Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

Construction and Building Materials 30 (2012) 753–760

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials


journal homepage: www.elsevier.com/locate/conbuildmat

Effects of humid environment on thermal and mechanical properties


of a cold-curing structural epoxy adhesive
Mariateresa Lettieri a,⇑, Mariaenrica Frigione b
a
Institute of Archaeological Heritage, Monuments and Sites, CNR–IBAM, Provinciale Lecce-Monteroni, 73100 Lecce, Italy
b
Department of Innovation Engineering, University of Salento, Via Arnesano, 73100 Lecce, Italy

a r t i c l e i n f o a b s t r a c t

Article history: The effects of exposure to different humid environments in a commercial cold-cured epoxy adhesive
Received 2 September 2011 were investigated. Samples were exposed up to one month to 55%, 75% and 100% relative humidity
Received in revised form 15 December 2011 (RH) or immersed in liquid water, at a constant temperature (23 °C). Weight changes, thermal and
Accepted 23 December 2011
mechanical properties before and at different stages of the aging, were discussed.
Available online 17 January 2012
In the examined aging conditions, absorbed water remained below 1% and saturation level was not
achieved. Plasticization, reactivation of curing reactions and erasure of physical aging were observed in
Keywords:
the specimens subjected to the different humidity regimes, all affecting both the thermal and the
Cold-curing adhesives
Durability
mechanical properties of the aged samples: while the Tg was influenced by plasticization mainly at
Hygrometric aging shorter times of exposure and by post-curing at longer treatment times, the mechanical characteristics
Water sorption were less affected by these phenomena. These effects were found more pronounced at humidity levels
DSC analysis higher than 75% RH. Doubly hydrogen-bonded water molecules linked to the network also influenced
Flexural mechanical properties the Tg of the system, while they did not affect noticeably their flexural properties. Finally, the effects
of water exposure can be regarded as equivalent to those of a thermal treatment at temperature around
the Tg, i.e. both leading to an erasure of the physical aging.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction conditions has detrimental effects on the epoxies, leading to prop-


erty changes as a consequence of physical and/or chemical transfor-
In recent years, epoxy resins have received considerable atten- mations. Actually, epoxy adhesives exposed to a wet environment
tion and they are commonly used as adhesives, coatings and matri- absorb water because they possess polar groups attracting water
ces for composites, due to their advanced chemical and mechanical molecules. Once inside, water may alter the properties of the poly-
properties. They found employment in civil engineering for struc- mer both in a reversible manner, through plasticization, and in an
tural applications and in cultural heritage conservation, as well. In irreversible manner, if hydrolysis, cracking or crazing occur.
these fields, practical and economical reasons force to use ‘‘cold- The lowering of the Tg due to water ingress into epoxy resins
curing’’ resins, that is epoxy systems able to achieve a suitable [4–11] can be particularly harmful for cold-curing epoxies whose
degree of cure and acceptable mechanical properties at ambient typical Tg, as already underlined, is not much higher than the ser-
temperature. Aliphatic amines are usually used as curing agents vice temperature. Therefore, if the service temperature approaches
for this purpose, since they are able to react with epoxies also at the Tg of the system, a lowering of the adhesion strength and of the
low temperature. However, many weeks of cure, if compared to mechanical properties of the adhesive could occur.
the few hours curing times required by the epoxy cured with a Both the kinetics and the effects of water absorption in epoxy sys-
source of heat, are necessary to provide a material with a reason- tems have been deeply investigated by means of various analytical
ably high degree of cross-linking, even though the conversion of techniques [4,5,12–23], the most of studies in literature report the
epoxy groups is never complete, and a moderate glass transition results of immersion treatments; minor attention has been given
temperatures (Tg), which generally does not exceed 60 °C [1–3]. to the action of the atmospheric vapor of varying relative humidity.
The excellent properties of epoxy adhesives may be modified Furthermore, whereas the exposure of epoxy resins cured at ele-
by the environment which often acts as a degrading agent. Of vated temperatures has been widely studied, the sorption process
particular relevance is the ambient humidity that in actual service in epoxy systems cured under ambient conditions has been scarcely
investigated. Starting from this lack of information, an experimental
⇑ Corresponding author. Tel./fax: +39 0832 422219. study was devoted to the effects of water absorption in a cold-cured
E-mail addresses: mt.lettieri@ibam.cnr.it (M. Lettieri), mariaenrica.frigione@ epoxy adhesive either exposed to different levels of relative humid-
unisalento.it (M. Frigione). ity or immersed in actual water. In order to better understand the

0950-0618/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2011.12.077
754 M. Lettieri, M. Frigione / Construction and Building Materials 30 (2012) 753–760

independent effects of water on the studied resin, synergistic influ- 2.2. Exposure procedures
ence of moisture and temperature was avoided. To this aim, any
In order to examine the effects of water ingress on the properties of the adhe-
treatment was performed in isothermal conditions (i.e. at the tem- sive, samples of resin were exposed to controlled humidity level (55% and 75% RH),
perature of 23 ± 2 °C). In fact, a synergistic effect of the different kept in saturated water vapor atmosphere (100% RH) or immersed in distilled
environmental agents cannot be excluded, thus their combined ef- water, at a constant temperature (23 ± 2 °C).
fects can result larger than the sum of their separate effects. The samples were treated after the maximum state of cure achievable at ambi-
ent temperature was reached, that is, as previously stated, the treatments were per-
The research was aimed at investigating the behavior of the
formed on samples cured for 3.5 months at 23 °C.
adhesive in a real service condition, rather than in the saturation In Table 1, the curing procedure and the treatments performed on all the sam-
stage, which can require years of exposure to be achieved [24]. ples before the tests, were summarized.
The changes, promoted by the absorption of water, on the proper- The exposure to moisture was realized in a desiccator: the ambient at 55% RH
ties of the adhesive has been evaluated, while the kinetic aspects of was attained through the water vapor evaporated from a saturated solution of
Mg(NO3)2; the conditions of 75% RH was obtained by means of a saturated solution
the process have been discussed in a previous work [24]. The expo- of NH4Cl, while 100% RH was reached using ultrapure water (conductiv-
sures have been prolonged up to one month of treatment. This time ity = 0.06 lS/cm at 25 °C, produced using a Milli-Q (Millipore S.p.A.) water system).
of exposure was, in fact, expected to yield amounts of water uptake The immersion treatment was performed placing the specimens into liquid ultra-
very small (lower than 1%) and far from the saturation level (found pure water. Each treatment was carried out up to 28 days.
Control samples were left in a dry atmosphere (in a desiccators with silica gel, at
to be around 1.8% of absorbed water, in the previous study [24]);
15% RH) to monitor the evolution of the properties of the epoxy system not exposed
however, this limited water content can be more significant from to water.
the application point of view because closer to that measured in
true environmental conditions, as experienced in a recent study 2.3. Test methods
on weathering of cold-curing epoxy adhesives [25].
In a previous research, the same epoxy adhesive taken into ac- 2.3.1. Gravimetric measurements
count in this study was investigated after 10 and 30 thermo The weight variations measured on both the samples exposed to moisture and
those immersed in water were used to assess the amount of water absorbed during
-hygrometric cycles [3,24]. In addition, 10-day and 28-day isothermal
any treatment. In all cases, the samples were periodically removed from the storage
treatments (50 °C) were used to examine the effects of temperature in box, wiped with a dry cloth and weighed in order to calculate the percentage of ab-
isolation [3]. For comparative purposes, the same time spans, i.e. 10- sorbed water. The analytical balance, described in Section 2.1, was used.
day and 28-day, were chosen to analyze the effects of water in order to After the measurements the samples were replaced in the container as soon as
possible. Throughout any treatment performed, the environmental parameters
complete the pattern of behavior.
were controlled by the humidity–temperature logger described in Section 2.1;
The thermal and the mechanical properties measured before the reported data were calculated as the mean value of measurements performed
and at different stages of the hygrometric treatments, are dis- on five samples.
cussed. The changes in weight were also evaluated throughout
the exposure to water. 2.3.2. Calorimetric analyses
Differential scanning calorimetry (DSC) was performed on untreated samples
just before the exposure/immersion treatments. The thermal dynamic scans were
2. Materials and methods
carried out between 5 °C and 210 °C with a heating rate of 10 °C/min, under nitro-
gen atmosphere (flow rate = 80 mL/min); a DSC 822, Mettler Toledo, calorimeter
2.1. Materials and curing conditions
was used. The glass transition temperature was determined as the transition mid-
point; the relaxation enthalpy and the residual heat of cross-linking reactions were
The material object of this study is a commercial epoxy adhesive produced and
evaluated from the peak area delimited by the tangent line to DSC curve. In order to
supplied by MAPEI S.p.A. (Italy). The system is representative of epoxy resins used
prove the repeatability of results, the calorimetric experiments were repeated at
for rehabilitation procedures as adhesive and it has been already employed for res-
least three times and the results averaged.
toration of both civil and monumental buildings [26].
Epoxy resins held below their Tg are subjected to ‘‘physical aging’’, that reflect
The epoxy resin is based on diglycidylether of bisphenol-A; the curing agent is a
the tendency of the system to approach the equilibrium state [28–30]. At each tem-
mixture of aliphatic and aromatic amines, i.e. polyethylenimine – m-xylenediamine
perature below its Tg, polymers are in the glassy state and they are frozen in a non-
– nonylphenol, provided ready-to-use. Nonylphenol is usually added to hardener
equilibrium condition. However, a slow structural relaxation toward the final ther-
mixtures to increase the rate of curing due to the presence of phenolic OH group;
modynamic equilibrium state at that temperature occurs. The presence of physical
the nonyl group, as aliphatic chain, can reduce the evaporation during the applica-
aging in a polymer isothermally aged can be measured in a dynamic calorimetric
tion procedure, and have plasticizing effects, as well.
experiment. An endothermic peak in the glass-transition region, whose position
Samples of the adhesive were prepared using the mixing ratio suggested by
and intensity depends on temperature and time of aging, is observed. The area of
suppliers, that is resin:hardener = 4:1 by weight. The exact amount of each compo-
this peak is related to the relaxation enthalpy (DHrel) and it increases with the aging
nent was weighed with an analytical balance (Sartorius, Mod. BP 221S) with an
time, becoming constant when the equilibrium is achieved.
accuracy of ±0.1 mg. The hardener was poured into the base resin and they were
gently stirred, avoiding the formation of air bubbles, until the mix was perfectly
homogeneous. Specimens of standard rectangular form (dimensions: 90  10  Table 1
4 mm) were obtained by pouring the mix into Teflon moulds of apposite shape. Curing and drying conditions before the treatments and procedures of exposure to
The samples were cured at a temperature of 23 ± 2 °C and a relative humidity water.
(RH) of 50 ± 5%. They were removed from the moulds after 26 h and maintained un-
der the same controlled conditions of temperature and humidity (i.e. T = 23 ± 2 °C System Curing Drying procedure Exposure to water
and RH = 50 ± 5%) for further 9 days. In fact, the curing time reported on the product conditions
data sheet is 7 days; however, in previous studies it has been found that curing H55 50% RH and 15% RH (silica gel) and 55% RH and 23 C° up to
times higher than those suggested by the suppliers are necessary to reach the max- 23 C° for 23 C° for 3 months 28 days
imum state of cure achievable by the epoxy adhesive cured at ambient temperature. 10 days
In particular, calorimetric and spectroscopic analyses revealed that, for the system H75 50% RH and 15% RH (silica gel) and 75% RH and 23 C° up to
object of this study, 3.5 months of cure are required to reach the maximum cross- 23 C° for 23 C° for 3 months 28 days
linking degree of the resin at ambient temperature [3], that does not correspond the 10 days
complete cure of the same system, i.e. the conversion of epoxy groups was not com- H100 50% RH and 15% RH (silica gel) and 100% RH and 23 C° up to
plete. The specimens were, then, moved in a desiccator with silica gel and stored at 23 C° for 23 C° for 3 months 28 days
room temperature for further 3 months before expose them to moisture or water. 10 days
During both the preparation of the specimens and the curing stage, the environ- WL 50% RH and 15% RH (silica gel) and Immersion in liquid water
mental conditions were monitored by means of a humidity–temperature logger. A 23 C° for 23 C° for 3 months at 23 C° up to 28 days
Digitron MonoLog (Mod. MLH) was used; it can collect temperature data in a range 10 days
from 20 °C to 50 °C (with an accuracy of ± 2 °C) and relative humidity data in a Control 50% RH and 15% RH (silica gel) and –
range from 5% to 100% (with a precision of ±5%). 23 C° for 23 C° for 4,5 months
The specimens’ shape and dimensions were chosen on the basis of ASTM 10 days
standard for flexural mechanical tests [27].
M. Lettieri, M. Frigione / Construction and Building Materials 30 (2012) 753–760 755

The calorimetric tests were repeated after a 12-day treatment on the H55 sys- Table 2
tem, and 10 days of exposure on the systems H75, H100 and WL. The thermal prop- Water uptake amount and thermal properties measured on epoxy adhesive before
erties were also measured after 28 days of exposure to water vapor or immersion and during the exposure to humidity or immersion in water.
in liquid water.
All tests were performed immediately after the end of any treatment, to mini- System Treatment time [days] WAa [%] Tgb [°C] DHresc [J/g] DHreld [J/g]
mize the desorption of water from the exposed samples. H55 0 0 57.1 ± 0.3 9.2 ± 0.3 8.7 ± 0.3
12 0.12 ± 0.01 53.9 ± 0.2 6.1 ± 1.7 10.0 ± 0.3
28 0.17 ± 0.01 51.9 ± 0.2 5.8 ± 1.3 6.6 ± 0.5
2.3.3. Flexural mechanical tests
H75 0 0 57.1 ± 0.3 9.2 ± 0.3 8.7 ± 0.3
Flexural mechanical tests were carried out in three points bending, using a
10 0.15 ± 0.04 51.8 ± 1.1 6.5 ± 1.8 6.8 ± 0.4
Lloyd LR-MK4 Dynamometer (load capacity of the testing rig 1 kN, class of the ma-
28 0.27 ± 0.04 51.5 ± 1.8 5.0 ± 1.9 7.7 ± 0.3
chine 1, maximum accuracy 0.54%, maximum repeatability 0.38%), following the
H100 0 0 57.1 ± 0.3 9.2 ± 0.3 8.7 ± 0.3
ASTM D 790-10 standard [27]; the experiments were performed at room tempera-
10 0.32 ± 0.01 52.8 ± 1.1 5.2 ± 0.8 3.9 ± 0.7
ture, with a support span of 60 ± 0.1 mm and a crosshead rate of 2 mm/min; flex-
28 0.59 ± 0.03 49.4 ± 1.6 3.9 ± 0.1 0.9 ± 0.6
ural mechanical properties (i.e. flexural modulus and maximum flexural strength)
WL 0 0 57.1 ± 0.3 9.2 ± 0.3 8.7 ± 0.3
were calculated as the average of five measurements. The mechanical tests were
10 0.24 ± 0.02 48.7 ± 2.5 5.1 ± 0.3 4.7 ± 1.6
performed at start of the exposure/immersion procedures and they were repeated
28 0.49 ± 0.03 51.7 ± 2.4 4.0 ± 0.1 3.5 ± 1.7
after the same treatment times used to monitor the thermal properties: 12-day
and 28-day treatments for the H55 system, and 10-day and 28-day treatments for a
Absorbed water amount.
the systems H75, H100 and WL. b
Glass transition temperature.
Also in this case, each experiment on aged samples was carried out immediately c
Residual heat of reaction.
at the end of the treatment. d
Relaxation enthalpy.

3. Results and discussion


Table 3
3.1. Un-exposed epoxy system Flexural mechanical properties measured in three point bending on epoxy adhesive
before and during the exposure to humidity or immersion in water.

The thermal and mechanical properties of the specimens iso- System Treatment time [days] Eflexa [GPa] rmaxb [MPa]
thermally cured for 3.5 months in a dry environment, were first H55 0 3.45 ± 0.66 93.4 ± 12.0
measured. 12 3.17 ± 0.17 97.8 ± 6.5
The un-exposed samples showed a Tg of 57.1 ± 0.4 °C; a relaxa- 28 3.16 ± 0.34 97.9 ± 15.7
H75 0 3.45 ± 0.66 93.4 ± 12.0
tion peak was noticed in the DSC thermogram and the correspond-
10 3.72 ± 0.20 92.1 ± 16.8
ing enthalpy was calculated to be 8.7 ± 0.3 J/g. A residual heat of 28 3.27 ± 0.21 105.7 ± 13.6
reaction was also observed and its average value was equal to H100 0 3.45 ± 0.66 93.4 ± 12.0
9.2 ± 0.3 J/g. A measurable residual heat of reaction indicates that, 10 3.59 ± 0.15 105.6 ± 15.7
even after long time of cure, that is 3.5 months, the cross-linking 28 3.00 ± 0.28 93.9 ± 11.0
WL 0 3.45 ± 0.66 93.4 ± 12.0
process carried out at ambient temperature did not reach the
10 3.75 ± 0.30 107.2 ± 13.4
completion. 28 2.94 ± 0.44 91.4 ± 13.2
The results of mechanical tests performed on the un-exposed a
Flexural modulus of elasticity.
samples are: a flexural modulus of elasticity of 3.45 ± 0.66 GPa, b
Maximum flexural strength.
an average maximum flexural strength of 93.4 ± 12.0 MPa.
The thermal and mechanical properties of the un-exposed sam-
ples are reported in Table 2 and Table 3, respectively, with the indi-
cation of the maximum deviation.
Both the calorimetric and mechanical tests were also performed
on the adhesive samples cured for 5 months and stored at 23 °C in
a dry atmosphere (i.e. 15% RH).

3.2. Weight variations

The percentage of absorbed water (WA) in any sample was cal-


culated from the equation:

WA ¼ ½ðM t  M 0 Þ=M 0   100

where Mt and M0 are the weights of the specimen after and before
the treatment, respectively.
In Fig. 1, the values of WA for the four treated epoxy systems as
a function of the exposure time, up to 28 days, are reported.
The sorption rate was very high during the first days of treat-
ment and the process resulted faster as the humidity level in- Fig. 1. Water absorption percentage (WA) as a function of the exposure/immersion
creased. Other authors found a similar trend [15]. The higher time.
sorption rate observed during the first day of treatment could ac-
count for stronger interactions between water and the polymer
network. In the dry resin, the hydrophilic sites are unoccupied and they
The research dealing with the sorption mechanism has sug- easily link the water, involving singly or doubly hydrogen-bonded
gested that the water molecules into the polymer have a double molecules [17,32], that is water molecules forming one or two
nature: unbound molecules confined into the free volume of the hydrogen bonds, respectively. In the absorption process into epoxy
system and molecules bound to hydrophilic sites along the poly- networks, water molecules firstly bind with the hydrophilic
mer [6,12,16,31–33]. groups, then diffuse into free volume elements as unbound water
756 M. Lettieri, M. Frigione / Construction and Building Materials 30 (2012) 753–760

[13,20,34]. Therefore, the hydrogen-bonds, which are energetic


favorable, appears earlier in the water absorption process and are
produced prior to the free water molecules [35]. As the water con-
centration increases, the number of polar groups on the network
available for hydrogen bonding remains unvaried, but they result
saturated by water and additional water molecules prefer to form
hydrogen bonds with other water molecules whereas with the po-
lar groups in the polymer [33,36,37].
A slower water uptake was measured for few days of exposure
suggesting that the absorption was retarded since several polar
sites were previously involved in the aforementioned hydrogen
bonds. At this stage, the entering water molecules have a higher
possibility to remain inside the epoxy network as unbound water.
The investigated exposure time (i.e. 28 days) was not sufficient
to achieve the saturation level in any of the examined conditions.
In an earlier study performed on the same epoxy adhesive cured
at ambient temperature and subjected to either immersion or
exposure to 100% RH, in fact, 60 weeks were necessary to reach Fig. 2. Water absorption percentage (WA) as a function of the relative humidity
the saturation condition [24]. level.
The highest amount of absorbed water was measured for the
system H100, that is the system exposed to 100% RH. The increase
in weight over the time of the system WL, immersed in liquid same table, the amounts of water absorbed by the specimens after
water, was comparable to that of samples H100, even if the im- different exposure to moisture or immersion times, are also
mersed samples showed lower amounts of absorbed water reported.
throughout the whole treatment time. As expected, the exposure to moisture, as well as the immersion
At 100% RH the chemical potential of the water vapor is the treatment, decreased the Tg of the epoxy resin, since the absorption
same as that of the liquid, hence sorption, diffusion and equilib- of water molecules gives rise to plasticizing effects.
rium moisture uptake should be the same as in the liquid phase Whereas the unbound water molecules are characterized by a
[34,38]. However, differences between water content in specimens higher mobility, due to the low interaction with the polymer net-
exposed to environment with 100% RH and those immersed have work, the bound molecules are immobilized, since an high energy
been already observed and attributed to the formation of micro- barrier have to be overcome [15,40] and are expected to be respon-
cracks or to reactions occurring between the liquid water and the sible for the plasticization [16,32,41,42]. These effects are mainly
polymer [34], phenomena that are reported to be more pro- attributed to the water forming single hydrogen bonds; on the
nounced in systems immersed into water. The initial difference other hand, doubly hydrogen-bonded molecules do not signifi-
in the sorption curves of immersed and 100% RH exposed speci- cantly contribute to plasticization, whereas produce opposite ef-
mens was found to be erased after about 12 weeks of aging [24]. fects as a consequence of pseudo-crosslinking resulting from the
In addition, both aging treatments led to very close diffusion coef- secondary bridging they form between the chain segments [9,17].
ficients (1.19  109 for immersed specimens and 1.98  109 for A general reduction in Tg was observed after each 10-day treat-
samples exposed to 100% RH) and water content profiles [24]. ment: H55, H75 and H100 exhibited very similar Tg, being the differ-
Similar water sorption profiles were observed for samples H55 ences within the experimental error; on the other hand, the Tg
and H75, in particular during the early stage of the absorption pro- value measured for WL was even slightly lower.
cess. Slightly different amounts of absorbed water were measured Different trends were observed as the treatment time increased.
in the same specimens just after several days of exposure (i.e. At 28 days of exposure, only the systems H55 and H100 showed an
20 days), coherent with the different levels of humidity exposure. additional decrease in Tg with respect to the 10-day treatment;
When the relative humidity is lower than 100%, fewer water similar values of glass transition temperature were measured for
molecules are in the air than there would be over a pure water sur- H75 samples treated for either 10 or 28 days. An appreciable in-
face. Therefore, progressively less marked effects in terms of water crease in Tg was found only in immersed specimens. The latter re-
absorbed are expected after exposure to 75% and 55% RH. In fact, in sult could be related to a post-curing process promoted by the
our case, the lower the humidity level, the lower the amount of ab- decrease in Tg upon moisture ingress in the partially cured resin.
sorbed water. In order to illustrate the different behavior of the It is reported that the lower Tg allows the polymer chains to be-
specimens at high humidity levels, in Fig. 2 the absorbed water come mobile, increasing the probability of a contact between the
amounts vs. the relative humidity of treatment, are reported. A lin- groups that have not reacted, thus promoting additional cross-link-
ear trend can be observed for exposure environments up to 75% ing reactions and making possible the completion of the curing
RH, while larger increases in absorbed water were measured as process [43–45]. In that case, an increase in Tg is expected.
the humidity level increased. An additional 25% increase in humid- The changes in thermal properties, as a consequence of water
ity level of exposure, up to 100% RH, produced, in fact, a doubled ingress, are manifest in DSC traces drawn in Fig. 3. In this figure
water absorption. Other authors have experienced a threshold in the glass transition region, the endothermic peak related to the
the water uptake process at relative humidity higher than 80% physical aging and the exothermic peak due to the residual heat
[34,39]. of reaction, are evidenced.
Finally, no change in weight was measured on control samples In Fig. 4 the Tg is reported as a function of the percentage of
kept in dry atmosphere. water uptake for the specimens exposed to different levels of
humidity or immersed in water. Within the early stage of the
3.3. Thermal properties absorption process, even low amounts of water gain were able to
produce an appreciable reduction in Tg. Closer Tg values were mea-
The thermal properties of the epoxy adhesive under study, sub- sured at larger weight gain. The behavior of Tg measured on speci-
jected to the hygrometric treatments, are listed in Table 2. In the mens immersed is somehow different from that of samples exposed
M. Lettieri, M. Frigione / Construction and Building Materials 30 (2012) 753–760 757

dependence on the humidity level of treatment was noticed: the


higher the percentage of relative humidity of exposure, the lower
the residual heat of reaction. The values found for H100 and WL coin-
cide in the whole range of observation. The same values of DHres
measured for H100 and WL would suggest comparable degrees of
cross-linking for these two systems. This result indicates that the
increase in Tg observed for the immersed samples between 10
and 28 days of exposure cannot be simply related to the post-curing
process. It can be hypothesized that the resumption of curing reac-
tions along with pseudo cross-linking brought about by the water
molecules forming bridging between chain segments, produce less
mobile polymer chains due to constrains into the structure.
Despite the fact that a reduction of the residual heat of reaction
was observed throughout each treatment, the Tg values were al-
ways lower than those measured on the untreated resin. This result
suggests that the plasticization undoubtedly prevailed, while the
post-cure produced minor effects on the thermal properties of this
epoxy adhesive.
In Fig. 5, the residual heat of reaction as a function of the water
Fig. 3. DSC traces of the 28-day aged samples.
gain, is reported. The DHres decreased as the amount of absorbed
water increased. As previously stated, the more mobile chains,
due to the platicization effect, allow to post-cure and increase
the cross-linking density of the systems. The ingress of few water
molecules during the first stage of the absorption process were
able to reduce roughly by half the residual heat of reaction. Then,
little additional decreases of DHres were measured even at much
greater amounts of sorbed water. In the latter case, a less mobile
polymer structure can be suggested once again: the bridging
formed between the water molecules and the polar sites in the re-
sin slow the motion of the polymer chains thus reducing the prob-
ability of a contact between the un-reacted groups. The total
disappearance of the peak due to residual curing reactions was
not experienced, proving that un-reacted groups are still present
in the treated resin samples.
By comparing the fitting curves for Tg and DHres vs. amount of
absorbed water reported in Figs. 4 and 5, respectively, a very sim-
ilar trend is observed. This to confirm that the calorimetric proper-
ties of cold-curing epoxy adhesive under study are significantly
affected by small quantities of sorbed water, while greater
Fig. 4. Glass transition temperature (Tg) as a function of absorbed water amount amounts do not appreciably influence these properties, at least
(WA).
for the investigated range of exposures.
The residual heat of reaction (DHres) on control samples, kept
to different levels of RH, in particular for low amounts of absorbed for 4.5 months at 15% RH, was still measured, being around
water. In the same Fig. 4, the curve used to fit the experimental Tg 7.7 ± 2.4 J/g. This value is comparable with that found before the
data of specimens exposed to different levels of humidity is also treatments, since the difference is within the experimental error.
shown. It is confirmed that the curve is able to fit also the Tg value
of the immersed specimens only at the longest immersion time.
The reductions in Tg observed at low amounts of sorbed water
can be ascribed to the prevailing of plasticization effects on the
post-curing. At low amount of sorbed water, in fact, strong
water–resin interactions are more likely to occur: in this case,
bound molecules, which are responsible of Tg decrease, prevail
[13,33,36,46]. On the other hand, as the water content increases,
the occurrence of a more rigid molecular structure, caused by both
the post-curing process and doubly hydrogen-bonded water mole-
cules, is able to counteract the decrease in Tg due to the plasticiza-
tion. Absorbed water molecules forming two hydrogen bonds
cause, in fact, an increase in Tg [9,35].
The Tg of the control samples, kept for 4.5 months at 15% RH
(equivalent to a 28-day treatment), remained unchanged (Tg =
56.9 ± 0.4 °C) with respect to the same parameter measured after
3.5 months of cure (i.e. before the treatments). Therefore the varia-
tions in Tg measured on the treated epoxy resin can be ascribed just
to the water action.
In examining the residual heat of reaction, it is worth noting that
all samples showed a reduction in DHres (Table 2 and Fig. 3). A Fig. 5. Residual heat of reaction (DHres) vs. absorbed water amount (WA).
758 M. Lettieri, M. Frigione / Construction and Building Materials 30 (2012) 753–760

In thermosetting resins, the increased segmental mobility, in-


duced by water ingress, can be the consequence of the motions
of both segments of unreacted molecules, which can be still pres-
ent in the cured system, and chain segment in cross-linked regions
of network [28]; hence, the erasure of physical aging can be also
observed [10,37,47] not only by heating the system above its Tg
[29,48,49], but also by the action of other factors, including tem-
perature, moisture and stress [50–52].
The relaxation enthalpy (DHrel), as a measure of the physical
aging phenomenon, is reported in Table 2 at different exposure/
immersion times. The immersed samples and the specimens ex-
posed to the highest humidity level exhibited similar trend of DHrel
vs. immersion/exposure time, that is a continuous decrease. How-
ever, the lowest relaxation enthalpy was measured on the systems
H100.
In the systems exposed to 75% and 55% RH, the exposure to
moisture produced no appreciable variations in DHrel. From data
in Table 2, the small differences in DHrel can be, in fact, attributed
to inevitable inaccuracy in the calculation of the small areas repre-
sentative of the relaxation enthalpy released during DSC scans, as Fig. 7. Relaxation enthalpy (DHrel), in logarithmic scale, vs. residual heat of reaction
already reported elsewhere [53]. (DHres).
In Fig. 6 the DHrel as a function of the absorbed water amount, is
reported. The trend to decrease can be observed, confirming the
partial erasure of physical aging occurring as a consequence of as consequence of water ingress, is likely to be measurable only
water ingress. for immersed and 100% RH exposed specimens, as discussed next.
The relaxation enthalpy (DHrel) in function of the residual heat
of reaction (DHres) is reported in Fig. 7. Increasing the immersion/ 3.4. Mechanical properties
exposure time, both DHrel and DHres decreased. In fact, water in-
gress promoted the resumption of curing reactions (producing a In literature, studies dealing with the effect of the immersion in
DHres decrease) and made the system less ‘‘dense’’ in terms of water on the mechanical properties of epoxy resins in bulk are re-
physical aging (as a consequence DHrel decrease). The data are ported. This treatment yields an initial increase in ultimate tensile
not on a straight line, but they exhibit a similar trend, irrespective strength, followed, at longer immersion times, by a decrease to val-
to the exposure regime used. ues similar to that of the unaged polymer [42,43]. The initial in-
The value of relaxation enthalpy (i.e. 9.4 ± 0.2 J/g) of the control crease in the ultimate tensile strength was explained in terms of
samples (kept for 4.5 months at 15% RH) is close to that measured an increase in cross-link density [42,54]. Later reductions in
on un-exposed specimens. strength, finally, were the result of degradation due to the presence
In conclusion, the variations in relaxation enthalpy as conse- of water. From the same studies, the Young’s modulus of the aged
quence of de-aging process are almost insignificant for the speci- epoxy resulted slightly lower than that of the control samples,
mens exposed to intermediate levels of humidity, while they increasing the reduction in modulus by increasing the immersion
become appreciable starting from the exposure to 100% RH The re- time [39,55,56]. The physical aging gives rise to continuous
sults observed for H100 and WL specimens testify the equivalence of changes of the material properties, with reduction in creep compli-
a low amount of water ingress and a moderate temperature, not ance, stiffening, reduction of ultimate elongation, increase in yield
exceeding the Tg of the resin, in the de-aging process of these res- strength [57–59].
ins. In a previous research, similar results were found, confirming The mechanical properties of the epoxy adhesive under study
that, during a hygrometric treatment, a de-aging process took place are listed in Table 3.
[3]. The influence on mechanical properties of de-aging, achieved The flexural modulus of elasticity (Eflex) was calculated as
follow:
3
Eflex ¼ ðL3 mÞ=4bd Þ
where L is the support span in mm, m is the slope of the initial
straight-line portion of the load–deflection curve in N/mm, b is
the width in mm of the tested specimen, and d is the sample thick-
ness in mm.
Comparing first the Eflex data as a function of exposure/immer-
sion time, the systems H75, H100 and WL exhibited a slight increase
in flexural modulus at shorter times, up to 9%. However, the differ-
ences in Eflex calculated on treated and untreated samples fell with-
in the range of the experimental error. A subsequent appreciable
decrease of this parameter was verified at longer exposure/immer-
sion times, the highest the humidity level the greater the reduction
in stiffness. The behavior observed after few days of exposure
could be ascribed to a more marked influence on the flexural mod-
ulus of the effects due to the reactivation of the curing reactions,
already witnessed by the decreased residual heat of reaction mea-
sured through the DSC analyses. Similar results were found by
Fig. 6. Relaxation enthalpy (DHrel) as a function of absorbed water amount (WA). other researchers in mechanical tensile tests: the exposure to
M. Lettieri, M. Frigione / Construction and Building Materials 30 (2012) 753–760 759

water yielded an initial slight increase in the elastic modulus, fol-


lowed by more marked plasticization effects (that is, elastic mod-
ulus decrease) at larger water uptake [42]. The samples exposed
to 55% RH exhibited values of flexural modulus Eflex slightly lower
than those calculated on the untreated system, irrespective to the
exposure time. It must be noted, however, that the std. deviation in
mechanical data, reported in Table 3, are in the order of 20%, but
only for un-exposed specimens.
In Fig. 8 the flexural modulus as a function of water absorbed
for any humidity regime used, is reported.
At lower water contents, i.e. up to 0.32%, Eflex fluctuated around
a mean value (3.46 GPa) very close to that calculated before the
treatments (3.45 GPa). The decrease of Eflex expected as a conse-
quence of the plasticization was most likely balanced by the in-
crease due to the post-curing process, resulting in not dissimilar
values of flexural modulus. At higher water uptake, a decrease up
to 15% was observed with respect to Eflex calculated on un-exposed
material. In this case, the cross-linking reactions were almost com-
pleted (minor changes were detected in DHres), then the plasticiz-
ing effect prevailed. Fig. 9. Stress–strain curves of the 28-day aged samples.
The maximum flexural strength (rmax) was calculated using the
following equation:
as a consequence of de-aging process. At short immersion/expo-
rmax ¼ ð3PLÞ=ð2bd2 Þ sure times its effects could have been counteracted by those pro-
duced by the reactivation of cross-linking reactions while they
where P is the load (in N) at a given point on the load–deflection are likely to result more marked at higher amounts of water up-
plot, L is the support span in mm, b is the width of the tested spec- take, where the lowest values of DHrel along with the greatest
imen expressed in mm, and d is the sample thickness in mm. reductions in elastic modulus were measured.
Average stress–strain curves found for the resin exposed to the The flexural modulus of the control samples left for 4.5 months
different regimes of moisture or immersed in water for 28 days, are at low relative humidity (at 15% RH) was measured to be
shown in Fig. 9. A linear relationship was observed in all cases, irre- 3.48 ± 0.67 GPa, the maximum flexural strength 95.0 ± 19.9 MPa.
spective to the treatment performed. The yield point was never ob- These results confirmed that steady values of the properties of
served and the maximum flexural strength coincided with the the epoxy resin were already attained before the exposure to water.
stress at break. A similar trend was found by other authors [42].
The values of rmax for the systems analyzed, reported in Table 3,
reveal a certain increase in strength at the lowest exposure/immer- 4. Conclusion
sion time, again ascribed to a post-curing process, even though
high values of maximum deviations in results can be noticed. At In the present study, the effect of the immersion of liquid water
the longest time, no difference is measured as a consequence of or the exposure to different levels of relative humidity on the ther-
the different regimes of exposure. It can be concluded that the ef- mal and mechanical properties of a commercial cold-cured epoxy
fect of short term (up to one month) exposure to water had an adhesive was analyzed and discussed. The observed changes were
insignificant effect on the strength of specimens. undoubtedly due to water action since the aged specimens were
As previously stated, mechanical properties can be also influ- exposed after a curing time suitable to stop, not necessarily to
enced by the erasure of physical aging especially in the epoxy complete, the cure at ambient temperature. In fact, no significant
adhesives exposed to the highest levels of water. A reduction in variations in properties of the control samples left for 4.5 months
both the elastic modulus and the maximum strength is expected at 15% RH were measured.
All the examined aging conditions were not able to give rise to
the saturation level in the investigated exposure time (i.e.
4 weeks). However, a short period of exposure was chosen in order
to be more similar to that experienced in true service conditions.
The amounts of absorbed water depended on humidity level of
exposure. Generally, the higher the relative humidity, the higher
the water uptake; however, larger increases in absorbed water
were measured above 75% RH.
The water ingress led to the plasticization of the adhesive, en-
hanced the reactivation of cross-linking reactions, as well as the
erasure of physical aging. In the latter case, the consequence of
water exposure was equivalent to the effects of a thermal treat-
ment at temperature around the Tg. This process was appreciable
again at regimes of humidity higher than 75%.
Plasticization influenced the Tg particularly at shorter times of
exposure; at longer treatment times, the same parameter was
more influenced by the effects of both the reactivation of curing
reactions and the doubly hydrogen-bonded water molecules.
Fig. 8. Flexural modulus of elasticity (Eflex) as a function of the absorbed water Regarding the mechanical properties, at lower amounts of
amount (WA). water absorption, the action of the post-cure balanced the effects
760 M. Lettieri, M. Frigione / Construction and Building Materials 30 (2012) 753–760

of the plasticization and, possibly, of de-aging. As a consequence, [27] ASTM D790. Standard test methods for flexural properties of unreinforced and
reinforced plastics and electrical insulating materials. West Conshohocken
the mechanical properties appear to be scarcely influenced by
(PA): ASTM International; 2010.
the exposure to humidity/immersion in water. However, at higher [28] Fraga F, Castro-Díaz C, Rodríguez-Nuñez E, Martínez-Ageitos JM. Physical
water uptake values, i.e. in correspondence of humidity levels aging for an epoxy network diglycidyl ether of bisphenol A/m-
higher than 75%, the effects of the de-aging process coupled with xylylenediamine. Polymer 2003;44:5779–84.
[29] Struik CE. Physical aging in amorphous polymers and other materials. New
those of the plasticization led to a certainly reduction in the stiff- York: Elsevier Scientific Publishing Company; 1978.
ness. A linear relationship in stress and strain curves was observed [30] Colombini D, Martinez-Vega JJ, Merle G. Dynamic mechanical investigations of
in all cases, irrespective to the treatment performed. the effects of water sorption and physical ageing on an epoxy resin system.
Polymer 2002;43:4479–85.
[31] Xiao GZ, Delamar M, Shanahan MER. Irreversible interactions between water
References and DGEBA/DDA epoxy resin during hygrothermal aging. J Appl Polym Sci
1997;65:449–58.
[1] Frigione M, Naddeo C, Acierno D. Cold-curing epoxy resins: aging and [32] Musto P, Ragosta G, Mascia L. Vibrational spectroscopy evidence for the
environmental effects I – Thermal Properties. J Polym Eng 2001;21:23–51. dual nature of water sorbed into epoxy resins. Chem Mater 2000;12:
[2] Frigione M, Aiello MA, Naddeo C. Water effects on the bond strength of 1331–41.
concrete/concrete adhesive joints. Constr Build Mater 2006;20:957–70. [33] Wu C, Xu W. Atomistic simulation study of absorbed water influence on
[3] Frigione M, Lettieri M, Mecchi AM. Environmental effects on epoxy adhesives structure and properties of crosslinked epoxy resin. Polymer 2007;48:5440–8.
employed for restoration of historical buildings. J Mater Civil Eng [34] Maggana C, Pissis P. Water sorption and diffusion studies in an epoxy resin
2006;18:715–22. system. J Polym Sci Pol Phys 1999;37:1165–82.
[4] Fernández-García M, Chiang MYM. Effect of hygrothermal aging history on [35] Li L, Zhang SY, Chen YH, Liu MJ, Luo DingYF, XW PuZ, et al. Water
sorption process, swelling, and glass transition temperature in a particle-filled transportation in epoxy resins. Chem Mater 2005;17:839–45.
epoxy-based adhesive. J Appl Polym Sci 2002;84:1581–91. [36] Wu P, Siesler HW. Water diffusion into epoxy resin: a 2D correlation ATR-FTIR
[5] Xiao GZ, Shanahan MER. Irreversible effects of hygrothermal aging on DGEBA/ investigation. Chem Phys Lett 2003;374:74–8.
DDA epoxy resin. J Appl Pol Sci 1998;69:363–9. [37] Soles CL, Chang FT, Gidley DW, Yee AF. Contributions of the nanovoid structure
[6] Musto P, Mascia L, Ragosta G, Scarinzi G, Villano P. The trasport of water in to the kinetics of moisture transport in epoxy resins. J Polym Sci Pol Phys
tetrafunctional epoxy resin by near-infrared Fourier transform spectroscopy. 2000;38:776–91.
Polymer 2000;41:565–74. [38] Sung NH. Moisture effects on adhesive joints. In engineered materials
[7] Prolongo SG, del Rosario G, Urena A. Comparative study on the adhesive handbook: adhesives and sealants, vol. 3, Materials Park (OH,USA): ASM
properties of different epoxy resins. Int J Adhes Adhes 2006;26:125–32. International Publication; 1990. p. 622–7.
[8] Netravali AN, Fornes RE, Gilbert RD, Memory JD. Effects of water sorption at [39] Loh WK, Crocombe AD, Abdel Wahaba MM, Ashcroft IA. Modelling anomalous
different temperatures on permanent changes in an epoxy. J Appl Polym Sci moisture uptake, swelling and thermal characteristics of a rubber toughened
1985;30:1573–8. epoxy adhesive. Int J Adhes Adhes 2005;25:1–12.
[9] Zhou J, Lucas JP. Hygrothermal effects of epoxy resin. Part II: Variations of glass [40] Diamant Y, Marom G, Broutman LJ. The effect of network structure on
transition temperature. Polymer 1999;40:5513–22. moisture absorption of epoxy resins. J Appl Polym Sci 1981;26:3015–25.
[10] Startsev OV, Krotov AS, Perov BV, Vapirov YM. Interaction of water with [41] Weir MD, Bastide C, Sung SP. Characterization of interaction of water in epoxy
polymers under their climatic ageing. In: Milan AIM, editor. Proc of the 4th by UV reflection spectroscopy. Macromolecules 2001;34:4923–6.
European conference of advanced materials and processes—EUROMAT 95. [42] Nogueira P, Ramìrez C, Torres A, Abad MJ, Cano J, Lòpez J, et al. Effect of water
1995. p. 245–254. sorption on the structure and mechanical properties of an epoxy resin system.
[11] Ameli A, Papini M, Spelt JK. Hygrothermal degradation of two rubber- J Appl Pol Sci 2001;80:71–80.
toughened epoxy adhesives: application of open-faced fracture tests. Int J [43] Kajomcheappunngam S, Gupta RK, GangaRao VS. Effect of aging
Adhes Adhes 2011;31:9–19. environment on degradation of glass-reinforced epoxy. J Compos Constr
[12] Mensitieri G, Lavorgna M, Musto P, Ragosta G. Water transport in densely 2002;6:61–8.
crosslinked networks: a comparison between epoxy systems having different [44] Pethrick RA, Hollins EA, McEwan I, MacKinnon AJ, Hayward D, Cannon LA.
interactive characters. Polymer 2006;47:8326–36. Dielectric, mechanical and structural, and water absorption properties of a
[13] Liu M, Wu P, Ding Y, Chen G, Li S. Two-dimensional (2D) ATRFTIR thermoplastic-modified epoxy resin: poly(ether sulfone)amine cured epoxy
spectroscopic study on water diffusion in cured epoxy resins. resin. Macromolecules 1996;29:5208–14.
Macromolecules 2002;35:5500–7. [45] Perrin FX, Nguyen MH, Vernet JL. Water transport in epoxy–aliphatic amine
[14] Ngono Y, Maréchal Y, Mermilliod N. Epoxy-amine reticulates observed by networks – influence of curing cycles. Eur Polym J 2009;45:1524–34.
infrared spectrometry. I: hydration process and interaction configurations of [46] Cotugno S, Larobina D, Mensitieri G, Musto P, Ragosta G. A novel spectroscopic
embedded H2O molecules. J Phys Chem B 1999;103:4979–85. approach to investigate transport processes in polymers: the case of water–
[15] Cotugno S, Mensitieri G, Musto P, Sanguigno L. Molecular interactions in and epoxy system. Polymer 2001;42:6431–8.
transport properties of densely cross-linked networks: a time-resolved FT-IR [47] Bao LR, Yee AF, Lee CYC. Moisture absorption and hygrothermal aging in a
spectroscopy investigation of the epoxy/H2O system. Macromolecules bismaleimide resin. Polymer 2001;42:7327–33.
2005;38:801–11. [48] Nichols ME, Wang SS, Geil PH. Creep and physical aging in polyamideimide
[16] Musto P, Ragosta G, Scarinzi G, Mascia L. Probing the molecular interactions in carbon fiber composite. J Macromol Sci Phys 1990;29:303–36.
the diffusion of water through epoxy and epoxy-bismaleimide networks. J [49] Nuñez L, Fraga L, Nuñez MR, Villanueva M, Rial B. TTT cure diagram epoxy
Polym Sci Pol Phys 2002;40:922–38. system diglycidyl ether of bisphenol A and m-xylylenediamine. J Therm Anal
[17] Zhou J, Lucas JP. Hygrothermal effects of epoxy resin. Part I: The nature of Calorim 2002;70:9–17.
water in epoxy. Polymer 1999;40:5505–12. [50] Hu H, Sun CT. The equivalence of moisture and temperature in physical aging
[18] Barral L, Cano J, López J, Nogueira P, Ramirez C, Abad MJ. Water sorption in of polymeric composites. J Compos Mater 2003;37:913–28.
tetrafunctional phenol novalac epoxy mixtures cured with diamine. J Therm [51] Zheng Y, Priestley RD, McKenna GB. Physical aging of an epoxy subsequent to
Anal 1998;52:823–30. relative humidity jumps through the glass concentration. J Polym Sci Pol Phys
[19] Ivanova KI, Pethrick RA, Affrossman S. Investigation of hydrothermal ageing of 2004;42:2107–21.
a filled rubber toughened epoxy resin using dynamic mechanical thermal [52] Lee JK, Hwang JY, Gillham JK. Erasure below glass-transition temperature of
analysis and dielectric spectroscopy. Polymer 2000;41:6787–96. effect of isothermal physical aging in fully cured epoxy/amine thermosetting
[20] Grave C, McEwan I, Pethrick RA. Influence of stoichiometric ratio on water system. J Appl Polym Sci 2001;81:396–404.
absorption in epoxy resins. J Appl Polym Sci 1998;69:2369–76. [53] Sciolti MS, Frigione M, Aiello MA. Wet lay-up manufactured FRPs for concrete
[21] Chin JW, Nguyen T, Aouadi K. Sorption and diffusion of water, salt water, and and masonry repair: influence of water on the properties of composites and of
concrete pore solution in composite matrices. J Appl Polym Sci their epoxy components. J Compos Constr 2010;14:823–33.
1999;71:483–92. [54] Gupta VB, Drzal LT, Lee CYC, Rich MJ. The temperature-dependence of some
[22] Ahmad Z, Ansell MP, Smedley D. Effect of nano and micro-particle additions on mechanical properties of a cured epoxy resin system. Pol Eng Sci
moisture absorption in thixotropic room temperature cure epoxy-based 1985;25:812–23.
adhesives for bonded-in timber connections. Int J Adhes Adhes [55] Zanni-Deffarges MP, Shanahan MER. Diffusion of water into an epoxy
2010;30:448–55. adhesive: comparison between bulk behaviour and adhesive joints. Int J
[23] Leger R, Roy A, Grandidier JC. Non-classical water diffusion in an industrial Adhes Adhes 1995;15:137–42.
adhesive. Int J Adhes Adhes 2010;30:744–53. [56] Lapique F, Redford K. Curing effects on viscosity and mechanical properties of a
[24] Frigione M, Lettieri M. Procedures conditioning the absorption/desorption commercial epoxy resin adhesive. Int J Adhes Adhes 2002;22:337–46.
behavior of cold-cured epoxy resins. J Polym Sci Pol Phys 2008;46:1320–36. [57] Kong ESW. Sub-Tg annealing studies of advanced epoxy-matrix graphite-fiber-
[25] Lettieri M, Frigione M. Natural and artificial weathering effects on cold-cured reinforced composites. J Appl Phys 1981;52:5921–5.
epoxy resins. J Appl Polym Sci 2011;119:1635–45. [58] Morgan RJ. The effect of thermal history and strain rate on the mechanical
[26] Balsamo A, Battista U, Herzalla A, Viskovic A. The use of aramidic fibers to properties of diethylenetriamine-cured bisphenol-A-diglycidyl ether epoxies. J
improve the structural behaviour of masonry structures under seismic actions. Appl Polym Sci 1979;23:2711–7.
In: Proc int congress archi 2000, UNESCO-ICOMOS; 2001. 6p. <http:// [59] Cook WD, Mehrabi M, Edward GH. Ageing and yielding in model epoxy
www.unesco.org/archi2000/pdf/balsamo.pdf>. thermosets. Polymer 1999;40:1209–18.

You might also like