Institute Theory I
General Overview
Introduction:
The research activities of the Institute encompass several key areas of condensed matter theory: (i) electronic and structural properties of complex systems ranging from large organic molecules and mesoscopic contacts to complex solids; (ii) electronic excitations and dynamical properties of molecular clusters, solids and solid surfaces, as well as the quasiparticle behavior of transition metals and oxides resulting from electronic correlations; (iii) nano-scale tribology, friction, plastic deformation, adhesion and brittle fracture.
The principal goal of these studies is to achieve a microscopic understanding of complex phenomena on an electronic and atomic scale. Although most of these topics involve fundamental issues associated with many-body interactions, they are nevertheless relevant to a wide host of practical applications in modern materials science. In fact, the majority of research topics is directly motivated by the aim of providing a quantitative or qualitative physical basis for possible future technologies.
In line with the range of topics investigated a wide variety of conceptual and computational methods is currently employed: density functional theory, molecular dynamics calculations, genetic algorithms, data mining techniques, time-dependent generalization of density functional theory, diagrammatic perturbation approaches, classical and Quantum Monte Carlo methods, exact diagonalization schemes, as well as analytical methods. Needless to say that the state-of-the-art use of these methods benefits greatly from the excellent computational facilities available at the research center. The unique combination of basic physics questions and possible practical applications, investigated with these theoretical tools, is the hallmark of modern condensed matter and computational materials science research.
Research Topics:
1. Ballistic electron transport through mesoscopic contacts:
Narrow contacts between electron reservoirs are the heart of modern storage devices. Modern technology allows one to make nanoscale connections which the electrons can pass phase-coherently. A method has been developed to determine the conductivity of such quantum-contacts from a Schroedinger equation, and was applied to devices studied experimentally. As a next step, this method will be extended to cover the case of connections via single molecules. (A. Bringer; in collaboration with J. Appenzeller)
2. Femtosecond spectroscopy of molecular clusters:
Via time-delayed short laser pulses it is possible to determine lifetimes of excited quantum states in the femtosecond range and to identify the decay mechanisms. This "pump-probe" technique is applied in the experimental Institute for Electronic Properties (IFF-IEE) to study the dynamics of molecular clusters. Previous theoretical interpretations of the spectra are still largely phenomenological and efforts to improve existing models for relaxation processes in quantum systems are necessary. In connection to recent experiments on Au2C0 clusters in the IEE the process of photon excitation and subsequent electronic and vibrational relaxation will be simulated by a suitable time dependent Schroedinger equation. (A. Bringer; in collaboration with P. Bechthold, M. Neeb)
3. Mathematical modelling using d genetic algorithm:
Many quantities of considerable interest (such as the electrical conductivity of a real material) cannot be predicted by first principles theory. However; by mining existing data for hidden correlations (using e.g. the genetic algorithm) it may be possible to generate mathematical models for such quantities that use readily computed descriptors and which are able to predict these quantities with a certain degree of reliability. (J. Harris; A. Bringer)
4. Computation of thermal and mechanical properties using density functional methods:
A body of recent work has shown that the DFT method with well-established exchangecorrelation energy functionals describes the small changes necessary to compute quantities such as the stiffness matrix (elastic moduli) of even quite complicated crystals. Furthermore, application of full Grueneisen theory gives a good description of thermo-mechanical properties like the coefficient of thermal expansion. These are currently supercomputer applications, however, requiring a very careful treatment of phonon spectra. A lower level of theory (Debye-Grueneisen) is available but so far has been tested only for some elemental metals. The current status of the programme is to establish procedures for implementing Debye-Grueneisen theory in general and testing its area of applicability. (J. Harris; in collaboration with J. Rodgers, Y. LePage, P. Schmidt)
5. Energy surfaces, structures and reactions of polymers:
The electronic and structural properties of organic molecules are evaluated using density functional theory and molecular dynamics methods with the aim of understanding in detail the mechanism of reactions between polymer chains, and the catalysts and additives used in their production. The density functional programs run on the T3E supercomputers and are very demanding of computer resources for systems of the complexity studied here. An example is provided by the reaction between the nucleophilic molecule lithium phenoxide (LiOPh) and a segment of a polycarbonate chain, which involves calculations for a total of 145 atoms. The resulting energy surfaces, however, provide a consistent and extendible data base for developing simplified models that allow calculations on much larger systems. Of particular interest are: (i) the refinement of classical force fields for use in molecular dynamics and Monte Carlo simulations, (ii) the development of a model of `living polymers' that allows one to study the polymerization process in polycarbonate over a wide range of temperatures and densities. This has lead to the identification of the nature of the polymerization transition. (R.O. Jones; in collaboration with J. Akola, P. Ballone, Bayer AG)
6. Dynamical Response at metal surfaces:
Quantum well behavior in adsorbed thin films is usually observed for electrons confined by the band gap of the underlying substrate. Recent photoemission spectra for Na films on Al(111), however, reveal such behavior also for the resonances induced by the large negative potential step between adlayer and substrate. Using time-dependent density functional theory it was shown that this observation is made feasible via the local screened photonfield which is enhanced and confined to the overlayer if the photon frequency is tuned to the Na collective modes. (A. Liebsch; in collaboration with S. Barman, K. Horn)
7. Magneto-optical Kerr effect for non-equilibrium electron distributions:
There is currently great interest in understanding the dynamical response of magnetic surfaces to excitations via ultrafast lasers (e.g., recent experiments in this area in the institute IFF-IEE). In order to investigate to what extent the Kerr signal provides information on the timevarying magnetization of the sample, the optical conductivity tensor is evaluated for non-equilibrium electron distributions generated by the initial pump laser. Of particular interest is the so-called bleaching effect (closing of certain excitation channels) at high laser intensities. (A. Liebsch; in collaboration with P. Oppeneer)
8. Trajectories and their discrete counterparts:
It is generally believed that trajectories can well be represented by their discrete counterparts, the so called Poincare maps (PM) which quite often are much easier to handle (e.g. numerically). It is shown, that in repelling systems (transient chaos) the PM gives misleading results. Even the sign of drifts can be erroneous. Modifying the PM by including the return time (time between two intersections) leads to correct results. At present correlation funtions and diffusion are computed and compared by using PM and the modified PM. (H. Lustfeld; in collaboration with Z. Kaufmann)
9. Characterization of tracer gas concentrations in the atmosphere:
In the atmosphere small scale fluctuations of tracer gas, in particular pollutant, concentrations occur signalizing that the concentrations are singular functions. It is shown that these singularities have to be characterized not by just one Holder exponent but by an ensemble classifying the strength of these singularities. The situation is quite analogous to that of trajectories: These have to be classified by an ensemble of Lyapunov exponents as well. Just one exponent is not sufficient. (H. Lustfeld; in collaboration with Z. Neufeld)
10. Friction, nano-scale tribology:
A research activity related to many practical applications concerns the area of tribology, in particular, friction and related topics such as plastic deformation, adhesion and brittle fracture. The influence of surface roughness on adhesion between elastic bodies is also investigated. Furthermore, Molecular Dynamics calculations are performed to understand the transition between boundary lubrication and hydrodynamic lubrication. An important result was obtained within a recently developed theory of contact mechanics between randomly rough surfaces, where the solids are assumed to deform elastically when the stress is below the yield stress, and plastically when the stress reaches the yield stress. Of key interest is the dependence of the (apparent) area of contact on the magnification. In most cases the area of real contact is proportional to the load. If the rough surface, however, is selfaffine fractal (Hurst exponent H) up to the lateral size L of the nominal contact area, and assuming no plastic deformation, then the real contact area is proportional to LH. (B.N.J. Persson; in collaboration with A. Volokitin, V. Samoilov, S. Zilbermann).
11. Rubber Friction
When rubber slides on a hard, rough substrate, the surface asperities of the substrate exert oscillating forces on the rubber surface leading to energy "dissipation" via the internal friction of the rubber. A recently developed theory shows how the resulting friction force depends on the nature of the substrate surface roughness and on the sliding velocity both for stationary and non-stationary sliding. Numerical results were obtained for the case when the substrate surface has a self affine fractal structure and are in good agreement with experimental observations. The theory is now used by tire companies in developing new rubber compounds for tires. At present the theory is extended to take into account the flash temperature in the rubber-substrate contact areas. (B.N.J. Persson)
12. Dynamical correlations in the electron gas:
To understand dynamical correlations in the electron gas beyond the RPA, lowest order corrections in the dynamically screened Coulomb interaction are evaluated numerically. The corresponding irreducible polarization diagrams have been known for 40 years but have resisted computation because. of their complexity. By analytical means the resulting 7-dimensional integrals were now reduced to 3-dimensional integrals. For (k, w) outside the particle-hole continuum, dynamical correlations consist of two-particle-hole-pair, two-plasmon and coupled particle-hole-pairplasmon excitations. The results in this range compare well with experiments. Presently the calculations are extended to (k, w) inside the particle-hole continuum with the aim of analyzing additional available data. ( K. Sturm; in collaboration with A. Gussarov, H. Lustfeld)
External Funding:
1. R.O. Jones receives support within the framework of the Kompetenzzentrum "Werkstoffmodellierung: Wege zum computergestiitzen Materialsdesign", 01.01.2000 31.12.2004, funded by the BMBF (50 %), Bayer AG, FZJ. Total grant: 1.929.240 DM, BMBF: 997.240 DM (Personell: 780.000 DM, travel: 52.000 DM, equipment 100.000 DM).
2. A. Liebsch receives support within the European Community Network on Ultrafast Magnetization Dynamics (total grant 1.500.000 Euro).
3. Between 1994 and 2002 Bo Persson obtained research grants from the DFG (about 100.000 DM) for several 3-months visits of Prof. A. Volokitin (Samara, Russia).
4. As a participant in a 5-year BMBF supported German-Israeli Cooperative Project on "Novel Tribological Strategies: from the Nano-to Meso-Scales" Bo Persson receives annual funding of 50 000 DM.
5. Bo Persson also receives a European Community grant (about 350.000 DM) within the network: SMART QUASICRYSTALS, 2001 (total grant 20.000.000 Euro).
Ansgar Liebsch
Acting Head since 01.08.2001