CdTe/ZnTe/GaAs Heterostructures for Single-Crystal CdTe Solar Cells

Amorphous CdS/single-crystal CdTe solar cells were grown on GaAs substrates by metalorganic chemical vapor deposition. The structures of the films and the electrical properties of the devices were characterized. Highly conducting arsenic-doped ZnTe was grown on GaAs(100) substrates as the buffer layer for CdTe growth. By use of a ～30-nm ZnTe buffer layer, a p-CdTe film with a doping level of ～5×10¹⁶ cm^?3 was achieved. The hole concentration of p-CdTe increased with increasing VI/II ratio under a high As concentration during growth. From temperature-dependent Hall transport measurements, the ionization energy of the As acceptor in the p-CdTe was estimated to be approximately 88 meV. Ohmic behavior of the junctions between CdTe/ZnTe and ZnTe/GaAs was also confirmed. The solar cell performance of this structure, for example an open circuit voltage of 0.63 V, could be improved if the crystal quality of the CdTe film is optimized and the dislocation density of the CdTe film is minimized.     

Tellurium-Rich CdTe,and the Effect of Tellurium Content on the Properties of CdTe

Co-evaporation of CdTe and Te has been reported to result in CdTe films with high hole concentrations. Higher carrier density should result in more efficient solar cells if the carrier lifetime is not effected. This achievement could have a large effect on CdTe technology, in which carrier density has been limited to the 10¹⁴–10¹⁵ cm^?3 range. Reproducing the work from the open literature and analyzing the films in more detail revealed that material with a high hole concentration can be obtained by co-evaporating CdTe and Te. However, analysis of these films indicated that the measured high carrier density is not because of doping of the CdTe base material but because of an integrated network of Te present as its own phase within the CdTe matrix.     

Iodine Doping of CdTe and CdMgTe for Photovoltaic Applications

Iodine-doped CdTe and Cd_1?x Mg _x Te layers were grown by molecular beam epitaxy. Secondary ion mass spectrometry characterization was used to measure dopant concentration, while Hall measurement was used for determining carrier concentration. Photoluminescence intensity and time-resolved photoluminescence techniques were used for optical characterization. Maximum n-type carrier concentrations of 7.4 × 10¹⁸ cm^?3 for CdTe and 3 × 10¹⁷ cm^?3 for Cd_0.65Mg_0.35Te were achieved. Studies suggest that electrically active doping with iodine is limited with dopant concentration much above these values. Dopant activation of about 80% was observed in most of the CdTe samples. The estimated activation energy is about 6 meV for CdTe and the value for Cd_0.65Mg_0.35Te is about 58 meV. Iodine-doped samples exhibit long lifetimes with no evidence of photoluminescence degradation with doping as high as 2 × 10¹⁸ cm^?3, while indium shows substantial non-radiative recombination at carrier concentrations above 5 × 10¹⁶ cm^?3. Iodine was shown to be thermally stable in CdTe at temperatures up to 600°C. Results suggest iodine may be a preferred n-type dopant compared to indium in achieving heavily doped n-type CdTe.     

Strain Reduction in Selectively Grown CdTe by MBE on Nanopatterned Silicon on Insulator (SOI) Substrates

Silicon-based substrates for the epitaxy of HgCdTe are an attractive low-cost choice for monolithic integration of infrared detectors with mature Si technology and high yield. However, progress in heteroepitaxy of CdTe/Si (for subsequent growth of HgCdTe) is limited by the high lattice and thermal mismatch, which creates strain at the heterointerface that results in a high density of dislocations. Previously we have reported on theoretical modeling of strain partitioning between CdTe and Si on nanopatterned silicon on insulator (SOI) substrates. In this paper, we present an experimental study of CdTe epitaxy on nanopatterned (SOI). SOI (100) substrates were patterned with interferometric lithography and reactive ion etching to form a two-dimensional array of silicon pillars with ～250 nm diameter and 1 μm pitch. MBE was used to grow CdTe selectively on the silicon nanopillars. Selective growth of CdTe was confirmed by scanning electron microscopy (SEM), atomic force microscopy (AFM), and X-ray photoelectron spectroscopy (XPS). Coalescence of CdTe on the silicon nanoislands has been observed from the SEM characterization. Selective growth was achieved with a two-step growth process involving desorption of the nucleation layer followed by regrowth of CdTe at a rate of 0.2 Å s^?1. Strain measurements by Raman spectroscopy show a comparable Raman shift (2.7 ± 2 cm^?1 from the bulk value of 170 cm^?1) in CdTe grown on nanopatterned SOI and planar silicon (Raman shift of 4.4 ± 2 cm^?1), indicating similar strain on the nanopatterned substrates.     

Effect of features of the technology of polycrystalline CdTe growth on the conductivity and deep level spectrum after annealing

The conductivity, morphology, and deep levels in polycrystalline CdTe are studied. Undoped p-CdTe is grown from the vapor phase by low-temperature methods of direct Cd and Te chemical reaction and CdTe vacuum sublimation at P _min. Chlorine-doped CdTe is also grown. The resistivity of the grown samples is ～105–109 Ω cm. After annealing in liquid cadmium or in cadmium vapor at ～500°C, the conductivity type changes, the free-carrier concentration in the undoped and doped samples increases to 4 × 10¹⁵ and ～2 × 10¹⁶ cm^?3, respectively. For all samples, a defect ground level of ～0.84 eV and continuous background are observed in DLTS spectra after annealing. A correlation between the primary-defect and free-carrier concentrations in undoped and doped CdTe is observed. Chlorine is a main residual impurity in the undoped samples. It is assumed that the defect is a complex including chlorine and observed structural defects in CdTe.     

MBE growth and characterization of in situ arsenic doped HgCdTe

We report the results of in situ arsenic doping by molecular beam epitaxy using an elemental arsenic source. Single Hg_1−xCd_xTe layers of x ∼0.3 were grown at a lower growth temperature of 175°C to increase the arsenic incorporation into the layers. Layers grown at 175°C have shown typical etch pit densities of 2E6 with achievable densities as low as 7E4cm⁻². Void defect densities can routinely be achieved at levels below 1000 cm⁻². Double crystal x-ray diffraction rocking curves exhibit typical full width at half-maximum values of 23 arcsec indicating high structural quality. Arsenic incorporation into the HgCdTe layers was confirmed using secondary ion mass spectrometry. Isothermal annealing of HgCdTe:As layers at temperatures of either 436 or 300°C results in activation of the arsenic at concentrations ranging from 2E16 to 2E18 cm⁻³. Theoretical fits to variable temperature Hall measurements indicate that layers are not compensated, with near 100% activation after isothermal anneals at 436 or 300°C. Arsenic activation energies and 77K minority carrier lifetime measurements are consistent with published literature values. SIMS analyses of annealed arsenic doping profiles confirm a low arsenic diffusion coefficient.     

On the behavior of Bi in a CdTe lattice and the compensation effect in CdTe:Bi

CdTe crystals of two types have been grown by the vertical Bridgman method: (i) crystals doped with Bi to ～10¹⁸ cm^?3 and (ii) double-doped (Bi + Cl) crystals with a Bi concentration of ～10¹⁸ cm^?3 and a Cl concentration of ～10¹⁷ cm^?3. The temperature dependences of the resistivity, photoconductivity, and low-temperature photoluminescence are investigated for the crystals grown. Analysis has shown that doping with Bi (crystals of the first type) leads to compensation of the material. The resistivity of the CdTe:Bi samples at room temperature, depending on the doping level, is varied in the range of 10⁵–10⁹ Ω cm. The hole concentration is determined by the acceptor level at E _v + 0.4 eV in lightly doped CdTe:Bi samples and by the deep center at E _v + 0.72 eV in heavily doped CdTe:Bi samples. Double doping leads to inversion of the conductivity type and reduces the resistivity to ～1 Ω cm. Heavily doped CdTe:Bi crystals and double-doped crystals exhibit the presence of acceptors with an ionization energy of 36 meV, which is atypical of CdTe.     

Novel Cu Nanowires/Graphene as the Back Contact for CdTe Solar Cells

Cu‐nanowire‐doped graphene (Cu NWs/graphene) is successfully incorporated as the back contact in thin‐film CdTe solar cells. 1D, single‐crystal Cu nanowires (NWs) are prepared by a hydrothermal method at 160 °C and 3D, highly crystalline graphene is obtained by ambient‐pressure CVD at 1000 °C. The Cu NWs/graphene back contact is obtained from fully mixing the Cu nanowires and graphene with poly(vinylidene fluoride) (PVDF) and N‐methyl pyrrolidinone (NMP), and then annealing at 185 °C for solidification. The back contact possesses a high electrical conductivity of 16.7 S cm^?1 and a carrier mobility of 16.2 cm² V^?1 s^?1. The efficiency of solar cells with Cu NWs/graphene achieved is up to 12.1%, higher than that of cells with traditional back contacts using Cu‐particle‐doped graphite (10.5%) or Cu thin films (9.1%). This indicates that the Cu NWs/graphene back contact improves the hole collection ability of CdTe cells due to the percolating network, with the super‐high aspect ratio of the Cu nanowires offering enormous electrical transport routes to connect the individual graphene sheets. The cells with Cu NWs/graphene also exhibit an excellent thermal stability, because they can supply an active Cu diffusion source to form an stable intermediate layer of CuTe between the CdTe layer and the back contact.     

Defects Study in Hg x Cd1−x Te Infrared Photodetectors by Deep Level Transient Spectroscopy

At high temperature, infra-red focal plane arrays are limited by their performance in operability, detectivity D ^* or noise equivalent temperature difference. Trap characterization and defect studies are necessary to better understand these limitations at high temperature. In this paper, we use deep level transient spectroscopy to study electrically active defects in mercury cadmium telluride n ⁺/p diodes. The material investigated has a cut-off frequency (λ _c) of 2.5 μm at 180 K and p doping performed with mercury vacancy. Trap energy signatures as well as capture cross-section measurements are detailed. A low temperature hole trap close to midgap is observed in the range 150–200 K with an activation energy around 0.18 ± 0.025 eV. A high temperature hole trap is also observed in the range 240–300 K with an activation energy of 0.68 ± 0.06 eV. A hole capture cross-section of 10^?19 cm² is obtained for both traps. The nature of the defects and their correlation with dark current are discussed.     

Effect of the arsenic cracking zone temperature on the efficiency of arsenic incorporation in CdHgTe films in molecular-beam epitaxy

Cd _x Hg_{1 ? x} Te films with x ≈ 0.22 and thickness of ～10 µm have been grown by molecular-beam epitaxy on gallium arsenide substrates and doped in situ with arsenic. Activation annealing of doped films provided p-type conduction with a hole density of up to 10¹⁷ cm^?3. The influence exerted by the arsenic cracking zone temperature on the efficiency of arsenic incorporation into the CdHgTe film was studied. A model describing the dependence of the arsenic concentration in the films on the arsenic cracking zone temperature was suggested. A comparison of the model and the experimental data demonstrated that the incorporation efficiency of diatomic arsenic is approximately two orders of magnitude higher than that of tetratomic arsenic.     

Study of deep-level defects and annealing effects in undoped and Sn-doped GaAs solar cells irradiated by one-MeV electrons

DLTS and C-V techniques have been employed to determine the defect energy levels and density, carrier capture cross sections, lifetimes and diffusion lengths in the Sn-doped and the undoped GaAs solar cells irradiated by one-MeV electrons under different electron fluences (10¹⁴ to 10¹⁶ cm^?2), fluxes (2 × 10⁹, 4 × 10¹⁰ e/cm²-s), and annealing conditions (150 ? T ? 230°C). The results show that density of both electron and hole traps will in general increase with incresing electron fluence and flux, and decrease with increasing annealing temperature and annealing time. Some distinct difference in defeat spectrum was observed in the undoped and the Sn-doped GaAs solar cells studied. The low temperature thermal annealing and the recombination enhanced annealing processes are found to be very effective in reducing the density of deep-level defects induced by one-MeV electrons. The results of our findings are discussed in detail in this paper.     

Specific features of transmutational doping of 30Si-enriched silicon crystals with phosphorus: Studies by the method of electron spin resonance

Electron spin resonance (ESR) is used to study the neutron transmutation doping of silicon crystals enriched with ³⁰Si isotope: phosphorus donors and radiation defects produced in the course of transmutational doping are observed. The ESR signals related to the phosphorus uncontrolled impurity in ³⁰Si before transmutational doping (the P concentration is ～10¹⁵ cm^?3) and phosphorus introduced by neutron irradiation with doses ～1 × 10¹⁹ cm^?2 and ～1 × 10²⁰ cm^?2 (the P concentrations are ～5 × 10¹⁶ and ～7 × 10¹⁷ cm^?3, respectively) are studied. As a result of drastic narrowing of the phosphorus ESR lines in ³⁰Si, the intensity of lines increased appreciably, which made it possible to measure the phosphorus concentration in the samples with a small volume (down to 10^?6 mm^?3). The methods for determining the concentration of P donors from hyperfine structure in the ESR spectra of isolated P atoms, exchange-related pairs, and clusters that consist of three, four, and more P donors are developed. In the region of high concentrations of P donors, in which case the hyperfine structure disappears, the concentration of P donors was estimated from the exchange-narrowed ESR line.     

Growth of tin-doped indium antimonide for magnetoresistors

Magnetoresistors made from n-type indium antimonide are of interest for magnetic position sensing applications. In this study, tin-doped indium antimonide was grown by the metalorganic chemical vapor deposition technique using trimethylindium, trisdimethylaminoantimony, and tetraethyltin in a hydrogen ambient. Using a growth temperature of 370°C and a pressure of 200 Torr, it was found that the electron density in tin-doped films varied from 3.3×10¹⁶ cm⁻³ to 4.0×10¹⁷ cm⁻³ as the 5/3 ratio was varied from 4.8 to 6.8. From secondary ion mass spectroscopy (SIMS) studies, it was found that this variation is not caused by a change in site occupancy of the tin atoms from antimony to indium lattice sites, but rather to a change in the total tin concentration incorporated into the films. This dependence of tin incorporation on stoichiometry could be used to rapidly vary the doping level during growth. Undoped films grown under similar conditions had electron densities of about 2×10¹⁶ cm⁻³ and electron mobilities near 50,000 cm²V⁻¹s⁻¹ at room temperature for films that were only 1.5 μm thick on a gallium arsenide substrate. Attempts to grow indium antimonide at 280°C resulted in p-type material caused by carbon incorporation. The carbon concentration as measured with SIMS increased rapidly with increasing growth rate, to above 10¹⁹ cm⁻³ at 0.25 μm/h. This is apparently caused by incomplete pyrolysis of a reactant at this low growth temperature. Growth at 420°C resulted in rough surface morphologies. Finally, it was demonstrated that films with excellent electron mobility and an optimized doping profile for magnetoresistors can be grown.     

Hall Effect Studies of AlGaAs Grown by Liquid-Phase Epitaxy for Tandem Solar Cell Applications

We report results from Hall effect studies on Al _x Ga_1?x As (x = 0.23–0.24) with bandgap energies of 1.76 ± 0.01 eV grown by liquid-phase epitaxy (LPE). Room-temperature Hall measurements on unintentionally doped AlGaAs revealed p-type background doping for concentrations in the range 3.7–5.2 × 10¹⁶ cm^?3. Sn, Te, Ge, and Zn-doped AlGaAs were also characterized to study the relationship between doping concentrations and the atomic fractions of the dopants in the melt. Temperature-dependent Hall measurements were performed to determine the activation energies of the four dopants. Deep donor levels (DX centers) were dominant for Sn-doped Al_0.24Ga_0.76As, but not for Te-doped Al_0.24Ga_0.76As. Comparison of the temperature-dependent Hall effect results for unintentionally and intentionally doped Al_0.24Ga_0.76As indicated that the impurity contributing to the p-type background doping had the same activation energy as Mg. We thus suggest a Te-doped emitter and an undoped or Ge-doped base to maximize the efficiency of Al _x Ga_1?x As (x ～ 0.23) solar cells grown by LPE.     

P-type doping of GaAs by carbon implantation

The electrical properties of C-implanted <100> GaAs have been studied following rapid thermal annealing at temperatures in the range from 750 to 950°C. This includes dopant profiling using differential Hall measurements. The maximum p-type activation efficiency was found to be a function of C-dose and annealing temperature, with the optimum annealing temperature varying from 900°C for C doses of 5 × 10¹³ cm⁻² to 800°C for doses ≥5 × 10¹⁴cm⁻². For low dose implants, the net p-type activation efficiency was as high as 75%; while for the highest dose implants, it dropped to as low as 0.5%. Moreover, for these high-dose samples, 5 × 10¹⁵ cm⁻², the activation efficiency was found to decrease with increasing annealing temperature, for temperatures above ∼800°C, and the net hole concentration fell below that of samples implanted to lower doses. This issue is discussed in terms of the amphoteric doping behavior of C in GaAs. Hole mobilities showed little dependence on annealing temperature but decreased with increasing implant dose, ranging from ∼100 cm²/V·s for low dose implants, to ∼65 cm²/V·s for high dose samples. These mobility values are the same or higher than those for Be-, Zn-, or Cd-implanted GaAs.     

GaSb: A New Alternative Substrate for Epitaxial Growth of HgCdTe

In this work, GaSb is proposed as a new alternative substrate for the growth of HgCdTe via molecular beam epitaxy (MBE). Due to the smaller mismatch in both lattice constant and coefficient of thermal expansion between GaSb and HgCdTe, GaSb presents a better alternative substrate for the epitaxial growth of HgCdTe, in comparison to alternative substrates such as Si, Ge, and GaAs. In our recent efforts, a CdTe buffer layer technology has been developed on GaSb substrates via MBE. By optimizing the growth conditions (mainly growth temperature and VI/II flux ratio), CdTe buffer layers have been grown on GaSb substrates with material quality comparable to, and slightly better than, CdTe buffer layers grown on GaAs substrates, which is one of the state-of-the-art alternative substrates used in growing HgCdTe for the fabrication of mid-wave infrared detectors. The results presented in this paper indicate the great potential of GaSb to become the next generation alternative substrate for HgCdTe infrared detectors, demonstrating MBE-grown CdTe buffer layers with rocking curve (double crystal x-ray diffraction) full width at half maximum of ～60 arcsec and etch pit density of ～10⁶ cm^?2.     

Growth of fully doped Hg1−xCdxTe heterostructures using a novel iodine doping source to achieve improved device performance at elevated temperatures

Band gap engineered Hg_1−xCd_xTe (MCT) heterostructures should lead to detectors with improved electro-optic and radiometric performance at elevated operating temperatures. Growth of such structures was accomplished using metalorganic vapor phase epitaxy (MOVPE). Acceptor doping with arsenic (As), using phenylarsine (PhAsH₂), demonstrated 100% activation and reproducible control over a wide range of concentrations (1 × 10¹⁵ to 3.5 × 10¹⁷ cm⁻³). Although vapor from elemental iodine showed the suitability of iodine as a donor in MC.T, problems arose while controlling low donor concentrations. Initial studies using ethyliodide (EtI) demonstrated that this source could be used successfully to dope MCT, yielding the properties required for stable heterostructure devices, i.e. ≈100% activation, no memory problems and low diffusion coefficient. Cryogenic alkyl cooling or very high dilution factors were required to achieve the concentrations needed for donor doping below ≈10¹⁶cm⁻³ due to the high vapor pressure of the alkyl. A study of an alternative organic iodide source, 2-methylpropyliodide (2 MePrI), which has a much lower vapor pressure, improved control of low donor concentrations. 2 MePrI demonstrated the same donor source suitability as EtI and was used to control iodine concentrations from ≈ 1 × 10¹⁵ to 5 × 10¹⁷cm⁻³. The iodine from both sources only incorporated during the CdTe cycles of the interdiffused multilayer process (IMP) in a similar manner to both elemental iodine and As from PhAsH₂. High resolution secondary ion mass spectroscopy analysis showed that IMP scale modulations can still be identified after growth. The magnitude of these oscillations is consistent with a diffusion coefficient of≈7 × 10⁻¹⁶cm²s⁻¹ for iodine in MC.T at 365°C. Extrinsically doped device heterostructures, grown using 2 MePrI, have been intended to operate at elevated temperatures either for long wavelength (8–12 smm) equilibrium operation at 145K or nonequilibrium operation at 190 and 295K in both the 3–5 μ and 8–12 μ wavelength ranges. Characterization of such device structures will be discussed. Linear arrays of mesa devices have been fabricated in these layers. Medium wave nonequilibrium device structures have demonstrated high quantum efficiencies and R₀A = 37 Ωcm² for λ_co = 4.9 μ at 190K.     

Si Doping of GaN in Hydride Vapor-Phase Epitaxy

Growth of GaN boules by hydride vapor-phase epitaxy (HVPE) is very attractive for fabrication of GaN substrates. Use of dichlorosilane as a source for Si doping of bulk GaN is investigated. It is shown that no tensile strain is incorporated into mm-thick, Si-doped GaN layers on sapphire substrates if the threading dislocation density is previously reduced to 2.5 × 10⁷ cm^?2 or below. High-quality GaN layers with electron densities up to 1.5 × 10¹⁹ cm^?3 have been achieved, and an upper limit of about 4 × 10¹⁹ cm^?3 for Si doping of GaN boules was deduced considering the evolution of dislocations with thickness. A 2-inch, Si-doped GaN crystal with length exceeding 6 mm and targeted Si doping of about 1 × 10¹⁸ cm^?3 is demonstrated.     

Carbon doped GaAs grown in low pressure-metalorganic vapor phase epitaxy using carbon tetrabromide

Carbon tetrabromide was used as carbon source for heavily p-doped GaAs in low pressure metalorganic vapor phase epitaxy (MOVPE). The efficiency of carbon incorporation was investigated at temperatures between 550 and 670°C, at V/III ratios from 1 to 50 and carbon tetrabromide partial pressures from 0.01 to 0.03 Pa. Hole concentrations from 8 × 10¹⁷ to 5 × 10¹⁹ cm⁻³ in as-grown layers were obtained. After annealing in nitrogen atmosphere at 450°C, a maximum hole concentration of 9 × 10¹⁹ cm⁻³ and a mobility of 87 cm²/Vs was found. At growth temperatures below 600°C, traces of bromine were detected in the layers. Photoluminescence mapping revealed an excellent doping homogeneity. Thus, CBr₄ is found to be a suitable carbon dopant source in MOVPE.     

Characterization of CdS Nanowires Self-Assembled in a Nanoporous Alumina Template

CdS nanowires were self-assembled in a thin film (～200 nm) anodic aluminum oxide template on an indium tin oxide-coated glass substrate via dc electrodeposition. Raman spectral studies were done to probe the vibrational properties of scattering CdS phonons. Strong 1 longitudinal optical (LO), 2 LO, and 3 LO peaks were observed at 302 cm^?1, 603 cm^?1, and 906 cm^?1 having an energy separation of 37 meV, which is in accordance with the CdS bulk values. The photoluminescence spectra showed improved intensity of emission on annealing of the CdS nanowires. Field-emission scanning microscopy confirms the growth of nanowires of diameters ranging from 10 nm to 25 nm for these templates. These diameters agreed with those extracted from the luminescence emission energies.