Search

We have found experimentally that high-purity Ge low-energy X-ray detectors have a relatively thick entrance window which renders them practically useless below ~ 2.3 KeV. A simple X-ray fluorescence experiment establishes clearly that the window is physically in the Ge material itself. Experiments with detectors made from different Ge crystals, and with Schottky barrier contacts of different metals indicate that the effect is due to a basic property of the transport of electrons near a surface. Theoretical considerations and a Monte Carlo calculation show that the window is caused by the escape of warm electrons which are the end product of a photo event. The mean free path of the electrons becomes longer as they lose energy by optical phonon collisions and they can be trapped at the surface before they are picked up by the electric field.

The availability of germanium with less than 1010 acceptors/cm3 has enabled us to fabricate and study the performance of germanium detectors of substantially larger volume than has hitherto been possible except by using lithium-drifting techniques. Detectors having a sensitive volume of up to 25 cm3 have been made. These detectors have a lithium diffused n+ contact and a metal barrier non-injecting back contact, and are totally depleted. The energy resolution of these detectors is similar to that of equivalent sized lithium-drifted detectors, but the high-purity detectors are easier to fabricate, and present considerably fewer handling problems.

Several important areas for the use of germanium spectrometers would be opened if these detectors could be successfully operated at temperatures considerably in excess of 77°K. Precise measurements on a series of detectors made from different high-purity germanium crystals and on one lithium-drifted germanium detector indicate that the resolution does not degrade significantly until nearly 150°K is reached. Similar-sized detectors, including the lithium-drifted device, exhibited essentially identical temperature-resolution relationships.

Results demonstrating the successful use of a gold surface-barrier as a thin-window p + termination of the intrinsic region of a lithium-drifted germanium detector are reviewed; these results are a rebutal to statements recently made by Dearnaley et al. in this journal.

A Li-diffused n+ surface layer is the best n+ contact developed up to the present time. However, the effective dead layer resulting from the Li-diffusion prevents the optimum stacking of detectors for counter telescopes. Diffusion temperatures as low as 160°C have been successfully used to produce effective dead layers as thin as 10 ?. Unfortunately, these dead layers increase significantly when the detector is annealed following radiation damage.