Computational studies of supercritical fluids

Supercritical carbon dioxide is an exciting solvent for chemical reactions. It is cheap, non-toxic, non-flammable and non-polluting, and its physical properties are greatly influenced by relatively small variations in temperature and pressure, which enable the reaction outcomes to be controlled. However, fewer chemicals are soluble in carbon dioxide than in solvents such as water, which restricts the use of supercritical carbon dioxide in chemical processes.To increase the solubility of organic molecules in carbon dioxide, hydrogen atoms in hydrocarbon chains can be replaced by fluorine atoms. Alternatively, fluorinated hydrocarbon molecules (HFCs) can be used as the basis of surfactants, which enable small droplets of water to exist within the carbon dioxide solvent. Chemical reactions can then take place inside the water droplets, even when the chemicals involved are not soluble in the surrounding carbon dioxide.Fluorinated molecules can have a low environmental impact, and a number of small HFCs have replaced ozone-depleting CFCs as refrigerants. Unfortunately, the high cost of producing larger fluorinated molecules has limited the commercial potential of HFC / carbon dioxide mixtures. Furthermore, the reason for the enhanced solubility of fluorinated molecules in carbon dioxide is not known, which makes it difficult to design more cost-effective alternatives to HFCs.The proposed research will tackle these theoretical and practical problems associated with HFC / carbon dioxide mixtures. The work will be carried out at the University of Nottingham, and is a new collaboration between established research groups with considerable expertise in the areas of molecular interactions (Wheatley), simulations (Hirst) and experimental measurements on supercritical carbon dioxide (Poliakoff and Ke).We shall begin by developing new computational methods, based on those already used successfully by the PI, to investigate the interactions between HFCs and carbon dioxide molecules with more accuracy and in more detail than previously possible. Experimental measurements of the phase equilibria, critical phenomena and thermodynamic properties of the mixtures, using state-of-the-art equipment developed by the experimental CoIs, will allow validation and improvement of these computational methods for mixtures involving small HFCs, and for the few large HFCs that are currently available.The calculated intermolecular interactions will be used in atomic simulations of the structures and physical properties of HFC / carbon dioxide mixtures. Hirst's experience of simulations of large organic molecules will ensure that the technical challenges of this work are overcome, and the demand that the simulations will place on computing resources will be met by the new 1024-node teraflop computing cluster at the University. The results obtained from these simulations will allow us to explore the reason for the enhancement of solubility by fluorination, and to produce new solubility data - which cannot currently be obtained from experiment - for a range of new HFC molecules. In the longer term, the research will allow us to predict the characteristics that make molecules soluble in supercritical carbon dioxide, and will therefore assist in the design of new, efficient, environmentally-friendly chemical solvents.