SUSHI

The 'SimUlation of Heterogeneous Systems and Interfaces' project (SUSHI) concerns the study of structural and thermodynamic properties of fluids and solids using Monte Carlo simulations. During the last few years, molecular simulations have been successfully applied to the understanding and prediction of many physico-chemical processes with applications to various fields such as the oil industry, food industry or biology. The methods have been extended to a wide range of applications thanks to the increase in computer power and methodological advances, such as the parallel tempering ensemble and statistical bias, giving access to a better exploration of the configurational space of the systems. This made Monte Carlo tools applicable to simple and 'complex systems'. However, systems of interest to the community, are of increasing complexity: nanomaterials, micro-emulsions, systems of biological interest, reactions are quite challenging in the simulation area. These systems are characterized by heterogeneities: large chemical and density differences appear between phases separated by interfaces. This introduces both practical and fundamental difficulties. For example, long range corrections have been developped mostly for homogeneous systems and have to be adapted to heterogeneous cases. In order to check for chemical equilibrium, local quantities like local chemical potential have to be defined, to ensure that the quantity is constant across phases. In the framework of the SUSHI project, we will study different systems. The study of explicit interfaces will be realized under the supervision of Prof. P. Malfreyt (partner 2, Univ. Blaise Pascal). A new method for the computation of the interfacial tension will be implemented and an original method will be proposed to determine the critical micellar concentration when surfactants are present. Comparison with experimental data will be undertaken by V. Lachet (partner 4, IFP). The investigation of chemical equilibrium under high pressure, motivated by the need to understand reactive systems in extreme conditions will be mainly performed under the supervision of E. Bourasseau (partner 3, CEA). Gases solubility in semi-crystalline polymers will be investigated by partners 1 and 4, under the supervision of B. Rousseau (partner 1, CNRS and Univ. Paris Sud 11). In order to account for the semi-crystalline nature of the polymer and to compute solubility under stress, an original adaptation of the osmotic ensemble will be used, in which the mechanical stress and the fugacity of the gas are imposed. During these studies, improvements of intermolecular potentials may be required. These improvements will be mostly made by partner 4 (with some help from Anne Boutin, partner 1). Following a succesful line of research ( winner of Fluid Simulation Challenge of AIChE, 2004), this work will be based on ab initio calculations by T. De Bruin (partner 4, IFP) as well as vapour-liquid equilibrium data. We believe that in order to extend the range of applicability of Monte Carlo simulations to complex systems, it is important to dispose of multipurpose Monte Carlo code. This is an ambitious project because multipurpose codes are more complicated to develop and maintain and may become slow and unefficient. However, it can be a powerfull tool from a scientific point of view, as evidenced by codes like Gaussian. Partners in this project have used and contributed for several years to the development of a shared Monte Carlo simulation code. The code is maintained by Jean-Marie Teuler (partner 1, C.N.R.S. and Univ. Paris Sud 11), specialised in software development and software maintaining, an unvaluable help in this project.