Computational Sciencing Materials Transformation! -Theoretical Understanding & Visual Understanding-

Research Background and Outline

We have been studying the frontiers of computational science in "'chemical' reaction dynamics", theoretically and visually.

While the word "chemistry" might bring to mind images of a 'laboratory', 'white robe', and 'flasks', one major field of today's computational science is occupied with such "chemistry," which seems at a glance to be quite irrelevant to computers. In fact, computational chemistry and theoretical chemistry, promoted by high use of "computer technology", are traditional in the world-class research conducted in Japan. This can be partially recognized from the Nobel Prize in Chemistry awarded to Kenichi Fukui in 1981 for the 'Frontier Electron Theory'.

Under the circumstances, in our lab, we are aiming at clarifying unknown chemical phenomena with developing some new computational scientific methods which are able to deal with "time-dependent and statistical characteristics", in addition to today's elaborate quantum chemical methods. 。

分子計算科学 [References]

Computational Sciences Series Vol.6, 2nd Part: Development of Computational Science
"Computational Science of Molecular Systems - Simulations for Multi-Body Systems Consisting of Electrons and Atoms -"
Chapter 3 Computational Science of Chemical Reactions (Kyoritsu Shuppan Co.,Ltd.)

Research themes

There are mainly two approaches in computational science of chemical reactions. One is the "chemical kinetics" approach. It determines the reaction mechanism of the product experimentally obtained and designs the reaction mechanism to obtain the product that is to be newly synthesized. This is done by establishing, according to the expected reaction mechanism, simultaneous differential equations concerning several kinds of densities of a molecular species (e.g., molecular number density or partial pressure, etc.) and then solving them. It can be said that this is an approach which is experimentally oriented to reproduce an experimental value such as the yield.

The other approach is the "quantum chemistry" approach that treats the molecule from the microscopic viewpoint and deals with the cleavage and the formation of chemical bonds that take place during the reaction from an atomic level. From this standpoint, the purpose is to understand the molecular structure of the product and the rearrangement mechanism of such atoms under the reaction, by determining a molecular chemical reaction at the atomic level, and connecting it to a new molecular synthetic design. Essentially, although a chemical phenomenon that happens in a test tube is unique, there are now two approaches according to the difference in standpoint.

In our laboratory we have been working mainly with the latter, and we have been promoting computational scientific research on material chemical reaction dynamics in condensed systems (i.e., such molecular aggregates as liquids or solids) on the basis of the techniques of quantum and theoretical chemistry. In 1991, we advocated the "chemical reaction" molecular dynamics method, which was not yet known worldwide, and developed the technique of incorporating the microsolvent effect and nonequilibrium nature from the first principle. Moreover, we developed the "molecular frictional theory" using a molecular hamiltonian, which was established on the basis of the electronic state theoretical information, and enabled us to consider the origin of a frictional phenomenon in molecular terms, i.e., intramolecular or intermolecular interactions. Furthermore, we developed the "free energy gradient method" as a structure optimization method for a solvated or adsorbed molecule, and succeeded for the first time in the full optimization of the transition state structure of a Menshutkin reaction in solution, paving the way for wide use.

Future and dreams

Recently, a number of ultra-fast chemical phenomena on the sub-pico to femto-second scale have come to be quantitatively observed experimentally. On the other hand, both "supercomputers and networking techniques" have enabled us to execute "ab initio" multi-scale simulations of complex materials phenomena. Now, we are aiming at constructing some computational scientific methodologies that enable us to understand such super-degrees-of- freedom material systems with an atomic resolution, which can be observed experimentally, and that lead to the microscopic elucidation of (i) "nonequilibrium phenomena" and (ii) an understanding "life phenomena". We are also dreaming of constructing some new theoretical methods to understand "them" continuously making a connection between the first "chemical kinetics" approach and the second "quantum chemistry" one in future.