ACS Spring Meeting and Exposition |
|
March 21-25, 2010, San Francisco, USA |
|
Please meet us during the ACS Spring Meeting and Exposition. Visit booth #332 (SCM) in the exhibition hall if you are interested in what our software can do for you. We will be presenting four lectures during the conference: 61 - Multiscale model for interfacial heterodispersity in crude oil dispersions Crude oils invariably contain a wide distribution of chemical compounds. When dispersed in a brine for exploration, recovery or transport, such emulsion will have to be stabilized by some surfactant / polymer mix; that may also be variable depending on the crude oil composition. We report on a novel class of scripted multiscale models, embedded in the Culgi framework (Chemistry Unified Lanuage Interface), that encompass both molecular and mesoscopic approaches. In particular, we discovered that emulsion surfaces are best described as a thick layer of heterodisperse crude oil components. We discuss the interfacial thermodynamics, with focus on emulsions derived from Canadian tar sands, and speculate on conditions for improving emulsification processes. Culgi is sponsored by an international consortium, including petrochemical industries and EU projects. 358 - Chemistry unified language interface for scripting multiscale modeling The Chemistry Unified language Interface (CULGI) is a C++ library for soft materials multiscale modeling, ranging from molecular to mesoscopic and materials informatics. Culgi is sponsored by an international industrial consortium and several large EU projects (Multimatdesign, Nanomodel and Selfmem) and the Dutch Space Agency. The library is organized as a 'computer game', where a user can manipulate individual chemical objects, either through manual interaction via a script or by some physical interaction model. Scripts can be written in Tcl or Python, or a specialized subset in simplified Culgi script. A special consideration that requires considerable investigation is the narrow path between writing a library for experts only, or a scripting interface than can be used by non-experts as well, similar to existing Graphical User Interfaces. We discuss the logic of Culgi, and present some applications from customizability in petroleum engineering, engineering plastics and surfactants design. 1 - Principles of stochastic quasi-Newton dynamics and applications in soft materials multiscale modeling We report a Stochastic Quasi-Newton (S-QN) method for sampling energy potential hypersurfaces in soft materials (CULGI). In the heart of the theory is a compound mobility matrix that responds to a landscape by a memory function, derived from QN theory. Our approach is a real-space generalization of Fourier acceleration algorithms that use filtering for the separation of different length and time scales. Applications of S-QN are ambitious, and include structure optimization, analysis of correlations and automated multiscale modeling. We report on the mathematical fundamentals of the method, the comparison with regular (L-)BFGS and force-biased Monte Carlo, and discuss applications to real-life problems in surfactant assembly, polymer conformation sampling and coarse-grained protein folding. A conspicuous property of the method is that slow modes and fast modes evolve equally fast, in principle. The Culgi consortium is sponsored by international industry and several EU projects (Multimatdesign, Nanomodel and Selfmem). 2 -Scripted multiscale modeling of heterodisperse industrial surfactants and amphiphilic polymers Industrial surfactants and polymers more than often have one property in common: the molecular distributions are extremely heterodisperse. In a real-life synthetic surfactant application one may easily find O(100) different molecule types; in random polymer melts, such number goes into many orders more, and likewise in crude oil dispersions. Modeling such complex mixtures puts extreme demands to theory and software, yet is very practical and useful. We report on a new scripting approach in the Culgi framework (Chemistry Unified Language Interface), where individual molecular topologies are assigned atomtypes automatically in molecular modelling, then mapped to a coarse grained model and further to self-assembly. The principle of the method is discussed, with special attention to automation. How far can one go? Some examples in detergents modeling and polyolefin blending are used as illustration. The Culgi consortium is sponsored by international industry and several large EU projects (Multimatdesign, Nanomodel, and Selfmem). You might also be interested in the following presentation: 248 - Effect of cellulose surface orientation on contaminant adhesion studied by molecular dynamics simulation Changes in the surface chemistry of cellulose could affect the adhesion of contaminants such as oils. We investigated the adhesion of a simple hydrophobic contaminant on different orientations of a crystalline cellulose surface in the presence of water with molecular dynamics (MD) simulations. Formamide is our initial representative contaminant since it has a hydrophilic amide group as well as a hydrophobic methyl group, the important features of most of the common oily contaminants (e.g. oleic acid). Results suggest that the 100 surface is more hydrophobic compared to the 001 surface; by changing the crystal orientation from 100 to 001, the number of oxygen atoms decreases by 66%, which impacts the surface energy and hydrophilicity. Formamide prefers to be hydrated rather than interact with either cellulose crystal surface. A comparison to other contaminants will be discussed; for instance, a hydrophobic contaminant with an anionic charge reveals other aspects of surface contamination. Please visit the booth #332 (SCM) in the exhibition hall if you are interested in what our software can do for you. < back |