Physica 135B (1985) 77-80 North-Holland, Amsterdam PROXIMITY-EFFECT STUDIES OF Nb-BASED BILAYERS WITH s-p, RARE-EARTH AND HEAVY-FERMION METALS L.H. GREENE, W.L. FELDMANN and J.M. ROWELL Bell Communications Research, 600 Mountain Ave., Murray Hill, NJ 07974, USA Utilizing a thickness-spread technique, proximity-effect studies of Nb juxtaposed with rare-earth alloys and a heavy- fermion metal (CeCu6) have been undertaken. The 20 to 1000A thick Nb films exhibit a bulk T~down to a 130/~ thickness and the mean free path at low temperature is limited by interface scattering. Preliminarydata and experimental directions are reported. 1. Background and introduction In our previous research on metallic multilayers consisting of Nb layered with magnetic (Er, Tm) and non-magnetic (Lu) rare-earth (RE) metals, we compared the magnetic and non-magnetic proximity effect in a superlattice geometry [1]. These results are reproduced in fig. 1, where the superconducting transition temperature, Tc, is plotted as a function of modulation wavelength, A for Nb/Er and Nb/Lu superlattices sputter-de- posited onto room-temperature sapphire. Note that T c is depressed with decreasing A, the effect being more dramatic in the magnetic case. There are two difficulties in interpreting these data. First, for each given periodicity, the actual Nb thickness is somewhat uncertain due to the growth technique, as discussed previously [1, 2], in that one end of an array of multilayers is grown Nb-rich. Since each data point in fig. 1 is taken from a different film, the Nb thickness uncertain- ty creates scatter in the data. Second, when growing a superlattice, the single deposition temperature, TD, is not necessarily optimum for each separate component. These difficulties can be alleviated by turning from a superlattice to a bilayer geometry. Once this geometry is adopted, spreads in either thick- ness or composition in one of the layers can be advantageous. A single sample contacted with parallel cross strips can generate a wide range of T c vs. thickness or composition measurements. Also, each layer can be grown at its optimum deposition temperature, provided that the time between depositions is short and the vacuum system is sufficiently clean so that an impurity layer does not form between the two components. We have recently begun bilayer studies, utiliz- ing a film-thickness spread of Nb. After the Nb is sputter-deposited onto single-crystal sapphire a thick overlayer of another material is grown. In order for these studies to be successful, several materials parameters must be brought under con- trol. First, the Nb film must be of high quality. It must be clean, high (near bulk) T c material, and no chemical interactions with the substrate can be tolerated. Second, any overlayer chosen which might be of interest in Nb-proximity studies must not chemically interact or interdiffuse with the Nb. Finally, and most important to these studies, the interface between the Nb and overlayer must be sharp and clean. This final constraint is most easily met when the deposition temperature of the two materials are identical so that the time between depositions is short (<1 s). The interface cleanliness is then comparable to that for super- lattices. Since this work is in its initial stages, we present here our expectations and preliminary data. Consider a single Nb film grown with thickness varying from less than one-tenth to about twice the low temperature coherence length of Nb (~500~,). If perfect, this wedge-shaped film would exhibit a bulk T c along its entire length, as 0378-4363 / 85 / $03.30 t~) Elsevier Science Publishers B.V. (North-Holland Physics Publishing Division)