PNNL

Abstract

Solubility is a critical physicochemical property of redox active materials due to its direct association with the energy density of redox flow batteries. A data-driven approach can significantly to accelerate the development of highly soluble redox active materials. To achieve this, it is crucial to collect large-quantity and high-quality solubility data to construct and validate prediction models. However, accurate solubility prediction remains challenging due to the scarcity of relevant solubility databases and limitations of high throughput screening methods commonly used in pharmaceutical research for RFB applications. In this study, a high-throughput experimentation process was developed, combining a HTP methodology for designing the workflow with a customized robotic platform for execution, to achieve highly automated and high-throughput solubility measurement. By mimicking the classical shake-flask method and standardizing the experimental process, we demonstrate the versatility of this method for thermodynamic solubility determination and screening the effects of additives on solubility in aqueous and non-aqueous RFB applications.