Superconducting Properties of Crystalline Films of Aluminum on Silicon

Abstract

For some years the properties of disordered and ultrathin films have been the subject of active investigation (see, e.g., Refs. 1). Until the work reported here there have been, as far as we know, no investigations of superconductivity in well-ordered and well-characterized ultrathin films. Perhaps the most significant feature of the present results is that the transition temperature of the Al films is not very much higher than the bulk value, even down to thicknesses near 30 A. Films evaporated with more gas have higher T;'s and films oxidized by exposure to air also have increased T;'s. Another interesting feature of these results is the effect of the initial stages of nucleation on the film properties. Films grown epitaxially have lower R 0 than films that are less crystalline. However, although the higher-resistance films have broader transitions, the T;'s are the same. Theseexperiments will be described more fully later. The method for preparation of epitaxial Al films on silicon has been previously described by Lander and Morrison 2 and we have followed their procedure with good success. Briefly, the silicon is heated in an ultrahigh-vacuum low-energy electron diffraction system in the 10-1 0-Torr range. V nder these conditions the surface reorients and forms the characteristic structure shown by the LEED photograph in Fig. 1(top). Next, one tries to deposit about one-third of a monolayer of Al to form the intermediate structure shown in Fig. 1 (bottom), and finally more Al is deposited and grown epitaxially on this structure. A LEED pattem from a good Al film is shown in Fig. 2. In Fig. 3 (bottom) we show the transition of a film grown by the Lander and Morrison prescription which is about 30 A thick. In Fig. 3 (top) we show the transition of a film prepared at exactly the same time and conditions, but on a part of the silicon which did not have the proper intermediate structure. The R 0 is ten times higher, but the Tc is comparable to the film with the smaller Ro. In general, films were prepared at pressures of about 1 x 10-9 Torr, and evaporation times were about 30 sec or less. The films were covered with about 50-100 A of Ge before exposure to air. Films with an order-of-magnitude Ionger evaporation time, or made at higher pressures, appear to have T", near 1.6 or 1.7°K. The lowest r; we attained, forafilm about 130-A thick, was 1.26°K, which is essentially explained by the differential contraction between Al and silicon.