B.1.1. Science Driver 1: Electronic and Magnetic Materials

B.1.1. Science Driver 1: Electronic and Magnetic Materials

[M. Jarrell (LSU), J. Perdew (Tulane), J. Ditusa (LSU)]

Program/Project Description. Strongly correlated materials are characterized by strong interactions between the electrons and ions, which yield, e.g., moment formation and the emergence of complex competing phases. They will serve as a new generation of materials for devices and electronics. The goal of SD1 is to develop and test new computational formalisms, algorithms and codes which will eventually enable the search for new correlated materials on the supercomputer.

Methods Developed/Employed. Investigators using the present generation of Density Functional Theory (DFT) and many body tools, including Dynamical Mean Field Theory (DMFT) and its cluster extensions, have made remarkable progress towards understanding these systems. Nevertheless, DMFT methods scale exponentially in the size of system treated explicitly, and DFT fails to describe strong correlation effect. We are developing and testing multiscale methods that circumvent this exponential scaling and more accurate DFT potentials.

Goals and Milestones for year two include the development and porting of GPU enabled multiscale codes, hyperGGA functionals, synthesis and study of metalorganics and ferroics, and the development and use of codes which combine DFT and manybody methods to enable first principles study of correlated materials.

Recruiting, Hiring and Project Coordination. In the first year of this project SD1 successfully filled the majority of the graduate student and post-doctoral positions allocated to this driver. In year two, we recruit ? postdocs and ? graduate students.

Research Accomplishments

Focus 1: Multiscale Methods for Strongly Correlated Materials

The computation and theoretical prediction of materials properties must confront the electron-electron interaction. Correlated many-body methods, including Quantum Monte Carlo (QMC) [43], are computationally inefficient for systems of many electrons. One way to deal with this problem is to perform multiscale calculations, where QMC is only used to treat correlations on the shortest length scales, with approximate methods to treat longer length scales. The other and more common way is to use the Kohn-Sham density functional theory (DFT) [44], an orbital-based approach in which the electron exchange-correlation energy is provided by a functional of the electron density that must be approximated in practical calculations. The remaining problems are then to improve the accuracy of the available approximations, to understand long-range correlations including van der Waals interactions, and to deal with the fact that even the exact Kohn-Sham band structure can underestimate the fundamental energy gap of a solid [45].

Figure 1. The hybridization function Γ is the order parameter for the localization transition. It goes to zero at the critical disorder strength Vc. Above Vc, electrons become localized within the length scale of the cluster
Figure 2. Convergence of the antiferromagnetic leading eigenvalue in the two-dimensional Hubbard model. The results obtained from the multi-scale approach converge faster than that obtained from the two length scale approach.

M. Jarrell and J. Moreno, with C.E. Ekuma, S. Feng,Z. Meng, C. Moore and R. Nelson (LSU), are developing multiscale methods for disordered and interacting systems. They have found that graphics processing units (GPU’s) greatly accelerate materials simulations, including simulations of Ising-model glasses [46,47] and QMC calculations. In collaboration with other researchers in the CTCI team they are developing QMC codes tuned [48,49] for the next generation of GPU supercomputers. To incorporate nonlocal correlations systematically, they have proposed a Cluster Typical Medium Theory (CTMT) that opens a new avenue to the study of Anderson localization in both model and real materials, unlike the coherent phase approximation [50] and its cluster extensions, including the DCA [51]. The idea is to extend the Typical Medium Theory [52], which replaces average quantities with typical values, to its cluster version (Fig. 2). They have also shown that size extrapolations of calculated properties over multiple scales can converge much better if three length scales are invoked: the shortest one for explicit correlation, an intermediate one treated perturbatively, and a longest one treated for the first time via mean field correlations [53](See Fig.2). Finally, they have used density functional theory to generate a band structure for (Ga,Mn)As and (Ga,Mn)N, then applied a Wannier-based downfolding method [54] to get effective interacting Hamiltonians.

The van der Waals interaction is a weak long-range attraction between two objects due to correlations among their fluctuating multipole moments. It is most important when the objects are not otherwise strongly bonded, as for two biological molecules or nanostructures. The accurate calculation of this interaction via many-electron wavefunctions or DFT is feasible only for a pair of atoms or small molecules. Thus standard intermolecular interactions are often based on an atom pair potential picture. Jianmin Tao, John P. Perdew, and Adrienn Ruzsinszky (Tulane) have shown [55-57] how to evaluate this interaction between two quasispherical objects accurately and efficiently, using just the electron densities and static dipole polarizabilities. They have found that the atom pair potential picture is correct at best for the interaction between two solid spheres, but not when one or both objects are spherical shells (e.g., fullerenes). Other work from the Perdew group [58-62] concerns improvements to semilocal and nonlocal DFT approximations.

Most, previous computations utilizing density functional theory (DFT) potentials reported band gaps that are 20-100% smaller than the corresponding, experimental values. Bagayoko, Zhao, and Williams (BZW) have introduced a computational method that circumvents, with first-principle calculations, the above band gap problem. In this reporting period, Bagayoko (Southern Univ.), Ekuma, and collaborators performed ab-initio, self-consistent calculations of properties of ZnO[65], ScN and YN [66], SrTiO3 [67], Ge [68], and InP [69]. These calculations utilized the enhancement of the method based on the work of Ekuma, Franklin (EF) and co-workers. The resulting BZE-EF method led to excellent agreement with experiment of the electronic (including band gaps), optical, transport, structural, and elastic properties of the above materials. Fundamental features of the method include (a) the search and attainment of optimal basis sets that verifiably yield the minima of the occupied energies of the systems under study and (b) the avoidance of over-complete basis set with the utilization of the Rayleigh theorem.

Focus 2: Correlated Organic and Ferroelectric Materials

Figure 3: Current vs. Voltage for Polythiophenes Containing an in-chain Cobaltabisdicarbollide . a) DFT/Green functions simualtions. b) experiment[3]. 1T, 2T, 3T=one, two, and three thiophenes.

Polythiophenes Containing In-Chain Cobalt Carborane centers:Experimental and computational explorations of cobalt carborane complexes that are covalently linked to polythiophenes were performedto study the electrical conductivity with the inclusion of boron clusters. J. Garno (LSU) performed atomic force microscopy (AFM) surface studies and conducting probe measurements of charge transport. Her results indicate that polymers with bithienyl and terthienyl behave like heavily doped semiconductors rather than pure semiconductors, while the current-voltage (I-V) profile for poly- thienyl exhibits no measurable current consistent with the insulating character of the film (submitted). Simulations of these structures by P. Derosa, and N. Ranjitkar (LaTech) employing Gaussian09[1] determined the spin state of these systems. By using a variety of density functionals in their simulations, they concluded that the system is in a spin singlet state. Their calculated conductivity of these structures using Green’s functions[3] compared well to the conducting probe AFM measurements [2] (Fig. 3).

Magnetic and Multiferroic Materials. S.Whittenburg(UNO) has expanded his micromagnetics code to include ferroelectric materials through the use of a Landau-Devonshire potential and correctly predicted the ferroelectric phase transitiontemperatures of BaTiO3. In addition, he can model the elastic properties of materials including stress-induced changes to the morphology. Thus,the electric or magnetic field induced stress, and the resulting shape changes,can be calculated and directly compared to the experimental results. G. Caruntu(UNO)has developed a novel experimental methodology for the local measurement of the strain-mediated magneto-elastic coupling in nanocomposite films. Here he employs an AFM tip to monitor the piezoresponse of a perovskite layer caused by the magnetostriction of a ferrite layer[4-8].L. Malkinski (UNO) has also assisted into the development of new technologies to grow multiwall microtubes of magnetic or magnetic and piezoelectric materials where the magnetic properties have been found to depend on the curvature of the films. Malkinski is also involved in the investigation of thin Fe1-xNix alloy films whose composition varies across its thickness and displayunusual hysteresis curves. In addition, he has explored liquid crystal/ferromagnetic nanoparticles composites where the liquid crystal device properties depend on the applied magnetic field.

The understanding and control of the magnetic properties of nano- and microscopic materials is important for a large range of applications from pharmaceuticals to magnetic storage. A. Burin (Tulane) has used

Figure 4: Optimized structure of an iron cluster of the spinel-type moiety

ORCA quantum chemistry software[9] (BPW91/LanL2DZ level) to model nanoscopic iron oxide clusters finding a high spin (S=12) ground state (Fig. 4)[10,11]. He will extend these calculations to include relativistic corrections[12] to rule out other, nearly degenerate, high spin states. Other calculations include the study of DNA base pair radical cations[13,14] and the modeling of electronic glasses exposed to electric fields[15]. Related experimental work comes from the team of Kucheryavy, Goloverda, and Kolesnichenko (Xavier U.) which has produced ultrasmall superparamagnetic iron oxide nanoparticles in a surfactant-free colloidal form with sizes ranging from 4 to 8 nm. This was accomplished by varying the nucleation and growth conditions, and using a sequential growth technique. Since the T1 relaxivity for magnetite and its oxidized form, estimated by NMR, was found to be similar, they concluded that oxidized magnetite would be preferred as a more stable and potentially less toxic MRI contrast agent.

R. Kurtz and P. Sprunger (LSU) investigated the magnetic properties of FeAl whereDFT calculations[16-23] predict a ferromagnetic ground state at odds with experiment[24-25]. However, DFT+U methods find a paramagnetic state when U, the correlation energy, is sufficiently strong[26]. More interestingly, DFT indicates that the bulk terminated and incommensurate FeAl(110) surfaces may exhibit ferromagnetic ordering of the Fe atoms, with moments enhanced compared to the bulk[23]. In addition to magnetometry showingthat the bulk is paramagnetic, theirsynchrotron X-ray magnetic circular dichroism (XMCD)measurements carried out at CAMDindicates no ferromagnetism at either the commensurate FeAl(110) or the surface reconstructed incommensurate FeAl2 surface[28-29].

J. DiTusa (LSU) is investigating transition metal silicides, germanides and gallium compounds to explore their interesting magnetic and electrical transport properties.These materials are interesting and important because they are relatively simply grown, have crystal structures that lack inversion symmetry, and range from good metals, to magnetic semiconductors and small band gap insulators. They have explored bulk crystals and crystalline nanowires demonstrating the accurate control of Co dopants in FeSi at the 0.5% substitution level, the ability to measure the conductance of 20 nm wide nanowires to temperatures below 300 mK (Fig. 5), and the discovery of interesting behavior in the Hall effect of, Fe3Ga4

Focus 3: Superconducting Materials

Iron-based superconductors and related materials. Since their discovery in 2008, iron-based superconductors have generated intense scientific interest since the complex interplay between magnetism and superconductivity in these materials suggests that the attraction needed to form bound electron pairs is provided by spin fluctuations [30-34]. A systematic investigation of transport, magnetic, and superconducting properties of the phase diagram of the chalcogenide material using resistivity, Hall coefficient, magnetic susceptibility, specific heat, and neutron scattering was reported in 2010 [35]. Leonard Spinu (UNO) is completing this picture by measuring the London penetration depth in single crystals at ultra-low temperatures as a function of temperature and Se concentration (25% to 45%). Results were presented at several conferences [36,37].

The Zhiqiang Mao group (Tulane) has synthesized a new layered iron pnictide CuFeSb [38]. In contrast with other iron pnictides and chalcogenides, this material exhibits a metallic ferromagnetic state with a Curie temperature of 375 K. This finding suggests that a competition between antiferromagnetic and ferromagnetic coupling may exist in iron-based superconductors. It also supports theoretical predictions [39,40] that the nature of the magnetic coupling within the iron plane depends on the height of the anion plane above the iron plane (~1.8 Å for the Sb plane in CuFeSb vs. ~1.4 in FeAs compounds).

In strongly-correlated materials there is typically a close coupling between structure, charge, and spin, leading to a competition among several phases at low temperature. Structural changes at interfaces can drive changes in material functionality. W. Plummer and V.B. Nascimento are measuring surface structure in complex materials via low energy electron diffraction (LEED) [41]. But the analysis of LEED data is an inverse problem: one must search for the surface structure that yields a given diffraction pattern. They have developed novel LEED codes that use global search algorithms [42] and can also tackle the structural determination of multiple terminated crystallographic surfaces. The codes have been tested successfully for BaTiOultra-thin films [42], and will next be applied to multi-phase (001) surfaces of the BaFeAsand Ba(FeCo)Asiron pnictide superconducting materials.

Changes in Research Directions, Future Work

There are no major changes in research directions.

References:

1. M. J. T. Frisch, G. W. Trucks, H. B. Schlegel,G. E. Scuseria,M. A. Robb, J. R. Cheeseman,G. Scalmani, V. Barone,B. Mennucci,G. A. Petersson, H. Nakatsuji,M. Caricato,X. Li,H. P. Hratchian, A. F. Izmaylov, J. Bloino,G. Zheng,J. L. Sonnenberg, M. Hada,M. Ehara,K. Toyota,R. Fukuda, J. Hasegawa,M. Ishida,T. Nakajima,Y. Honda,O. Kitao,H. Nakai,T. Vreven,J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark,J. J. Heyd, E. Brothers, K. N. Kudin,V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,S. S. Iyengar,J. Tomasi,M. Cossi,N. Rega,N. J. Millam, M. Klene,J. E. Knox, J. B. Cross,V. Bakken,C. Adamo,J. Jaramillo,R. Gomperts,R. E. Stratmann, O. Yazyev,A. J. Austin,R. Cammi,C. Pomelli,J. W. Ochterski, R. L. Martin, K. Morokuma,V. G. Zakrzewski, G. A. Voth,P. Salvador,J. J. Dannenberg, S. Dapprich,A. D. Daniels, Ö. Farkas,J. B. Foresman,J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian 09, Wallingford, CT. (2009).
2. B. Fabre, E. H. Hao, Z. M. LeJeune, E. K. Amuhaya, F. Barriere, J. C. Garno, and M. G. H. Vincente, Polythiophenes Containing In-Chain Cobaltabisdicarbollide Centers. ACS Applied Materials and Interfaces2, 691–702 (2010).
3. P. A. Derosa, and J.M. Seminario, Electron transport through single molecules: Scattering treatment using density functional and green function theories. J. Phys. Chem. B 105, 471-481 (2001).
4. G. Caruntu, A. Yourdkhani,M. Vopsaroiu, G. Srinivasan, Probing the local strain-mediated magnetoelectric coupling in multiferroic nanocomposites by magnetic field-assisted piezoresponse force microscopy,Nanoscale 4, 3218-3227 (2012).
5. A. Yourdkhani,A. K. Perez, C. K. Lin, G. Caruntu, Magnetoelectric Perovskite-Spinel Bilayered Nanocomposites Synthesized by Liquid-Phase Deposition, Chem. Mater.22, 6075-6084 (2010).
6. D. Damjanovic, Ferroelectric, dielectric and piezoelectric properties of ferroelectric thin films and ceramics,Rep. Prog. Phys.61, 1267-1324 (1998).
7. S. V. Kalinin, D. A. J. Bonnell,Contrast mechanism maps for piezoresponse force microscopy,Mater. Res.17, 936-939 (2002).
8. C. Harnagea,A. Pignolet, M. Alexe,D. Hesse,Possibilities and limitations of voltage-modulated scanning force microscopy: Resonances in contact mode, Integrated Ferroelectrics 60, 101-110 (2004).
9. F. Neese, ORCA – an ab initio, Density Functional and Semiempirical program package, Version 2.4. Max-Planck-Insitut für Bioanorganische Chemie, Mülheim and der Ruhr, (2004).
10. M. Castro, D. R. Salahub, Density-functional calculations for small iron clusters: Fen, Fen+, and Fen- for n≤5, Phys. Rev. B 49, 11842 (1994).
11. G. L. Gutsev, C.W. Bauschlicher Jr., Electron Affinities, Ionization Energies, and Fragmentation Energies of Fen Clusters (n = 2−6): A Density Functional Theory Study, J. Phys. Chem. A 107, 7013-7032 (2003).
12. C. J. Milios, A. Vinslava, W. Wernsdorfer, S. Moggach, S. Parsons, S. P. Perlepes, G. Christou, E. K. Brechin, A Record Anisotropy Barrier for a Single-Molecule Magnet, Journal of the American Chemical Society 129, 2754-2755 (2007).
13. A. L. Burin, A. K. Kurnosov, Fluctuator model of memory dip in hopping insulators, J. Low Temp. Phys., 167, 318-328 (2012),
14. A. L. Burin, A. K. Kurnosov, Fluctuator model of memory dip in hopping insulators, European Superconductivity News Forum, CR25 (
15. Sarah L. Tesar, John M. Leveritt III, Arkady A. Kurnosov, Alexander L. Burin, Temperature dependence for the rate of hole transfer in DNA: Nonadiabatic regime, Chemical Physics 393, 13–18 (2012).
16. V.L. Moruzzi, P.M. Marcus, Magnetic-Structure of Ordered FeAl and FeV, Phys. Rev. B, 47, 7878-7884(1993).
17. V. Sundararajan, B.R. Sahu, D.G. Kanhere, P.V. Panat, G.P. Das, Cohesive, electronic and magnetic-properties of the transition-metal aluminides FeAl, CoAl, and NiAl, J. Phys.-Condes. Matter, 7 6019-6034(1995).
18. J. Zou, C.L. Fu, Structural, electronic, and magnetic-properties of 3d transition-metal aluminides with equiatomic composition, Phys. Rev. B, 51 2115-2121 (1995).
19. D.A. Papaconstantopoulos, K.B. Hathaway, Predictions of magnetism in transition metal aluminides, J. Appl. Phys., 87, 5872-5874(2000).
20. G.P. Das, B.K. Rao, P. Jena, S.C. Deevi, Electronic structure of substoichiometric Fe-Al intermetallics, Phys. Rev. B, 66184203 (2002).
21. P.G. Gonzales-Ormeno, H.M. Petrilli, C.G. Schon, Ab-initio calculations of the formation energies of BCC-based superlattices in the Fe-Al system, Calphad 26, 573-582(2002).
22. S.H. Yang, M.J. Mehl, D.A. Papaconstantopoulos, M.B. Scott, Application of a tight-binding total-energy method for FeAl,J. Phys.-Condes. Matter, 14, 1895-1902(2002).
23. J.L. Pan, J. Ni, B. Yang, Stability of FeAl(110) alloy surface structures: a first-principles study, European Physical Journal B, 73, 367-373(2010).
24. P. Shukla, M. Wortis, Spin-glass behavior in iron-aluminum alloys – microscopic model, Phys. Rev. B, 21 159-164(1980).
25. J. Bogner, W. Steiner, M. Reissner, P. Mohn, P. Blaha, K. Schwarz, R. Krachler, H. Ipser, B. Sepiol, Magnetic order and defect structure of FexAl1-x alloys around x=0.5: An experimental and theoretical study, Phys. Rev. B, 58 14922-14933(1998).
26. P. Mohn, C. Persson, P. Blaha, K. Schwarz, P. Novak, H. Eschrig, Correlation induced paramagnetic ground state in FeAl, Phys. Rev. Lett., 87 196401(2001).
27. L. Hammer, H. Graupner, V. Blum, K. Heinz, G.W. Ownby, D.M. Zehner,Segregation phenomena on surfaces of the ordered bimetallic alloy FeAl, Surf.