The composition and structural properties of a single crystal of the Cu-Ag-In-Se system are analyzed. Laue diffraction shows a single crystal while XRD diffraction and EDAX composition indicate two crystalline phases and two compositions close to Cu0.97Ag0.03In1.75Se2.84 and Cu0.95Ag0.05In2Se3.5 with lattice parameter, a = 5.770 angstrom and a = 5.790 angstrom and c/a congruent to 2.0 respectively. Impedance spectroscopy is carried out at temperatures up to 120 degrees C in a sequential annealing in order to obtain the electrical properties. A motion of two ions is observed and two ionic resistances and activation energies are computed in the 0.15-0.17 eV range and 0.52 eV, respectively. In the successive annealing, the impedance spectra change, probably due to a non-reversible process in the sample. After the impedance analysis, composition measurements and the structural analysis show a massive motion of Ag + Cu and In ions in the slice. These motions produce different phases with very different compositions in different regions. Due to the high disorder in Cu and In sublattices and to the high number of (2V(Cu) + In-Cu) defect pairs, these ions are easily moved, leading to the formation of an InSe crystalline phase. Ions are rearranged in the chalcopyrite phase region, along with the transformation of In3+ into In2+ chemical species accompanied by the corresponding electron conduction capture. These changes are responsible of the non-reversibility of the process. These results would allow to understand the highest solar energy conversion efficiencies of up to 20.3% observed in CuIn1-xGaxSe2 (CIGS) thin films obtained using a three-stage co-evaporation process. In these films, the CIGS layer reaches a copper rich composition and a quasi-liquid Cu2-ySe phase is formed which enhances crystallization of the absorber layer and also affects the distribution of the group III elements throughout the layer. (C) 2012 Elsevier B.V. All rights reserved.

Two single crystals of the same ingot and a polycrystal with compositions close to CuIn 2 Te 3.5 , CuIn 3 Te 5 and CuIn 4 Te 6 , respectively, have been studied using impedance spectroscopy at different temperatures in the 20–120 °C range. The negative imaginary impedance part, − Z ** , as a function of the real impedance part, Z * , (Nyquist plots) can be fitted to one or two semicircles associated with the Warburg diffusion and related one or two mobile ions, depending on the slice composition. In dc current and using the current intensity decay method at 20 °C, the conductivity as a function of time permits one to compute the diffusion coefficients of one or two mobile ions observed in impedance spectroscopy. The diffusion coefficients of the Cu ion are higher than those in CuInSe 2 and are associated with the number of copper vacancies, V Cu , and indium–copper antisites, In Cu , in the Cu sublattice. EDAX measurements in the slice profile, before and after the electrical analysis, confirm a massive motion of Cu atoms along the slice thickness. In all samples, there is a region with composition close to CuIn 1.7 Te 3 and a new atomic distribution, different in each sample depending on the initial composition and their structure. The ionic motion permits one to understand the behaviour of chalcopyrite thin films when they are grown with a composition gradient (from a Cu-rich chalcopyrite to an In-rich one).

Tin doped indium oxide (ITO) and fluorine doped tin oxide (FTO) thin films have been prepared by one step spray pyrolysis. Both film types grown at 400°C present a single phase, ITO has cubic structure and preferred orientation (400) while FTO exhibits a tetragonal structure. Scanning electron micrographs showed homogeneous surfaces with average grain size around 257 and 190nm for ITO and FTO respectively.

Copper Aluminium dioxide CuAlO2 (CAO) thin films have been grown by spray pyrolysis technique on different substrates: glass, zinc oxide and silicon. Different annealing in air at 450-500 degrees C temperature range have been carried out. The thin film structure has been studied from the experimental X-ray diffraction patterns and different morphologies have been found which depend on the substrate type and the grown conditions. The CuAlO2 phase has been found on ZnO substrate after annealing at 500 degrees C in air during two hours. The CuAlO2 grain size was around 30 nm. Optical measurements of a sample on glass substrate after annealing at 500 degrees C have been realized. They have shown a film transmittance around 80% in the visible zone with a band gap near 3.75 eV in a good agreement with literature.

Fluorine doped tin oxide (FTO) thin films have been prepared by spray pyrolysis technique with no further annealing. Films with 2.5% of fluorine grown at 400 degrees C present a single phase and exhibit a tetragonal structure with lattice parameters a = 4.687 angstrom and c = 3.160 angstrom. Scanning electron micrographs showed homogeneous surfaces with average grain size around 190 nm.The films are transparent in the visible zone and exhibit a high reflectance in the near infrared region. The best electrical resistivity was 6.3 x 10(-4) Omega cm for FTO with 2.5% of fluorine. The ratio of transmittance in the visible to the sheet resistance are in the 0.57 x 10(-2)-1.96 x 10(-2) Omega(-1) range.

In order to obtain single-phase thin films of the system Cu–In–Te with optoelectronic properties adequate for solar cells, electrodeposition techniques were used on substrates of molybdenum supported by glass. Different annealings in Te atmosphere have been done that affect the Te concentration and In/Cu atomic ratio. Single chalcopyrite phase appears in two ranges of composition where the In/Cu atomic ratio varies between 0.21–0.76 and 0.90–3.46, respectively. Morphology, cell lattice parameters and electrical resistance for single-phase samples depend strongly on the composition in the annealed samples. The cell parameters ranges are a=6.141–6.183Å and c=12.201–12.375Å.

Several CuIn3Te5 single-crystal samples grown by the vertical Bridgman method have been studied. Changes in composition of the In/Cu atomic ratios in the 3.45–1.88 range have been detected along the ingots by X-ray energy-dispersive analysis. In the first-to-freeze region, Cu excess and In and Te depletions appear. From X-ray powder diffraction data we conclude that the crystals show the P-chalcopyrite-type structure, space group P4̄2c, with lattice parameters a=6.1713Å and c=12.329Å, similar to that found in CuIn3Se5. The optical energy gaps, as determined by spectral ellipsometry, lie in the 1.83–1.93 eV range and change with composition. Micro-Raman spectra and infrared reflectance measurements confirm the presence of the crystallographic phase CuIn3Te5 and indicate a highly ordered Te sublattice and disordered Cu and In sublattices. Conduction types n and p are found, suggesting the possibility of intentional doping. The crystal density and gaps seem appropriate for their use in gamma-ray detectors.

The possible application of amorphous carbon nitride thin films to prevent the multipactor effect is discussed. The main difficulty lies in the increase of the secondary electron emission coefficient after air exposure. Adsorption processes on amorphous carbon nitride surfaces due to air exposure and subsequent changes due to thermal annealing have been studied by means of x-ray photoemission spectroscopy with the purpose of understanding the changes of secondary electron emission characteristics.

The dielectric constant of CuGaSe2 as a function of stoichiometric deviations has been obtained from photomodulated spectral ellipsometry measurements. Optical gaps have been computed by fitting experimental and differentiated data of the imaginary part of the dielectric constant to function ε2 of the Lorentz model and to the second derivative of ε2. In each sample, three transition energy values have been found in the 1.65–1.72, 1.83–1.95, and 2.97–3.14 eV ranges. The presence of point defects reduces the energy gap values. For nonstoichiometric samples, changes in the first energy gap values have been analyzed as a function of the displacement of the position of the anion in the unit cell. The shifts in the valence band structure have been analyzed and it is concluded that the difference between the first and second transition energies, (Eg2−Eg1), is also affected by stoichiometric deviations, so that the Γ5v(2) level in samples with point defects is closer to Γ4v(2) level than in stoichiometric ones. Changes in (Eg3−Eg2) and (Eg3−Eg1) are related to Se position.

Two alternative crystal structures have been identified in evaporated films grown from synthesized CuInSe2. The first one, tetragonal CuInSe2 was obtained by ‘flash’ evaporation and the second one, the ordered vacancy compound CuIn2Se3.5 by e-beam evaporation. They had optical gaps of 1.0 for the first and 1.24 eV for the second. The conductivities were p- and n-type, respectively; the last one remained n-type even after annealing in selenium atmospheres. The combination of both films in the heterostructure Mo/CuInSe2/CuIn2Se3.5/CdS/ITO has been obtained and is at present being studied as a solar cell.

Polycrystalline CuIn2Se3.5, CuIn3Se5 and CuIn5Se8 have been synthesized in vacuum under different cooling conditions resulting in different phases. X-ray diffraction and Rietveld refinement of the diagrams have been used to characterize the different structures obtained. The two former compounds were refined using two different types of structure: stannite (S.G.: I4̄2m) and the proposed P–chalcopyrite (S.G.: P4̄2c). The determined bond distances after refinements indicate a higher tetragonal distortion for stannite type structure than P–chalcopyrite structure. These results seem to indicate that the Cu–In–Se structure with these two stoichiometries belongs to the P4̄2c S.G. On the other hand, for the CuIn5Se8 compound a far different XRD diagram was obtained where all the reflections have been assigned to two structures, named γ-phases, one trigonal (γ) and one hexagonal (γH).

Cu-In-Se-Te thin films have been produced by electrodeposition on molybdenum electrodes supported on glass substrates. The composition of the films is a function of the growth conditions. The best conditions for a near-stoichiometric composition are a deposition potential of -0.5 V vs. SCE and a time greater than 4000 s. After electrodeposition the samples are amorphous. However, crystallization of the films can be achieved by various thermal treatments, for example, annealing in a Te atmosphere at 550 degrees C for 30 minutes followed by annealing under vacuum at 400 degrees C for 1 h. This produces crystalline quaternary phases, as confirmed by X-ray diffraction. In order to study the optical properties of the films, some samples were grown on indium tin oxide (ITO) electrodes and annealed at 550 degrees C for 30 min. The observed absorption coefficients showed two regions corresponding to two energy gaps having values of E-g1 = 0.81 eV and E-fg1 = 1.28 eV. The same films exhibited photoelectrochemical responses with flatband potentials of -0.15 V. All samples exhibited p-type conduction.

The dependence of structural parameters and force constants for Cu–Se, Ga–Se, and In–Se bonds on compositional deviations in CuIn0.5Ga0.5Se2 have been studied. The composition gradient along the ingot was obtained by a single fusion at 1150 °C of the components and subsequent slow cooling in a still ampoule placed in a vertical furnace. All along the sample, a single chalcopyrite phase is present and its composition along its length was found by energy dispersive analysis of x-ray measurements on slices. Unit cell parameters, anion displacement, and Cu occupation fraction in its sublattice were analyzed by x-ray powder diffraction and Rietveld refinement methods. The anion displacement found is a function of the Cu defect in its sublattice. The existence of associated defects, i.e., two Cu vacancies and one Ga in Cu site, [2V(Cu)+GaCu], is proposed to explain the Cu defect in its sublattice and the changes in lattice parameters. This leads to the existence of BIII vacancies (BIII=In+Ga), and interstitial Cu up to 8 at. % that also cause changes in the structural parameters. Infrared reflectance measurements led to the imaginary dielectric constant determination which, fitted to a Lorentz function, permitted to obtain atomic vibration modes. Using the model of Neumann for chalcopyrites, the values of force constants for Cu–Se, Ga–Se, and In–Se bonds were computed. These appear to increase when the occupation of each sublattice increases.

CuInSe2 and CuGa1−xInxSe2, with x=0.25, 0.5 and 0.75, films have been grown by 'flash' evaporation. Films 2 μm thick over Mo coated glass have shown columnar grains, single chalcopyrite phase and adequate optical and electrical properties as photovoltaic semiconductor. The film composition could be changed by modifying the deposition parameters, such as crucible and substrate temperatures and deposition rate. Inherent problems in this technique, such as 'spitting', have been removed by changing the dimensions of the system and crucible type. A solar cell formed by depositing on top of the CuInSe2 a Cu-poor layer of composition close to CuIn2Se3.5 has shown about 6% efficiency.

The low-temperature oxidation during deposition by evaporation of molybdenum thin films has been investigated. Analysis by x-ray photoelectron spectroscopy and x-ray diffraction reveals that small differences in the substrate temperature during deposition may give rise to important changes in the final composition and structure of the molybdenum oxide. Changes in binding energy and line shape of the Mo 3d5/2−Mo 3d3/2 doublet attributed to oxygen incorporation have been studied. Two principal steps can be distinguished, with a transition temperature of ∼310 °C. Up to substrate temperatures of ∼310 °C, the superficial Mo remains almost unaffected, with some oxygen dissolved. At ∼310 °C, mixing of Mo0 metal and molybdenum oxide (Moδ+0<δ<4) clusters or islands is observed. Finally, above this temperature a surface layer of molybdenum oxide, Mo6+, is formed. In addition, an abrupt change in d100 interplanar parameter of Mo is observed.

CdS thin films have been prepared by chemical bath deposition. As-deposited films are cubic and show a sulfur deficiency. From the transmittance and reflectance data analysis direct band gaps (Eg) ranging from 2.180 to 2.448 eV have been obtained. Air and vacuum annealed samples show a decrease in the band gap. The refractive index (n) lies in the range 1.61–2.34. A dependence of band gap on composition has been observed and the possible reasons are discussed.

The effect of Cu and In vacancies on the structural parameters of near stoichiometric CuInSe2 samples has been analyzed finding that while both change the Se position and the interatomic distances, In vacancies seem to affect more strongly to these parameters. The presence of this kind of defects has been studied by PL and electrical conductivity as a function of temperature measurements. Band gap energy values have been obtained from reflectance measurements and are compared to those predicted by Jaffe and Zunger equation [Phys. Rev. B 29 1882]. The structural contribution term needs to be corrected for small stoichiometric deviations around CuInSe2.

A method to obtain the band gap energy value from reflectance measurements has been developed for bulk materials with direct and indirect gaps. In an energy region around the gap value, the experimental reflectivity is fitted by two functions of the type R=A0+A1′(1/hν)2−A1′Eg×(1/hν)3 and R=A0+B[(hν−Eg)/(hν)]4, for direct and indirect transitions, respectively, in an energy range around the gap value. This implies that the absorption coefficient, α, shows a dependence with the energy α=(A/hν)(hν−Eg)m with m=1/2, 2 for direct and indirect gaps, respectively. The method has been checked in several materials of known energy gaps including Si and Ge.

The dependence of the structural parameters on compositional deviations of CuGaSe2 has been studied. These deviations have been induced along an ingot by a single fusion of the components at 1150 degrees C and subsequent slow cooling in a stationary ampoule in a vertical furnace. All along the sample a single chalcopyrite phase is present and a compositional gradient along the ingot was found by energy-dispersive analysis of X-rays (EDAX) measurements, the upper part being rich in Ga (series B) and the lower part in Cu (series A), with Cu/Ga ratios of 0.95 and 1.1 respectively. A hypothesis of the existence of two phases in the melt is proposed to explain these facts. The unit cell parameters, anion displacement and Cu and Ga occupation numbers in their sublattices were analysed by X-ray powder diffraction and Rietveld refinement methods. In series A the occupation numbers are near stoichiometry, while in series B a Cu defect appears. In both series, changes in unit cell parameter are related to changes in Cu content, suggesting the presence of a fraction of Cu ions either as interstitials or at Ga sites when Cu is in excess, or of Cu vacancies in its sublattice when there is a Cu deficiency. (C) 1998 John Wiley & Sons, Ltd.