Solvation effects on alanine dipeptide: A MP2/cc-pVTZ//MP2/6-31G** study of (Phi, Psi) energy maps and conformers in the gas phase, ether, and water.

The effects of solvation on the conformations and energies of alanine dipeptide (AD) have been studied by ab initio calculations up to MP2/cc-pVTZ//MP2/6-31G**, utilizing the polarizable continuum model (PCM) to mimic solvation effects. The energy surfaces in the gas phase, ether, and water bear similar topological features carved by the steric hindrance, but the details differ significantly due to the solvent effects. The gas-phase energy map is qualitatively consistent with the Ramachandran plot showing seven energy minima. With respect to the gas-phase map, the significant changes of the aqueous map include (1) the expanded low-energy regions, (2) the emergence of an energy barrier between C5-beta and alpha(R)-beta(2) regions, (3) a clearly pronounced alpha(R) minimum, a new beta-conformer, and the disappearance of the gas-phase global minimum, and (4) the shift of the dominant region in LEII from the gas-phase C7(ax) region to the alpha(L) region. These changes bring the map in water to be much closer to the Ramachandran plot than the gas-phase map. The solvent effects on the geometries include the elongation of the exposed N-H and C=O bonds, the shortening of the buried HN--CO peptide bonds, and the enhanced planarity of the peptide bonds. The energy surface in ether has features similar to those both in the gas phase and in water. The free energy order computed in the gas phase and in ether is in good agreement with experimental studies that concluded that C5 and C7(eq) are the dominant species in both the gas phase and nonpolar solvents. The free energy order in water is consistent with the experimental observation that the dominant C7(eq) in the nonpolar solvent was largely replaced by P(II)-like (i.e., beta) and alpha(R) in the strong polar solvents. Based on calculations on AD + 4H(2)O and other AD-water clusters, we suggest that explicit water-AD interactions may distort C5 and beta (or alpha(R) and beta) to an intermediate conformation. Our analysis also shows that the PCM calculations at the MP2/cc-pVTZ//MP2/6-31G** level give good descriptions to the bulk solvent polarization effect. The results presented in this article should be of sufficient quality to characterize the peptide bonds in the gas phase and solvents. The energy surfaces may serve as the basis for developing of strategies enabling the inclusion of solvent polarization in the force field.