Here we have given a very brief introduction to the challenge that bond-breaking reactions pose for electronic structure theory. Obviously advances in this area would greatly aid in the theoretical prediction of potential energy surfaces, activation barriers, and reaction mechanisms. Molecular dynamics simulations of bond-breaking events need reliable energetics to be useful. Continued efforts by our research group and others seek improved reliability and speed in the computation of these processes.
We have performed a more thorough investigation on the choice of active space, including comparisons to full CI for various theoretical methods, as presented in J. S. Sears and C. D. Sherrill, Mol. Phys. 103, 803-814 (2005). This study adds interesting cases like the simultaneous breaking of both bonds in H2O, or breaking the triple bond in N2.
We have also written two introductory articles for non-specialists on this topic. Please see
