Hartree-Fock complete basis set limit properties for transition metal diatomics

Abstract

Numerical Hartree-Fock (HF) energies accurate to at least 1 microhartree are reported for 27 diatomic transition-metal-containing species. The convergence of HF energies toward this numerical limit upon increasing the basis set size has been investigated, where standard nonrelativistic all-electron correlation consistent basis sets and augmented basis sets, developed by Balabanov and Peterson [J. Chem. Phys. 123, 064107 (2005)], were employed. Several schemes which enable the complete basis set (CBS) limit to be determined have been investigated, and the resulting energies have been compared to the numerical Hartree-Fock energies. When comparing basis set extrapolation schemes, those in the form of exponential functions perform well for our test set, with mean absolute deviations from numerical HF energies of 234 and 153 microE(h), when the CBS limit has been determined using a two-point fit as proposed by Halkier et al. [Chem. Phys. Lett. 302, 437 (1999)] on calculations of triple- and quadruple-zeta basis set qualities and calculations of quadruple- and quintuple-zeta basis set qualities, respectively. Overall, extrapolation schemes in the form of a power series are not recommended for the extrapolation of transition metal HF energies. The impact of basis set superposition error has also been examined.

Application of computational modeling to analyze reactive organotin(IV) species

“…One may expect that such a sequential layering of optimizations would result in a COM that is even closer to being ideal than any one of the COTs could have computed individually. However, it seems that a compounding of errors, as are endemic to the application of computational modeling theory regarding COTs, causes a COM resulting from sequential layering (as described above) to either (a) drift even further from the ideal, or (b) match the same COM that would have been modeled if the last COT applied was the only COT applied and was applied only once [13,18,22,23].…”

“…These default values do not include a complete computational representation of tin(IV)'s properties, especially as each tin(IV) atom relates to the other atoms of the molecule (e.g., correlation effects, packing effects, and intramolecularity), which include intramolecular primary and secondary effects (where primary is an atom being affected by a tin atom directly and secondary is an atom being affected by an atom that is being affected by a tin atom directly). The lack of full computational treatment of these properties, as compared to the default values applied, results in a loss of COM accuracy relative to the ideal, as will be seen here [13,18,22,23].…”

“…Yet, solid-state molecular properties tend to be experimentally determined in a crystalline phase [18]. This phase difference can significantly increase the complexity of determining gas phase-based COMs that model solid-state crystalline molecular properties, especially for molecules (such as those studied herein) that are organometallic, larger, or not well studied [13,19]. For this research, various COTs were used (defined herein) to examine the various molecules [18].…”

“…COT validation is a detailed comparative analysis over a range of COTs that quantifies the range of reliability of each COM per each COT applied to the molecule of interest and relative to ideal values (experimentation or lowest global energy of formation, described above) [13,19,22,23]. COTs may be selected to represent the rungs on Jacob's ladder or other considerations [24,25].…”

“…Much of this is laboratory research. Yet, there is significant growth in the usage and application of computational modeling research of organotin(IV) molecules [6][7][8]13].…”