Group leader: Bang-Gui Liu,
Group members: Ding-Sheng Wang, Yan-Mei Yu(since Aug. 2004)
Research field: Computational Physical Research of Solids and Nanostructured Materials
We study physical properties of solids and nanostructured materials by using first-principle calculations and numerical simulation methods. Our main researches are summarized as four series in the following.
[1] First-principle study and design of half-metallic ferromagnets compatible with semiconductors:
Since electronic spins would likely be used for information process and storage, half-metallic ferromagnetic materials are believed to be useful in achieving novel functional devices that can work well even at nanoscales. On experimental side, zincblende phases of MnAs, CrAs, and CrSb, although being not ground-state phases, have been fabricated successfully, but other zincblende transition-metal pnictides have not been realized. This fact implies that these zincblende epitaxial layers can be grown with the crystal symmetry of substrates only when some necessary conditions are satisfied. Theoretically, this requires that they should be stable mechanically against crystal deformations, and their total energy should not be too high with respect to corresponding ground-state phases or, in another word, their metastable energies should be low enough. We first found the half-metallic ferromagnetism in zincblende phases of CrSb using accurate first-principle calculations. We proposed a reasonable explanation for the half-metallic ferromagnetism in this kind of materials. Furthermore, we explored systematically half-metallic ferromagnets in zincblende transition-metal chalcogenides. We found three better zincblende half-metallic ferromagnets: CrSe, CrTe, and VTe. Compared to the zincblende half-metallic pnictides, these three have similar electronic structures, but have wider half-metallic gaps, better shear elastic constants, and much lower metastable energies. Their stability are even better than the zincblende CrAs that has been fabricated successfully. We also found theoretically half-metallic ferromagnetism in nine of transition-metal pnictides and chalcogenides and in transition-metal doped ZnTe and CdTe.
[2] Magnetism of metal nanoclusters:
Because there exist orbital correlations and sophisticate surface structures in clusters, it is very difficult to explain their magnetism by using first-principle calculations or semi-empirical methods. Using a unified TB method, we have treated successfully the electronic interactions, surface orbital shifts, and orbital correlation. This method was used to study magnetism of Ni nano-clusters of 10-700 atoms. Orbital magnetic moment, being frozen in crystals, was found to enhance substantially in clusters, and play the main role in moment oscillation with cluster size. Spin and orbital moments both were found to increase at the surfaces of clusters depending approximately most on atomic coordination. For clusters less than 20 atoms, the change of d-electron number induced by quantum confinement affected the magnetism substantially.
[3] Theoretical electronic states, magnetism, and phonons of nanostructured materials:
Since carbon nanotubes are unique and very useful, we continued our ab initio research in this field by systematically exploring phonon dispersions of small single-wall carbon nanotubes. Force constants were calculated by norm-conserving pseudopotentials under local density approximation. Dynamical matrices were constructed by cumulant force constant method rather than by interatomic force constant method. Residue forces in relaxed nanotube structures were filtered out by a linear fitting force-constant scheme. Thanks to these improved methods, our phonon dispersions, especially their low energy parts, are much better than existing results, and the phonon soft-modes completely disappear. Raman and IR active modes were reanalyzed and our phonon dispersions agree excellently with the experimental observations. Some issues in low-temperature specific heat and neutron-scattering experiments were well explained with our phonon dispersion. In addition, we showed by using a simple model important effect of electronic interaction in nanostructures on their electronic states; clarified the forming mechanism of intriguing nanomagnetism of the regular arrays of Fe nanostripes on W(110); studied dynamical behaviors of spins of nanostructures under external fields or currents.
[4] Growth of nanostructures on surfaces and phase-field method for large growth systems:
Using numerical simulation methods, we studied some typical growth systems on surfaces, and explored effects of strains induced by lattice mismatch on growth. We, cooperating with En-Ge Wang’s group, found that adsorbed CO molecules affected substantially the growth of Pt islands, appropriate concentration of CO was able to control shape and orientation of the Pt islands. We proposed to use phase filed model to simulate epitaxial growth of two- and three-dimensional islands or thin films on surfaces. A phase-field variable is used to describe stable atomic islands or atomic layers in thin films, and adatoms are described by a local density variable. These two variables satisfy two time-dependent coupled differential equations determining growth processes and final patterns. This method can describe time-dependent atomic growth morphology as kinetic Monte Carlo method does, and at the same time can reflects large-scale properties such as scaling laws, as mean field methods do. It has the merits of these two kinds of methods, but overcomes their defects, can be used to simulate large-scale growth of real nanostructures. We also explored effective mathematical description of strain in hetero-epitaxial growth systems, and its effects on growth processes and morphology. This method has been applied successfully to submonolayer growth systems, growth of large-scale nanostructures, and growth of specific nanostructures on templates. Local growth morphology and mean-field scaling laws both can be obtained in a unified way, and are consistent with experimental observations. This method can be used to growth systems that can not be described conveniently in term of atomic layers.
2. List of main publications or patents (less than 10)
[1] Wen-Hui Xie, Ya-Qiong Xu, Bang-Gui Liu and D. G. Pettifor: Half-metallic ferroma- gnetism and structural stability of zincblende phases of the transition-metal chalcogenides，Phys. Rev. Lett. 91, 037204 (2003).
[2]Bang-Gui Liu: Robust half-metallic ferromagnetism in the zincblende CrSb,Phys. Rev. B 67, 172411 (2003) (http://arxiv.org/ : cond-mat/0206485).
[3]Ya-Qiong Xu, Bang-Gui Liu and D. G. Pettifor: Half-metallic ferromagnetism of MnBi in zincblende structure, Phys. Rev. B 66, 184435 (2002).
[4]Wen-Hui Xie, Bang-Gui Liu, and D. G. Pettifor: Half-metallic ferromagnetism in transition-metal pnictides and chalcogenides with wurtzite structure, Phys. Rev. B 68, 134407 (2003).
[5]Lin-Hui Ye, Bang-Gui Liu, Ding-Sheng Wang, and R.-S. Han: Ab initio phonon dispersions of single-wall carbon nanotubes, Phys. Rev. B 69, 235409 (2004).
[6]X.-G. Wan, L. Zhou, J.-M. Dong, T.-K. Lee, and Ding-Sheng Wang: Orbital Polari- zation, Surface Enhancement and Quantum Confinement in Nano-cluster Magnetism, Phys. Rev. B 69, 174414 (2004).
[7]Jing Wu, E.-G. Wang, K. Varga, Bang-Gui Liu, S. T. Pantelides, and Z.-Y. Zhang: Island shape selection in Pt(111) submonolayer homoepitaxy without or with CO as adsorbates, Phys. Rev. Lett. 89, 146103 (2002).
[8]Yan-Mei Yu and Bang-Gui Liu: Phase-field model of island growth in epitaxy, Phys. Rev. E 69, 021601 (2004).
[9]Yan-Mei Yu and Bang-Gui Liu: Self-organized formation of regular nanostripes on vicinal surfaces, Phys. Rev. B 70, N20 (15 Nov 2004).
[10] Bang-Gui Liu: “Half-metallic ferromagnetism and stability of transition metal pnictides and chalcogenides”, in 'Lecture Notes in Physics' series books: 'Half-metallic Alloys: Fundamentals and Applications', edited by P. H. Dederichs and I. Galanakis,Springer, to appear 2004/2005.
3. List of invited talks at International Conferences (Less than 5)
[1] Bang-Gui Liu: First-principle study and design of half-metallic ferromagnets compatible with semiconductors. The Joint Conference of 'The 6th International Conference of Compu- tational Physics (ICCP6)' and 'Conference of Computational Physics 2003 (CCP2003)', Jade Palace Hotel, Beijing, China, 23-28 May 2004. Plenary Invited Report.
[2] Ding-sheng Wang: Linear and nonlinear optical properties of borate crystals as calcu- lated from the first principles. First Conference of Asian Consortium Computational Mate- rials Science (ACCMS-1), Bangalore, India, 29 Nov - 1 Dec, 2001. Invited Report.
[3] Bang-Gui Liu: Half-metallic ferromagnetism and stability of zincblende transition- metal pnictides and chalcogenides. The 11th International Conference of Composite/NANO Engineering, Hilton Head Island, South Carolina, USA, August 8-14, 2004. Invited Report.
4. Service to the International/Domestic Professional Societies and/or Journals
[1] Wang Ding-Sheng: Chinese Physics Letters, managing editor.
5. Awards & Honors
[1] Bang-Gui Liu (the fourth awardee): Team award of vitally important innovation for 2001-2002 (in Chinese: 2001－2002年度中国科学院重大创新贡献团队), Chinese Academy of Sciences, 2003.
[2] Bang-Gui Liu (the fourth awardee): Kinetical growth of thin films / nanostructures: theoy and experiment (in Chinese: 原子尺度的薄膜/纳米结构生长动力学：理论和实验), The Beijing Award of Science and Technology, The First Class, 2003.
Proposer: Bang-Gui Liu
Members: Ding-Sheng Wang and Yan-Mei Yu
Research Field: Computational physical
research of solids and nanostructured materials
Novel solids and nanostructured materials and their physical properties make one of most important research fields in the world, being promising to be applied in next generation of electronics and computer science. Basic units of current computer chips already reach 100 nm in size and future basic devices concerned should be at most this size, even being tens of nanometers or smaller. For these applications one needs not large bulk solid crystal materials but solid thin film or layer materials with thickness of less than 100nm, or wire or dot materials of similar sizes. Enhanced quantum effect, surface effect, interfacial effect, and electronic interaction in these nanosystems make their electronic, magnetic, and other physical properties substantially different from those in bulk crystalline solids, more complicated. On the other hand, there should exist many new phenomena in the small systems and should appear a lot of new varieties of materials, which makes more opportunity. Scatterings at these scales are reduced so that electrons dynamically become ballistic, and therefore electronic spins can transport coherently and even be used for transporting and treating information. For future spin-based nanodevices, one needs high-spin-polarized materials compatible with existing semiconductor technology. In addition to their potential for applications, these systems themselves are very important for basic scientific research and would lead to discovery of new phenomena and new concepts. With computer becoming more and more powerful, computational physics becomes more and more important. With various high-accuracy and high-efficiency computational methods being available, one not only can study physics of solid materials with accurate first-principle and large-scale numerical simulation methods, but also can predict or design convincingly new solid materials with first-principle methods. Under this large background, it is very important to study the physical properties of solids and nanostrucutred materials by using computational physical methods. These, with experimental study, will promote research in this important field.
Too many things are included in the physical properties of solids and nanostrucutred materials. We hope to concentrate on important issues that are most suitable to computational research. In future three years we shall focus our research on the following three inter-correlated aspects.
It has been found that many of zincblende (or wurtzite) transition-metal pnictides and chalcogenides are half-metallic ferromagnets with 100% spin polarization without taking spin-orbit effect into account. Although being only meta-stable, three of these zincblende phases have been fabricated experimentally. Therefore, it is highly desirable to explore new high-spin-polarized solid materials that not only are compatible with semiconductor technology but also can be fabricated at least as thin films or layers of nanometer thickness. We shall continue to explore new promising high-spin-polarized solid materials satisfying these conditions by including ternary or more components. Furthermore, we shall explore composite structures of these materials and appropriate semiconductors. On the other hand, we shall study unique electronic, spin, and other properties, both static and dynamical, of these solid and composite materials. Our expected goal is to find useful high-spin-polarized solid or composite materials that can be realized at nanometer scales by using accurate first-principle calculations, and to clarify the unique electronic and spin properties, including dynamical transport property, of the solid and composite materials.
We shall continue to study nanomagnetism of nano-clusters and regular arrays of nanostructures on solid surfaces such as Fe nano-stripes on W(110). We shall study dynamical behaviors of spins of nano-systems, especially field or spin-polarized current induced coherent spin reversal. We aim at clarify the mechanism behind these nanomagnetism or the coherent spin reversal. We also plan to study electronic states and magnetism of quantum dots and other typical nanostructures, with the emphasis being on calculating interacting electronic states and magnetic configurations by solving directly quantum mechanical equations concerned. We aim at clarify high-spin states and properties concerned in these nanostructures.
We shall continue to develop our phase-field method, and apply it to study growth of important nanostrucutred materials, including three-dimensional islands. We plan to improve and extend it and finally make it suitable to simulate general growth systems since the phase-field does not essentially need the description in term of layers and, most importantly, this method can describe not only atomic-scale morphology during growth but also large-scale mean-field scaling laws. In addition, The essence of phase-field makes it suitable to study of microscopic dynamical property during surface phase transitions. We shall explore along this direction and hope to finally propose a phase-field theory for studying microscopic dynamical properties of surface phase transitions.
We are capable of accomplishing all tasks for this proposed research. We have led in the research direction: half-metallic ferromagnetism and stability of transition pnictides and chalcogenides. Our works in this direction are included in Springer’s Lecture Notes In Physics series books as the whole second chapter in a new volume on theoretical and experimental progresses of half-metals (see the report for last three years for details).
We have three faculty members in our group. Professor Ding-Sheng Wang is a senior expert in the community of first-principle research of solids and surfaces. He will advise generally the works in this group and will perform some part of them. Bang-Gui Liu is good at first-principle study of solid materials, Momte Carlo simulation of epitaxial growth, and phenomenological study of electronic and spins in solids and nanostructures. He will be responsible for this group and mainly perform with students the first aspect and some part of the second aspect of the proposed works. Some of these works will be done with collaboration with Professor S G Louie in Berkeley. Yan-Mei Yu has worked for several years with phase-field methods. The third aspect will be performed mainly by her with some help from Bang-Gui Liu and maybe others outside of this group.
Calculation jobs without requiring parallelization will be done with the PC cluster of this group. Jobs requiring parallel computations will be done in the IBM p690 computers of the Laboratory. Large-scale massive parallel computations will be done in suitable supercomputers through international cooperation.