Zimeng Cheng

A Study of Light-Sensitive Ideality Factor and Voltage-Dependent Carrier Collection of CdTe Solar Cells in Forward Bias

Two features that distinguish CdS/CdTe solar cells from traditional c-silicon cells have been widely confirmed by experiment: 1) The ideality factor of the diode current is sensitive to photon flux input, and 2) the photocurrent is voltage dependent. Here, numerical simulation of one-sided p-type CdTe junctions has been performed in order to interpret the experimental characteristics under forward bias. In these calculations, we assumed a distribution of trapping states within the CdTe bandgap. Detailed discussions have been divided into two parts: light-dependent forward current and voltage-dependent carrier collection. The former has demonstrated that recombination loss increases with increased optical generation. As a result, the diode current, in a certain forward bias region, depends on illumination. With this mechanism, the ideality factor of simulated J-V curves shows a variation consistent with experimental curves. Carrier collection loss under voltage bias is the combined result of carrier diffusion, drift, and recombination during transport. Simulations that are based on our assumption have demonstrated that collection loss increases at higher forward bias and depends on wavelength, carrier mobility, and carrier lifetime.

Detection of electron emission as DLTS signal in CdTe solar cells

This work identifies an incongruity in the detection of the minority carrier signal in CdTe solar cells during the deep level transient spectroscopy (DLTS) measurement. Use of quasi-Fermi level instead of Fermi level of majority carriers to estimate the probability of emitting carriers seems to correct the ambiguity. During the experiment, signals from minority carrier traps (electron traps) were detected by using a long filling pulse time instead of an electron injection pulse. The DLTS measurements of CdTe solar cells observed a single electron trap with energy level EE1 = 0.47 eV, and two hole traps with energy levels, EH1 = 0.17 eV and EH2 = 0.27 eV. The possibility of any impact from the back contact was excluded, and the phenomenon was clarified by the simulation. It was further observed that when the condition of quasi-Fermi level is considered, the results of calculated probability were significantly different from that of the results that used only Fermi level of majority carriers. The simulation...

Analysis of alloying and built-in voltage in thin film chalcogenide solar cells using modulation spectroscopy

Modulation spectroscopy has been performed on superstrate CdS/CdTe solar cells prepared by close-spaced sublimation. Photoreflectance (PR) measurements were conducted near the CdTe E0 band edge using a variable-wavelength probe beam in conjunction with a chopped pump beam. A fit to the observed PR spectrum was made using two third-derivative, low-field lineshapes. The band gap of the active material near the junction is 1.46 eV (CdTe1-xSx). The envelope function varied as 1 - V/Vbi· The PR spectrum shifted with a temperature coefficient of -0.31 meV/K. Photoreflectance spectroscopy offers a powerful method of determining absorber band gaps, alloying, and built-in voltage in thin-film solar cells.

The growth model of CdTe thin film by close spaced sublimation

Close spaced sublimation (CSS) has attractive features for high-rate deposition of CdS/CdTe thin film solar modules. CSS has been successfully implemented in commercial production of CdTe solar panels with size of 0.6 m × 1.2 m and 1.2 m × 1.6 m. A CSS growth model is a useful tool to guide the CdTe thin film deposition and the design of CSS. In this paper, a growth model was developed for CdTe deposition including diffusion limited, sublimation limited and reaction limited with the presence of oxygen. The model can explain the effect of source and substrate temperature, ambient gas pressure and the separation between source and substrate on the growth rate of CdTe by CSS. The model may prove very useful for optimization of the CSS process with respect to its design variables (pressure, temperature, etc.).

Large area Cds thin film grown by chemical bath deposition

Cadmium Sulfide (CdS) thin films are often deposited on glass substrates coated with TCO layers by the close-spaced sublimation (CSS) or sputter techniques in industrial because in-line production integration. It is seldom reported that CdS is deposited by the chemical bath deposition (CBD) batch process. The bottleneck of CBD for commercial application is its low production rate and waste water treatment. In this paper, we report how to produce different thickness large area (100cm2) CdS thin film by CBD. We study how to increase the production rate. The influence of the solution temperature and concentration on the thickness of CdS is investigated.

CdTe growth model by close spaced sublimation

Close spaced sublimation (CSS) has attractive features for high-rate deposition of CdS/CdTe thin film solar modules. It is necessary to have a solid growth model to explain their growth. In past 20 years, there are several growth models which are conflict each other. In this paper, a growth model was developed for CdTe deposition. The model can explain the effect of source and substrate temperature, ambient gas pressure and the separation between source and substrate on the growth rate of CdTe by CSS.

Injection dependent minority carrier lifetime and defect configuration in thin film CdTe solar cells

Although in CdS/CdTe thin film solar cells, experiments showed that the minority carrier lifetime depends on excess carrier concentration, it is not known that under steady state how the lifetime is affected by the defects. High defect concentration in CdS/CdTe solar cells violates the assumption needed to simplify of the Shockley Read Hall (SRH) recombination equation. Also, with high defect concentrations and thus short carrier lifetime, the excess carrier concentration under illumination can be equivalent to or even lower than the defect concentrations. By simplifying SRH equation considering high defect concentration, it is found that in neutral region, the lifetime is excess carrier dependent. By simultaneously solving the equations of charge neutrality, charge conservation and SRH recombination in the neutral region, the minority carrier lifetime as a function of generation rate can be calculated. The measured minority carrier diffusion length in a CdS/CdTe solar cell, as determined from the steady-state photo-generated carrier collection efficiency, shows the predicted transition of minority carrier lifetime versus optical injection level. A numerical fitting of the indirectly-measured minority carrier lifetime by assuming the minority carrier mobility gives a non-intuitive picture of the p-n junction with a low free hole concentration but a narrow depletion region width.

Evidence for CdTe 1−x S x compound formation in CdTe solar cells from high-precision, temperature-dependent device measurements

High-precision, temperature-dependent device measurements have been made on CdTe solar cells fabricated by close-spaced sublimation on CdS-coated SnO2:F-glass substrates. In particular, the temperature dependence of Voc was measured from 110 K-310 K, taking care that the wavelengths of the incident light did not overlap the absorber band edge. The conventional deduction of the absorber bandgap Eg was clarified by taking into account its temperature dependence. Depending upon fabrication conditions, it was found that the 0 K extrapolated bandgap of the region of the absorber in which most recombination occurs was 1.33 e V, i.e. significantly less than that of CdTe. It was concluded that this region consists of the compound CdTe1-xSx.

Investigation of defects in N + -CDS/P-CdTe solar cells

Two sets of samples (CdTe solar cells) were investigated at -1V within a temperature range of 300K-393K. We discuss Shockley-Read-Hall recombination /generation (SRH R-G) as applied to CdTe and the assumptions used to yield trap energy levels in CdTe. Observed activation energies of the J-V characterization made with Cu-containing back contact in one sample shows one slope while in another sample shows two distinct slopes in Arrhenius plot of ln (J0T-2) vs. 1000/T. Using published identification of the physical trap with energy level, we conclude that one sample does not have hole traps while the other cell shows deep levels corresponding to substitutional impurities of Cu and positive interstitial Cui2+.

The steady state occupancy and Effective Fermi Level of P-N junction

Explicit difference between occupation probability of the transition Gibbs free energy level (TGFEL) of semiconductor under equilibrium and that under steady state is pointed out. Our own computer simulation results are presented to show the detailed steady state occupancy of TGFELs under forward and reverse bias cases.. The result shows that Quasi Fermi levels and capture cross section of defects which is due to columbic interaction or other effects determine the different occupation probability in the region where Quasi Fermi levels are present. An Effective Fermi Level EFeff is defined for these two cases of the semiconductor P-N junction, with the presence of two Quasi Fermi levels.