Cluster Linux
FILICO

In recent years, the European chemical industry community, following the corresponding changes in US and Japan, has experienced a significant transformation from resource-based towards knowledge-based approach, which implies a transition from low value products towards materials with high added value and considderable complexity in term of chemical structure, molecular architecture, specific functionality and more environmental-friendly approach. In order to design, characterise and understand all these new materials and preocesses, a number of new methods and techniques are being developed and used during last years, particularyl in US and Japan, but also in the European Community.

One of the most promising methods to address thermodynamic problems of fundamental and practical interest in the chemical industry is undeniable molecular modelling. These techniques have the advantage of much broader applicability than less fundamental based approaches. This is desirable for many reasons, the most important being that such methods can be used to predict properties well beyond the range of systems and thermodynamic state conditions used to fit the molecular parameters of the model. Molecular modelling methods are based on the application of sophisticated statistical mechanics techniques, developed primarily in the physics and chemistry communities, byt increasingly developed and applied in the chemical engineering community. Molecular modelling is a methodology which has experienced explosive growth in the US chemicla engineering community over the past decade, as chemical engineers have increasingly souhgt to understand as many phenomena as possible at the molecular level.

The molecular modelling methods are based on the knowledge of the inter- and intra-molecular forces between molecules, as well as the molecular structure of the system, which allows including in the model the most relevant microscopic effects (molecular shape, association interactions, dispersive forces, electrostatic contributions, etc.). The thermodynamic behaviour of the system can be predicted with confidence, over a wide range of state conditions, using well stablished statistical mechanics techniques which in most cases implies the use of several approximative schemes. A high predictibility is expected since the parameters are independent of thermodynamic conditions. Since the molecular properties are well defined, the same model can be solver exactly, within the statistical errors, using molecular simulation techniques. The comparison between theoretical predictions ans simulation results allow checking if approximations made in the theory are suitable. Direct comparison between simulation results and experimental data also provides valuable information on the ability of the molecular model to predict the behaviour of real systems. Finally, the proposed theory can be used with confidence to predict the thermodynamic behavior of the system over a wide range of state conditions.

The group has been involved in the use and development of the molecular-based Statistical Associating Fluid Theory or SAFT equation of state. The so-called Soft-SAFT equation of state accounts for mixtures of homonuclear and heteronuclear associating Lennard-Jones chains. The SAFT formalism is based on Werhteim's thermodynamic perturbation theory for spherical associating fluids. The SAFT approach constitutes one of the most sucessful and versatile molecular-based equations of state to deal with systems of industrial interest as complex as associating substances (hydrogen bonding), chain-like molecules (including polymers), amphiphiles, etc. The key point for the success of this equation is the inclusion of the most important microscopic effects in the free energy of the system: attractive and repulsive interactions, chain-like contribution and associating intermolecular interactions among others. The group has also improved and developed other versions of SAFT, such as the SAFT-HS and the SAFT-VR formalisms.