Origin of high predictive capabilities in transition-state modeling

Autor:	F. M. Menger, M. J. Sherrod
Rok vydání:	1990
Předmět:	Force constant Colloid and Surface Chemistry Chemistry Stereochemistry Gaussian orbital Ab initio General Chemistry Statistical physics State (functional analysis) Biochemistry Catalysis
Zdroj:	Journal of the American Chemical Society. 112:8071-8075
ISSN:	1520-5126 0002-7863
DOI:	10.1021/ja00178a033
Popis:	Rates of 15 acid-catalyzed lactonizations, calculated by transition-state modeling, have been reported to correlate well with experimental rates (r=0.95). It is now found that the ability of transition-state modeling to predict rates depends less on the accurate portrayal of a transition state than on how closely the associated parameters coincide with a family of parameter sets that happens to give a good correlation. Success of transition-state modeling relies on unrealistic force constants (e.g., stretching parameters for partial bonds at 50% of ground-state equivalents) arbitrarily and fortuitously assigned to the transition structures. When these force constants are replaced by ab initio derived parameters, the correlation degenerates into scatter. Rate correlations near unity are achievable with «nonsense» force fields created by our FUDGIT software
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::0bea49fc993be1f339ae1d127042d17b https://doi.org/10.1021/ja00178a033 Zobrazit plný text záznamu