ABSTRACT OF THE TALK
ReaxFF is a bond order potential developed by van Duin and Goddard et al., which attracts wide interest with fast growing applications in recent years. Combined with molecular dynamics (MD), ReaxFF is promising in investigating the complex chemistry of large scale systems (~ 1000 - ~10,000 atoms) at high temperature, for no pre-defined reaction pathways are required. However, the computing performance and the capability for analyzing reaction details becomes critical for such applications. This talk will review the methodology development and applications of ReaxFF MD for practical pyrolysis simulation of large organic models. The creation of the first graphics processing unit (GPU)-enabled code (GMD-Reax) allows for efficient ReaxFF MD simulations of models with ~10,000 atoms, which is made possible by one magnitude speed up achieved by GMD-Reax over the CPU ReaxFF MD codes available on desktop workstation with a single GPU attached. To uncover the complex chemistry events, the first code for comprehensive chemical reaction analysis of ReaxFF MD simulations, VARxMD, was developed to catch the overall species and chemical reactions. The combination of GMD-Reax and VARxMD provides a way for direct observing the time evolution of species and underlying reactions for intermediate and radicals in large scale ReaxFF molecular dynamics simulations. GMD-Reax and VARxMD have been employed in ReaxFF simulations of very challenging pyrolysis problems of coal (Hailaer brown coal, Fugu sub-bituminous coal and Liulin bituminus coal), biomass (cellulose and lignin) and polymer (HDPE), as well as for fuel oxidation of bio-oil and RP-1 with multi-components. These applications suggests that large scale ReaxFF MD is a very promising approach for understanding complex reaction mechanisms that is hardly accessible by experiments or other computational methods.
BIOGRAPHY
Xiaoxia Li, B.S. 1985, Tsinghua University; M.S. 1988, Institute of Chemical Metallurgy, Chinese Academy of Sciences; Professor since 2006, Institute of Process Engineering, Chinese Academy of Sciences. She has been working on chemical databases and internet chemistry search engines. Recently she proposed large scale ReaxFF molecular dynamics approach for understanding complex chemical reactions by creating the GPU-accelerated code of GMD-Reax and code of VARxMD for automated reaction analysis, which have been applied to challenging problems for revealing mechanism of complex pyrolysis and combustion systems, including pyrolysis of polymers, biomass and coal, as well as pyrolysis and oxidation of bio-oil and hydrocarbon fuel.
Xiaoxia Li now serves for a number of academic organizations. She is Vice-Chairman, Committee of Computer Chemistry, Chinese Chemical Society; Executive member, Chinese National Committee for CODATA; Trustee, Chemical Structure Association Trust; and Committee member of Women Chemists, Chinese Chemical Society.
Bi-Weekly Seminar Series by the NYU-ECNU Center for Computational Chemistry at NYU Shanghai