CdZnTe substrate producibility and its impact on IRFPA yield

Abstract

The need for cost effective production of HgCdTe infrared detectors and focal plane assemblies has led to increased attention to the availability of high quality large-area CdZnTe substrates. Reasonable yield of large-area substrates (≥4 cm×6 cm format) is necessary for fabrication of focal plane assemblies (FPAs) now in production, and for future infrared (IR) detectors which are growing in size and complexity. Raytheon’s infrared materials producibility (IRMP) program has addressed this issue, after identifying critical drivers of FPA yield coming from substrates, and targeted certain improvements in substrate process steps for highest impact on large-area substrate yield. Three specific areas of improvements in the substrate process were addressed: (1) compounding of a large 6 kg charge of CdTe; (2) vertical Bridgman growth of 92 mm diameter CdZnTe boules in both quartz and pyrolytic boron nitride (PBN) crucibles; and (3) optimized Cd overpressure control during growth and cool-down of the boule. It was shown that the Cd overpressure and the cooling schedule had the strongest effects on defect populations. The resulting improvements include a 33% increase in wafer yield per unit starting weight, an estimated 50% reduction in substrate cost per cm², better morphology of epitaxial HgCdTe layers, and improved yield of satisfactory IR detectors. The criteria for selecting substrates have also improved as a result of this work. In addition, photovoltaic detectors were fabricated on wafers from a variety of sources, and tested. Results compare favorably with those on baseline (earlier process) substrates.