Polymer Journal, Vol. 28, No. 5, pp 446--451 (1996)

Synthesis of High Molecular Weight Nylon 6 by Anionic Polymerization of e-Caprolactam

Kazue UEDA, Kazunobu YAMADA, Makoto NAKAI, Tsunetoshi MATSUDA, Masahiro HosooA, and Kazuo T AI

Research & Development Center, UNITIKA LTD., 23 Kozakura, Uji, Kyoto 611, Japan

(Received October 20, 1995)

ABSTRACT: To obtain high molecular weight nylon 6, the anionic polymerization of e-caprolactam was investigated in terms of water content in monomer and concentration of catalyst (Grignard reagent; ethyl magnesium bromide) and chain initiator (monofunctional; N-acetyl-e-caprolactam or bifunctional; N,N'-adipoyl-bis-e-caprolactam). The weight average molecular weight, Mw=7.0 x 10 5 (the intrinsic viscosity; [ry] = 11.2) was obtained using the monofunctional chain initiator, and Mw=9.5 x 10 5 (['1] = 14.1) was found using the bifunctional chain initiator under the following conditions: water content, 0.013 mol%; concentration of catalyst, 0.1 mol%; concentration of chain initiator, 0.03 mol% (monofunctional), 0.015 mol% (bifunctional); polymerization temperature, 150°C. In the present study, the molecular weight of the polymer is governed by quantities of active molecular chains. Since water mainly terminates the propagation of molecular chains, minimizing water content in e-caprolactam is indispensable to maximizing the molecular weight of the polymer. KEY WORDS High Molecular Weight I Nylon 6 I Anionic Polymerization I e-Caprolactam I Water Content I Catalyst I Grignard Reagent I Chain Initiator

The anionic polymerization of s-caprolactam has been Seiyaku Ltd.) as solvent for viscosity measurement and extensively investigated because of the high yield and hexafluoro-2-propanol (HFIP; Ishizu Seiyaku Ltd.) as high rate of polymerization. 1 - 4 Fundamental characters the mobile phase in gel permeation chromatography such as various catalysts, 2 •5 •6 propagation mech­ were of reagent grade and used without further pu­ anism 7 - 15 and kinetics16 - 26 are well known. Sebenda rification. et al. 7 • 16 - 22•2 7 studied the propagation mechanism and kinetics of the polymerization, and propose the free ion Polymerization mechanism and side reaction mechanism. Recently, Water in s-caprolactam was removed to less than Stehlicek and Puffr reported that Grignard reagents as 0.013 mol% by bubbling dry argon gas for 5 h in a anionic catalysts gave much higher rates of the po­ three-neck flask at 120oc. Water content was controlled lymerization than sodium-s-caprolactamate. 26 On the by adding water to this s-caprolactam and measured other hand, Puffr and Vladimirov synthesized the high by Karl Fischer auto titration equipment (Hiranuma molecular weight nylon 6 using potassium-s-caprolacta­ Industrial Co., AQV-5S ). The chain initiator was injected mate as the catalyst at 180oC. 6 The studies directed into s-caprolactam with controlled water content. The toward high molecular weight nylon 6, however, were mixture of monomer and initiator was poured into a not performed on the Grignard reagents. In addition, glass tube (inside diameter 20 mm, volume 20 ml) the influence of water content on polymerization was not beforehand dried by heating above 100°C under 10 Pa investigated in spite of the general concept that water of vacuum. The catalyst (EtMgBr) solution was injected content is very important. In this paper, we describe the into the tube via a syringe, and the tube was sealed method for synthesizing high molecular weight nylon 6 after unnecessary solvent was evaporated at 10 Pa. by the anionic polymerization of s-caprolactam. Dis­ Polymerization was carried out with shaking in an oil cussed is the importance of water content and con­ bath at 130-190°C (thermostated at± 1.0°C). centration of the catalyst (Grignard reagent) and mono­ As the polymerization of s-caprolactam proceeded, the or bifunctional chain initiator. viscosity of the system increased, and fluidity dis­ appeared. We called this point the solidification point, EXPERIMENTAL when the fluidity of the system disappeared, with glass tube in tilted position. We define the time from the Materials beginning of the polymerization to the solidification point Industrial fiber grade s-caprolactam as monomer, ethyl as the solidification time. Since it was confirmed that magnesium bromide (EtMgBr; Aldrich Chemical Co., conversion almost reached 20% at the solidification time, Inc.) as catalyst in 3.0 moll- 1 diethyl ether solution or this time was conveniently used for estimation of reaction 1.0 moll- 1 tetrahydrofuran solution and N-acetyl-s­ rate. The polymerization was stopped at 2.5 times (using caprolactam (Ac-CL; Tokyo Chemical Industry Co., 0.5mol% EtMgBr) or 6 times (using 0.1 mol% EtMgBr) Ltd.) as monofunctional chain initiator were used for of the solidification time. It was confirmed that at that the polymerization. N,N'-adipoyl-bis-s-caprolactam time conversion was about 100%. Termination of the (Ad-CL) as bifunctional chain initiator was prepared polymerization was carried out by quenching to dry from s-caprolactam and adipoyl-chloride (Aldrich Co., ice/methanol temperature (-78°C). The product was Inc.) in the presence of pyridine, and the product was shaved with a boring machine and analyzed. Mono­ purified by recrystallization. 28 96% sulfuric acid (Ishizu mer and oligomer in the polymer were extracted by 446 Synthesis of High Molecular Weight Nylon 6

1.5,------, 300r------. "2 '\ I 200 \ . Q) D 0> E 0 ..., _J c: 0

(.) 100 <;::: 32 0 -0.5 4 5 6 (/) Log Mw

Fi_gure 1. Relation between log[IJ] and logMw. ['1] was determined 0.1 0.2 0.3 With polymer concentration below 0.2gdl- 1 (5 points) using H 2S04 Concentration of Ac-CL (mol%) at 25°C. Mw was measured by LS using HFIP at 30°C. Figure 2. Effects of water content on the solidification time, where the solidification time is plotted against the concentration of the chain boiling water in a flask for 5 h. The monomer content initiator (Ac-CL). Polymerization conditions: temperature, 150°C; in the extract was measured by liquid chromatogra­ concentration of catalyst (EtMgBr), 0.5mol%; water content, phy (Column; Waters Cl8 Purecil, mobile phase; 0.1 M (e), 0.038mol% (D), 0.063mol% (.&.), 0.088mol% (0), NaH2P04 + H 3P04 , pH 3.0), and conversion was calcu­ respectively. lated from change in monomer content before and after polymerization. 5.0 r----r------,

Molecular Weight Determination 4.0 The molecular weight of the polymers was convention­ allr intrinsic viscosity measurement ([1]]; "'6 3.0 96 Yo sulfunc ac1d, at 25°C). The concentration for this X measurement was below 0.2 g dL - 1 because of non­ 2.0 D Newtonian behavior in the ordinary concentration of the 1 29 measurement (less than 0.5 g dL - ). Light scattering 1.0 (LS; Wyatt LTD., DAWN-F) measurements for the polymer with various [17] were carried out in HFIP to obtain weight average molecular weight (Mw). o.o 0.1 0.2 0.3 Figure 1 shows the relation between [17] and M of Concentration of Ac-CL (mol%) the polymer in this study. Linear regression analysis was applied to the data and coefficients of Mark­ Figure 3. Effects of water content on Mw, where Mw is plotted against the concentration of the chain initiator (Ac-CL). Polymerization Houwink-Sakurada equation were determined as fol­ conditions: temperature, 150oC; concentration of catalyst (EtMgBr), lows: 0.5mol%; water content, 0.013mol% (e), 0.038mol% (D), 0.063mol% (.). 0.088mol% (0), respectively. The thick solid line (1) (-)is the calculated Mw. Mw of the polymers under various polymerization con­ ditions were calculated by eq 1 using observed [17]. water contents; 0.013, 0.038, 0.063, and 0.088mol%, Gel permeation chromatography (GPC; TSK gel respectively. The concentration of catalyst (EtMgBr) was GMH-H+G7000H) measurements were carried out kept constant at 0.5mol%. As the Ac-CL concentration with HFIP as the mobile phase. In general, GPC measure­ decreases, the solidification time gradually increases at ments give Mw, the number average molecular weight water Solidification time becomes longer w1th mcreasmg water content. At low concentration of (Mn) and polydispersity (P=MwfMn). Calibration of the column could not be carried out, since we had no the chain initiator, the solidification time becomes steeply longer at every water content. Figure 3 shows M w of the sample of standard nylon 6 with monodispersed polymer as a function of Ac-CL concentration under the molecular weight. Therefore, only P reliably obtained same conditions as Figure 2. The relation between M from GPC measurement was used. Details of the analysis and Ac-CL concentration is shown by a curved plot will be published elsewhere. 30 with a maximum peak. As water content decreases, the maximum Mw increases. Mw (4.0 x 10 5) was maximum RESULTS AND DISCUSSION at0.07mol% of Ac-CLand0.013mol% water content. The thick solid line in Figure 3 indicates the calculated Influence of Water Content The influence of water content on anionic polymeriza­ M w assuming eq 2. tion is demonstrated in Figures 2 and 3. Here in Figure P·McL (2) 2 the experimental results of the solidification time of [I]/[M] the polymerization are shown as a function of the concentration of the chain initiator (Ac-CL) at various Here, [I] is the concentration of the chain initiator Polym. J., Vol. 28, No. 5, 1996 447 K. UEDA et a/.

Table I. Reaction mechanisms of the anionic polymerization of £-caprolactam3 ' 27 ' 31

H 0 e EtMgBr 0 .. 0 ( i)

0 0 0 8 II II II + 0 CH3-co ( ii)

e H 0 0 + 0 (iii)

(iv)

1 (mol g- ), and [M] is the initial concentration of the 100 ,------6.0 monomer (molg- 1), and MeL is the molecular weight of e-caprolactam. P means polydispersity (P=MwfMn), the value of which was estimated to be 2.3 by GPC .Ec 80 measurements. As for this polymerization, it was pro­ 4.4.0 QJ OJ') posed that the initiation reaction proceeds according E 60 • • • 0 +=' to the eq i and ii in Table I (the main reactions), 3 ·27•31 • r c 6 X and the molecular chains extend with 1 as an initiator. 0 ""§ 40 $: Therefore, if the polymerization proceeds ideally, the () 2.0 0 number of chains corresponds to the number of initiators, :Q 0 20 represented by eq 2. The smaller amount of initiator (f) 1) might provide the higher molecular weight.

In Figure 3, the right side of the peak top in each o curve closely follows the calculated line. The maximum 0.0 0.1 0.2 0.3 0.4 0.5 value of M w obtained by these experiments (water content Concentration of EtMgBr (mol%) 0.013 mol%) is 4.0 x 10 5 , which corresponds to Mw= 3.7 x 10 5 calculated from the concentration of Ac-CL Figure 4. Effects of concerntration of catalyst (EtMgBr) on the solidification time and Mw. Polymerization was carried out at: (0.07mol%). Therefore, it is clarified that the high temperature, 150°C; concentration of chain initiator (Ac-CL), molecular weight nylon 6 is obtained along with the line 0.07 mol%; water content, 0.013 mol%. by controlling the water content and the concentration of chain initiator Ac-CL. On the anionic polymerization with Grignard reagents, In Figure 3, the left side of the peak top in each curve it was revealed the water content and the chain initiator deviates appreciably from the calculated line. The concentration controls the molecular weight of the solidification time at low Ac-CL concentration (especially product. below 0.05 mol%) becomes longer such as in Figure 2 even at the water content 0.013mol%. The conver­ Effects of Concentration of Catalyst sion reaches only 70-85% in the left side of the peak To clarify the effects of the catalyst, the solidification top. The long solidification time is the result of low time and M w of the product versus the concentration of concentration of the growing chain end. Consequently, catalyst EtMgBr were investigated in the presence of the side reaction like eq iii in Table I with slow reaction 0.013 mol% water, which is shown in Figure 4. The rate3 becomes significant, which is generally ignored in initiator Ac-CL concentration was kept constant at polymerization using a high concentration of the chain 0.07 mol%. Although the solidification time becomes initiator. The influence of water is not ignored for the longer below 0.1 mol% of EtMgBr, M w of the polymer anionic polymerization even at 0.013mol%, which is is constant at about 3.9 x 10 5 ([17] = 6.6). It is suggested comparable with the concentration of Ac-CL (0.03- that the molecular weight of the product does not mainly 0.07mol%). Water causes the side reactions like eq iv in depend on the concentration of the catalyst. To achieve Table I and stops the propagation of the molecular chain. the high molecular weight, however, the polymerization The deviation between the experimental Mw and the should be ideally initiated by the minimum concentration calculated M w is caused by .all factors described above. oft formed with molar equivalent of the catalyst EtMgBr 448 Polym. J., Vol. 28, No. 5, 1996 Synthesis of High Molecular Weight Nylon 6

100 .···t:s···--····· 600 8.0 ,-... 80 7.0 ...... ,iR 500 '2 c:: 60 6.0 0 "§ I 400 Q) "' Q) 40 5.0 6 > E c:: 0 300 X (.) 20 """c:: 4.0 0 3.0 == 0 200 0 100 200 300 400 500 :!2 2.0 0 Polymerization time (min) C/) 100 1.0 Figure 5. Time--conversion curves at various polymerization tem­ perature. Polymerization conditions: temperature, 130oC ( 0 ), 140°C 0 0 (6), 150oC (D), 170°C ( +), 190oC (.), respectively; concentration of 0.00 0.02 0.04 0.06 0.08 0.10 catalyst (EtMgBr), 0.1 mol%; chain initator (Ac-CL), 0.05 mol%; water content, 0.013mol%. Concentration of Ac-CL (mol%) Figure 7. Plots of solidification time and Mw against the concentra­ 6.0 tion of the monofunctional chain initiator (Ac-CL). Polymerization conditions: temperature, 150°C; concentration of catalyst (EtMgBr), 5.0 0.1 mol%; water content, 0.013 mol%. in temperature. This means that the polymerization rate 4.0 '? increases with temperature. The conversion at 190°C does 0 not reach 100% (about 97%), since an equiliblium 3.0 X monomer exists probably. 32 Figure 6 shows Mw of the polymer versus the == 2.0 polymerization time at 130--190°C using Ac-CL. The 1.0 time to reach the steady state of M w decreases with increase in temperature. M w of the polymer synthesized 0 at 150-190°C was about 5.0 x 10 5 and at 130 and 140°C, 0 100 200 300 400 500 only 3.3 x 10 5 and 4.4 x 10 5 respectively. Polymerization time (min) In spite of decrease in side reactions at lower tem­ perature (130 and 140°C), it takes a very long time to Figure 6. Effects of the polymerization temperature on Mw, where reach 100% conversion, since the main reaction be­ Mw is plotted against the polymerization time at various temperatures. Polymerization conditions: temperature, 130°C ( 0 ), 140oC (6), 150oC comes slower than that at higher temperature. Mw of (D), 170°C ( +), 190oC (.), respectively; concentration of catalyst the polymers synthesized at 170aC and 190°C are same (EtMgBr), 0.1 mol%; chain initiator (Ac-CL), 0.05 mol%; water as at 150°C despite the increase in polymerization rate. content, 0.013 mol%. The increase of Mw after reaching 100% conversion at 170ac is almost within experimental error. It is con­ and the chain initiator Ac-CL. Consequently, the sidered that the lower temperature is favorable to de­ concentration of EtMgBr is required to approximate the crease side reactions, so that 150°C is the most suitable concentration of Ac-CL closely and not to increase the temperature of the polymerization. side reaction or not to decrease the polymerization rate. We conclude the concentration of the catalyst should be Maximum Mw by Monofunctional Initiator reduced to 0.1 mol%. The polymerization was carried To obtain higher molecular weight nylon 6, polym­ out at 0.1 mol% of EtMgBr. erization was carried out at 0.1 mol% of EtMgBr and 0.013 mol% of water content. The solidification time and Effects of Temperature Mw of the polymer as a function of the concentration The catalytic system cons1stmg of the catalyst of the chain initiator (Ac-CL) are shown in Figure 7. (EtMgBr) and the chain initiator (Ac-CL or Ad-CL) may As the concentration of the chain initiator decreases, Mw provide advantageous polymerization at lower temper­ of the polymer increases gradually above 0.03 mol% ature. Polymerization temperature is an important fac­ of Ac-CL. Mw maximum is 7.0 x 10 5 (['7] = 11.2), at tor for controlling the reaction rate of the main reaction 0.03mol% of Ac-CL concentration. Below 0.03mol% and side reaction. The effects of temperature on the of Ac-CL, M w of the polymer decreases. This is similar polymerization were studied and the results are shown to the results in Figure 3. The solidification time in­ in Figures 5 and 6. Here, the concentration of catalyst creases monotonously and steep increase are observed (EtMgBr) is 0.1 mol%, the concentration of chain below the chain initiator 0.03 mol%. This behavior is initiator (Ac-CL) is 0.05 mol% and water content is similar to that in Figure 2. Consequently, by controlling 0.013 mol%. Figure 5 shows the relation between the the catalyst concentration, the increase of M w was more conversion and the polymerization time at 130--190aC in the region of the lower concentration of the chain using Ac-CL. The conversion increases monotonously initiator than that of Figure 3 (0.07 mol% of Ac-CL). with polymerization time for all temperatures. The time The observed maximum value of Mw (7.0 x 10 5) is to reach 100% conversion becomes short with increase somewhat lower than the calculated value of Mw Polym. J., VoL 28, No. 5, 1996 449 K. UEDA et a/.

200 CONCLUSION

c 8.0 To obtain high molecular weight nylon 6, the anionic I 150 polymerization of e-caprolactam was explored. Maxi­ Q) "'6 mum molecular weights of Mw=7.0x 105 ([17]=11.2) E 6.0 using the monofunctional chain initiator (Ac-CL) and X -c: Mw=9.5x105 ([17]=14.1) using the bifunctional one 0 100 3: "iii 4.0 (Ad-CL) were achieved. It was clarified that the molecular () t;::: weight of the polymer depends on quantities of active :2 molecular chains, and minimizing water content is 0 50 CJ) 2.0 necessary for maximal molecular weight, since water mainly terminates the propagation of the molecular 0 0 chains. The essential conditions for maximum M w are; 0.00 0.02 0.04 0.06 0.08 0.10 (1) water content in e-caprolactam should be below Concentration of Ad-CL (mol%) 0.013 mol%, (2) relation between the concentration of the catalyst and the chain initiator is important, and Figure 8. Plots of solidification time and Mw against the concentra­ tion of the bifunctional chain initiator (Ad-CL). Polymerization 0.1 mol% of the catalyst (EtMgBr) and 0.03 mol% conditions: temperature, 150°C; concentration of catalyst (EtMgBr), (monofunctional) or 0.015 mol% (bifunctional) of the 0.1 mol%; water content, 0.013mol%. chain initiator are required, (3) 150oC is the most suitable polymerization temperature without decrease of the (8.6 x 105) assuming that the concentration of Ac-CL is polymerization rate and without increase of side re­ 0.03 mol%. The reason is attributed to the increase of actions. side reactions like eq iii in Table I. Higher M w than 7.0 x 105 using the monofunctional chain initiator (Ac­ REFERENCES CL) could not be achieved, since water in e-caprolactam (even 0.013 mol%) and in EtMgBr solution may not be 1. 1. Sebenda, "Comprehensive Chemical Kinetics," Vol. 15, C. H. ignored in these low concentration of the catalyst and Bamfordand and C. F. H. Tipper, Ed., Elsevier, Amsterdam, 1976, p 279. the chain initiator. It is experimentally difficult to reduce 2. H. Sekiguchi, "Ring-Opening Polymerization," Vol. 2, K.1. Ivin the water content below 0.013 mol%. Therefore, it is and T. Saegusa, Ed., Elsevier, London, 1984, p 832. recognized that the highest molecular weight of the 3. 1. Sebenda, "Comprehensive Polymer Science," Vol. 3, C. F. H. polymer using the monofunctional chain initiator (Ac­ Tipper and C. H. Bamford, Ed., Pergamon Press, Oxford, 1988. 4. R. Puffr, "Lactam-Based Polyamides," Vol. 1, R. Puffr and V. CL) is Mw=7.0 x 105 . Kubimek, Ed., CRC Press, Boca Raton, 1991. 5. 1. Karger-Kocsis and L. Kiss, J. Polym. Sci., Polym. Symp., 69, Maximum Mw by Bifunctional Initiator 67 (1981). Polymerization using the bifunctional chain initiator 6. R. Puffr and N. Vladimirov, Makromol. Chern., 194, 1765 (1993). (Ad-CL) was carried out to achieve further high mo­ 7. 1. Sebenda, J. Macromol. Sci. Chern., A6, 1145 (1972). lecular weight. Figure 8 shows the solidification time 8. G. Champetier and H. Sekiguchi, J. Polym. Sci., 48, 309 (1960). 9. H. Sekiguchi and B. Coutin, J. Po/ym. Sci., Chern. Ed., 11, and Mw of the polymer as a function of the concentration 1601(1973). of the chain initiator Ad-CL at 0.1 mol% of EtMgBr. 10. T. M. Frunze, V. A. Kotel'nikov, T.V. Volkova, V. V. Kurashev, As the concentration of the chain initiator decreases, Mw S. P. Davtyan, and V. V. Korshak, Dokl. Akad. Nauk SSSR, 255, of the polymer increases gradually above 0.015 mol% of 612 (1980). 11. T. M. Frunze, V. A. Kotel'nikov, T. V. Volkova, V. V. Ad-CL. Mw maximum is 9.5x105 ([17]=14.1), which Kurashev, S. P. Davtyan, and I. V. Stankevich, Acta Polym., 32, is achieved at 0.015mol% of Ad-CL concentration. Mw 31 (1981). of the polymer decreases below 0.015mol% of Ad-CL. 12. T. M. Frunze, V. A. Kotel'nikov, T. V. Volkova, and V. V. This is similar to the results in Figure 7. The solidifica­ Kurashev, Eur. Polym. J., 17, 1079 (1981). tion time is also similar to that in Figure 7 except that the 13. N. Mougin, C. A. Veith, R. E. Cohen, and Y. Gnanou, time is about half. Macromolecules, 25, 2004 (1992). 14. H. Tani and T. Konomi, J. Polym. Sci. A-1, 6, 2295 (1968). The concentration of the bifunctional chain initiator 15. R. Z. Greenley, 1. C. Stauffer, and 1. E. Kurz, Macromolecules, (Ad-CL) providing maximum Mw is 0.015mol%, which 2, 561 (1969). is half the monofunctional chain initiator (Ac-CL, 16. E. Sittler and 1. Sebenda, Collect. Czech. Chern. Commun., 33, 0.03 mol%), since Ad-CL has 2 active molecular chain 270 (1968). 5 17. E. Sittler and 1. Sebenda, Collect. Czech. Chern. Commun., 33, ends per molecule. Maximum Mw=9.5 x 10 (Ad-CL, 3182 (1968). 0.015 mol%) obtained from the bifunctional chain 18. 1. Sebenda, J. Polym. Sci. C, 23, 169 (1968). initiator is lower than the calculated Mw (1.7 x 106 ). This 19. J. Sebenda and V. Kouri!, Eur. Polym. J., 8, 437 (1972). considerable deviation between the observed and 20. 1. Sebenda, A. Stiborova, L. Lochman, and Z. Bukac, Org. Prepr. Proced. Int., 12, 289 (1980) [Chern. Abstr., 94, 16165 (1981)]. calculated M w may be due to the relative increase of side 21. J. Stehlicek and J. Sebenda, Eur. Polym. J., 22, 769 (1986). reactions such as eq iii in Table I. The highest molecular 22. 1. Stehlicek, J. Labsky, and 1. Sebenda, Collect. Czech. Chern. weight of the polymer obtained by the bifunctional chain Commun., 32, 545 (1987). initiator is Mw=9.5 x 105 . 23. T. M. Frunze, S. P. Davtyan, V. A. Kotel'nikov, T.V. Volkova, and V. V. Kurashev, Vysokomo/. Soedin., Ser. B, 23,388 (1981).

450 Polym. J., Vol. 28, No. 5, 1996 Synthesis of High Molecular Weight Nylon 6

24. V. G. Frolov, Vysokorno/. Soedin., Ser. B, 25, 134 (1983). Polon. Sci., ser. Sci. Chirn., 14, 751 (1966). 25. S. Barzakay, M. Levy, and D. Vofsi, J. Polyrn. Sci. A-1, 5, 965 29. M. Turuta, "Polyamide-jusi," Nikkan-k6gy6-shinbun, Tokyo, (1967). 1961. 26. J. Stehlicek and R. Puffr, Macrornol. Chern., 193, 10,2539 (1992). 30. K. Ueda, K. Hattori, M. Nakai, and K. Tai, to be published. 27. J. Sebenda, Preprints of Scientific Papers I, IUPAC International 31. S. Barzakay, M. Levy, and D. Vofsi, J. Polyrn. Sci. B, 3, 601 Symposium on Macromolecular Chemistry, Tokyo-Kyoto, 1966, (1965). p 251. 32. H. Yumoto, J. Chern. Phys., 29, 1234 (1958). 28. S. Chrzczonowlcz, J. Kowalski, and M. Wlodarczyk, Bull. Acad.

Polym. J., Vol. 28, No. 5, 1996 451