1/f noise in very-long-wavelength infrared Hg1−xCdx Te detectors

This paper investigates 1/f noise performance of very-long-wavelength infrared (VLWIR) Hg1-xCdxTe (cutoff wavelengths λc = 15 µm and λc = 16 µm) photodiodes at 78 K, where detector current is varied by changing detector area, detector bias, and illumination conditions. Holding detector bias and temperature constant, the 1/f noise current is proportional to the detector current. Significant nonuniformity is observed in the noise data for each detector area because of the varying detector quality. Defects are presumed resident in the detectors to produce greater nonuniformity in 1/f noise as compared to dark current at 100-mV reverse bias. For λc = 16 µm, 4-µm-radius, diffusion-limited diodes at 78 K and 100-mV reverse bias, the average dark current is Id = 9.76 ± 1.59 × 10-8A, while the average noise current measured at 1 Hz is in = 1.01 ± 0.63 × 10-12A/Hz1/2. For all detector areas measured, the average ratio in 1-Hz bandwidth is αD = in/Id = 1.39 ± 1.09 × 10-5. The 1/f noise was also measured on one diode as a function of detector-dark current as the applied bias is varied. In the diffusion-limited portion of this detector's current-voltage (I-V) curve, to about 130 mV, the 1/f noise was independent of bias. For this diode, the ratio αD = in/Id = 1.51 ± 0.12 × 10-5. The 1/f noise associated with tunneling currents is a factor of 3 greater than the 1/f noise associated with diffusion currents, αT = in/IT = 5.21 ± 0.83 × 10-5. In addition, 1/f noise was measured on detectors held at - 100 mV and 78 K under dark and illuminated conditions. The measured ratios αp ∼ αD ∼ 1.5 × 10-5 were about the same for the dark and photon-induced diffusion currents. Therefore, the diffusion current appears to have a unique value of α as compared to the tunneling current. This may be indicative of unique noise-generation mechanisms associated with each current.