UCLA

When an excess electron is solvated by water molecules, it forms a well-known species named the hydrated electron. Although hydrated electrons serve as intermediates in many important reactions, such as radiation, oxidation, and electron transfer reactions, their hydration structure is unknown. This is important because the reactivity of the hydrated electron is strongly coupled with its solvation structure, making it impossible to accurately model chemical reactions involving this species because the structure is unknown. For the past decade, there has been an upsurge of interest in simulating hydrated electron structures with ab initio methods, specifically with density functional theory (DFT). All DFT-generated hydrated electron structures, regardless of the functional of choice, produced a distinct cavity region where no waters enter and a highly structured first hydration shell. Although several groups have claimed that such DFT-generated structures are ‘correct’, there has been no attempt to rigorously test their accuracy by making direct comparisons to experiments. Thus, the first focus of this thesis lies in evaluating two different DFT-based hydrated electron models that are widely accepted in the literature: a 0 K minimalist model that replaces explicit waters with a polarizable continuum and a fully-periodic molecular dynamics model. This thesis provides proof that DFT calculations make predictions that fail when directly compared with the experiment, a clear indication that the DFT-generated hydration structures cannot be correct. The second focus of this thesis explores how hydrated electrons behave in presence of ions. Experimentally, the hydrated electron’s absorption spectrum is known to blue-shift in electrolyte solutions, but a molecular explanation for this shift is unknown. To understand hydrated electron/electrolyte interactions, both the mixed quantum/classical (MQC) and DFT hydrated electron models were paired with a single Na+ to examine their detailed pairing interactions. For MQC models, the predicted Na+-induced blue-shift turns out to be highly sensitive to the simulation parameters of both the electron and the cation. The electron model that has the weakest hydration shell and the smallest cavity produce the blue-shift that is in best agreement with experiment. Most importantly, the DFT electron, with its strongly structured hydration shell, predicts a spectral red-shift upon ion-pairing, another clear proof that DFT-based simulations predict an incorrect hydration structure. In summary, this thesis shows clearly that DFT is not capable of correctly capturing the essence of this simple, yet complex, system.