Thermal degradation kinetics of poly(O,O‐diethyl‐O‐allylthiophosphate‐co‐acrylonitrile) in nitrogen

The nonisothermal degradation kinetics of the copolymer poly(O,O‐diethyl‐O‐allylthiophosphate‐co‐acrylonitrile), which was synthesized with O,O‐diethyl‐O‐allylthiophosphate and acrylonitrile, were studied by thermogravimetry/derivative thermogravimetry techniques. The kinetic parameters, including the activation energy and the pre‐exponential factor of the copolymer degradation process, were calculated by the Kissinger and Flynn–Wall–Ozawa methods. The thermal degradation mechanism of the copolymer was also studied with the Satava–Sestak method. The results show that the activation energies were 138.17 kJ/mol with the Kissinger method and 141.63 kJ/mol with the Flynn–Wall–Ozawa method. The degradation of the copolymer followed a kinetic model of a phase boundary reaction and the kinetic equation could be expressed as G(α) = 1 ? (1 ? α)⁴ [where G(α) is the integral function of conversion and α is the extent of conversion of the reactant decomposed at time t]. The reaction order was 4. © 2009 Wiley Periodicals, Inc. J Appl Polym Sci, 2010     

Synthesis of polystyrene modified with fluorine atoms: Monomer reactivity ratios and thermal behavior

Homo‐ and copolymers of 4‐fluorostyrene (FSt) and styrene (St) were synthesized with different feed ratios using free radical bulk polymerization with azobisisobutyronitrile (AIBN) as initiator. It yielded series of (co)polymers with various amounts of included FSt, P(St‐co‐FSt) (5–50 mol%) and PFSt. The effect of the initiator concentration on the molecular weights of the homopolymers, that is, PSt and PFSt was investigated. Copolymer compositions were determined by nuclear magnetic resonance spectroscopy. The relative reactivity ratios of both comonomers were determined by applying the conventional linearization methods of Jaacks (J), Finemann–Ross (F–R), inverted Finemann–Ross (IF–R), and Kelen‐Tüdos (K–T). The reactivity ratios values of St and FSt obtained from J plot are 1.06 and 0.84, F–R plot are 1.18 and 1.06, IF–R 1.01 and 0.86, and K–T plot 1.04 and 0.88, respectively. Thermal properties of prepared (co)polymers, that is, glass transition temperature (T_g) and thermal stability, were determined from differential scanning calorimetry and thermogravimetrical measurements. The lack of significant influence of FSt comonomer content on T_g of (co)polymers was observed. Additionally, the thermal degradation kinetics of obtained PSt and PFSt was studied by thermogravimetric analysis. Kinetic parameters such as the thermal decomposition activation energy (E) and frequency factor (A) were estimated by Ozawa model [E(O) and A(O), respectively] and Kissinger model [E(K) and A(K), respectively]. The activation energy and the frequency factor of PFSt (253 kJ/mol) were higher than PSt (236 kJ/mol). The resulting activation energies estimated using the two methods were quite close. POLYM. ENG. SCI., 54:1170–1181, 2014. © 2013 Society of Plastics Engineers     

Evaluation of nonisothermal crystallization kinetic models for linear poly(phenylene sulfide)

The nonisothermal crystallization kinetics of linear Poly(phenylene sulfide) (PPS) was studied with differential scanning calorimetry. Ozawa theory, Jeziorny model, and Mo equation were applied to describe the crystallization kinetics and to determine the crystallization parameters and mechanism of the linear PPS resin. The crystallization activation energies were also calculated using Kissinger formula and Flynn‐Wall‐Ozawa equation, respectively. According to the Ozawa model, it is found that instantaneous nucleation takes place during crystallization of PPS; the Ozawa exponent m is 3 in initial stage of crystallization; as the crystallization temperature decreases, the value of m reduces, and the growth rate of crystal almost keeps a constant. The Avrami exponent n obtained from Jeziorny model fluctuate around 1.84. Based on the Jeziorny model, the crystallization rate increases with increasing the cooling rate, but it does not change any longer when the cooling rate rise to a certain value. Mo equation also exhibits great advantages in treating the nonisothermal crystallization kinetics of PPS. The activation energy E of nonisothermal crystallization process of PPS is calculated to be −162.73 kJ/mol by the Kissinger formula, and the mean value of E determined by Flynn‐Wall‐Ozawa equation is −152.40 kJ/mol. © 2011 Wiley Periodicals, Inc. J Appl Polym Sci, 2011     

Curing kinetics and structural changes of a of di[(N‐m‐acetenylphenyl) phthalimide] ether/[(methyl) diphenylacetylene] silane copolymer

Differential scanning calorimetry, Fourier transform infrared (FTIR) spectroscopy, and ¹³C‐NMR were used to characterize the curing kinetics and structural changes of a copolymer of di[(N‐m‐acetenylphenyl) phthalimide] ether (DAIE) and [(methyl) diphenylacetylene] silane (MDPES). The results show that the apparent activation energy (E) and reaction order (n) calculated according to the Kissinger method were nearly the same as those calculated according to the Ozawa method. E was 160.4 kJ/mol and n was 0.96 with the Kissinger method, and E was 158.1 kJ/mol and n was 0.95 with the Ozawa method. The FTIR and solid‐state ¹³C‐NMR results also indicate that with increasing curing temperature, the peaks assigned to Si? H and C?C bonded to phenylene carbons decreased, broadened, and finally vanished, whereas the peaks assigned to the C?C carbons and phenyl carbons increased and broadened. Crosslinking reactions in the curing of the DAIE/MDPES copolymer were possible due to the hydrosilylation reaction and the Diels–Alder reaction. © 2006 Wiley Periodicals, Inc. J Appl Polym Sci 100: 2126–2130, 2006     

Effect of particle size on thermal decomposition and devolatilization kinetics of melon seed shell

Thermogravimetric analysis (TGA) and devolatilization kinetics of melon seed shell (MSS) at different particle sizes (150?µm and 500?µm) and at different heating rates (10, 15, 20, and 25?°C/min) were investigated with the aid of TGA. The results of the TGA analysis show that the TGA curves corresponding to the first and third stages for 150?µm particle sizes exhibited some bumps that developed at the first and third stages of pyrolysis. It was also observed that at constant heating rate, the maximum peak temperature increases as the particle sizes increase from 150 to 500?µm, whereas 500?µm particle sizes exhibited higher peak temperatures compared to 150?µm particle sizes. The resulting TGA data were applied to the Kissinger (K), Kissinger–Akahira–Sunose (KAS) and Flynn–Wall–Ozawa (FWO) methods and kinetic parameters (activation energy, E and frequency factor, A) were determined. The E and A obtained using K method were 74.27?kJ mol^?1 and 3.84?×?10⁵?min^?1 for 150?µm particle size, whereas for 500?µm particle size were 97.12?kJ mol^?1 and 3.74?×?10⁷?min^?1, respectively. However, the average E and A obtained using KAS and FWO methods were 82.35?kJ mol^?1, 1.29?×?10⁷?min^?1, and 88.50?kJ mol^?1, 1.32?×?10⁷?min^?1 for 150?µm particle sizes. While for 500?µm particle sizes, the E and A were 108.46?kJ mol^?1, 3.14?×?10⁹?min^?1, and 113.05?kJ mol^?1, 7.56?×?10⁹?min^?1, respectively. It was observed that E and A calculated from FWO and KAS methods were very close and higher than that obtained by K method. It was observed that the minimum heat required for the cracking of MSS particles into products is reached later at higher peak temperatures since the heat transfer is less effective as they are at lower peak temperatures.     

PE/PE‐g‐MAH/Org‐MMT nanocomposites. II. Nonisothermal crystallization kinetics

The nonisothermal crystallization kinetics of high‐density polyethylene (HDPE) and polyethylene (PE)/PE‐grafted maleic anhydride (PE‐g‐MAH)/organic‐montmorillonite (Org‐MMT) nanocomposite were investigated by differential scanning calorimetry (DSC) at various cooling rates. Avrami analysis modified by Jeziorny, Ozawa analysis, and a method developed by Liu well described the nonisothermal crystallization process of these samples. The difference in the exponent n, m, and a between HDPE and the nanocomposite indicated that nucleation mechanism and dimension of spherulite growth of the nanocomposite were different from that of HDPE to some extent. The values of half‐time (t_1/2), K(T), and F(T) showed that the crystallization rate increased with the increase of cooling rates for HDPE and composite, but the crystallization rate of composite was faster than that of HDPE at a given cooling rate. Moreover, the method proposed by Kissinger was used to evaluate the activation energy of the mentioned samples. It was 223.7 kJ/mol for composite, which was much smaller than that for HDPE (304.6 kJ/mol). Overall, the results indicated that the addition of Org‐MMT and PE‐g‐MAH could accelerate the overall nonisothermal crystallization process of PE. © 2004 Wiley Periodicals, Inc. J Appl Polym Sci 91: 3054–3059, 2004     

A study on the curing kinetics of epoxycyclohexyl polyhedral oligomeric silsesquioxanes and hydrogenated carboxylated nitrile rubber by dynamic differential scanning calorimetry

A new organic–inorganic hybrid material was prepared through reactive blending of hydrogenated carboxylated nitrile rubber (HXNBR) with epoxycyclohexyl polyhedral oligomeric silsesquioxanes (epoxycyclohexyl POSS). The structure of the composite was characterized by Fourier transform infrared spectroscopy (FTIR) and solid‐state ¹³C Nuclear Magnetic Resonance spectra (solid‐state ¹³C‐NMR). The differential scanning calorimetry (DSC) at different heating rates was conducted to investigate the curing kinetics. A single overall curing process by an nth‐order function (1 ? α)ⁿ was considered, and multiple‐heating‐rate models (Kissinger, Flynn–Wall–Ozawa, and Crane methods) and the single‐heating‐rate model were employed. The apparent activation energy (E_a) obtained showed dependence on the POSS content and the heating rate (β). The overall reaction order n was practically constant and close to 1. The isoconversion Flynn–Wall–Ozawa method was also performed and fit well in the study. With the single‐heating‐rate model, the average E_a for the compound with a certain POSS content, 66.90–104.13 kJ/mol was greater than that obtained with Kissinger and Flynn–Wall–Ozawa methods. Furthermore, the calculated reaction rate (dα/dt) versus temperature curves fit with the experimental data. © 2011 Wiley Periodicals, Inc. J Appl Polym Sci, 2012     

Thermal Behavior and Thermolysis Kinetics of the Explosive Trans‐1,4,5,8‐Tetranitro‐1,4,5,8‐Tetraazadecalin (TNAD)

Trans‐1,4,5,8‐Tetranitro‐1,4,5,8‐Tetraazadecalin (TNAD), a cyclic nitroamine, has been studied with regard to the kinetics and mechanism of thermal decomposition, using thermogravimetry (TG), IR spectroscopy, and pressure differential scanning calorimetry (PDSC). The IR spectra of TNAD have also been recorded, and the kinetics of thermolysis has been followed by non‐isothermal TG. The activation energy of the solid‐state process was determined by using the Flynn‐Wall‐Ozawa method. Compared with the activation energy obtained from the Ozawa method, the reaction mechanism of the exothermic process of TNAD was classified by the Coats‐Redfern method as a nucleation and nuclear growth (Avrami equation 1) chemical reaction (α=0.30–0.60) and a 2D diffusion (Valensi equation) chemical reaction (α=0.60–0.90). E_a and ln A were established to be 330.14 kJ mol⁻¹ and 29.93 (α=0.30–0.60) or 250.30 kJ mol⁻¹ and 21.62 (α=0.60–0.90).     

Influence of azobenzene units on imidization kinetic of novel poly(ester amic acid)s and polymers properties before and after cyclodehydration

In this article, the imidization reaction kinetic of novel poly(ester amic acid)s with azobenzene units as side groups was studied by dynamic experiments by means of differential scanning calorimetry. Polymers differ in the number of chromophore moieties in their repeating unit and position in which azobenzene group is attached to the polymer chain. The kinetic parameters of poly(ester amic acid)s conversion to poly(ester imide)s was compared with data calculated for parent polymer, that is, without azobenzene groups. For the first time to our knowledge, the imidization kinetic of polymers with side azobenzene groups was studied. Kinetic parameters, such as the activation energy and frequency factor were estimated with the by Ozawa model [(E(O) and A(O)), respectively] and Kissinger model [(E(K) and A(K), respectively]. The values of activation energy determined with both models were in the range 167.1–198.3 kJ/mol. The lowest activation energy of imidization reaction exhibited polymer in which azobenzene units were placed between amide linkages. Polymers were characterized by FTIR, ¹H‐NMR, X‐ray, and UV–vis methods. The glass transition temperature of resultant poly(ester imide)s was in the range of 217–237°C. The presence of chromophore units slightly decreased T_g and significantly improved their solubility and optical properties. © 2010 Wiley Periodicals, Inc. J Appl Polym Sci, 2010     

Nonisothermal crystallization behavior of LLDPE/glass fiber composite

The nonisothermal crystallization behavior of linear low-density polyethylene (LLDPE)/glass fiber (GF) composite was investigated by differential scanning calorimetry (DSC). It was observed that the crystallization temperature peak (T_p) of LLDPE composite containing 5.0 wt % GF (LLDPE/GF5) was higher than that of the pure LLDPE at various cooling rates. The half-time of crystallization (t_1/2) of LLDPE/GF5 composite was shortened under the effect of GF. The nonisothermal crystallization kinetics of LLDPE and LLDPE/GF5 composite were analyzed through the Avrami, Ozawa, and Mo equations. The results indicated that the data of the nonisothermal crystallization for LLDPE and LLDPE/GF5 composite calculated based on the Ozawa equation did not have the good linear relationship, but the nonisothermal crystallization behaviors of LLDPE and LLDPE/GF5 composite could be successfully described by the modified Avrami and Mo methods. The crystallization rate Z_c of the modified Avrami parameter of LLDPE/GF5 composite was higher than that of pure LLDPE at the same cooling rate. The Mo parameter F(T) of LLDPE/GF5 composite was lower than that of LLDPE at the same degree of crystallinity. Through the Kissinger equation, the activation energies E_d of LLDPE and LLDPE/GF5 composite were evaluated, and their values were 312.3 and 251.2 kJ/mol, respectively. © 2008 Wiley Periodicals, Inc. J Appl Polym Sci, 2008     

Transformation of propane and CO to carboxylic acid and ester by highly active CaCl2 catalyst

The functionalization reaction of propane with CO to afford carboxylic acids and an ester by CaCl₂ catalyst in the presence of K₂S₂O₈ and CF₃COOH has been studied. The reaction gave isobutyric acid as the main product and n-butyric acid and isopropyl trifluoroacetate as by-products. Atmospheric pressure of propane underwent the reaction with 30 atm of CO pressure at 80 °C for 24 h, giving about 95% total yield based on propane. The activation and thermodynamic parameters have been determined to be E_a = 130.3, 138.0 and 153.8 kJ/mol; A = 7.14 × 10¹³, 5.83 × 10¹⁴ and 5.80 × 10¹⁶ 1/s; ΔH ^‡ = 128.0, 134.7 and 150.5 kJ/mol; ΔS ^‡ = 10.3, 28.6 and 66.8 J/mol K and ΔG^‡ ₂₅₃ = 124.4, 123.9 and 126.9 kJ/mol for the products of isobutyric acid, n-butyric acid and isopropyl trifluoroacetate, respectively. This revised version was published online in July 2006 with corrections to the Cover Date.     

Sillimanite‐mullite transformation observed in synchrotron X‐ray diffraction experiments

High‐resolution synchrotron powder X‐ray diffraction (XRD) experiments were conducted to clarify the transformation of sillimanite to mullite (mullitization) and determine the mullitization temperature (T_c). We were able to distinguish sillimanite and mullite in the XRD patterns, despite their very similar crystallographic parameters, and to detect the appearance of small mullite peaks among sillimanite peaks. Analysis of the Johnson‐Mehl‐Avrami (JMA) equation for mullitization ratio (ζ) revealed that at temperatures T≥1240°C the mullitization had the same kinetics. The activation energy E at T≥1240°C obtained from the Arrhenius plot was 679.8 kJ mol^?1. In analysis using a time‐temperature‐transformation diagram for mullitization, a mullitization curve of ζ=1% can be described as where t is time, n is a reaction‐mechanism‐dependent parameter determined as 0.324 by JMA‐analysis, k₀ is the frequency factor, E_A is the activation energy for atomic diffusion, and represents the activation energy for nucleation. The results of fitting the data to this equation were T_c=1199°C, A=3.9×10⁶ kJ mol^?1 K^?2, E_A=605 kJ mol^?1, and k₀=3.65×10¹⁵. We conclude that the boundary between sillimanite and mullite+SiO₂ in the phase diagram is ~1200°C.     

Cure kinetics of a glass/epoxy prepreg by dynamic differential scanning calorimetry

Dicyandiamide (DICY)‐cured epoxy resins are important materials for structural adhesives and matrix resins for fiber‐reinforced prepregs. Dynamic differential scanning calorimetry (DSC) with heating rates of 2.5, 5, 10, and 15°C/min was used to study the curing behavior of the epoxy prepreg Hexply 1454 system, which consisted of diglycidyl ether of bisphenol A, DICY, and Urone reinforced by glass fibers. The curing kinetic parameters were determined with three different methods and compared. These were the Kissinger, Ozawa, and Borchardt–Daniels kinetic approaches. The lowest activation energy (76.8 kJ/mol) was obtained with the Kissinger method, whereas the highest value (87.9 kJ/mol) was obtained with the Borchardt–Daniels approach. The average pre‐exponential factor varied from 0.0947 × 10⁹ to 2.60 × 10⁹ s⁻¹. The orders of the cure reaction changed little with the heating rate, so the effect of the heating rate on the reaction order was not significant. It was interesting that the overall reaction order obtained from all three methods was nearly constant (≅2.4). There was good agreement between all of the methods with the experimental data. However, the best agreement with the experimental data was seen with the Ozawa kinetic parameters, and the most deviation was seen with the Borchardt kinetic parameters. © 2010 Wiley Periodicals, Inc. J Appl Polym Sci, 2011     

Halogeno‐modified polystyrene: monomer reactivity ratios,thermal behaviour and flammability

Chemical modification based on incorporation of flame retardants into polymer backbones was used in order to reduce flammability of polystyrene (PSt). The halogeno‐substituted styrenes: 4‐chlorostyrene (ClSt), 4‐bromostyrene (BrSt) and 2,3,4,5,6‐pentafluorostyrene (5FSt) were applied as reactive flame retardants. Homo‐ and copolymers of these halogeno‐substituted styrenes and styrene (St) were synthesized with various feed ratios using free radical bulk polymerization with azobisisobutyronitrile as a initiator. This yielded series of (co)polymers with various amounts of included ClSt, BrSt and 5FSt (5–50 mol% of modified St). Copolymer compositions were determined using ¹H NMR spectroscopy. The relative reactivity ratios of the used comonomers were determined by applying conventional linearization methods. The Jaacks (J) method was used for systems including BrSt and ClSt monomers whereas the Fineman–Ross method was additionally used to confirm the values of reactivity ratios of St–5FSt. The reactivity ratios of comonomer pairs obtained from J plots were 0.75 and 0.38 (St–ClSt), 1.65 and 0.46 (St–BrSt), 0.44 and 0.42 (St–5FSt). Glass transition temperature and thermal stability of obtained (co)polymers were determined using differential scanning calorimetry and thermogravimetric analysis (TGA), respectively. The thermal degradation kinetic of PSt, PClSt, PBrSt and P5FSt was studied applying TGA. Kinetic parameters such as thermal decomposition activation energy (E) and frequency factor (A) were estimated using Ozawa and Kissinger models. The resulting activation energies estimated using these two methods were quite close. The values of activation energy (kJ mol^?1) increased in the following order: PClSt (E(O) = 216.1) < PSt (E(O) = 219.9) < PBrSt (E(O) = 224.7) < P5FSt (E(O) = 330.9). A pyrolysis combustion flow calorimeter was applied as a tool for assessing the flammability of the synthesized (co)polymers. © 2014 Society of Chemical Industry     

Nonisothermal cure of bisphenol epoxy resin with a nonlinear aliphatic polyamine hardener

The amino terminated polypropylenimine dendrimer (DAB‐dendri‐(NH₂)₄) was employed as a new nonlinear aliphatic curing agent for diglycidyl ether of bisphenol A (DGEBA). Nonisothermal curing reaction kinetics of DGEBA/DAB was investigated with a differential scanning calorimeter (DSC). The apparent reaction activation energy E_a is about 56.7 kJ/mol determined using the Kissinger equation, and a two‐parameter (m, n) autocatalytic model ([icirc]Sesták–Berggren equation) was confirmed to be able to well simulate the reaction kinetics in the light of the Málek method. In addition, the relation between reaction activation energy E_a and curing degree α was obtained by applying model‐free isoconversional analysis with the Kissinger‐Akahira‐Sunose (KAS) method. As α increases, E_a reduced quickly from >80 kJ/mol to ≈60 kJ/mol up to a ≈ 15%, then decreased slowly to 55 kJ/mol till a ～ 75%, and finally dropped to 44 kJ/mol at full conversion. © 2011 Wiley Periodicals, Inc. J Appl Polym Sci, 2011     

Nonisothermal crystallization behavior of a luminescent conjugated polymer,poly(9,9‐dihexylfluorene‐alt‐2,5‐didodecyloxybenzene)

The nonisothermal crystallization kinetics of poly(9,9‐dihexylfluorene‐alt‐2,5‐didodecyloxybenzene) (PF6OC12) from the melt were investigated using differential scanning calorimetry under different cooling rates. Several analysis methods were used to describe the nonisothermal crystallization behavior of PF6OC12. It was found that the modified Avrami method by Jeziorny was only valid for describing the early stage of crystallization but was not able to describe the later stage of PF6OC12 crystallization. Also, the Ozawa method failed to describe the nonisothermal crystallization behavior of PF6OC12. However, the method developed by combining the Avrami and Ozawa equations could successfully describe the nonisothermal crystallization kinetics of PF6OC12. According to the Kissinger method, the activation energy was determined to be 114.9 kJ mol^?1 for the nonisothermal melt crystallization of PF6OC12. Copyright © 2006 Society of Chemical Industry     

Initial rate study on the vapor-phase condensation of aniline to diphenylamine

The vapor-phase condensation of aniline to diphenylamine has been studied in a downflow differential reactor at 370–426°C and under atmospheric pressure using H-151 ALCOA catalyst. An initial rate equation based on the Langmuir-Hinshelwood model, i.e. surface reaction control with a similar dual site mechanism, has been proposed: r_o = (k₁K_A²p_Ao²)/ (1 + K_Ap_Ao)² The effects of different mass transfer resistances, of longitudinal and of radial dispersion have been taken into consideration. The rate constant value at 404°C is 1.488 × 10^?7 (k mol)/ (s) (kg cat) [5.358 × 10^?4 (mol)/ (h) (g cat)] and the activation energy is calculated to be 123.3 kJ/mol [29.45 k cal/mol] in the working temperature range.     

Kinetic parameter determination of vegetable oil oxidation under Rancimat test conditions

In this research, five different vegetable oils were oxidized at four different temperatures (373, 383, 393, and 403 K) under Rancimat test conditions. An increasing rate of oxidation could be observed as temperature increased. The natural logarithms of the kinetic rate constant (k value) varied linearly with respect to temperature, with the temperature coefficients (T_Coeff) ranging from 6.95×10^–2 to 7.40× 10^–2 K^–1 for the vegetable oils. On the basis of the Arrhenius equation and the activated complex theory, frequency factors (A), activation energies (E_a), Q₁₀ numbers, activation enthalpies (ΔH⁺⁺), and activation entropies (ΔS⁺⁺) for oxidative stability of the vegetable oils were calculated. The A, E_a, Q₁₀, ΔH⁺⁺, and ΔS⁺⁺ values for the vegetable oils ranged from 6.38×10³ to 28.03×10³ h^–1, from 86.86 to 92.42 kJ/mol, from 2.08 to 2.18, from 83.64 to 89.20 kJ/mol, and from –116.66 to –104.35 J/mol K, respectively.     

Kinetics of thermal degradation of poly(p‐phenylene benzobisoxazole)

The kinetics of thermal degradation of poly (p-phenylen benzobisoxazole) (PBO) were studied by thermogravimetric analysis (TG) in dynamic nitrogen gas at four different heating rates: 5, 10, 15, 20°C/min. The activation energy calculated by Kissinger Method was 352.19 kJ/mol, and the mean value of activation energies evaluated by Flynn-Wall-Ozawa Method was 338.32 kJ/mol. The degradation kinetic model of PBO followed the mechanism of random scission of weak bonds of PBO molecule and impact of the active groups obtained from the broken bonds, Mampel Power equation with integral form G(α) = α^3/2 and differential form . And the mathematical equation of kinetic compensation effect was ln A = 0.1365 E_a − 1.4102. © 2006 Wiley Periodicals, Inc. J Appl Polym Sci 103: 3675–3679, 2007