(189cn) High-Throughput in silico Screening of Candidate Compounds for Deep Eutectic Solvents

Deep eutectic solvents (DES) are a class of generally low-cost, environmentally benign solvents that have
diverse chemical applications. Similar in characteristics to ionic liquids (IL), they contain a mixture of
cations and Lewis acid anions, which form the hydrogen bond acceptor (HBA) species, and Lewis or
Bronsted acids, which form hydrogen bond donors (HBD). The H-bond links between the donors and the
acceptors allow for melting-point depression in the eutectic mixture and hence lower melting points
compared to ILs. Additionally, their easy synthesis from inexpensive starting chemicals along with low
vapor pressures and flammability make them attractive systems. Our work concerns the development of
novel DESs for use in supercapacitors as well as of the underlying molecular modeling techniques. Key
target properties our research focuses on are wide electrochemical windows (ECWs), high ionic
conductivities, and low melting points, which can be tuned at the molecular level. We are employing our
automated high-throughput screening infrastructure to characterize large-scale candidate libraries. We
utilize a diverse set of (multi-scale) molecular simulation techniques that range from first-principles
electronic structure theory, to classical molecular dynamics, to cheminformatics, to custom-built
machine learning prediction models. We will detail our mining efforts of the resulting data sets to
extract rational design principles for next-generation DES lead candidates.