The frequency-dependent conductivity of laser-deposited YBa2Cu3C>7-s thin films shows an onset of midinfrared absorption at -140 cm -1 and structure in the 400-500-cm ~! region. These low-energy absorptions occur both above and below 7V, making them unlikely to be the superconducting gap in the usual BCS sense. The absorption across the gap is weak because the high-r c materials are in the clean limit; this weak absorption is masked by the midinfrared absorption.

A metallurgically stable and laterally uniform contact to n-GaAs with an enhanced barrier height (0.99 V) and an ideality factor of 1.10 has been achieved with a NiAl bimetallic metallization. This barrier height, as measured by the forward current-voltage technique after annealing for 20 s at 650 °C, is higher than the reported barrier heights of refractory metallizations to n-GaAs. Auger electron spectroscopy (AES) sputter profiles reveal an Al-Ga exchange reaction after high-temperature (500–950 °C) rapid thermal annealing. From these results, the barrier height enhancement is attributed to the formation of an Al1−xGaxAs layer at the NiAl/n-GaAs interface. The thermal stability and low electrical resistivity of the NiAl phase, the enhanced barrier height on n-GaAs, and the ease of patterning the as-deposited Ni/Al/Ni layered structure by lift-off techniques make NiAl a very promising gate contact material for GaAs metal-semiconductor field-effect transistors and related devices.

Superconducting ceramics are produced in crystalline form by, for example, the crystal pulling technique. The seed crystal is prepared from another ceramic material which is non-superconducting but has a similar molecu lar structure as the superconducting ceramic to be formed. By virtue of the similarity in the crystalline structure, single crystals of superconducting ceramics can be easily produced on the seed crystal.

Our paper was unclear about the fits shown in Fig. 1. Above T c we used a Drude part, a sum of three Lorentzian oscillators, and 6oo; below T c we used a zero-width (. The fits were carried out at each temperature. The caption to Fig. 1 gave the Drude results at each temperature but only a rough average of the Lorentzian parameters. The actual Lorentzian (o el , (o p u and // (in cm " *) which were used for the upper panel are as follows: at 300 The parameters interact strongly, so that any could be changed by about 10% and the others adjusted to give fits as nearly as good as shown in the paper. What is outside this generalization is that with decreased temperature L \ appears to shift down and L 2 to shift up; both narrow somewhat. This can be seen in the data of Figs. 2 and 3.Only the Drude parameters were used in further analysis in the paper, so that this erratum does not affect the major conclusion of the paper.(a)