J. MATER. CHEM., 1991, 1(3), 371-374 37 1 High-surface-area Resins derived from 2,3=Epoxypropyl Methacrylate cross-linked with Trimethylolpropane Trimethacrylate P. D. Verweij" and David C. Sherringtonb a Department of Chemistry, Gorlaeus Laboratories, Leiden University, P.O. Box 9502, 2300 RA Leiden, The Netherlands Department ofPure and Applied Chemistry, University of Strathclyde, Glasgow G1 IXL, UK Suspension copolymerization of 2,3-epoxypropyl methacrylate (glycidyl methacrylate, GMA) and 2-ethyl-2- (hydroxymethyl)-propan-l,3-diol trimethacrylate (trimethylolpropane trimethacrylate, TRIM) has been performed in several solvents as porogens, i.e. cyclohexanol-dodecan-1-01 9/1 v/v, octan-2-one, n-butyl acetate, pxylene, toluene, ethyl acetate, benzonitrile, cyclohexanone and dodecan-1-01.The GMA :TRIM ratio in the monomer mixture and the monomer: porogen ratio have been varied. The B.E.T. surface area, pore volume and number of unreacted carbon-carbon double bonds are strongly dependent on the solvent used. Both the B.E.T. surface area and the pore volume decrease rapidly with increasing GMA :TRIM ratio when the cyclohexanol-dodecan- 1-01 mixture is used as the porogen. In contrast, the monomer: porogen ratio hardly affects the B.E.T. surface area of the resulting copolymers. The pore volume decreases with increasing monomer: porogen ratio in the interval studied. A very large pore volume, i.e. 1.86 cm3 g-', was found when octan-2-one was used as the porogen, with GMA :TRIM =1:1 and monomer :porogen 1:3.Similarly a substantial B.E.T. surface area of ca. 175 m2 g-' was achieved using the same GMA:TRlM ratio and octan-2-one as the porogen with monomer :porogen of 1:2. Keywords: 2,3-Epoxypropyl methacrylate; Copolymer; Cross-linking; Surface area Cross-linked 2,3-epoxypropyl methacrylate-ethylene glycol dimethacrylate (GMA/EDMA) resins containing the epoxy group have been extensively studied. They have been used as ion exchangers after modification of the epoxy group with amines.'S2 Alternatively, diol-containing resins derived from these polymers can be applied as protecting groups for aldehydic functions through the formation of acetal linkage^.^ Horak et al. investigated the effect of a number of polymeriz- ation variables, e.g.the concentration of the porogen in the dispersed phase, the concentration of the cross-linking agent in the monomer mixture and the polymerization temperature, on the specific B.E.T. surface area, the pore size, the porosity and the mechanical properties of GMA/EDMA polymer^.^ These properties are expected to be of great importance in the kinetic performance of ion exchangers. In general terms, rather high levels of cross-linker are required in order to produce resins with high surface areas,' and as a result the content of functional comonomer has to be restricted. In order to allow the maximum content of GMA simultaneously with a high surface area it was decided to investigate resins in which the trifunctional cross-linker TRIM was used in place of EDMA.Rosenberg et aL6 studied the physical properties of the products of the polymerization of TRIM and the copolymeriz- ation of TRIM and methyl methacrylate (MMA). B.E.T. surface areas were found to vary with the solubility parameter of the inert solvent used in the dispersed phase. High B.E.T. surface areas were found for the TRIM homopolymers; how- ever, for the TRIMiMMA copolymers the B.E.T. surface area decreases rapidly with the amount of MMA in the monomer mixture. Recently, Walenius and Flodin reported the synthesis of a GMA/TRIM polymer and the reaction of the epoxy substitu- ents with aliphatic amino compound^.^ In this paper the effect of some polymerization variables in the copolymerization of TRIM and GMA on the physical properties of the products is described.The polymers have been studied by surface area (B.E.T.) measurements, mercury porosimetry and solid-state I3Ccross-polarization and magic- angle spinning nuclear magnetic resonance spectroscopy (CP MAS 13C-NMR). Experimental Starting Materials All monomers and solvents were commercially available and were used without further purification. Polymerization The suspension polymerization mixtures consisted of GMA and TRIM and an inert solvent or solvent mixture as a porogen. The following solvents were used: cyclohexanol- dodecan- 1-01 9/1 v/v, octan-2-one, n-butyl acetate, p-xylene, toluene, ethyl acetate, benzonitrile, cyclohexanone and dodecan-1-01. The aqueous phase was a solution (0.1 wt.%) of Biozan Gum R (xanthan gum, Hercules Powder) in water.The GMA :TRIM ratio was varied from 1:3 to 3 :1 and the monomer :porogen ratio from 1 :1to 1 :3. The polymerization took place at 80 "C for 8 h and the products were purified by extraction in a Soxhlet apparatus with acetone and dried in UQCUO at 60 "C. More details of the polymerization procedure have been published already.2" B.E.T. Surface Area Measurements, Mercury Porosimetry and CP MAS 13C-NMR The B.E.T. surface area was obtained from N2 absorption measurements according to the B.E.T. method' using a Micro- meritics Accusorb 2100E. The pore volume and pore size were determined by mercury porosimetry using a Micromeritics Autopore I1 9220. Solid-state CP MAS 13C-NMR (cross- polarization and magic-angle spinning) was applied to deter- mine the amount of unreacted carbon-carbon double bonds in the polymers.The spectra were obtained on a Bruker MSL 400 spectrometer, operating at 100.6 and 400.1 MHz for I3C and 'H, respectively. The sample spinning rate was 5000 Hz. 3 72 J. MATER. CHEM., 1991, VOL. 1 The cross-polarization contact time was 0.8ms,6 with 3 s recycle delays between successive scans. Generally 1200 scans were employed. Results and Discussion Physical parameters of the polymers are listed in Table 1. Fig. 1 shows the B.E.T. surface area and the pore volume as functions of the amount of GMA in the monomer mixture with monomer :porogen = 1:2 and the cyclohexanol-dodecan-1-01 mixture as the porogen.Both the B.E.T. surface area and the pore volume were found to decrease rapidly with increasing GMA concentration, i.e. falling TRIM content. The pore-size distribution for P4 (GMA :TRIM =3 :1) (Fig. 2) shows a sharp maximum at 0.035 pm, which was found to broaden and to be displaced towards larger pore sizes when the amount of GMA in the monomer mixture decreased. Furthermore, the distributions indicate that a considerable number of pores with diameters of ca. 0.015 pm is mainly responsible for the large surface areas of P2 (GMA :TRIM = pore diametedpm Fig. 2 Pore size distribution for values of GMA (%) in the monomer mixture: (a) 25%; (b) 50%; (c) 75%; monomer :porogen = 1 :2; por- ogen =cyclohexanol-dodecan- l-ol 9/1 v/v 1 :3) and P3 (GMA:TRIM= 1 :l), while the large pore vol- umes of these polymers are mainly due to pores with diameters larger than 0.030nm.Fig. 3(a)and (b)show the B.E.T. surface area and the pore volume as a function of the monomer concentration in the organic phase for cyclohexanol-dodecan- l-ol 9/ 1 and octan- 2-one as the porogen, respectively, and GMA :TRIM = 1 :1. I I I I 20 40 60 80 100 The monomer concentration was found to have a large effect GMA in monomer mixture (YO) on the pore volume. When octan-2-one was used as the porogen a very large pore volume, i.e. 1.86 cm3 g-', was Fig. 1 B.E.T. surface area (0)and pore volume (0)us. GMA (%) found with monomer :porogen = 1:3. The effect of the in the monomer mixture; monomer :porogen = 1:2; porogen =cyclo-monomer :porogen ratio on the B.E.T.surface area was much hexanol-dodecan- 1-01 9/ 1 v/v less pronounced, suggesting that mainly large pores are formed Table 1 Physical and structural parameters of the GMA/TRIM polymers polymer GMA:TRIM M: P" solvent pore vol./cm3 g-' surface areab/m2 g-' surface areac/m2 g-' NMR unreacted C=C (%) P1 1:l 1:1 cycl-dod 9/1 0.38 121 144 3 P2 1:3 1:2 cyd-dod 9/1 1.28 339 267 6 P3 1:l 1 2 cycl-dod 9/1 1.13 1 40 223 2 P4 3: 1 1:2 cycl-dod 9/1 0.57 41 130 0 P5 1:1 1:3 cycl-dod 9/1 1.12 128 120 2 P6 I:1 1:I octan-2-one 0.65 127 173 5 P7 1:1 I :2 octan-2-one 1.31 174 245 3 P8 3: 1 1:2 octan-2-one 0.97 39 73 0 P9 1:l 1:3 octan-2-one 1.86 149 225 1 P10 1:1 1:2 n-butyl acetate 1.27 170 199 3 P11 1:l 1 :2 p-xylene 1SO 139 266 2 -P12 3: 1 1:2 p-xylene 1.47 2 51 P13 1:l 1 :2 toluene 1.02 145 192 4 P14 1:l 1:2 ethyl acetate 0.66 110 176 4 P15 1:l 1:2 benzonitrile 0.07 <1 44 4 P16 1:l 1:2 cyclohexanone 0.16 0.2 82 4 d d d 5P17 1:l 1:2 dodecan- l-ol 'M :P =monomer :porogen; determined by N2 adsorption according to the B.E.T.method; determined by mercury porosimetry; not determined, very fine powder. J. MATER. CHEM., 1991, VOL. I -I I I 1 I 1 I I 20 30 40 50 20 30 40 50 monomer in organic phase (%) Fig. 3 B.E.T. surface area (0)and pore volume (0)us. monomer (%) in the organic phase; GMA :TRIM = 1 : 1; porogen =cyclohexa-nol-dodecan-1-01 9/l v/v (a),octan-2-one (b) when the monomer concentration is low.This is confirmed by the relatively large average pore diameter when monomer :porogen = 1 :3, i.e. 37 and 33 nm for cyclohexanol- dodecan- l-ol 9/ 1 v/v and octan-2-one as the solvent, respect- ively, and also by the pore-size distributions (Fig.4A and B for the cyclohexanol-dodecan- 1-01 mixture and octan-2-one as the solvent, respectively). Solid-state CP MAS 13C-NMR spectroscopy was used in order to determine the amount of unreacted carbon-carbon double bonds. Carbonyl groups conjugated with a double bond have a lower chemical shift (166 ppm) than the unconju- gated, reacted ones (176ppm).6 In Table 1 the amount of unreacted methacrylate groups is listed for the polymers. The number of unreacted double bonds was found to be low and dependent on the porogen used during the polymerization. The values are lower than those normally found with styrene/ divinylbenzene (DVB) resins,' and confirm the observations of Rosenberg et aL6 The difference may be associated with the enhanced flexibility of the TRIM cross-linker versus DVB.Fig. 5 shows the dependence of the amount of unreacted double bonds on the amount of GMA in the monomer mixture. For GMA :TRIM =3 :1 no double bonds remain in the resin after polymerization, while the number of double 0 25 50 75 GMA in monomer mixture (%) Fig.5 Number of unreacted double bonds us. GMA (YO)in the monomer mixture; monomer :porogen = 1:2; porogen =cyclohexa-nol-dodecan- l-ol 911 v/v bonds increases with increasing amount of TRIM, as found for the copolymerization of TRIM and methyl methacrylate (MMA) in ethyl acetate,6 where the number of unreacted double bonds reaches a maximum of 16.7% in the case of TRIM= 100 vol.%.In Fig. 6 the number of unreacted double I I I I 1 1 0.1 0.01 1 0.1 0.01 20 30 40 50 pore diameteripm monomer in organic phase (%) in the Fig.6 Number of unreacted double bonds us. monomer (YO)Fig. 4 Pore size distribution for values of the monomer (YO) in the organic phase: (a) 25%; (b) 50%; (c) 75%; GMA: TRIM = 1: 1; por-organic phase; GMA :TRIM = 1:1; porogen =cyclohexanol-ogen =cyclohexanol-dodecan- 1-01 9/1 v/v (A), octan-Zone (B) dodecan-1-01 9/1 v/v [0,(a)] or octan-2-one [O, (b)] bonds is drawn as function of the amount of monomer in the organic phase for cyclohexanol-dodecan- 1-01 (a)and octan- 2-one (b)as the solvents.In both cases the number of unreacted double bonds was found to increase with increasing monomer concentration. This was previously found by Rosenberg et aL6 for the polymerization of TRIM, and is probably due to the effect that a larger amount of solvent introduces an improved mobility of the methacrylate substituents during the polymerization. Conclusions It can be seen from the results that the copolymerization of GMA and TRIM in different solvents yielded a wide variety of polymers. Some of the polymers, i.e. those made in octan- 2-one, n-butyl acetate, p-xylene, toluene and cyclohexanol- dodecan-1-01 9/1 v/v, show high porosity, while the beads made in benzonitrile and cyclohexanone were found to be non-porous.Both the B.E.T. surface area and the pore volume were found to decrease with increasing GMA:TRIM ratio in the monomer mixture when the cyclohexanol-dodecan- 1-01 mixture was used as the porogen. The pore volume also decreases with increasing monomer :porogen ratio in the organic phase. The B.E.T. surface area is a maximum at monomer :porogen = 1 :2. Using n-butyl acetate and octan-2- one as the porogen resins, substantial surface areas are achievable (170-175 m2 g-') while maintaining a GMA con- tent of 50%. We are indebted to the Netherlands Organization for Scientific Research (NWO) for their financial support and to Mr. C.Erkelens for his assistance with the NMR experiments. J. MATER. CHEM., 1991, VOL. 1 References (a) J. Kalal, F. Svec, E. Kalalova and Z. Radova, Angew. Makromol. Chem., 1976, 46, 93; (b) E. Kalalova, Z. Radova, J. Kalal and F. Svec, Eur. Polym. J., 1977, 13, 293; (c) E. Kalalovi, J. Kalal and F. Svec, Angew. Makromol. Chem., 1976, 54, 141; (d) F. Svec, H. Hrudkova, D. Horak and J. Kalal, Angew. Makromol. Chem., 1977, 63, 23; (e) F. Svec, D. Horak and J. Kalal, Angew. Makromol. Chem., 1977, 63, 37; (f)E. Kalalova, V. Beiglova, J. Kalal and F. Svec, Angew. Makromol. Chem., 1978, 72, 143; (g) F. Svec, J. Kalal, E. Kalalova and M. Marek, Angew. Makromol. Chem., 1980, 87, 95; (h) F. Svec, E. Kalalova, M. Tlusthkova and J. Kalal, Angew.Makromol. Chem., 1980,92, 133; (i) F. Svec and A. Jehlickova, Angew. Makromol. Chem., 1981, 99, 1I; (j)J. Kalal, E. Kalalovi, L. Jandova and F. Svec, Angew. Makromol. Chem., 1983, 115, 13; (k) F. Svec and A. Jehlickova, Angew. Makromol. Chem., 1984, 121, 127; (I) F. Svec, E. Kalalovi and J. Kalal, Angew. Makromol. Chem., 1985, 136, 183. (a) D. Lindsay and D. C.Sherrington, React. Polym., 1985, 3, 327; (b)D. Lindsay, D. C. Sherrington, J. Greig and R. Hancock, J. Chem. Soc., Chem. Commun., 1987, 1270; (c) D. Lindsay, D. C. Sherrington, J. Greig and R. Hancock, React. Polym., 1990, 12, 59; (d) D. Lindsay, D. C. Sherrington, J. Greig and R. Hancock, React. Polym., 1990, 12, 75. J. M. J. Frkchet, E. Bald and F. Svec, React. Polym., 1982, 1, 21. (a)D. Horik, F. Svec, M. Bleha and J. Kalal, Angew. Makromol. Chem., 1981, 95, 109; (b) D. Horak, F. Svec, M. Ilavsky, M. Bleha, J. Baldrian and J. Kalal, Angew. Makromol. Chem., 1981, 95, 117. R. L. Albright, React. Polym., 1986, 4, 155. (a)J-E. Rosenberg and P. Flodin, Macromolecules, 1986,19, 1543; (b)J-E. Rosenberg and P. Flodin, Macromolecules, 1987,20, 1518; (c) J-E. Rosenberg and P. Flodin, Macromolecules, 1987,20, 1522. M. Walenius and P. Flodin, Br. Polym. J., 1990, 23, 67. S. Brunauer, P. H. Emmett and E. Teller, J. Am. Chem. Soc., 1938, 60, 309. A. Guyot and M. Bartholin, Prog. Polym. Sci.,1982, 8, 277. Paper 0/05020B; Receiued 8th November, 1990