CMS seminar2005
Molecular Dynamics Simulation for Metallic Glasses
Dr. Masato Shimono
【Date & Time】21 December 2005 (Wed), 3:30 pm - 5:00 pm
【Place】6F seminar room, Sengen site
【Speaker】Dr. Masato Shimono
【Affiliation】Particle Simulation & Thermodynamics group, NIMS-CMSC
【Title】Molecular Dynamics Simulation for Metallic Glasses
【Abstract】
The glassy state of metals, which is realized by devising alloy combinations, is attracting attention as a new material because it shows electromagnetic and mechanical properties different from those of normal crystalline states. However, not only the mechanism of the vitrification but also the structure of the glass state are not well understood at present. In this regard, the molecular dynamics (MD) method is a simulation method that tracks the movement of individual atoms, and is therefore suitable for elucidating the structure and properties of metallic glasses. However, there is a drawback that the simulation can only be performed in an extremely short time (at most microseconds) due to the limitations of the computer's ability. The following is an example of an attempt to evaluate the glass forming ability of an alloy by the MD method under such constraints.
【Contact】Dr. Yoshitaka Tateyama
【Place】6F seminar room, Sengen site
【Speaker】Dr. Masato Shimono
【Affiliation】Particle Simulation & Thermodynamics group, NIMS-CMSC
【Title】Molecular Dynamics Simulation for Metallic Glasses
【Abstract】
The glassy state of metals, which is realized by devising alloy combinations, is attracting attention as a new material because it shows electromagnetic and mechanical properties different from those of normal crystalline states. However, not only the mechanism of the vitrification but also the structure of the glass state are not well understood at present. In this regard, the molecular dynamics (MD) method is a simulation method that tracks the movement of individual atoms, and is therefore suitable for elucidating the structure and properties of metallic glasses. However, there is a drawback that the simulation can only be performed in an extremely short time (at most microseconds) due to the limitations of the computer's ability. The following is an example of an attempt to evaluate the glass forming ability of an alloy by the MD method under such constraints.
【Contact】Dr. Yoshitaka Tateyama
Ab initio molecular dynamics simulation of molecular radicals in solution
Prof. Michiel Sprik
【Date & Time】16 November 2005 (Wed), 3:30 pm - 5:00 pm
【Place】6F seminar room, Sengen site
【Speaker】Professor Michiel Sprik
【Affiliation】Department of Chemistry, University of Cambridge
【Title】Ab initio molecular dynamics simulation of molecular radicals in solution
【Abstract】
Radical species in solution are of interest because of their high chemical reactivity. The theoretical description of the solvation of radicals is therefore a suitable topic for all-atom density functional methods. It is also a challenge because some of these methods, in particular the density functionals used in ab initio molecular dynamics simulation ("Car-Parrinello") tend to overestimate the reactivity which may go as as far as the radical wrongly attacking the solvent. In this talk we present some of our recent work in this area, Systems we have studied (in collaboration with others) include the hydroxyl (OH), peroxyl (O$_2$H), superoxide (O$_2^-$) and vitamin C radicals in water and also molecular ions (TTF, TH, quinones) in non-aqueous solvents (acetonitril, methanol). We will focus on the solvation and structure of the unpaired electron (or hole).
【Contact】Dr. Yoshitaka Tateyama
【Place】6F seminar room, Sengen site
【Speaker】Professor Michiel Sprik
【Affiliation】Department of Chemistry, University of Cambridge
【Title】Ab initio molecular dynamics simulation of molecular radicals in solution
【Abstract】
Radical species in solution are of interest because of their high chemical reactivity. The theoretical description of the solvation of radicals is therefore a suitable topic for all-atom density functional methods. It is also a challenge because some of these methods, in particular the density functionals used in ab initio molecular dynamics simulation ("Car-Parrinello") tend to overestimate the reactivity which may go as as far as the radical wrongly attacking the solvent. In this talk we present some of our recent work in this area, Systems we have studied (in collaboration with others) include the hydroxyl (OH), peroxyl (O$_2$H), superoxide (O$_2^-$) and vitamin C radicals in water and also molecular ions (TTF, TH, quinones) in non-aqueous solvents (acetonitril, methanol). We will focus on the solvation and structure of the unpaired electron (or hole).
【Contact】Dr. Yoshitaka Tateyama
Quantum Monte Carlo methods for the surface chemistry of oxide materials
Prof. M. J. Gillan
【Date & Time】9 November 2005 (Wed), 3:30 pm - 5:00 pm
【Place】6F seminar room, Sengen site
【Speaker】Professor M. J. Gillan
【Affiliation】Physics and Astronomy Department and London Centre for Nanotechnology, University College London
【Title】Quantum Monte Carlo methods for the surface chemistry of oxide materials
【Abstract】
Oxide materials are important for many reasons: catalysis, corrosion, gas-sensing, coatings, etc... Density functional theory (DFT) has been used for nearly 15 years to study the energetics and structure of oxide surfaces, and the adsorption and reactions of molecules on these surfaces. However, it has become clear over the past five years that the usual DFT methods are not very accurate or reliable. For example, the adsorption energies of molecules are often in error by up to 0.5 eV (up to 10 kcal/mol). This poor accuracy is not good enough for many practical purposes. This seminar will describe our recent efforts at UCL to apply quantum Monte Carlo techniques (QMC) to these problems. The basic ideas of QMC will be summarised, and examples of their use for small molecules will be shown. Some recent technical developments in basis sets for QMC will then be outlined. Then, the results of our recent QMC calculations on the prototype oxide MgO will be presented, and I will also present our encouraging results on the MgO (001) surface. Finally, I plan to mention preliminary QMC work on the adsorption of the water molecule on the MgO (001) surface.
【Contact】Dr. Tsuyoshi Miyazaki
【Place】6F seminar room, Sengen site
【Speaker】Professor M. J. Gillan
【Affiliation】Physics and Astronomy Department and London Centre for Nanotechnology, University College London
【Title】Quantum Monte Carlo methods for the surface chemistry of oxide materials
【Abstract】
Oxide materials are important for many reasons: catalysis, corrosion, gas-sensing, coatings, etc... Density functional theory (DFT) has been used for nearly 15 years to study the energetics and structure of oxide surfaces, and the adsorption and reactions of molecules on these surfaces. However, it has become clear over the past five years that the usual DFT methods are not very accurate or reliable. For example, the adsorption energies of molecules are often in error by up to 0.5 eV (up to 10 kcal/mol). This poor accuracy is not good enough for many practical purposes. This seminar will describe our recent efforts at UCL to apply quantum Monte Carlo techniques (QMC) to these problems. The basic ideas of QMC will be summarised, and examples of their use for small molecules will be shown. Some recent technical developments in basis sets for QMC will then be outlined. Then, the results of our recent QMC calculations on the prototype oxide MgO will be presented, and I will also present our encouraging results on the MgO (001) surface. Finally, I plan to mention preliminary QMC work on the adsorption of the water molecule on the MgO (001) surface.
【Contact】Dr. Tsuyoshi Miyazaki
ntroduction to the NMTO method: Direct generation of Wannier functions and its applications
Dr. Atsushi Yamasaki
【Date & Time】25 October 2005 (Tue), 10:30 am - 12:00 am
【Place】6F seminar room, Sengen site
【Speaker】Dr. Atsushi Yamasaki
【Affiliation】Max-Planck-Institut fuer Festkperforschung, Stuttgart
【Title】ntroduction to the NMTO method: Direct generation of Wannier functions and its applications
【Abstract】
Recently new generation of the LMTO method is developing, so-called NMTO, Nth-order muffin-tin orbital. The keywords, which is nearly connected with each other, are as follows: (1)minimal basis set (by massive downfolding) (2)short-ranged orbitals (by screening transformations) (3)accurate and robust (by overlapping MT-potentials and a polynomial approximation to the energy dependence of the partial-wave set) Those are explained briefly. With the massive downfolding and energy-selective partial waves, it is possible to generate Wannier functions directly, instead of via projection from the delocalized Bloch states. NMTO Wannier functions can be used to visualize chemical bonding and to construct effective (Hubbard-type) Hamiltonians, which are including chemical trends and material dependencies correctly, for strongly correlated systems. The results of the series of 3d(t$_{2g}$)$^1$ perovskites by means of LDA+DMFT are presented. If time permits, other results are also given. These MTO's have significant advantages over those used in the past.
【Contact】Dr. Igor Solovyev
【Place】6F seminar room, Sengen site
【Speaker】Dr. Atsushi Yamasaki
【Affiliation】Max-Planck-Institut fuer Festkperforschung, Stuttgart
【Title】ntroduction to the NMTO method: Direct generation of Wannier functions and its applications
【Abstract】
Recently new generation of the LMTO method is developing, so-called NMTO, Nth-order muffin-tin orbital. The keywords, which is nearly connected with each other, are as follows: (1)minimal basis set (by massive downfolding) (2)short-ranged orbitals (by screening transformations) (3)accurate and robust (by overlapping MT-potentials and a polynomial approximation to the energy dependence of the partial-wave set) Those are explained briefly. With the massive downfolding and energy-selective partial waves, it is possible to generate Wannier functions directly, instead of via projection from the delocalized Bloch states. NMTO Wannier functions can be used to visualize chemical bonding and to construct effective (Hubbard-type) Hamiltonians, which are including chemical trends and material dependencies correctly, for strongly correlated systems. The results of the series of 3d(t$_{2g}$)$^1$ perovskites by means of LDA+DMFT are presented. If time permits, other results are also given. These MTO's have significant advantages over those used in the past.
【Contact】Dr. Igor Solovyev
Interpretation of Hund's multiplicity rule
Dr. Ryo Maezono
【Date & Time】12 October 2005 (Wed), 3:30 pm - 5:00 pm
【Place】6F seminar room, Sengen site
【Speaker】Dr. Ryo Maezono
【Affiliation】First-Principles Simulation group (II), NIMS-CMSC
【Title】Interpretation of Hund's multiplicity rule
【Abstract】
Hund's multiplicity rule is investigated for atomic systems using quantum Monte Carlo methods. Our calculations give an accurate account of electronic correlation and obey the virial theorem to high accuracy. This allows us to obtain accurate values for each of the energy terms and therefore to give a convincing explanation of the mechanism. We find that the energy gain in high spin states with respect to low spin states is due to the greater electron-nucleus attraction in the higher spin state, in accordance with Hartree-Fock calculations and studies including correlation.
【Contact】Dr. Yoshitaka Tateyama
【Place】6F seminar room, Sengen site
【Speaker】Dr. Ryo Maezono
【Affiliation】First-Principles Simulation group (II), NIMS-CMSC
【Title】Interpretation of Hund's multiplicity rule
【Abstract】
Hund's multiplicity rule is investigated for atomic systems using quantum Monte Carlo methods. Our calculations give an accurate account of electronic correlation and obey the virial theorem to high accuracy. This allows us to obtain accurate values for each of the energy terms and therefore to give a convincing explanation of the mechanism. We find that the energy gain in high spin states with respect to low spin states is due to the greater electron-nucleus attraction in the higher spin state, in accordance with Hartree-Fock calculations and studies including correlation.
【Contact】Dr. Yoshitaka Tateyama
New density-functional (DF) MD methods for redox (electron-transfer) reactions
Dr. Yoshitaka Tateyama
【Date & Time】6 October 2005 (Thu), 3:30 pm - 5:00 pm
【Place】6F seminar room, Sengen site
【Speaker】Dr. Yoshitaka Tateyama
【Affiliation】First-Principles Simulation group (I), NIMS-CMSC
【Title】New density-functional (DF) MD methods for redox (electron-transfer) reactions
【Abstract】
Redox (electron transfer) reaction in solution is a vital process in many interesting phenomena such as corrosion, battery, fuel cell, catalysis, photosynthesis. However, few ab-initio methods can compute the key properties of redox reactions. We have developed new theoretical approaches, 'grand-canonical DF-MD' and 'DF-MD+energy gap formula' methods, for quantitative calculation of such quantities. Besides, we demonstrated the validity of these two approaches by using model redox reactions of transition-metal complexes in aqueous solution. In this talk, the methodology, physical implication and typical accuracy will be presented.
【Contact】Dr. Taizo Sasaki
【Place】6F seminar room, Sengen site
【Speaker】Dr. Yoshitaka Tateyama
【Affiliation】First-Principles Simulation group (I), NIMS-CMSC
【Title】New density-functional (DF) MD methods for redox (electron-transfer) reactions
【Abstract】
Redox (electron transfer) reaction in solution is a vital process in many interesting phenomena such as corrosion, battery, fuel cell, catalysis, photosynthesis. However, few ab-initio methods can compute the key properties of redox reactions. We have developed new theoretical approaches, 'grand-canonical DF-MD' and 'DF-MD+energy gap formula' methods, for quantitative calculation of such quantities. Besides, we demonstrated the validity of these two approaches by using model redox reactions of transition-metal complexes in aqueous solution. In this talk, the methodology, physical implication and typical accuracy will be presented.
【Contact】Dr. Taizo Sasaki
All electron calculations in condensed matter: the GAPW approach
Dr. Marcella Iannuzzi
【Date & Time】3 October 2005 (Mon), 3:30 pm - 5:00 pm
【Place】6F seminar room, Sengen site
【Speaker】Dr. Marcella Iannuzzi
【Affiliation】Physical Chemistry Institute, University of Zurich
【Title】All electron calculations in condensed matter: the GAPW approach
【Abstract】
Our recent work focused on the development of methods to perform Kohn--Sham calculations using accurate basis sets on large systems, including condensed matter systems (periodic boundary conditions). The Gaussian Augmented Plane Wave method (GAPW) is a hybrid method that combines the advantages of the plane waves approach with the expansion of the KS orbitals in terms of localized functions, thus achieving significant improvements in accuracy and efficiency. Moreover, it can be extended to all electron calculations, which are mandatory to reproduce with high accuracy those properties, such as magnetic response or the x-ray absorption/emission spectra, which depend critically on the details of the wavefunctions in the core regions. Some results are presented about x-ray spectroscopy simulations for water systems. For the calculation of properties like the excitation energies and the chemical shifts, we reformulated the generalized density functional perturbation theory in terms of the GAPW scheme. Preliminary results are given to demonstrate accuracy and performance of the method.
【Contact】Dr. Yoshitaka Tateyama
【Place】6F seminar room, Sengen site
【Speaker】Dr. Marcella Iannuzzi
【Affiliation】Physical Chemistry Institute, University of Zurich
【Title】All electron calculations in condensed matter: the GAPW approach
【Abstract】
Our recent work focused on the development of methods to perform Kohn--Sham calculations using accurate basis sets on large systems, including condensed matter systems (periodic boundary conditions). The Gaussian Augmented Plane Wave method (GAPW) is a hybrid method that combines the advantages of the plane waves approach with the expansion of the KS orbitals in terms of localized functions, thus achieving significant improvements in accuracy and efficiency. Moreover, it can be extended to all electron calculations, which are mandatory to reproduce with high accuracy those properties, such as magnetic response or the x-ray absorption/emission spectra, which depend critically on the details of the wavefunctions in the core regions. Some results are presented about x-ray spectroscopy simulations for water systems. For the calculation of properties like the excitation energies and the chemical shifts, we reformulated the generalized density functional perturbation theory in terms of the GAPW scheme. Preliminary results are given to demonstrate accuracy and performance of the method.
【Contact】Dr. Yoshitaka Tateyama
First Principles Computational Studies of Hydrogen Storage in Pure Mg and in Novel Mg Nanocomposites
Prof. Sean C. Smith
【Date & Time】14 September 2005 (Wed), 10:30 pm - 11:30 pm
【Place】6F seminar room, Sengen site
【Speaker】Professor Sean C. Smith
【Affiliation】Centre for Computational Molecular Science, The University of Queensland
【Title】First Principles Computational Studies of Hydrogen Storage in Pure Mg and in Novel Mg Nanocomposites
【Abstract】
We have recently been intensively involved in first principles computational modelling of hydrogen dissociative chemisorption, diffusion and desorption with pure Mg and Mg-based nanocomposite materials, with a focus on the development of effective hydrogen storage media. A key challenge to be addressed is the integration of catalytic effects for each of these components of the hydrogen storage process within a single, easily manufacturable material. Our calculations utilise density functional theory with the generalised gradient approximation, coupled with the nudged elastic band method for determination of reaction paths and barriers. The computational results assist in the interpretation of experimental hydrogen storage studies for nanocomposites produced by ball-milling.
【Contact】Dr. Hidehiro Onodera
【Place】6F seminar room, Sengen site
【Speaker】Professor Sean C. Smith
【Affiliation】Centre for Computational Molecular Science, The University of Queensland
【Title】First Principles Computational Studies of Hydrogen Storage in Pure Mg and in Novel Mg Nanocomposites
【Abstract】
We have recently been intensively involved in first principles computational modelling of hydrogen dissociative chemisorption, diffusion and desorption with pure Mg and Mg-based nanocomposite materials, with a focus on the development of effective hydrogen storage media. A key challenge to be addressed is the integration of catalytic effects for each of these components of the hydrogen storage process within a single, easily manufacturable material. Our calculations utilise density functional theory with the generalised gradient approximation, coupled with the nudged elastic band method for determination of reaction paths and barriers. The computational results assist in the interpretation of experimental hydrogen storage studies for nanocomposites produced by ball-milling.
【Contact】Dr. Hidehiro Onodera
Computer simulation of grain growth in three dimensions by the phase field model
Dr. Yoshihiro Suwa
【Date & Time】21 July 2005 (Thu), 3:30 pm - 5:00 pm
【Place】6F seminar room, Sengen site
【Speaker】Dr. Yoshihiro Suwa
【Affiliation】Particle Simulation & Thermodynamics group, NIMS-CMSC
【Title】Computer simulation of grain growth in three dimensions by the phase field model
【Abstract】
The grain structure of polycrystalline materials containing metals is a key factor in determining physical properties. Since it is difficult to incorporate the geometrical features of the structure directly into the analytical theory of grain growth, computer simulation is often used, and various models such as the Monte Carlo method have been proposed. In recent years, attempts have been made to apply the phase-field method, which has made remarkable progress in the field of computational materials science, to grain growth simulation. In this presentation, we report the Zener pinning effect of the second phase dispersed particles on the grain growth, in addition to the calculation results of normal grain growth and abnormal grain growth (secondary recrystallization), comparing the simulation results of three-dimensional polycrystalline growth by the phase-field method with those by other methods.
【Contact】Dr. Toshiyuki Koyama
【Place】6F seminar room, Sengen site
【Speaker】Dr. Yoshihiro Suwa
【Affiliation】Particle Simulation & Thermodynamics group, NIMS-CMSC
【Title】Computer simulation of grain growth in three dimensions by the phase field model
【Abstract】
The grain structure of polycrystalline materials containing metals is a key factor in determining physical properties. Since it is difficult to incorporate the geometrical features of the structure directly into the analytical theory of grain growth, computer simulation is often used, and various models such as the Monte Carlo method have been proposed. In recent years, attempts have been made to apply the phase-field method, which has made remarkable progress in the field of computational materials science, to grain growth simulation. In this presentation, we report the Zener pinning effect of the second phase dispersed particles on the grain growth, in addition to the calculation results of normal grain growth and abnormal grain growth (secondary recrystallization), comparing the simulation results of three-dimensional polycrystalline growth by the phase-field method with those by other methods.
【Contact】Dr. Toshiyuki Koyama
Terahertz Emission using High Temperature Superconductors
Prof. Masashi Tachiki
【Date & Time】14 July 2005 (Thu), 3:30 pm - 5:00 pm
【Place】8F large seminar room, Sengen site
【Speaker】Professor Masashi Tachiki
【Affiliation】NIMS
【Title】Terahertz Emission using High Temperature Superconductors
【Abstract】
The structure of high temperature superconductors comprise stacks of naturally-built Josephson junctions. These junctions sustain a new form of excitation, i.e. Josephson plasmas (falling in the terahertz range) which lie within the superconducting energy gap and is thus protected from Landau damping. If a method of exciting this state can be devised, it will therefore offer a promising way of emitting electromagnetic fields in the Terahertz range. The above scenario can be formulated into a set of nonlinear equations. I present solutions of these equations utilizing the Earth Simulator, and compare them with recent experiments.
【Contact】Dr. Akihiro Tanaka
【Place】8F large seminar room, Sengen site
【Speaker】Professor Masashi Tachiki
【Affiliation】NIMS
【Title】Terahertz Emission using High Temperature Superconductors
【Abstract】
The structure of high temperature superconductors comprise stacks of naturally-built Josephson junctions. These junctions sustain a new form of excitation, i.e. Josephson plasmas (falling in the terahertz range) which lie within the superconducting energy gap and is thus protected from Landau damping. If a method of exciting this state can be devised, it will therefore offer a promising way of emitting electromagnetic fields in the Terahertz range. The above scenario can be formulated into a set of nonlinear equations. I present solutions of these equations utilizing the Earth Simulator, and compare them with recent experiments.
【Contact】Dr. Akihiro Tanaka
Merging first-principles and model approaches: GW+DMFT
Dr. Ferdi Aryasetiawan
【Date & Time】23 June 2005 (Thu), 3:00 pm - 4:30 pm
【Place】6F seminar room, Sengen site
【Speaker】Dr. Ferdi Aryasetiawan
【Affiliation】AIST-RICS
【Title】Merging first-principles and model approaches: GW+DMFT
【Abstract】
First-principles calculations reveal a lot more details about the system than model calculations can hope to do. On the other hand, model approaches are theoretically more sophisticated than first-principles approaches. By combining the GW method, which is a successful first-principles method beyond the LDA, and dynamical mean-field theory (DMFT), which is traditionally applied to study strongly correlated systems in the model context, we obtain a first-principles scheme that incorporates the strength of DMFT in treating systems with strong onsite correlations. The new scheme was tested in ferromagnetic nickel with encouraging results.
【Contact】Dr. Hiori Kino
【Place】6F seminar room, Sengen site
【Speaker】Dr. Ferdi Aryasetiawan
【Affiliation】AIST-RICS
【Title】Merging first-principles and model approaches: GW+DMFT
【Abstract】
First-principles calculations reveal a lot more details about the system than model calculations can hope to do. On the other hand, model approaches are theoretically more sophisticated than first-principles approaches. By combining the GW method, which is a successful first-principles method beyond the LDA, and dynamical mean-field theory (DMFT), which is traditionally applied to study strongly correlated systems in the model context, we obtain a first-principles scheme that incorporates the strength of DMFT in treating systems with strong onsite correlations. The new scheme was tested in ferromagnetic nickel with encouraging results.
【Contact】Dr. Hiori Kino
Charge localization phenomena in synthetic DNA
Prof. Mauro Boero
【Date & Time】9 June 2005 (Thu), 1:30 pm - 2:30 pm
【Place】6F seminar room, Sengen site
【Speaker】Professor Mauro Boero
【Affiliation】Institute of Physics, University of Tsukuba
【Title】Charge localization phenomena in synthetic DNA
【Abstract】
Charge transfer in DNA is currently the subject of intense theoretical and experimental investigation. This is due both to a possible use of DNA as a component in nanoelectronic and electrochemical devices and to the fundamental role of conductivity in the oxidative damage and mutations of DNA. By using Car-Parrinello molecular dynamics, we study the mechanism of positive charge and electron hole localization in a laboratory realizable radical cation Z-DNA crystal[1]. We find that at room temperature structural deformation are not sufficient to provide an efficient localization mechanism. Instead we find evidence for both an ion-gated and proton-coupled mechanism[2]. Namely, a hole h+ can be localized by two mechanisms: (i) proton shift or (ii) fluctuations in the solvation shell[3] (with some warning). Between these two scenarios, the proton-coupled charge transfer mechanism seems to provide the best agreement and the key to interpret EPR and H/D substitution experiments. Due to the large size of the full quantum system, this calculation was performed on the Earth Simulator (ES) computer facility[4]. Work is now in progress on a hybrid QMMM system, coupled to metadynamics, to work out charge localization reaction paths and related activation barriers: few details will be given as closing remarks.
【Contact】Dr. Hiori Kino
【Place】6F seminar room, Sengen site
【Speaker】Professor Mauro Boero
【Affiliation】Institute of Physics, University of Tsukuba
【Title】Charge localization phenomena in synthetic DNA
【Abstract】
Charge transfer in DNA is currently the subject of intense theoretical and experimental investigation. This is due both to a possible use of DNA as a component in nanoelectronic and electrochemical devices and to the fundamental role of conductivity in the oxidative damage and mutations of DNA. By using Car-Parrinello molecular dynamics, we study the mechanism of positive charge and electron hole localization in a laboratory realizable radical cation Z-DNA crystal[1]. We find that at room temperature structural deformation are not sufficient to provide an efficient localization mechanism. Instead we find evidence for both an ion-gated and proton-coupled mechanism[2]. Namely, a hole h+ can be localized by two mechanisms: (i) proton shift or (ii) fluctuations in the solvation shell[3] (with some warning). Between these two scenarios, the proton-coupled charge transfer mechanism seems to provide the best agreement and the key to interpret EPR and H/D substitution experiments. Due to the large size of the full quantum system, this calculation was performed on the Earth Simulator (ES) computer facility[4]. Work is now in progress on a hybrid QMMM system, coupled to metadynamics, to work out charge localization reaction paths and related activation barriers: few details will be given as closing remarks.
【Contact】Dr. Hiori Kino
Photonic Anderson Model -Hybridization theory between localized mode and free propagating mode for light scattering from a dielectric sphere-
Dr. Junichi Inoue
【Date & Time】2 June 2005 (Thu), 3:30 pm - 5:00 pm
【Place】6F seminar room, Sengen site
【Speaker】Dr. Junichi Inoue
【Affiliation】NIMS-ICYS / NIMS nanomaterials lab.
【Title】Photonic Anderson Model -Hybridization theory between localized mode and free propagating mode for light scattering from a dielectric sphere-
【Abstract】
Light scattering from a homogeneous dielectric sphere is discussed in terms of hybridization between a localized mode excited inside the dielectric sphere and free propagating modes in vacuum. This theory is a photonic counterpart of the Anderson model in electron systems, yielding rigorous theoretical foundation of the heavy photon concept, which was numerically proposed for almost flat photonic bands. The magnitude of the hybridization is analytically expressed. The localized mode is identified with the photon virtual bound state. In order to confirm the validity of the present theory, a comparison is made between the present theory and conventional numerical calculation for results of the photonic density of states.
【Contact】Dr. Akihiro Tanaka
【Place】6F seminar room, Sengen site
【Speaker】Dr. Junichi Inoue
【Affiliation】NIMS-ICYS / NIMS nanomaterials lab.
【Title】Photonic Anderson Model -Hybridization theory between localized mode and free propagating mode for light scattering from a dielectric sphere-
【Abstract】
Light scattering from a homogeneous dielectric sphere is discussed in terms of hybridization between a localized mode excited inside the dielectric sphere and free propagating modes in vacuum. This theory is a photonic counterpart of the Anderson model in electron systems, yielding rigorous theoretical foundation of the heavy photon concept, which was numerically proposed for almost flat photonic bands. The magnitude of the hybridization is analytically expressed. The localized mode is identified with the photon virtual bound state. In order to confirm the validity of the present theory, a comparison is made between the present theory and conventional numerical calculation for results of the photonic density of states.
【Contact】Dr. Akihiro Tanaka
Correlated Electron-Ion Dynamics: quantum molecular dynamics with exchange of energy between ions and electrons
Dr. David Bowler
【Date & Time】26 May 2005 (Thu), 2:30 pm - 4:00 pm
【Place】6F seminar room, Sengen site
【Speaker】Dr. David Bowler
【Affiliation】NIMS-ICYS / Dept. of Physics and Astronomy, University College London
【Title】Correlated Electron-Ion Dynamics: quantum molecular dynamics with exchange of energy between ions and electrons
【Abstract】
Correlated Electron-Ion Dynamics (CEID) is an extension of molecular dynamics that allows us to introduce in a correct manner the exchange of energy between electrons and ions. The formalism is based on a systematic approximation (Small Amplitude Moment Expansion - SAME). I will introduce the formalism and discuss two research directions: implementation with open boundaries, and application to inelastic tunneling spectroscopy. (This work was done in collaboration with A.J.Fisher(UCL), A.P.Horsfield(UCL), C.Sanchez(Queen's Univ. of Belfast) and T.Todorov(Queen's Univ. of Belfast).)
【Contact】Dr. Yoshitaka Tateyama
【Place】6F seminar room, Sengen site
【Speaker】Dr. David Bowler
【Affiliation】NIMS-ICYS / Dept. of Physics and Astronomy, University College London
【Title】Correlated Electron-Ion Dynamics: quantum molecular dynamics with exchange of energy between ions and electrons
【Abstract】
Correlated Electron-Ion Dynamics (CEID) is an extension of molecular dynamics that allows us to introduce in a correct manner the exchange of energy between electrons and ions. The formalism is based on a systematic approximation (Small Amplitude Moment Expansion - SAME). I will introduce the formalism and discuss two research directions: implementation with open boundaries, and application to inelastic tunneling spectroscopy. (This work was done in collaboration with A.J.Fisher(UCL), A.P.Horsfield(UCL), C.Sanchez(Queen's Univ. of Belfast) and T.Todorov(Queen's Univ. of Belfast).)
【Contact】Dr. Yoshitaka Tateyama
Construction of low-energy effective Hamiltonians using Wannier functions formalism
Dr. Igor Solovyev
【Date & Time】18 May 2005 (Wed), 4:00 pm - 5:30 pm
【Place】6F seminar room, Sengen site
【Speaker】Dr. Igor Solovyev
【Affiliation】First-Principles Simulation group (II), NIMS-CMSC
【Title】Construction of low-energy effective Hamiltonians using Wannier functions formalism
【Abstract】
There is a large class of compounds whose electronic properties are predetermined by the behavior of a limited number of bands located near the Fermi level and well isolated from the rest of the states, and those electronic structure cannot be described properly using conventional methods based on the local-density approximation (LDA). Several typical examples are the perovskites (SrVO$_3$, LaTiO$_3$, etc.), V$_2$O$_3$, Y$_2$Mo$_2$O$_7$, NaCoO$_2$, and many others. The source of the problem is known to be the on-site Coulomb correlations, whose form is greatly oversimplified in LDA. Therefore, many attempts have been done in order to incorporate the physics of on-site Coulomb correlations in LDA. I will discuss the systematic procedure for constructing the effective low-energy Hubbard-type Hamiltonian for these systems using results of first-principle calculations. This procedure consists of three parts: (1) The derivation of the kinetic-energy part using the downfolding method. (2) Construction of Wannier functions. (3) Calculation of screened Coulomb interactions in the basis of Wannier functions. I will also show applications for transition-metal oxides.
【Contact】Dr. Yoshitaka Tateyama
【Place】6F seminar room, Sengen site
【Speaker】Dr. Igor Solovyev
【Affiliation】First-Principles Simulation group (II), NIMS-CMSC
【Title】Construction of low-energy effective Hamiltonians using Wannier functions formalism
【Abstract】
There is a large class of compounds whose electronic properties are predetermined by the behavior of a limited number of bands located near the Fermi level and well isolated from the rest of the states, and those electronic structure cannot be described properly using conventional methods based on the local-density approximation (LDA). Several typical examples are the perovskites (SrVO$_3$, LaTiO$_3$, etc.), V$_2$O$_3$, Y$_2$Mo$_2$O$_7$, NaCoO$_2$, and many others. The source of the problem is known to be the on-site Coulomb correlations, whose form is greatly oversimplified in LDA. Therefore, many attempts have been done in order to incorporate the physics of on-site Coulomb correlations in LDA. I will discuss the systematic procedure for constructing the effective low-energy Hubbard-type Hamiltonian for these systems using results of first-principle calculations. This procedure consists of three parts: (1) The derivation of the kinetic-energy part using the downfolding method. (2) Construction of Wannier functions. (3) Calculation of screened Coulomb interactions in the basis of Wannier functions. I will also show applications for transition-metal oxides.
【Contact】Dr. Yoshitaka Tateyama