Abstract

Our photoconductive detector is basically a p-i-n-FET, integrated into a single device. Its function is based on the light-induced increase of the conductance in the n-layer of a p-i-n structure, which is provided with two ohmic contacts to the n-layer and one to the p-layer. The goal of designing this device is to obtain the largest possible photoconductive response in the n-channel with a minimum number of photons incident on the detector. Apart from optimizing the quantum efficiency of the p-i-n diode, this goal can be achieved by minimizing the capacitance of the device. Therefore we use a sophisticated sample design with a small detection area, which surrounds a spatially separated large absorption area. Figure 1 shows a SEM picture of the device. The n-layer of the large absorption area is always depleted (etched down with respect to the detection area) and thus does not contribute to the device capacitance.¹ Because of the ‘giant ambipolar diffusion constant of such structures,² the diffusion of the photogenerated carriers from the inner absorption area to the surrounding detection area is fast enough (e.g., τ _diff < 50 ps if the diameter of the absorption area is less than 20 μm). This design allows the combination of narrow contact spacings, necessary for a high photoconductive gain, with small RCRC and diffusion time constants. This results in high gain-bandwidth values (>20 GHz). The 3-dB frequency is adjustable by the integrated load resistor.

© 1995 Optical Society of America