RING SIZE AND SOLVENT EFFECTS ON SESQUIBICYCLIC HYDRAZINE SELF-ELECTRON-TRANSFER RATES

Self-electron-transfer (self-ET) rate constants k(ex) for sesquibicycl ic hydrazines were determined by NMR line broadening at 25 degrees C i n CD3CN: 6(0/+)(21/21), 18 500 +/- 500; 4(0/+)(21/u22), 2290 +/- 90; 2 (0/+)(22/u22), 12 100 +/- 100 +/- 100; 7(0/+)(23/u22), 3370 +/- 140; 8 (0/+)(22/u23), 3840 +/- 110; 3(0/+)(22/22), similar to 700 (dec); and 9(0/+)(23/22), 580 +/- 25 M(-1) s(-1). Differences in Delta G(ex)(doub le dagger) for the first five compounds correlate linearly with differ ences in inner shell vertical relaxation enthalpies calculated by AM1, and it is concluded that changes in Marcus lambda(in) dominate the ch anges in k(ex). The last two compounds show smaller k(ex) values than predicted by AM1, which incorrectly calculates the twist at the NN bon d of the neutral forms. The effects of solvent on k(ex) are reported f or 8(0/+)NO(3-)(22/u23), solvent polarity range chloroform to methanol , and 4(0/+)PF(6-)(21/u22) and 8(0/+)PF(6-)(22/u23), solvent polarity range pyridine to acetonitrile. Plots of the Eyring free-energy barrie r Delta G(ex)(double dagger) vs solvent polarity parameters E(T)(30) a nd gamma give estimates of the ratio of the inner sphere to total Marc us reorganization barrier, lambda(in)(T)/lambda(T), of 0.78 (vs E(T)(3 0)) and 0.75 (vs gamma) for 2(0/+)(22/u22) and 0.82 (vs E(T)(30)) and 0.75 (vs gamma) for 4(0/+)(21/u22). The ratios obtained are unreasonab ly large for the 8(0/+)(22/u23) data (in the range 0.85-0.89), and it is concluded that 22/u23(+) is unstable enough under the experimental conditions to make these ratios inaccurate. It is concluded that self- ET for sesquibicyclic hydrazines is strongly diabetic.