Research Center for CD-FMat

breadcrumb navigationOrganization >  Research Center for Computational Design of Advanced Functional Materials  >  Research Teams > First-principles multi-scale simulation Team

First-Principles Multi-Scale Simulation Team (FPMS)

We develop first-principles computational methods and programs and extend them toward multi-scale simulation. To solve practical issues, we try to establish the best scheme and, where appropriate, utilize digital-data-driven methods.

Team's Research Theme

Development and application of QMAS (a first-principles computational code for materials science)

We have been developing a first-principles computational code "QMAS" based on the density-functional theory (DFT) using plane wave basis sets and the projector augmented-wave (PAW) method and applying it to various problems in the materials science field. Recently, we focus on computational features such as maximally-localized Wannier orbitals, two-component spinor formalism and electron-positron two-component DFT.

Figure represents maximally-localized Wannier orbitals constructed from the electronic states contributing the spontaneous polarization. λ=0 and λ=1 correspond to the paraelectric and ferroelectric phases, respectively.

ツ図1
Fig.Wannier orbitals constructed
[S. Ishibashi and K. Terakura, JPSJ 83, 073702 (2014)]

Study of ionic conduction by means of first-principles molecular dynamics

To understand properties of molten salts and electrolytes in batteries, it is very important clarify the ionic conduction mechanism. To simulate realistic systems, large-scale and long-period calculations are required, and we have been developing novel computational techniques.

Figure represents a snapshot in simulation on Mg(BH4)1.6(AlH4)0.4. We performed first-principles molecular dynamics calculations on the high-temperature phase of Mg(BH4)2 and found that mobility of Mg2+ ions is dramatically enhanced when a part of anions by substituting AlH4- for part of anions.

ツ図2
Fig. represents a snapshot in simulation on Mg(BH4)1.6(AlH4)0.4
[T. Ikeshoji et al., RSC Adv. 4, 1366 (2014)]

Development of quantum-chemical computational methods to study practical huge systems

In order to apply quantum chemical calculations to large systems for improving materials properties and/or exploring new features, we have been developing the fragment molecular orbital (FMO) method and related computational techniques. It is expected that ten to a hundred times large molecules can be handled compared to standard computational techniques. We apply this method to clarify dominant factors in realistic phenomena. Recently, we have developed a two-stage parallelization scheme for QM/MM calculations and accomplished faster replica-exchange dynamics. As an application, we investigated isomerization reactions in solvent for H2O2 by 2 ns molecular dynamics calculations.

ツ図1
Fg. An example of a fragment split

Electronic theory for correlated materials and its application to permanent magnets

Since novel high-performance magnets and transition-metal compounds are attracting much attention, we have been developing computational techniques to deal with "correlated" materials and applying them to practical issues. Recently, in order to find novel materials, we try to establish a proper scheme to deal with non-uniformity as well as to increase accuracy.

Figure represents the crystal structure of a permanent magnet compound NdFe12N. We investigated, by first-principles calculations, Ti-substitution effects (To stabilize the crystal structure, a part of Fe atoms are substituted by Ti) and the mechanism of magnetic anisotropy enhancement by N introduction.

ツ図1
Fig. represents the crystal structure of a permanent magnet compound NdFe12N
[T. Miyake et al., JPSJ 83, 043702 (2014)]

Staff Members

position & name
Team Leader
Takashi MIYAKE
Chief Senior Researcher
Shoji ISHIBASHI
Senior Researcher
Eiji TSUCHIDA
Senior Researcher
Dmitri FEDOROV
Postdoctoral Researcher
Taro FUKAZAWA
Postdoctoral Researcher
Nobuya SATO
Postdoctoral Researcher
Arpita VARADWAJ