Large scale molecular simulation

Mike Gillan, Large Scale Molecular Simulation, in Mol. Simul., 14 (4-5), Gordon 8c Breach, Lausanne, 1995. 

Valiev M, BylaskaE, Govind N, Kowalski K, Straatsma T, Van Dam H, Wang D, Nieplocha J, Apra E, Windus T, de Jraig W (2010) NWChem a comprehensive and scalable open-source solution for large scale molecular simulations. Comp Phys Commun 181 1477... 

Gao, G. Large Scale Molecular Simulations With Apphcation to Polymers and Nanoscale Materials. Ph.D. Thesis, California Institute of Technology, Pasadena, CA, USA,... 

S. K. Gray, D. W. Noid and B. G. Sumpter, Symplectic integrators for large scale molecular dynamics simulations A comparison of several explicit methods , J. Chem. Phys., Vol 101, no 5, 4062-72, 1994. 

Yamaguchi, Y. and Gspann, J., Large-Scale Molecular Dynamics Simulations of Cluster Impact and Erosion Processes on a Diamond Surface," Phys. Rev. B, Vol. 66, 2002, pp. 155408-1-10. 

Figure 6.1 Computer applications in catalysis research range all the way from understanding the role of molecular active intermediates to large-scale process simulations.

Kramers idea was to give a more realistic description of the dynamics in the reaction coordinate by including dynamical effects of the solvent. Instead of giving a deterministic description, which is only possible in a large-scale molecular dynamics simulation, he proposed to give a stochastic description of the motion similar to that of the Brownian motion of a heavy particle in a solvent. From the normal coordinate analysis of the activated complex, a reduced mass pi has been associated with the motion in the reaction coordinate, so the proposal is to describe the motion in that coordinate as that of a Brownian particle of mass g in the solvent. 

More realistic models were subsequently developed to incorporate many-body effects and realistic reaction mechanisms. Implementation of these in reduced dimensionality, small-scale simulations demonstrated the importance of the inclusion of these effects [161,162] however, the complexity of the functions and accompanying computational requirements prevent their use in large-scale molecular dynamics simulations in the condensed phase. 

The possible accumulation of negative ions at the air/ water interface was first predicted by Perera and Berkow-itz,8 who found out from molecular dynamics simulations, surprisingly, that the large anions (Cl , Br , and I ) are expelled from water clusters to their interface. Their predictions are supported by the recent large-scale molecular dynamics simulations for the air/water interface of various electrolyte solutions, which reveal that, when the polarization of ions and water molecules is explicitly taken into account, the large anions are accumulated near the interface.9... 

Though not instrumental in nature, another important technique in the polymer arsenal is large-scale computer simulation experiments. These have proved especially useful over the last several years in, for example, molecular-level simulations of polymer mechanical properties [42] and of the transport properties of concentrated polymer solutions [43], Polymers are in many ways ideal objects for this level of simulation study although it is difficult to have accurate detailed knowledge of local interactions, as mentioned earlier, much polymer behavior is dominated by nonlocal interactions that can be much more adequately represented. 

Progress was also reported in modeling the reaction and transportation processes on fuel cell catalysts and through membranes, using multiple paradigms as well as starting from first principle quantum mechanics to train a reactive force field that can be applied for large scale molecular dynamics simulations. It is expected that the model would enable the conception, synthesis, fabrication, characterization, and development of advanced materials and structures for fuel cells . 

D. C. Rapaport, Comput. Phys. Rep., 9,1 (1988). Large-Scale Molecular Dynamics Simulation Using Vector and Parallel Computers. 

In principle, there is no upper limit for the computing resources needed in MD simulations. The modeled systems can always be made larger than the largest systems studied so far. The models can also be made more accurate and brought closer to the fundamental physics. MD simulations can always be performed to cover longer and longer time periods. These aspects are the motivations behind large-scale Molecular Dynamics simulations and we will return to these issues in this Chapter. 

The term large scale in connection to Molecular Dynamics simulations has suffered from a very severe inflation ever since it was invented, in the beginning of the vector-supercomputer era in early 80 s [11-13]. This, of course, has been unavoidable due to the very rapid development in computer technology. For each new generation of hardware and software the limits for how demanding simulations can be carried out have been pushed further away. In this Section an attempt is made to give a semi-quantitative and less diffuse content to large scale MD simulations. 

In the case of large scale Molecular Dynamics simulations, the hardware used has evolved from the vector-based mainframes and supercomputers to parallel computers of different design. Today, high performance computing (HPC) for large scale MD is synonymous with parallel computing. 

S. J. Zhou, D. M. Beazley, P. S. Lomdahl, and B. L. Holian, Large-Scale Molecular Dynamics Simulations of Three-Dimensional Ductile Failure, Phys. Rev. Lett., 78 (1997), 479-482. 

M. R. S. Pinches, D. J. Tildesley, and W. Smith, Large Scale Molecular Dynamics on Parallel Computers Using the Link-Cell Algorithm, Mol. Simul.,6(1991), 51-87. 

Wavelets and multiscale description of surfaces and interfaces MULTISCALE SIMULATION METHODS FOR SOLIDS Large-scale molecular dynamics methods Coupling methods... 

Davis, J., Ozsoy, A., Patel, S., Taufer, M. Towards Large-Scale Molecular Dynamics Simulations on Graphics Processors, Springer, Berlin/Heidelberg, 2009. 

Recently, we have identified the dynamical hinge points of the L1L ribo-zyme using large-scale molecular dynamics (MD) simulations [65]. We have departed from an analysis of the two crystallized conformers and shown using over 600 ns of MD simulations that the transition between on and off conformational states can be almost entirely described by changes in only four virtual torsion angles. Virtual torsions are formed along the virtual bonds between C4- and P atoms and have been shown to be able to discriminate between major RNA folds [66,67],... 

In large scale molecular dynamics simulations an 80 20 binary mixture of Lennard-... 

In Section 3, we showed that large-scale molecular-dynamics (MD) simulations could be used to study the effect of impacts upon perfect crystals of high-explosive diatomic molecules whose interactions are modeled by the reactive empirical bond-order (REBO) potential. We showed that perfect crystal shock simulations lead to detonation above a threshold impact velocity, with characteristics that satisfy the ZND theory of detonations. To see if the threshold for initiation of chemical reaction can be lowered, we also introduced a variety of defects into our samples. 

Large-scale molecular dynamics simulations are producing information on shock-induced reactions on picosecond (ps) to nanosecond (ns) time scales and approaching micron spatial scales. We describe experiments using ultrafast laser methods to produce experimental data on similar time and space scales to help benchmark the simulations as well as motivate their expansion to larger scales and more complicated materials. 

Another important issue is that large-scale molecular dynamics simulations are not poised to get this energy partitioning right without being "taught" the dynamical information that comes from the kinds of calculations discussed here. 

Tang, P., Xu, Y. (2002). From the Cover Large-scale molecular dynamics simulations of general anesthetic effects on the ion channel in the fully hydrated membrane The implication of molecular mechanisms of general anesthesia. Proceedings of the National Academy of Sciences of the United States of America, 99(25), 16035-16040. 

---