The Possibility of Flux Flow Spectroscopy.

In this Letter we examine the feasibility of performing spectroscopic measurements of the excitations reflected as a resonance in the dynamic susceptibility xs $ q, vd using flux flow devices. These devices can be fabricated from high temperature superconductors and used to probe the properties of another material (the specimen) placed in close proximity. A broad range of wave vector, $ q, and frequency, v, of the modes representing the specimen are potentially detectable with this technique. In fact, it should prove possible to examine a range of $ q amounting to a substantial fraction of the Brillouin zone (BZ). The proposed technique is complimentary to neutron diffraction but with the remarkable advantage that extremely small specimens (mass , 10214 g) may be investigated, making accessible submicroscopic systems. The idea of using the ac Josephson effect to perform in situ spectroscopy, wherein the specimen under study is the junction barrier material itself, was proposed some time ago [1]. Electron-spin resonance (ESR) spectroscopy was envisioned and was later successfully demonstrated using conventional superconducting tunnel junctions [2,3]. Subsequently, Goldman et al. [4] and Barnes and Mehran [5] developed the theory further to include measurements of xsq, vd in concentrated magnets. The present proposal is simpler in that the “device” is nothing more than a thin-film superconductor, and the sample need not form an integral part of the device—it need only be placed in close proximity. The use of high-Tc materials for device fabrication has the obvious advantage that an expanded temperature range for studies is made available. Furthermore, most of the characteristic quantities are larger. The theoretical maximum Josephson frequency (nmax) is limited by the zerotemperature superconducting energy gap to be 2Ds0dyh; for the cuprates [Ds0d , 30 meV] this implies nmax , 15 THz. The maximum theoretical wave vector is, for simple flux flow, ,1yj, and for a weak link geometry, ,lLyj, where j and lL are the superconducting coherence length and London penetration depth, respectively. Each experimental element involved in the presently proposed experiment has already been demonstrated, although never together. The self-detection of electromag-