Extension of MNDO to d orbitals: Parameters and results for the halogens

Abstract

A recently proposed extension of the MNDO formalism to d orbitals has been parameterized for the halogens CI, Br, and I. Extensive test calculations indicate slight consistent improvements for normalvalent molecules and dramatic improvements for hypervalent molecules, in comparison with established MNDO‐type methods without d orbitals. The mean absolute errors in calculated heats of formation are 3.9 kcal/mol for 155 normalvalent compounds and 2.8 kcal/mol for 23 hypervalent compounds. The predicted structures of the hypervalent molecules are qualitatively correct, with a mean absolute error of 2° in 19 bond angles.

Hypervalent CompoundsBased in part on the article Indium: Inorganic Chemistry by Dennis G. Tuck which appeared in the Encyclopedia of Inorganic Chemistry, First Edition .

“…The ambiguities in the original definition have led to several changes. Today, no one would regard NH 4 + as hypervalent, while those that consider PF 5 and SF 6 hypervalent would classify SiF 5 − and SiF 6 2− in the same way. At this point, it is clear that whether or not a species is hypervalent depends not just on valence, but on the total number of valence electrons: a hypervalent molecule or ion features an atom surrounded by a total of more than eight valence electrons -both bonding and nonbonding.…”

“…It is an empirical rule based on the formulas of most compounds known in the early part of last century, and there is no reason why there should not be exceptions such as SF 6 and PCl 5 , as was indeed recognized by Lewis. The term hypervalent was introduced in a classic paper by Musher 1 in 1969 to categorize 'molecules and ions formed by elements in Groups 15 -18 of the periodic table in any of their valences other than their lowest stable chemical valence of 3, 2, 1, and 0 respectively'. This paper led to a surge of interest in both the theoretical and experimental aspects of species that appear to have 'expanded octets.…”

“…If so, there can be no doubt about the hypervalency of PF 5 (1) and SF 6 (2). But what about F 3 PO (3) and F 2 SO (4)?…”

“…Furthermore, if the definition of hypervalency is restricted to , then this artificially excludes species such as SiF 5 − and SiF 6 2− , which are isoelectronic and isostructural with PF 5 and SF 6 , respectively. The ambiguities in the original definition have led to several changes.…”

Performance of the semiempirical PM3 (tm) method in the geometry optimization of transition metal complexes

“…However, parametrization of these methods for transition metal systems is not trivial. Some attempts have, nevertheless, been made to introduce d-orbitals in the traditional semiempirical methods, and one can mention MNDO/d, 11,12 SAM/1, 13,14 PM3-d, 15 ZINDO, 16,17 and EHMO-ASED. 18 None of these methods has, however, reached a widespread use, maybe because of poor distribution of the codes.…”