Brief Review of Advances in Developments of Living Polymers

by Michael Szwarc

Hydrocarbon Research Institute, University of Southern California

テトラヒドロフランのような極性非プロトン溶媒の中でナトリウムナフタリンのラジカルアニオンからス チレンモノマーへの1電子移動,それに続くスチレンのラジカルアニオンの2量化を開始反応とするリビン グアニオン重合が発見され,分子量のそろった単分散ポリスチレンやブロックコポリマーの合成はリビング アニオン重合によって可能となった.それから35年が過ぎたが,ビニルモノマーのカチオン重合法による リビングポリマー,あるいは金属ポルヒリンを用いた環状エーテル類のリビングポリマーの合成など,リビ ングポリマーの合成は大きく発展してきた.ここではリビングポリマーの開発の進歩について述べる.

Polymerization is referred to as living when the free of homo polymers, were the only products. growing polymers retain their capacity of growth Thus, a novel method of synthesis of well character for a time sufficiently long to allow the operator to ized block-polymers was discovered. complete a contemplated task, i.e. when the reaction Synthetic advantages of living polymerization is propagated with virtual exclusion of termination and chain transfer. It is beneficial if all the polymer- Living polymerization provides synthetic polymer ic molecules start their growth simultaneously, chemists with powerfull tools for design polymer because then the resulting polymers are of uniform chains of requested specification, namely : size. (1) A control of the molecular mass of the for- The feasibility of such a polymerization was med polymers and a method of producing polymers demonstrated 35 years ago by simple and most of uniform size. convincing experiments' described here briefly. The (2) A new technique for preparing block- reactor, shown in Fig. 1, contained 60 ml of-3 .10' polymers of desired sequence of blocks of requested M THF solution of sodium naphthalenide (the initi- size and composition with exclusion of homo- ator) to which 10 gr. of styrene was added by polymers. crushing a breakseal. The polymerization ensued instantly and was over in seconds. The viscosity of the resulting solution was estimated by the time of fall, —5 s, of a plunger placed in a side tube filled with the polymer solution by tilting the reactor. After a while, additional 10 gr of styrene in 60 ml of THF was added. The subsequent polymerization was over again in seconds and the viscosity of the solution increased, as indicated by the increased of the time of fall of the plunger to 20 s. Since the concentration of the polymer did not change, this proved that the previously formed polymers retained their activity and grew longer on the addi- tion of the monomer. Repetition of these experi- ments using isoprene as a second monomer Equipment for demonstrating living character confirmed these conclusions since block-polymers, of polymerization.

798高 分 子40巻12月 号(1991年) (3) A method of attachment of desired func- butadiene-styrene-butadiene nor the di-block poly- tional groups to one or both ends of a polymer chain mers behave in this uncommon way. Similar prop- since the addition of a proper reagent easily con- erties are claimed for the cationically prepared verts the active end of a living polymer into a indene-iso-butene-indene ter-block polymers'. desired functional group''. Films of di-blocks of styrene and butadiene of a The control of molecular mass of the produced sufficiently high molecular mass have irridescent polymers is of obvious advantage. Living polymers colors', although the homo-polymers are colorless. can be "fed" by monomer until they attain the The color depends on the molecular mass of the desired size. Their number average momecular blocks and changes reversibly on touching the film, mass increases proportionally with conversion, pro- This phenomenon results from the incompatibility vided that all of them begin their growth simultane- of the blocks which, being bonded and unable to ously. This requirement is met when the rate of separate, form paralel lamellae 1000 A thick. initiation is faster of or equal to the rate of propaga- These act as an optical grating producing interfer- tion. Alternatively, polymers of uniform molecular ence colors. Such an effect is not shown by the size are formed by feeding with monomer the previ- blends of the homo-polymers since being not linked ously prepared "monomeric" or oligomeric living by chemical bond they can separate into large irreg- polymers. For example, cumyl potassium, a living ular domains. monomeric "poly-a-methylstyrene", grows to a liv- Another example revealing the effect of linking ing n-meric poly- a-methylstyrene on addition of incompatible blocks by chemical bond is shown by a-methylstyrene. The complex of 1-phenylethyl the di-block films casted from good solvents for one chloride with tin tetrachloride, CH3CH (Ph) Cl • block but poor for the other. The same styrene- SnC14 is a living monomeric "polystyrene"' yielding butadiene di-block forms a flexible film when casted on feeding with styrene a living n-meric--- from a solvent good for polybutadiene but bad for CH,CH (Ph) Cl SnCl., polystyrene. The metallated poly-styrene, whereas a glassy film is formed in the iso-butyric ester, (CH3) 2C (COOCE13) , Cat+, is a reverse case. In the former solvent polystyrene living monomeric "polymethyl methacrylate"4 form- blocks form granulae in a continuous matrix of the ing on the addition of the monomer a living-- flexible poly-butadiene, whereas polystyrene is the CH2C- (CH,) (COOCH3) , Cat+ polymethyl metha- continuous phase in the latter solvent. crylate. Nuyken5 used recently a stable solution of Many di-block polymers are effective emulsifying CH3CH (OR) I complexed with quaternary ammo- agents, especially when composed by hydrophilic nium salt to produce, by feeding it with monomer, a and hydrophobic blocks. Graft and comb-shape similarly complexed living iodide of polyvinyl ether, polymers are prepared by a modified living polymer etc. In all these cases the resulting polymers are of technique', and many of them are most useful as uniform size since all of them start their growth adhesive and elastomers. Similar procedures' lead simultaneously. to star-shape polymers valuable in control of Living polymerization is unique in allowing prep- rheological properties of lubricants and as an agent aration of block polymers of desired architecture, strengthening various polymeric materials. free of homo-polymers. The styrene-butadiene- Much interest is shown nowaday in well defined styrene is a commercial ter-block polymer produced macro-cyclic polymers prepared by linking the by living anionic technique having the interesting active ends of hi-functional living polymers with and useful thermo-plastic properties. It is moldable appropriate "linking" agents, e.g. di-carbanions of at higher temperatures but behaves as a tough living linear polystyrene converted into rings by a, cross-linked rubber at ambient conditions'. This w-dibromo-p-xylene or dimethyldichlorosilane'°. remarkable and reversible change of its properties is Their rheological behavior differs drastically from caused by the formation of rigid micro-spheres of that of linear polymers of the same degree of poly- glassy polystyrene on cooling the melt, and these act merization, for example their respective glass as giant cross-links. Significantly, neither the temperarature decreases with the degree of poly-

高分子40巻12月 号(1991年)799 merization" whereas it increases for linear poly- their rate of propagation is slown down on the mers. Significantly, it makes a difference whether addition of crown ethers or cryptates" that convert cyclic polymers of Mn-10,000 dalton are linked by the tight pairs into the loose ones. The retarding ortho rather than by para-dibromoxylene". effect of these powerful cation solvating agents Numerous functional polymers were prepared indicates the participation of push-pull effect, the from living anionic or cationic polymers. The prob- anion and cation cooperate in the monomer addition lems encountered in the preparation of functional process. polymers by anionic technique were discussed in a A more convincing evidence for the operation of recent review by Quirk". Novel kind of functional push-pull mechanism is provided by the reaction of polymers, referred to as macromers, were reported ethylene oxide with anions of substantially delocal- 11 years ago". Such polymers are terminated by ized charge. The undissociated ion-pairs of 9- units acting as monomers, e.g., methylfluorenyl" or fluoradenyl" cleave the oxir- ----CH=CHPh or --CH=C(CH3)C0 OCH3, ane faster than their free ions, the addition of cryptates retards the reaction, whereas the addition and could be copolymerized with small monomers of common cation salts, that convert the free anion yielding polymers with predetermined branches. into ion-pairs, accelerates its course. Polymers endowed with two functional end-groups, Living anionic polymerization of polymethyl known as telechelics, are most useful as chain exten- methacrylates is more complex, the active anions tion agents. For example, the diols formed by the form allylic-enolates solvated by penultimate poly- cationic ring-opening polymerization of cyclic mer units". Consequently, two centers may be ethers are used in preparation of urethane foams. involved in the propagation : the C- or the 0-. The

Interesting amphilic gels were prepared by polymer- participation of 0- leads to some side reactions izing acrylic monomers with bi-macromers of which destroy the living character of this polymer- cationicaly produced living polyisobutene terminat- ization, especially in low polarity solvents. These ed on both ends by acrylic units. Such gels swell in undesired reactions are eliminated, or diminished, water and in organic solvents", and the character of by the addition of LiCl". The action of this salt is their surface varies with the nature of the environ- not well understood, it reacts probably with the ment. The hydrophilic segments dominate in their carbonyl group of the active anion and reduces its surface when the gel is emersed in water, whereas enolic character. Anionic polymerization of methyl the hydrophobic segments accumulate in the surface metacrylate yields polymers of narrow molecular in an oily medium. mass distribution when large cations are employed in polar solvents at low temperature. A remarkable The nature of living polymers improvement of the quality of the resulting poly- Anionic techniques were the first utilized in the mers is achieved by substituting tetrabutyl ammo- preparation of living polymers. The most adaptable nium for the alkali ions", the elimination of alkox- monomers for such polymerizations are : styrene ides is then greatly reduced. and its derivatives, various dienes, methacrylates Species, which are in equilibrium with the grow- and acrylates, small ring monomers such as ethyl- ing polymers but do not propagate themselves, are ene oxide, propylene sulfide, hexamethylcyclo- referred to as dormant polymers. The aggregates trisiloxane, some lactones, etc. A variety of species formed by lithium polystyrene or polydienes in participate in such reactions. In ethereal solvents hydrocarbon solvents are the classic examples of free anions and solvent separated ion-pairs propa- dormant polymers. Their participation in polymer- gate the polymerization of styrene, the dienes, and ization affects the kinetics of the reaction. The their derivatives, while the contribution of the most kinetics of propagation seems to be peculiar in abundant tight pairs is negligible. In contrast, the systems involving dormant triple ions, its rate is tight —0-Alk+ ion-pairs are more reactive than the then constant independent of the concentration of loose solvated pairs in polymerizations of oxiranes, living polymers. A treatement of such systems is

800高分子40巻12月号(1991年) outlined in ref. 21 (pp. 115-119). Not surprisingly, the addition of alcohol to a Within a decade after discovery of anionic living living polymer leads to a similar reaction yielding polymers it was shown that cationic ring-opening the adduct : polymerization involves also living polymers". These reactions are propagated by onium ions as well as by covalent end-groups, an important obser- vation first reported by Saegusa23. The reversible interconversion of onium ions into covalent species Alcohol acts here as a chain-transfer agent, it led to a fascinating research reviewed by Penczek regenerates the catalyst and produces a hydroxyl and his colleagues". It should be stressed that the terminated polymer. Hence, the reaction continues onium ions are relatively stable, and are more inspite of the presence of alcohols. This behavior of abundant than the covalent species in most of the the system induced Inoue to refer to it as "immortal equilibrated systems. In contrast, the dormant cova- polymerization", a debatable term. lent polymers are the more abundant in systems Another interesting development resulted from involving the esters derived from carbenium ions. the work of Webster and his colleagues" who repor- In the last decade several interesting living poly- ted a polymerization of methyl methacrylate initiat- mer systems have been discovered. The most versa- ed by silyl ketene acetal, SiR3O-C (OH3) =C (CH3) 22 tile catalysts prepared by hydrolysis of zink and activated by a suitable nucleophile. The following aluminum alkyls were developed by Tsuruta25 and scheme seems to account for this process referred to by Vandenberg". Their work , was extended by as the group transfer polymerization. Teyssie who showed" that these bimetallic alkox- ides , catalyse living polymerization e -caprolactone. Even more effective is the catalyst developed by Inoue based on a metalloporphyrine shown below.

He polymerized a variety of oxiranes using this catalyst, and produced polymers of narrow molecu- lar mass distribution and a well characterized di- and tri-block polymers with exclusion of homo- polymers. These achievements demonstrate the M denotes here the monomer and Nu- the activat- value of this method. Moreover, the interesting co- ing nucleophile. Alternatively, the reaction might polymerization of CO2 with oxiranes deserves spe- result from splitting Me3SiOAc (0Ac- is the nucleo- cial stressing. The above polymerizations proceed phile) followed by the conventional anionic presumably by a coordination mechanism operating polymerization of methyl methacrylate. in other similar reactions. The highly polar mono- mers become first coordinated with the Al center, and thereafter inserted into the C-0 bond, e.g.,

—> TPhPorAl—OCH2CH2 OR

高分 子40巻12月 号(1991年)801 Indeed, the addition of Me3SiOAc slows down the propagation, as could be expected on the basis of the second mechanism postulating the equilibrium between the initiated polymers and the growing polymethyl metacrylate anions. Two observations need stressing. To maintain a good quality of the polymers produced by the group transfer polymerization it is necessary to keep the ratio of (nucleophile) (initiator) low, at about 0.01. Hence, only 1% of the initiated polymers grows at any time. Again, the addition of Me3SiOAc, that slows down the propagation, improves the quality of the resulting polymers. These two facts imply that Lamela structure resulting in interference of light some reaction caused by the interaction of two and therefore leading to irridercent colors very- growing polymers, or their intermediates, vitiates ing with molecular size of blocks. the polymerization. Hence, the quality of the poly- mers is improved by reducing its rate relative to that of propagation (i.e., by decreasing the concen- tration of the growing species) . Still another, most promising development in liv- ing polymerization was reported in the last few years, namely the living ring opening metathesis of cyclic olefines". The ring expansion of cyclic olefines induced by complexes of transition metals have been known for a while". Their polymeriza- tion is assumed to result from the formation of complexes of carbene, i.e.,

transfer were observed, i.e., this is a truly living polymerization. Their claim is supported by the where Met denotes a complexed transition metal. recently reported" preparation of block polymers The cyclobutane derivative is in equilibrium with formed on the addition of another strained mono- carbene that propagate the reaction. Most of the mer to the dormant polynorbornene. complexed catalysts react with the unstrained C=C The Ti based catalyst is not unique. A similar bonds of the polymers in a process leading to process involving a tantalum based catalyst was destruction of the chain reaction. The breakthrough reported a year later". A molibdenum based cata- in the Grubbs work came from his choice of a lyst allows for polymerization of functional catalyst, sufficiently active to cause the metathesis monomers" but unfortunately the polymerization is of the stained C=C bonds but not active enough to not living. The polymerization of cyclobutene by react with the unstrained C=C bonds. Specifically, these catalysts is also not living. The resulting the reaction of norbornene induced by bis- polymers have a very broad molecular mass distrbu- (cyclopenatadienyl) -titane-cyclobutane proceeds tion, the Mw/Mr, ratio is greater than 3, because the smoothly at —65°C according to the scheme : initiation is so much slower than propagation. Real- It stops when the reacting mixture is chilled izing the shortcoming of this system Grubbs sear- below 0°C, but resumes when the temperature is ched for the ways of slowing down the propagation raised. According to the authors no termination or and achieved this goal by using strong Lewis

802高 分 子40巻12月 号(1991年) bases". The tungsten based catalyst in conjunction the lifetime of the active polymers, a unimolecular with trimethyl or dimethylphenyl-phosphine turned one, e.g. : out to be most successful". The polymerization yielded a linear monodispersed polybutadiene containing 80% of cis-linkages and hydrogenation produced linear monodispersed polyethylene. The variety of transition metals and profusion of and a bimolecular one known as chain transfer to available ligands make this kind of polymerization monomer, e.g. : most promising. For example, while the Ti catalyst has to be dried and kept in an inert atmosphere a rutenium based catalyst, described later", allows for polymerization in aqueous solutions. Let it be remarked, passingly, that such catalysts led to Both these reactions have to be retarded to allow polymerization of cyclo-octatetraen to poly- for the formation of living polymers. This calls for acetylene a polymer conducting on doping. A reduction of charge density on the protons which recent review of this field was published by could be achieved by increasing the charge density Schrock". on the carbon atom by strengthening its interaction Let me close this presentation with a discussion of with the counter-ion. However, such an increase, the recently developed living cationic polymeriza- which might lead eventually to the formation of a C tion. The living cationic polymerization was report- -A covalent bond , retards also the propagation, an ed in 1984 by Higashimura and his coworkers" and undesirable result. Hence, a compromise has to be 2 years later a similar claim was made by Kennedy struck, as always in life. and his associates". Both systems have common Let it be remarked at this point that it is beneficial features deserving a thorough examination. to have an indefinitely stable molecule (an ester) In my very first paper' introducing the concept of that is in equililibrium with the labile carbenium ion living polymers I stated explicitly : "living polymers having a limited lifetime 1- that determines its shelf are not immortal". What is demanded, therefore, time. from polymers referred to as living? Two conditions have to be met : (1) they have to last for a long time (long shelf time) , (2) their propagation should allow for the addition of many monomer molecules without being interrupted by termination or chain- transfer. Keeping it in mind let see what are the In such a system the lifetime (shelf time) of the problems perturbing a living cationic polymeriza- polymer increases, being given in the absene of tion monomer by K-'17- instead of -r. Therefore, we Let us ask first why a conventional cationic might desire to have K as small as possible. But polymerization is not living. Consider for example then a problem arises, a small K means a low the polymerization of isobutene initiated by SbC15 fraction of the actually growing species and hence a and an equivalent amount of water. The propagat- too slow polymerization. Consequently, a compro- ing species has the structure : mise has to be struck again, K0 provides an indefinitely stable but dead polymer. CH2C+(CH3) 2, SbC150F1 The value of K affects in a way the charge Although the positive sign is placed on the carbon density on the p hydrogen atoms and the rate of HA atom, the charge is spread throughout the whole elimination. The decreasing nucleophilicity of molecule, being quite substantial on the ,8 protons counter-ion A- and increasing basicity of the mono- (--1O%). This leads to two undesired reactions that mer slows down this undesired process and destroys the activity of the end-group and shorten increases the rate of propagation. Keeping this in

高分子40巻12月 号(1991年)803 mind, Higashimura and Sawamoto realized that stabilization of I- (the counter-ion) is needed to convert the non-living polymerization of vinyl ethers initiated by HI into a living one. This goal was achieved by employing iodine (1— >13i39" or the more efficient ZnI," as the anion stabilizing agent. In fact, Higashimura stressed the necessity of adjustment of the nucleophilicity of the anion to the basicity of the monomer in order to produce living polymers. A similar situation arises in the living polymeriza- tion of isobutene studied by Kennedy. The dormant polymer is the t-chloride, ---CH2C (CHO —C1 ABA block polymers in which the A block forms (this was not realized in the early work of Kennedy a discreate mezo-phase acting as a giant cross-link. when the acetate or ether was used as the initial starting material) . The nucleophilicity of Cl- is too high making the chloride inactive, virtually un- Under such conditions the resulting polymers are ionized. When the Cl- are complexed with BO, or composed of giants and liliputs. Keeping this TiC14 anions of low nucleophilicity are formed. This requirement in mind, we realize that it is beneficial renders the transfer sufficiently slow and the propa- for k, and kc to be high, while their ratio, K, is kept gation sufficiently fast, but leaves most of the poly- constant. mers in their inactive dormant state resulting in a Finally, another method yielding living polymers long shelf time. was contemplated", namely stabiliztion of the Solvating power of the medium is another factor growing carbenium ion by Lewis bases. The under- affecting the process. Increasing polarity of the lying principle of this method is not clear. The solvent increases the rate of propagation but also of complexing of carbenium ion with ethers or amines transfer. For example, vinyl carbazole (VCZ) is an converts it partially into stable but hardly reactive extremely basic monomer and its polymerization in onium ions. Thus an equilibrium is established, e.g., toluene yields living polymers even with the non- stabilized I- anion" but not with the more nucleo- philic C1-. However, in a more polar methylene chloride I- anion becomes too nucleophilic and has to be stabilized by the addition of quaternary ammo- nium salts to yield living polymer. The low solubil- retarding polymerization, increasing the shelf time ity of the readily polarized vinyl carbazole and the of the system, but hardly affecting the ratio of the tendency of its polymer to aggregate introduces rates of trnsfer to propagation. The beneficial effect complications that confuse the issue. of this procedure might be due the shortening the Let us pass to some kinetic problems. We wish to activity period of the actually growing polymers, keep the value of K at the appropriate level. Should the effect mentioned previously. One may speculate it be done by decreasing k or increasing lig In a also that in the process of conversion of the inactive process involving the active and dormant polymers onium ion into the active carbenium ion the Lewis the rate of exchange between them has to be fast, base is still in the vicinity of the latter and sterically otherwise the molecular mass distribution widens. prevents the elimination of HA or the transfer to This requirement is easily appreciated when one monomer. Whatever the explanation the effect is considers a system in which the rate of exchange is real and was confirmed by other investigators". so slow that the dormant polymers had no chance to The interesting effect of monomer on the longev- grow before all the monomer had been consumed. ity of living polymers was discussed in recent

804高 分 子4◎ 巻12月 号(1991年) papers". A clear example of such a phenomenon and J. Smid : Elsevier P131. was reported recently by Choi, Sawamoto and 19) Ph. Teyssie et al. : Makromol. Chem., Symp., 32, 61, 1990 Higashimura" who studied the decay of living 20) M.T. Reetz et al, : Angew. Chem., 27, 1373, 1988. polyvinyl ether in the absence of monomer and 21) M. Szwarc : Adv. Polymer Sci., 49, 1, 1983. observed their stabilization by its presence. The 22) a. C.E.H. Bawn, R.M. Bell, A. Ledwith : Polymer, 6, most probable explanation invokes the dual charac- 95, 1965. ter of vinyl ethers which may act as Lewis bases and b. M.P. Dreyfuss, P. Deyfuss : Polymer, 6, 93, 1965, J. Polymer Sci., 4, 2179, 1966. as vinyl monomers. Thus the initial interaction 23) T. Saegusa : Makromol. Chem., 175, 1199, 1974, 177, yields an oxonium ion which rearranges under the 2271, 1976. influence of another oncomming monomer into the 24) S. Penczek, P. Kubisa, K. Matyjaszewski : Adv. carbenium ion. This complex mechanism is Polymer Sci., 37, 1, 1980. 25) R. Sakata, T. Tsuruta : Makromol. Chem., 40, 64, proposed since the kinetic studies showed the propa- 1960. gation to be first order in the monomer. 26) E.J. Vanderberg : J. Polymer Sci., 47, 486, 1960. In conclusion, I hope that this brief review of 27) M. Osgan, Ph. Teyssie : Polymer Lett., B5, 789, 1967. living polymerization may induce you to further 28) a. T. Aida, S. Inoue Macromolecules, 14, 1162, 1166, 1981. studies of this interesting systems understanding of b.S. Inoue et al.: Macromolecules, 21, 1195, 1988. which was greatly foster by Japanse workers. 29) O.W. Webster et al. : J. Amer. Chem. Soc., 105, 5783, 1983. References 30) R.H. Grubbs, L.R. Gilliom : J. Amer. Chem. Soc., 108, 733, 1986. 1) M. Szwarc, M. Levey,R. Milkovich : J. Amer. Chem. Soc., 78, 2656, 1956. 31) K.J. Ivin "Olefin Metathesis" 1985, Academic Press, Publ. 2) M. Szwarc : Nature, 178, 1168, 1956. 32) R.H. Grubbs, W. Tumas : Science, 243, 907, 1989. 3) Y. Ishihama, M. Sawamoto, T. Higashimura : Polym. Bull, 23, 361, 1990. 33) K.C. Wallace, R.R. Schrock : Macromolecules, 20, 448, 1987. 4) L. Lochmann et al. : J. Polymer Sci., 17, 1727, 1979. 5) 0. Nuyken, H. Kroner : Makromol. Chem., 191, 1, 34) J.S. Murdzek, R.R. Schrock Macromolecules, 20, 2460, 1987. 1990. 6) G. Holden, R. Milkovitch : U.S. Patent 3, 231, 635, 35) R.H. Grubbs et al. : Polymer Prepr., 32, 1, 467, 1991. 1966. 36) B.M. Novak, R.H. Grubbs : J. Amer. Chem. Soc., 110, 7542, 1988. 7) J.P. Kennedy et al. : Polymer Prepr., 32, 1, 310, 1991. 37) F.L. Klavetter, R.H. Grubbs : J. Amer. Chem. Soc., 8) E. Vanzo : J. Polymer Sci., 4, 1727, 1966. 9) P. Rempp, E. Franta, J.E. Hertz : Adv. Polymer Sci., 110, 7807, 1988. 86, 145, 1988. 38) R.R. Schrock : Acc. Chem. Reserch, 23, 158, 1990. 10) a. D. Geiser, Hocker : Macromolecules, 13, 653, 1980. 39) a. M. Miyamoto, M. Sawamoto, T. Higashimura : b. T. Roovers, M. Toporowski : Macromolecules, Macromolecules, 18, 265, 1984. b. M. Sawamoto et al. : Macromolecules, 20, 916, c. G. Hild, C. Strazielle, P. Rempp : Europ. Polymer 2693, 1987. J., 19, 721, 1983. 40) a. R. Faust, J.P. Kennedy : Polymer Bul., 15, 317, 11) T.E. Hogen-Esch : to be published. 12) R.P. Quirk et al.: Makromol. Chem. Symp., 32, 47, 1986. 1990. b. M.K. Mishra, J.P. Kennedy : Polymer Bul., 17, 7, 1987. 13) G.O. Schulz, R. Milkovitch : J. Appi. Polymer Sci., 27, 4773, 1982, J. Polymer Sci., 22, 3795, 1984. c. J.P. Kennedy J. Macromolec. Sci., A24, 933, 1987. 14) B. Ivan, J.P. Kennedy, P.W. Mackey : Polymer 41) a. M. Sawamoto, C. Okamoto, T. Higashimura : Macromolecules, 20, 2693, 1987. Prepr., 33, 2, 215,1990. 15) A. Deffieux, P. Sigwalt, S. Boileau : Europ. Polymer b. T. Higashimura, et al. : Makromol. Chem., Symp., 13/14, 457, 1988. J., 20, 77, 1984. 16) P. Sigwalt, S. Boileau :J. Polymer Sci., Symp., 62, 51, 42) O.W. Webster et al. : Macromolecules, 23, 1918,1990. 1978. 43) M. Szwarc : "Recent Advances in Anionic Polymer- ization" p.93, 1987, T.E. Hogen-Esch and J. Smid, 17) C.J. Chang, R.F. Kiesel, T.E. Hogen-Esch J. Amer. Chem. Soc., 95, 8446, 1973. edts., Elsevier Publ. 18) A.H.E. Muller : in "Recent Advances in Anionic 44) W.O. Choi, M. Sawamoto, T. Higashimura : J. Polymerization" p. 205, 1987, Edts. T.E. Hogen-Esch Polymer Sci., 28, 2924, 1990.

高分子40巻12月 号(1991年)805