M. Emziane

The distribution of impurities in the interfaces and window layers of thin-film solar cells

We report a systematic multielement study of impurities in CdS window layers by dynamic and quantitative secondary-ion-mass spectrometry (SIMS) with high depth resolution. The study was carried out on CdTe∕CdS solar cell structures, with the glass substrate removed. The analysis proceeded from the transparent conductive oxide free surface to the CdTe absorbing layer with a view to examining the influence of the CdCl2 heat treatment on the distribution and concentration of impurities in the structures. Special attention was paid to the impurities present in the CdS window layer that may be electrically active, and therefore affect the characteristics of the CdTe∕CdS device. It was shown that Cl, Na, and Sb impurities had higher concentrations in CdS following cadmium chloride (CdCl2) heat treatment while Pb, O, Sn, and Cu conserved the same concentration. Furthermore, Zn, Si, and In showed slightly lower concentrations on CdCl2 treatment. Possible explanations of these changes are discussed and the results...

A combined SIMS and ICPMS investigation of the origin and distribution of potentially electrically active impurities in CdTe/CdS solar cell structures

Quantitative, dynamic secondary ion mass spectrometry (SIMS) combined with inductively coupled plasma mass spectrometry (ICPMS) was used to study the origin and distribution of impurities in CdTe/CdS/In2O3:F/glass solar cell structures. Particular emphasis was put on the potentially electrically active impurities that may possibly originate from either the CdTe starting material or the cadmium chloride (CdCl2) post-deposition annealing-induced activation, and that are likely to affect the device performance. Structures were grown using CdTe starting material of 5N and 7N purity, and were analysed by SIMS, both before and after CdCl2 heat treatment. Depth profiles of the elements: Cl, O, Cu, Na, In, Sb, Sn, Si, Zn, Pb and S were made. Around 91% of the impurities detected by ICPMS in the CdCl2 powder used were found to consist of species known to show electrical activity in CdTe. The origin of most of the impurity species present in the cell structures was elucidated, and it was shown that Pb, Sn, Cu and Zn were not coming from the CdTe or CdCl2 starting materials. However, Na and Cl turned out to originate from the CdCl2 processing. The potential interdiffusion of elements, such as S, Te, Si, In and Na from the layers of the structures was also highlighted and is discussed.

In situ oxygen incorporation and related issues in CdTe/CdS photovoltaic devices.

CdTe∕CdS∕SnO2∕ITO:F solar cell devices were investigated using quantitative secondary ion mass spectrometry (SIMS) depth profiling. They were grown on sapphire substrates and potentially active impurity species were analyzed. The SIMS data were calibrated for both CdS window layer (grown by sputtering) and CdTe absorber layer (deposited by close-space sublimation). For comparison, some of the samples were grown with and without oxygen incorporation into the CdTe layer during its deposition, and with and without postgrowth cadmium chloride (CdCl2) annealing in air and chemical etching. These devices were back contacted using Mo∕Sb2Te3 sputtered layers. It was shown that for CdTe and CdS layers there was a correlation between the concentrations of oxygen and chlorine. In situ oxygen incorporation in the CdTe layer yielded a substantial improvement in the device parameters and achieved an efficiency of 14% compared to 11.5% for devices fabricated in the same conditions without oxygen incorporation in CdTe. I...

Efficiency improvement in thin-film solar cell devices with oxygen-containing absorber layer

The CdTe∕CdS solar cell devices were grown using a dry process consisting of sputtering for the transparent conducting oxide and CdS window layers, and close-space sublimation for CdTe absorber layer. These devices were back contacted using Mo∕Sb2Te3 sputtered layers following the CdCl2 activation process carried out in air. It was shown that when oxygen is intentionally introduced in the CdTe layer during its growth, this leads to a significant improvement in all the device parameters yielding an efficiency of 14% compared to 11.5% for devices fabricated in the same conditions but without intentional oxygen incorporation in CdTe. The data obtained were not altered following a light soaking. The devices were investigated by quantitative secondary ion mass spectrometry, which allowed insight into the distribution and amount of oxygen and chlorine within the entire device structure. Both impurities showed an increased concentration throughout the CdTe absorber layer.

Role of substrate and transparent conducting oxide in impurity evolvement in polycrystalline thin-film devices

A comparison of as-grown and processed CdTe∕CdS solar cell structures deposited on sapphire substrate has been undertaken with those grown on glass. The device structures were depth-profiled using quantitative secondary ion mass spectrometry. It was shown that while Si concentration profiles are similar to those for structures grown on glass, Na was more than one order of magnitude lower when sapphire was used instead of glass, showing that Na diffused from the glass. It was also found that there was no measurable diffusion of Sn from the SnO2 layer into CdTe, and that the former played an important role in preventing the diffusion of In from In-containing transparent conducting oxide layer. Cl, O, Br, and F species were also investigated and while Cl and O were found to be independent of the nature of the substrate used, Br and F were shown to be affected by the processing.

Impurity analysis of CdCl2 used for thermal activation of CdTe-based solar cells

We present an impurity analysis of cadmium chloride (CdCl2) using inductively coupled plasma mass spectrometry. Six batches of CdCl2 with different nominal purity ranging from 99.999% to 95% were analysed. Each batch turned out to be different in terms of the nature and concentration of impurities. However, all batches generally contained four impurity elements present with concentrations that were characteristic for that element, these being Na (~25 ppm (parts per million)), In (< 1 ppm), Pb (~0.5 ppm) and Te (< 3 ppm), regardless of the nominal purity or the supplier. In all the batches investigated, the major part of the impurities detected (between 83% and 91%) was found to comprise potentially electrically active dopants in CdTe. These impurities are therefore of importance to CdTe/CdS thin film solar cells, where the thermal activation is based on the use of CdCl2.

On the Origins of Impurities in CdTe-Based Thin Film Solar Cells

We present a multi-element study of impurities in CdTe/CdS photovoltaic cells, aimed at understanding their origins and impact on devices. Our investigation was based on calibrated secondary ion mass spectrometry (SIMS) depth profiling, with Na, Zn, Sn, O, Sb, Si, Cl, In, Cu and Pb being studied. The solar cell structures were grown by sputtering and close-space sublimation, and some of them were further processed (CdCl2 anneal and Br2-methanol etch) for the purpose of comparison. Using source materials of different purity allowed us to establish the origins of impurities. We found that some elements increased in concentration upon processing, namely Cl (x100), Na (x10), and In (x1.5). The concentrations of Si, Cu, Zn, Sn and Sb found in a processed device were largely unchanged – and are similar to those found in a high purity single crystal CdTe reference sample.

Effects of Impurities in CdTe/CdS Structures: Towards Enhanced Device Efficiencies

We report on a comprehensive study of the effect of impurities in thin film CdTe/CdS PV structures. Dynamic, quantitative SIMS and ICPMS spectrometry have been used to analyze starting materials, device layers and final device structures. Source materials of different purity were used: CdCl2 (5N to 95%), CdTe (7N to 5N), CdS (4N) and TCO (5N). Structures grown on glass have been compared with those grown on sapphire. Impurity profiles were obtained for O, Na, Si, Cl, Cu, Zn, Sn, Sb, In, Pb, S, Cd and Te. The profiles for Cu, Zn, Sn and Pb were found to be independent of the CdTe starting material and post-growth CdCl2 treatment. Cl, O, Na and Si had higher levels in structures grown with 5N CdTe source material. Interdiffusion of Te and S was enhanced in devices grown with 5N CdTe. In was found to originate from the TCO. Na originated principally from the glass but was also introduced by the use of lower grade CdCl2. The Si profile was independent of the substrate used. Intentional incorporation of 62 during the CdTe growth led to an enhanced concentration of both O and Cl in finished devices. This resulted in an improvement in all device parameters and an increase in efficiency from 11.5% to 14%

Does CdTe Deposition Affect the Impurity Profile in Sputtered CdS Window Layers

We report a multi-element study of impurities in CdS window layers by dynamic and quantitative SIMS. Two CdS/TCO/glass samples, grown separately using nominally the same conditions, were considered. In2O3:F grown on soda lime glass was used as TCO followed by a sputtered CdS layer. One of the samples was subsequently used as a substrate for growth of CdTe by CSS. SIMS was carried out on both samples, and O, Na, Si, Cl, Sb, In, Zn, Sn, Pb, Cu, Te, S and Cd were depth profiled. It was shown that before CdTe growth, most of the impurity elements showed flat levels in the CdS ranging from 2-3◊10 20 cm -3 for Zn and O to 2-3◊10 17 cm -3 for Na, Cl, Sb and Te. Si was found to segregate at the CdS/TCO interface with a maximum level of 10 18 cm -3 . However, following CdTe growth, the impurities in the CdS layer showed higher concentrations and different profile shape compared to those before CdTe growth. Some of the impurities also showed a diffusion-like profile following the CdTe growth as compared to before. Possible explanations of these changes are discussed in terms of the purity of the starting materials and the growth environments, as well as the diffusion from the TCO and glass.

Ion implantation of the window and front contact layers and its effect on polycrystalline photovoltaic devices

Potential dopants sodium and chlorine ions were used to implant CdS window layers, and boron and lead ions were implanted into the transparent conducting oxide (TCO) layers that serve as front contact for CdTe/CdS solar cell devices grown on soda-lime glass substrates. The implantation doses were in the range 1014 to 1015 cm−2 and the peak concentration achieved was 1020 cm−3 for all the species implanted. The distribution of these species was investigated by secondary ion mass spectrometry (SIMS) in the entire device structures using a quantitative approach. The photovoltaic parameters measured on the resulting CdTe/CdS solar cells indicate that the device behaviour is primarily governed by the implantation-induced structural damage of the device structure layers.