Minority carrier diffusion lengths and mobilities in low-doped n-InGaAs for focal plane array applications

Abstract

The hole diffusion length in n-InGaAs is extracted for two samples of different doping concentrations using a set of long and thin diffused junction diodes separated by various distances on the order of the diffusion length. The methodology is described, including the ensuing analysis which yields diffusion lengths between 70 -85 µm at room temperature for doping concentrations in the range of 5 − 9 × 10 15 cm −3 . The analysis also provides insight into the minority carrier mobility which is a parameter not commonly reported in the literature. Hole mobilities on the order of 500 − 750 cm 2 /V·s are reported for the aforementioned doping range, which are comparable albeit longer than the majority hole mobility for the same doping magnitude in p-InGaAs. A radiative recombination coefficient of (0.5±0.2)×10 −10 cm −3 s −1 is also extracted from the ensuing analysis for an InGaAs thickness of 2.7 µm. Preliminary evidence is also given for both heavy and light hole diffusion. The dark current of InP/InGaAs p-i-n photodetectors with 25 and 15 µm pitches are then calibrated to device simulations and correlated to the extracted diffusion lengths and doping concentrations. An effective Shockley-Read-Hall lifetime of between 90-200 µs provides the best fit to the dark current of these structures.

Heavy and light hole minority carrier transport properties in low-doped n-InGaAs lattice matched to InP

“…Results from devices from two different wafers than the device reported here were reported in Ref. 13 and are similar to the present results. A fit to measured collected current densities for an applied bias of 0.3V to Eq.…”

“…The hole density under the injecting diode is p inj , which is dictated by the applied bias, and the hole concentration at the collecting diode is the equilibrium value p 0 since the collecting diode is maintained at zero applied bias. Using these boundary conditions, and assuming no interface recombination, solving the diffusion equation leads to the solution in terms of the collected hole current density J p (W) as a function of interdiode separation W given as 13,14 …”

“…11,12 The method is demonstrated using an n-type InP:Si/intrinsic InGaAs/n-type InP:Si a) Electronic mail: alexandre.walker@nrc-cnrc.gc.ca double-heterostructure which is used to make pin short wave infrared detectors and focal plane arrays. 13 Furthermore, this method is shown to be capable of separating the contributions of heavy and light holes, and, with a few assumptions, determining the lifetimes and diffusion constants (i.e. mobilities).…”

“…The basic physics is summarized in detail elsewhere. 13,14 d 1 Essentially, forward biasing one line diode injects minority carriers into the InGaAs layer leading to diffusion of carriers through the InGaAs channel. This can be seen in Fig.…”

“…This is in agreement with other studies conducted on low-doped InGaAs. 13,[19][20][21] While technically this method cannot distinguish between bulk recombination in the layer and surface recombination at the hetero junction interfaces, the apparent dominance of radiative recombination justifies the assumption of negligible surface recombination. The lifetimes for heavy holes are more accurate than those for light holes.…”