Given the increasing prevalence of computational tools across the chemical sciences, it is critical that these topics be included in undergraduate chemistry curricula. In particular, machine learning (ML) methods have been applied to a wide range of problems across different areas of chemistry. One approach for incorporating computational and ML activities into the chemistry curriculum is to provide companion modules to common class topics or laboratory experiments. Here we report a new pedagogical activity that introduces chemistry students to ML regression analysis in the context of the common physical chemistry laboratory experiment on the absorption spectra of conjugated cyanine dyes. In the classic version of this experiment, students are taught to model cyanine dye absorption using the Kuhn free electron model based on the quantum particle in a box. In the new computational activity, students develop different ML regression models to predict the wavelength of maximum absorption of cyanine dyes. This exercise introduces students to the workflow of ML regression analyses, including data preprocessing, model training, and model evaluation. After completion of the regression models, students are prompted to use their results to gain insight into what molecular properties affect the wavelength of maximum absorption; through this process, students are able to generate hypotheses that lead them naturally to the particle in a box model. This activity is designed to be completed in a single lab period by upper-division chemistry students with little previous programming experience. To facilitate adoption of this activity, we provide a sample dataset and reference implementations in Python and Mathematica. Although intended as a complement to the classic wet lab experiment, this activity could be performed on its own. Alternatively, it could be extended to add additional content in the areas of ML, cheminformatics, electronic structure calculations, or programming.

Rediscovering the particle in a box: Machine learning regression analysis for hypothesis generation in physical chemistry lab

Where thousands of chemistry professionals meet to share ideas and advance scientific and technical knowledge. Spring 2023 abstracts, posters, and presentations will launch at the end of March.

ACS Spring 2023

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.