Abstract
Two techniques for removing the background in inelastic electron tunneling spectroscopy (IETS) have been examined in detail. The first of these uses a polynomial least‐squares routine to fit a smooth curve to the data on either side of the IETS peak, while the second relies on differential inelastic electron tunneling spectroscopy (DIETS) to subtract one spectrum from another in order to remove some unwanted IETS peaks. Both procedures were applied to spectra obtained from Al‐Pb junctions with the lead normal and superconducting. One finds that it is easier to determine the background when the normal rather than superconducting lead is used, as the superconducting line shape has a broad undershoot. The intensity results shown here indicate that failure to consider the undershoot may result in errors as large as 30%. A comparison of theoretical and experimental line shapes is made for data with both normal and superconducting electrodes.