We have studied deep-level impurities in CdTe/CdS thin-film solar cells by capacitance–voltage (C–V), deep-level transient spectroscopy (DLTS), and optical DLTS (ODLTS). CdTe devices were grown by close-spaced sublimation. Using DLTS, a dominant electron trap and two hole traps were observed. These traps are designated as E1 at EC−0.28 eV, H1 at EV+0.34 eV, and H2 at EV+0.45 eV. The presence of the E1 and H1 trap levels was confirmed by ODLTS. The H1 trap level is due to Cu-induced substitutional defects. The E1 trap level is believed to be a deep donor and is attributed to the doubly ionized interstitial Cu or a Cu complex. The E1 trap is an effective recombination center and is a lifetime killer.

We have studied deep-level impurities in p+–n GaInNAs solar cells using deep-level transient spectroscopy (DLTS). These films were grown by atmospheric- and low-pressure metalorganic vapor-phase epitaxy. The base layer is doped with silicon, and the emitter layer is zinc doped. Two types of samples have been studied: samples were grown with and without the addition of oxygen impurity. Two electron traps were found in all samples. These are designated as: E1, at EC−0.23–EC−0.27 eV, E2 at EC−0.45 eV, and E2* at 0.77 eV. With the addition of oxygen impurity, DLTS showed additional traps designated as E3 (electron) at EC−0.59 eV and H3 (hole) at EV+0.59 eV. Using secondary ion mass spectroscopy, the oxygen concentration was found to be about 2–3×1019 and 1×1017 cm−3 in two sets of samples. However, only samples containing oxygen contained the two near-midgap levels (E3 and H3). We present evidence that these levels are associated with the oxygen defect. As we change the dc bias voltage, the E3 trap disappears in unison with the appearance of the H3 trap. Furthermore, E3 and H3 trap levels have comparable capture cross sections. This oxygen-related trap is an effective recombination center. The measured Shockley–Hall–Read lifetime for this center is about 0.6 μs.

No abstract

No abstract

Technologically the electrochemical deposition method through the influence of potential, temperature, pH and composition of reactants offers excellent control lover the properties of semiconductors Using a potentiostatic approach, the films of CdS were deposited on t i n oxide coated glass substrates at merent conditions. The films were found to be smooth, uniform and adherent 114th a small grain size. X-ray &tion analysis indicated a hexagonal phase rather than the cubic phase.Electrochemica1 deposition parameters were studied to

Abstract. This paper presents and discusses the first Deep-Level transient spectroscopy (DLTS) data obtained from measurements carried out on both Schottky barriers and homojunction devices of GaInNAs. The effect of N and In doping on the electrical properties of the GaNInAs devices, which results in structural defects and interface states, has been investigated. Moreover, the location and densities of deep levels related to the presence of N, In, and N+In are identified and correlated with the device performance. The data confirmed that the presence of N alone creates a high density of shallow hole traps related to the N atom and structural defects in the device. Doping by In, if present alone, also creates low-density deep traps (related to the In atom and structural defects) and extremely deep interface states. On the other hand, the copresence of In and N eliminates both the interface states and levels related to structural defects. However, the device still has a high density of the shallow and deep traps that are responsible for the photocurrent loss in the GaNInAs device, together with the possible short diffusion length.

No abstract

We have studied deep level impurities in p+−n GaInNAs solar cells using secondary ion mass spectroscopy (SIMS), capacitance–voltage (C−V), and deep-level transient spectroscopy (DLTS). These films were grown by atmospheric and low-pressure metalorganic vapor phase epitaxy. The base layer is doped with silicon and the emitter layer is Zn doped. Two types of devices have been studied: devices grown with and without the addition of oxygen impurity. Using SIMS, the oxygen concentration was found to be about 2−3×1019 and 1×1017 cm−3, respectively. C−V measurements at temperatures below 190 K have revealed that carrier freeze out occurs in high oxygen samples, whereas we did not observe this phenomenon in low oxygen devices. In addition to observation of several trap levels in all samples, we observed two additional near midgap traps designated E3 (electron) at EC −0.59 eV and H3 (hole) at EV +0.59 eV only in high oxygen devices. We present evidence that these levels (E3 and H3) are associated with the oxygen defect and are an effective recombination center. We observed a logarithmic correlation between the concentration of the oxygen recombination center and the device quantum efficiency. From this correlation, the hole diffusion length is 0.06–0.15 and 0.12–0.31 μm with and without back reflection, respectively. We found that the hole diffusion length is strongly dependent on the concentration of the oxygen recombination center. We conclude that the oxygen recombination center is a lifetime-limiting defect and, therefore, controls the hole diffusion length.