Abstract
The hydrolysis of phosphorus fluoridates was studied by semiempirical and ab initio methods. The reaction proceeds through a metastable intermediate separated from the reactants and products by transition structures. Both methods gave similar qualitative results; however, the first transition (formation of the trigonal bipyrimid intermediate) occurred earlier and the second (loss of F-) later with semiempirical methods. Including solvation in the calculations is critical. The AMSOL program generated reasonable energies especially when the limitations of the semiempirical Hamiltonian are considered. The Onsanger model was not a significant improvement over calculations on isolated molecules.