Anomalous phase shift between the angular dependence of the β 5′-H hfs constants produced by the phosphate groups in C4′ deoxyribose radicals: an electronic rather than a structural effect

Abstract

The angular dependence of the β 5′-H hyperfine splitting (hfs) constants in the 2′-deoxyribose 1′-amino 3′,5′-biphosphate C4′-radical (1) was previously computed to follow individually the relation a_H5i′ = B_0i + B_1i cos θ_i + B_2i cos²θ_i (i = a, b) at the UHF/6-31G//ROHF/3-21G level. The phase shift (Δφ = 141°) between the angular dependence of the two a_H5′ constants was found to be much larger than the normal value (Δφ = 120°). This apparently anomalous behavior was attributed to a distortion of geometry around C5′ owing to the presence of the phosphate groups. By contrast, it is demonstrated that a strongly electronegative β-substituent such as oxygen can produce a large phase shift. In the present case the anomalous phase shift is due exclusively to the electronic effect of the β-oxygen of the 5′-phosphate, whereas the 3′-phosphate has little effect. The a_H5i′ values depend strongly on the dihedral angle ω between the singly occupied C4′-2p_z orbital and the C5′-O bond and a unique relation, a_H5′ = A + B₁ cos θ + B cos²θ + C cos θ cos ω, accounts for the angular dependence of both the a_H5′ constants. The angular dependence does not change significantly upon replacing the phosphates with hydroxy groups both at the UHF/6-31G//ROHF/3-21G and B3LYP/6-311G**//UHF/6-31G* level. Thus, the poor agreement found between the hfs constants computed for this model radical and the experimental hfs constants of C4′ radicals determined with EPR spectroscopy can not be due to replacement of the phosphates with hydroxy groups as previously suggested.