[NH:1]1[C:10](=[O:11])[C:9]2[NH:8][CH:7]=[N:6][C:5]=2[N:4]=[C:2]1[NH2:3].S([O-])([O-])(=O)=[O:13].[NH4+].[NH4+].C[Si](C)(C)N[Si](C)(C)C.[C:28]([O:31][CH2:32][CH:33]([O:39][CH2:40]Cl)[CH2:34][O:35][C:36](=[O:38])[CH3:37])(=[O:30])[CH3:29].[C:42]1([CH3:48])C=CC=CC=1>>[C:42]([NH:3][C:2]1[NH:1][C:10](=[O:11])[C:9]2[N:8]=[CH:7][N:6]([CH2:40][O:39][CH:33]([CH2:34][O:35][C:36](=[O:38])[CH3:37])[CH2:32][O:31][C:28](=[O:30])[CH3:29])[C:5]=2[N:4]=1)(=[O:13])[CH3:48]A mixture of 2.57 g (18 mmole ) of guanine, 1.8 g of ammonium sulfate, and 126 ml of hexamethyldisilazane was stirred at reflux under N2. The solid gradually dissolved. After 2 days the solution was cooled and concentrated under high vacuum. The viscous, residual oil was dissolved in about 28 ml of dry toluene and maintained under N2 as a solution of 5 g (22.3 mmole) of chloromethyl 1,3-diacetoxy-2-propyl ether in 12 mol of dry toluene was added. The resulting solution was heated at reflux under N2 for 1.5 hours. It was then cooled, concentrated, and dried under high vacuum. The viscous, orange residual oil was treated with 30 ml of water and 30 ml of saturated sodium bicarbonate solution. The mixture was swirled with warming on a steam bath for 5 minutes, during which time the residue solidified. After cooling, the solid was collected on a filter and washed with a small volume of water. Although this cream-colored solid (4.6 g) gave a single spot on TLC (80:20:2 CHCl3 --MeOH--H2O), NMR showed that it contained 10-15% of the 7-alkylated isomer in addition to the desired 9-isomer. The material was combined with 0.6 g of similar material from other runs and was suspended in 184 ml of acetic anhydride. The mixture was heated at 97° for 18 hours by which time nearly all of the solid had dissolved, and TLC showed complete acetylation. The reaction was cooled and concentrated under vacuum. Treatment of the residue with 200 ml of methylene chloride gave a solid, which was isolated by filtration and washed once with methylene chloride. Recrystallization from methylene chloride gave 0.55 g of pure 9-isomer. The filtrates were passed through a column containing 40 g of silica gel. Elution with 97:3 CH2CL2 --MeOH gave 4.5 g of partially purified product. Upon recrystallization from methylene chloride (about 45 ml), 3.0 g of pure 9-isomer was obtained.