Toward Deterministic Construction Of Low Noise Avalanche Photodetector Materials

Over the past 40+ years, III-V materials have been intensively studied for avalanche photodetectors, driven by applications including optical communications, imaging, quantum information processing, and autonomous vehicle navigation. Unfortunately, impact ionization is a stochastic process that introduces noise, thereby limiting sensitivity and achievable bandwidths, leading to intense effort to mitigate this noise through the identification of different materials and device structures. Exploration of these materials has seen limited success as it has proceeded in a largely ad hoc fashion due to little consensus regarding which fundamental properties are important. Here, we report an exciting step toward deterministic design of low-noise avalanche photodetector materials by alternating the composition at the monolayer scale; this represents a dramatic departure from previous approaches, which have concentrated on either unconventional compounds/alloys or nanoscale band-engineering. In particular, we demonstrate how to substantially improve upon the noise characteristics of the current state-of-the art telecom avalanche multipliers, In0.52Al0.48As grown on InP substrates, by growing the structure as a strain-balanced digital alloy of InAs and AlAs layers, each only a few atomic layers thick. The effective k-factor, which has historically been considered a fundamental material property, was reduced by 6-7 x from k= 0.2 for bulk In0.52Al0.48As to k= 0.05 by using the digital alloy technique. We also demonstrate that these "digital alloys" can significantly extend the photodetector cutoff wavelength well beyond those of their random alloy counterparts. Published by AIP Publishing.