HgTe/HgCdTe Superlattices Grown on CdTe/Si by Molecular Beam Epitaxy for Infrared Detection

Journal of Electronic Materials, Jun 2004 by Selamet, Y, Zhou, Y D, Zhao, J, Chang, Y, Et al

The general behavior of the in-plane electrical properties of an HgTe/CdTe SL is the same as for other narrow bandgap semiconductors as demonstrated by Hall measurements. Figure 7 shows in-plane Hall data (carrier concentration and mobility versus temperature) of one of the single SL layers. Other layers have similar results. The Hall measurements were, carried out as a function of temperature using the Van der Pauw geometry. As seen in Fig. 7, the carrier concentration in this SL layer is in the low 10^sup 15^ cm^sup -3^ range, with n-type mobilities exceeding 10^sup 4^ cm^sup 2^/Vs.

TEST-STRUCTURE DEVICE FABRICATION AND CHARACTERIZATION

A test structure was fabricated by mesa etching using a bromine methanol solution. The conventional method of converting an undoped-HgCdTe top layer to p^sup ^ is by As ion implantation. After implantation, the structure needs to be annealed at high temperatures to activate the As and reduce extensive surface and lattice damage caused by high-energy ion bombardment. Such a high-temperature (425-450�C) ac-tivation and annealing procedure may have adverse effects on the SL interface sharpness and the junction location. Our recent study3 showed that SLs could not withstand high-temperature annealing because of layer interdiffusion. Longer anneal times increase the interdiffusion of the SL layers. A best case scenario involves annealing at temperature below. 235�C for short periods of time, which will not be sufficient for As activation. Consequently, doping with As appears impractical, so we used Au for the p-type conversion of the top HgCdTe layer. Windows in the CdTe cap layer were opened photolithographically for selective gold diffusion. A very thin, ultrahigh-purity Au layer was deposited and sequentially annealed at low temperatures to diffuse the Au into the structure. This method is known to dope samples p-type.4 However, a systematic study to optimize such junction formation is required to achieve state-of-the-art performance. The primary goal in this study was to form p-n junctions at low enough temperatures to preserve the SL integrity, so that the SL-based detector properties, such as spectral cutoffs, can be evaluated.

The fabricated devices were mounted in a liquid N^sub 2^ dewar with a ZnSe window, and spectral-response measurements both at 80 K and room temperature were carried out. Figure 8a and b shows representative spectral-response curves for one of the devices at 80 K and room temperature, respectively. A metal screen was used to reduce the background flux. The device gives a strong signal even at room temperature. In Fig. 8a, two main response peaks are observed. The position of the broader peak at longer wavelength is about 5 �m, which we will refer to as peak 1. The cutoff of this peak is 7 �m. When compared with layers with similar structures and material growth conditions, we conclude that the only possible origin of this signal is the HgTeTHgCdTe SL. Several sub-peaks on this peak may correspond to other intersubband transitions. The position of the other peak (peak 2) is at 2.7 �m, with a cutoff at 3 �m. When compared with the previous layer-growth data, we conclude that it comes from the HgCdTe layers. We also measured the room-temperature spectral response. Assuming peak 2 is due to HgCdTe, we used the Hansen and Schmidt formula5 to predict the room-temperature positions of the peak and cutoff. The 80-K cutoff of peak 2 gives a predicted value of the room-temperature cutoff at 2.85 �m. From Fig. 8b, we see that these two peaks merge together, and there is a shoulder peak at the position of 2.6 �m with an estimated cutoff at 3 �m, in reasonable agreement with the predicted value. A comparison of the room-temperature spectral-response curve to the room-temperature FTIR data of Figs. 5 and 6 also suggests that this shoulder peak comes from the HgCdTe-alloy layer, and the broader peak to the long wavelength direction (peak 1) comes from the HgTe/ HgCdTe SL. Figure 8b shows the position of this broader peak to be at 4.5 �m with a cutoff at 5.4 �m. The room-temperature cutoffs for the HgCdTe alloy and HgTe/CdTe SLs are in good agreement with the ones obtained from room-temperature FTIR measurements. These assignments are corroborated by calculated absorption and transmission spectra.