I. Cognitive and instructional factors relating to students' development of personal models of chemical systems in the general chemistry laboratory. II. Solvation in supercritical carbon dioxide/ethanol mixtures studied by molecular dynamics simulation

Abstract

Part I. Students' participation in inquiry-based chemistry laboratory curricula, and, in particular, engagement with key thinking processes in conjunction with these experiences, is linked with success at the difficult task of "transfer" - applying their knowledge in new contexts to solve unfamiliar types of problems. We investigate factors related to classroom experiences, student metacognition, and instructor feedback that may affect students' engagement in key aspects of the Model-Observe-Reflect-Explain (MORE) laboratory curriculum - production of written molecular-level models of chemical systems, describing changes to those models, and supporting those changes with reference to experimental evidence - and related behaviors. Participation in introductory activities that emphasize reviewing and critiquing of sample models and peers' models are associated with improvement in several of these key aspects. When students' self-assessments of the quality of aspects of their models are solicited, students are generally overconfident in the quality of their models, but these self-ratings are also sensitive to the strictness of grades assigned by their instructor. Furthermore, students who produce higher-quality models are also more accurate in their self-assessments, suggesting the importance of self-evaluation as part of the model-writing process. While the written feedback delivered by instructors did not have significant impacts on student model quality or self-assessments, students' resubmissions of models were significantly improved when students received "reflective" feedback prompting them to self-evaluate the quality of their models. Analysis of several case studies indicates that the content and extent of molecular-level ideas expressed in students' models are linked with the depth of discussion and content of discussion that occurred during the laboratory period, with ideas developed or personally committed to by students during the laboratory period being likely to appear in students' post-laboratory refined models. These discussions during the laboratory period are primarily prompted by factors external to the students or their laboratory groups such as questions posed by the instructor or laboratory materials. Part II. Solvation of polar molecules within non-polar supercritical carbon dioxide is often facilitated by the introduction of polar cosolvents as entrainers, which are believed to preferentially surround solute molecules. Molecular dynamics simulations of supercritical carbon dioxide/ethanol mixtures reveal that ethanol molecules form hydrogen-bonded aggregates of varying sizes and structures, with cyclic tetramers and pentamers being unusually prevalent. The dynamics of ethanol molecules within these mixtures at a range of thermodynamic conditions can largely be explained by differences in size and structure in these aggregates. Simulations that include solute molecules reveal enhancement of the polar cosolvent around hydrogen-bonding sites on the solute molecules, corroborating and helping to explain previously reported experimental trends in solute mobility.