Bulletin of the American Physical Society
APS March Meeting 2017
Volume 62, Number 4
Monday–Friday, March 13–17, 2017; New Orleans, Louisiana
Session A37b: Metal Insulator Phase Transitions I: Theory |
Hide Abstracts |
Sponsoring Units: DCMP Chair: Nicola Lanata, National High Magnetic Field Laboratory: Florida State University Room: 384 |
Monday, March 13, 2017 8:00AM - 8:12AM |
A37b.00001: Resolving the chicken-and-egg problem in VO$_2$: a new paradigm for the Mott transition Oscar Najera, Marcello Civelli, Vladimir Dobrosavljevi, Marcelo Rozenberg We consider a minimal model to investigate the metal-insulator transition in VO$_2$. We adopt a Hubbard model with two orbital per unit cell, which captures the competition between Mott and singlet-dimer localization. We solve the model within Dynamical Mean Field Theory, characterizing in detail the metal-insulator transition and finding new features in the electronic states. We compare our results with available experimental data obtaining good agreement in the relevant model parameter range. Crucially, we can account for puzzling optical conductivity data obtained within the hysteresis region, which we associate to a novel metallic state characterized by a split heavy quasiparticle band. Our results show that the thermal-driven insulator-to-metal transition in VO$_2$ is entirely compatible with a Mott electronic mechanism, solving a long standing "chicken-and-egg" debate and calling for further research of ``Mottronics'' applications of this system. [Preview Abstract] |
Monday, March 13, 2017 8:12AM - 8:24AM |
A37b.00002: Nature of the metal--insulator transition in oxide interfaces Michael Osofsky, Joseph Prestigiacomo, Sandra Hernández-Hangarter, Anindya Nath, Virginia Wheeler, Scott Walton, Rachel Myers-Ward, Clifford Krowne, Kurt Gaskill, Konrad Bussmann, Kristin Charipar, Christopher Chervin, Debra Rolison, Michael Veit, Yuri Suzuki One of the many unusual properties of several two-dimensional (2D) oxide interface systems (e.g., LaAlO$_{\mathrm{3}}$/SrTiO$_{\mathrm{3}})$ is the presence of a metal--insulator transition (MIT). This feature contradicts the famous prediction of Abrahams, et al. that all two-dimensional systems must be insulating. Since the MIT is a quantum phase transition (one that occurs at T$=$0K) the transport properties should be independent of the chemical and structural details of the system. Indeed, recent work has demonstrated that a generic phase diagram for the 2D MIT can be constructed for two very different systems: 1) highly disordered RuO$_{\mathrm{2}}$ nanoskins and 2) plasma-functionalized graphene. This phase diagram consists of three regions: metallic, weakly localized insulator with conductivity, conductivity\textasciitilde logT, and strongly localized insulator. We will present details of the transport properties of the disordered RuO$_{\mathrm{2}}$ nanoskins and plasma-functionalized graphene near their respective MITs. We will then present transport results for several gated oxide interface systems near their MITs and compare them with those for the RuO$_{\mathrm{2}}$ nanoskins and functionalized graphene. [Preview Abstract] |
Monday, March 13, 2017 8:24AM - 8:36AM |
A37b.00003: Tuning the metal-insulator transition of VO2 by introducing W dopants via a combinatorial approach Yangang Liang, Seunghun Lee, Xiaohang Zhang, Ichiro Takeuchi We have systematically studied the structural phase transition and the electronic properties of composition spread V$_{\mathrm{1-x}}$W$_{\mathrm{x}}$O$_{\mathrm{2}}$ (0 $\le $ x $\le $ 0.037) thin films fabricated on silicon (001) and c-cut sapphire substrates through combinatorial pulsed laser deposition of a V$_{\mathrm{2}}$O$_{\mathrm{5}}$ target and a WO$_{\mathrm{3}}$ target. Our in-situ temperature-dependent x-ray diffraction measurements reveal a gradual change in the film structure from a monoclinic phase to a tetragonal phase via an intermediate mixture of the two as the concentration of tungsten increases from 0{\%} to 3.7{\%} at 300 K. At 358 K, the film is found to be in a tetragonal phase for the entire composition range we studied. The results also suggest that the volume of the unit cell increases as the concentration of tungsten increases. Electrical transport results further show that both the phase transition temperature and the width of the hysteresis loop decrease with the increasing of the concentration of tungsten. Especially, epitaxial V$_{\mathrm{1-x}}$W$_{\mathrm{x}}$O$_{\mathrm{2}}$ films fabricated on c-cut sapphire substrates show narrower hysteresis loop compared to textured V$_{\mathrm{1-x}}$W$_{\mathrm{x}}$O$_{\mathrm{2}}$ films fabricated on Si (100) substrates. In addition, the Hall effect measurements on the epitaxial V$_{\mathrm{1-x}}$W$_{\mathrm{x}}$O$_{\mathrm{2\thinspace }}$thin films at various temperature points provide important information for the change in the electronic structure upon increasing the concentration of tungsten. [Preview Abstract] |
Monday, March 13, 2017 8:36AM - 8:48AM |
A37b.00004: Structural energetics of VO$_2$ under strain Chanul Kim, Chris Marianetti Attaining a proper first-principles description of structural energetics in VO$_2$ is a necessary condition for fully characterizing the metal-to-insulator transition. Of the existing methods which have been employed for total energies, only DFT+$U$ applied under unorthodox conditions (ie. non-spin-polarized (NSP), with an unreasonably small $U$) has shown promise. In particular, this DFT+$U$ properly captures the qualitative and quantitative energy difference between the monoclinic (M$_1$) and rutile (R) phases. We explain why this unorthodox procedure works by studying a minimal model of the structural energetics based on the Peierls-Hubbard model; where the exact solution may be compared to Hartree-Fock, single-site dynamical mean-field theory (DMFT), and cluster DMFT. Furthermore, we use this DFT+U approach to calculate the strain phase diagram at T=0, including the R, M$_1$, and M$_2$ phases. [Preview Abstract] |
Monday, March 13, 2017 8:48AM - 9:00AM |
A37b.00005: The optical gap in VO$_2$ insulating phases is dominated by Coulomb repulsion Christopher Hendriks, Eric Walter, Henry Krakauer, Tyler Huffman, Mumtaz Qazilbash Under doping, tensile strain or heating, vanadium dioxide (VO$_2$) transforms from an insulating monoclinic (M1) to a metallic rutile (R) phase, progressing through intermediate insulating triclinic (T) and magnetic (M2) phases. Broadband optical spectroscopy data have been obtained \footnote{T. J. Huffman {\em et al.} arXiv:1608.08227.} on the T and M2 phases in the {\em same} sample. While only half the V atoms are dimerized in M2 compared to M1 and T, the measured optical gap is essentially unaltered by the first-order structural phase transition between them. Moreover, the optical interband features in the T and M2 phases are remarkably similar to those previously observed in the well-studied M1 phase. This shows that the electronic structure is insensitive to the lattice structure. Our ab-initio HSE optical conductivity calculations on the insulating phases of VO$_2$ are in excellent agreement with the experimental measurements. We will discuss the choice of $\alpha$, the fraction of exact exchange. As the energy gap is insensitive to the different lattice structures of the three insulating phases, we rule out Peierls effects as the dominant contributor to the opening of the gap. Rather, the energy gap arises from intra-atomic Coulomb correlations. [Preview Abstract] |
Monday, March 13, 2017 9:00AM - 9:12AM |
A37b.00006: Destabilizing Mott insulators with currents: theory and application to Ca$_2$RuO$_4$ Giuliano Chiriaco, Andrew Millis Nakamura, Maeno et al. (Scientific Reports, 2013) showed that under an electric field ($\sim40\,\rm{V/cm}$), the metal insulator transition temperature decreases, so that $\rm{Ca_2RuO_4}$ remains metallic at temperatures well below the equilibrium transition temperature. They further showed that this phenomenon is not due to the Joule heating, and argued that it arose from a nonequilibrium correlation effect. In this work we investigate the issue theoretically, using a one dimensional density wave model and considering Zener tunnelling and in-band and interband relaxation processes. Boltzmann-type transport equations are derived for conduction and valence band, including photon (up to quadratic order) and phonon relaxation terms. Even for modest fields a substantial population imbalance may be induced and consequently destabilize the gap. This work is supported by the NSF DMR 1308236 grant. [Preview Abstract] |
Monday, March 13, 2017 9:12AM - 9:24AM |
A37b.00007: Phase diagram of alkali-doped fullerides: A rotationally-invariant slave-boson perspective Aldo Isidori, Massimo Capone We study the phase diagram of alkali-doped fullerides ($\rm A_3C_{60}$ with $\rm A=K, Rb, Cs$) as a function of the local Coulomb interaction $U$ and the phonon-mediated Jahn-Teller coupling $J$ for various levels of electron filling. In these materials, the Jahn-Teller coupling between electrons and the vibrational modes of the $\rm C_{60}$ molecules effectively reverses the sign of the Hund's coupling, providing a source for a local s-wave pairing mechanism. Using the rotationally-invariant slave-boson formalism we investigate the phase transitions between metallic (superconducting) states and different types of Mott insulating states at either large $U$ or large $J$, revealing a correlation-induced enhancement of superconductivity in proximity of the Mott localization mechanism. [Preview Abstract] |
Monday, March 13, 2017 9:24AM - 9:36AM |
A37b.00008: Orbital-selective Mott transition in Sr$_2$Mn$_3$As$_2$O$_2$ Vaideesh Loganathan, Andriy Nevidomskyy Sr$_2$Mn$_3$As$_2$O$_2$ is a layered material composed of alternating cuprate-like MnO$_2$ layers and MnAs layers similar to iron pnictides [1]. Recent neutron-scattering measurements have revealed a quasi-2D Neel-AF order in the MnO$_2$ layer, along with a G-type AFM order in the MnAs layer. To better understand the experimental findings, we have performed first-principles DFT$+U$ calculations to explore the electronic structure in this material. We find the MnAs layer to be a simple Slater insulator due to the AF ordering. The MnO layer displays more correlated electron behavior that affects the transport properties. We observe a Mott transition in the MnO layer arising from the $d_{x^2-y^2}$ orbitals, reminiscent of cuprate superconductors. To study the layer- and orbital-selective Mott transition, we map the Bloch wave-functions onto Wannier orbitals with $d_{x^2-y^2}$ character. The resulting tight-binding model forms a basis for the Hubbard Hamiltonian, which we investigate using the Variational Cluster Approximation.\\ {[1]} C.-W. Chen {\it et al.}, "Orbital selective Mott transition in layered Sr$_2$Mn$_3$As$_2$O$_2$ single crystals" (under review) [Preview Abstract] |
Monday, March 13, 2017 9:36AM - 9:48AM |
A37b.00009: Resistive Switching in Ordered Insulators: Thermal or Electronic Mechanism? Jiajun Li, Camille Aron, Gabriel Kotliar, Jong Han We investigate the dramatic switch of resistance in ordered correlated insulators, when driven out of equilibrium by a strong voltage bias. Starting from a microscopic description, we present a driven-dissipative mechanism that explains and reproduces the characteristic features of resistive switching (RS), such as the S-shaped I-V curves with hysteresis and the formation of hot conductive filament during the switch. We discuss the tangled relationship between filament growth and negative differential resistance, and also how crystallographic structure and disorder contribute to this mechanism. The distribution function computed at the RS shows that the thermal and electronic mechanisms are compatible: nonequilibrium excitations created by the Landau-Zener tunneling over a self-consistent gap have the first moment fully consistent with the Joule heating scenario. [Preview Abstract] |
(Author Not Attending)
|
A37b.00010: New DMFT capabilities in CASTEP Evgeny Plekhanov, Vincent Sacksteder, Phil Hasnip, Matt Probert, Stewart Clark, Cedric Weber, Keith Refson We present the first implementation of Dynamical Mean-Field Theory in UK's major ab-initio code CASTEP [1]. This implementation: i) is modular; ii) allows great flexibility in choosing local basis set for downfolding/upfolding of self-energy; iii) permits wide choice of impurity solvers (including external solver libraries); and iv) gives the user a possibility to use several self-consistency schemes and calculate total energy and forces. We explain in details the theoretical framework used. We benchmark our implementation on several strongly-correlated insulating systems with d- and f-shells: $\gamma$-Ce and Ce$_2$O$_3$ by using Hubbard I and CTHYB-QMC solvers. Our results appear to be in excellent agreement with the reference data published previously in the literature [2,3,4]. [1] E. Plekhanov, et al. in preparation (2016). [2] L. Pourovskii, et al. Phys. Rev. B 76, 235101 (2007). [3] B. Amadon, et al. Phys. Rev. B 77, 205112 (2008). [4] J. Kuneš, et al. Phys. Rev. Lett. 99, 156404 (2007). [Preview Abstract] |
Monday, March 13, 2017 10:00AM - 10:12AM |
A37b.00011: Analytical results for locally interacting systems: How strong fluctuations of emergent gauge fields affect charge-blocking physics Nobuhiko Taniguchi We present an analytical treatment that captures non-perturbative effect of strong local correlation for the multi-level dot in the environment. Through the DMFT idea, this gives a universal mechanism of how dynamical fluctuations induce the metal-insulator transition without symmetry breaking. Using a Keldysh functional of the quantum $U(1)$-rotor plus fermion model, we re-examine the effect of emergent $U(1)$-gauge fields, whose strong quantum dynamical fluctuations invalidate the standard quadratic approximation around the saddle-point. Our analysis originates from the analysis of how to evaluate exactly the atomic correlation for the multi-level dot in terms of Keldysh functionals, which is nontrivial because of its quartic nature though one can readily achieve it by the operator method. It helps us identify a strong-coupling effective action that describes charge-blocking physics, and learn how to treat correctly large gauge fluctuations, particularly its compactness. We then investigate the effect of connecting the interacting dot with the environment (= the leads) and see gauge fluctuations have the self-energy acutely diverge and reduce near the Fermi level, which brings a new quasiparticle peak at low temperature. [Preview Abstract] |
Monday, March 13, 2017 10:12AM - 10:24AM |
A37b.00012: Nonlocal correlations in the orbital selective Mott phase of a one dimensional multi-orbital Hubbard model Shaozhi Li, Nitin Kaushal, Yan Wang, Elbio Dagotto, Steven Johnston, Yanfei Tang, Gonzalo Alvarez, Alberto Nocera, Thomas Maier In recent years the multi-orbital Hubbard model has been widely studied by using the dynamical mean field theory (DMFT), which neglects spatial fluctuations and nonlocal correlations. However, it is currently not known how important additional nonlocal correlations may be, particularly in one dimension where DMFT is least accurate. In this talk, we present a determinant quantum Monte Carlo and density matrix renormalization group study of the non-local correlations in a three-orbital Hubbard model defined on an extended one dimensional chain. We focus on a parameter regime that hosts an orbital selective Mott phase (OSMP) and an orbitally ordered insulating state. In the OSMP, we show that the momentum dependence of electronic properties is strong for the itinerant electrons and weak for the localized electrons. In addition, although electrons are localized in the orbitally ordered insulating phase, there are short range orbital correlations at finite temperature. In short, these momentum dependent quantities and the orbital order indicate a degree of non-local correlations, which suggests that non-local effects, neglected in single-site DMFT approaches, can be important. [Preview Abstract] |
Monday, March 13, 2017 10:24AM - 10:36AM |
A37b.00013: Two-dimensional quantum percolation on anisotropic lattices Brianna Dillon Thomas, Hisao Nakanishi In a previous work [Eur. Phys.J B {\bf 87}, 286 (2014)], we calculated the transmission coefficient of the 2D quantum percolation (QP) model and found exponentially localized, power-law localized, and delocalized regimes. However, the existence of a delocalized state remains controversial. We note that many works claiming only localization in 2D QP are based on highly anisotropic 2D strips, whereas our work is based on an isotropic geometry. To understand the difference in our results and the anisotropic strip results, we apply our direct calculation of the transmission coefficient to highly anisotropic strips of varying widths at three energies and a range of dilutions. For parameters overlapping those used in other works on highly anisotropic strips, our results are consistent with localization found in those works. However, for low dilutions we find the localization length does not converge as the strip width increases toward the isotropic limit. This indicates a delocalized state for small disorder that lies outside of the parameter space investigated by such previous studies. We additionally calculate the inverse participation ratio of the lattices and find that it too is consistent with a phase transition from delocalized to localized states at the same dilutions. [Preview Abstract] |
Monday, March 13, 2017 10:36AM - 10:48AM |
A37b.00014: Origin of the subpeaks at the Hubbard band edges Seung-Sup B. Lee, Jan von Delft, Andreas Weichselbaum The spectral function of the infinite-dimensional Hubbard model at half filling exhibits a triple-peak structure in metallic phase, consisting of a heavy-fermion quasiparticle peak and lower/upper Hubbard bands. Recent studies using dynamical mean-field theory (DMFT) report that, on top of the triple-peak structure, sharp subpeaks appear at the inner edges of the Hubbard bands, when the system is close to the metal-to-insulator transition [1-6]. However, the origin of the subpeaks has not been clarified yet. Here we propose an effective theory for the subpeaks, and support the theory by DMFT calculations using the numerical renormalization group (NRG) as impurity solver, of one- and two-band models. The dynamics of the particle/hole excitations, expressed as the projections of fermion operators, is the key mechanism developing the subpeaks.\\ \\ {[1]} M.~Karski, C.~Raas, and G.~S.~Uhrig, Phys.~Rev.~B {\bf 72}, 113110 (2005).\\ {[2]} M.~Karski, C.~Raas, and G.~S.~Uhrig, Phys.~Rev.~B {\bf 77}, 075116 (2008).\\ {[3]} R.~Z\v{i}tko and T.~Pruschke, Phys.~Rev.~B {\bf 79}, 085106 (2009). \\ {[4]} Y.~Lu et al., Phys.~Rev.~B {\bf 90}, 085102 (2014).\\ {[5]} F.~A.~Wolf et al., Phys.~Rev.~B {\bf 90}, 115124 (2014).\\ {[6]} M.~Ganahl et al., Phys.~Rev.~B {\bf 92}, 155132 (2015). [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700