Estimations of global warming potentials from computational chemistry calculations for CH ₂F ₂ and other fluorinated methyl species verified by comparison to experiment

In this work, the global warming potential (GWP) of methylene fluoride (CH ₂F ₂), or HFC-32, is estimated through computational chemistry methods. We find our computational chemistry approach reproduces well all phenomena important for predicting global warming potentials. Geometries predicted using the B3LYP/ 6-311g** method were in good agreement with experiment, although some other computational methods performed slightly better. Frequencies needed for both partition function calculations in transition-state theory and infrared intensities needed for radiative forcing estimates agreed well with experiment compared to other computational methods. A modified CBS-RAD method used to obtain energies led to superior results to all other previous heat of reaction estimates and most barrier height calculations when the B3LYP/6-311g** optimized geometry was used as the base structure. Use of the small-curvature tunneling correction and a hindered rotor treatment where appropriate led to accurate reaction rate constants and radiative forcing estimates without requiring any experimental data. Atmospheric lifetimes from theory at 277 K were indistinguishable from experimental results, as were the final global warming potentials compared to experiment. This is the first time entirely computational methods have been applied to estimate a global warming potential for a chemical, and we have found the approach to be robust, inexpensive, and accurate compared to prior experimental results. This methodology was subsequently used to estimate GWPs for three additional species [methane (CH ₄); fluoromethane (CH ₃F), or HFC-41; and fluoroform (CHF ₃), or HFC-23], where estimations also compare favorably to experimental values.