This meeting was held in The Robin Brooke Centre of St.Batholomews's Hospital, Smithfield, London on 14th December 1998. About 100 delegates, including many students, attended this useful overview of the methods of molecular modelling and review of academic and industrial applications. It was organised by the Industrial Physical Chemistry Group within the Industrial Affairs Division of the Royal Society of Chemistry.
Nick Quirke (Imperial College) explained that although reality is complex we have to make simplifying assumptions to make any progress in the field. An early method was to consider molecules as hard spheres and to model the force between them in terms of a potential made up of two parts, a long term attraction and a hard core repulsion. Liquids contain many thousands of molecules, to simulate the liquid properties one takes an ensemble of many molecules in all possible configurations weighted by the probability of that configuration occuring. Clearly using more molecules in the simulation gives a closer approximation to reality but takes more computing time. Since each molecule experiences a similar force to its neighbours the calculation for each one is identical to that for the next, making this an ideal problem for a parallel computer. In Monte Carlo methods each molecule is perturbed from its original position by a small amount and the energy of whole system calculated. By choosing those configurations which have lower energy one hopes to eventually arrive at an equilibrium. Molecular dynamics can be calculated by considering how the ensemble behaves over time. The first papers in this field were published in the 1950s, and a reference list is given at the end of this report. people are now attempting the simulations of surface wetting, including the addition of particulates to polymers during production, where it is essential for wetting to take place.
Dominic Tildesley, Unilever, spoke on atomistic and mesoscale modelling in an industrial environment. At Port Sunlight they simulate the materials, the chemical processing plants and the supply chain. He discussed the prediction of small crystal shape in shampoos. Some crystal shapes make the shampoo gleam more than others, which is apparently a great selling point with their customers. (Personally, I keep my eyes shut when shampooing my hair so I have not observed this phenomenon.) His simulations involve the prediction of the change in morphology as dopant is added to give the desired shape. Unilever have a large detergent manufacturing business and are working on improving the anti-microbial properties of their products by simulations.
Richard Catlow, Royal Institution, spoke on simulations in materials research, where they model the surface properties and large scale properties such as the elasticity or the diffusion through structures. They have modelled micro porous silicalite a catalyst for hydrocarbon synthesis and also glassy materials by starting from a crystalline state and using molecular dynamics to 'heat' the materials until it 'melts'. They are modelling zeolites for use as catalysts and looking at how the organic molecules 'dock' into the holes in the structure. Diagrams of zeolites can be found on the CD-ROM given to delegates containing Vol 19 no 5,6 (1997) of the journal 'Molecular Simulation'.
Keith Gibbons, North Carolina State University, discussed applications in chemical engineering including phase equilibria and confined flow, such as that through oil pipelines and at a much smaller scale through porous glasses used in chromatography.
After lunch Peter Coveney, Queen Mary and Westfield College, explained the importance of merging the atomistic scale models smoothly with those of the mesoscale. Andrew Leach, Glaxo Wellcome, spoke of structure based library design and use of combinatorial chemistry to simulate many chemicals during their search for better drugs. They may simulate thousands of molecules but few will have any biological activity; the problem is to find just those which are worth synthesising and testing.
Ian Gould, Imperial College, continued this theme of combinatorial chemistry
but he is applying it to modelling proteins and protein drug interactions.
He emphasised the importance of using many methods of structure
determination, X-rays, neutrons and NMR, having good visualisation tools and
the modelling of the proteins in their environment, usually water, where the
proteins are not rigid, and so are continually folding and refolding. He is
also working on the modelling of photosynthesis in order to produce a
better type of solar cell which can use many different wavelengths of light
as plants do.
David Nicolaides, MSI, talked about MesoDyn, an Esprit
project (no. 228650) a collaboration between several European partners who
are modelling lubricant flow round complex shapes. All the applications will
be in the public domain, look at the MSI website for further details of
their applications. ( http://www.msi.com )
Conclusions
The day contained a wide range of talks on a wide range of applications. Atomistic techniques are now reaching maturity with several software codes available. Computer power is increasing so that such codes can be run on the desktop. Mesoscale applications are becoming possible, but one problem is that much of this work is not freely accessible because it is done in a commercial environment. It is now becoming possible to simulate non-equilibrium problems.
Kate Crennell
References:
Readers are reminded that the Dec 96 issue of
'Crystallography news' (p 46, 47) had a review of
'Understanding
Molecular Simulation. from Algorithm to Applications,
by D Frenkel
and B.Smit published in 1996 by the Academic Press.
I am grateful to Prof N.Quirke for supplying the references below as suitable textbooks for those just entering this field, and the list of references to early papers in this field.
Books
Molecular Simulation early references and journals
Journals
Papers in this field are published in the journal 'Molecular
Simulation' by Gordon and Breach Science Publishers. E-mail:
[email protected]
Web site
http://www.gbhap.com/Molecular_Simulation/
Delegates received a copy of the special CD-ROM issue Volume 19 no 5-6 (1997), which gives a flavour of the type of papers; it can be viewed with an Internet browser on most microcomputers.
Papers with greater graphical content may be published in the 'Journal of Molecular Graphics and Modelling' published by Elsevier. The journal home page is at http://www.elsevier.nl/locate/jmgm/ The full text is available on-line via ChemWeb (to join go to http://ChemWeb.com ) You can receive the table of contents of each issue in advance by email by registering at http://www.elsevier.nl/locate/ContentsDirect
Societies
The Molecular Graphics and Modelling Society, MGMS, edits the journal and organises meetings and workshops. Their url is http://www.yorvic.york.ac.uk/~mgms
CCP5 is the Daresbury Laboratory's Collaborative Computational Project on Molecular Dynamics, Monte Carlo and Lattice Simulations of Condensed Phases. They maintain a large software library and organize workshops on simulation problems, which can be found at URL http://www.dl.ac.uk/CCP/CCP5/main.html
Click here to return to BCA homepage