Polymer Degradation and Stability 91 (2006) 2960e2967 www.elsevier.

com/locate/polydegstab

Thermal degradation of polycarbonate, poly(vinyl acetate) and their blends

lu b,* Tamer Uyar a, Alan E. Tonelli a, Jale Hacalog
a

Fiber and Polymer Science Program, College of Textiles, Campus Box 8301, North Carolina State University, Raleigh, NC 27695-8301, USA b Department of Chemistry, Middle East Technical University, Ankara, 06531, Turkey Received 5 April 2006; received in revised form 8 August 2006; accepted 14 August 2006 Available online 16 October 2006

Abstract We have recently developed a novel approach for intimately mixing thermodynamically incompatible polymers, which utilizes the formation of inclusion compounds (ICs) formed with host cyclodextrins (CDs), followed by removal of CD and coalescence of the common guest polymers into a blend. In this paper direct insertion probe pyrolysis mass spectrometry (DIP-MS) analyses of polycarbonate (PC), poly(vinyl acetate) (PVAc) and PC/PVAc blends, obtained by coalescence from their inclusion compounds formed with host g-CD (coalesced blend) and by coprecipitation (physical blend), have been performed. Variations in the thermal stabilities of the coalesced polymers were recorded both by TGA and DIP-MS and compared to the corresponding as-received polymers. It has been determined that for both coalesced and physical blends of PC/PVAc, CH3COOH formed by deacetylation of PVAc above 300  C, reacts with PC chains decreasing their thermal stability. This process was more effective for the physical blend, most likely due to enhanced diffusion of CH3COOH, produced by deacetylation of PVAc, into the PC domains, where it can further react producing low molecular weight PC fragments bearing methyl carbonate chain ends. 2006 Elsevier Ltd. All rights reserved.
Keywords: Polycarbonate; Poly(vinyl acetate); Blend; Cyclodextrin; Inclusion compound; Pyrolysis mass spectrometry

1. Introducton Interest in studying polymer blends has considerably increased due to their signicant industrial applications. Materials with improved characteristics are produced by blending two or more polymers in order to combine their properties for certain purposes. The conventional methods for producing polymer blends are melt blending, co-precipitation, and solvent casting [1]. When two or more polymer components are blended together, it is usual to obtain phase-separated materials, due to a small entropy and the generally unfavorable enthalpy of mixing when there are no specic interactions (hydrogen bonds, van der Waals and/or polar interactions, etc.) between the polymer components. In the case of poor mixing, the blends exhibit properties inferior to their component
* Corresponding author. E-mail address: jale@metu.edu.tr (J. Hacalo glu). 0141-3910/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.polymdegradstab.2006.08.028

polymers due to the poor degree of interfacial adhesion between the phase-separated components, which create a multiplicity of defects in the system [2]. In addition to the properties of each polymer component and their compositions, the eventual characteristics of a polymer blend depend most importantly on the miscibility of the component polymers in the blend [1e3]. The preparation of miscible polymer blends depends on many factors, such as solvent, temperature, weight/molar ratio of the components, molecular weight, and specic interactions [3]. Cyclodextrins (CDs), cyclic starch oligomers, are truncated molecular cones consisting of 6, 7, and 8 glucose units, and are named alpha (a-), beta (b-) and gamma (g-) cyclodextrins (CDs), respectively (see Fig. 1). Although the depth of the cav ), their cavity diities for the three CDs is the same (w7.9 A , respectively [4]. Because of ameters are w6, 8, and 10 A their unusual structure, CDs may host both polar and nonpolar guests, including small molecules as well as polymers [5e11].

T. Uyar et al. / Polymer Degradation and Stability 91 (2006) 2960e2967

2961

0.95 nm

0.78 nm

0.57 nm 0.78 nm

-CD

-CD

-CD

d
Polymer 1 Polymer 2

time RT organic solvent Inclusion Complexation CD Filter

Coalescing Process water Removal of CD with Blend of intimately mixed coalesced polymer chains in extended conformations

Polymer-CD-IC crystals

Fig. 1. (a) g-CD chemical structure; (b) approximate dimensions of a-, b- and g-CDs; schematic representation of (c) polymerecyclodextrineinclusion compound (polymereCDeIC); (d) polymereCDeIC formation, the coalescence process, and the coalesced polymer blend.

Our own research studies have shown that CDs may act as hosts in the formation of inclusion compounds (ICs) with various guest polymers [7e19]. PolymereCDeICs represent crystalline compounds obtained by threading of the CD molecules onto the guest polymer chains. The included polymers are conned to the narrow, continuous CD channels, and so are necessarily highly extended and segregated from neighboring polymer chains by the walls of the CD stacks (see Fig. 1c). We have shown that coalescence of guest polymers from their CDeIC crystals can result in a signicant improvement in their bulk physical properties caused by modication of their structures, morphologies and even conformations compared to those observed in their normally produced bulk samples. Recently, we have developed a novel approach for mixing thermodynamically incompatible polymers by rst obtaining a common polymereCDeIC. Subsequently, the guest polymers are coalesced from their common CDeIC crystals by removal of the CD host [15e19]. This method has successfully produced intimate blends of poly(3-caprolactone) (PCL)/ poly(L-lactic acid) (PLLA) [15], poly(ethylene terephthalate) (PET)/poly(ethylene 2,6-naphthalate) (PEN) [16], polycarbonate (PC)/poly(methyl methacrylate) (PMMA) [17,18], (PC)/ poly(vinyl acetate) (PVAc) [18], PMMA/PVAc [18] and the ternary blend PC/PMMA/PVAc [19]. These well-mixed binary blends achieved by coalescence from their common g-CDeICs exhibited thermal stabilities

and thermal degradation mechanisms that were distinct from those of the pure component polymers and their binary blends obtained by both solvent-casting and co-precipitation methods. The details of thermal degradation characteristics of coalesced and physical blends of PMMA/PVAc, and PC/ PMMA have been studied by direct pyrolysis mass spectrometry and discussed in our recent publications [18,20,21]. In the literature, studies on the PC/PVAc blend system are limited. Sato et al. reported that PC forms partially miscible blends with poly(vinyl alcohol) (PVA) and partially saponied PVAc due to the existence of side-chain hydroxyl groups, which improves their miscibility by intermolecular hydrogen bonding, but PVAc was found to be immiscible with PC [22]. We have recently shown that thermally stable and intimately mixed blends of PC/PVAc can be obtained by formation of and coalescence from their common IC with g-CD [18]. In this paper, the details of direct pyrolysis mass spectrometry analyses of PVAc, PC, and the binary PVAc/PC blend when coalesced (c-) from their g-CDeICs are investigated and their thermal behaviors and degradation mechanisms are reported. The thermal characteristics of c-PVAc, c-PC and the c-PVAc/PC blend were characterized by comparing their direct pyrolysis mass spectrometry results with those for g-CD, as-received PVAc and PC, PVAc/PC physical mixture, and their inclusion compounds.

2962

T. Uyar et al. / Polymer Degradation and Stability 91 (2006) 2960e2967

2. Expermental Details regarding the materials used, the formation of PVAc/PC polymereg-CDeinclusion compounds (polymere g-CDeICs), the coalescence of PVAc/PC polymer blend from their CDeICs (coalesced blend) and the preparation of binary blends of PVAc/PC by co-precipitation (physical blend) are given in our previous reports [18,14]. Thermal analyses of the samples were performed with a PerkineElmer Pyris 1 TGA instrument. The thermal decompositions of the samples were recorded between 25 and 650  C. The heating rate was 20  C/min and nitrogen was used as a purge gas. Our direct insertion probe pyrolysis mass spectrometry (DIP-MS) system consists of a 5973 HP quadruple mass spectrometer coupled to a JHP SIS direct insertion probe pyrolysis system. Samples (0.01 mg) were pyrolyzed in ared glass sample vials. The temperature was increased at a rate of 10  C/min and the scan rate was 2 scans/s. 3. Results and discussion The thermal degradation of PC takes place by several reaction pathways [23e25]. In the initial stage of pyrolysis, around 400e500  C, intramolecular exchange reactions, yielding cyclics, and hydrolysis reactions, generating phenolic end groups and CO2, and decarboxylaton of the carbonate groups, producing ether bridges, are active. Products due to methyl losses are also detected, as CeC bond dissociation of the isopropylidene linkage is energetically easiest in this temperature range [23]. The thermal degradation pathways of PC around 400e500  C, as given in the literature, are shown in Scheme 1. The pyrolysis mass spectra of as-received and coalesced PC samples were in accordance with the proposed mechanism and were both dominated by several peaks, among which the peaks at m/z 508, 405 and 213 Da were the most intense. The peak at 508 Da corresponds to cyclic dimer produced by an intramolecular exchange reaction (Scheme 1a). Loss of CO2 from cyclic carbonates generates the fragment ion having m/z 405 Da, whereas hydrolysis reactions play a role in the production of the 213 Da fragment ion (Scheme 1b and c). For coalesced PC, low molecular weight PC based products

were detected at lower temperatures indicating a decrease in thermal stability for the coalesced PC samples. The decrease in thermal stability was attributed to loss of polycarbonate chain ends and/or units strongly interacting with the small amount of remaining g-CD; the hydroxides of the degradation products of g-CD interact with the carbonate linkages of the PC chains and generate PC chains bearing OH end groups [21]. It is known that PVAc degrades in two stages, around 360 and 440  C, corresponding to deacetylation and disintegration of the polyolenic backbone, respectively [26,27]. Pyrolysis of the coalesced sample indicated signicant decrease in relative yields of polyolenic products. Crosslinking of the polyolenic backbone with the decomposition products of reminant g-CD, involving two or more OH groups generated during the rst thermal degradation stage, has been proposed [20]. The total ion current (TIC) curves, the variation of total ion yield as a function of temperature, of physical and coalesced binary blends of PC/PVAc are shown in Fig. 2 together with those for the homopolymers for comparison. The pyrolysis mass spectra recorded at the shoulders and at the TIC maxima are also included in this gure. TGA analysis of binary blends revealed two degradation stages at 369 and 538  C for the physical mixture and at 377 and 551  C for the coalesced blend indicating slight changes compared to the corresponding homopolymers [14,18]. The rst maxima present in the TIC curves of the physical and coalesced PC/PVAc blends at 370 and 380  C, respectively, are quite similar to the Td values in the corresponding TGA curves (Fig. 2d.i and d.ii). For the coalesced sample, a shoulder around 325  C and a second maximum on the high temperature tail at 440  C can also be seen. As the maximum attainable temperature with our present pyrolysis system is 445  C, the highest temperature weight losses detected in TGA curves could not be studied. Analyses of the pyrolysis mass spectra of the coalesced binary blend indicated evolution of g-CD around 325  C together with some PC based fragments. Actually, for all the coalesced samples, weak shoulders around 300e360  C, are detected and associated with evolution of g-CD based products indicating that all the samples contained small amounts of remnant g-CD [14,18,20e22].

a
O C O O

intramolecular exchange
O n O C O O n

b
O C O O

CO2 elimination (1-3 shift)

O

c
O O CO

+ H2O O - CO2 OH

Scheme 1.

T. Uyar et al. / Polymer Degradation and Stability 91 (2006) 2960e2967

43 d.ii.c-PC/PVAc 213

2963
213

325 C

43 213

380 C

440 C

91 50 250 450

405 508 250 450

100

300

500

50

370 C
d.i.p-PC/PVAc 43 50 213 270 250 450

c.i. c-PC

18 213 60

310 C
213 18 405 508 333

405 333

440 C
508 449

100 c.i.PC

250

400

100

250 400 405 213

440 C
508

intensity

333 100 250 400

b.ii. c-PVAc

60 43 73

340 C

43 60

370 C

18 91 43 141

440 C

145 b.i. PVAc 50 100 150 50 43 60

78 100

128 100 43 91 141 215 200 300

360 C

440 C

78 50 100

141 100 200 300 400

a. -CD 60 73

310 C

73

370 C
85 213 145 163

97

145

250

350

50

100

150

50

100

150

Temperature C
Fig. 2. TIC curves (left) and pyrolysis mass spectra (right) recorded at the maximum of the peaks and shoulders in the TIC curves of (a) g-CD, (b) as-received PVAc, (c) coalesced PVAc, (d) as-received-PMMA, (e) coalesced PMMA, (f) PMMA/PVAc physical mixture, and (g) coalesced PMMA/PVAc blend.

The relative intensities of PC based product peaks were noticeably changed in the pyrolysis mass spectra of both blends. The increase in the relative intensity of the 213 Da product peak was the most signicant. Another point that should be noted was the decrease in the relative intensities of PVAc based fragments in the case of the physical blend. In Fig. 3I and II single ion pyrograms of some characteristic thermal degradation products of as-received and coalesced

PC, PVAc, and their physical and coalesced blends are shown. The evolution proles of thermal degradation products bearing OH groups in the coalesced blend showed trends quite similar to those of the coalesced homopolymers. Yet, the yield of the 213 Da product increased signicantly around 380  C. On the other hand, for the physical blend, the changes in the evolution proles of the PC based products were much more drastic; not only OH bearing products, but all PC based products started to

2964

T. Uyar et al. / Polymer Degradation and Stability 91 (2006) 2960e2967

I. Polycarbonate based products c-PC a b c-PC/PVAc blend

c PC/PVAc phy.blend d

PC

x27 x67.2 x27 x2.8

m/z=508

x1.4

x30.8

x13

x1

m/z=405

x2.1 x12 x37 x105 m/z=482

x1.9 x5.9 x10 x4.6

m/z=91

x1 x1
250 350 450 250 350 450

x1
250 350 450

x1.6
250 350 450

m/z=213

II. PVAc based products

c-PVac

b c-PC/PVAc blend

c PC/PVAc phy. blend

PVac

x26.7

x5.9

x10

x14.7

m/z=91

x1
250 350 450

x1.4
250 350 450

x7
250 350 450

x1
250 350 450

m/z=43

Temperature C
Fig. 3. Evolution proles of I. PC based products recorded during the pyrolysis of (a) coalesced PC, (b) coalesced PC/PVAc blend, (c) physical PC/PVAc blend, and (d) as-received-PC, and II. PVAc based products from (a) coalesced PVAc, (b) coalesced PC/PVAc blend, (c) physical PC/PVAc blend, and (d) as-received-PVAc.

appear in the pyrolysis mass spectra at lower temperatures. The maximum in the pyrograms of OH bearing products was recorded at 370  C. Unlike all those recorded from other PC containing samples, the homopolymers and coalesced blend, the evolution proles of 508 and 405 Da fragments had a maximum around 410  C. Furthermore, compared to all other samples, the relative intensities of PVAc based products diminished noticably, including the ratio of CH3COO and CH3CO fragments, for the physical blend. Two different types of chemical reactions occurring in PC/ PVAc blends may be proposed by analogy to PC/PMMA blends [21, 28e30]. Unlike the coalesced PMMA/PC blend,

however, the rst possibility, the ester-ester exchange reaction between PC and PVAc that may lead to in situ formation of a graft copolymer is not very likely, because the methyl ester groups of PMMA are replaced by polyvinylester groups [21]. If a graft copolymer was produced, PC chains bearing C2H3 groups should be evolved during the pyrolysis. The lack of product peaks bearing C2H3 groups for both PC/PVAc blends is in accordance with this expectation, indicating that no estereester exchange reaction between PC and PVAc takes place during thermal degradation. Another possibility is an exchange reaction between carbonates of PC and CH3COOH generated during the thermal

T. Uyar et al. / Polymer Degradation and Stability 91 (2006) 2960e2967

CH3 C-O H O O C O O C O O O C O CH3 OH

2965

-CO2 O C O O O C OH n O O C O O n OH

Scheme 2.

degradation of PVAc chains above 300  C. If this is the case, then due to the diffusion of CH3COOH at the interface or even into the PC domains, low molecular weight PC chains bearing OH and acetate chain ends, as shown in Scheme 2, may be produced. If such a reaction takes place, the relative intensities of CH3COOH based peaks should change; the relative intensity of COOH, CO2, and CH3COOH peaks should decrease with respect to the characteristic product peaks of the polyolenic backbone. On the other hand, products bearing methyl carbonyl groups should be detected. Possible degradation products due to the exchange reactions between PVAc or CH3COOH and carbonate functional groups of PC in PC/PVAc blends are listed in Table 1. For both blends signicant increase in relative intensities of 228, 482 and 213, 467 Da products was noted, indicating that generation of PC chains bearing OH end groups was enhanced in the presence of PVAc. Furthermore, generation of 270 and 312 Da products was detected from both blends, being more signicant for the physical blend. In Fig. 4, evolution proles of some characteristic products for the physical and coalesced blends are shown. The CH3COOH based products, namely CH3CO, COOH, CO2, and CH3COOH, were quite intense for the coalesced blend and all showed identical evolution proles with maxima at 370  C. Shoulders around 370  C, corresponding to the maximum in the evolution proles of CH3COOH based products, were present in the evolution proles of PC based products bearing OH or OCOCH3. On the other hand, the evolution of 508 and 405 Da products showed quite similar trends with those from as-received and coalesced PC samples. However, unlike the homopolymers, the 508 Da product due to straight chain dimer followed similar trends with the OH or OCOCH3 bearing PC based products.

Although the composition of both blends (PVAc:PC repeating mole ratio) was identical, a signicant decrease in the relative intensities of CH3COOH based products was noted for the physical blend. The decrease was more than 24-fold for COOH and CH3COOH and about 7-fold for CH3CO. For the physical blend, superimposable maxima in the evolution proles of CH3CO and 213, 270, 312, 467 Da products at 370  C were detected. Evolution of all other CH3COOH based products occurred at slightly lower temperatures; the maximum in the single ion pyrograms was observed at 360  C for COOH and CH3COOH. On the other hand, an increase in the relative intensities of 270 and 312 Da products associated with OCOCH3 bearing PC chains was detected, as can be seen in Fig. 4. The evolution of 508 and 405 Da products also shifted to lower temperatures giving maximum yield around 390  C, while, as in the case of the coalesced blend, the acyclic dimer showed similar trends with OH and OCOCH3 bearing products. Thus, it can be concluded that the CH3COOH generated during the thermal degradation of PVAc chains reacts with PC decreasing its thermal stability. The variations in relative yields of CH3COOH and OCOCH3 bearing PC chains indicate that the diffusion of CH3COOH into the PC domains was much more effective for the physical blend. On the surface, this appears to be inconsistent with the intimate mixing and phase segregation, respectively, of PC and PVAc chains in the coalesced and precipitated blends [18]. However, while the precipitated blend is phase-segregated it is also totally amorphous, while the coalesced blend contained crystalline PC domains, which both DSC and X-ray observations demonstrated were absent after cooling from the melt. As a result, during continuous heating in the DIPMS apparatus, it may be more difcult for CH3COOH

Table 1 Possible degradation products due to the exchange reaction between CH3COOH and carbonate functional groups of PC in PC/PVAc blends X OH
X
OC O O

Y OH OeCOeCH3 OeCOOH OeCOeCH3 OeCOeOH OeCOOH

m/z (Da) 228, 270, 272, 312, 314, 316, 482 526 528 568 570 566

Y n

OeCOeCH3 OeCO-OH

2966

T. Uyar et al. / Polymer Degradation and Stability 91 (2006) 2960e2967

Physical PC/PVAc blend

coalesced PC/PVAc blend

x29.6

x19.2

m/z=509

x29.6

x10.4

m/z=508

x11.8

x3.1

m/z=467

x13.9

x4.9

m/z=405

x99.0

x320

m/z=312

x5.7

x11.8

m/z=270

x1.0

x1.0

m/z=213

x7.2

x1.1

m/z=43

x52.9

x2.3

m/z=60

x39.0

x1.6

m/z=45

x14.6 300 350 400 450

x4.6

m/z=44

Temperature C
x25.4 300 350 400 450 m/z=73

Temperature C
Fig. 4. Evolution proles of some characteristic products recorded during the pyrolysis of (a) physical PC/PVAc blend and (b) coalesced PC/PVAc blend.

produced by deacetylation of PVAc above 300  C to diffuse into at least the portion of PC that was initially crystalline in the coalesced blend until after melting at w250  C, compared with the initially totally amorphous, though phase-segregated precipitated blend. It was not possible to detect any change, even if it occurred, in the thermal stability of the polyolenic backbone bonds generated as a result of deacetylation of PVAc during the thermal degradation of the coalesced blend, as was noted in the

case of coalesced PVAc [20]. Various thermal degradation products of PC merged with the unsaturated fragments from the thermal degradation of polyolenic backbone bonds. 4. Conclusions Direct insertion probe pyrolysis mass spectrometry analyses of coalesced and physical PC/PVAc blends indicated that CH3COOH formed by deacetylation of PVAc above 300  C

T. Uyar et al. / Polymer Degradation and Stability 91 (2006) 2960e2967

2967

reacts with PC chains decreasing their thermal stability. The process was more effective for the physical blend, most likely due to enhanced diffusion of CH3COOH into the PC domains, where it can react producing low molecular weight PC species bearing methyl carbonate chain ends. The initial partial PC crystallinity of the coalesced blend only becomes amorphous above w250  C, and so the diffusion of CH3COOH into the only recently molten PC chains may be retarded. Acknowledgements The authors are grateful to the National Textile Center (US Dept. of Commerce), North Carolina State University, and TUBITAK through the research funds TBAG-1691 and TBAG-1997, for their nancial support. We also appreciate the Direct Insertion Probe Pyrolysis Mass Spectrometry (DIP-MS) experiments performed by Evren Aslan. References
[1] Utracki LA, editor. Polymer alloys and blends e thermodynamics and rheology. New York: Oxford University Press; 1990. [2] Krause S. In: Paul DR, Newman S, editors. Polymer blends. New York: Academic Press; 1978. p. 15. [3] Ellis TS. In: Shonaike GO, Simon GP, editors. Polymer alloys and blends. New York: Marcel Dekker; 1999. p. 209. [4] Harada A, Li J, Kamachi M. Nature 1994;370:126. [5] Storsberg J, Aert HV, Roost CV, Ritter H. Macromolecules 2003;36(1):50e3. [6] Kamiya M, Mitsuhashi S, Makino M, Yoshioka H. J Phys Chem 1992;96:95e9. [7] Huang L, Allen E, Tonelli AE. Polymer 1999;40(11):3211e21. [8] Lu J, Shin ID, Nojima S, Tonelli AE. Polymer 2000;41(15):5871e83.

[9] Shuai X, Porbeni FE, Wei M, Shin ID, Tonelli AE. Macromolecules 2001;34(21):7355e61. [10] Rusa CC, Rusa M, Gomez M, Shin ID, Fox JD, Tonelli AE. Macromolecules 2004;37(21):7992e9. [11] Bullions TA, Wei M, Porbeni FE, Gerber MJ, Peet J, Balik M, et al. J Polym Sci Part B Polym Phys 2002;40(10):992e1012. [12] Huang L, Tonelli AE. J Macromol Sci Rev Macromol Chem Phys 1998;C38(4):781. [13] Wei M, Davis W, Urban B, Song Y, Porbeni FE, Wang X, et al. Macromolecules 2002;35:8039. [14] Uyar T, Rusa CC, Hunt MA, Aslan E, Hacaloglu J, Tonelli AE. Polymer 2005;46:4762e75. [15] Rusa CC, Tonelli AE. Macromolecules 2000;33(15):5321e4. [16] Bullions TA, Edeki EM, Porbeni FE, Wei M, Shuai X, Rusa CC, et al. J Polym Sci Part B Polym Phys 2003;41(2):139e48. [17] Wei M, Tonelli AE. Macromolecules 2001;34:4061. [18] Uyar T, Rusa CC, Wang X, Rusa M, Hacaloglu J, Tonelli AE. J Polym Sci Part B Polym Phys 2005;43(22):2578e93. [19] Rusa CC, Uyar T, Rusa M, Wang X, Hunt MA, Tonelli AE. J Polym Sci Part B Polym Phys 2004;42(22):4182e94. [20] Uyar T, Aslan E, Tonelli AE, Hacaloglu J. Polym Degrad Stab 2006;91:1e11. [21] Uyar T, Oguz G, Tonelli AE, Hacaloglu J. Polym Degrad Stab 2006;91:2471e81. [22] Sato M, Ooki A, Inamura I, Kubo Y. Macromol Rapid Commun 2002;23:362e5. [23] Montaudo G, Carroccio S, Puglisi C. J Anal Appl Pyrolysis 2002;64:229e47. [24] Jang BM, Wilke CA. Polym Degrad Stab 2004;86:419e30. [25] Polli H, Pontes LAM, Araujo AS. J Therm Anal Calorimetry 2005;79:383e7. [26] Guyot A, Bert M. J Appl Polym Sci 1973;17:753. [27] Gupta MC, Viswanath SG. Ind Eng Chem Res 1998;37:2707. [28] Rabeony M, Hseih DT, Garner RT, Peiffer DG. J Chem Phys 1992;97:4505e11. [29] Debier D, Jonas AM, Legras R. J Polym Sci Part B Polym Phys 1998;36:2197e210. [30] Montaudo G, Puglisi C, Samperi FJ. J Polym Sci Part A Polym Chem 1998;36:1873e84.