptschke, P.; Krause, B.; Stange, J.; Munstedt, H. Elongational viscosity and foaming behavior of PP modified
by electron irradiation or nanotube addition. Macromol. Symp. 2007, 254, 400408. [CrossRef]
53. Bresson, G.; Jumel, J.; Shanahan, M.E.R.; Serin, P. Statistical aspects of the mechanical behaviour a paste
adhesive. Int. J. Adhes. Adhes. 2013, 40, 7079. [CrossRef]
54. Nassiet, V.; Habas, J.P.; Hassoune-Rhabbour, B.; Baziard, Y.; Petit, J.A. Correlation between viscoelastic
behavior and cooling stresses in a cured epoxy resin system. J. Appl. Polym. Sci. 2006, 99, 679690. [CrossRef]
55. Grassie, N.; Scott, G. Polymer degradation and stabilisation. Polymer 1988, 29, 1725.

2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).