In:
The Journal of Chemical Physics, AIP Publishing, Vol. 99, No. 7 ( 1993-10-01), p. 5306-5315
Abstract:
Coupled cluster calculations at the CCSD(T)/[5s4p3d/4s3p] and CCSD(T)/[5s4p3d2 f1g/4s3p2d] level of theory are reported for reactions X+H2→XH+H [X=F (1a), OH (1b), NH2 (1c), and CH3 (1d)] utilizing analytical energy gradients for geometry, frequency, charge distribution, and dipole moment calculations of reactants, transition states, and products. A careful analysis of vibrational corrections leads to reaction enthalpies at 300 K, which are within 0.04, 0.15, 0.62, and 0.89 kcal/mol of experimental values. For reaction (1a) a bent transition state and for reactions (1b) and (1c) transition states with a cis arrangement of the reactants are calculated. The cis forms of transition states (1b) and (1c) are energetically favored because of electrostatic interactions, in particular dipole–dipole attraction as is revealed by calculated charge distributions. For reactions (1a)–(1d), the CCSD(T)/[5s4p3d2 f1g/4s3p2d] activation energies at 300 K are 1.1, 5.4, 10.8, and 12.7 kcal/mol which differ by just 0.1, 1.4, 2.3, and 1.8 kcal/mol, respectively, from the corresponding experimental values of 1±0.1, 4±0.5, 8.5±0.5, and 10.9±0.5 kcal/mol. For reactions (1), this is the best agreement between experiment and theory that has been obtained from ab initio calculations not including any empirically based corrections. Agreement is achieved after considering basis set effects, basis set superposition errors, spin contamination, tunneling effect and, in particular, zero-point energies as well as temperature corrections. Net corrections for the four activation energies are −1.05, −0.2, 1.25, and 0.89 kcal/mol, which shows that for high accuracy calculations a direct comparison of classical barriers and activation energies is misleading.
Type of Medium:
Online Resource
ISSN:
0021-9606
,
1089-7690
Language:
English
Publisher:
AIP Publishing
Publication Date:
1993
detail.hit.zdb_id:
3113-6
detail.hit.zdb_id:
1473050-9
Bookmarklink
