Computer Aided Molecular Design of alternative coolant molecules for ethylene glycol

Coolants play an important role in engineering and automotive applications with ethylene glycol being a very commonly used coolant. Despite its widespread use, ethylene glycol is highly toxic and poses significant health risks to humans and the environment. As a result, there is a growing need for safer alternative coolants without compromising on performance. Advancements in computer aided molecular design (CAMD) coupled with global optimization can provide a powerful new way of discovering new materials with specifically desired physical and chemical properties. In this study, we hypothesize that candidate molecules produced via CAMD can reduce the volume of coolant required in a conventional car engine. To do this, we researched alternative coolants to ethylene glycol based on properties such as specific heat capacity at constant pressure, molecular weight, toxicity, and danger to the environment as determined by the United States Environmental Protection Agency (EPC). This problem was then formulated as a Mixed Integer Nonlinear Programme (MINLP) using constrained global optimisation to estimate the thermodynamic properties of candidate molecules based on physical and auxiliary constraints. By comparing the candidate molecules from the solver to the research we determined the solver had found similar molecules to currently viable safer coolant alternatives. The model predicted heat capacities within a 5% error margin of their experimental values on the National Institute for Technology and Standards (NIST) database. The best alternative candidate molecule found was isobutane, R-600a, commonly used as a coolant. This work highlights the use of CAMD as a prototyping technique in the chemical synthesis design process, and it makes the process more iterative by speeding up the selection of candidate molecules to save time and reduce the cost of research and development (R&D).