Barrier height calibration

With the introduction of two parameters in Eq. (4-15), the EH-MOVB method can be calibrated to reproduce exactly the experimental barrier height and the desired reaction energy. 

The theoretical results for the classical saddle-point and barrier height are summarized in Table X. Based on the FCI calibration calculations, the MRCI(322)(2p) + Q calculations with 7 electrons correlated (i.e., excluding F 2s correlation) are expected to reproduce the result of an FCI calculation in a nearly complete one-particle basis set. Since F 2s correlation decreases the barrier, this MRCI(322)(2p) -I- Q barrier, when corrected for the Cl superposition error (SE), represents an absolute upper bound of... 

In this review, we have shown how FCI benchmark calculations can be used not only to provide general criteria by which approximate n-particle space treatments can be judged, but also to provide detailed calibration as to which treatments are appropriate in specific cases. In this way it is possible to attach confidence limits to the computed results obtained when such a treatment is applied in a large ANO basis. This ability to estimate realistic error bars for calculated quantities is very important when the results are to be used in interpreting experiment or in other calculations, and we can expect more use to be made of it in the future. We have described several applications of this technique to problems in spectroscopy (such as the values for NH and CN) and chemistry (F + H2 barrier height). 

Using the EVB approach, which was calibrated to the aqueous reaction [28], Warshel et al. also reproduced the experimental free energy of activation for the catalyzed (OMP) reaction in ODCase. The EVB energy profile and barrier for the decarboxylation reaction in the gas phase was not reported. Although the computed barrier heights for the enzyme reaction are similar in the two computational studies [16, 28], a major difference is that the... 

As illustrated by the F+H2 reaction, even the best calculations may require some scaling to reproduce experimental barrier heights, exothermicities, or reaction rates. The global surfaces obtained at a lower level of theory can also be adjusted by comparing with more accurate calibration calculations carried out at critical points on the surface. For example, the CCI-l-Q potential for the F4-H2 reaction co d be modified by comparison with more accurate MRCI calculations. 

The calibration plots can be used to convert the barrier voltage axis into a mobility axis, and from the widths of the negative derivative peaks at the half height, ion mobility resolving power = Kf/AK ) can be determined. In the presented data some of the routinely obtained mobility resolving power (Rniob) values observed at various peaks were = 30 at angiotensin II2+, = 25 at flbrinopeptide A 3+,... 

---