Epoxidized Glycidyl Ester of Soybean Oil as Reactive Diluent for Epoxy Resin

Rongpeng Wang and Thomas Schuman Department of Chemistry, Missouri University of Science and Technology, Rolla, MO 65409

Abstract
Epoxidized glycidyl esters of soybean oil (EGS) have been synthesized and used as reactive diluents for partial replacement of a commercial, bisphenol A-based epoxy resin (DGEBA). The EGS merits include a higher epoxy content and lower viscosity than the epoxidized triglyceride soybean oil (ESO). Thermosetting resins were fabricated from DGEBA systems blended with various amounts of EGS and ESO, using 4-methyl-1,2cyclohexanedicarboxylic anhydride as a curing agent and 2-ethyl-4-methylimidazole as catalyst. The curing behavior and glass transition were monitored by differential scanning calorimetry (DSC), the performance of thermosetting resins was studied by measurement of thermal stability and flexural properties. The results indicate that EGS resins provide better compatibility, intermolecular crosslinking, and yield materials that are stronger than materials obtained using ESO. However, the EGS resin systems significantly reduce viscosity compared to either pure DGEBA or ESO-blended DGEBA counterparts. Therefore, EGS derived from renewable sources holds potential to enable fabrication of complex, shaped epoxy composites for structural applications.

1. Introduction
Epoxy resin is one of the most important thermosetting polymers and widely used in coatings, adhesives and composites due to its excellent mechanical strength, outstanding chemical resistance, good thermal and electrical properties, and low shrinkage upon cure.1 Most commercially epoxy resins are relatively high viscosity liquids or solids. Cured pure epoxy resins are rigid and brittle materials with low impact resistance. Adding linear elastomeric or thermoplastic additives can increase the toughness; however, this invariably results in a corresponding decrease in resin flow and processing difficulty.2 Reactive diluents are used for reducing and controlling the viscosity of epoxy resins to improve wetting and handling characteristics because in the liquid-moulding technologies like resin transfer moulding or pultrusion, the viscosity and resin flow are critical to achieving a quality laminate.3 Recent trends toward lower
Correspondence to: T. Schuman (tschuman@mst.edu)

VOC, higher solids epoxy formulations have also resulted in increased utilization of reactive diluents.1 Phenyl glycidyl ether and n-butyl glycidyl ether are efficient and widely used diluents, but they have been losing interest due to toxicity, volatility or obnoxious odor issues. Industrial trends show increasing interest in longer chain reactive diluents, e.g., C12C14 alkyl glycidyl ethers, neopentyl glycol diglycidyl ether, diglycidyl ether of polypropylene glycol. These longer chain reactive diluents can also behave as flexibilizing agents to increase thermosetting polymer elongation and impact resistance, though there is often a tradeoff of tensile strength, glass transition temperature and chemical resistance.1 Since petroleum resources are limited, polymers based on vegetable oils are of great interest because they are renewable and can significantly contribute to a more sustainable development.4-6 Epoxidized soybean oil (ESO) has attracted great interest because of a plentiful soybean supply in the United States and therefore of relatively low cost. ESO can be crosslinked into thermosetting polymers by various curing agents.7 However, due to a low oxirane content and sluggish reactivity of internal epoxy groups, the cured ESO normally has a low crosslinking density, resulting from partially unreacted ESO and saturated fatty acid chains that act to plasticize and degrade the thermal and mechanical properties of cured resin. Most ESO industrial uses are limited to nonstructural applications such as plasticizers or stabilizers for poly vinyl chloride (PVC)8, oil-base coatings9 with low strength requirements.10 ESO has a moderate viscosity and offers good miscibility with epoxy resins.11 So ESO or their derivatives can be used as reactive diluents for the partial replacement of epoxy resin to decrease the overall cost and improve the processability.2,12,13 Generally, the mechanical strengths of ESO blended resins are not comparable to those of pure, non-modified epoxy resins, while their toughnesses are better due to the introduction of a two phase structure.14-17 Only those oils of poly-unsaturated fatty acid content, especially soybean or linseed oils that can produce dense epoxy functionality resins, are capable to produce satisfactory properties.7,18,19 Epoxidized vegetable oils (EVO) of low oxirane values either are not reactive or impart waxy, non-curing properties to the resin system. In this research, epoxy resins of epoxidized glycidyl ester (EGS) derived from soybean oils were synthesized and examined. The goals were to remove as much as possible the plasticizing effect of saturated components, to further increase the oxirane content, and to minimize viscosity to permit use as an efficient reactive diluent. We reason that EGS with the addition of a terminal epoxy group (glycidyl), which is readily accessible to nucleophilic attack, should further support reactivity compared with standard, commercial ESO. The increase in oxirane content, consequent reduction of the ESO molecular size, and a facilitated removal of the saturated component should each provide a denser intermolecular crosslinking structure and yield a thermosetting resin material with improved properties.

2. Experimental
2.1. Materials Refined, food grade soybean oil was purchased from a local grocery store. ESO was purchased from Union Carbide Corporation. Epichlorohydrin, methylene chloride, meta-chloroperoxybenzoic acid (MCPBA), sodium carbonate, sodium bicarbonate, sodium sulfite, and anhydrous sodium sulfate were purchased from Fisher Scientific. Cetyltriethylammonium bromide (CTEAB), 2-ethyl-4-methylimidazole, and 4-methyl-1,2-

cyclohexanedicarboxylic anhydride (MHHPA) were purchased from Aldrich. Commercial DGEBA was supplied by Momentive with trade name EPON Resin 828. Mold release agent Chemlease 41-90 EZ was purchased from Chem-Trend, Inc. 2.2. Soap and free fatty acid preparation Soybean oil derived free fatty acids (FFA) were made via acid neutralization of soap. Vegetable oil was reacted with sodium hydroxide to generate soap, then acidified with sulfuric acid to pH<2. The lower layer was separated, washing the top FFA layer using hot water and finally the organic layer dried using anhydrous sodium sulfate. The iodine value of the soybean FFA was 133. Generation of the FFA also facilitated removal of saturated acids. Freshly prepared FFA was dissolved in acetone based on the weight ratio of 1:6, then transferred to a flask and purged with N2 gas, cooled to -20 C for overnight, and the formed crystals were removed by vacuum filtration. The procedure could be repeated several times until no further crystals were generated. Iodine value of the fully refined unsaturated soybean FFA was 145. To the FFA/acetone solution of weight ratio of 1:10, 110% (an excess) sodium hydroxide solution (10N) was added dropwise. The reaction lasted for 4 hours and was sparged by N2 gas to prevent air oxidation of the soap. The formed soap was readily filtered by vacuum filtration and dried using in vacuo rotary evaporation and then by heating the soap powder at 110 C for 1 hr. 2.3. Epoxidized glycidyl ester preparation Dry soap and epichlorohydrin, at a mol ratio of 1:10, were heated to reflux and a phase transfer catalyst CTEAB at 2 equiv.-% to equivalent soap was added and refluxed for 30 min, cooled and centrifuged, the solution was decanted to a flask. Excess epichlorohydrin was removed using in vacuo rotary evaporation. The prepared glycidyl ester had an oxirane value of 4.4%.

Glycidyl ester (GE) and sodium carbonate were mixed with methylene chloride. MCPBA dissolved in methylene chloride was added dropwise at a reaction temperature below 15C for 4 hours to complete epoxidation. The reaction mixture was washed with 10% sodium sulfite and then by 10% sodium bicarbonate and water. Methylene chloride was removed by in vacuo rotary evaporation and the product EGS was dried using anhydrous sodium sulfate. Epoxidation also was accomplished by an analogous performic acid process except using toluene as solvent instead of benzene.20 2.4. Characterization Infrared spectra (IR) were conducted on a Nicolet Nexus 470 E.S.P. spectrophotometer. 1H NMR spectra were obtained on a Varian VXR 400 MHz spectrometer using DMSO-d6 as solvent. Iodine value measurements were based on ASTM Method D554-95. Oxirane value was measured using AOCS Method Cd 9-57. Viscosity was tested on a Brookfield LVDV-III+ Ultra Rheometer at 25C. 2.5. Curing reaction The weight ratios of EGS/ESO to DGEBA resin blend chosen for the present work were 0:100 (pure DGEBA), 10:90, 30:70, 50:50, 70:30; and 90:10. Stoichiometric weight of MHHPA curing agent and 1 wt% (based on epoxy part) of 2-ethyl-4-methylimidazole were added to the epoxy resin blend. After mixing by a high shear mixer for 10 min, the mixture was degassed for 30 min, then poured into a mold sprayed with a mold release agent. Curing was performed at 145 oC for 15 hrs for all blends except ESO-DGEBA (90:10) blend, which was firstly pre-cured at 145 oC for 10 min then mixed again and poured into the mould for fully cure at 155 oC for 15 hrs, because of the low reactivity of ESO. The postcure for all samples were performed at 175 oC for 1 hr. The mixtures weighing 2 to 3 mg encapsulated in aluminum hermetically sealed pan were also cured on a differential scanning calorimetry machine at a heating rate of 10 oC/min from 40-250 oC to study the cure behavior of the different formulations. 2.6. Thermal Tests DSC (Q2000, TA Instrument) was used to determine the glass transition temperature (Tg) of cured resin. Measurement was carried out over a temperature range from -40 to 180 C at a heating rate of 20 C/min. Samples were first preheated to 180 C and quenched with liquid N2 to remove any thermal history. TGA (Q50, TA Instrument) was used to determine the thermal stability of cured resin. Measurement was performed from 30 to 750 C at a heating rate of 10 C/min under an ambient air flow environment.

2.7. Mechanical Tests The flexural properties were determined according to the ASTM method D790 on an Instron 4469 universal testing machine. The modulus was determined in a three-point bending mode, with a sample dimension of 102 mm12.7 mm3.2 mm. The span was 50.8 mm, the crosshead speed was set at 12.7 mm/min.

3. Results and Discussion

3.1. Preparation of Epoxidized Glycidyl Ester

Scheme 1. Synthetic route to EGS. (Soybean oil and fatty acids are shown as simplified structures containing only oleic acid though they also contain saturated and polyunsaturated fatty acids. See the text for detail.) Scheme 1 shows the synthetic route to EGS, oleic acid generalizing a soybean fatty acid chain. Preparation of mixed FFA from triglyceride is straightforward and well-developed. Acetone was used as a low boiling recoverable solvent to prepare soap and effect low temperature crystallization. A slight excess of NaOH with higher concentration was desirable when preparing soap from FFA because unsaturated FFAs are prone to dissolve in acetone rather than react with base and also unsaturated FFA soap is more soluble in water.21 Carefully dried and finely powdered soap resulted in greater yields of glycidyl esters of fatty acids.22

A low solubility of soap in epichlorohydrin suggested a phase transfer catalyst was needed to accelerate the reaction. With CTEAB catalyst, the consumption of soap was completed within half an hour under reflux condition. Glycidyl esters can also be prepared directly from FFA in EPCH rich medium, and then subsequently dehydrohalogenating with alkali, but the yield and purity were lower than the soap process.21 The epoxidation of glycidyl ester of soybean oil (GES) was carried out using MCPBA or in situ generated performic acid. The former was more efficient, however, due to the low solubility of MCPBA in methylene chloride, large amounts of recoverable solvent was required for the epoxidation.

Figure 1. IR spectra of soybean oil, mixed-FFA, GES and EGS Figure 1 shows the FT-IR spectra of soybean oil, mixed FFA, GES and EGS. The band at 3008 cm-1 was attributed to the C-H stretching of =CH in unsaturated fatty acid, such as oleic acid, linoleic acid or linolenic acid. New bands at 910 cm-1 and 852 cm-1 in the spectrum of GES, with the disappearance at 937 cm-1 in spectrum of mixed-FFA means the occurrence of glycidyl group. The conversion of double bonds to epoxy was confirmed by the disappearance of the 3008 cm-1 band in GES, and the appearance of band at 752 cm-1 in EGS.

Figure 2. 1H NMR spectrum and structural assignments of FFA, GE and EGS. Table 1. General physical properties of epoxy resins Oxirane oxygen EEW (g/100g sample) 10.1 158 6.9 232 8.6 186

Epoxy Resin EGS ESO DGEBA

Viscosity at 25 oC (mPaS) 70 430 13000

Figure 2 shows the 1H NMR spectra of mixed-FFA, GES and EGS, where linoleic acid was used as a generalized compound for structural assignments. The spectra showed no evidence of side reactions in preparing GES using soap process, nearly quantitative conversion of double bonds to epoxy groups, and no oxirane ring opening during the epoxidation of GES to EGS using MCPBA, i.e., showed complete conversion but a lack of side reactions. General properties of EGS product compared to ESO, ELO, EGL and DGEBA is shown in Table 1.

3.2. Viscosity reducing ability

Figure 3. Viscosity of DGEBA with various EGS/ESO concentrations.

It was found that EGS had inherently lower viscosity than ESO. EGS has an extra glycidyl group and lower molecular weight compared to ESO, which is a triglyceride and has oligomeric behavior. The viscosity reducing abilities of EGS and ESO were tested at different concentrations replacement of DGEBA resin, which has a high viscosity at 13000 mPaS (see Figure 3). ESO and EGS all showed good miscibility with DGEBA; however, EGS exhibited a much better viscosity reducing ability than ESO. Only 30 wt% of EGS reduced the DGEBA resin viscosity to value below 1000 mPaS, which is indispensable for many applications. At least 50 wt% of ESO was needed to reduce DGEBA resin to the same viscosity.

3.3.

Curing of Reaction

Figure 4. Dynamic thermograms of DGEBA-EGS/ESO-MHHPA systems

Table 2. DSC results of curing DGEBA-EGS/ESO-MHHPA systems

Composition Pure DGEBA 10wt% EGS 30wt% EGS 50wt% EGS 70wt% EGS 90wt% EGS Pure EGS H, (J/g) 355.3 347.7 348.2 340.2 336.6 334.8 321.5
*

H, (kJ/mol)

Tonset, ( C) 133.7 138.0 138.1 136.9 136.5 136.7 141.3

Tpeak, ( C) 151.9 156.6 157.7 159.2 160.9 163.3 167.6

66.4 63.8 61.8 58.4 55.9 53.9 52.5

10wt% ESO 30wt% ESO 50wt% ESO 70wt% ESO 90wt% ESO Pure ESO

359.9 320.6 300.2 281.4 250.1 230.0

68.5 63.4 61.7 60.3 55.9 52.6

134.4 135.1 140.3 143.2 147.6 182.6

153.9 156.0 159.8 167.8 205.4 215.9

* based on the total number of epoxy groups.

Differential scanning calorimetry was applied to study the curing behavior of the blended epoxy resins, as shown in Figure 3, the exothermic peaks were characteristic of the epoxy and anhydride curing reactions. Integration of these peaks allows the determination of the enthalpy of curing reaction(H), onset curing temperature (Tonset) and exothermic peaks (Tpeak). The results are shown in Table 2. From Figure 3, the pure DGEBA and ESO all show single exothermic reaction peaks at 152 oC and 216 oC, respectively. The higher Tpeak value of ESO means a slower reaction rate, which was also confirmed by a lower H value in Table 2. A lower oxirane content of ESO and also internal epoxy function groups react sluggishly with MHHPA curing agent. The addition of ESO to DGEBA lead to a shifting of Tpeak and Tonset to higher values. With a decrease of H value, two partially overlapped peaks were clearly observed, especially for 50 wt% ESO or higher replacement, which suggested that there are decreasing levels of ESO miscibility in the DGEBA. Non-homogenous mixing will prevent complete cure of the epoxy resin. ESO has internal, hindered epoxy groups whereas DGEBA has glycidyl groups of less steric hindrance and more reactive than the internal epoxy groups of ESO. Similar results have also been reported by Altuna23 and Boquillon.24

10

The prepared EGS resin showed a quite different and interesting curing behavior. The neat EGS shows two overlapped peaks, similar to the blend of DGEBA and ESO, which is believed to be due to the inherent different reactivity of glycidyl and internal epoxy function groups. The Tpeak and Tonset value of EGS were more than 40 oC lower than ESO, which indicated EGS is much more reactive than ESO. Increased addition of EGS to DGEBA also lead to shifting of Tpeak to higher values, but the Tonset remained nearly constant, and only a 16 oC increase of Tpeak was observed for 90 wt% EGS replacement compared to pure DGEBA, while it was 54 oC for 90 wt% ESO replacement. The H (J/g) also followed a similar trend. The higher oxirane content of EGS, which bears glycidyl groups like DGEBA, appears to facilitate a more homogenous three dimension polymer structure upon curing compared to ESO blends. One note of interest was that a small amount of replacement, e.g., 30 wt% EGS or below, or 10 wt% ESO, had little effect on the H or Tpeak values of DGEBA, which may be related to homogeneity and compatibility with the DGEBA. 3.4. Thermal Properties

Figure 5. The change of glass transition temperatures with various ESO/EGS contents

11

The glass transition temperature (Tg) is considered one of the fundamental characteristics as it relates to polymer properties and processing. For a polymer to serve as a useful plastic, its Tg must be appropriately higher than the temperature of its intended work environment.25 The aliphatic amine,7,26 or boron trifluoride diethyl etherate27,28 cured ESO polymers generally have very low Tg (even below 0 oC). Aromatic amine29, cycloaliphatic amine,2 thermal latent initiator30 or anhydride31 cured ESO polymers generally have higher Tg, but it is still rare to see a cured ESO6 with Tg above 60 oC. Low Tg represents a low crosslink density, as mentioned above. Due to the low oxirane content of ESO, sluggish reactivity of internal epoxy groups with nucleophilic curing agents, the self intermolecular crosslinking of ESO, and unreactive saturated components like stearic acid, palmitic acid or myristic acid that act as plasticizers, which further decrease the polymer Tg. DSC and Dynamic Mechanical Analysis (DMA) are widely used to characterize the Tg. It is necessary to note here that for most thermosetting plastics, the DMA measurement based on the tan peak at a frequency of 1 Hz generally occurs at a temperature as much as 15-20 oC above Tg as measured by dilatometry or DSC.32 The change of cured epoxy resin Tg as measured by DSC is shown in Figure 5. Not surprisingly, the MHHPA cured pure EGS had a much higher Tg at 88 oC, which was nearly 40 oC higher than ESO though still lower than the pure DGEBA resin. Addition of ESO or EGS lead to a decrease of Tg, for smaller replacement, e.g., below 30 wt%, the Tg values of ESO-DGEBA or EGS-DGEBA systems were quite similar, which indicated the Tg behavior was mainly determined by the DGEBA structure. Further increase the replacement contents of ESO or EGS, the Tg values decreased rapidly, especially for the ESO system. The inherent aliphatic long chain structure of ESO and the higher saturated content and lower epoxy groups make it difficult to produce a densely crosslinked structured polymer as some segments will vibrate more freely upon thermal stress.12 It was also found that neat ESO or higher replacement (above 50 wt%) of ESO-DGEBA thermosetting polymer showed a broad transition from the glassy to the rubbery state. Similar behavior was also found in ELO replacement of di-glycidyl ether of bisphenol F (DGEBF) resin,33 which was not found in EGS-DGEBA system. The plasticizing effect of saturated fatty acids in the network32 and/or the different reactivity of ESO and DGEBA lead a broad distribution of polymer molecular weight and indicat a heterogeneous polymer network was formed.34 Figure 6 presents the TGA curves as a function of temperature for the cured epoxy resin. Since ESODGEBA resin had similar thermal stability with EGS-DGEBA resin, only the latter is shown here. TGA results indicated all cured EGS-DGEBA resins appear thermally stable to temperatures below 300 oC. Addition of EGS lead to an earlier onset of degradation. All the resins similarly presented two stage degradation behavior. The first stage of decomposition, from 350 to 450 oC, is believed to be due to the pyrolysis of the crosslinked epoxy

12

resin network, decomposition of unreacted MHHPA, and dehydration of hydroxyl groups. The second loss stage from ~450 to 600 oC was considered to be the complete decomposition of the smaller fragments like cyclized or aromatic degradation byproducts as indicated by a decrease of char residue when EGS component was increased in DGEBA.

Figure 6. TGA analysis of cured EGS-DGEBA blends compared to pure EGS and pure DGEBA. 3.5. Mechanical Performance The flexural properties of the cured resin system with varying ESO/EGS content were determined and the results were listed in Table 3. As can be seen from Table 3, 10 wt% of ESO or EGS replacement of DGEBA lead to an improvement in flexural modulus, although further increases of ESO/EGS content lead to a decrease of flexural modulus. Similar results were also reported in an amine cured soy-based epoxy resin system.34 The flexural stress of EGS-DGEBA exhibited a gradual decrease until 50 wt%, then followed an abrupt degradation. While for ESO-DGEBA, such degradation occurs at 30 wt%. As discussed previously, two overlapping peaks in the curing curves determined by DSC were observed when high contents of EGS/ESO were added into DGEBA. Different reactivity or reduced compatibility of
13

ESO/EGS with DGEBA, may ultimately prevent the ESO/EGS resin from fully participating in the crosslinking. Therefore, the resulting crosslinked thermosets may become increasingly plasticized by partially reacted ESO/EGS resins at high contents, leading to a decrease of flexural strength. EGS has higher oxirane value and was observed to be more reactive with DGEBA compared to ESO, so a higher content of EGS replacement of DGEBA was achieved with less degradation of the mechanical strength. Due to the inherent lower epoxy content and sluggish reactivity of internal epoxy function groups, a lower ESO content replacement was required to prevent additional sacrifice of mechanical performance. Table 3. Flexural properties of cured DGEBA-EGS/ESO resin Composition Flexural StrengthMPa Flexural ModulusMPa Pure DGEBA 10wt% EGS 30wt% EGS 50wt% EGS 70wt% EGS 90wt% EGS 10wt% ESO 30wt% ESO 50wt% ESO 70wt% ESO 90wt% ESO 138.1 133.0 125.2 121.9 91.0 81.3 124.2 120.8 107.0 81.4 60.0 2945.0 3162.1 2837.8 2829.1 2608.2 1687.6 2977.2 2822.8 2471.0 2270.8 1652.8

4. Conclusion
EGS resin materials were produced from soybean oils with reduced saturated FFA fraction content. The products were characterized and showed high oxirane contents and were more reactive than ESO. The EGS blends were cured by MHHPA and their thermosetting polymer Tgs measured in comparison to control ESO cured in similar fashion. The EGS resin systems had significantly reduced viscosity compared to their pure DGEBA or ESO-blend epoxy counterparts. The products displayed glass transitions that were a fairly simple function of oxirane content with some added influence of glycidyl versus internal oxirane reactivity. The products displayed improved Tgs and mechanical properties compared to their ESO counterparts and, in addition to an inherently low viscosity and efficient viscosity reduction, should therefore be more attractive as

14

a reactive diluent.

For instance, EGS derived from renewable sources could further enable defect-free

fabrication of complex, shaped epoxy composites for structural composite applications.

5. References
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Mark, H. F., Ed. Encyclopedia of Polymer Science and Technology, Wiley-Interscience: New York, 2004. Czub, P., Macromolecular Symposia 242, 60 2006. Varley, R. J.; Tian, W., Polymer International 53, 69 2004. Raquez, J. M.; Delglise, M.; Lacrampe, M. F.; Krawczak, P., Progress in Polymer Science 35, 487 2010. Cheng, X.; Chen, Y.-X.; Du, Z.-L.; Zhu, P.-X.; Wu, D.-C., Journal of Applied Polymer Science 119, 3504 2011. Tan, S. G.; Chow, W. S., Polymer-Plastics Technology and Engineering 49, 1581 2010. Xua, J. X.; Liu, Z. L.; Erhan, S. Z., Journal of the American Oil Chemists' Society 81, 813 2004. May, C. A., Ed. Epoxy Resins Chemistry and Technology, Marcel Dekker, Inc.: New York, 1988. Raghavachar, R.; Sarnecki, G.; Baghdachi, J.; Massingill, J., Journal of Coatings Technology 72, 125 2000. Chandrashekhara, K.; Sundararaman, S.; Flanigan, V.; Kapila, S., Materials Science and Engineering: A 412, 2 2005. Gupta, A. P.; Ahmad, S.; Dev, A., Polymer Engineering & Science 51, 1087 2011. Gelb, L.; Ault, W.; Palm, W.; Witnauer, L.; Port, W., Journal of the American Oil Chemists' Society 37, 81 1960. Czub, P., Macromolecular Symposia 245-246, 533 2006. Kar, S.; Banthia, A. K., Materials and Manufacturing Processes 19, 459 2004. Miyagawa, H.; Misra, M.; Drzal, L. T.; Mohanty, A. K., Polymer Engineering & Science 45, 487 2005. Jin, F.-L.; Park, S.-J., Materials Science and Engineering: A 478, 402 2008. Tan, S. G.; Chow, W. S., J Therm Anal Calorim 101, 1051 2010. La Scala, J.; Wool, R. P., Polymer 46, 61 2005. Meier, M. A. R.; Metzger, J. O.; Schubert, U. S., Chemical Society Reviews 36, 1788 2007. Gan, L. H.; Ooi, K. S.; Goh, S. H.; Chee, K. K., Journal of Applied Polymer Science 46, 329 1992. Maerker, G.; Saggese, E.; Port, W., Journal of the American Oil Chemists' Society 38, 194 1961. Kester, E. B.; Gaiser, C. J.; Lazar, M. E., The Journal of Organic Chemistry 08, 550 1943. Altuna, F. I.; Espsito, L. H.; Ruseckaite, R. A.; Stefani, P. M., Journal of Applied Polymer Science 120, 789 2011. Boquillon, N.; Fringant, C., Polymer 41, 8603 2000. Stevens, M. P., Polymer Chemistry : An Introduction; Oxford University Press 1999. Lu, P.; University of Missouri-Rolla, Rolla: 2001. Liu, Z.; Doll, K. M.; Holser, R. A., Green Chemistry 11, 1774 2009. Liu, Z.; Erhan, S. Z., Journal of the American Oil Chemists' Society 87, 437 2009. Earls, J. D.; White, J. E.; Lpez, L. C.; Lysenko, Z.; Dettloff, M. L.; Null, M. J., Polymer 48, 712 2007. Park, S.-J.; Jin, F.-L.; Lee, J.-R., Macromolecular Rapid Communications 25, 724 2004. Gerbase, A.; Petzhold, C.; Costa, A., Journal of the American Oil Chemists' Society 79, 797 2002. Tanrattanakul, V.; Saithai, P., Journal of Applied Polymer Science 114, 3057 2009. Miyagawa, H.; Mohanty, A. K.; Misra, M.; Drzal, L. T., Macromolecular Materials and Engineering 289, 629 2004. Miyagawa, H.; Misra, M.; Drzal, L. T.; Mohanty, A. K., Journal of Polymers and the Environment 13, 87 2005.

15