CdZnTe Wafers Research Articles

Distribution of Te inclusions in a CdZnTe wafer and their effects on the electrical properties of fabricated devices

We quantified the size and concentration of Te inclusions along the lateral- and the growth-directions of a ∼6 mm-thick wafer cut axially along the center of a CdZnTe ingot . We fabricated devices, selecting samples from the center slice outward in both directions, and then tested their response to incident X-rays. We employed, in concert, an automated IR transmission microscopic system and a highly collimated synchrotron X-ray source that allowed us to acquire and correlate comprehensive information on Te inclusions and other defects to assess the material factors limiting the performance of CdZnTe detectors.

Study on the Microstructure of Au/CdZnTe Interface by HRTEM

As a new type of detector material, the CdZnTe (CZT) crystal has the advantages such as a wide energy gap, a large average atomic number, the strong ability to stop X- ray and high resistivity. The CZT is not only the best epitaxial substrate to the growth process of infrared detector material—HgCdTe, but also can be widely used in the detector production of X-ray fluorescence analysis, nuclear waste monitoring, airport and port security testing, industrial detection, medical diagnosis, astrophysics research, etc. So it has very broad prospects. Compared with the traditional NaI detector, the CZT detector has a better energy resolution. Compared with Si and Ge detector, it can work at room temperature. So it absorbed great attention in recent years. The structure of the interface between the metal electrode and CZT has great effect on the performance of detector. Au electrode film was prepared on high resistance CZT single crystal by using vacuum evaporation technology in this research. The interface of Au/CZT was investigated by high resolution transmission electron microscope (HRTEM). The results showed that there existed a disordered layer with the thickness of 5-10nm in the interface between the Au electrode and CZT wafer which was only mechanical polished (as opposed to polished and etched) in our present experimental condition. That could be approved as an amorphous layer through diffraction pattern. On the other hand, the interface was flat, smooth and no amorphous layer on which CZT wafer was polished and etched.

Anisotropic Damage Mechanism during Grinding of CdZnTe Wafers

The grinding wheels with grit size of #600 and #2000 are employed to grind CdZnTe (110) and (111) planes, and ground surface and subsurface damages were investigated. The experimental results show the damage types are affected by the crystallographic orientation and grit size. When grinding CdZnTe wafers with grit size of #600, due to different angle between the cleavage plane and ground surface, the surface morphology and directions of cracks on the (111) plane are different from those on the (110) plane. When grinding CdZnTe wafers with grit size of #2000, the (111) plane has less plastic deformation than the (110) plane, and there is obvious direction of stacking faults on subsurface.

Effect of Mechanical Anisotropy on Grinding of CdZnTe Wafers

In this study, nanoindentation and nanoscratching were employed to research on the mechanical anisotropy of CdZnTe (110) and (111) planes, and their effects on the grinding of CdZnTe wafers were studied. The results suggest that the primary slip system is responsible for the anisotropy of hardness and frictional coefficient. It is easy to slip along directions on (110) plane and directions on (111) plane during scratching, hence the frictional coefficient is the lowest compared to that of other directions, and high surface quality could be obtained during grinding along these directions. The problem of embedded abrasives, which is due to the soft nature of CdZnTe crystal, could be solved when using the grinding process instead of the lapping process.

Damage mechanisms during lapping and mechanical polishing CdZnTe wafers

CdZnTe wafers were machined by lapping and mechanical polishing processes, and their surface and subsurface damages were investigated. The surface damages are mainly induced by three-body abrasive wear and embedded abrasive wear during lapping process. A new damage type, which is induced by the indentation of embedded abrasives, is found in the subsurface. When a floss pad is used to replace the lapping plate during machining, the surface damage is mainly induced by two-body abrasive and three-body abrasive wear, and the effect of embedded abrasives on the surface is greatly weakened. Moreover, this new damage type nearly disappears on the subsurface.

Characterization of CdZnTe Crystals Grown Using a Seeded Modified Vertical Bridgman Method

In this paper, several 60 mm diameter CdZnTe crystal ingots, containing large-size single crystals, were grown using (111) or (211) orientation seeds by the modified vertical Bridgman method. The Zn concentration distribution along the axial direction of the ingots was measured by near-infrared transmission spectroscopy at room temperature. The partition coefficient of Zn during the growth was calculated to be about 1.2. Zinc distribution uniformity measurements were carried out using photoluminescence mapping measurement at 80 K, which showed that the band gap variation in the CdZnTe wafers is less than 0.004 eV. IR microscopy showed that the diameters of the Te inclusions present in the material are in the range 6-9 mum, and the density of the inclusions is 1~3times10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">5</sup> cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-3</sup> . IR transmission measurements in the wave number region from 500 to 4000 cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-1</sup> demonstrate that the IR transmittance of CdZnTe wafers is higher than 60%. Current-voltage measurements were performed on test structures fabricated using thermally evaporated Au contacts deposited on as-grown crystals, which revealed bulk resistivity values of 2~5times10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">10</sup> Omegamiddotcm. Typical leakage currents for the planar devices were <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">~</i> 4 nA at a field strength of 1500 Vcm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-1</sup> . The electron and hole mobility-lifetime products were evaluated using alpha particle irradiation. The obtained typical (mutau) <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">e</sub> and (mutau) <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">h</sub> values for the as-grown CdZnTe were 2.3times10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-3</sup> cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> V <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-1</sup> and 1.5times10 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-4</sup> cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> V <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">-1</sup> , respectively.

Material removal mechanism of precision grinding of soft-brittle CdZnTe wafers

Cd0.96Zn0.04Te (111) wafers were precisely ground by #800, #1500, #3000, and #5000 diamond grinding wheel. For comparison, Cd0.96Zn0.04Te (110) wafers were machined by lapping, mechanical polishing, and chemical mechanical polishing. High-resolution environmental scanning electron microscopy equipped with energy dispersive spectroscopy and optical interference surface profiler both were employed to investigate the surface quality and material removal mechanism. The results show that the material removal mechanism of #800 grinding wheel is abrasive wear, fatigue wear, and adhesive wear, and that of #1500 is abrasive wear and fatigue wear. Both the material removal mechanism of #3000 and #5000 grinding wheel are abrasive wear, leading to the excellent ductile removal precision grinding. While the material removal mechanism of CMP on CdZnTe wafers is firstly chemical resolving reaction and secondly mechanical carrying action. Moreover, precision grinding exhibits high-efficiency character and eliminates the imbedding of free abrasives of Al2O3 and SiO2.

Influence of CdZnTe/Au Contact on the Wire Bonding Property between the Down-Lead and Au Electrode

According to the need to make high quality X-ray or γ-ray detectors made by high resistance CdZnTe, the wire bonding technology between the Au contact layer of CdZnTe wafer and the down-lead has been studied in this paper. The influence of welding parameters and surface treatment of CdZnTe wafer on the welding quality between Au contact layer and down-lead have been discussed in detail, such as chemical and mechanical polishing technology of CdZnTe wafer before welding, thickness of the contact layer of CdZnTe wafer, welding energy, welding pressure, welding time, and etc. The results showed that the successful jointing between Au contact layer of CdZnTe and down-lead can be achieved when applying the suitable welding parameters and pretreatment technology of CdZnTe wafer. The optimized welding parameters for CdZnTe wafer were as follows: welding power 2w, welding pressure 60×10-3 kg, welding time 20ms and the power for fire ball formation 1.5w.

Comparison of CdZnTe crystals grown by the Bridgman method under Te-rich and Te-stoichiometric conditions and the annealing effects

Two categories of CdZnTe (CZT) wafers were grown, one by the modified vertical Bridgman method (MVBM) under Te-rich and the other under Te-stoichiometric conditions. The effects of annealing in saturated Cd vapor on CZT properties such as precipitate, infrared (IR) transmission, etch pit density (EPD) and full-width at half-maximum (FWHM) value of X-ray rocking curve were systematically discussed. As-grown CZT wafers grown under Te-rich conditions contain large-size Te precipitates with density of high-10 3–low-10 4 cm −2 and IR transmission less than 60%. The micron-size Te precipitates were entirely removed and IR transmission was increased to over 60% after annealing based on the Cd-diffusion-induced thermomigration mechanism. In comparison, as-grown wafers grown under Te-stoichiometric conditions contain Cd precipitates with density of high-10 5 cm −2 and IR transmission ∼60%. After annealing, the precipitate density was reduced to low-10 4 cm −2 and IR transmission exceeded 60%, indicating a simple and yet novel approach for eliminating Cd precipitates. The FWHM and EPD of CZT wafers were not significantly influenced by annealing. In short, crystal growth under Te-rich conditions was proved to be a practical technique for producing high-quality CZT.

Effects of mosaic structure on the physical properties of CdZnTe crystals

Mosaic structure in CdZnTe crystals was identified by using scanning electron microscopy (SEM), then the effects of mosaic structure on the physical properties were studied by means of high resolution X-ray diffraction (HRXRD), I–V characterization, and Fourier transform infrared (FT-IR) measurements. X-ray studies suggest that mosaic structure has a slightly different orientation from the main part of the sample. In FTIR measurements, IR transmittance of CdZnTe wafer with mosaic structure in the wavenumber between 500 and 1100cm−1 is only about 4%. From 1100 to 4000cm−1, it increases gradually to about 50%. Analysis of the low IR transmittance is discussed using the theory of free carrier absorption. I–V curve of CdZnTe wafer with mosaic structure exhibited piezoresistance characteristic. This characteristic can be considered as a transition to the space charge limited regime due to carriers injecting at the mosaic structure. By analogy with Schottky barrier behavior, we calculated interface barriers of CdZnTe wafer containing mosaic structure and the mosaic structure free wafer, which are 0.78 and 0.98eV, respectively.

The relationship between stress and photoluminescence of Cd 0.96Zn 0.04Te single crystal

Photoluminescence (PL) and high-resolution X-ray diffraction (HRXRD) experiments have been carried out on strained Cd 0.96Zn 0.04Te single crystals to reveal the correlation between luminescence spectrum and stress in CdZnTe wafers. The tensile stress introduced by quenching was measured by X-ray diffraction methods. The photoluminescence spectra were found to move to higher energies, and the intensity of neutral donor bound exciton (D 0, X) peak of CdZnTe wafer was increased after quenching. The reasons for these variations are suggested to be determined by the change of the energy band gap E g of the CdZnTe wafer and the increase of crystal defects. The in-plane biaxial stress of 1 GPa results in a shift of the excitonic PL spectrum of 7.05–12.03 meV. In particular, the tensile strain reduces the formation energy of Te Cd antisite defects, which are the shallow donors located at 0.01 eV below the conduction band. Therefore, high density of Te Cd antisite defects demonstrated by the higher (D 0, X) emission are generated due to the high tensile strain in the quenched specimen.

Study of dislocations in CdZnTe single crystals

AbstractExperimental studies have been conducted to investigate the influence of dislocations on the properties of CdZnTe crystals. Dislocations were introduced into CdZnTe wafers by means of bending deformation at elevated temperature. The average infrared (IR) transmittance of CdZnTe wafers after deformation is decreased from 64 to 44%. The polarization absorption of dangling‐bond electrons in dislocations should be responsible for this decrease of IR transmittance. In photoluminescence measurements, the shallow donor–acceptor pair transition peak at 1.557 eV is detected in CdZnTe after deformation. In the defect‐related region, a new band (Dcomplex) located at 1.508 eV appears, which should be attributed to defect levels introduced by dislocations. Meanwhile, leakage current at 15 V is increased from 10–8 to 10–4 A. The analysis suggests that the leakage current of CdZnTe after deformation is dominated by the Poole–Frenkel effect. Te and Cd dislocations created at two faces introduce different electrical characteristics. (© 2007 WILEY‐VCH Verlag GmbH &amp; Co. KGaA, Weinheim)

Effect of annealing on the residual stress and strain distribution in CdZnTe wafers

The effect of annealing on residual stress and strain distribution in CdZnTe wafers was studied based using an X-ray diffraction (XRD) method. The results proved the effectiveness of annealing on the reduction of the residual stress and strain. By the means of transmission electron microscopy (TEM) and infrared (IR) transmission analyses, it was found that dislocation gliding, decreases in the size of the Te precipitates, dispersing of Te precipitates, composition homogenization, and point defects recombination contributed to a reduction of the residual stress and strain during annealing of the wafer. Additionally, the larger residual stress in CdZnTe wafers introduced bigger lattice misfits. Thus, for more the residual stress and strain in the CdZnTe wafer, the IR transmission will be lowered.

Residual stress and strain in CdZnTe wafer examined by X-ray diffraction methods

A non-destructive method based on X-ray diffraction was developed to measure the stress and strain in CdZnTe single crystal near the surface. From the experimental results and calculations, the residual stresses in CdZnTe single crystal were determined to be σ1=30 MPa, σ2=14 MPa, and τ12=-4 MPa, respectively. The residual stress derived from the measurement strain in CdZnTe was thought to be composed of the thermal stress, the misfit stress, and the mechanical stress. The distributions of non-uniform strain in a CdZnTe wafer are about 3.9%, while the distributions of uniform strain in the CdZnTe wafer are 0.5%, much smaller than those of the non-uniform strain.

The surface leakage currents of CdZnTe wafers

The surface leakage currents (SLCs) and surface sheet resistances (SSRs) of CdZnTe (110), (111) A and (111) B surfaces after etching with Br–MeOH solution, chemo-mechanical polishing (CMP) and passivation were measured in the parallel stripe model, respectively. Meanwhile the surface compositions were determined by X-ray photoelectron spectroscopy (XPS). Te enrichment introduced by etching with Br–MeOH resulted in the increase of the SLCs of CdZnTe wafers. After chemo-mechanical polishing, Te enrichment was removed, and SLCs decreased. CdZnTe (111) B without Te enrichment possesses higher SLC than that of (111) A, and (110) surface has the lowest SLC, which should be attributed to the lower surface dangling bonds. Passivation treatment with NH4F+H2O2 is an effective method to decrease SLCs of CdZnTe, by which the SLC was decreased two orders.

Dislocation-induced IR absorption and photoluminescence emission in Cd 0.96Zn 0.04Te

Dislocations were introduced into CdZnTe wafers by the means of bending deformation at an elevated temperature. The average IR transmittance of CdZnTe wafers was found to be decreased to 44% after deformation from 64% before. The polarization absorption of dangling-bond electrons in dislocations should be responsible for the decrease of IR transmittance. In photoluminescence measurement, the shallow donor–acceptor pair transition peak at 1.557 eV and its first longitudinal optical phonon replica were detected in CdZnTe after deformation. In defect-related region, a new band D complex located at 1.508 eV appeared, which should be attributed to defect levels introduced by dislocations.

Effect of surface treatments on the electrical and optical properties of CdZnTe single crystal

The effects of surface treatments on the electrical and optical properties of CdZnTe (CZT) crystals were studied experimentally. The IR transmittance of CZT after cutting is only 15%, after chemically polishing with Br-MeOH it increased to 63%. The reflection of surface roughness and absorption of surface-damaged layer should be responsible for the variation. After aging and passivation,CZT possesses low IR transmittance. After fine polishing, the surface oxide layer was removed and the IR transmittance increased. CZT wafers with high roughness possess low surface sheet resistance, and the Te enrichment introduced by etching with Br-MeOH results in high surface leakage current. The effects of etching and passivation on surface trap state have been analyzed with photoluminescence spectra. Etching can deprive effectively the surface damage layer, and passivation can minimize the surface trap state density.

Impurity segregation in CdZnTe by photoluminescence mapping

Low temperature photoluminescence (PL) mapping is employed to retrospectively monitor the processes taking place during crystal growth. Time evolution of the solidification of the melt is derived from zinc concentration map, which is calculated from energy gap obtained from free-exciton (FX) reabsorption in the PL spectra. Investigation of the distribution of PL intensities of specific defects shows that at least relative segregation coefficients can be calculated. This method has a potential application in preparation and characterization of pure and homogeneous CdZnTe wafers and also in identification of unknown defects. (© 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

Study on the annealing of Cd1－xZnxTe in In vapor

In order to meet the requirements of the design of radiation detectors, CdZnTe (or Cd1-xZnxTe) wafers grown by vertical Bridgman method w ere annealed in the vapor of In. The nature of this treatment is a diffusion pro cess, thus it is meaningful to relate the change of resistivity to the diffusion parameters. A transformation model correlating resistivity and conduction type of CdZnTe with the main diffusion parameter —— diffusion coefficient——is put forward in this paper. Combining the model with the analysis of our experimenta l data (namely DIn = 5.17 10-9, 2.625×10-10 an d 3.455×10-11cm2·s-1), the values of the diff usion coefficient of In in Cd0.9Zn0.1Te at 1073, 973 and 873K have been given for the first time, which coincide closely with those in Cd Te given by different authors. With the data, the effect of annealing time on th e change of resistivity and conduction type for Cd0.9Zn0.1 Te slice, which is annealed in saturated In vapor at 1073, 973 and 873K, have be en simulated and good consistency acquired. This work suggests an alternative wa y to determine the diffusion coefficient in semiconductor material, and also ena bles us to analyze the diffusion process quantitatively and predict the annealin g result.

Surface passivation of CdZnTe wafers

The chemical polishing process and the subsequent passivation process of CdZnTe wafers were studied. The treatment effects were tested through XPS analysis and I–V measurement. The chemical etching in 2%Br–MeOH solution may effectively remove the damaged layer and improve the ohmic contact between CdZnTe wafer and Au electrodes. CdZnTe wafers after Br–MeOH etching were passivated with five different passivants respectively. It was found that the surface leakage currents of all CdZnTe wafers passivated with different passivants were reduced by 1–2 orders. CdZnTe wafer passivated in NH 4F/H 2O 2 solution showed the best passivation efficiency because the enriched Te on the surface was fully oxidized to TeO 2, which results in the thickest oxide layer, the most stoichiometric surface and the least leakage current. The surface of CdZnTe wafer is Te-rich after passivated in NH 4F/H 2O 2 or H 2O 2 solution and Cd-rich after passivated in KOH or KOH/H 2O 2 solution.