(321q) Prediction of Global Warming Potentials through Computational Chemistry – Testing Robustness of Methodology through Experimental Comparisons

Global warming is a scientifically based concern regarding addition of natural and anthropogenic based chemicals to the troposphere where the species can trap energy in the infrared region. Predicting global warming potentials requires highly accurate rate constant measurements for the reactions of the chemicals with hydroxyl radicals, which is the first and rate limiting step in environmental degradation. Radiative forcing, the amount of energy that can be captured by the chemicals per square meter of exposed area for a given concentration, requires spectroscopic information about peak locations and intensities, which are then aggregated into absorption cross sections. These values are then used in atmospheric modeling simulations to determine the radiative forcing. Both kinetic and spectroscopic measurements have many potential experimental difficulties, which makes predicting global warming potentials (GWPs) from theory attractive.

We build on our previous work from last year by examining more chemicals using theoretical chemistry to predict GWPs. Last year's work investigated CH₂F₂ and found excellent comparison to experiment for predicting all intermediate steps for GWPs, including kinetic degradation rates with hydroxyl radical under low temperature tropospheric conditions, atmospheric lifetime estimates, radiative forcing in the atmospheric window, and overall GWPs at 20 year, 100 year, and 500 year time horizons.

In this work, improvements are made in the kinetic estimation procedures because a cancellation of errors in the original CH₂F₂ work led to accurate kinetic estimates. In particular, variational transition state theory was used as opposed to the original transition state theory. In the original work, the Wigner tunneling correction was used, but this tended to underestimate tunneling corrections due to the use of B3LYP/6-31g* calculations and the limitations of the model. To correct for this, higher level calculations were done at the CCSD and MP4SDQ levels for use with the small curvature tunneling (SCT) approach. The lowest real vibrational mode along the reaction coordinate was found to be a hindered rotor so this was corrected. Finally, the commonly used structural factor was more rigorously assessed with the approach of Truhlar, et al.

The addition of the four kinetic improvements led to more robust and accurate prediction of rate constants for a host of other chemicals, including CH₄, CH₃F, CHF₃, CH₃F₃, and CH₂FCF₃. We find good agreement for all kinetic based parameters for these species compared to experimental values. Radiative forcing estimates are also in good agreement with available experimental results. Finally, we now have a larger database of chemicals where we have verified our methodology of accurately predicting global warming potentials completely from theory.