Mid-Wave Infrared Resonant Cavity Detectors

The dominant issue confronted by designers and users of mid-wave infrared (MWIR) detectors is
their large dark current and its noise-producing fluctuations. MWIR dark current is more than a million times that of short-wave (telecom) IR detectors. Several decades of work on dark current reduction via attention to material quality and epitaxial structure are nearing maturity. Significant improvements in dark current require new approaches. Our approach reduces dark current by reducing the volume of the detector’s absorber layer. A thin absorber layer is located inside an optical cavity formed by two mirrors – the so-called Resonant Cavity Detector (RCD). Optical absorption of the thin layer is typically ~1%, which would produce unacceptably small quantum efficiency (QE) in a single-pass detector. In an RCD, the light passes through the thin absorber many times (~50-100) as it re-circulates within the optical cavity, which produces good QE, and also low dark current. RCDs with QE~50% have achieved nearly 10× reduction in dark current and we expect to eventually reduce dark current by 100×.The RCD approach also affects frequency bandwidth (speed) and spectral bandwidth. RCDs absorbers are ~100× thinner than conventional detectors, which reduces minority carrier transit time, thus the frequency bandwidth is expected to increase to > 10 GHz. The spectral bandwidth, Δ, of RCDs is greatly reduced due to interference of the recirculating light with itself. RCDs with Δ ~ 20 nm have been produced, and Δ ~ 3 nm seems feasible. The narrow Δ and low dark current makes RCDs ideal for narrow band applications, such as detection of lasers or sensing spectrally sharp absorption features of gases. The RCD’s spectral response can be tuned 50-100 nm with detector temperature or the light’s angle of incidence. Finally, we discuss the tolerance on the epitaxial growth, which is significantly tighter for RCDs (1% layer thickness control) than for conventional MWIR detectors.