CHAPTER 2: Understanding Entrainer Effects in Supercritical CO₂ Check Access

Despite decades of experiments adding small amounts (typically 1–5%) of solvents as ‘entrainers’ to increase the solute solubility in supercritical CO₂ (scCO₂), we have no coherent picture of how to predict a good entrainer for a given new solute. With no agreed standards regarding the data that need to be gathered and presented in digital format, it is very difficult to mine the literature, so a lot of hard-won data is under-utilised by the community. Moreover, with a wide choice of equations to fit whichever aspects of the data interest individual researchers, we have no single language to quantify entrainer efficiency, which prevents us from having an easy way to scan the literature for trends. Finally, with competing explanations for how entrainers work, there is no agreed methodology for looking at the core mechanism in order to gain deeper insight into entrainer action on a molecular scale. We show here that a rather straightforward re-analysis of many datasets gives us a single, universal, ‘good enough’ indicator of entrainer efficacy. In addition, the unambiguous identification of the key effect behind entrainer efficiency (entrainer–solute interactions) allows us to focus efforts towards the ideal of creating predictive tools. Two modest steps would help advance the field. First, to gather/present data in a way that can readily be analysed using the unambiguous theory from first principles. Second, rather than gathering more data on ad hoc solute–entrainer combinations, a carefully chosen set of solutes and entrainers would afford a rich set of big data, which could be usefully mined by theoreticians interested in providing predictive tools. Despite the current limitations, a single hypothesis confirms what has been stated tentatively in many publications: most entrainers will have little beneficial effects on most solutes and only solute–entrainer pairs with strong hydrogen bonding interactions can give large entrainer enhancements.