“…41,42 Despite the well-known issue that B3LYP may be inadequate to account for the dispersion interaction which is a significant contribution to the interaction between CBPQT and TTF/ DNP, 16,19 it should still be adequate in describing the difference between the interactions at different oxidation states, since the variation of the dispersion interaction for different oxidation states should be minor compared to the variation of the electrostatic interaction. At the final geometry, the solvation energy in AN (dielectric constant 35.69, density 0.7857 g/cm 3 at 20°C, 43 probe radius 2.18 Å) is also calculated in B3LYP/ 6-31G** (since the solvation energy is not very sensitive to the choice of basis set 42 ) using the Poisson-Boltzmann continuum-solvation model. 44,45 The following atomic radii are used (in Å): 42 H 1.15, C 1.9, N 1.6, O 1.6, F 1.682, P 2.074, and S 1.7 (which has been changed from 1.9 to reproduce the experimental reduction potentials of TTF).…”