As-Received CdZnTe Substrate Contamination

Abstract

State-of-the-art as-received (112)B CdZnTe substrates were examined for surface impurity contamination, polishing damage, and tellurium precipitates/ inclusions. A maximum surface impurity concentration of Al = 7.5 9 10 14 , Si = 3.7 9 10 13 , Cl = 3.12 9 10 15 , S = 1.7 9 10 14 , P = 7.1 9 10 13 , Fe = 1.0 9 10 13 , Br = 1.9 9 10 12 , and Cu = 4 9 10 12 atoms cm À2 was observed on an asreceived 6 9 6 cm wafer. As-received CdZnTe substrates have scratches and residual polishing grit on the (112)B surface. Polishing scratches are 0.3 nm in depth and 0.1 lm wide. The polishing grit density was observed to vary from wafer-to-wafer from $5 9 10 6 to 2 9 10 8 cm À2 . Te precipitate/inclusion size and density was determined by near-infrared automated microscopy. A Te precipitate/inclusion diameter histogram was obtained for the near-surface (top $140 lm) of a 6 9 6 cm substrate. The average areal Te precipitate/inclusion density was observed to be fairly uniform. However, there was a large density of Te precipitates/inclusions with a diameter significantly greater than the mean. Te precipitate/inclusion density>10 lm diameter = 2.8 9 10 3 cm À3 . The large Te precipitates/inclusions are laterally non-uniformly distributed across the wafer.

Characterization of (111)B and (211)B CdZnTe Substrates for HgCdTe Growth

“…The as-received substrates were characterized, then polishing and/or etching processes were performed on the substrates, and they were characterized a second time. We found that the number of particles and defects on the as-received and etched (211)B substrates were large and agreed with the numbers found by Benson et al 8,9 However, our interpretation of these features is different. Furthermore, the (111)B substrates gave the surprising result that there were very few features on the as-received substrate, but after the growth preparation etch the density of morphological defects/features on the surface increased considerably.…”

“…The number density here was in agreement with that of Benson et al for (211)B Nippon substrates, but while they found the particles to be silica grit before etching and CdZnTe after Br:methanol etching, we believe they are largely Te precipitates. 8,9 In our study, silica grit particles were mostly found close to the beveled edges, but had a low density. A small agglomeration of such grit particles is shown in Fig.…”

“…measured an RMS roughness of 0.4 nm on as-received (211)B substrates. 8 FTIR transmission spectra were recorded from a 5 × 5 grid on the as-received SubC. The transmission spectra and a transmission map at 500 cm -1 are shown in Fig.…”

“…Benson et al have studied (211)B substrates from JX Nippon Mining & Metals Corporation, Japan (Nippon), which are generally considered state-of-the-art substrates for HgCdTe growth. [7][8][9][10] They found that the as-received substrates had a significant amount of impurity contamination, particles, Te precipitates and scratches on the surface, and they concluded the substrates were not epi-ready out of the box.…”