Dehalogenation is among the most important processes involved in contaminant fate, but despite all the work that has been done on the kinetics of dehalogenation, there are few linear free energy relationships (LFERs) that can be used to explain or predict rates of dehalogenation by environmental reductants. Previously, we summarized kinetic data for dehalogenation of chlorinated alkanes and alkenes by zero-valent iron (Fe0) and showed that correlation analysis of these data with published two- electron reduction potentials did not give a simple relationship. In this study, we report successful LFERs based on estimated lowest unoccupied molecular orbital (LUMO) energies calculated from semiempirical (AM1 and PM3) and ab initio methods (6-31G*) and one-electron reduction potentials. Solvation effects can be modeled with COSMO and incorporated into semiempirical estimates of E(LUMO), but this did not improve the correlation with k. The best LFER (log k = -5.7-1.5 E(LUMO)) explains 83% of the variability in surface area-normalized rate constants (k) with ab initio LUMO energies. The LFER is improved by correcting for statistical bias introduced by back transformation from log-linear regression models. New kinetic data for six compounds are compared with rate constants predicted using the unbiased LFER.
|Original language||English (US)|
|Number of pages||8|
|Journal||Environmental Science and Technology|
|State||Published - Oct 1 1998|
ASJC Scopus subject areas
- Environmental Chemistry