Growth Mechanism in Atomic Layer Epitaxy (I) Re-evaporation of Cd and Te from CdTe(111) Surfaces Monitored by Auger Electron Spectroscopy

The mechanism of atomic layer epitaxy (ALE) of cadmium telluride has been studied. Auger electron spectroscopy is used to measure the isothermal re-evaporation rates of elemental Cd and Te deposits on the (111)A and (111)B surfaces of CdTe substrates. The results include an observation that the sticking coefficients of Cd and Te are smaller than unity at the growth temperatures typical of CdTe ALE. After desorption the substrates are left partially covered: 35% by a Cd overlayer on the (111)B surface and 72% by Te on the (111)A surface. The re-evaporation rates of Cd and Te experience a drastic change near the substrate-deposit interface. These rates appear two orders of magnitude smaller than those of bulk-like amorphous Cd and Te solids. The activation energies for reevaporation of the near-interface layer region are estimated to be: 1.5 eV for Te on the (111)A face, 1.0 eV for Te on (111)B and 0.5 eV for Cd on (111)B. It has also been shown that AES can be used to identify the polarity of the CdTe(111) surfaces. The relative difference in peak-to-peak intensity ratios of Cd MNN to Te MNN for (111)A and (111)B is (11 ± 2)%.     

Ultrahigh Vacuum Atomic Layer Epitaxy of Ternary II-VI Semiconductor Compounds

A concise discussion concerning the UHV ALE growth of ternary II-VI compounds is presented in this paper. Simultaneous reflection mass spectrometry (REMS) and reflection high energy electron diffraction (RHEED) measurements of the surface kinetic and structural parameters, respectively, governing the UHV ALE growth of Cd_1-xZn_xTe and Cd_1-xMn_xTe heteroepitaxial films are reported. In addition, a Monte-Carlo-based method for simulation of the UHV ALE process of CdTe (the model-compound for this growth technique) has been used for investigation of the Cd cation's fluxes reflected from the growing epilayer surface in different phases of the ALE process. The Cd⁺ ion-related REMS signals measured during CdTe growth have been compared with the simulation results.     

Growth mechanism in atomic layer epitaxy (III) reevaporation of Cd and Te from CdTe (111) surfaces and thick elemental deposits monitored by quadrupole-mass spectrometry

Quadrupole-mass spectroscopical (QMS) studies on isothermal reevaporation of Cd and Te species from the CdTe (111) surface have been performed in two extreme cases. The first concerns the reevaporation of thick, bulk-like non-crystalline Cd and Te films deposited in high vacuum at room temperature on the (111) surface, whereas in the second case evaporation of the constituent species from the bare single crystalline (111) surface has been investigated. The fluxes of the species desorbing in high vacuum (10⁻⁶ Pa) have been monitored with QMS and the desorption temperatures have been measured with a thermocouple mounted as near as possible to the sample surface. The following values E_a(Cd)_bulk = 1.13 ± 0.12 eV, E_a(Te)_bulk = 1.64 ± 0.18 eV and E_a(Cd)₍₁₁₁₎ = 1.13 ± 0.06 eV, E_a(Te)₍₁₁₁₎ = 1.92 ± 0.13 eV of the activation energies for these two cases have been determined from the slopes of the Arrhenius plots. Using these experimental values, the numbers of atomic bonds N_Cd and N_Te occurring in the atomic aggregates of quasi-gas molecules forming the near surface quasi-gas transition layers have been estimated. For Cd quasi-gas molecules 2 ≤ N_Cd ≤ 5, whereas for Te molecules 3 ≤ N_Te ≤ 10. However, no prediction concerning the number of atoms creating the quasi-gas molecules could be made on the basis of the QMS investigations. It has also been shown that Cd atoms evaporate from the bare single crystalline CdTe (111) surface with an activation energy that is equal (in the limits of the experimental error) to the activation energy for sublimation of Cd atoms from pure, non-crystalline, bulk Cd pieces. The analogous activation energy measured for Te atoms is about 20% larger than that of the relevant sublimation process. This result confirms the fact that Te atoms are bound much stronger in the CdTe crystal lattice than Cd atoms.     

AES Investigation of Chemical Treatment Effect on CdTe and CdZnTe Surfaces

{211} polar surfaces of a Cd_0.96Zn_0.04Te chemically treated single crystal were investigated by Auger electron spectroscopy depth profiling. It was found that the use of a discriminatory “black-white” etchant causes the formation of a thick layer with cadmium having been severely depleted at both (211) A and (211) B faces. The thickness of this layer is larger by a factor of 2 for the matt black face then for the bright and reflecting face. AES investigation of chemical treatment influence on (100) CdTe epitaxial layer surface has been performed. It has been established that etching in Br/methanol solution removes S, Cl and O containing contaminants from the surface, but simultaneously forms a Cd-depleted layer.     

Comprehensive characterization of CdTe and (Cd,Zn)Te single crystals by a chemical etchant

Using a known chemical etchant low- and high angle boundaries and lamellar twins can be seen on CdTe and (Cd, Zn)Te crystal ingots as a whole as well as on slices with naked eyes. Also the polarity of {111} samples can be determined in this way. Etch pits are produced on cut and polished surfaces independent of their crystallographic orientation. A new modified etchant was used to study the low angle subgrain structure on (111)Te surfaces.     

High-resolution photoluminescence studies of (211) CdTe grown on (211)B GaAs substrate

Sharp and rich photoluminescence lines accociated with free exciton (FE), excitons bound to neutral acceptors (A⁰X) and donors (D⁰X) in molecular beam epitaxially (MBE) grown (211) CdTe/(211)B GaAs have been reported for the first time. The results show that the (211) CdTe/(211)B GaAs grown under optimized conditions could have as high a crystal perfection as those grown on lattice-matched substrates.     

Phase-field simulations of Te-precipitate morphology and evolution kinetics in Te-rich CdTe crystals

Te precipitates are one of principal defects that form during cooling of melt-grown CdTe or CZT crystals when grown Te-rich. Many factors such as the kinetic properties of intrinsic point defects (vacancy, interstitial, and antisite defects); stresses associated with the lattice mismatch between precipitate and matrix; temperature gradients and extended defects (dislocations, twin and grain boundaries); non-stoichiometric composition; thermal treatment history all affect the formation and growth/dissolution of Te precipitates in CdTe. A good understanding of these effects on Te precipitate evolution kinetics is technically important in order to optimize material processing and obtain high-quality crystals. This research develops a phase-field model capable of investigating the evolution of coherent Te precipitates in a Te-rich CdTe crystal undergoing cooling from the melt. Cd vacancies and Te interstitials are assumed to be the dominant diffusing species in the system, which is in two-phase equilibrium (matrix CdTe and liquid Te inclusion) at high temperatures and three-phase equilibrium (matrix CdTe, Te precipitate, and void) at low temperatures. Using available thermodynamic and kinetic data from experimental phase diagrams and thermodynamic calculations, the effects of Te interstitial and Cd vacancy mobility, cooling rates and stresses on Te precipitate, and void evolution kinetics are investigated.     

In situ monitoring of CdTe nucleation on GaAs (100) using spectroscopic ellipsometry

In situ spectroscopic ellipsometry was used to monitor the nucleation behavior of CdTe grown on vicinal GaAs (100) substrates by organometallic vapor phase epitaxy. CdTe was grown on GaAs (100) substrates of exact and 2° off towards 110 orientations. A spectroscopic ellipsometer was used to collect in situ data at 44 wavelengths from 4000–7000 Å. The Bruggeman's effective medium approximation was employed to determine the variation of the epilayer volume fraction with thickness, which was an indirect way of monitoring the expected island growth behavior. The Stranski-Krastonov (layer plus island) mode of growth was clearly observed for CdTe growth. The growth on the 2° off substrate was also “denser” than that on exact (100), which implied that coalescence of the islands occurred at lower thickness. This was expected since island nucleation is most favored along the ledges on the surface whose spacing decreases with increasing misorientation. A simple nucleation model, assuming cylinder-like islands, was able to fit the experimental data quite well, lending support to the island growth model.     

Synthesis and spectrum stability of high quality CdTe quantum dots capped with stearate groups in N-oleoylmorpholine solvent

The stearate-capped CdTe quantum dots (QDs) have been first prepared via direct reaction of cadmium stearate with Te powder in N-oleoylmorpholine solvent, which was a kind of clean, air-stable and conveniently synthesized acylamide, and can readily dissolve precursors cadmium stearate and Te powder at a relative low temperature. The as-prepared CdTe QDs exhibited size-dependent optical properties, steep absorbance edge and narrow photoluminescence full width at half maximum. The high-resolution transmission electron microscopy images and X-ray diffraction revealed that the highly monodisperse CdTe QDs were of regular spherical morphology with zinc blende crystal structure displaying mean sizes of about 4 nm. The energy dispersed spectrometry measurement indicated the presence of Cd and Te, with the Cd:Te ratio being close to 1:1. Fourier transform infrared transmission spectra confirmed the existence of stearate on the CdTe QDs surfaces. The experimental results also demonstrated that the stearate-capped CdTe QDs had an unexpected good stability.     

Calculation of Diffusion-limited Growth of Hg1−x Cdx Te Epilayers on CdTe from Te-rich Melts by Step-Cooling LPE

Calculations of layer thicknesses and composition profiles in Hg_1−xCd_xTe layers on CdTe substrates for the growth from Te-rich melts have been carried out for liquidus temperatures of 460 °C, 480 °C, and 500 °C. This has been made on the basis of the multicomponent diffusion model of SMALL and GHEZ and the solid-liquid phase relation of BRICE . It could be shown that growth velocity increases only slightly with rising liquidus temperature. On the other hand, the interdiffusion velocity of Hg and Cd in the solid increases remarkably at a higher temperature. Therefore, to get layers with a constant x-value a higher supersaturation of the melt is necessary. The x-value decreases with rising supercooling by about 0.003 K⁻¹. To demonstrate the thermodynamically and kinetically advantageous properties of CdTe as substrate material, comparative calculations for a “hypothetical” HgTe substrate have been involved.     

Refinement of the atomic structure of CdTe single crystals

A comparative refinement of the anharmonic and disordered models of the atomic structure was performed based on precision X-ray diffraction data sets collected from a CdTe single crystal at 295, 376, 491, and 583 K. Although both models provide the asymmetric contributions of Cd and Te to the structure factors, the anharmonic model seems to be more advantageous.     

Properties of Hg1‐xCdxTe epitaxial films grown on (211)CdTe and (211)CdZnTe

Hg_1‐xCd_xTe (MCT) epitaxial films have been grown employing single crystalline substrates of CdTe and Cd_0.96Zn_0.04Te with (211)Cd and (211)Te crystalline orientations. The Isothermal Vapor Phase Epitaxy (ISOVPE) technique without Hg overpressure has been used for the epitaxial growth. Substrates and films were characterized by optical microscopy, chemical etching and x ray diffraction (Laue technique). The electrical properties were determined by Hall effect measurements. The characterization results allowed to evaluate the crystalline quality of MCT films. (© 2004 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Compensation Mechanism in Vanadium and Gallium Doped CdTe and (Cd,Zn)Te

Vanadium and gallium doped CdTe crystals were grown from the vapour phase (modified Markov method) and (Cd_0.9Zn_0.1)Te: V from the melt (vertical Bridgman). The crystals were characterized by photoinduced current transient spectroscopy (PICTS), photoluminescence (PL) and time dependent charge measurements (TDCM). Transitions from different charge states (V²⁺/V³⁺) of the vanadium donor have been observed in the V-doped crystals by PICTS. A shallow donor level (dE = 0.068 eV) and the Ga A-center have been identified by PICTS and PL measurements in CdTe:Ga. In case of V-doping high resistivity is achieved all over the crystal while Ga-doping results in a high resistivity region only in the middle of the crystal. Calculation of the resistivity by means of a compensation model shows that for both dopants an additional not observed deep donor has to be assumed in order to describe the resistivity distributions.     

Si和GaAs衬底上的ZnTe、CdTe分子束外延材料的晶向倾角

本文采用X射线双晶衍射二次测量法对φ76mm Si(211)和GaAs(211)B衬底上生长的ZnTe和CdTe外延层的晶向倾角进行了测量,发现对于Si和GaAs衬底,外延层的[211]均绕外延层与衬底的[0-11]复合轴朝[111]倾斜,其晶向倾角与晶格失配呈线性关系;通过实际测量验证了在外延层探测到的[133]峰代表[211]关于[111]旋转180°的[255]孪晶向.     

CdTe(110)表面原子与电子结构的第一性原理研究

采用基于密度泛函理论的第一性原理计算了CdTe(110)表面的原子和电子性质.结果表明,CdTe(110)理想表面在禁带中出现两个明显的表面态,弛豫后表层Cd原子和Te原子p态电子发生转移,Cd原子趋向于sp2平面杂化构型,Te原子趋向p3杂化的锥形构型.经过表面弛豫大大降低了表面能,增大了表面功函数,表面占据态和表面空态分别被推进价带顶之下和导带底之上,导致弛豫表面没有明显的表面态.     

High-quality CdTe(100)/GaAs(100) grown by hot-wall epitaxy using gold tube radiation shield

High-quality CdTe(100) layers grown on GaAs(100) substrates by hot-wall epitaxy using a gold tube radiation shield are reported for the first time. From the investigation of thermal properties, we find that the gold tube radiation shield is more effective in heat confinement and temperature stability than a stainless steel tube radiation shield. The CdTe lattice parameters perpendicular to the interface decrease as the layer thickness increases by strain relaxation. We obtain 89 arc sec full width at half maximum of the X-ray double-crystal rocking curve for a 15 μm thick CdTe layer which is the smallest value reported to date. Exciton emission and donor-acceptor pair emission along with longitudinal optical (LO) phonon replicas are obtained from PL measurements, confirming the good quality of the crystal.     

Direct observation of sublimation process on a CdTe(1 1 1) surface using an ultrafine particle

A particle with a CdTe core and carbon mantle was produced by the advanced carbon-coating method which enables direct coverage with a carbon layer using an electron microscope. The coagulated particles containing approximately 30–200 CdTe particles produced by the gas evaporation method were covered with a carbon layer of about 7 nm thickness at 300°C. By heating these particles above 500°C, the sublimation process of a part of the CdTe particle can be directly captured by high-resolution transmission electron microscopy and recorded in real time on videotape. Sublimation on the CdTe(1 1 1) surfaces occurred in the step flow mode of two (1 1 1) layers. It was observed that two (1 1 1) zinc-blende layers changed to the (0 0 0 2) würtzite configuration unit just before sublimation. The condensation of CdTe on the sublimated particle surface and growth of CdTe in the carbon layers were also captured in the video image. These sublimation processes were discussed in terms of the existence of the polarity of II–VI compounds.     

Effects of surface reconstruction on CdTe/GaAs(001) interface structure

CdTe/GaAs(001) heterostructures were fabricated by molecular beam epitaxy on chemically etched and thermally deoxidized GaAs(001) substrates, as well as GaAs(001) (3×1) buffer layers grown in situ by molecular beam epitaxy. Different growth protocols were also explored, leading to Te-induced (6×1) or (2×1) surface reconstructions during the early growth stage. High-resolution cross-sectional transmission electron microscopy was used to examine the final interface structure resulting from the different substrate preparations, and surface reconstructions. The (2×1) surface reconstruction led to pure (001) growth, while the (6×1) reconstruction led to an interface which included small (111)-oriented inclusions. In addition, deposition on etched and deoxidized GaAs(001) wafers led to preferential CdTe growth within etch pits and resulted in a macroscopically rough interface region.     

Microstructure of interfacial HgTe/CdTe superlattice layers for growth of HgCdTe on CdZnTe (2 1 1)B substrates

Transmission electron microscopy has been used to characterize the microstructure of HgTe/CdTe superlattices (SLs) grown by molecular beam epitaxy on CdZnTe(2 1 1) B substrates. The purpose of these intermediate layers was to improve the quality of subsequent HgCdTe (MCT) epilayers intended for infrared detectors. The observations confirmed that the SLs smoothed out the surface roughness of the substrate, and showed that threading dislocations were prevented from reaching the MCT epilayers. High-quality growth of MCT on CdZnTe using the HgTe/CdTe interfacial layers has been demonstrated.     

Se-doped AlGaAs grown on GaAs(111)A by molecular beam epitaxy

The electrical properties of Se-doped Al_0.3Ga_0.7As layers grown by molecular beam epitaxy (MBE) on GaAs(111)A substrates have been investigated by Hall-effect and deep level transient spectroscopy (DLTS) measurements. In Se-doped GaAs layers, the carrier concentration depends on the misorientation angle of the substrates; it decreases drastically on the exact (111)A surface due to the re-evaporation of Se atoms. By contrast, in Se-doped AlGaAs layers, the decrease is not observed even on exact oriented (111)A. This is caused by the suppression of the re-evaporation of Se atoms, by Se---Al bonds formed during the Se-doped AlGaAs growth. An AlGaAs/GaAs high electron mobility transistor (HEMT) structure has been grown. The Hall mobility of the sample on a (111)A 5° off substrate is 5.9×10⁴ cm²/V·s at 77 K. This result shows that using Se as the n-type dopant is effective in fabricating devices on GaAs(111)A.