Are ``GAPT Charges{''} Really Just Charges?

Autor(es)

Richter, Wagner E. ^[1] ; Duarte, Leonardo J. ^[2] ; Bruns, Roy E. ^[2]

Afiliação do(s) autor(es)

^[1] Univ Tecnol Fed Parana, Dept Chem Engn, BR-84017220 Ponta Grossa, Parana - Brazil

^[2] Univ Estadual Campinas, Inst Chem, BR-13083861 Campinas, SP - Brazil

• Tipo de documento: Artigo Científico
• Fonte: JOURNAL OF CHEMICAL INFORMATION AND MODELING; v. 61, n. 8, p. 3881-3890, AUG 23 2021.
• Citações Web of Science: 0

Resumo

Generalized atomic polar tensor (GAPT) has turned into a very popular charge model since it was proposed three decades ago. During this period, several works aiming to compare different partition schemes have included it among their tested models. Nonetheless, GAPT exhibits a set of unique features that prevent it from being directly comparable to ``standard{''} partition schemes. We take this opportunity to explore some of these features, mainly related to the need of evaluating multiple geometries and the dynamic character of GAPT, and show how to obtain the static and dynamic parts of GAPT from any static charge model in the literature. We also present a conceptual evaluation of charge models that aims to explain, at least partially, why GAPT and quantum theory of atoms in molecules (QTAIM) charges are strongly correlated with one another, even though they seem to be constructed under very different frameworks. Similar to GAPT, infrared charges (also derived from atomic polar tensors of planar molecules) are also shown to provide an improved interpretation if they are described as a combination of static charges and changing atomic dipoles rather than just experimental static atomic charges. (AU)

• Processo FAPESP: 18/24844-7 - Uso de tensores polares atômicos e parâmetros QCT para treinar um modelo de machine learning e prever constantes de Hammett
• Beneficiário: Leonardo José Duarte
• Modalidade de apoio: Bolsas no Exterior - Estágio de Pesquisa - Doutorado Direto
• Processo FAPESP: 18/08861-9 - Aplicação do modelo QTAIM / CCTDP e machine learning para a previsão de reatividades químicas
• Beneficiário: Roy Edward Bruns
• Modalidade de apoio: Auxílio à Pesquisa - Regular