Abstract:

Students in undergraduate Analytical Chemistry courses often encounter difficulties when solving common quantitative analytical problems. A common issue is the tendency to apply formulas in a procedural or mechanical way without fully understanding the underlying chemical concepts, such as equilibrium relationships, dilution processes, or stoichiometric calculations, or how the different variables involved influence the final result. In addition, errors related to units and numerical conversions are frequent, and small mistakes in factor application can lead to incorrect results even when the general approach to the problem is correct. As a result, students may find it difficult to confidently assess whether their solutions are valid.

An interesting approach to addressing these challenges is the development of interactive web-based tools using generative artificial intelligence (AI) systems, which are capable of producing functional HTML files or similar web-based resources within seconds. This allows students to design web pages adapted to specific analytical problems, while also providing them with a practical tool to verify and self-evaluate their solutions.

In the present study, the potential of developing web-based tools using generative AI to improve students’ understanding of common analytical problems was explored. To carry out the activity, students enrolled in the Chemistry and Pharmacy degree programs from the University of Valencia (Spain) were divided into small groups, which used ChatGPT or Microsoft Copilot to generate HTML files capable of performing the calculations associated with Analytical Chemistry exercises. Students’ satisfaction with the activity, together with other aspects of the educational experience, was subsequently evaluated. The results indicate that the activity was positively received by a large proportion of the students and contributed to an improved understanding of the subject matter.

Keywords:

Artificial Intelligence, problems resolution, web-based sources, self-evaluation tools.