UC Riverside

In the first project, we modeled low barrier hydrogen bond systems, which have a unique energy profile. Classification of a hydrogen bond as a low barrier hydrogen bond requires assignment of the experimental vibrational spectrum, as well as prediction of the zero point energy. We modeled these systems using a 2D potential energy surface that explored the symmetric and asymmetric stretch of the proton transit. The resulting predicted spectra and zero point energies helped classify three molecules as hydrogen bonds or low barrier hydrogen bonds.

In addition, solvation effects were explored. First, a traditional implicit polarizable continuum model, and standard explicit microsolvation techniques were used to explain why only two out of four isoenergetic gas phase conformers appeared in infrared multiple photon dissociation spectra. Our calculations suggest that this resulted from an experimental artifact. We showed that the two configurations were much more stable in water. Since the ion delivery system was electrospray, it was concluded that the solvent effectively determined the structure of the gas phase ions, since the barrier to isomerization was quite high.

While these simple models were able to explain those experiments, a much more robust and accurate model for solvent effects is needed, especially for systems with nonhomogeneous environments (such as protein binding sites). Solvent structure and features of the environment are completely neglected in both of the rudimentary models. Quantum Mechanical (QM)/ Molecular Mechanical (MM) sampling of solvent configuration space is effective, but extremely expensive, due to the high cost of the repeated QM calculations. Our model samples the configuration space of the solvent using inexpensive, but configurationally accurate, molecular dynamics, then includes the information from several timesteps in one single expensive QM calculation. A mathematically rigorous coarse-graining scheme translates and scales the multipoles, polarizabilities and the dispersions to grid points around the solute region, and then computes the energy of the system in a single QM/MM calculation. The results are extremely promising