J. MATER. CHEM., 1991, 1(6), 1051-1056 Synthesis of Aromatic Polyethers by the Scholl Reaction Part 9.t-Cation-Radical Polymerization of 4,4'-Bis(2=naphthoxy)diphenylSulphone Virgil Percec* and James H. Wang Department of Macromolecular Science, Case Western Reserve University, Cleveland, OH 44106-2699, USA The synthesis and cation-radical polymerization of 4,4'-bis(2-naphthyloxy)diphenyl sulphone (3)are presented. For long reaction times, the soluble fractions of the polymers resulting from 3 exhibit a multi-modal molecular- weight distribution consisting of three peaks of very different molecular weights. The highest molecular-weight fraction usually has a number-average molecular weight (Mn)of ca. 106g rn0l-l relative to a polystyrene calibration, while the anof the lowest molecular-weight fraction is usually <10 000 g m0l-l.The multimodal molecular-weight distribution of the polymers synthesized from 3 was suggested to be due to the participation in reactions of several reactive positions on the 2-naphthyloxy rings. The reaction between the most reactive C, positions takes place during the early stage of the polymerization. However, once the concentration of C, positions decreases at a number-average molecular weight of ca. 10 000 g mol-l, intermolecular reactions between the less reactive positions such as C5, C, and C, of the low-molecular-weight polymer chains causes the formation of a branched polymer fraction of very high molecular weight and also a small amount of an insoluble fraction. A polymer fraction of intermediary molecular weight is also obtained.Keywords: Cation-radical polymerization; 4,4'-Bis(2-naphthyloxy)diphenyl sulphone; Aromatic polyether sul-phone ; Scholl reaction Polyether sulphones and pol yether ketones are conventionally synthesized by variants of aromatic nucleophilic substitution or aromatic electrophilic substitution reaction^.'^^ Two novel approaches to the synthesis of these functional aromatic pol yethers have been demonstrated by cation-radical poly-merization of bis(ary1oxy) derivatives by the Scholl reac-tion 10-18 and by the homocoupling of aryl halides by Nio- catalysed reaction^.'^.^^ The cation-radical polymerization of bis(4-phenoxyphenyl) sulphone, 4,4'-bis(pheny1thio)diphenyl sulphone, and bis(4-phenoxyphenyl) sulphone substituted with various electron-donating groups leads to pol yether sulphones of low molecular weight only." This is caused by the low poly- merizability of these monomers.The polymerization of 13-bis(phen0xy)pentanes substituted with methyl groups is also accompanied by proton-transfer reactions from the benzylic groups." The cation-radical polymerization of bis( 1 -naph- thyloxy) monomers leads to high-molecular-weight polymers when the monomers are bisC4-( l-naphthyl- ox y)phen y1] sulphone,' 4,4- bis( 1-naphth ylox y)benzophenone,' a,o-bis[4-( 1-naphthyloxyphenylsulphonyl]perfluoroalkane~,~~ bis[4-( l-naphthyl~xy)phenyl]methane,'~ 1,3-and 1,4-bis[4-( 1 -naphthyl~xy)phenylmethyl]benzene,'~ 2,2'-and 3,3'-bis(1-naphthylo~y)biphenyl,~~ 1,3-bis( 1 -naphthyloxy) benzene,15 and cc,o-bis(l-naphthyloxy)alkanes.12,21The bis( 1 -naphthyloxy) monomers investigated so far are summarized in Scheme 1.This article describes the synthesis and the cation-radical polymerization of the first member from the series of bis(2- naphthyloxy) monomers, i.e. bis[4-(2-naphthyloxy)phenyl] sulphone. Experimental Materials 2-Naphthol (1) (99%, Aldrich), FeCl, (anhydrous, Aldrich), and K2Co3 (Fisher) were used as received. Bis(4-chlorophenyl) p Part 8. V. Percec, J. H. Wang and Y. Oishi, J. Polym. Sci., Polym. Chem. Ed., in the press. sulphone (2, 98%, Aldrich) was recrystallized from toluene. Dimethyl sulphoxide (DMSO, Fisher) was distilled from CaO.Nitrobenzene (PhN02, Fisher) was distilled from CaH, under nitrogen. Other solvents were used as received. Techniques 200MHz 'H NMR spectra were recorded on a Varian XL- 200 spectrometer. All spectra were recorded in CDC1, with TMS as internal standard. Gel permeation chromatography (GPC) measurements were performed on a Perkin-Elmer series 10 LC instrument equipped with an LC-100 column oven, an LC 600 autosampler, and a Nelson Analytical 900 series data station. The measurements were made using a UV detector set at 254 nm, chloroform as solvent (1 ml min-', 40"C), a set of PL-gel columns (500 and 1048,), and a calibration plot constructed with polystyrene standards (Sup- elco). Purity was similarly determined by high-performance liquid chromatography (HPLC)/GPC using a 100 8, PL-gel column.A Perkin-Elmer DSC-4 differential scanning calor- imeter equipped with a TADS 3600 data station was used to determine the glass-transition temperatures of the polymers at a heating rate of 20 "C min-'. The glass-transition tempera- tures were read at the middle of the change in the heat capacity of the second heating scan. Synthesis of 3 2-Naphthol (20.00 g, 138.7 mmol) was dissolved in a mixture of 200 cm3 dry dimethyl sulphoxide and 67 cm3 dry toluene. K2C03 (23.00 g, 166.4 mmol) was added subsequently. The reaction flask was equipped with a nitrogen inlet, a ther-mometer, and a Dean-Stark trap equipped with a condenser. The reaction mixture was heated at 155 "C until no more water was collected in the Dean-Stark trap.Then the mixture was allowed to cool to 70 "C, and bis(4-chlorophenyl) sul- phone (18.33 g, 63.8 mmol) was added. The reaction mixture was stirred at 155 "C for 6 h. The cooled reaction mixture was poured into 2 dm3 cold water, and extracted with chloro- form three times (200 cm3 each). Chloroform was evaporated Ar= +SO2+ Scheme 1 Structure of bis( 1 -naphthyloxy) monomers polymerized by cation-radical reactions on a rotary evaporator. The resulting solid was dissolved in 100 cm3 toluene, washed five times with 2 mol dm-, aqueous NaOH, and then with water until neutral, and subsequently dried over anhydrous MgS04. The evaporation of toluene led to a light-red solid. The solid was purified twice by column chromatography (basic alumina, methylene chloride).Methyl- J. MATER. CHEM., 1991, VOL. 1 ene chloride was evaporated on a rotary evaporator and the resulting residue crystallized when left to stand, yielding 19.97g (62%) of 3. Purity (HPLC), 99.6%; m.p. 134-135 "C; 'H NMR (CDCI,, TMS): dH, 7.07 (d, 4 H, Ph-H meta to the SOz group), 7.21 (dd, 2 H, 1-H of the naphthalene unit), 7.44- 7.52 (m, 6 H, 3-, 6-, 8-H of the naphthalene unit), 7.75 (m, 2 H, 4-H of the naphthalene unit), 7.86-7.96 (m, 8 H, Ph-H ortho to SOz group and 5-and 7-H of the naphthalene unit). Polymerization Experiments All polymerizations were performed under nitrogen in dry nitrobenzene, using FeC1, as oxidant. The detailed polymeriz- ation conditions are summarized in Table 1.A typical poly- merization example is provided below. Compound 3 (0.45 g, 1.0 mmol) was dissolved in 1.0 cm3 dry nitrobenzene placed in a 25 cm3 three-necked flask equipped with a nitrogen inlet-outlet, and an addition funnel. A solution of 0.65 g FeC1, dissolved in 2.5 cm3 dry nitroben- zene was added dropwise under a stream of nitrogen over a 20 min period. The reaction mixture was stirred at 60 "C for 24 h. The content was precipitated into 200 cm3 methanol acidified with 2% HC1. The precipitate was filtered and washed with boiling methanol and dried in uacuo yielding 0.49 g (98%) of polymer. The polymer contains some insoluble fraction. The GPC chromatogram of the soluble fraction of the polymer displayed two peaks.The first peak represents 7% of total area: &fn=0.9 x lo6 g mol-', Mw&fn=3.4. The second peak represents 93% of total area; fin=4200 g mol-', &fw/&fn=2.9.In several cases the progress of the polymeriz- ation reaction was monitored by GPC as described below for experiment 7 from Table 1. After 5, 30, 60, 300 and 1020min, using a syringe fitted with a loin? needle and previously flushed with nitrogen, a sample of 0.1 cm3 of the polymerization mixture was with- drawn and subsequently precipitated into 2.0 cm3 methanol. The separated precipitate was washed twice with methanol. Methanol was evaporated and the resulting precipitate was dried in uacuo. The dried precipitate was analysed by GPC. Results and Discussion The synthesis of 3 is shown in Scheme2.It consists of the aromatic nucleophilic substitution of bis(4-chlorophenyl) sul- phone (2) by 2-naphthol (1). The cation-radical polymerization of 3 is presented in eqn. (2) from Scheme 3. The polymerization reactions were performed in nitrobenzene and were initiated by FeCl, oxi- dant. The expected structure of the repeating structural unit is shown as 4. The results of the polymerization experiments are summarized in Table 1. The most prominent character of the soluble fractions of t 1 in=2.54 cm. w 1 Scheme 2 Synthesis of 3 J. MATER. CHEM., 1991, VOL. 1 Table 1 Polymerization of 3; polymerization time =24 h ~~~ ~ monomer solution FeCl, solution polymer reaction experiment monomer/ temperature/ yield" M,bl no.mmol PhN02/cm3 FeCl,/mmol PhN02/cm3 "C (YO) g mol-' Mw/Mnb T,/ "C ~ ~ ~~ 1 1.o 1 .o 4.0 2.5 25 86 27 x lo6 1.1 (6%) - 7 100 -' (94%) 2 1.o I .o 4.0 2.5 60 98 0.9 x lo6 3.4 (7%) 233 4200 2.9 (93%) 3 1 .o 1.o 4.0 1.5 25 93 19 x lo6 1.2 (4%) 259 7700 -' (96%) 4 I .o 2.0 4.0 2.5 25 70 21 x lo6 1.2 (10%) - 7500 -' (go0/,) 5 1.o 2.0 4.0 4.0 25 71 18 x lo6 1.2 (6%) 274 9000 -' (94%) 6 1 .o 4.0 4.0 2.5 25 58 16 x lo6 1.3 (12%) 274 8700 -' (88%) 7 1.o 0.8 4.0 2.5 60 95 3.8 x lo6 2.3 (42%) - 3700 1.8 (58%) 8 1.o 2.0 4.0 2.5 60 93 20 x lo6 1.2 (7%) 0.9x lo6 1.8 (3%) - 5000 1.7 (goo/,) 9 1.o 1.o 4.0 2.5 100 99 64 x lo6 1.9 (8%) 2.3 x lo6 2.0 (2%) - 4300 1.7 (goo/,) " The polymer yield includes both the soluble and the insoluble fractions.The a,,and M,/M, were determiend from the soluble fraction. Multi-peak molecular weight distributions were observed on the gel permeation chromatograms. The a,and Mw/M, values were calculated and listed for the individual peaks. 'The peak overlapped with a minor peak. 3 L -In 4 Scheme 3 Cation-radical polymerization of 3 the polymers is their multi-modal molecular weight distri- bution. The polymer produced with long reaction times usually has one peak of very high molecular weight (Mn>l x lo6 g mol- '), a peak of intermediary molecular weight, and a low-molecular-weight peak (I@, <10 000 g mol-'). The a,, MW/M,and the area of the individual peaks are listed in Table 1.The GPC traces of several representative polymers derived from 3 are presented in Fig. 1. These polymers were synthe- sized under different experimental conditions. Curve (a) rep-resents the polymer (experiment 1, Table 1) synthesized at room temperature. The main peak has an M,value of 7100 g mol-'. It represents 94% of the total area. The minor peak (6% of total area) has an M, value of 27 x lo6 g mol-' and Mw/Kf,of 1.1 relative to polystyrene standards. Curve (b) displays the GPC trace of a polymer synthesized at the identical monomer and FeCl, concentrations but at an elev- ated temperature (60 "C, experiment 2, Table 1). The polydis- persity of the main peak is broadened.Curves (c) and (d) are I I I I5 10 15 20 elution volurne/cm3 Fig. 1 GPC traces of polymers synthesized from 3 under different reaction conditions: (a)experiment 1; (b)experiment 2; (c) experiment 4; (d)experiment 6; (e)experiment 7 (all from Table 1) the GPC traces of polymers synthesized at lower monomer concentrations. The highest percentage of the high-molecular- weight fraction was observed in curve (e), which has 42% of the high-molecular-weight fraction with M, of 3.8 x lo6 g ' mol-'. Besides the easily distinguishable peaks, the GPC traces shown in Fig. 1 also exhibit a shoulder located between the two major peaks. An example of a polymer with three distinguishable peaks on their GPC traces is shown in Fig. 2 (experiment 8, Table 1).The three peaks have calculated values of relative M, of 20 x lo6, 0.9 x lo6 and 5000 g mol- ',respectively. The ratio of the areas of these three peaks is 7 :3 :90. In order to observe the time sequence of the formation of the individual peaks of different molecular weights, the pro- gress of the polymerization was followed by taking samples I I I I5 10 15 20 elution volurne/cm3 Fig. 2 GPC trace of polymer derived from 3 (experiment 8, Table 1) at different reaction times. The GPC analysis of these collected samples provides the molecular-weight distribution of the polymers. An example is presented in Fig. 3 for the samples taken from experiment 7 (Table 1). Curve (a) displays the molecular-weight distribution of the sample taken at 5 min of polymerization. The polymer consists mostly of oligomers.There is no trace of high-molecular-weight fraction detectable by GPC. As the polymerization continues, the sample taken at 30 min (b)has a multi-modal molecular-weight distribution containing three main peaks. The polymer contains a fraction of broad low molecular weight, a fraction of very high molecular weight and a shoulder of intermediate molecular weight. The subsequent GPC traces are shown in (c)and (d) for t =300 and 1020 min. The molecular weight distributions of the peaks obtained at longer reaction times [(c) and (d)] are very similar to those obtained at 30 min (b)except for the decrease in the amount of oligomers. This indicates that the ~~I I I 10 15 20 elution volume/crn3 Fig.3 GPC traces of polymer samples taken at different reaction times, t/min: (a)5; (b)30; (c) 300, (d) 1020 (experiment 7, Table 1) J.MATER. CHEM., 1991, VOL. 1 resulting polymers change their molecular-weight distri-butions negligibly with the reaction time. The glass-transition temperatures of the polymers synthe- sized from 3, range from 233 to 274°C. The transition temperatures are influenced by the molecular weight and the molecular-weight distribution of the polymer samples. The 200 MHz 'H NMR spectrum of a representative poly- mer derived from 3 is presented in Fig. 4. The assignment of the resonances was based on the expected structural unit 4 (Scheme 3). In spite of the presence of polymers of totally different molecular weights, i.e.the high- and low-molecular- weight fractions, no differentiation can be made based on the 'H NMR spectra for the clear assignment due to each fraction. This is expected since the resonances of high-molecular-weight fraction should be low owing to their low concentration in the polymers (typically <lo%, Table 1). Consequently, the resonances resulting from the high-molecular-weight fraction probably overlap with those from the major resonances resulting from the low-molecular-weight fraction. The multi-modal molecular-weight distribution of the poly- mers synthesized from 3 is most probably due to the presence of several reactive positions in the 2-naphthyloxy ring of the monomer. These reactive positions have different reactivities and concentrations, and consequently may react to a different extent at different stages of the polymerization generating three independent polymeric species. The relative reactivity of the different ring positions of 2-methylnaphthalene cation radical was studied in nucleo- philic substitution reactions using cyanide anions.22 The iso- lated substitution product ratios are summarized in structure 5 in Scheme 4.This suggests the following order of reactivity: C1& C8zC5>C4.These positional reactivities were rational- ized by the molecular-orbital calculations.The electron density IQ ICHCI3 7.26 (c1 7.21-7.39\ I 6 Fig. 4 'HNMR spectrum (CDCl,, TMS, 200 MHz) of a representative polymer derived from 3 7% 60% 0.151 0.197mCH3 6% 4% 0.143 0.159 5 6 Scheme 4 Substitution products distribution and LUMO electron density of 2-methylnaphthalene J.MATER. CHEM., 1991, VOL. 1 1055 of the lowest occupied molecular orbital (LUMO)of 2-methyl- naphthalene cation radical is summarized as structure 6 in (3) Scheme 4.It was proposed that the nucleophilic attack took place preferably at the ring positions with high LUMO electron density. However, the positive charge distribution of 3 7 the 2-methylnaphthalene cation radical did not explain the observed relative reactivity.22 The C1 position is also the most Ar = reactive position on the 2-substituted naphthalene ring in the aromatic electrophilic substitution reactions.', Based on these results, we assume that the C1 position represents the most reactive position of 3 in the cation-radical polymerization.This assumption is supported by the fact that 1,l '-binaphthyL2,2'-diol is the predominant product from the oxidative dimerization of 2-naphth01.~~~~~ 7+3 The mechanism of the cation-radical polymerization of 3 is presented in Scheme 5. The single-electron transfer oxi- dation of monomer 3 by FeC1, oxidant leads to the monomeric cation radical 7, eqn. (3). The radical-substrate ~o~pling~~~~~ between the cation radical 7 and monomer 3 generates the 8 dimeric cation radical 8 eqn. (4). There are three possible pathways for the formation of neutral dimer The first pathway is shown in eqn.(S) and (6), and consists of the single-electron transfer between the dimeric and the mono- meric cation radicals, eqn.(9,leading to the dimeric dication 9, eqn.(S), followed by the subsequent elimination of two 8+7 +3 protons, eqn. (6). The second pathway refers to the oxidation of the dimeric cation-radical 8 by FeC13 oxidant to form the dication 9, eqn. (7). The first step of the third pathway is the elimination of one proton from the dimeric cation radical 8, 9 yielding a dimeric radical 11, eqn. (8). The dimeric radical 11 is subsequently oxidized to its cation 12 by FeCl,, eqn. (9). The single-proton elimination from 12 leads to the neutral dimer 10, eqn. 10. Owing to the high reactivity of the C1 position, the reaction 9 sequence shown in Scheme S occurs at the early stage of the polymerization. However, once the concentration of the C1 position of the 2-naphthyloxy rings is decreased corresponding to a number-average molecular weight of ca.10 000 g mol- (Fig. 3), competitive reactions can occur. 10 The low-molecular-weight polymer chains can undergo intermolecular reactions at the less reactive positions such as C5,C, and C4 which are now available in much higher 8 Fe3 9 concentrations than the position C1. An example for such a reaction is presented in Scheme 6 for the intermolecular dimerization between two C, positions. These intermolecular reactions can lead to soluble branched polymer fractions of very high molecular weight. The narrow polydispersity of the 8 -H+ highest molecular-weight fraction (Table 1) indicates the branched nature of this polymer fraction. The reactions between extensively branched soluble polymer fractions gener- ate the insoluble three-dimensional polymer network.A very low amount of insoluble polymer fraction was present in all 11 polymers that resulted from 3. At low concentrations of C1 groups cyclization reactions are also possible. Conclusions The C1 position of 3 is the most reactive position of 3. 11 (9) Therefore, the cation-radical polymerization of 3 takes place by dehydrogenative C1-Cl coupling of 3. However, once the concentration of C1 functional groups decreases at a number-average molecular weight of ca. 10 000 g mol-' the intermolecular reactions of the less reactive positions such as the C5, C, and C4 carbons of the naphthalene rings of the 12 4' - 12 10 low-molecular-weight polymer chains lead to a soluble frac- Scheme 5 Mechanism of the cation-radical polymerization of 3 tion of branched polymers of molecular weight of ca.lo6 g mol-These branched reactions generate a small concen- tration of insoluble polymer fraction. 1056 J. MATER. CHEM., 1991, VOL. I 3 J. B. Rose, in Recent Advances in Mechanistic and Synthetic Aspects of Polymerization, ed. M. Fontanille and A. Guyot, 2 -Ar 4 Reidel, Dordrecht, 1987, p. 207. R. May, in Encyclopedia of Polymer Science and Engineering, ed. H. F. Mark, N. M. Bikales, C. G. Overberger and G.Menges, 5 Wiley, New York, 2nd edn., 1988, vol. 12, p. 313. J. E. Harris and R. N. Johnson, in Encyclopedia of Polymer Science and Engineering, ed. H. F. Mark, N. M. Bikales, C. G. Overberger and G.Menges, Wiley, New York, 2nd edn., 1988, 13 6 vol. 13, p. 196. P. A. Staniland, in Comprehensive Polymer Science, ed. G. Allen 2 FeCS -2 FeC12, -2 CI-, -2 H+ 1 7 8 9 and J. C. Bevington, Pergamon, Oxford, 1989, vol. 5, p. 483. F. Parodi, in Comprehensive Polymer Science, ed. G. Allen and J. C. Bevington, Pergamon, Oxford, 1989, vol. 5, p. 561. Y. Imai and Y. Oishi, Prog. Polym. Sci., 1989, 14, 173. V. Percec, J. H. Wang and R. S. Clough, Makromol. Chem., Macromol. Symp., in the press. 10 V. Percec and H. Nava, J. Polym. Sci., Polym.Chem. Ed., 1988, 26, 783. I1 V. Percec, J. H. Wang and Y. Oishi, J. Polym. Sci., Polym. Chem. Ed., 1991, 29, 949. 12 V. Percec and J. H. Wang, Polym. Bull., 1991, 25, 9. 13 V. Percec and J. H. Wang, Y. Oishi and A. E. Feiring, J. Polym. Sci., Polym. Chem. Ed., 1991, 29, 965. (11) 14 V. Percec, J. H. Wang and S. Okita, J. Polym. Sci., Polym. Chem. Ed., 1991, 29, 000. 15 16 V. Percec, S. Okita and J. H. Wang, Macromolecules, submitted. V. Percec, J. H. Wang and S. Okita, J. Polym. Sci., Polym. Chem. Ed., submitted. 17 18 V. Percec and J. H. Wang, Makromol. Chem., Macromol. Symp., in the press. V. Percec, J. H. Wang and Y. Oishi, J. Polym. Sci., Polym. Chem. Ed., submitted. 19 I. Colon and G. T. Kwiatkowski, J. Polym. Sci., Polym. Chem. 14 20 Ed., 1990, 28, 367. M. Ueda and F. Ichikawa, Macromolecules, 1990, 23, 926. Scheme 6 The formation of branched polymers during the cation- radical polymerization of 3 Acknowledgment is made to Amoco Performance Products 21 22 23 R. G. Feasy, A. Turner-Jones, P. C. Daffurn and J. L. Freeman, Polymer, 1973, 14, 241. K. Yoshida and S. Nagase, J. Am. Chem. SOC., 1979, 101,4268. P. H. Gore, A. S. Siddiquei and S.Thorburn, J. Chem. SOC., Perkin Trans., 1972, 1, 1781. for financial support of this research. References 24 9.zLJ 26 27 H. Musso, Angew. Chem., 1963,75, 965. M. Tisler, Org. Prep. Proc. Znt., 1986, 18, 19. V. D. Parker, Ado. Phys. Org. Chem., 1983, 19, 131. 0. Hammerich and V. D. Parker, Adv. Phys. Org. Chem., 1983, 1 2 S. Maiti and B. K. Mandal, Prog. Polym. Sci., 1986, 12, 111. M. J. Mullins and E. P. Woo, J. Macromol. Sci.-Rev. Macromol. 20, 55. Chem. Phys., 1987, C27, 313. Paper 11032475; Received 1st July, 1991