Positron-defect profiling in Cd^sub 1-x^Zn^sub x^Te wafers after saw cutting

Journal of Electronic Materials, Jun 2003 by McNeil, Sean P, Lynn, Kelvin G, Weber, Marc H, Szeles, Csaba, Soundararajan, Raji

Near-surface damage induced by saw cutting of ingots of Cd^sub 1-x^Zn^sub x^Te was investigated by positron-defect depth profiling. The damage extends to several micrometers depth and depends on the cutting apparatus. The samples were polished and etched repeatedly, followed each time by positron-depth profiling. New subsurface damage created during the polishing process is observed. No new damage is observed after etching. Positron-depth profiling is suggested as a diagnostic tool to monitor the quality of sample surfaces.

Key words: Surface damage, saw cutting, semiconductors, CdZnTe, positrons, defects, depth profile

INTRODUCTION

In many environments, the surface quality of a sample is critical to the performance of optical and electronic devices. Single crystals are grown, sliced, diced, and polished prior to their incorporation in the device. Both the cutting and the polishing processes can leave near-surface damage. In radiation detectors fabricated from large bandgap materials, such as Cd^sub 1-xZn^sub x^Te, near-surface damage will degrade the energy resolution. In laser crystals, damage may cause excess heat that ultimately can destroy surface coatings. To avoid production-line problems, an excess amount of material is removed after cutting, thereby lowering the production yield. Alternatively, the material can be annealed or etched to remove damage. However, in some cases, such as Cd^sub 1-x^Zn^sub x^Te, cadmium vacancies are created during the annealing process, rendering the material useless for radiation-detector applications. Corners and edges are rounded off by etch steps. An electric field applied to collect charges created from incident radiation can no longer be controlled precisely. It is imperative to limit the surface treatment.

POSITRON-ANNIHILATION MEASUREMENTS

Positrons, the antiparticles to electrons, can be used to profile near-surface damage in a sample.1-5 With a beam of monoenergetic positrons with energies from several eV to 60 keV positrons can be implanted into a sample where the mean implantation depth is a function of beam energy and sample density. Prior to their annihilation with electrons, they diffuse through the material. At defects, the local electron-momentum distribution and density is altered compared to a defect-free crystalline region, creating regions of repulsive or attractive potentials for the positrons. The latter are predominantly vacancies, clusters thereof, or extended defects like dislocations (referred to later as open volume-type defects), all of which act as traps for positrons. At room temperature, detrapping is unlikely. The momentum distribution of the electrons at the site of annihilation determine the annihilation line width by way of Doppler shifts, which over the sum of many annihilations cause Doppler broadening. A sharp narrow line is indicative of low-momentum electrons from delocalized conduction-band states, while electrons from valence bands and deeper lying states broaden the annihilation line. The ratio from the former to the latter increases at open volume-type defects. Hence, the sharpness of the annihilation line is an indicator of the presence of such defects. The concentration of positron-trapping defects also influences the diffusion length of the positrons.

The implantation energy controls the mean implantation depth of an implantation profile that is very similar to the first derivative of a Gaussian bell curve. The mean depth is proportional to the empirical relation E^sup 1.6^/[rho], where E is the incident-positron energy, and p the sample density. In a typical measurement, the "sharpness" of the annihilation line for a given mean implantation depth is assessed by comparing the peak height of the line to the total area of the peak. Defect-depth profiles down to several [mu]m can be performed. Vacancy concentrations to about 1 part per million are detectable. The extent of a near-surface damage region can be deconvoluted from the measured depth profiles. More detailed descriptions of the interaction of positrons with solids can be found in the review by Krause-Rehberg and Leipner.1 Positron beams are discussed, for example, in Ref. 2.

In the work presented here, the thickness of near-surface damage regions in slices of Cd^sub 1-x^Zn^sub x^Te is determined by positron-depth profiling. The effects of polishing and etching of the samples will be shown. A large number of alternate and well-established techniques, such as x-ray diffraction, low-temperature photoluminescence, deep-level transient spectroscopy, and capacitance measurements, to name just a few, are very powerful tools for defect detection and characterization. While positron-annihilation tools are among the least well-established tools, they offer the advantages of depth dependence and sensitivity to open-volume defects. Absolutely no sample preparation, such as contact depositions, nor thin slices of samples are required. The intent of this contribution is to demonstrate the capability of the technique. A detailed correlation of positron results with surface treatment reaches beyond the scope of this paper. Excellent examples of positron near-surface damage studies can be found in work by Borner et al.6 and Krause-Rehberg et al.7 on sawcut GaAs wafers. Wang et al. report a study on Nd: YAG surface treatment.8