Research

AlInAsSb Avalanche Photodiodes: Recently, we have been working on avalanche photodiodes fabricated from a new material, AlInAsSb, a collaboration with Seth Bank’s group at the University of Texas. This material is grown as a digital alloy. Digital alloys are essentially a short-period, multicomponent generalization of superlattices, where the superlattice period is reduced sufficiently that charge carrier wavefunctions integrate over sufficiently many periods that the material properties begin to approximate those of the bulk, random alloy.  However, the periodicity gives rise to key differences in the electrical and optical properties that promise to enable radical new material and device capabilities. 

For the past 50 years, there have been efforts to develop a solid-state alternative to photomultiplier tubes. Staircase APDs proposed by Capasso, 1983, utilize sequentially bandgap-graded regions. Under reverse bias, their energy band diagrams look like a staircase from which the device gets its name. The operating principle is that each time an electron transits a stair step, it will have sufficient energy to impact ionize. However, for 40 years, all attempts to demonstrate a staircase APD failed. Initially, the material system, AlxGa1-xAs/GaAs was used to fabricate the staircase band structures. Unfortunately, the AlxGa1-xAs/GaAs conduction band discontinuity is insufficient to impact ionize GaAs, particularly for high-energy electrons scattered to satellite valleys. The AlxIn1-xAsySb1-y material system, on the other hand, is well suited for the staircase APD structure. The direct bandgap is widely tunable, and the change in bandgap occurs almost entirely in the conduction band. Using this material system, we reported [S. D. March, et al., Nat. Photonics vol 15, pp. 468–474, 2021], the first successful operation of a staircase APD, achieving confirmation of 2N gain scaling with the number of staircase steps, N. The bandgap steps function analogously to the dynodes of a photomultiplier tube, creating a more deterministic gain process with a resultant reduction in gain fluctuations and, thus, lower noise. There have been numerous theoretical studies of staircase APDs. They have consistently predicted that the detector noise would be the same as a photomultiplier tube. This is even less than the noise of an ideal conventional APD. Our measurements of the staircase APD noise confirm that it does indeed exhibit the ultra-low noise of a photomultiplier, five times lower than the lowest noise reported for a conventional avalanche photodiode. We are currently working on arrays of these detectors for next-generation imaging.

We have also found that conventional structure AlInAsSb avalanche photodiodes exhibit noise comparable to Si, the standard for low-noise APDs for the past 50 years. A problem for optical fiber communications is that Si does not work at the wavelengths used in fiber optic systems (1.3 µm to 1.6 µm). Many approaches involving novel materials and structures have been tried to achieve Si-like noise for the telecommunications wavelengths, and none succeeded until our AlInAsSb digital alloy work. This work is a breakthrough for optical receivers for short-to medium-distance optical communications such as data centers. Recently, with the advance of LIDAR systems in highly populated urban areas, eye safety has become a growing concern. Higher laser power must be used to allow for a more extended range and higher resolution detection, which increases the potential for eye damage. The lasers and detectors that are the most attractive candidates have been commercially developed for optical communications. They operate near 1.55 µm wavelength.  To circumvent the eye-damage issue, longer wavelength lasers can be employed, allowing for higher power operation and, thus, better LIDAR imaging. The 2-µm window is ideal for this application, and the lasers are available. However, suitable detectors are not. Imaging is a second area of interest to the DoD agencies. Recently, there has been a lot of interest in multispectral imaging, i.e., detectors in different spectral ranges, somewhat like the cones in the human eye. Again, 2 µm is a critical wavelength. Using the AlInAsSb digital alloy, we have successfully demonstrated AlInAsSb avalanche photodiode with low dark current and the same low noise (comparable to or better than Si) that operate to wavelengths as long as 3 µm.

High-Power, High-Speed Photodiodes: We have developed high-power, high-speed photodiodes for several years for analog optical links. Applications include the replacement of point-to-point microwave links, antenna remoting (e.g., the Atacama Large Millimeter Array in Chile), generation of millimeter-wave frequencies, beam-forming networks for phased array antennas, photonic processing of microwave signals, and precision microwave spectroscopy. Using a modified uni-traveling carrier (MUTC) photodiode that we have developed, we have achieved record output power and linearity in a broad frequency range (10 GHz to > 120 GHz). One of our collaborators is the National Institute of Standards and Technology. Recently, they have used our photodiodes to demonstrate coherent down-conversion of the phase of a state-of-the-art optical clock, 1×10-18 absolute stability on a 10 GHz microwave signal, and <1×10-19 frequency offset through the down-conversion process. This is the record absolute stability of microwaves and the first demonstration of microwaves with better absolute stabilities than Cs fountain clocks. Our colleagues indicate that this has the potential to become a new time standard.

Single Photon Detection: We actively use avalanche photodiodes for single photon detection. This has involved developing new photodetector structures optimized for single photodetection and novel quenching circuits to enable higher detection rates. A primary application is quantum optics and quantum communications. We are participants in three NSF quantum centers.