Theoretical study of the enol imine <--> enaminone tautomeric equilibrium in organic solvents

The tautomeric enol imine <--> enaminone (phenol <--> quinone) equilibrium of the 1-hydroxy-2-naphthaldehyde Schiff base (2-phenyliminomethyl-naphthalen-1-ol) was investigated by density functional theory (B3LYP) and ab initio (MP2) methods in the IEF-PCM polarizable continuum dielectric solvent approximation and by a combined ab initio + FEP/MC study by considering an explicit solvent model. Special emphasis was put on the effect of solvation on this equilibrium by using an apolar (CCl4), polar aprotic (CH3CN), and polar protic (CH3OH) solvent. Compared with experimental tautomerization Gibbs free energies, the IEF-PCM/B3LYP calculations apparently overestimate the stability of the quinone form both when the 6-31G(d,p) and the 6-311++G(d,p) basis sets are applied. IEF-PCM/MP2 studies with the above basis sets predict the preference of the aromatic phenol tautomer, in contrast to the experiment in methanol and acetonitrile solvent. Calculation of the total relative free energy as DeltaG(tot) = DeltaE(int)(IEF-PCM/QCISD(T)/6-31G(d)) + DeltaG(solv, FEP/MC) + DeltaG(thermal) provided agreement with the experimental values up to 0.6 kcal/mol in the three solvents, and the predominant tautomer was always correctly predicted. In-solution relevant atomic charges, derived by a fit to the molecular electrostatic potential generated by the IEF-PCM/B3LYP/6-31G(d,p) wave function, show strong dependence on the fitting procedure (CHELPG or RESP) and are fairly insensitive to the chemical nature of the actual solvent. Use of the CHELPG charges in FEP/MC simulations revealed to be superior in comparison with the use of the RESP charge set.