|
Yongbin Lee Assistant Scientist - Ames Laboratory - USDOE
|
![]() |
||
|
Education
Experience
Publications
Research Interests Thanks to the increasing computational power and developed algorithms, first principles methods are able to handle complex materials that have more than hundreds atoms per cell and to produce accurate enough results that are able to predict material properties and to provide theoretical analysis of fine and delicate experimental results. Modern calculations are also able to screen many possibilities - like thought experiments - fast enough to guide new experiments. My research is focused on the theoretical investigation of electronic and magnetic properties of complex novel materials (mostly, but not limited, rare-earth compounds) through first principles methods. X-ray Magnetic Circular Dichroism (XMCD) & X-ray Resonant Magnetic Scattering (XRMS) Using X-ray techniques for investigation of magnetic properties of materials is not an easy task because of the many orders of magnitude stronger charge scattering. However, with new theoretical insights (e.g. sum rules) and the advent of new generation photon sources and techniques, x-ray methods are becoming a powerful tool for magnetic properties investigation. X-ray Magnetic Circular Dichroism (XMCD) and X-ray Resonant Magnetic Scattering (XRMS) are fairy new techniques that involve transitions from well understood core levels to unoccupied valence states. The measured spectra can yield information about the spin polarization and orbital moments of final empty states. Furthermore, the information obtained is element and orbital specific since the technique requires scanning through specific absorption edges. Since, in the rare-earth compounds, 5d electrons play an essential role in coupling the rare-earth 4f moments with each other (RKKY mechanism) and in coupling the rare-earth 4f and transition metal 3d moments, it is important to understand 5d states for investigation of magnetic properties. The L-edges spectra that involve electronic dipole transitions from 2p1/2 (L2) and 2p3/2 (L3) core levels to the empty 5d states contain information about 5d states. However the interpretations of L-edge spectra are not simple and not straightforward tasks because of solid-state effects - such as 4f-5d exchange, spin-orbit interactions, crystal field, and band structure effects. While two sum rules derived from atomic models and applied to the experimental data are able to produce rather accurate information for local d-state moments in 3d and 4d elements, they have failed for rare-earth 5d states because of the strong exchange interaction between 4f-5d states and corresponding strong 2p-5d transition matrix element effects. Therefore, it is important to consider solid-state effects with a high degree of accuracy for complete understanding and interpreting of L-edge spectra of rare earth containing compounds. However most of theoretical investigations have employed a rather simple parameterized atomic model that does not include solid-state effects completely and uses parameters that are hard to justify. While first principles calculations are able to consider solid-state effects accurately, because of the complexity, there are few groups in the world (can be counted by figures of one hand) able to seriously investigate the physics of XMCD(XRMS) L-edge spectra of rare earths with first principles. I believe no other group right now has the experience and insight that we have gained in the last five years, and there are now opportunities to push XMCD/XRMS to a new level for quantitative analysis of rare earth related magnetism. There was a problem, the so-called “branching ratio” problem - ratio between L3 and L2 spectra intensity. We were able to show quantitatively that the spectrum is affected greatly by energy and spin dependent matrix elements, with the 5d band spin-orbit interactions playing the dominant role in determining the ratio. This was a completely missing component of the atomic model proposal which had previously attempted to explain the ratio. Er2Fe17 has interesting features in its XMCD L-edge spectra (of Er) that depend on temperature. At high temperature - higher than the Er 4f electron ordering but lower than Fe 3d electron ordering – the L3 and L2 spectra have same sign; but with decreasing temperature (below the Er 4f ordering temperature), the intensity of the L3 spectra becomes much stronger and the L2 spectra changes sign. At low enough temperature that Er 4f electrons are completely ordered the sign of L3 and L2 become completely opposite. Without the calculation, people attributed this to spin reorientation phenomena and magnetic anisotropy, but in my calculation we successfully demonstrated that these trends originate from the competition between the Fe 3d and Er 4f influence on the Er 5d state ordering rather than by magnetic anisotropy or spin reorientation. Our work showed the strong solid-state effects on the L-edge spectra of rare earths and the importance of first principles calculations for the interpretation of spectra. Iron-Arsenic Based Superconductivity Since its superconducting transition temperature was rather high to be explained within Bardeen-Cooper-Schrieffer(BCS) theory, the discovery of superconductivity on the Iron-Arsenic based materials brought great excitement and attracted very intensive investigations to understand the mechanism of pairing and basic properties. To contribute these efforts and to support experimental results (ARPES, Neutron, X-ray scattering) we have performed the electronic structure calculations of magnetic, non-magnetic, doped, undoped, and two different structures for RFeAsO1-xFx (R=La, Nd, Pr), AFe2As2 (A=Ba, Sr, Ca). With Fermi surface and Band structure calculation we successfully supported the ARPES group who performed the first ARPES measurement for the NdFeAsO1-xFx. For CaFe2As2 the pressure experiment (by Neutron, X-ray scattering group) discovered new phase structure - so called collapsed tetragonal phase – that has no magnetic ordering and has more than 12% reduction in c/a ratio compared to ambient pressure structure. This discovery is very important and excited result because it may provide a plausible pathway for superconducting phase transition. For supporting this discovery, we have performed total energy calculations and we found that the orthorhombic states is the ground state for ambient pressure and with 5% volume reduction and c/a= 2.67 (experimentally 5% volume reduction and c/a =2.65) the collapsed tetragonal structure is the one. Our calculation shows beautiful agreement with the pressure experiment. We are also investigating effects of Fe-As phonon mode on the phase instability of collapsed tetragonal structure and superconducting ordering. Structure & Magnetic Phase Transition (Gd5Si2Ge2) Gd5Si2Ge2 is a fascinating material as a strong candidate for near room temperature magnetic refrigeration. The magneto-structural phase transition in this material involves large shear displacements of Gd-containing slabs along the “a” direction of the orthorhombic unit cell. In this work we confirmed quantitatively the idea that this large movement breaks the covalent bonds between Ge atoms and weakens the magnetic coupling between slabs. Through first principles calculations, we were able to explain that the Ge site magnetic moment (it is interesting because Ge is not usually considered a magnetic material) that was observed by the X-ray Magnetic Circular Dichroism (at the K edge of Ge) method originates from the hybridization between Ge p-orbital states and Gd d-orbital states. Furthermore we were able to show quantitatively the relation between the weaken bonding and the magnetic coupling between the slabs through Gd-Gd exchange constant calculations. Positive Exchange Bias – LSMO/SRO Bilayer Exchange bias is created by interfacial magnetic interactions. The magnetic hysteresis loop of ferromagnetic layer that is deposited on another magnetic material is shifted along the field axis and this shift is represented as an exchange field. In this work, we investigated the physical origins that give the observed anti-ferromagnetic exchange interaction of La2/3Sr1/3MnO3/SrRuO3(LSMO/SRO) bilayers. Each material by itself is ferromagnetic. We performed LDA+U calculations to properly treat the localized Mn d states and used a supercell calculation for the layered structure. To compare positive and negative exchange interactions, we calculated both ferro and anti-ferro spin configurations and found that anti-ferro spin configuration between layers had lower total energy. We also identified the importance of the hybridization of interfacial O 2p with 3d states of Mn atoms and 4d states of Ru. The hybridization which is affected by the Mn spin flipping produces lower energy bonding states of interfacial O atoms in the anti-ferromagnetic spin configuration, which is the origin of the preferred anti-ferromagnetic interlayer coupling. This work was motivated by experiments performed at the University of Wisconsin. Ferromagnetic Shape Memory and Martensitic Phase Transition – Ni2MnGa Ni2MnGa is attractive because it exhibits a shape memory effect related to a martensitic transition, which can be controlled by external magnetic fields as well as by temperature (practical range) for use in sensors and actuators. In this work, we were interested in a precursor phenomenon related to the premartensitic phase transition. Neutron scattering of this material observed that one of phonon branch has a strong temperature dependent dip. By temperature dependent Fermi surface and generalized susceptibility calculations, we identified that this soft mode is directly related to Fermi-surface nesting effect. Super Hard Materials & Elastic Properties – AlMgB14 Ames Laboratory Scientists discovered the super-hard property of AlMgB14 of which hardness could be reach to 40-46 GPa (Diamond ~ 70GPa, BN ~ 45-50 GPa) with small chemical modification. We calculated the electronic structure and elastic properties for both the ideal 64 atoms/cell material and the real 62 atoms/cell material with 2 vacancies at metal sites. To check small chemical modification, we also calculated Si added cases at the vacancies sites. The calculated shear modulus is consistent with the empirical proportionality observed between shear moduli and micro-hardness. |
|||
|