Search

Anomalous secondary-ion mass spectroscopy (SIMS) profiles of copper in thin pieces of HgCdTe are explained using the model used for diode formation by ion milling and ion implantation. In this model, the SIMS ion beam injects mercury interstitials into the HgCdTe as it etches the HgCdTe. The interstitials fill metal vacancies and kick copper off the metal lattice sites. The copper interstitials then diffuse either to the surface being etched, where it is removed and detected by the SIMS instrument, or deeper into the HgCdTe, where it annihilates vacancies. Good agreement between model predictions and experimental SIMS profiles are obtained.

The high-operating-temperature (HOT) midwave infrared (MWIR) n-on-p detector has been pursued using the high-density vertically integrated photodiode (HDVIP®) architecture. In this device, arsenic-doped HgCdTe grown by liquid-phase epitaxy (LPE) is used, passivated on both surfaces with interdiffused CdTe. Dark current densities on these diodes as low as 2.5 mA/cm2 normalized to a 5 μm cutoff at 250 K have been demonstrated. 1/f noise at 1 Hz, measured at 250 K, is found to be 6 × 10−11 A/rHz-cm measured on a cutoff of 4.43 μm. These results agree with the theoretical predictions for the devices made.

Hg1−x Cd x Te samples of x ~ 0.3 (in the midwave infrared, or MWIR, spectral band) were prepared by molecular beam epitaxy (MBE) for fabrication into 30-μm-pitch, 256 × 256, front-side-illuminated, high-density vertically-integrated photodiode (HDVIP) focal plane arrays (FPAs). These MBE Hg1−x Cd x Te samples were grown on CdZnTe(211) substrates prepared in this laboratory; they were ~10-μm thick and were doped with indium to ~5 × 1014 cm−3. Standard HDVIP process flow was employed for array fabrication. Excellent array performance data were obtained from these MWIR arrays with MBE HgCdTe material. The noise-equivalent differential flux (NEΔΦ) operability of the best array is 99.76%, comparable to the best array obtained from liquid-phase epitaxy (LPE) material prepared in this laboratory.

Exposure to specific damage introduced by either ion implantation or ion milling converts p-type short wavelength infrared (SWIR) HgCdTe to n-type in a manner similar to the conversions in medium wavelength infrared (MWIR) or long wavelength infrared (LWIR) mercury cadmium telluride. However, the depth of conversion for SWIR Hg1−xCdxTe, with x=0.48, is approximately 300% smaller when compared to the depth of conversion for MWIR HgCdTe for an identical degree of ion milling. The depth of conversion, or the n/p junction depth, tracks linearly the extent of surface removals by ion milling when the metal vacancy concentration is held invariant. These results can be correlated to the interaction between metal vacancies and a product of the lattice damage process resulting from ion milling. The observation of a linear dependence of this depth on the degree or time of ion milling rules out the existence of a diffusive barrier in the transfer of this product for both MWIR and SWIR HgCdTe.

High operating temperature (HOT) detector concepts using midwave infrared (MWIR) (x∼0.3) p-type HgCdTe operating at temperatures within the thermoelectric cooler range are of significant interest at the present time. However, it is apparent that much work remains to be done in the areas of material, diode passivation, and diode formation technologies before the “holy grail” of photon detection at room temperature for all infrared wavelengths is achieved. Over the years, at DRS, we have developed a technology base for both n- and p-type HgCdTe materials parameters that are relevant to photodiode design and fabrication. This paper will discuss data that we have taken recently on minority carrier lifetime in MWIR and long wave infrared (LWIR) HgCdTe, particularly p type, and how it compares to current theories of Auger 7, radiative, and Shockley-Read recombination in this material. Extrinsic group IB (Cu, Au) and group V (arsenic) p-type dopants were used, together with group III (In) for n-type. The impact of the data on future HOT detector work is discussed.

Transmission spectra of liquid-phase epitaxy (LPE) Hg1-xCdxTe with Cd mole fractions in the range of 0.23

Arsenic is known to be an amphoteric impurity that may occupy either sublattice in HgCdTe depending upon sample annealing. As an acceptor in low concentrations, it offers several features that are attractive for the fabrication of certain n + -on-p detector diode structures. The epitaxial growth of arsenic-doped HgCdTe from a Te-rich melt can fulfill the requirements for application in a variety of devices where low vacancy concentrations and low defect densities are critical requirements in minimizing dark currents. These devices may include the high operating temperature (HOT) detectors operated in a strong nonequilibrium and reverse bias mode to suppress the Auger-generated dark currents. For the materials’ growth process to be effective, the segregation coefficient determining the incorporation of arsenic from the Te-rich melt needs to be established. This coefficient was measured during these investigations and was observed to vary with arsenic concentration. Within the range of interest, this parameter varied between 8×10−6 and 1×10−4. These extremely small values limit the doping that can be achieved to

Arsenic incorporated in HgCdTe epifilms at levels ranging from 6×1014 to 4×1016 cm−3, and activated completely as an acceptor, converts into a donor upon introduction of specific damage introduced by ion milling, ion implantation, or ECR. The results could be correlated to the formation of a complex between the arsenic and a product of the lattice damage process. If this product is assumed to be an atom, which could be an Hg interstitial, then a direct mass balance relationship dictates the formation of one Hg interstitial for each 49,000 atoms of Hg removed by the damage process. A very similar relationship is observed in the annihilation of metal vacancies pre-existing in the solid by damage introduced by ion milling. The deactivated arsenic, present as a donor, can be restored back to complete activation as an acceptor upon annealing for relatively short durations at 120°C.

The recently developed Te-rich, liquid-phase-epitaxy growth technology for low arsenic-doped mid-wavelength infrared (MWIR) HgCdTe with p-type doping concentrations

Hg 1− x Cd x Te( x ∼0.22) samples grown by LPE on CdZnTe(111B)-oriented substrates were exposed to various doses of thermal neutrons ( ∼1.0×10 16 −1.7×10 16 n/cm 2 ) and subsequently annealed for ∼24 h in Hg overpressure to remove damage and reduce the presence of Hg vacancies. Extensive magnetotransport measurements were performed on these samples as part of an investigation into the use of elemental transmutation for efficient p-type doping of this material. The data were analyzed using a multi-carrier approach which incorporates various scattering mechanisms and the presence of two conduction channels of differing alloy content to describe the changes in the transport properties due to neutron irradiation.

Many methods for the employment of (Hg,Cd)Te alloys employ anneals at temperatures

Arsenic is an amphoteric impurity in Hg1-xCdxTe. Under tellurium-rich conditions, it tends to occupy metal-sublattice sites and act as a donor; while under mercury-rich conditions, it tends to occupy the tellurium sublattice and act as an acceptor. In this process, mercury from the ambient, first, displaces an arsenic atom (donor) occupying a metal-sublattice site. This arsenic atom, in turn, displaces a tellurium atom, taking residence on the tellurium sublattice (acceptor). A second mercury atom combines with this excess tellurium atom at a surface or other crystalline discontinuity. A quantitative knowledge of the equilibrium reaction is important both to understand point defects and doping in HgCdTe. An investigation of this equilibrium under selected mercury pressures spanning the entire range from the tellurium-saturated phase limit to the mercury-saturated phase limit reveals that a majority of the arsenic remains on the sites in the tellurium sublattice even under tellurium-saturated conditions for arsenic concentrations less than 2 × 1016 cm-3. With increasing mercury pressure, the relative population of the arsenic atoms on sites in the tellurium sublattice monotonically increases, reaching 100% for the mercury-saturated limit.

The kinetics of the activation of arsenic in Hg1−xCdxTe into an acceptor state is discussed based on a new model.1 In this model, it is assumed that the rate-limiting step is the out-diffusion of the excess tellurium that is produced as an arsenic atom jumps from a metal sublattice site to a tellurium sublattice site. In this paper, the model is extended to incorporate additional data from the literature. It is shown that arsenic activation can be accomplished by mercury-rich anneals in reasonable times, over a wide range of temperatures, extending as low as 250 C.

A kinetic model is developed for the activation of group V impurities in Hg1xCdxTe alloys as acceptors. The model assumes that mercury interstitials are introduced at the surface and diffuse into the alloy. There, they displace group V impurities residing on the metal sublattice and place them on the tellurium sublattice, generating tellurium interstitials. These tellurium interstitials then diffuse back to the top surface, or to climbing dislocations. The rate-controlling process is the out-diffusion of the tellurium interstitials. A key finding is that the conversion rate is inversely proportional to the impurity concentration. The model is found to be in good agreement with published data for the activation of arsenic.

The optical absorption coefficient of Cd1−yZny Te near the fundamental band edge was measured at room temperature using transmission spectroscopy. Like in other II–VI semiconductors, it was found that the absorption coefficient exhibits an exponential dependence on incident photon energy according to Urbach’s rule. It was also found that the exponential parameters depend on composition, y, of Cd1−y ZnyTe. A technique is described for determining the composition of Cd1−y ZnyTe from optical transmission spectroscopy. This technique has been implemented in the manufacturing of Cd1−yZnyTe substrates for lattice matched epitaxial growth of HgCdTe.

The operability of long-wavelength p-on-n double layer heterojunction arrays for 40K low-background applications has long been limited by the wide variation in pixel-to-pixel zero bias resistance (Ro) values. Diodes on test structures showing lower performance, with Ro values below 7 × 106 ohm at 40K, usually contained gross metallurgical defects such as dislocation clusters and loops, pin holes, striations, Te inclusions, and heavy terracing. However, diodes with Ro values between 7 × 106 and 1× 109 ohm at 40K contained no visible defects. To study the “invisible” performance-limiting defects (i.e. defects that cannot be revealed by etching), a good correlation between the dynamic resistance at 50 mV reverse bias (R50) value at 77K and the Ro value at 40K was first established, and then used as a tool. The correlation allowed measurements of a large number of devices at 77K, rather than relying exclusively on time-consuming measurements at 40K. Interesting results regarding Ro values at 40K, such as insensitivity to low-density dislocations, mild degradation from Hg vacancies, severe degradation from Hg interstitials, and correlation with junction positioning, were obtained from specially designed experiments.

Excellent high-operating-temperature infrared photodiodes in the medium-wavelength infrared spectral band with cutoff wavelengths ∼5 μm at 77 K were fabricated on Hg1−xCdxTe samples (x∼0.30) prepared by liquid-phase epitaxy in a tellurium-melt reactor. The samples were doped with indium to ∼1×1014 cm−3 and gold to ∼5×1015 cm−3. Thick planar diodes (∼80 μm thick) and thin cylindrical diodes (∼10 μm thick) of the n-on-p type were fabricated, and they gave comparable, excellent detector dark values within a factor of two. At 130 K, dark currents as low as 5×10−7 A/cm2 were obtained.

Impurity levels were tracked through the stages of substrate and liquid phase epitaxy (LPE) layer processing to identify sources of elements which degrade infrared photodetector performance. Chemical analysis by glow discharge mass spectrometry and Zeeman corrected graphite furnace atomic absorption effectively showed the levels of impurities introduced into CdZnTe substrate material from the raw materials and the crystal growth processes. A new purification process(in situ distillation zone refining) for raw materials was developed, resulting in improved CdZnTe substrate purity. Substrate copper contamination was found to degrade the LPE layer and device electrical properties, in the case of lightly doped HgCdTe. Anomalous HgCdTe carrier type conversion was correlated to certain CdZnTe and CdTe substrate ingots.

A new etch system is described which produces pits on the technologically important B face of (111) and (211) CdTe and CdZnTe which are commonly used in mercury cadmium telluride (MCT) epitaxy. A ratio of approximately 10 wide: 1 deep is achieved with this etch allowing its use without removing excessive material. Examples of the use of this etch are given and a comparison is made with the Nakagawa, A face etch system which is in common use to characterize this family of materials. A screening protocol is discussed which integrates the use of etch pitting into the manufacture of substrates for use in epitaxial MCT applications. Comparisons are made between CdZnTe substrates grown using the horizontal and vertical Bridgman techniques.

Impurites were tracked from raw material purification through to CdZnTe processing in an effort to identify the sources of elements which impact on IR photodetector performance. Chemical analyses by GDMS and ZCGFAA effectively showed the levels of impurities introduced into CdZnTe substrate material from the manufacturing processes. A new purification process (ISDZR) for raw materials was developed, resulting in improved CdZnTe substrate purity. Substrate CU contamination was found to have detrimental effects on LPE layer and device electrical properties for lightly doped HgCdTe.

Gold-doped Hg1−xCdxTe samples of x=0.2067 (in the very-long-wavelength infrared spectral band, with cutoff wavelengths ∼13.2 μm at 77 K) were prepared by tellurium-melt liquid-phase epitaxy. The samples were doped with indium to ∼2×1014 cm−3 and gold to ∼7×1015 cm−3, and were characterized by secondary ion mass spectroscopy, Hall measurements, and minority carrier lifetime measurements. State-of-the-art minority carrier lifetime of ∼0.82 μs was obtained.

A study of the electrical properties of p-type, low-arsenic-doped Hg1−xCdxTe (x∼0.3, carrier concentration ∼1×10+15 cm−3) suitable for fabrication of high-operating-temperature photodetectors was carried out. The arsenic-doped HgCdTe samples were prepared by molecular beam epitaxy using an elemental arsenic source and were characterized by Hall measurements at 77 K after the samples were subjected to four different arsenic activation annealing schemes with annealing temperatures ranging from 300 to 450 °C. For comparison purpose, a sample doped to low-10+16 cm−3 was also prepared and subjected to the same annealing schemes. Although the four annealing schemes had little influence on the Hall data of the 10+16 cm−3 sample, they had significant impact on the Hall data of the 10+15 cm−3 sample. Furthermore, it was found that the Hall data could not be explained by any scattering mechanisms but could be satisfactorily modeled by a two-layer model in which an n-type skin layer is included.

The High Operating Temperature Auger suppressed infrared detector concept is being pursued using the high density vertically integrated photodiode (HDVIP ® ) architecture and an n + -p device structure. Dark current densities as low as 2.5 mA/cm 2 normalized to a 5 µm cutoff at 250K have been demonstrated on these diodes. These dark currents imply minority carrier lifetimes in excess of 300µsec. 1/f noise in these devices arises from the tunneling of charge into the passivation interface, giving rise to a modulation in the surface positive charge and hence to the width of the depletion region in the p-side of the device and a modulation in the total dark current. The measured 1/f noise is in agreement with the predictions of this model, with very low noise being observed when the lifetimes are high. Keywords: Mercury cadmium telluride, infrared detector, focal plane array, high operating temperature, HOT 1. INTRODUCTION 1.1 Motivation In the past, most photon detecting infrared Focal Plane Arrays (FPAs) have operated at a temperature of 77K or below. There is a considerable system advantage if the operating temperature of the FPA can be increased, namely that the resulting Dewar and Cooler package into which the FPA is integrated will be considerably lighter, smaller, and thus cheaper. In addition the power required to cool the FPA can be decrease by a factor of ~5-10, as illustrated in Figure 1. Implementation of this technology enables new system concepts and functionality.

Measurements have been performed of the carrier concentrations in vacancy-doped Hg1−xCdxTe with x=0.22, 0.29, 0.45, and 0.5. Anneals to establish the carrier concentrations were performed on both the mercury- and tellurium-rich sides of the phase field. When these results were added to earlier data for x=0.2 and 0.4, and assuming that all vacancies are doubly ionized, then vacancy concentrations for all values of x and anneal temperature can be represented by simple equations. On the mercury side of the phase field, the vacancy concentrations varied as 2.50×1023(1−x) exp[−1.00/kT] for low concentrations, and as 3.97×107(1−x)1/3ni2/3 exp[−0.33/kT] for high concentrations, where ni is the intrinsic carrier concentration. On the tellurium rich side, the vacancy concentrations varied as 2.81 × 1022(1−x) exp[−0.65/kT] for low concentrations and as 1.92×107(1−x)1/3ni2/3 exp[−0.22/kT] for high concentrations.

An experimental investigation of the artificially structured In x Ga 1−x Sb/InAs strained layer superlattice (SLS) system proposed by Mailhiot and Smith in 1987 for long-wavelength infrared detection was carried out. The SLS structures were grown on GaSb(100) substrates using MBE and were characterized by reflection high-energy electron diffraction, Auger electron spectroscopy, double- crystal X-ray diffraction, Roman spectroscopy, transmission electron microscopy and Fourier transform infrared spectroscopy. Far infrared absorption was observed at a wavelength of 12.6 μm in a 25-period, 150 A/150 A [In 0.2 Ga 0.8 Sb/InAs]‖GaSb(100) sample, thereby demonstrating the feasibility of this material system for long-wavelength infrared detection.

An overview of the DRS HDVIP architecture for realization of large-area infrared focal plane arrays (IRFPAs) is given. Improvements needed to meet more stringent application requirements are discussed and modeled. Both theoretical and experimental data are presented.

An overview on DRS' approaches towards realization of HgCdTe photonic infrared detectors based on DRS's proven HDVIP technology is given. The first approach involves the use of a silicon microlens array attached to the detector array, and the second reduction of dark currents in each detector itself. Recent progress is presented.

We report on results of laboratory and field tests of dual- band MWIR/LWIR focal plane arrays (FPAs) produced under the Army Research Laboratory's Multidomain Smart Sensor Federated Laboratory program. The FPAs were made by DRS Infrared Technologies using the HgCdTe material system and by BAE Systems using QWIP technology. The HgCdTe array used the DRS HDVIPTM process to bond two single-color detector structures to a 640 X 480-pixel single-color read-out integrated circuit (ROIC) to produce a dual-band 320 X 240 pixel array. The MWIR and LWIR pixels are co-located and have a high fill factor. The images from each band may be read out either sequentially (alternating frames) or simultaneously. The alternating frame approach must be used to produce optimal imagery in both bands under normal background conditions. The QWIP FPA was produced using MBE-grown III-V materials. The LWIR section consisted of GaAs quantum wells and AlGaAs barriers and the MWIR section used InGaAs quantum wells with AlGaAs barriers. The detector arrays were processed with three ohmic contacts for each pixel allowing for independent bias control over both the MWIR and LWIR sections. The arrays were indium bump-bonded to an ROIC (specifically designed for two color operation) which puts out the imagery from both bands simultaneously. The ROIC has variable gain and windowing capabilities. Both FPAs were tested under similar ambient conditions with similar optical components. The FPAs were subjected to a standard series of laboratory performance tests. The relative advantages and disadvantages of the two material systems for producing medium-format dual-band FPAs are discussed.

The optical transfer matrix formulation was applied to the analysis of thin solid film multilayer structures used in infrared detection device technology. Our analysis is similar to optimization techniques applied to optical multilayer coatings, i.e., performing simulation of spectra and obtaining best fits to experimental spectral data (targets). We present, as a example, the application of the transfer matrix method to determine the thickness of HgCdTe films attached with epoxy to CdZnTe carriers, using normal incidence reflectance spectra. MID Infrared reflection spectra were obtained using a NICOLET Magna IR 750 Fourier Transform Infrared Spectrometer, equipped with a SPECTRATECH, Inc., microscope, taking care to operate in the linear response range of the detector. The obtained thickness values of HgCdTe films and epoxy layers agree very well with those measured by photographic techniques.

Liquid phase epitaxy(LPE) of (Hg,Cd)Te thick films from large tellurium-rich solutions was used to produce very large quantities of high quality films used for the fabrication of focal plane arrays. The composition uniformity is found to be ix=±O.OO1 for x=O.223 over an area ranging from 43cm2 to 54cm2. The reproducibility for 2OO growth runs

A 1/f noise model is developed for reverse biased diodes based on McWhorter’s concept of charge tunneling into semiconductor states at passivation layer interfaces [A. L. McWhorter, in Semiconductor Surface Physics, edited by R. H. Kingston (University of Pennsylvania Press, Philadelphia, 1957), pp. 207–228]. The charge modulates the width of semiconductor surface charge layers on either side of the junction, resulting in fluctuations in dark current from these volumes due to the net difference in depletion and diffusion current generation rates per unit volume in the semiconductor. The 1/f spectrum associated with the fluctuating surface charge translates into a 1/f spectrum in thermally generated diode dark current. The model is applied to midwavelength infrared HgCdTe N+/P diode architectures.

HgCdTe heterostructures consisting of a thin n-type widegap (250 meV or 5 μm cutoff) layer deposited on an n-type narrowgap (100-125 meV or 10-13 μm cutoff) layer offer the promise of very high performance metal-insulator-semiconductor (MIS) photocapacitors for long wavelength infrared (LWIR) detection. Molecular Beam Epitaxy (MBE) is a candidate growth technology for these two layer films due to its fine control in composition, thickness and doping concentration. The critical materials issues are reducing the defect content associated with twins in the grown layers, achieving low net donor concentrations in the widegap layer, and avoiding the formation of misfit dislocations at the HgCdTe heterointerface. This paper will report on our recent progress in these directions.

We report on an investigation into the molecular‐beam epitaxial growth and characterization of InAs/InxGa1−xSb (0≤x≤0.4) heterojunctions and strained‐layer superlattices (SLS) for long‐wavelength infrared detector applications. All structures have been grown on GaSb(100) substrates with period thicknesses of the SLS in the range 25–180 A. Interface properties of single heterojunction structures have been characterized by x‐ray photoelectron spectroscopy and secondary‐ion mass spectroscopy depth profiles, and structural properties of the SLS have been assessed by transmission electron microscopy and x‐ray diffraction analysis. For growth at an optimal temperature (390 °C) the SLS are of high structural quality, with no evidence of compound formation at either (GaSb/InAs, InAs/GaSb) interface, or of post‐growth interdiffusion. However, significant levels of As (7%) are found to be present in nominally grown InxGa1−xSb layers, resulting from incorporation from background. Fourier transform infrared absorptio...

We present recent progress in the development of metal‐insulator‐semiconductor (MIS) heterostructure detectors grown by molecular beam epitaxy (MBE); these heterostructure films should significantly improve the available well capacity for MIS long wavelength infrared (LWIR) detectors. Recent MBE grown HgCdTe(112)Te heterostructures show a 4× increase in available charge capacity and a 2× decrease in net donor density relative to our MBE HgCdTe(001) layers. To obtain full advantage of the MIS heterostructure however, another 2.5×increase in the MWIR breakdown field (0.8–2.0 V/μm) and a 2× decrease in the MWIR donor density (1×1015–5×1014 cm−3) is necessary. Since the MWIR breakdown field does not appear to be limited by compositional micro‐inhomogeneities or by compound twin defects in the MBE HgCdTe layers, we suspect either dopant micro‐inhomogeneities or native point defects are responsible for the significant tunnel currents measured in our MWIR MBE HgCdTe.

The precipitation of tellurium in alloys of Hg1−xCdxTe with x = 0.2 to 0.3 prepared under conditions of excess tellurium by the solid state recrystallization method have been studied using chemical defect etching and transmission eloctron microscopy. Precipitation of tellurium occurred during the quench from the recrystallization anneal. During the initial part of the quench, precipitation occurred as a result of nucleation of precipitates on pre-existing dislocations. This precipitation resulted in the formation of dislocation helices with precipitates in the interior of the helix loops. Both monoclinic and trigonal phases of tellurium were found in these precipitate defects, as well as the twinned matrix phase. Later in the quench, an additional nucleation mechanism, either homogeneous or heterogeneous on an impurity occurs resulting in the formation of complex defects in the bulk. During a subsequent post-anneal at temperatures of approximately 300°C, the tellurium precipitates dissolve, and interstitial perfect and faulted dislocation loops are formed. Both the precipitation process and the subsequent post-annealing result in considerable multiplication of dislocations.

A new gettering mechanism is proposed for substitutional impurities which diffuse by an interstitial process. In this mechanism, an externally imposed gradient of self interstitials generates a gradient in impurity interstitials leading to the segregation of fast diffusing impurities to low interstitials, high vacancy regions. Data are presented to support this model in (Hg, Cd)Te alloys, as well as explicitly rule out gettering by dislocations, by segregation to precipitates, or by enhanced solubility arising from the interaction of the impurity with a varying Fermi level.

Resonant tunneling has been demonstrated through a variety of molecular‐beam epitaxially grown HgCdTe heterostructures. Single‐quantum‐well, double‐barrier resonant tunneling structures with a large variation in quantum well widths and barrier thicknesses were investigated. We also present results on transport through two strongly coupled superlattices, which exhibit negative differential resistance due to energy filtering by the superlattice minibands.

The surface kinetics that govern molecular‐beam epitaxial (MBE) growth of HgCdTe films places strict boundaries on the optimal deposition conditions to be selected. This paper will examine the limits imposed by the surface kinetics on the Hg flux, substrate orientation, and substrate temperature to be chosen for MBE HgCdTe film growth. The range of Hg flux available to grow single crystalline HgCdTe (as determined by reflection high‐energy electron diffraction) will be seen to vary significantly with the (112)Te, (112)Cd, and (001) growth orientations. The HgCdTe(112)Te films are deposited with a high Hg condensation coefficient at 200 °C (0.04–0.10), have an excellent surface morphology, and avoid the cellular twins that are found in HgCdTe(111)Te layers. However, each of the three orientations result in faceted growth at substrate temperatures near 200 °C; at this substrate temperature, HgTe begins to desorb from the growth surface which leads to a preferential exposure of the lower energy (111) planes....

Several results of our experimental program studying the relationship between the properties of the silicon starting material and N‐channel MOS dynamic RAM circuit performance is described. The thermal history and oxygen and carbon concentrations control the silicon material properties and much of the subsequent circuit performance. The beneficial effect of intrinsic gettering due to precipitation of oxygen in the bulk on MOS dynamic RAM refresh performance will be shown. Equivalent thermal processes will result in similar circuit performance only when the silicon material is equivalent. Some circumstantial evidence is also presented for a degrading effect of high concentrations of carbon in the starting material.

Brief summaries of the precipitation of tellurium in (Hg,Cd)Te during the quench from the recrystallization anneal of the solid state recrystallization process and the subsequent dissolution of these precipitates during the postanneal are given. It is shown that if vacancies and interstitials are considered to be the only important point defects in the lattice, the observed mechanisms of precipitation and annihilation require that tellurium in excess of stoichiometry must be accommodated by metal vacancies. Diffusion from the mercury ambient during postanneal then must occur by metal interstitials. Enhanced self‐diffusion on the metal sublattice is shown to occur along grain and subgrain boundaries. The most significant source of dark current in CID imagers fabricated on (Hg,Cd)Te is shown to be a sufficiently high dislocation density in a pixel.

The existence region of the Hg0.8Cd0.2Te phase field has been re-examined using three techniques: Hall measurements, p-n junction depth measurements on material which had been annealed to produce an n-type skin following equilibration at high temperature, and quantitative modelling of experimental observations of gettering of silver in (Hg,Cd)Te. The results we obtain are generally in good agreement with previously reported measurements, although some systematic disagreement is found, particularly under tellurium saturated conditions for temperatures less than 450‡C.

Junction profiles of B+‐implanted Hg1−xCdxTe are examined using reverse bias capacitance–voltage measurements. All devices are fabricated on residual copper‐doped, low vacancy concentration, p‐type bulk material with a net doping of 1.5×1015 cm−3. The junction profiles showed the as‐implanted junctions to be abrupt with one side of the junction doped in the low‐ to mid‐1014 cm−3 range. These data and related circumstantial evidence are consistent with an n−/p junction after implantation. Type conversion in the immediate vicinity of the junction is postulated to result from indiffusion of interstitial Hg that is created by implant damage, the annihilation of group‐II site vacancies by the Hg, and the associated movement of the highly mobile copper acceptor away from the low vacancy concentration region via an interstitial gettering mechanism. An immobile background donor concentration produces the n− region in the copper denuded region. During thermal anneal treatments, grading of the junction and improvem...

In this paper we show that anodic oxide/Hg0.8Cd0.2Te interface is a two‐layered structure with a thin CdTeO3 layer next to the Hg0.8Cd0.2Te substrate and a second thicker layer of CdTeO3, TeO2, HgTeO3, and HgTe particles. For a 700‐A‐thick anodic oxide, the thin CdTeO3 layer was 30–50 A thick and the second interface layer was ∼120–150 A thick. The large positive fixed charge measured on metal‐insulator semiconductor devices which have been fabricated with the anodic oxide may be caused by a large concentration of positively charged oxygen vacancies associated with the HgTe particles in the second interface layer. The low interface trap densities are probably due to the thin CdTeO3 layer which forms a good interface with HgCdTe.

Precipitation of tellurium in solid (Hg,Cd)Te has been investigated. Review of phase equilibria data and review of conditions required for solid‐state precipitation demonstrates that the conditions necessary for precipitation exist along the tellurium‐saturated phase limit for (Hg,Cd)Te. Metallographic techniques have been used to demonstrate the existence of precipitate particles up to a density of 1012 cm−3 in solid‐state recrystallized (Hg,Cd)Te. The effects of such particles on the properties of (Hg,Cd)Te are discussed. It is shown that remnants of the Te precipitate structure would be expected to persist in post‐annealed (Hg,Cd)Te.

The angular misorientation of GaAs epitaxial layers grown on silicon substrates by molecular beam epitaxy has been measured using x-ray diffraction. A significant misorientation, or tilt, between the epitaxial layer and the substrate has been observed. The magnitude of the tilt depends on the initial substrate orientation, the silicon substrate type (float zone or Czochralski), postgrowth annealing, and epitaxial layer thickness. In almost all cases, the sense of the tilt is such that the GaAs 〈001〉 lies between the surface normal and the silicon 〈001〉. While the presence of interfacial dislocations with Burgers vectors that are approximately parallel to the heterointerface does predict a tilt between the substrate and the epitaxial layer, the sense of the tilt that arises from these dislocations is opposite to that observed experimentally. A model, based on the relief of misfit by dislocations inclined approximately 45 degrees to the interface, is proposed that correctly describes the observed tilt.

HgTe grown by MBE onto CdTe substrates under excess Hg conditions was found to lead to a columnar void-twin microstructure, with the twins nucleating from the voids. Grown under Hg-deficient conditions, the microstructure of MBE HgTe consisted of columnar polycrystals. Nucleation of the polycrystals occurred only at the commencement of growth. Possible growth and twinning mechanisms are discussed.

An interstitial model is developed to quantitatively account for the anomalous effects which accompany the diffusion of phosphorus in silicon from high surface concentrations. Phosphorus is assumed to diffuse as both a positively charged (intrinsic diffusion) and an uncharged interstitial (high concentration case). Silicon is assumed to diffuse as both an uncharged (instrinsic) and a negatively charged interstitial (high phosphorus concentrations). The effect of internal fields on charged species is included. An excellent fit is found with published experimental data if a surface enhancement factor, analogous to that for the case of oxidation enhancement, is used. Flat-top diffusion is modelled by assuming the formation of a dislocation array, which alters the point defect equilibria.

Many methods for the preparation of (Hg,Cd)Te alloys rely on a low temperature processing step to convert the as‐grown p‐type material to n‐type, or to otherwise adjust the concentration of native acceptors. During this anneal, tellurium precipitates in the material are annihilated by in‐diffusing mercury, resulting in a substantial multiplication of dislocations. For substantially long anneals (>1 day at 270 °C) the depth of the p–n junction is found to vary as the square root of the anneal time and inversely as the square root of the excess tellurium concentration. Rapidly diffusing impurities such as silver are gettered out of the skin and into the remaining vacancy‐rich core. The kinetics of these processes are analyzed for self‐diffusion on the metal sublattice involving only vacancies, only interstitials, and for a mixed vacancy–interstitial model. Comparison with experimental data shows best agreement with the mixed interstitial–vacancy model.