Integrating Large Language Models into the Chemistry and Materials Science Laboratory Curricula

Abstract

How might LLMs impact student learning in laboratory settings, especially regarding critical thinking, core concepts, and scientific identity? What are the ramifications of LLMs for course delivery strategies, including course design, scaffolding, and assessment? Can LLMs facilitate the incorporation of alternative delivery strategies into the undergraduate curriculum, such as course-based undergraduate research experiences (CUREs), to potentially broaden access to authentic research? Figure 1. Demonstration of data visualization and analysis generated entirely by natural language prompting using ChatGPT-4o. The first example highlights data analysis from a general chemistry absorption spectroscopy laboratory, in which students are asked to (a) generate a calibration curve and (b) determine the concentration of an unknown analyte using Beer’s law. The second example illustrates a chemical kinetics experiment familiar to upper division physical chemistry laboratories. Students (c) collect and visualize time-dependent absorbance data, relate the measured absorbance to analyte concentration with Beer’s law, and then (d) linearize the data to determine the reaction order, rate constant, and integrated rate law. The authors gratefully acknowledge support from the National Science Foundation under Grant No. DMR-2333654. The authors thank the invited workshop speakers: Allyson Fry-Petit, Bo Wu, Taylor Sparks, Boris Kiefer, Susan Gentry, Sam Cox, Mayk Caldas Ramos, and Shruti Badhwar. The authors also gratefully acknowledge all workshop participants for their attendance. This article references 31 other publications. This article has not yet been cited by other publications.