Electron Multi-Injector: A Strategy for Improving Quantum Efficiency of Infrared Photodetectors

Summary form only given. The ultimate limit of the sensitivity of a photodetector can be expressed by its normalized detectivity as [1]: D* = [(λ)/(hc)] √{[(Ao)/(Ae)]} η√{[1/(2 (G+R)t')]} where for a given optimized detector material and thickness, the ratio of optical to electronic area of the detector, A
_o
/A
_e
, is a crucial parameter for the detector performance. By leveraging the degree of freedom intrinsic to this ratio, Electron Injector (EI) technology has proven capable of surpassing the current performance of commercial short-wave infrared (SWIR) cameras [2]. Unlike conventional photon detectors and imagers, which are based on planar sensing elements made of a stack of semiconductors, EI uses low-dimensional charge confinement in a 3D structure to achieve highly sensitive photon detection. Recently, theoretical and experimental work [3,4] has motivated the push towards further shrinking in size of the electronic area in order to achieve yet higher sensitivity. However, this improvement in sensitivity comes at the cost of a decrease in quantum efficiency, as the diffusion length of the minority carriers limits the area over which a photogenerated carrier can be collected at the device junction before recombining [5]. We demonstrate an EI detector architecture that allows to alleviate the detrimental effect on quantum efficiency of small electron injectors, significantly increasing the area fill factor of the detectors while maintaining a small device capacitance. The proposed architecture encompasses the use of multiple small electron injectors to form a single pixel of the FPA: such injectors are separated in space so as to span a larger portion of the optical area of the absorber, without the need for large electronic area. These devices exhibit a more than 2-fold increase in fill factor at no apparent costs for the responsivity and speed of the devices, thanks to the modest increase in electronic area. The presented strategy is fully compatible with the hybrid integration of the FPA with a Si ROIC, and can be further developed to optimize the number, spacing and geometry of the injectors, in order to reach the ideal tradeoff point between increase in fill factor and decrease in detector speed and sensitivity.