How constraint programming can help chemists to generate Benzenoid structures and assess the local Aromaticity of Benzenoids

• Published in: Constraints : an international journal Vol. 27; no. 3; pp. 192 - 248
• Main Authors: Carissan, Yannick, Hagebaum-Reignier, Denis, Prcovic, Nicolas, Terrioux, Cyril, Varet, Adrien
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.07.2022; Springer Nature B.V; Springer Verlag
• Subjects: Aromaticity; Artificial Intelligence; Benzenoids; Carbon; Chemical properties; Chemical Sciences; Chemistry; Chemists; Computer Science; Enumeration; Hexagons; Operations Research/Decision Theory; Optimization; or physical chemistry; Quantum chemistry; Search engines; Theoretical and; Graph variables and constraints; Theoretical chemistry; Modeling; Constraint programming
• ISSN: 1383-7133; 1572-9354
• DOI: 10.1007/s10601-022-09328-x

Benzenoids are a subfamily of hydrocarbons (molecules that are only made of hydrogen and carbon atoms) whose carbon atoms form hexagons. These molecules are widely studied in theoretical chemistry and have a lot of concrete applications. Then, there is a lot of problems relative to this subject, like the enumeration of all its Kekulé structures (i.e. all valid configurations of double bonds). In this article, we focus our attention on two issues: the generation of benzenoid structures and the assessment of the local aromaticity. On the one hand, generating benzenoids that have certain structural and/or chemical properties (e.g. having a given number of hexagons or a particular structure from a graph viewpoint) is an interesting and important problem. It constitutes a preliminary step for studying their chemical properties. In this paper, we show that modeling this problem in Choco Solver and just letting its search engine generate the solutions is a fast enough and very flexible approach. It can allow to generate many different kinds of benzenoids with predefined structural properties by posting new constraints, saving the efforts of developing bespoke algorithmic methods for each kind of benzenoids. On the other hand, we want to assess the local aromaticity of a given benzenoid. This is a central issue in theoretical chemistry since aromaticity cannot be measured. Nowadays, computing aromaticity requires quantum chemistry calculations that are too expensive to be used on medium to large-sized molecules. In this article, we describe how constraint programming can be useful in order to assess the aromaticity of benzenoids. Moreover, we show that our method is much faster than the reference one, namely NICS.