DFG Research Group: Project 5

Ultrasonic Investigations of Electronic and Structural Instabilities

Prof. Dr. Volker Müller, Dr. Dietrich Maurer

  1. Ultrasonic Investigations of Electron-Lattice Interaction

    For many of the earlier transition metal compounds like VO2 and V2O3, which undergo a temperature-dependent metal-insulator (MI) transition, the role of electron-lattice interaction is widely unknown. More precisely, it is still an open question whether the transition is mainly driven either by electron-electron correlations or by lattice instabilities. For a deeper insight into the underlying physics, a knowledge of the dynamical properties of the lattice, at least in the vicinity of the transition temperature, would be quite informative. In this context ultrasonic investigations of single crystals are particularly useful, since they allow for studies of the elastic behaviour and its dependence on lattice symmetry in detail. Measurements by means of ultrasound are usually performed on single crystals of several mm in length. However, larger single crystals show a notorious tendency of crack formation accompanying the MI-transition in VO2 as well as in V2O3. That is the reason, why investigations right through the MI-transition by means of conventional ultrasonic methods are doomed to fail. In contrast to single crystals, epitaxial films of these compounds are very resistant against ageing effects. We therefore decided to perform sound velocity measurements of acoustic bulk waves through such films. Although there are numerous difficulties which have to be overcome by experiment, some measurements on epitaxial V2O3-films were already successfully performed. To our knowledge, this is the first time that the sound velocity of this compound has been determined right through its MI-transition. Quite astonishingly, the obtained sound velocity data reveal a considerable softening of the V2O3-lattice within the insulating state, i.e. well below the transition temperature. Of course, the accuracy of the measurement is not as high as is commonly achieved by use of single crystals. Nevertheless, in our case the resolution is sufficiently high to point out clearly that the elastic modulus of the antiferromagnetic insulating phase is about 5% lower on average than the corresponding modulus of the paramagnetic metallic phase. As preliminary conclusion it should be emphasized, that the overall temperature dependence is indicative more of structural changes than of electron-electron correlations as dominant driving force of the MI-transition in V2O3.

  2. Ultrasonic Microscopy

    When cooling through the MI-transition, single crystalline VO2 and V2O3 develop extended polydomain structures. We have studied the formation of domains in these compounds in detail by means of optical microscopy, mostly done under polarized light, and ultrasonic microscopy, in particular. Optical studies of these samples, however, suffer from the quite small penetration depth, which therefore restricts investigations of the samples to near-surface regions. In contrast, images from the interior of opaque material can be obtained by ultrasonic microscopy. Structural changes accompanying the MI-transition especially in the bulk were investigated this way. Most interestingly, the transformation kinetics observed in the bulk differs markedly from that found in the near-surface regions by means of optical methods. As a further advantage of ultrasonic microscopy, also surface sound can be excited locally by use of an acoustic lens. This, in particular, allows for sound velocity measurements of surface acoustic waves on samples of only some tenth of æm in length. It is this possibility, which became quite important in view of the fact, that larger single crystals of the MI-compounds show strong ageing effects after thermal cycling through the transition temperature. Only crystallites sufficiently small are able to perform the MI-transition in a reversible manner. The respective upper critical size, however, approaches 50 - 100 micrometer in length for VO2 but only a few micrometer in the case of V2O3. Consequently, sound velocity measurements of surface waves microscopically excited could be solely done on VO2 at the moment. Investigation of this compound reveals a strong elastic anisotropy of the metallic phase, which, however, disappears almost completely, when the insulating state is developed. Below the MI-transition of VO2, an unusually strong stiffening of the lattice takes place on further cooling, which is not yet understood.