Rational Design of Benzyl-Type Protecting Groups Allows Sequential Deprotection of Hydroxyl Groups by Catalytic Hydrogenolysis

Abstract

Benzyl protection of a hydroxyl group is one of the most frequently used procedures in synthesis because of the mild conditions involved in its removal by catalytic hydrogenolysis.1-3 The synthesis of polyhydroxylated compounds often requires orthogonal protecting strategies to distinguish between hydroxyl groups. It would be highly desirable to develop a range of benzyl-type protecting groups with different reactivities that can be sequentially removed via catalytic hydrogenolysis. This requires a detailed understanding of the mechanism of the cleavage of the benzyl oxygen bond by the palladium hydrogen species. Recently, we have determined the amphipolar nature of the palladium hydrogen bond (modes a, Mδ+ Hδ-, or b, MδHδ+) in both homogeneous4 and heterogeneous5 hydrogenation of alkenes. This has led us to test whether the electronic properties of the aromatic group can influence the rate of cleavage, which should in turn guide the development of hydroxyl protecting groups with different reactivities. The results in Table 1 show that the rate of debenzylation can be dramatically affected by the electronic properties of the aromatic ring. The substitution of the electron-withdrawing trifluoromethyl group onto the aromatic ring severely retards debenzylation under 1 atm of hydrogen. In contrast, there is considerable acceleration by electrondonating substituents, which suggests that the benzylic carbon bears a partial positive charge in the transition state. The hydrogenolysis of benzyl alcohols carried out in acetic acid has shown that protonation of the hydroxyl group is essential for the cleavage of the carbon-oxygen bond.6 Under the neutral conditions in our study, the reaction may occur by protonation of the benzyl oxygen atom, through the operation of mode b, MδHδ+, to give a positively charged benzylic carbon. Alternativly, it is possible that palladium could act as a Lewis acid and coordinate to the benzyl oxygen atom to promote the same electron-deficient transition state (mode a, Mδ+ Hδ-). The large difference in reactivity within this range of substituted benzyl groups suggests that they can be sequentially deprotected, therefore proving useful in multistep synthesis. To test the synthetic application of these groups, competition experiments were conducted on model systems with two differently substituted benzyl groups attached to ethanediol (Scheme 1a). Surprisingly, the benzyl group was cleaved first in competition with any of the substituted benzyl groups. This phenomenon has been observed with the 4-methoxybenzyl group (PMB); however, no explanation was proposed.7,8 The results with the linker experiments (Scheme 1a) seem to contradict those obtained when only one benzyl group is involved (Table 1). Surface scientists have determined that the aromatic ring lies flat on the metal surface for optimal coordination.9,10 It is possible that substitution on the aromatic ring could have an adverse steric effect that would interfere with the planar geometry required for effective binding and thus reduce its affinity for the metal surface. The linker experiments show that the limited number of active sites on the palladium surface could lead to a competition for adsorption sites between substituted and unsubstituted benzyl groups. This may explain why the least substituted benzyl group, although not electronically favored, can still be preferentially cleaved. It is clear that for the rational design of selective benzyl type protecting groups both electronic factors and adsorption must be taken into account. For synthetic purposes, it would be desirable to find a more labile group than the benzyl group for protection of the hydroxyl functionality. We anticipated that the 2-naphthylmethyl (NAP) group would fulfill these criteria: it is electron rich and should have a (1) Greene, T. W.; Wuts, P. G. M. In Protective Groups in Organic Synthesis; John Wiley & Sons, Inc.: New York, 1991. (2) (a) Czernecki, S.; Georgoulis, C.; Provelenghiou, C. Tetrahedron Lett. 1976, 3535. (b) Iverson T.; Bundle K. R. J. Chem. Soc., Chem. Commun., 1981, 1240. (3) Czech, B. P.; Bartsch, R. A. J. Org. Chem. 1984, 49, 4076. (4) Yu, J.; Spencer, J. B. J. Am. Chem. Soc. 1997, 119, 5257. (5) Yu, J.; Spencer, J. B. J. Org. Chem. 1997, 62, 8618. (6) Kieboom, A. P. G.; De Kreuk, J. F.; Van Berkum, H. J. Catal. 1971, 20, 58. (7) Srikrishna, A.; Viswajanani, J. A.; Sattigeri, J. A.; Vijaykumar, D. J. Org. Chem. 1995, 60, 5961. (8) Sajiki, H.; Kuno, H.; Hirota, K. Tetrahedron Lett. 1997, 38, 399. (9) Lin, R. F.; Koestner, R. J.; Van Hove, M. A.; Somorjai, G. A. Surf. Sci. 1983, 161. (10) Held, G.; Bessent, M. P.; Titmuss, S.; King, D. A. J. Chem. Phys. 1996, 11305. Table 1a