Robustness of Threshold Collision-Induced Dissociation Simulations for Bond Dissociation Energies

The threshold collision-induced dissociation (T-CID) method is the workhorse for gas-phase bond dissociation energy (BDE) measurements. However, T-CID does not measure BDEs directly; instead, BDEs are obtained by fitting simulated data to the experimental data. We previously observed several large discrepancies between the computed and experimental BDEs. To analyze the reliability of the experimental values, we previously reported a study of the dissociation rate models in the simulation. Here, we report a study of the collision simulation part, specifically in the L-CID (ligand CID) program. We show that the BDE values are robust even to intentionally introduced mistakes in the simulations, varying in most cases by less than 3 kcal mol(-1). The most significant exception is the collisional energy transfer (CET) simulation, which led to deviations larger than 10 kcal mol(-1). However, we found that the BDEs obtained with explicitly simulated CET distributions deviated by only 3 kcal mol(-1) from those simulated with the original model. Collectively, our results suggest that the T-CID-derived BDE values are robust and are likely to be accurate.