Isodesmic Reactions for Transition States:  Reactions of Cl Atoms with Methane and Halogenated Methanes

The performance of the technique of isodesmic reactions for transition states (IRTS) has been analyzed via application to 17 reactions of abstraction of hydrogen atoms from methane and halogenated methanes by Cl atom. A variety of quantum chemical methods and basis sets was used. The calculated energy barriers demonstrate linear correlations with those derived from modeling of the experimental rate constant data, in agreement with the prediction based on the IRTS formalism. The results of the study confirm the validity of the technique of isodesmic reactions for transition states for calculation of reaction rates and demonstrate the existence of method-specific systematic errors in calculations of reaction barriers. The technique of isodesmic reactions is directed at factoring out and eliminating these systematic errors. The predictive ability of the technique is directly related to the quality of the observed correlations. Average and maximum deviations from the best fit lines on the correlation plots depend on the quantum chemical method used. The highest quality correlation (the least amount of scatter, average deviation of 1.5 kJ mol^-1 and maximum deviation of 3.5 kJ mol^-1) was obtained with the BH&HLYP/6-311+(3df,2pd)//BH&HLYP/6-311(d,p) combination of single-point energy//geometry optimization methods. Use of higher level methods such as spin-projected PMP4, QCISD(T), and CCSD(T) results in small systematic errors (1.4−5.9 kJ mol^-1) but larger scatter on the plots of the calculated barriers vs “experimental” barrier correlations (maximum deviations of 5.6−6.8 kJ mol^-1).