Exploring the Photochemistry of Oxaloacetic Acid and Diethyl Oxaloacetate using Experimental and Non-adiabatic Molecular Dynamics Methods

Secondary organic aerosols (SOA) have significant impact on climate, air quality, and human health. Thus, it is critical to study the chemical fate of SOA particles, which is driven in part by photochemistry. Oxaloacetic acid (OAA) constitutes a broader class of keto-carboxylic acids commonly observed in the atmospheric environment. Our experiments showed that OAA undergoes both decarboxylation reaction and photodissociation. UV-Vis spectroscopy was used to capture the time decay of characteristic π*←n carbonyl band of OAA in aqueous solution mimicking conditions in cloud droplets. Vapor pressure of OAA is too low to study in the gaseous phase, therefore, more volatile diethyl oxaloacetate, was used to examine gas phase photochemistry associated with the carbonyl group of OAA. Photochemical processes involving OAA may unfold on picosecond time scales, making photochemical dynamics extremely hard to observe in experiments. Therefore, theoretical method simulating the excited state is of central importance to photochemistry in organic aerosol formations. Nonadiabatic radiationless transitions between-electron states are notoriously hard to control and proceed through “dark”, experimentally elusive conical intersections (CIs). For ground state, the Born-Oppenheimer (BO) approximation, which separates the electronic and nuclear degrees of freedom, is generally accurate. However, the BO approximation breaks down for excited states where two or more BO potential energy surface (PES) are strongly coupled. To capture the excited states more accurately, nonadiabatic molecular dynamics (NAMD) is employed, which demonstrates more efficient sampling of nonadiabatic processes accounting for nuclear quantum effects. The use of both experiments and NAMD simulations will make it possible to establish the atomistic mechanism of the photochemical processes, unveil ultrafast reaction pathways and dynamics, and determine observables such as yields and long-lived intermediates.