Close-Up Look at Electronic Spectroscopic Signatures of Common Pharmaceuticals in Solution

Abstract

Simulating electronic properties and spectral signals requires robust computational approaches that need tuning with the system's peculiarities. In this paper, we test implicit and fully atomistic solvation models for the calculation of UV−vis and electronic circular dichroism (ECD) spectra of two pharmaceutically relevant molecules, namely, (2S)-captopril and (S)-naproxen, dissolved in aqueous solution. Room temperature molecular dynamics simulations reveal that these two drugs establish strong contacts with the surrounding solvent molecules via hydrogen bonds. Such specific interactions, which play a major role in the spectral response and are neglected in implicit approaches, are further characterized and quantified with natural bond orbital methods. Our calculations show that simulated spectra, and especially ECD, are in good agreement with experiments solely when conformational and configurational dynamics, mutual polarization, and solute−solvent repulsion effects are considered.

The Role of Hydrogen Bonding in the Raman Spectral Signals of Caffeine in Aqueous Solution

“…, respectively, which are associated with moderate or mild interactions, unlike the strong interaction found for charge-assisted HBs [51,67]. As is also true for the reported UV-Vis and ECD spectra of captopril and naproxen [67], configurations with a large cumulative |E (2) d→a | have spectra that better match the experiments. The individual stabilization that those charge transfers provide to the system is quantified in Figure S1 in the Supplementary Materials.…”

“…This is carried out using the NBO analysis [54][55][56] in the NBO7 program [92]. Similar QM/FQ NBO analyses have been conducted for common pharmaceuticals [51,67] and hypoxanthine [48] in solution. Conventional bond critical points (BCP) of the electron density and localized Natural BCPs topological descriptors are computed with the Natural BCP (NBCP) analysis as implemented in NBO7 [92].…”

Computational Spectroscopy of Aqueous Solutions: The Underlying Role of Conformational Sampling

“…The modeling of the spectral signals originating from a molecular system dissolved in an aqueous solution is extremely challenging because of the presence of specific solute–solvent interactions, such as hydrogen bonding (HB). − The most reliable methods for these systems are quantum mechanics (QM)/molecular mechanics (MM) approaches, in which the solute is described at the QM level, while solvent molecules are modeled using a classical MM force field (FF). ,− These methods are generally coupled to a configurational sampling of the solute–solvent phase space, allowing for a description of the dynamical aspects of the solvation phenomenon. The quality of computed spectral signals thus depends on the accuracy of the QM level, the quality of treatment of the QM/MM interaction, which should include polarization to get a physically consistent picture, ,,− and the reliable sampling of the solute’s configurational space. While the refinement of QM/MM interactions and the QM level has attracted much interest in the context of computational spectroscopy, the role of accurate configurational sampling has generally been overlooked and performed employing classical molecular dynamics (MD) simulations. , The latter is a good compromise between computational cost and the accuracy of the sampling ,,,, and, along with density functional theory (DFT), has been used in multiple works to obtain spectra in solution. − However, anytime a molecular system is characterized by multiple metastable states in equilibrium, the exploration of the whole phase space can result frustrated by the presence of kinetic bottlenecks .…”