Exciting research opportunities lie in the interfaces between disciplines. Even within a single field of research, the combination of complementary techniques brings new insights into complex problems. that cannot be obtained by either of the methods alone. There is a long tradition of constant interplay between experiment and theory in condensed matter physics and materials science research. In this interdisciplinary area, Rrecent advances in experimental techniques, promise to bring the interplay of experimental and theoretical research to new levels of quantitative detail.

On the experimental front, recent advances in instrumentation and techniques are taking place using advanced X-ray sources such as the ESRF. In the field of computational material theory, novel algorithms and faster computers are opening up uncharted territories in materials research. The combination of these two approaches is illustrated in the following articles, highlighted in this note below.

The first paper, by Weissker et al. [1], studies the frequency and momentum dependence of silicon’s dielectric function in a broad region of frequency and momentum transfer. Silicon, a textbook material, is an excellent model system for testing the accuracy of different theoretical approaches. The measurements use high-resolution, non-resonant, inelastic X-ray scattering. The theoretical methods used include non-self-consistent many-body perturbation theory in the GW approximation as well as the time dependent density functional theory in the LDA (local density approximation). The latter approach turns out to be remarkably successful in accounting for all the main features of the data, provided that an energy dependent lifetime broadening is added, as suggested in ref. [2]. The calculations relate the different features of the experimental data to the quasiparticle bands and the plasmon mode of the material. Theory is required to make this connection since off diagonal elements of the microscopic dielectric function (which are generally known as local field effects) give rise to non trivial features in the frequency and wavevector dependence of the macroscopic dielectric constant.

The second paper, by Slezak et al. [3], determined the in-plane phonon density of states of monolayers at and near a clean Fe (100) surface in ultra-high high-vacuum conditions. Phonon spectroscopy is made This break-through became possible by the combination of nuclear inelastic scattering, availability at third-generation synchrotron radiation sources, and detectors advaavailable at the ESRF. Grazing-incidence nuclear scattering allows the determination of the surface phonon density of states.nced surface science techniques. Modern implementations of density functional theory, together with the direct method [4], allow the evaluation of the phonon spectra in materials with large unit cells. The studies of ref. [3] use a thick slab containing 29 iron layers.

The combined experimental/theoretical study establishes that, while the second layer of the surface vibrates at almost the same frequencies as the bulk, the first layer is markedly different. The surface vibrational amplitudes are clearly enhanced relative to the bulk, while the surface phonon frequencies are substantially softened (of the order of 30%) relative to the bulk values. The power of the experimental setup to select the contribution from the first layer and of different phonon polarisation states is bolstered by the theoretical capabilities allowing the identification of each phonon branch and its weight in the phonon density of states.

The third paper, by Raymond et al. [5], studies PuCoGa5, a remarkable material which undergoes a direct transition from a very incoherent metallic state, characterised by Curie-like susceptibility, to an unconventional superconducting state at low temperatures. To determine the degree of correlation in this material, i.e. to which extent the f electrons in Pu are localised, Raymond et al. present a combined experimental and theoretical study of the phonon spectra.

The effects of correlation on the phonon spectra are modelled by varying the Hubbard parameter U of the LDA+U method. The experimental determination of phonon properties in small crystals containing actinide elements is made possible by inelastic X-ray scattering. Comparing theory and experiment, a value of U=3 is was determined for 242Pu. This type of calculation had only been previously carried out on much simpler materials such as NiO and MnO [6]. The results of Raymond et al. support the idea that strong correlations are indeed present in such Pu materials. The combined experimental and theoretical study of Raymond et al. supports the idea that strong correlations are indeed present in 242Pu materials.

In summary, advances in computational physics and in experimental instrumentation are enabling a much closer interaction between theory and experiment. This interplay of theory and experiment holds promise for reaching a deeper and more quantitative understanding of the physical properties of complex materials.

G. Kotliar, Center for Materials Theory, Rutgers University (USA)



[1] H. Weissker et al. PRL 97, 237602 (2006).
[2] S. Rahman and G. Vignale, PRB 30, 6951 (1984).
[3] T. Slezak et al., PRL 99, 66103 (2007).
[4] K. Parlinski et al., PRL 78, 4063 (1997).
[5] S. Raymond et al., PRL, 96, 237003 (2006).
[6] S. Savrasov and G. Kotliar, PRL 90, 056401 (2003).