2026/05/05 by Wassja A. Kopp, Philipp Kuboth, Tom-Luka Zwarg +3 · 1 voice
Chemistry · Computer Science · Engineering · #Augmented Reality Applications #Experimental Learning in Engineering #Various Chemistry Research Topics
paper · doi:10.26434/chemrxiv.15002719/v1
openalex publication_date 2026/05/05 · openalex created_date 2026/05/06 · openalex updated_date 2026/07/14
Chemistry students are required to understand abstract and structurally complex concepts such as high-dimensional potential energy surfaces (PESs) and densely connected chemical reaction networks (CRNs). In teaching, these systems are usually reduced to textbook figures or screen-based visualizations, which necessarily simplify their spatial and topological structure. This can hinder students’ ability to recognize transition states as saddle points and make it difficult to maintain an overview of complex network connectivity. We designed and implemented an augmented reality (AR) teaching unit in a Master-level chemistry course in which PESs and CRNs are treated as shared, navigable three-dimensional learning objects. The unit extends existing computational coursework through two guided activities, namely exploring model PES as room-scale energy landscapes and navigating a literature-based pentane oxidation mechanism to identify branching pathways and temperature-dependent radical formation routes. An exploratory classroom evaluation (N = 9) combined learningrelated self-assessment with established measures of usability (inspired by the system usability scale SUS), perceived workload (NASA-TLX), and simulator comfort (SSQ items). Students reported the strongest perceived benefits for CRN-related outcomes, particularly understanding connectivity and collaborative discussion supported by a shared spatial reference. Reported gains for PES interpretation were positive but more moderate. Usability ratings indicated that the system could be learned quickly, while workload and eye strain highlighted practical constraints for session duration and onboarding. These results suggest that collaborative AR can complement conventional computational instruction by supporting structural and system-level sense-making for complex reactivity representations, especially when learning relies on discussion and shared interpretation.