Mathematical and chemistry properties of geometry-based invariants

TL;DR

This paper explores the mathematical and chemical properties of geometry-based invariants like the Sombor index, analyzing their extremal graph structures and demonstrating their usefulness in modeling thermodynamic properties of alkanes.

Contribution

It introduces and investigates the properties of new geometry-based invariants and applies them to predict thermodynamic properties of chemical compounds.

Findings

Identified extremal graphs for various invariants under given constraints.

Showed the invariants' effectiveness in modeling thermodynamic properties.

Validated the chemical applicability of these invariants.

Abstract

Recently, based on elementary geometry, Gutman proposed several geometry-based invariants (i.e., $SO$, $SO_{1}$, $SO_{2}$, $SO_{3}$, $SO_{4}$, $SO_{5}$, $SO_{6}$). The Sombor index was defined as $SO(G)=\sum\limits_{uv\in E(G)}\sqrt{d_{u}^{2}+d_{v}^{2}}$, the first Sombor index was defined as $SO_{1}(G)= \frac{1}{2}\sum\limits_{uv\in E(G)}|d_{u}^{2}-d_{v}^{2}|$, where $d_{u}$ denotes the degree of vertex $u$. In this paper, we consider the mathematical and chemistry properties of these geometry-based invariants. We determine the maximum trees (resp. unicyclic graphs) with given diameter, the maximum trees with given matching number, the maximum trees with given pendent vertices, the maximum trees (resp. minimum trees) with given branching number, the minimum trees with given maximum degree and second maximum degree, the minimum unicyclic graphs with given maximum degree and girth, the minimum connected graphs with given maximum degree and pendent vertices, and some properties of maximum connected graphs with given pendent vertices with respect to the first Sombor index $SO_{1}$. As an application, we inaugurate these geometry-based invariants and verify their chemical applicability. We used these geometry-based invariants to model the acentric factor (resp. entropy, enthalpy of vaporization, etc.) of alkanes, and obtained satisfactory predictive potential, which indicates that these geometry-based invariants can be successfully used to model the thermodynamic properties of compounds.