Search

As thin-film cadmium telluride (CdTe) solar cells gain prominence, one particular challenge is optimizing contacts and their interfaces to transfer charge without losses in efficiency. Back contact recombination is still significant and will prevent CdTe solar technology from reaching its full potential in device efficiency, and transparent back contacts have not been developed for bifacial solar technology or multijunction solar cells. To address these challenges, this study investigates sputtered CuxZn1-xS as a p-type semi-transparent back contact material to thin-film polycrystalline CdTe solar cells at Cu concentrations x = 0.30, 0.45, and 0.60. This material is selected for its high hole conductivity (160-2120 S cm-1), wide optical band gap (2.25-2.75 eV), and variable ionization potential (approximately 6-7 eV) that can be aligned to that of CdTe. We report that without device optimization, CdTe solar cells with these CuxZn1-xS back contacts perform as well as control cells with standard ZnTe:Cu back contacts. We observe no reduction in external quantum efficiency, low contact barrier heights of approximately 0.3 eV, and carrier lifetimes on par with those of baseline CdTe. These cells are relatively stable over one year in air, with VOC and efficiency of the x = 0.30 cell decreasing by only 1 and 3%, respectively. Using scanning electron microscopy and scanning transmission electron microscopy to investigate the CuxZn1-xS/CdTe interface, we demonstrate that the CuxZn1-xS layer segregates into a bilayer of Cu-Te-S and Zn-Cd-S, and thermodynamic reaction calculations support these findings. Despite its bilayer formation, the back contact still functions well. This investigation explains some of the physical mechanisms governing the device stack, inspires future work to understand interfacial chemistry and charge transfer, and elicits optimization to achieve higher-efficiency CdTe cells.

Thin film solar cells often have high densities of dopants and native defects that result in carrier compensation and local electrostatic fields, which produce band edge fluctuations observable as tails in luminescence and absorption spectra. The analysis here shows that these fluctuations can reduce the open-circuit voltage (Voc) by tens to hundreds of millivolts and alter the Voc temperature coefficients. Although the extrapolation of Voc to 0 K is often used to gauge interface recombination, reductions in this value can also be explained by band edge fluctuations. In addition, these fluctuations can explain pronounced anomalous trends between carrier concentration and Voc in state-of-the-art CdSeTe:As solar cells, representing a significant departure from device models that do not include them. The analysis can be applied to other thin film technologies. To improve CdSeTe solar cell performance, it is critical to assess band edge fluctuations and address carrier compensation and defect chemistry to mitigate their impact. [ABSTRACT FROM AUTHOR]

Using two-photon tomography, carrier lifetimes are mapped in polycrystalline CdTe photovoltaic devices. These 3D maps probe subsurface carrier dynamics that are inaccessible with traditional optical techniques. They reveal that CdCl2 treatment of CdTe solar cells suppresses nonradiative recombination and enhances carrier lifetimes throughout the film with substantial improvements particularly near subsurface grain boundaries and the critical buried p-n junction.

Fluctuations refer to inhomogeneity in the distribution of donors and acceptors at the nanometer scale and occur in many compound solar cell materials such as Cu(In,Ga)Se2, Cu2ZnSn(S,Se)4, and CdSexTe1−x. In this work, numerical simulations show that these fluctuations produce not only electrostatic potential variation, but also, local changes in the carrier density and effective bandgap. For a CdSexTe1−x absorber doped with arsenic, simulations and cathodoluminescence data within single grains demonstrate how donor and acceptor densities—consistent with capacitance-voltage and secondary-ion mass-spectrometry data—produce tails in photoluminescence, quantum efficiency, and absorption measurements. Using multiple theoretical approaches, we demonstrate that the fluctuations can hinder expected performance gains from increased carrier density, and we describe the significant open-circuit voltage deficit observed in the CdSexTe1−x:As solar technology. Our results demonstrate that it is critical to characterize and reduce carrier compensation to realize a higher efficiency. [ABSTRACT FROM AUTHOR]

As thin-film and silicon solar technologies mature, questions emerge about the upper bounds of thin-film solar performance and realistic experimental paths to reach them. Directions include increasing absorber hole density and bulk lifetime, improving the junction interface, reducing back-surface recombination, and implementing a back-surface electron reflector. Textbook solutions of idealized p-n junctions create a powerful conceptualization of solar cells as predominantly minority-carrier-driven devices. We demonstrate that thin films are distinct, and models often fail to capture the important role of majority-carrier lifetime, leading to contradictions with lifetime measurements and overestimates of potential device improvement from back-surface passivation and/or reflectors. Furthermore, we identify methods to probe majority-carrier lifetime and re-examine the degree to which back-surface passivation and electron reflectors can increase efficiency for a range of common thin-film interface and absorber properties, using current and emerging CdTe technology as an example. Results indicate that a practical approach is to focus first on improving front-interface recombination velocity and the absorber properties, and then on implementing the back-surface passivation or reflector, which can ultimately allow thin-film solar technology to reach 28% efficiency. [ABSTRACT FROM AUTHOR]

We determine the grain-boundary (GB) recombination velocity, S G B , and grain-interior (GI) lifetime, τ G I , parameters in superstrate CdS/CdTe thin-film solar cell technology by combining cathodoluminescence (CL) spectrum imaging and time-resolved photoluminescence (TRPL) measurements. We consider critical device formation stages, including after CdTe deposition, CdCl2 treatment, and Cu diffusion. CL image analysis methods extract GB and GI intensities and grain size for hundreds of grains per sample. Concurrently, a three-dimensional CL model is developed to simulate the GI intensity as a function of τ G I , S G B , grain size, and the surface recombination velocity, S surf . TRPL measurements provide an estimate of S surf for the CL model. A fit of GI intensity vs. grain size data with the CL model gives a self-consistent and representative set of S G B and τ G I values for the samples: S G B (τ G I) = 2.6 × 106 cm/s (68–250 ps), S G B (τ G I) = 4.1 × 105 cm/s (1.5–3.3 ns), and S G B (τ G I) = 5.5 × 105 cm/s (1.0–3.8 ns) for as-deposited, CdCl2-treated, and CdCl2- and Cu-treated samples, respectively. Thus, we find that the CdCl2 treatment both helps to passivate GBs and significantly increase the GI lifetime. Subsequent Cu diffusion increases GB recombination slightly and has nuanced effects on the GI lifetime. Finally, as a partial check on the S G B and τ G I values, they are input to a Sentaurus device model, and the simulated performance is compared to the measured performance. The methodology developed here can be applied broadly to CdTe and CdSeTe thin-film technology and to other thin-film solar cell materials including Cu(In1-xGax)Se2, Cu2ZnSnS4, and perovskites. [ABSTRACT FROM AUTHOR]

As single-junction silicon solar cells approach their theoretical limits, tandems provide the primary path to higher efficiencies. CdTe alloys can be tuned with magnesium (CdMgTe) or zinc (CdZnTe) for ideal tandem pairing with silicon. A II-VI/Si tandem holds the greatest promise for inexpensive, high-efficiency top cells that can be quickly deployed in the market using existing polycrystalline CdTe manufacturing lines combined with mature silicon production lines. Currently, all high efficiency polycrystalline CdTe cells require a chloride-based passivation process to passivate grain boundaries and bulk defects. This research examines the rich chemistry and physics that has historically limited performance when extending Cl treatments to polycrystalline 1.7-eV CdMgTe and CdZnTe absorbers. A combination of transmittance, quantum efficiency, photoluminescence, transmission electron microscopy, and energy-dispersive X-ray spectroscopy clearly reveals that during passivation, Mg segregates and out-diffuses, initially at the grain boundaries but eventually throughout the bulk. CdZnTe exhibits similar Zn segregation behavior; however, the onset and progression is localized to the back of the device. After passivation, CdMgTe and CdZnTe can render a layer that is reduced to predominantly CdTe electro-optical behavior. Contact instabilities caused by inter-diffusion between the layers create additional complications. The results outline critical issues and paths for these materials to be successfully implemented in Si-based tandems and other applications. [ABSTRACT FROM AUTHOR]

Thin‐film solar cells such as CdTe are a major commercial photovoltaic technology, with more than 25 GW installed worldwide and levelized costs of electricity competitive with fossil fuels. Further progress may result from integrating CdSeyTe1−y absorbers with MgxZn1−xO contacts, but the device efficiency is difficult to maximize due to coupled dependence on chemical composition of both alloys. Herein, a high‐throughput approach is demonstrated to codesign chemical compositions in alloyed MgxZn1−xO/CdSeyTe1−y thin‐film solar cells, using combinatorial libraries of PV devices with orthogonal composition gradients in CdSeyTe1−y absorbers and MgxZn1−xO contacts. It is found that the solar cell performance is a strong and coupled function of both elemental compositions, with efficiency up to 17.7% (VOC = 836 mV, fill factor = 69%, JSC = 30.6 mA cm−2) at atomic compositions of Mg/(Mg + Zn) ≈18% and average Se/(Se + Te) ≈4%. These performance trends among >100 devices are explained by >100 ns lifetime of photoexcited charge carriers at the MgxZn1−xO/CdSeyTe1−y interface where strong Se accumulation is also observed. This study reports the optimal compositions of the commercially relevant MgxZn1−xO/CdSeyTe1−y solar cells and demonstrates a general approach to codesigning performance of alloyed thin‐film solar cells and other optoelectronic devices. [ABSTRACT FROM AUTHOR]

Two-dimensional simulations are performed to investigate the impact of grain boundaries (GBs) on Cu(In, Ga)Se2 (CIGS) solar-cell performance. A downshift of the valance-band energy, as predicted by the observed Cu depletion at CIGS Gbs, can effectively block holes from the GB region and allow efficiencies comparable to GB-free material.

We conducted T = 6K cathodoluminescence (CL) spectrum imaging with a nanoscale electron beam on beveled surfaces of CdTe thin films at the critical stages of standard CdTe solar cell fabrication. We find that the through-thickness CL total intensity profiles are consistent with a reduction in grain-boundary recombination due to the CdCl2 treatment. The color-coded CL maps of the near-band-edge transitions indicate significant variations in the defect recombination activity at the micron and sub-micron scales within grains, from grain to grain, throughout the film depth, and between films with different processing histories. We estimated the grain-interior sulfur-alloying fraction in the interdiffused CdTe/CdS region of the CdCl2-treated films from a sample of 35 grains and found that it is not strongly correlated with CL intensity. A kinetic rate-equation model was used to simulate grain-boundary (GB) and grain-interior CL spectra. Simulations indicate that the large reduction in the exciton band intensity and relatively small decrease in the lower-energy band intensity at CdTe GBs or dislocations can be explained by an enhanced electron-hole non-radiative recombination rate at the deep GB or dislocation defects. Simulations also show that higher GB concentrations of donors and/or acceptors can increase the lower-energy band intensity, while slightly decreasing the exciton band intensity. [ABSTRACT FROM AUTHOR]

We conducted cathodoluminescence (CL) spectrum imaging and electron backscatter diffraction on the same microscopic areas of CdTe thin films to correlate grain-boundary (GB) recombination by GB "type." We examined misorientation-based GB types, including coincident site lattice (CSL) ∑=3, other-CSL (∑=5-49), and general GBs (∑>49), which make up ≥47%-48%, ≥6%-8%, and ≥44%-47%, respectively, of the GB length at the film back surfaces. Statistically averaged CL total intensities were calculated for each GB type from sample sizes of ∼97 GBs per type and were compared to the average grain-interior CL intensity. We find that only ≥16%-18% of ∑=3 GBs are active non-radiative recombination centers. In contrast, all other-CSL and general GBs are observed to be strong non-radiative centers and, interestingly, these GB types have about the same CL intensity. Both as-deposited and CdCl2-treated films were studied. The CdCl2 treatment reduces non-radiative recombination at both other-CSL and general GBs, but GBs are still recombination centers after the CdCl2 treatment. [ABSTRACT FROM AUTHOR]

One of the main limiting factors in CdTe solar cells is its low p-type dopability and, consequently, low open-circuit voltage (VOC). We have systematically studied P and As doping in CdTe with first-principles calculations in order to understand how to increase the hole density. We find that both P and As p-type doping are self-compensated by the formation of AX centers. More importantly, we find that although high-temperature growth is beneficial to obtain high hole density, rapid cooling is necessary to sustain the hole density and to lower the Fermi level close to the valence band maximum (VBM) at room temperature. Thermodynamic simulations suggest that by cooling CdTe from a high growth temperature to room temperature under Te-poor conditions and choosing an optimal dopant concentration of about 1018/cm3, P and As doping can reach a hole density above 1017/cm3 at room temperature and lower the Fermi level to within ~0.1 eV above the VBM. These results suggest a promising pathway to improve the VOC and efficiency of CdTe solar cells. [ABSTRACT FROM AUTHOR]