Hello, A yearly allocation for the LCRC cluster has been requested with the following updated information: Submitter/PI: Mark Pfeifle Project Name: pozdyn Division: CSE Project title: Dissociation Dynamics of Primary Ozonides in Alkene Ozonolysis Associated funding: DAAD (German Academic Exchange Service) Other Systems: Linus (CSE) (no fixed time allocation, depending on current total load) Science: Ozonolysis is a major tropospheric degradation process for alkenes. In the first reaction step, a cyclic primary ozonide (POZ) is formed which is subsequently decomposed into a Criegee intermediate (CI) and a carbonyl compound. The CI population is initially hot due to the exothermic POZ formation. This nascent energy distribution of the CI is essential to adequately model the yields of promptly decomposing and stabilized CI. These quantities in turn are of considerable interest because the thermalized CI acts as an atmospheric oxidant for various species and enhances aerosol formation. It is commonly assumed that the excess energy is distributed statistically among the fragments in the spirit of phase-space theory. However, this assumption is questionable when the exit channel potential energy surface is strongly repulsive, as is the case with POZ decomposition. In this project, direct dynamics calculations for C2H4 + O3 are carried out. Besides concerted POZ cleav age, a step-wise diradical pathway contributes comparably to the reaction flux. Trajectory calculations, initiated at both exit transition states, serve as a basis to test the statistical partitioning assumption and to provide appropriate input distributions for CI master equation calculations. Project description: So far, direct dynamics calculations using density functional theory (DFT) have been successfully carried out for the concerted and step-wise primary ozonide decomposition pathways. Significant non-statistical effects have been found. However, the reliability of DFT especially for the step-wise channel is uncertain because of the multi-reference character of the diradical structure. In the next stage, we want to treat the step-wise pathway with strongly correlated electronic structure methods. Such methods are expected to be more accurate than DFT. Promising candidates are multi-configurational 2nd order perturbation theory (MCPT2) and restricted active-space 2nd order perturbation theory (RASPT2), which are implemented in Molpro and Columbus, and particularly multiconfiguration pair-density functional theory (MC-PDFT), implemented in openmolcas. Before running actual trajectories, the accuracy, performance and memory usage of the different methods and program packages will be thoroughly investigated. The accuracy can be tested by carrying out a series of energy calculations on the DFT minimum-energy pathway between the transition state and products and comparing the results with a reference calculation (e.g. multi-reference configuration interaction, MRCI). Parallelization is another topic of interest, even though the problem of Monte Carlo simulations itself is embarrasingly parallel, because several individual trajectories can be independently run on multiple cores on one node. The trajectory integration will be carried out with a locally modified version of DiNT (http://tcg.cse.anl.gov/papr/codes/dint.html), where the computational cost for this step is negligible compared to the quantum-chemical energy and gradient calculations within Molpro, Columbus, or openmolcas. Industry partnership: Project URL: Current FY Hours Used: undetermined amount New FY Requested allocation: 200000 Q1: 200000 Q2: 0 Q3: 0 Q4: 0 Justification: Storage requirements: Thank You, The LCRC Accounts System