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

High-quality HgTe/CdTe super-lattices (SLs) and device structures incorporating them were grown by molecular beam epitaxy (MBE) on CdTe/Si substrates. In-situ techniques, such as reflection, high-energy electron diffraction and spectroscopic ellipsometry, were extensively used to rigorously control the growth parameters. The full width at half maximum (FWHM) of x-ray doublecrystal rocking curves (DCRCs) were 100-150 arcsec, comparable to those of HgCdTe alloys grown on the same type of substrates. The room-temperature Fourier transform infrared (FTIR) spectrum exhibits two-dimensional features characteristic of SLs. Trial devices in a p^sup +^-n^sup -^-n^sup +^ format were fabricated by diffusing gold in order to further evaluate the HgTe/CdTe SL performance. Gold diffusion was chosen to fabricate photovoltaic junctions in order to preserve the structural integrity of the SLs during the device processing. Though no attempt was made in the current study to optimize the junction properties by Au diffusion, this method has proven to be very useful for rapid preliminary evaluation. The measured spectral-response and detectivity data indicate the possibility to fabricate photovoltaic devices on an HgTe/CdTe SL, although further work is needed to optimize the p-n junction fabrication.

Key words: CdTe/Si substrates, x-ray, rocking curves, superlattices, MBE

INTRODUCTION

An infrared (IR) photon detector operating at temperatures higher than the current ones will eliminate the expensive cryogenic-cooling equipment, improve the reliability of imaging systems, and greatly improve the yield. To achieve background limited performance (BLIP), the cooling requirements of current IR detectors depend on the cutoff wavelength. Theoretically, the detectors need to be cooled to about 180 K to 20 K for cutoff wavelengths varying from the midwavelength infrared (MWIR) to very long wavelength infrared (VLWIR) region, respectively. Increasing the BLIP operation temperature of these detectors, especially in the long wavelength region, is highly desirable and critical for various space applications.

Increasing minority carrier-recombination lifetimes would lead to higher operating temperatures without performance degradation. There are two intrinsic carrier-recombination mechanisms: Auger and radiative. The third limiting mechanism, Shockley-Read-Hall recombination, is due to defects and impurities in the sample that give rise to trap levels in the energy gap. The effects of this limiting mechanism can be reduced by optimizing growth conditions and/or post-growth thermal treatment. The Auger recombination rate strongly depends on the carrier concentration (increases as the square of the carrier concentration at low and moderate concentrations and at a slower rate at higher concentrations when carriers are degenerate). The radiative recombination rate, on the other hand, increases only linearly with the carrier concentration and is not the limiting mechanism throughout most of the operating temperature region for long wavelength detectors.

In the VLWIR region, extrinsic materials, such as doped Si and Ge with impurity band conduction, are typically used as IR detectors. However, these materials have millimeter-scale absorption lengths and cannot be used to form large area arrays. Also, these detectors operate at very low temperatures (20 K or below), limiting their applications. Intrinsic detectors have several inherent advantages over extrinsic detectors. They have higher quantum efficiencies, absorption coefficients, and operating temperatures, leading to large format arrays. However, achieving VLWIR cutoff control with intrinsic alloy materials is very difficult. At these wavelengths, HgCdTe has a very narrow bandgap, and any small variation in the Cd mole fraction translates into a large variation in the cutoff wavelength. The HgTe/ CdTe superlattices (SLs) are relatively more robust against composition variations because the cutoff wavelength is controlled by the well thicknesses. Layer thickness can be controlled down to atomic scales with molecular beam epitaxy (MBE) growth.

The HgTe/CdTe SLs have additional advantages including a near lattice match (~0.3% mismatch) between HgTe and CdTe and a common anion, Te. The HgTe/CdTe SLs can be more stable than HgCdTe alloys because of a spatial separation of HgTe and CdTe layers.1

The HgTe/CdTe SL-based IR photovoltaic detectors are expected to have lower tunneling currents because of a larger effective mass in the growth direction, which is inversely related to the CdTe barrier width, d^sup b^. For a large d^sup b^, the perpendicular effective mass is also large and tunneling current low, whereas the opposite is true when db is small. However, if the effective mass is too large, then perpendicular transport is no longer possible. Therefore, a compromise is required to achieve a reduced tunneling current with a sufficiently large vertical transport. This requires the perpendicular effective mass to be about ten times larger than the in-plane effective mass and d^sup b^ [approximate]5 nm.