Introduction

Epoxies are important thermosetting polymers widely used for a variety of purposes such as civil engineering applications, airframe materials and cultural heritage conservation [1,2,3,4]. Epoxies exhibit low shrinkage during curing; ease of cure and processing; chemical, solvent, impact and moisture resistance; and excellent adhesion to many substrates. However, they typically have low thermal stability [5,6,7]. Epoxy resins are crosslinked with aliphatic and cyclo-aliphatic amines as curing agents and are used extensively in protective coating applications [8]. Our interest in these resins arises from their application in timber composite materials which are increasingly being used in outdoor structural applications. The evaluation of durability is important to find optimum ways to enhance the resistance of epoxy adhesive system against environmental conditions [9]. In outdoor applications especially, degradation of epoxy resins can not only detract from the appearance of the products but lead to decreased performance and then failure [10]. Moisture, rain, temperature extremes, ultraviolet irradiation (UV) and pollutants are the main environmental factors that leading to irreversible material degradation, primarily by facilitating oxidation reactions [11,12,13,14,15]. Aromatic epoxies are more sensitive to incident UV irradiation than aliphatic epoxies because of the aromatic chromophore [16,17,18,19]. Moisture from humid environments can degrade the mechanical and chemical properties of epoxy resins, which can absorb up to 7% of their weight in water [20]. This swelling can cause load yielding and fracture of the resin, as well as increasing access to oxygen which causes oxidation reactions. Exposure of epoxy resin to moisture can both cause extreme brittleness (by chain crosslinking) and loss of strength (by chain scission) reducing the flexural modulus of elasticity and tensile strength [21, 22]. High temperature is another factor that can accelerate the degradation of epoxy resins. Increased temperature generally accelerates the rate of unwanted reactions and further facilitates such reactions because above the glass transition temperature of polymer segments oxygen diffusion is greatly enhanced [23, 24]. To date, studies of accelerated weathering of epoxy resins have been limited, especially studies explicitly comparing aliphatic and aromatic epoxy resins. Considering the potential for any aromatic group to act as a chromophore and absorb UV radiation, beginning degradation phenomena, we wished to determine whether replacing aromatic groups with aliphatic groups could lead to any improvement in environmental stability. To date, studies of accelerated weathering have not compared two compounds similar except for aromaticity under an extended weathering regime. Previous researchers have used relatively mild conditions (e.g., DGEBA/hemp fibre composites at up to 1000 h alternating 1 h of UV irradiation with 2 h of heating to 50 °C under moist conditions [25]), or have used harsh conditions over shorter periods of time. Over 26 h of accelerated weathering by UV irradiation at 60 °C, the carbonyl and hydroxyl regions of a DGEBA resin were found to increase significantly [26]. On comparison of three types of epoxy adhesives exposed to accelerated weathering through alternating UV irradiation at 70 °C and heating under moist conditions at 50 °C, for exposure times up to 200 h, an aliphatic epoxy adhesive type (butyl glycidyl ether of bisphenol A) was found to have more resistance overall than polyvinyl acetate and methylated melamine formaldehyde [1]. In this work, aromatic (based on the diglycidyl ether of bisphenol A, DGEBA) and aliphatic (based on the hydrogenated diglycidyl ether of bisphenol A, HDGEBA) but otherwise identical resins were exposed to the same extended regime of alternating UV and high-temperature moisture exposure, and their mechanical properties, thermal degradation, and surface chemistry evaluated as a function of time.

Experimental

Materials

The epoxy resins used in this study consisted of the glycidyl ether of bisphenol A (DGEBA, Sika Ltd., ‘Sikadur 330A’) provided by Sika Australia Pty. Ltd., and hydrogenated glycidyl ether of bisphenol A (HDGEBA, CVC Thermoset Co., ‘Epalloy 5000’) was supplied from Brentage Australia Pty Ltd., Sydney, Australia. The hardener was 2,2,4-trimethylene-1,6-hexadiamine with hydrocarbon solvent (TMDA, Sika Ltd. ‘Sikadur 330B’) provided by Sika Australia Pty. Ltd., The chemical structures of the commercial epoxies and curing agent were verified using NMR and GC-MS and are shown in Fig. 1.

Fig. 1
figure 1

Chemical structures of materials utilized in this study

Full size image

Sample preparation

DGEBA or HDGEBA epoxy was mixed with TMDA at a ratio of 20 g curing agent per 100 g of epoxy (a stoichiometric ratio of approximately 4:1 epoxy to curing agent). The mixture was stirred using a wooden rod for 5 min to reduce any entrapped bubbles. The mixtures were poured into plastic moulds and allowed to cure at room temperature for 24 h, followed by post-curing at 50 °C for 5 h for DGEBA specimens and 10 h for HDGEBA samples.

UV irradiation and moisture exposure

Both epoxies were treated using an accelerated weathering chamber equipped with fluorescent UV lamps to generate UVA radiation of maximum intensity at 340 nm. The specimens were exposed to UV irradiation for one week then turned over and exposed for a further week after which they were exposed to 100% humidity at elevated temperature (50 ± 5 °C) for two weeks. Specimens were tested after this procedure. This procedure was then repeated, to give total accelerated weathering times of 1, 2, 3, 4 and 6 months for each specimen.

Characterization

The epoxy specimens before and after degradation were analysed by FTIR, TGA, mechanical testing, and SEM to evaluate the effects of moisture, heating, and UV irradiation on their degradation.

Fourier transform infrared (FTIR)

The FTIR spectra of epoxies were tested with a Perkin Elmer Spectrometer Type Two, Model L1600300 (Perkin-Elmer, Llantrisant, UK). The spectra were measured in the range of 700–4000 cm−1. The carbonyl and hydroxyl indexes (Eqs. 1 and 2) were measured to assess the degree of oxidation in the epoxy resins before and after exposure to accelerated weathering [16]:

$$ \mathrm{Carbonyl}\ \mathrm{index}=\frac{I_{abs}\mathrm{at}\ 1654\ {\mathrm{cm}}^{-1}}{I_{abs}\mathrm{at}\ 2877\ {\mathrm{cm}}^{-1}}\times 100\% $$
(1)
$$ \mathrm{Hydroxyl}\ \mathrm{index}=\frac{I_{abs}\mathrm{at}\ 3300\ {\mathrm{cm}}^{-1}}{I_{abs}\mathrm{at}\ 2877\ {\mathrm{cm}}^{-1}}\times 100\% $$
(2)

Tensile strength tests

Tensile strength testing was performed using a universal testing machine (UTM) (Industrial Series DX 300KN, Instron Ltd., Bayswater, Victoria, Australia). The specimens were tested according to ASTM D638 using the dog-bone shape of all specimens with dimensions 50 mm length, 20 mm width, and 2 mm thickness. The testing speed was 10 mm/min. Specimens were positioned vertically between the grips of the testing machine, and stress-strain curves were plotted during the test. Five specimens were tested for each set of conditions.

Thermogravimetric analysis (TGA)

The thermogravimetric instrument (TA Instruments, Inc. Hi-Res TGA 2950, USA) was calibrated with calcium oxalate. Specimens of 5–10 g mass were placed in an aluminum pan and then heated from 30 to 600 °C at a rate of 10 °C/min.

Scanning Electron microscopy (SEM)

Scanning Electron Microscope (JEOL 6020, USA) was used to investigate structural changes in two types of epoxy before and after accelerated weathering; the specimens were prepared by gold coating.

Results and discussion

FTIR

FTIR analysis was performed to monitor the changes in chemical compositions for the two epoxy resins before and during different accelerated weathering times. Previous studies have shown that carbonyl and hydroxyl groups are generated by degradative oxidation reactions following chain scission and hydrogen abstraction from the polymer backbone [27]. FTIR spectroscopy of the surface of specimens exposed to different durations (1, 2, 3, 4 and 6 months) detected an increasing relative intensity of a hydroxyl group in the 3300 cm−1 band and a carbonyl group in 1654 cm−1 (Figs. 2, 3, 4 and 5).

Fig. 2
figure 2

FTIR comparison for DGEBA epoxy before and after exposure to different accelerated weathering times

Full size image
Fig. 3
figure 3

FTIR comparison for HDGEBA epoxy before and after exposure to different accelerated weathering times

Full size image
Fig. 4
figure 4

Carbonyl indices of DGEBA (○) and HDGEBA (□) epoxy before and after exposure to different accelerated weathering times

Full size image
Fig. 5
figure 5

Hydroxyl indices of DGEBA (○) and HDGEBA (□) epoxy before and after exposure to different accelerated weathering times

Full size image

For both resins, there is an initial reduction in the C-H stretching region at about 2900 cm−1, after which little change occurs in this region. The growth of the carboxyl and hydroxyl peaks occurs after the diminution of the resonance in this region [28].

The carbonyl index of unweathered DGEBA specimens was much higher than for HDGEBA specimens (Fig. 4); the hydroxyl index was also higher for the unweathered DGEBA specimen but to a much less marked extent (Fig. 5). Over the course of weathering, the carbonyl and hydroxyl indices increased for both epoxies, with the extent of DGEBA degradation as evidenced by FTIR increasing both in absolute and relative terms. The hydroxyl and carbonyl indices did not increase smoothly, with the increase in relative value compared to HDGEBA arising primarily in the last few months of accelerated weathering. These results indicate that on a molecular level, HDGEBA-derived resins have better resistance than DGEBA-derived resins to the UV, temperature, and moisture conditions employed.

Tensile tests of DGEBA and HDGEBA

Stress-strain curves were obtained for five epoxy resin specimens of each material before exposure to accelerated weathering and at each exposure time (1, 2, 3, 4, 6 months). All tensile test curves are given in the Supplementary material. The overall behaviour of stress at break and elongation at break as functions of accelerated weathering time are shown in Figs. 6 and 7, respectively.

Fig. 6
figure 6

Stress at break for DGEBA (○) and HDGEBA (□) epoxies before and after exposure to accelerated weathering

Full size image
Fig. 7
figure 7

Elongation at break for DGEBA (○) and HDGEBA (□) epoxies before and during exposure to accelerated weathering

Full size image

Before exposure to accelerated weathering conditions, the DGEBA epoxy specimens had significantly higher tensile strength compared with the HDGEBA specimens (Fig. 6). However, the effect of six months of accelerated weathering was only a small loss of tensile strength of HDGEBA samples, in contrast to DGEBA epoxy specimens whose strength significantly decreased over this time. Specimens exposed to accelerated weathering consistently gave more elongation at break than unweathered specimens for both epoxy resin types (Fig. 7). After six months of weathering, HDGEBA-TMDA epoxy specimens exhibited a greater elongation at break (7.5 ± 0.3%), than did the DGEBA-TMDA specimens (2.0 ± 0.1%). Table 1 shows the comparison between our results and previous studies. The results in Table 1 exhibited the highest tensile strength (38 ± 0.5 MPa) obtained from this study’s specimens, compared with previously reported tensile strength of Sikadur 330 specimens in the range 30 to 32 MPa. Values of elongation at break were found to be 0.90 ± 0.21 while the Sikadur company data sheet reports 0.9% and previous researchers obtained 1.5%, 1%, and 0.7% [29,30,31].

Table 1 Mechanical properties of DGEBA-TMDA epoxy (Sikadur 330)
Full size table

While absolute values of elongation at break increased for both epoxy types as exposure to accelerated weathering increased, and absolute values of tensile strength decreased, there was no significant different in the relative elongation at break or tensile strength value; between one and six months the DGEBA resin had a tensile strength a factor 2.4–3.2 greater than HDGEBA, and an elongation at break a factor of 2.2–2.9 less than HDGEBA, with no clear trend over time.

Thermogravimetric analysis

Typical thermograms and derivatives of the DGEBA-TMDA epoxy and HDGEBA-TMDA epoxy investigated are given in Figs. 8 and 9, respectively. The thermal stability of each composition showing the onset temperature (Ti), the temperature at the maximum rate of decomposition (Tmax) and the residual mass for the weathered and control specimens are summarised in Tables 2 and 3. The initial mass loss before 150 °C is most likely due to loss of relatively volatile organic solvents present in the initial epoxy formulation, rather than moisture, as it is greater before weathering rather than after exposure to moist conditions. It can be seen that the specimens before and after exposure show a one-stage decomposition process independent of the exposure time, whereas the amount of residue at 600 °C strictly depends on the exposure time. Before accelerated weathering, HDGEBA specimens had a higher Ti (410 °C), than DGEBA specimens (395 °C) (Figs. 8 and 9). The onset of decomposition significantly decreases with increasing time of exposure, more so with DGEBA than HDGEBA. After exposure times of one month and two months, there were no observed changes in the weight loss of both two epoxies. The Tmax of the two epoxies also decreased with exposure (from 395 °C to 379 °C for DGEBA and from 410 °C to 390 °C for DGEBA). Unweathered HGEBA-TMDA sample had more residue at 500 °C (36.5) than unweathered DGEBA-TMDA sample (25.3), and this trend continued to be evident after six months of accelerated weathering, although the absolute amount of residue decreased (14.3% and 7.3%) respectively (Tables 2 and 3).

Fig. 8
figure 8

TGA of DGEBA-TMDA epoxy before (dotted line) and after (solid line) exposure to 6 months accelerated weathering

Full size image
Fig. 9
figure 9

TGA of HDGEBA-TMDA epoxy before (dotted line) and after (solid line) exposure to 6 months accelerated weathering

Full size image
Table 2 TGA results of DGEBA epoxy before and after different accelerated weathering times
Full size table
Table 3 TGA results of HDGEBA epoxy before and after different accelerated weathering times
Full size table

Structural changes (SEM)

Micrographs of DGEBA and HDGEBA epoxy specimens before and after exposure to accelerated weathering are shown in Figs. 10 and 11. The micrographs show no significant changes on the surface after exposure to up to two months accelerated weathering for both DGEBA and HDGEBA epoxies, but voids at the surface can be seen over longer periods of accelerated weathering for the DGEBA epoxies. There was no significant evidence of any surface structural changes in the HDGEBA-based epoxies even after six months of accelerated weathering.

Fig. 10
figure 10

SEM micrographs of DGEBA-TMDA epoxy. a – before exposure, b – 1 month exposure, c – 3 months exposure, d – 6 months exposure to accelerated weathering

Full size image
Fig. 11
figure 11

SEM micrographs of HDGEBA-TMDA epoxy. a – before exposure, b – 1 month exposure, c – 3 months exposure, d – 6 months exposure to accelerated weathering

Full size image

Conclusions

Both epoxy resins were significantly degraded after six months exposure, compared with unweathered specimens. In general, the aliphatic epoxy resin (HDGEBA) exhibited more resistance after exposure to accelerated weathering, although it started out with poorer physical properties. Fewer chemical changes to HDGEBA were seen on exposure to accelerated weathering, compared with the aromatic epoxy resin (DGEBA) which showed a marked increase in relative carbonyl and hydroxyl index values. DGEBA had better physical properties initially, and over time the poorer physical properties of HDGEBA were not compensated by significantly improved retention of mechanical properties on weathering. The thermal stability of the HDGEBA-based epoxies was less affected by accelerated weathering in comparison with DGEBA-based epoxies. SEM tests confirmed the degradation of two epoxies after accelerated weathering but showed less effect on the HDGEBA-TMDA epoxy after six months exposure. These results are qualitatively expected, since DGEBA contains a UV chromophore which will cause it to be more sensitive to UV radiation and will be liable to form excited states which can act as loci for degradation. The loss of strength of both epoxy resins was about 50% after six months of weathering, and the elongation at break increased for both DGEBA-TMDA and HDGEBA-TMDA epoxy specimens by similar factors under accelerated weathering, showing the greater degradation at the level of functional groups and surface imperfections was not reflected in a significantly greater level of structural degradation. This suggests that the UV chromophore of DGEBA is unlikely to be significantly accelerating degradation under the conditions observed, and will not make DGEBA-based resins unsuitable for outdoor applications.