Thermal stability and degradation of diglycidyl ether of bisphenol A epoxy modified with different nanoclays exposed to UV radiation

Polymer Degradation and Stability 98 (2013) 759e770

Contents lists available at SciVerse ScienceDirect

Polymer Degradation and Stability journal homepage: www.elsevier.com/locate/polydegstab

Thermal stability and degradation of diglycidyl ether of bisphenol A epoxy modified with different nanoclays exposed to UV radiation Alfred Tcherbi-Narteh, Mahesh Hosur*, Eldon Triggs, Shaik Jeelani Department of Material Science Engineering, Tuskegee University, 104 James Center, Tuskegee, AL 36088, USA

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 July 2012 Received in revised form 12 December 2012 Accepted 13 December 2012 Available online 3 January 2013

The primary focus of this study was to evaluate the effects of different montmorillonite nanoclays (MMT) on the thermal stability and degradation of epoxy composites exposed to UV radiation and elevated temperatures. Diglycidyl ether of bisphenol A (DGEBA) epoxy resin, SC15 was reinforced with three different montmorillonite nanoclays, NanomerÒ I.28E, CloisiteÒ 10A and CloisiteÒ 30B. Thermal properties of modified DGEBA nanocomposites were characterized. Subsequently, neat and nanocomposites were subjected to 500 h of UV radiation and characterized to determine the effects of various nanoclays on the degradation. Addition of nanoclays increased the thermal properties compared to the unmodified composite and better retention of material properties after exposure to UV radiation. Viscoelastic properties increased with addition of nanoclays in both unexposed and UV radiation exposed samples. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: MMT e montmorillonite nanoclay DGEBA e diglycidyl ether of bisphenol A UV radiation Degradation Thermal stability

1. Introduction Montmorillonite nanoclays have been widely used as fillers to enhance material properties of different polymers matrices used in fiber reinforced polymer (FRP) composites fabrications. These FRP composites have gained wide acceptance and currently being used as replacements for traditional isotropic engineering materials in many applications across many industries. This is due to the fact that FRP composites offer high stiffness to weight ratio, anti corrosive properties and therefore used in fuel efficient and weight critical applications. In many of these FRPs, matrix is modified to enhance properties of composites using different fillers. In recent years, more attention has been focused on montmorillonite nanoclays as studies have shown that with addition of these clays many desirable properties have been enhanced. Properties such as mechanical [1e4], thermal [5e7] and barrier [8,9] have all been realized with less than 10 wt. % loading of the clay. Montmorillonite nanoclays are more abundant and relatively cheaper compared to other nanoparticles. The inherent clay structure inhibits the flow of moisture and other volatiles in the event of moisture exposure or elevated temperatures leading to improvements in fire retardancy properties which was first reported by

* Corresponding author. Tel.: þ1 334 724 4220. E-mail addresses: [email protected], [email protected] (M. Hosur). 0141-3910/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.polymdegradstab.2012.12.013

Blumstein in 1965 [10]. Addition of nanoclays into polymers has also exhibited a potential of slowing down the degradation mechanisms often present during UV radiation exposure and elevated temperatures [11,12]. However, there are reports in literature [13e16] showing increased activities of photo-oxidation and chemical degradation reactions caused by the presence of montmorillonite nanoclay in polymeric composites. The influence of MMT on degradation of polymeric nanocomposites has been attributed to the presence and behavior of transition elements during UV radiation interaction [15,16]. These transition elements in MMT undergo redox reactions which can vary widely during photo-initialized chemical reactions leading to different outcomes of different MMT systems [17]. Most of the desirable properties in polymer layered silicate nanocomposites have been mainly driven by the type and surface modification of clay along with the quantity and degree of dispersion [1,18e20]. Another important factor is the interaction between clay particles and different polymer molecules to form strong interfacial bonding based on the processing parameters. Studies have shown these factors to significantly affect the microstructure of the epoxy resin [21e23], its curing kinetics [24,25] and finally the sought out properties. Some processing parameters such as curing time and temperature can all be affected due to the effects of the clay on viscosity [23] and mobility of reacting species to form the network chain that determines the properties of the final composite. It is therefore,

760

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

important to understand the chemistry of both the clay and polymer to choose compatible systems. The application needs of FRP composites have compelled them to become inescapable targets for environmental attacks such as UV radiation, elevated temperature, moisture, humidity, acid rain and many times a combination of these factors simultaneously acting on the material leading to property deteriorations over time. In many of these environmental attacks, the fiber reinforcements are the least affected, and given the viscoelastic nature of polymers, it is therefore essential to protect FRP against such attacks using bottom-up approach by reinforcing the matrix. For example, UV radiation on the surface of such materials creates a vicious cycle of hydrogen abstraction from the polymer molecules initiating at the surface of the material [11,12,26]. This leads to the formation of free radicals which initiate other reactions causing brittleness and consequently lower the molecular weights and loss of load bearing capacity and thermal diffusivity, hence overall deterioration in material properties. It is worth mentioning that interactions of UV rays with polymers differ from one polymer to the other [26], causing either random chain scission or increase in cross-linking density and consequently leads to variations in brittleness and reduction in material strength. As mentioned earlier, organomodified nanoclays have shown to posses the ability to minimize such activities, hence suitable for use as fillers in polymers for outdoor applications. In the current work, different montmorillonite nanoclays with different surface modifications were used in modifying diglycidyl ether of bisphenol A epoxy resin (SC15), and the effect of clays on thermal properties were studied. Subsequent studies was carried out on the degradation effects of UV radiation and elevated temperatures on each set of samples, as UV radiation has been identified as one of environmental factors besides moisture to affect most of desirable properties of polymer based composites by setting the stage for other environmental attacks on the polymers. 2. Experimentation 2.1. Materials Commercially available diglycidyl ether of Bisphenol A epoxy resin, (SC15) obtained from Applied Poleramic Inc was modified with three different types of nanoclays. The epoxy resin is a two part system, low viscosity, room temperature curing system with low shrinkage property and highly stable at elevated temperatures. The SC15 (Part B) is a cycloaliphatic amine curing agent. Organo nanoclays used as fillers were montmorillonite nanoclays, a 2:1 phyllosilicate with approximate chemical composition of A0.3(Al1.3Mg0.7)[Si4]O10$(OH)2$xH2O, where “A” is the exchangeable cation, Kþ, Naþ, or 0.5Ca2þ. Typical structure of montmorillonite nanoclay is shown in Fig. 1, and the various clays used were NanomerÒ e I.28E from Sigma Aldrich, CloisiteÒ 10A and 30B from Southern Clay Products Inc. These nanoclays are commercially available with different surface modifications and coefficient of exchange capacity (CEC) value. The hydrophilic clay platelets are incompatible with hydrophobic polymers; hence the surfaces were modified by exchanging some of the cation to make the clay more compatible with polymers. The CEC of NanomerÒ is 93.7 meq/100 g and modified with 25e30 wt. % of quaternary trimethyl stearyl ammonium while the CloisiteÒ 10A and 30B were modified with quaternary dimethyl, benzyl, hydrogenated tallow ammonium and quaternary dimethyl dihydrogenated tallow ammonium respectively. The CEC of CloisiteÒ 10A and 30B were 125 meq/100 g and 90 meq/100 g respectively. Typical chemical structure of the DGEBA e SC15 is shown in Fig. 2.

Fig. 1. Typical structure of montmorillonite nanoclay [2].

2.2. Sample preparation Measured amount of clay was pre-heated in a conventional oven at 100  C for about 2 h and allowed to cool prior to mixing with SC15 part A. The clay tends to absorb moisture in open air due to the hydrophilic nature and therefore heating was necessary to get rid of any trapped moisture. Various clay/epoxy systems under study were fabricated using 2 wt. % clays. During fabrication, measured amount of SC 15 and MMT clay were poured into a beaker and stirred for about 24 h using magnetic stirring in ambient temperature. Stoichiometry quantity of SC15 part B was then added (100 part A: 30 part B), stirred mechanically and desiccated to eliminate any entrapped bubbles due to stirring. The desiccated mixture was poured into molds and allowed to cure at room temperature for 24 h, followed by 4 h post curing in a conventional oven at 100  C to fabricate epoxy/clay nanocomposites. Unmodified epoxy samples were also fabricated using identical process and tested for comparison. 2.3. Characterization 2.3.1. Microstructural X-ray diffraction (XRD) studies were done on the various clay samples and their respective epoxy nanocomposites using RigakuDMAX-2000 with Cu Ka radiation of wavelength l ¼ 1.54 nm, operating at 40 KV and 30A. Basal spacing of the pure MMT samples were determined, and used as a baseline to study the dispersion state of the clay in the epoxy nanocomposites. XRD scan counts were accumulated every 0.01 (2q) from 2 to 40 at a scan rate of 0.2 during the study. Degree of dispersion of the clay structure in various nanocomposites after dispersion into epoxy was characterized by

Fig. 2. Typical chemical structure of DGEBA (SC15).

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

thermal expansion was determined from the slope of dimensional change versus temperature utilizing

comparing the diffraction pattern of the nanocomposites and that of the powder. Further studies on morphological changes due to addition of different MMT were performed to compliment XRD studies using high-resolution transmission electron microscope (TEM). Samples for TEM studies were microtomed from each epoxy nanocomposite mold using a diamond cutter and collected on the surface of water and transferred onto 200 Cu mesh at room temperature. TEM micrographs from the study are presented and analyzed.

(1)

where L is the original length, dL, change in length per unit change in temperature, dt is change in temperature. 2.3.4. Thermal stability Thermogravimetric analysis (TGA) was performed on all samples to determine the thermal stability of the nanocomposites due to the addition of different nanoclay using Q500 from TA Instruments Inc. TGA scans were performed under nitrogen environment, with purge flow rate of 60 ml/min from 30 to 650  C at a heating rate of 10 C/min to determine the thermal parameters of each composite and compared to the neat system. These parameters include onset and decomposition temperatures, and temperatures at which each material loses 50% of its original weight along with residue for each sample set. Three samples were tested and the average values are reported. During TGA runs as the samples are heated various products are formed from each sample and the activation energies involved can be determined through the statistical average values of all energies involved in the degradation process. Furthermore, activation energy of decomposition can be related to the strength of the various bonds being broken within the polymer chain network as the materials are subjected to higher

2.3.3. Thermo mechanical Thermo-mechanical analysis (TMA) was performed on unconditioned and UV conditioned samples to determine the effects of clay on coefficient of thermal expansion (CTE) according to ASTM D696 [28] using TA instruments Q 400 with an expansion probe. Samples were heated from 30 to 180  C at a heating rate of 5  C/min and the slope of the curve pre and post transition were determined as the CTE pre and post glass transition. The equipment was purged with dry nitrogen gas at a flow rate of 50 ml/min. Coefficient of

(c)

(a)

10000 Cloisite 10A Nanoclay Cloisite 10A/SC15 Nanocomposite

MMT - I.28E MMT - CL 10A MMT - CL-30B

8000

Intensity (au)

Intensity

1 dL  L dt

f ¼

2.3.2. Viscoelastic property Dynamic mechanical analysis (DMA) was performed on neat and nanocomposites using TA instruments Q800 equipment. The DMA was operated in a dual cantilever mode with a frequency of 1 Hz and an amplitude of 15 mm. Samples were heated from 30  C to 180  C at a ramping rate of 5 C/min and the peak of the tan delta curve was used in determining the glass transition temperature according to ASTM D4065-06 [27].

15000

761

10000

6000

4000

5000 2000

0 5

10

15

20

25

0

30

0

10

2 Theta

20

30

40

2-Theta

(d)

(b) 20000

15000 Cloisite 30B Nanoclay Cloisite 30B/SC15 Nanocomposite

Nanomer I.28E Nanoclay Nanomer I.28E/SC15 Nanocomposite

Intensity (au)

Intensity (au)

15000

10000

10000

5000 5000

0

0

10

20

2-Theta

30

40

0

0

10

20

30

40

2-Theta

Fig. 3. a. X-ray diffraction patterns of (a) powdered NanomerÒ I.28E, CloisiteÒ 10A and CloisiteÒ 30B and (bed) comparison with their respective Nanocomposites.

762

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

temperatures. The activation energy and pre-exponential factor gives an insight of the complex nature of polymer decomposition and can be determined using several approaches. In the current study, activation energy of each system was determined using FlynneWalleOzawa [29,30] method which required a minimum of three scans; hence subsequent scans were performed on each sample set at 5, 20 and 30  C/min. Several methods have been cited in literature for determining activation energies that require multiple scans, such as Kissinger [31] Coats and Redfern [32] and some others that require a single scan as in HorowitzeMetzger [33].

a relatively short time. Thermal properties were characterized and compared to their respective unexposed properties. The presence of carbonyl and carboxylic acids were monitored through Fourier transform Infrared (FTIR) spectroscopy on each sample set before and after UV radiation exposure. In addition, energy of UV photons, comparable to dissociation energy of most bonds in polymers [26], once absorbed excites the molecules on the surface. This phenomenon leads to the breaking of bonds in the polymer chains forming free radicals, which further destroy other bonds present in the molecules.

3. UV radiation conditioning

4. Results and discussion

To study the possibility of montmorillonite nanoclay inhibiting the destructive effects of ultra violet radiation (UV), and thermal stability, post cured samples were subjected to UV aging using UV accelerated weathering chamber e QUV/SE (Ohio). UV rays generated from 340 nm fluorescent were allowed to continuously incident on the surfaces of the various nanocomposites and a control sample (neat) over a period of 500 h. Intensity of the UV radiation was set to 0.90W/m2 at 60  C to have intense UV effect at

4.1. Morphology Morphology of various nanocomposites formed was studied using X-ray diffraction (XRD) technique which has been found to be a convenient way of determining dispersion state of nanoclay fillers in nanocomposites. XRD measures the average distance between layers of nanoclay and their relative stacking order, providing valuable quantitative information about the degree of dispersion.

Fig. 4. Transmission electron micrograph of (a) e NanomerÒ I.28E, (b) e CloisiteÒ10A and (c) e CloisiteÒ30B nanocomposites.

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

a

2500 Neat Epoxy Epoxy/Cloisite 10A Epoxy/Nanocor I.28E Epoxy/Cloisite 30B

2000

Storage Modulus, MPa

Intergallery spacing of clay platelets in nanocomposites once obtained are compared with that of the clay particles. XRD results for various nanocomposites in the current studies are shown in Fig. 3. As can be seen, there were distinct peaks characteristic of various nanoclays used, although similar peaks were obtained for the CloisiteÒ 10A and 30B, there were slight distinctions. The resulting nanocomposites showed different pattern (Fig. 3b). The absence of sharp diffraction peaks observed in the powdered clay suggests a lack of order in nanoclay platelets observed in the epoxy/ clay samples. This is also indicative of small mass transfer of epoxy resin into the layers of the nanoclay platelets showing a change in the interlayer spacing of montmorillonite powders forming fully dispersed nanoclay nanocomposites. However, it was unclear if obtained nanocomposite is fully intercalated or exfoliated and therefore high-resolution TEM micrographs were obtained throughout the samples to ascertain and compliment results from XRD studies shown in Fig. 4. Morphological studies from TEM micrographs showed loose or incompact intercalated tactoids especially for SC15/MMT I.28E (Fig. 4a). The micrographs obtained from SC15/MMT CL10A represent a typical exfoliated and intercalated structure in ordered stacked. From Figs. 3 and 4, NanomerÒ I.28E seemed to be well dispersed relative to the others based on the disappearance of a slight hump observed in the diffraction pattern and more loose clay platelets relative to CloisiteÒ 10A and CloisiteÒ 30B nanocomposites.

763

1500

1000

500

0

60

120

180

o

Temperature, C

b

1.0 Neat Epoxy Epoxy/Cloisite 10A Epoxy/Nanocor I.28E Epoxy/Cloisite 30B

0.8

Typical DMA analysis curves of various samples prior to conditioning are shown in Fig. 5, where storage modulus and damping effects (tan d) of each sample are plotted as functions of temperature. Three samples from each set were run and summary of pre and post conditioned results are presented in Table 1. There was an overall increase in storage modulus with the infusion of organo nanoclay into the epoxy resin under study. CloisiteÒ 30B system showed an increase of about 13% in storage modulus while NanomerÒ I.28E and CloisiteÒ 10A both gave a modest increase of about 5% at 30  C. It was also observed that the elastic region of CloisiteÒ 10A was shorter compared to the other systems. For example, at 80  C, CloisiteÒ 10A lost about 26% of its storage modulus at 30  C, while both CloisiteÒ 30B and NanomerÒ I.28E systems lost about 17%, and the unmodified SC15 showed a decrease of about 20%. Glass transition temperatures on the other hand showed a different trend, there was modest increase of about 6 and 5% for NanomerÒ I.28E and CloisiteÒ 30B systems respectively, and a decrease of about 4% for that of CloisiteÒ 10A compared to the neat system. Studies [1,3,5] conducted on epoxy-clay systems have reported increased glass temperatures in some cases whiles others have reported decreased in Tg. Generally, the addition of clay affects the chemistry of the epoxy composition due to interaction of the ions present on the surface of the clay and epoxy molecules. This leads to various enhancements based on the polymer molecules. Chan et al. [34], Lan et al. [35] and several others [36,37] showed that epoxies in the presence of smectite clays sometimes undergo self-polymerization which affects the chemistry and ultimately, the cure reaction leading to property enhancements. During this selfpolymerization, some clay particles may act as catalysts promoting cross-linking while others inhibit. From the above observation, it can be seen that each clay type had an influence on the glass transition temperature. NanomerÒ I.28E showed the highest Tg, possible due to a delay in the cross-linking during curing and therefore as the sample went through the thermal cycle there was further curing leading to higher Tg. Post conditioned samples tested showed a slight increase in glass transition temperatures for all samples including the neat.

Tan δ

0.6

4.2. Viscoelastic properties

0.4

0.2

0 50

100

150 o

Temperature, C Fig. 5. (a). Storage modulus of SC15 and SC15/I.28E/CloisiteÒ10A/30B nanocomposites. (b). DMA thermograms for determination of glass transition temperature.

This increase indicated that samples underwent further post curing increasing cross-linking density and formation of strong bonds. UV radiation exposure may have created a condition for much needed post curing or residual cross-linking of the polymer chain molecules in CloisiteÒ 10A, while having the opposite effect by making the other systems more brittle, hence the result of lower storage modulus. Among the nanocomposites, storage modulus decreased about 6% for NanomerÒ I.28E, 2% for CloisiteÒ 30B systems, while a 5% increase was observed with CloisiteÒ 10A system. It is worth noting that giving the same processing parameters, the addition of CloisiteÒ 10A may have slowed the chemical reaction involving the opening of the epoxide ring and further interaction with the amine hardener. This confirms the previous explanation given for lower storage modulus in unconditioned SC15/CloisiteÒ 10A composition. 4.3. Thermo-mechanical analysis Summary of thermomechanical analysis (TMA) test results using 3 samples from each set are presented in Table 2 and typical thermograms in Fig. 6. Results from unconditioned samples indicated the addition of different MMT clays increased the coefficient of thermal expansion (CTE) pre-transition for each system compared to the neat and up to 5.5% increase post transition. The

764

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

Table 1 DMA results of unconditioned and conditioned samples. Sample

Storage modulus, MPa Uncond

Neat MMT I.28E Cloisite 10A Cloisite 30B

2052.33 2161.00 2163.00 2313.33

Glass transition, Tg ( C)

% change UV Cond.

   

93.09 59.62 46.11 27.15

2009.13 2080.67 2281.00 2248.67

   

Uncond 2.10 6.39 5.46 2.80

19.37 83.48 49.00 55.04

increase could be attributed to the fact that the MMT clay increased the intermolecular chain length causing a decrease in the strength of the bonds. An increase in intermolecular distance results in weak intermolecular force of attraction, hence lower bond strength. The CTE increased in the order of 20, 30 and 25% for I.28E, CloisiteÒ 10A and 30B systems respectively. As the samples were conditioned, we speculate that the samples underwent further post curing resulting in stronger bond formation as evident in the DMA post conditioned results. This was also apparent in TMA results of post conditioned samples where there was significant decrease in CTE for all nanophased samples and dismal change in the neat system (Table 2). Lower CTE reported in Table 2 also indicates a volumetric shrinkage due to the formation of stronger bond compared to their respective unconditioned samples. The data also suggests the samples being dimensionally stable after UV radiation since dimensional changes were lower compared to their respective unconditioned samples. CTE values of post conditioned samples showed a mere decrease of 2% for neat system, an indication that the exposure to UV radiation had little effect on the structure of the system unlike the nanophased. Furthermore, these results can be interpreted as incomplete cured systems for all nanophased samples, as studies have shown that for modified epoxy systems, cure kinetic parameter may vary from that of a neat system. Given that the same processing parameters were used for all sample fabrication, it becomes clear that the systems used in this study may not have been fully cured. It is safe to assume that further exposure to UV radiation beyond this point could initiate the destructive effects of UV radiation in the neat system while nanophased samples may still have thermal stability causing delayed onset. Further exposure may be required to substantiate this assumption.

113.11 117.64 106.61 116.30

   

% change UV Cond.

3.24 0.77 0.74 0.46

112.01 121.98 118.73 117.85

   

0.48 0.23 1.16 0.33

0.97 3.69 11.37 1.33

at 10  C/min, and at various heating rates are presented in Tables 3 and 4 respectively. The onset of decomposition (Ti) provides important information about thermal stability of a compound as it shows the temperature at which noticeable mass loss begins. From Fig. 7 and summary of the results in Table 3, it can be seen that the onset of decomposition varied significantly among the various compositions and the decomposition mechanism was complex and varied from one composition to the other. Decomposition profile of SC15 and SC15/NanomerÒ I.28E appeared to be similar, and involves two-step decomposition, with decomposition temperatures of 368 and 370  C respectively. Compositions of SC 15/CloisiteÒ 10 and 30B showed a somewhat single peak with a broad shoulder with peak temperatures of 348 and 346 respectively as shown in Fig. 9. Decomposition occurred at approximately 14, 15, 16 and 17% weight loss for neat, SC15/I.28E, SC15/CloisiteÒ 10A and SC15/ CloisiteÒ 30B respectively. To further understand the decomposition profile of the clays, TGA was performed on the clays and the results shown in Fig. 7. From the results, it can be seen that although all clays were montmorillonite, the decomposition profile varied, possibly due to the different type of surface modifications. Thermograms of CloisiteÒ 10A showed two distinct peaks at 222 and 284  C, however after mixing with epoxy, there was only one prominent peak, possibly due to interaction between clay particles and epoxy molecules. Decomposition in the final composites started at a relatively lower temperature for samples with CloisiteÒ 30B and higher for both unmodified and NanomerÒ I.28E modified samples. At the end of the TGA ran, unmodified SC15 had the least residue at the end followed by CloisiteÒ 30B, NanomerÒ I.28E and finally CloisiteÒ 10A with the most residue. A similar trend in residue was also observed with powdered clay TGA shown in Fig. 8. 4.5. Determination of activation energy of decomposition

4.4. Thermogravimetric analysis Thermal decomposition parameters were studied for various SC15 compositions through thermogravimetric analysis (TGA) using dynamic heating mode. Decomposition kinetics of epoxies has been widely studied using different epoxies and several methods [29e33] to determine the kinetic parameters. Typical TGA thermograms of samples at a heating rate of 10  C/min and their corresponding derivative curves are shown in Fig. 7. Summary of the thermal stability of each composition showing the onset temperature (Ti), temperature at 50% weight loss (T50) and maximum decomposition temperatures (Tp) including residue for all samples

To determine the kinetic parameters such as activation energy and pre-exponential factor, TGA was performed at different heating rates of 5, 10, 20 and 30  C/min. Generally, thermal decomposition of polymeric materials is a complex phenomenon involving several mechanisms acting separately or concurrently, hence it is very difficult to define degradation mechanism for a particular system. Activation energy provides important information for the optimization and design parameters for extensive use of such materials in elevated temperature environment. As expected in all samples, onset and decomposition temperatures increased with increasing heating rate (Table 4). Activation

Table 2 Summary of thermo-mechanical analysis. Sample

Coefficient of thermal expansion CTE, mm/(m  C) Pre e Tg

Neat MMT- I.28E Cloisite 10A Cloisite 30B

75.50 90.74 98.81 94.79

   

2.45 2.58 2.92 4.11

% change

UV pre e Tg 73.76 72.10 65.45 74.45

   

1.65 2.83 2.82 5.52

Coefficient of thermal expansion CTE, mm/(m  C) Post e Tg

2.30 20.54 33.76 21.46

181.80 189.63 191.15 189.30

   

% change

UV post e Tg 0.42 1.15 0.38 1.84

185.16 188.17 188.07 195.93

   

14.31 1.56 4.15 2.45

1.85 0.77 1.61 3.50

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

Fig. 6. (a). TMA thermograms of unexposed samples. (b). TMA Thermogram of samples exposed to 500 h UV radiation.

Fig. 7. (a). TGA thermograms of weight loss as a function of temperature for SC15 and SC15/I.28E/CloisiteÒ10A/30B. (b). Derivative weight loss curve for SC15 and SC15/I.28E/ CloisiteÒ10A/30B.

energy was evaluated based on data obtained from the thermal parameters at maximum rate of decomposition for each heating rate. In the current study, FlynneWalleOzawa method was used in analyzing the data obtained from multiple heating rates for unconditioned and conditioned samples. 4.6. Kinetic analysis In most thermal reaction involving solid compounds, the rate of chemical degradation or conversion (da/dt) can be expressed as a linear function of temperature-dependent constant k, and a function of conversion, a as

df ¼ Kf ðfÞ dt

765

(2)

The rate constant K(T) has been described by Arrhenius expression


E K ¼ Aexp  RT

(3)

where A is the pre-exponential factor, E is the activation energy of decomposition, R is the universal gas constant (R ¼ 8.314 kJ/mol), and T is the absolute temperature. Therefore Equation (2) can be expressed as


df E ¼ Af ðfÞexp  dt RT

(4)

Table 3 Summary of TGA data at 10  C/min. System

SC15 I.28E CL-10A CL-30B

Ti,  C

%

Uncond.

UV

355.72 355.10 330.72 315.72

305.05 320.55 310.55 312.91

14.24 9.73 6.10 0.89

T50,  C

%

Uncond.

UV

406.5 404.41 376.43 358.61

352.39 363.82 352.6 351.97

13.31 10.04 6.33 1.85

Tp,  C

%

Uncond.

UV

367.67 368.87 348.51 328.58

328.73 335.16 325.01 327.535

10.59 9.14 6.74 0.32

Residue, %

%

Uncond.

UV

3.77 5.75 5.86 5.33

2.60 4.30 4.52 4.27

30.93 25.30 22.88 19.89

766

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

Table 4 Decomposition parameter of SC 15 and MMT/SC15 compositions.

SC-15 Epoxy

2 wt% I.28E/SC-15

2 wt% Cloisite 10A/SC-15

2 wt% e Cloisite 30B/SC-15

Tonset C

5 10 20 30 5 10 20 30 5 10 20 30 5 10 20 30

336.53 355.72 379.34 400.31 327.04 355.10 378.54 395.50 313.69 330.72 358.87 359.27 301.39 315.72 335.46 346.03

T50 C

223



Tdecomp C

382.62 406.50 423.46 444.35 388.20 404.41 429.43 441.81 363.15 376.43 399.32 403.38 345.24 358.61 378.66 388.89

348.47 367.67 387.69 408.78 353.40 368.87 423.40 435.28 332.05 348.51 369.87 375.74 314.95 328.58 346.14 357.79

3.80 3.77 3.26 3.11 6.81 5.75 4.81 4.05 6.79 5.86 4.60 2.39 4.58 5.33 4.63 4.68


 df A E ¼ f ðfÞexp  b dT RT

b

ZT

(6)

4.7. FlynneWalleOzawa method In this method activation energy of decomposition can be determined using an integral form of Equation (5) without the knowledge of the reaction order. The method applies Doyle’s approximation for the integration and can be expressed as;



 AE E  2:315  0:4567 gðaÞR RT

Cloisite 10A Cloisite 30B Nanomer I.28E

C

0.2

0.1 610

o

C 640

581

o

o

C

C

0

200

400

600

800

Fig. 8. Thermal decomposition profile of powdered NanomerÒ I.28E, CloisiteÒ10A and 30B.

(5)


E dT exp  RT

o

C

o

0

log b ¼ log

317 o

Temperature, C

Integrating equation (5) yields integrated form of conversion dependence function.

A

C

274

0

When samples are heated at different rates, the degree of conversion varies significantly based on the amount of heat being transferred into the solid and can be evaluated as a function of temperature and heating rate (b). Introducing the heating rate (b ¼ dT/dt) and rearranging Equation (4) yields;

gðaÞ ¼

o

Residue, % o

Heating rate,  C/min



Derivative Weight Change, %/ C

System



0.3

(7)

Activation energy of decomposition can be determined from the slope of the plot of logarithm of the heating rate (b) versus the

reciprocal of temperature (1/T) for a constant g(a) which based on Equation (7) is 0.4567(E/R). In all the TGA runs, maximum conversion occurred between 15 and 18%. However, there were still chemical conversion taking place and therefore activation energy for each system was determined with conversion a ¼ 0.05e0.90. Results of activation energy at the various conversion based on FlynneWalleOzawa are shown in Table 5 for unconditioned and conditioned samples. From the results, it can be seen that activation energies of decomposition increased in all nanophased samples compared to the neat system. This also gives a clear indication of the strength of the bonds formed within each nanocomposite with respect to the neat. As samples were exposed to UV radiation the activation energy of decomposition decreased for neat system by nearly 9%, while a relative increase was observed in all nanophased samples. The results once again collaborate inadequate cross-linking in the entire nanophased composites and therefore the exposure enhanced the bond strength in the nanocomposites rather than diminish as in the case of the neat samples. Lower activation energy reported for the neat conditioned samples compared to their respective unconditioned counterparts indicated that the strength of the chemical bonds is becoming relatively weaker with the exposure. It was also clear from the results that onset of material degradation due to UV radiation exposure had already commenced in the neat sample and may have been delayed in the nanocomposites as evident by the reduction in activation energy of decomposition

Table 5 Activation of energies of decomposition by FlynneWalleOzawa method for conversions from a 0.05 to a 0.90. Conversion, a

0.05 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 Activation Energy, KJ/mol

0wt%

I.28E

CL-10A

CL-30B

Uncond.

UV

Uncond.

UV

Uncond.

UV

Uncond.

UV

117.00 124.90 125.20 125.50 131.50 138.00 145.00 150.00 154.80 158.00 136.99

105.30 115.60 116.10 114.30 119.30 125.50 131.80 136.50 140.20 143.70 124.83

113.70 144.20 142.30 145.60 154.90 167.50 179.90 194.50 210.90 228.60 168.21

148.30 152.80 148.40 153.10 158.10 165.40 175.10 185.50 197.20 218.10 170.20

121.60 128.90 127.90 132.80 141.80 153.60 164.80 177.60 190.50 200.50 154.00

110.30 127.00 131.40 133.50 142.70 152.40 163.60 174.70 186.00 199.80 152.14

146.10 150.60 148.00 147.70 159.00 169.10 179.20 188.50 202.70 217.60 170.85

161.30 166.90 158.70 158.10 163.50 172.30 179.80 188.50 202.00 215.50 176.66

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

During fabrication of epoxy resin composites, there is an inherent production of carbonyl and carboxylic acid by-products due to reaction of epoxy resin and curing agent. These compounds in the presence of UV radiation and air can cause thermal and photo-oxidation degradation, reducing the viability of these polymeric materials over time. Photo- and thermo-oxidation in polymers leads to formation alkyl radicals, which interact with polymer molecules to form hydroperoxides by causing chain scission and other forms of degradation mechanisms [11,12,17,26]. The influence of MMT on degradation mechanism varies, based on the type and quantity of clay loadings and the polymer system being used, hence different thermal stabilities enhancement have been reported [16,38e41]. Role of different MMT in polymer degradation have also been found to be affected by different environment, as oxygen may tend to promote photo-oxidation reactions [38]. Thus the presence of MMT can increase the rate of chemical decomposition and accelerate photo-oxidation reaction due to transition metals in MMT and their interaction with UV radiation [13e15,42,43]. Behavior of these transition elements in MMT in redox reactions can vary widely during photo-initialized chemical reactions leading to different outcomes of different MMT systems [16]. To fully establish the influence of different MMT nanoclay on the photo-oxidation and thermal degradation of exposed samples under investigation, FTIR studies were conducted on all samples including the neat. From the spectrum, characteristic bands of SC15 epoxy resin were observed at various wavenumbers shown in Fig. 9(a, b), with varying intensities based on the addition of different MMT clays. Wavenumbers at 1032 and 1097 and 1238 cm1 are characteristic bands for CeO stretching of saturated aliphatic primary alcohols [44,45] which were observed to decrease in intensity with different clay content. The bands at wavenumbers 1454 and 1511 cm1 are attributed to the aromatic ring stretching of C¼C, characteristic of DGEBA epoxy systems [44,46]. The band at 1511 cm1 may also represent NHþ 3 deformation from the cycloaliphatic amine curing agent. However, unlike the bands observed at with CeO, there was restrictive stretching with addition of Nanomer I.28E and Cloisite 10A with increased stretching with CloisiteÒ 30B MMT. This may be possible due to the increased molecular distance with CloisiteÒ 30B MMT leading to lower intermolecular force of attraction. Results from unconditioned samples showed that addition of MMT tapered off the formation of the carbonyl stretching radicals especially in samples with CloisiteÒ 10A MMT. UV radiation degradation activities were monitored through the formation of carbonyl, alkoxy, hydroperoxide and other radical groups formed during conditioning. Signs of material degradation can be physical involving surface cracks discoloration, loss of weight, embrittlement while chemical occurs due to mutual interaction involving free radicals formation such as hydroperoxide side groups and residual peroxide catalysts leading to changes in the chemical structure of the polymer. Results of FTIR spectroscopy on unconditioned and UV radiation conditioned samples are presented in Fig. 10(aed) with baseline correction showing the various peaks representing carbonyl (C¼O), hydroxyl and other groups being monitored from each SC15/MMT samples for comparison. Various peaks and their corresponding FTIR band assignments are presented in Table 6 for both unconditioned and conditioned

(a) 2920 cm

Unmodified SC15 Composite SC15 / I.28E Nanocomposite SC15 / CL-10A Nanocomposite SC15 / CL-30B Nanocomposite

0.050

2856 cm 3744 cm

Absorbance

4.8. FTIR analysis

samples. As the samples were conditioned, several molecular activities were observed in each sample as shown in Fig. 10 with respect to their unconditioned counterpart. From Fig. 11 it is evident that in all the samples there was an increase in intensity of the broad peak observed between wavenumbers 3000e3600 cm1 which corresponds to characteristic peaks for aliphatic alcohols stretching, symmetrical primary and secondary amines (eNH2 and eNH) and amides, alkyl acetylene (C¼C) and carboxylic acids. The broad shape of the peak suggests that several hydroxyl and carboxylic species were formed simultaneously during UV exposure as the absorbance increased within this range with the exception of CloisiteÒ 30B samples in which these activities were suppressed compared to its unconditioned counterpart. Particularly, the wavenumber 3387 cm1 assigned to depolarize primary amine (NH2) symmetrical stretching, which was present in all unconditioned samples and conditioned except CloisiteÒ 30B in which it had disappeared after UV conditioning. There was a slight increase in intensity of peaks observed at 2926 cm1 assigned to eCH2 stretching for unmodified and CloisiteÒ 30B samples, while a relative decrease was observed for NanomerÒ I.28E and CloisiteÒ 10A samples. The absorption band of carbonyl radicals was present in all samples prior to conditioning and increased in intensity for all post conditions samples. The intensities of the peaks at 1720 cm1 ascribed to combination of aliphatic carboxylic and ketones varied for each sample as indicated in Table 6. At 1602 cm1 there was a distinct peak in unconditioned neat samples which was attributed to the presence of hydroxyl

3371 cm

0.025

0

3800

3600

3400

3200

3000

2800

2600

-1

Wavenumber, cm

(b)0.20 1097 cm

0.15

1032 cm

824 cm

1238 cm

757 cm

1511 cm

Absorbance

values reported. It becomes clear from the test that material degradation might have already began in the neat and barely with CloisiteÒ 10A samples as marked by the reduction in activation energies of 9 and 1.20% respectively.

767

0.10

1454 cm

0.05 Unmodified SC15 Composite SC15 / I.28E Nanocomposite SC15 / CL-10A Nanocomposite SC15 / CL-30B Nanocomposite

0 2000

1800

1600

1400

1200

Wavenumber, cm

1000

800

600

-1

Fig. 9. FTIR spectrum of unconditioned samples (a) between 2600 and 3800 cm1 and (b) between 550 and 2000 cm1.

768

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

Fig. 10. FTIR spectrum of unexposed and exposed (a) neat system, (b) NanomerÒ I.28, (c) CloisiteÒ10A and (d) CloisiteÒ30B Nanocomposites.

groups interacting with carbon atoms to form enol compounds. These enol compounds are unstable and hence disappeared during exposure to UV radiation to form ketones due to the high electro negativity of oxygen atoms. After conditioning, the peak at 1602 cm1 shifted slightly to wavenumbers 1608e1644 cm1 indicating the presence of NHþ 2 radicals from primary amine deformation and was present in all post conditioned samples except neat [44]. The intensity of peaks around wavenumber 1241 cm1 shifted to 1256 cm1 and decreased in samples with CloisiteÒ 10A and NanomerÒ I.28E compared to neat system. This peak is characteristic of saturated CH¼CH bonds and stretching of epoxide (eCeO) bonds within the composite [46]. It also indicated that the surface chemicals of MMT reacted to form saturated CH¼CH bonds hence stronger bond were form causing less

vibration. Also several activities were observed in post conditioned samples between wavenumbers 1319e1574 cm1, which represents the stretching of CeO bonds and OH [46] deformation, possible due to the energy absorbed from UV photons. The wavenumber at 1168 cm1 represents the CeO stretching of saturated aliphatic tertiary alcohols with less vibration in samples with CloisiteÒ 30B, and rather pronounce in samples with CloisiteÒ 10A. The above results indicate that photo-degradation due to UV radiation exposure was influenced by the different clays based on their surface modifications, which was measured by the intensities of carbonyl and carboxylic group vibrations. The extent of the influence of each clay on the degradation mechanisms was different hence different clay systems may have different results as reported by Pandey et al. [12], and several others [15,39e41,47].

Table 6 FTIR band assignments and absorption intensities for unconditioned and conditioned samples. Wavenumber, cm-1

3844 3733 2917 1710 1503 1232 1080 799 e 825

Characteristic band

OeH stretching Hydroperoxides, Carboxylic acids Asymmetric CH2 stretching C¼O stretch C¼C aromatic stretching, Amide II mode CeOeC asymmetric stretching Alkyl peroxide, CeO stretching, Primary amine

Absorbance e unconditioned

Absorbance e UV conditioned

SC15

I.28E

CL-10A

CL-30B

SC15

I.28E

CL-10A

CL-30B

0.0065 0.0120 0.0530 0.0215 0.0798 0.1153 0.1515

0.0106 0.0124 0.0512 e 0.0653 0.0819 0.1293

0.0161 0.0185 0.0443 0.0290 0.0719 0.0798 0.1154

0.0202 0.0399 0.0573 0.0357 0.0891 0.1001 0.1449

0.0065 0.0120 0.0333 0.0751 0.0530 0.0987 0.1855

0.0130 0.0186 0.0456 0.0825 0.0653 e 0.1613

0.0173 0.0253 0.0516 0.1080 0.0835 e 0.1540

0.0201 0.0304 0.0328 0.0633 0.0633 0.0603 0.0933

0.1216

0.0893

0.0835

0.1053

0.1287

0.1041

0.1136

0.0663

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

(a) 0.20 797 cm

1168 cm

0.15 1710 cm

Absorbance

1256 cm

0.10

0.05

Unmodified SC15 Composite SC15 / I.28E Nanocomposite SC 15 / CL-10A Nanocomposite SC 15 / CL-30B Nanocomposite

0

2000

1500

1000 -1

Wavenumber, cm

(b)

Absorbance

0.05

Unmodified SC15 Composite SC15 / I.28E Nanocomposite SC 15 / CL-10A Nanocomposite SC 15 / CL-30B Nanocomposite

0

3500

3000 -1

Wavenumber, cm

Fig. 11. FTIR spectrum of conditioned samples (a) between 550 and 2000 cm1 and (b) between 2600 and 3800 cm1.

5. Conclusion Effects of different montmorillonite nanoclays on the properties on DGEBA epoxy resin were studied using various techniques. Based on the results of various characterization methods the following conclusions can be drawn; - Different montmorillonite nanoclays have different effects on the curing of epoxy resin to form nanoclay nanocomposite - Thermal properties of nanocomposites are driven by the surface modification - Surface modification can affect the interaction of clay/epoxy molecules with UV rays leading to better retention of material strength post UV condition. Acknowledgment The author would like to acknowledge the following funding agencies for their financial support in this study; AL-EPSCoR and ONR, Alabama commission on Higher Education for fellowship. References [1] Pavlidou S, Papaspyrides CD. A review on polymer-layered silicate nanocomposite. Prog Polym Sci 2008;33(12):1119e98. [2] Alexandre M, Dubois P. Polymer-layered silicate nanocomposites: preparation, properties and uses of a new class of materials. Mater Sci Eng R 2000; 28(1e2):1e63.

769

[3] LeBaron PC, Wang Z, Pinnavaia TJ. Polymer-layered silicate nanocomposites: an overview. Appl Clay Sci 1999;15(1e2):11e29. [4] Kumar MA, Reddy KH, Reddy YVM, Reddy RG, Naidu SV. Improvement of tensile and flexural properties in epoxy/clay nanocomposites reinforced with weave glass fiber reel. Int J Polymer Mater 2010;59(11):854e62. [5] Lakshmi MS, Narmadha B, Reddy BSR. Enhanced thermal stability and structural characteristics of different MMT-Clay/epoxy-nanocomposite materials. Polym Degrad Stab 2008;93(1):201e13. [6] Leszczynska A, Njuguna J, Pielichowiski K, Banerjee JR. Polymer/montmorillonite nanocomposites with improved thermal propertiesdpart II. Thermal stability of montmorillonite nanocomposites based on different polymeric matrixes. Thermochim Acta 2007;454(1):1e22. [7] Messersmith PB, Giannelis EP. Synthesis and barrier properties of poly (caprolactone) e layered silicate nanocomposites. J Polym Sci Pol Chem 1995; 33(7):1047e57. [8] Manias E, Polizos G, Nakajima H, Heidecker MJ. Fundamentals of polymer nanocomposite technology in flame retardant polymer nanocomposites. Hoboken, NJ: Wiley; 2007. [9] Gilman JW. Flammability and thermal stability studies of polymer layeredsilicate (clay) nanocomposites. Appl Clay Sci 1999;15(1e2):31e49. [10] Blumstein A. Polymerization of absorbed monolayers: II. Thermal degradation of the inserted polymers. J Polym Sci Pol Chem 1965;3(7):2665e72. [11] Woo RSC, Kim JK, Leung CKY, Chen YH, Zhu HG, Li J. Environmental degradation of epoxy-organoclay nanocomposites due to UV exposure. Part I photodegradation. Compos Sci Technol 2007;67(15e16):3448e56. [12] Pandey JK, Ranghunatha RK, Pratheep KA, Singh RP. An overview on the degradability of polymer nanocomposites. Polym Degrad Stab 2005;88(2): 234e50. [13] Shokrieh MM, Bayat A. Effects of ultraviolet radiation on mechanical properties of glass/polyester composites. J Compos Mater 2007;41:2443e55. [14] Morlat-Therias S, Mailhot B, Gonzalez D, Gardette J-L. Photooxidation of polypropylene/montmorillonite nanocomposites. 2. Interactions with antioxidants. Chem Mater 2005;17:1072e8. [15] Qin H, Zhang Z, Feng M, Gong F, Zhang S, Yang M. The influence of interlayer cations on the photo-oxidative degradation of polyethylene/montmorillonite composites. J Polym Sci Pol Phys 2004;42:3006e12. [16] Morlat S, Mailhot B, Gonzalez D, Gardette J-L. Photo-oxidation of polypropylene/montmorillonite nanocomposites. 1. Influence of nanoclay and compatibilizing agent. Chem Mater 2004;16:377e83. [17] Rabek JF. Photodegradation of polymers: physical characteristics and applications. Berlin: Springer; 1996. [18] Xidas PI, Triantafyllidis KS. Effect of the type of alkylammonium ion clay modifier on the structure and thermal/mechanical properties of glassy and rubbery epoxyeclay nanocomposites. Eur Polym 2010;46(3):404e17. [19] Kornmann X, Lindberg H, Berglund LA. Synthesis and characterization of thermoset-clay nanocomposites. J Polym 2001;42:1303e10. [20] Fan-Long J, Kyong-Yop R, Soo-Jin P. Surface treatment of montmorillonite on the thermal stabilities of bisphenol-A diglycidyl dimethacrylate nanocomposites. Mater Sci Eng A 2006;435-436:429e33. [21] Ray SS. Rheology of polymer/layered silicate nanocomposites. Ind Eng Chem 2006;12(6):811e42. [22] Krishnamoorti R, Ren J, Silva AS. Shear response of layered silicate nanocomposites. J Chem Phys 2001;114(11):4968e73. [23] Knauert ST, Douglas JF, Starr FW. The effect of nanoparticle shape on polymernanocomposite rheology and tensile strength. J Polym Sci Pol Phys 2007; 45(14):1882e97. [24] Roman F, Montserrat S, Hutchinson JM. On the effect of montmorillonite in the curing reaction of epoxy nanocomposites. J Therm Anal Calorim 2007; 87(1):113e8. [25] Carrasco F, Pagès P. Thermal degradation and stability of epoxy nanocomposites: influence of montmorillonite content and cure temperature. Polym Degrad Stab 2008;93(5):1000e7. [26] Guillet JE. Fundamental processes in the UV degradation and stabilization of polymers. Pure Appl Chem 1972;1(30):135e44. [27] ASTM standard D4065-06. Standard test method D4065-06 standard practice for plastics: dynamic mechanical properties: determination and report of procedures. West Conshohocken, PA: American Society of Testing and Materials; 2006. [28] ASTM standard D696-08. Standard test method for coefficient of linear thermal expansion of plastics. West Conshohocken, PA: American Society of Testing and Materials; 2008. [29] Flynn JH, Wall LA. A quick, direct method for the determination of activation energy from thermogravimetric data. J Polym Sci Pol chem 1966;4(5):323e8. [30] Ozawa T. Kinetic analysis of derivative curves in thermal analysis. J Therm Anal Calorim 1970;2:301e24. [31] Kissinger HE. Reaction kinetics in differential thermal analysis. J Anal Chem 1957;29(11):1702e6. [32] Coats AW, Redfern JP. Kinetic parameters from thermogravimetric data II. J Polym Sci Pol Lett 1965;3(11):917e20. [33] Horowitz HH, Metzger G. A new analysis of thermogravimetric traces. J Anal Chem 1963;35(10):1464e8. [34] Chan Y-N, Juang T-Y, Liao Y-L, Dai SA, Lin J-J. Preparation of clay/epoxy nanocomposites by layered-double-hydroxide initiated self-polymerization. Polym J 2008;49(22):4796e801.

770

A. Tcherbi-Narteh et al. / Polymer Degradation and Stability 98 (2013) 759e770

[35] Lan T, Kaviratna PD, Pinavaia TJ. Epoxy self-polymerization in smectite clays. J Phys Chem Solids 1995;57(6):1000e10. [36] Benson Tolle T, Anderson DP. The role of preconditioning on morphology development in layered silicate thermoset nanocomposites. J Appl Polym Sci 2003;91(1):89e100. [37] Pustkova P, Hutchinson JM, Roman F, Montserrat S. Homopolymerization effects in polymer layered silicate nanocomposites based upon epoxy resin: implications for exfoliation. J Appl Polym Sci 2009;114(2): 1040e7. [38] Singh B, Sharma N. Mechanistic implications of plastic degradation. Polym Degrad Stab 2008;93(3):561e84. [39] Zhang X, Loo LS. Synthesis and thermal oxidative degradation of a novel amorphous polyamide/nanoclay nanocomposite. Polym 2009;50(12): 2643e54. [40] Kumanayaka TO, Parthasarathy R, Jollands M. Accelerating effect of montmorillonite on oxidative degradation of polyethylene nanocomposites. Polym Degrad Stab 2009;95(4):672e6. [41] Leszczynska A, Njuguna J, Pielichowski K, Banerjee JR. Polymer/montmorillonite nanocomposites with improved thermal properties Part I. Factors

[42]

[43]

[44]

[45]

[46] [47]

influencing thermal stability and mechanisms of thermal stability. Thermochim Acta 2007;453(2):75e96. Qin H, Zhao C, Zhang S, Chen G, Yang M. Photo-oxidative degradation of polyethylene/montmorillonite nanocomposites. Polym Degrad Stab 2003; 81(3):497e500. Tidjani A, Wilkie CA. Photo-oxidation of polymeric-inorganic nanocomposites: chemical, thermal stability and fire retardancy investigations. Polym Degrad Stab 2001;74:33e7. Sales RCM, Diniz MF, Dutra RCL, Thim GP, Dibbern-Brunelli D. Study of curing process of glass fiber and epoxy resin composite by FT-NIR, photoacoustic spectroscopy and luminescence spectroscopy. Mater Sci 2011;46:1814e23. Delor-Jestin F, Drouin D, Cheval P-Y, Lacoste J. Thermal and photochemical ageing of epoxy resin e influence of curing agents. Polym Degrad Stab 2006; 91:1247e55. Socrates G. Infrared and Raman characteristic group frequencies. 3rd ed. Chichester, New York: John Wiley and Sons; 2001. Morlat-Therias S, Fanton E, Tomer NS, Rana S, Singh RP, Gardette J-L. Photooxidation of vulcanized EPDM/montmorillonite nanocomposites. Polym Degrad Stab 2006;91(12):3033e9.