Chris S. Ferekides

High efficiency CSS CdTe solar cells

Cadmium telluride (CdTe) has long been recognized as a strong candidate for thin film solar cell applications. It has a bandgap of 1.45 eV, which is nearly ideal for photovoltaic energy conversion. Due to its high optical absorption coefficient essentially all incident radiation with energy above its band-gap is absorbed within 1‐2 mm from the surface. Thin film CdTe solar cells are typically heterojunctions, with cadmium sulfide (CdS) being the n-type junction partner. Small area efficiencies have reached the 16.0% level and considerable efforts are underway to commercialize this technology. This paper will present work carried out at the University South Florida sponsored by the National Renewable Energy Laboratory of the United States Department of Energy, on CdTe/CdS solar cells fabricated using the close spaced sublimation (CSS) process. The CSS technology has attractive features for large area applications such as high deposition rates and efficient material utilization. The structural and optical properties of CSS CdTe and CdS films and junctions will be presented and the influence of some important CSS process parameters will be discussed. q 2000 Elsevier Science S.A. All rights reserved.

CdTe solar cells with efficiencies over 15

High efficiency CdTe/CdS thin film solar cells have been prepared on SnO2 coated borosilicate glass substrates. The CdS films are deposited by the Chemical Bath Deposition (CBD) technique from an aqueous solution containing cadmium acetate, ammonium acetate, ammonia and thiourea in the temperature range of 80–90°C. The CdTe films are deposited by the Close Spaced Sublimation (CSS) technique from CdTe powder of 99.999% purity. Doped graphite paste is used as the ohmic contact to the CdTe and indium is used as the contact to the SnO2. Conversion efficiencies of over 15% have been achieved as a result of optimization of a number of processing steps.

RF sputtered back contacts for CdTe/CdS thin film solar cells

One of the critical issues associated with the fabrication of thin film CdTe/CdS solar cells is the formation of a stable, low-resistance back contact. This paper reports on the formation of a back contact to CdTe cells using RF sputtering of a Cu/sub x/Te target. The cells used in this study are of the commonly used superstrate configuration glass/SnO/sub 2//CdS/CdTe. The back contact was formed by depositing a thin Cu/sub x/Te film by RF sputtering followed by the deposition of a metal. The devices were characterized using I-V, C-V, and spectral response measurements. Fill factors in excess of 70% have been obtained using the Cu/sub x/Te/metal contact. The best device measured at the National Renewable Energy Laboratory exhibited a conversion efficiency of 14.9%.

Effects of thermal stressing on CdTe/CdS solar cells

The effects of thermal stressing on CdTe/CdS thin film solar cells have been investigated. Cadmium telluride solar cells fabricated by close-spaced sublimation, and contacted with Cu-based back contacts have been subjected to temperatures ranging from 70 to 120/spl deg/C in inert ambient for over 3500 hours. The average starting V/sub oc/ and ff for all stressed devices were 840 mV and 72% respectively. The devices were periodically removed from the stress environment and their light and dark J-V characteristics were measured. It has been found that all cells exhibited some degree of degradation, which was accelerated with temperature. Changes in device characteristics appeared to be gradual for temperatures in the range of 70-90/spl deg/C. In all cases a significant portion of the observed degradation took place within the first 500-600 hours. SIMS analysis indicated that the stress process resulted in Cu-accumulation near the CdTe/CdS junction, suggesting that Cu is at least partially responsible for the observed junction degradation.

Pulsed UV laser annealing of polycrystalline CdTe

Presented here are the results of a three dimensional, finite element simulation that models pulsed, ultraviolet (UV) laser annealing of polycrystalline CdTe. The model considers heat generated by the absorption of a 25 ns, 248 nm laser pulse normally incident to a 5 μm thick CdTe thin film deposited on a polycrystalline alumina substrate. In particular, focus is on the spatial and temporal distribution of temperature from laser fluences that achieve a sub-melting condition. The model shows that there are very large temperature gradients both in depth and in-plane directions. These predictions, as well as the onset of melting, are confirmed with cross sectional scanning electron microscopy. Additionally, the model predicts that the heat generated dissipates rapidly after the pulse has ended. This has implications if pulse trains are to be used experimentally.

The effects of CdS processing and glass substrates on the performance of CdTe solar cells

Cadmium sulfide films prepared by RF sputtering and close spaced sublimation (CSS) have been used for the fabrication of CdTe/CdS thin film solar cells on borosilicate glass substrates. The CdTe layer was prepared by CSS at high processing temperatures (600/spl deg/C). CdS films prepared by the chemical bath deposition process (CBD) were deposited on tin oxide coated soda lime glass substrates. For these devices the CSS CdTe films were prepared at low substrate temperatures (<550/spl deg/C). Devices prepared at low processing temperatures (CdTe-CSS/CdS-CBD) on soda lime glass substrates exhibited efficiencies in excess of 13% as measured under AM 1.5 conditions at the National Renewable Energy Laboratory.

Cu effects on CdS/CdTe thin film solar cells prepared on flexible substrates

Thin film CdS/CdTe solar cells of the substrate configuration (foil/back contact (BC)/CdTe/CdS/TCO) have been prepared on the molybdenum (Mo)-coated stainless steel substrates. Copper (Cu) is introduced in the cell by briefly immersing the device in a CuCl solution at different stages of the fabrication process in order to study the effect of Cu on cell performance. The structural and morphological properties of the CdTe and CdS films have been investigated. CdS/CdTe solar cells with and without Cu incorporation are characterized. The results suggest that cell performance can be improved by incorporating Cu under certain conditions. Solar cells with efficiencies above 7% have been prepared in this study.

Recent advances in thin film CdTe solar cells

CdTe thin film solar cells have been fabricated on a variety of glass substrates (borosilicate and soda lime). The CdS films were deposited to a thickness of 500–2000 A by the chemical bath deposition (CBD), rf sputtering, or close spaced sublimation (CSS) processes. The CdTe films were deposited by CSS in the temperature range of 450–625u2009°C. The main objective of this work is to fabricate high efficiency solar cells using processes that can meet low cost manufacturing requirements. In an attempt to enhance the blue response of the CdTe cells, ZnS films have also been prepared (CBD, rf sputtering, CSS) as an alternative window layer to CdS. Device behavior has been found to be consistent with a recombination model.

Photocurrent Spectral Distribution and Relaxation in CdS/CdTe Heterojunctions

Photocurrent spectral distribution and relaxation in CdS/CdTe heterojunctions Sergiu Vatavu, Petru Gasin, Iuliana Caraman. Thin film CdS/CdTe solar cells have proved their importance in use as solar energy converters achieving 16.5 % efficiency. Thin film CdS/CdTe heterojunctions have been deposited by close spaced sublimation method (CSS), onto SnO2 covered glass plates (2o2 cm 2 ). The thickness of CdS and CdTe thin films was 0.3-1.6 uf06dm and 2.3-6.6 uf06dm respectively. For to enhance CdS/CdTe heterojunctions photoelectrical parameters and photosensitivity, the deposition procedure was followed by annealing in presence of CdCl2 at 420uf0b0C for 15-60 min. Ni has been used as back contact to CdTe. After annealing, the photoelectrical parameters at 100 mW/nm2 are: Uoc=0.73 V, Isc=21.7 mA/cm 2 , ff=0.43. The analysis of the photosensitivity for samples with different thicknesses of the component layers in the 78-293K temperature range showed, that the increase of the photoconductivity for photon energies ħw>1.9 eV is determined by the generation-recombination processes in CdS x Te 1-x interface layer, and in the 1.42 eV-1.85 eV region, by light absorption mechanism in CdTe. At direct biases of the CdS/CdTe heterojunctions (2Uoc), the photocurrent spectral distribution is determined by the generation-recombination processes in CdS layer. The annealing of the heterojunction in presence of CdCl2 results in photocurrent increase more than one order of magnitude. CdS/CdTe at reverse biases, having a constant sensitivity in visible and near infrared spectral region, can be used in photometry as radiation detector. The analysis of the absorption spectra and the photocurrent spectral distribution at direct biases resulted in determination of the holes mean free path in CdTe layer, being equal to 1.05 mm. The ratio of the ambipolar diffusion coefficient and surface recombination velocity for CdS/CdTe interface is 0.3 uf06dm and the diffusion length is 0.6 uf06dm at 78K. The same parameters for Ni/CdTe interface are: 0.025 uf06dm and 0.05 uf06dm respectively. The phococurrent relaxation curves have been studied for illumination throught both sides of the heterojunctions at different biases, temperatures and wavelength, determining different penetration depths and materials excitation. Nonequilibrium charge carriers lifetime has been determined. The thermal annealing of the CdS/CdTe heterojunctions in presence of CdCl2 determine the nonequilibrium charghe carrier lifetime increase by more than one order of magnitude from 22 uf06ds up to 280 uf06ds.

Junction mechanisms in CdS/CdTe solar cells

Shockley-Read-Hall (SRH) recombination theory has been used to describe the performance of CdTe solar cells. The recombination center profile dominates performance and is found to vary significantly among devices with nominally the same efficiency. For most devices the profile results in a voltage dependent diode factor rather than a single A,J/sub 0/ pair that can describe performance. The observed narrow range for V/sub oc/ of 0.83-0.85 volts for high efficiency devices is attributed to two effects: the containment of the recombination centers between the imrefs and a proposed correlation between V/sub bi/ and the recombination center density.