1,2:5,6-Di-O-isopropylidene-3,4-bis-O-(4-vinylbenzoyl)-d-mannitol (1a) and -d-glucitol (1b) with styrene (St) were polymerized using AIBN in toluene at 60 °C to yield copolymers (2a and 2b, respectively) consisting essentially of the cyclic repeating and styrene units. In order to confirm asymmetric induction in 2, the removal of the template from 2 was carried out using KOH in aqueous MeOH, and then the resulting hydrolyzed copolymer was treated with diazomethane for conversion into poly[(methyl 4-vinylbenzoate)-co-styrene] (3). Copolymers 2a and 2b showed a negative specific rotation, whereas a positive rotation was observed for 3a and 3b. The CD spectrum of 1a exhibited a negative Cotton effect at 300.0 nm and a positive one at 267.3 nm as well as a negative exciton coupling of 1b. According to the exciton chirality method, 1a and 1b had a negative chirality and its benzoate chromophores, the two 4-vinylbenzoyl groups, were twisted counterclockwise. On the other hand, the CD spectrum of 3a (1a/St = 0.30/0.70) and 3b (1b/St = 0.27/0.73) showed a split Cotton effect with a positive one at 254.9 nm and a negative one at 237.1 nm for 3a and with a positive Cotton effect at 254.7 nm and a negative one at 232.7 nm for 3b. This result indicated that 3 had a positive chirality in which two 4-vinylbenzoyl groups were twisted clockwise, i.e., the vicinal (methyl 4-vinylbenzoate) groups in the main chain were an (S,S)-configuration.
The anionic cyclopolymerizations of 1,2:5,6-dianhydro-3,4-di-O-methyl-d-mannitol (1) and 1,2:5,6-dianhydro-3,4-O-isopropylidene-d-mannitol (2) have been carried out using potassium tert-butoxide (t-BuOK) and potassium hydroxide (KOH). For the polymerization of 1, a well-defined carbohydrate polymer consisting of (1→6) linked 2,5-anhydro-3,4-di-O-methyl-d-glucitol units was synthesized through a regio- and stereoselective mechanism. When a /[t-BuOK] molar ratio of 20 was used in toluene for 48 h, the yields and number-average molecular weights (Mn) of the polymers gradually increased with polymerization time. The Mn of the polymer varied with the molar ratio of monomer to initiator, and a linear relationship between them was found. The degree of polymerization was larger than that estimated from the molar ratio, resulting in an initiator efficiency of about 55%. KOH was also effective for converting monomer 1 to a gel-free polymer but was not as active as t-BuOK. The rate of polymerization was rather slow, and the polymerization was not complete, even after 100 h. The presence of a crown ether, 18-crown-6, in the cyclopolymerization allowed the Mn of the polymer to approach the value estimated from the /[t-BuOK] molar ratio. On the other hand, monomer 2 tended to form a gel in the polymerization process, so soluble polymers were isolated only at early stages of the polymerization.