Search

We report for the first time on direct measurements using atomic force microscopy (AFM) of electric field induced attraction and adhesion forces associated with soiling on photovoltaic (PV) modules. Real dust particles and silica spheres as surrogate to simulating dust particles were glued to AFM probe cantilevers. The electric field induced force (Fes) was measured via AFM force–distance ( f – z ) curves, where the electric field was generated by applying a voltage (Vs) to a simulated PV module. Fes and van der Waals (Fvw) force contributions could be separated with the f – z curves. The results show that Fes ∼2.5 μ N on dust particles is ∼5 times larger than Fvw ∼0.5 μ N at even Vs = −100 V (i.e., similar to large-module operating voltages). Fes increases by an order of magnitude as the applied potential increases from Vs = −100 V to Vs = −500 V. These adhesion forces are by far the strongest that we have measured using the AFM technique for “initial” contact of particles. Furthermore, unlike the more typical short-range forces of Fvw and liquid bridge, Fes extends sub-millimeters to a millimeter beyond the PV module surface, creating large attraction forces to even uncharged dust particles (i.e., via induced dipoles) in the air. These results indicate that the high voltages typically used with PV arrays today will attract more dust particles from the air, hold the dust particles to the surface very strongly, and potentially induce other PV module surface effects, all of which could increase the power production losses because of soiling.

In this paper, we employ transmission electron microscopy coupled with energy-dispersive X-ray spectrometry (EDS) to study the structure and chemistry of cadmium telluride (CdTe) thin-film solar cells with different extents of degradation. The studied regions originate from the same photovoltaic mini-module, which was subjected to one-sun light exposure at a temperature of 100 °C for 400 h to induce light and heat degradation. EDS maps reveal a discontinuous CdS layer and particles rich in O and S but Te-depleted within the CdTe absorber layer after the light and heat stress. These features could act as enhanced recombination centers, resulting in decreased photovoltaic conversion efficiency. Additionally, the most degraded CdTe sample shows a strong accumulation of Na precisely localized in the discontinuous part of the CdS layer. The local Na concentration is determined to be ∼16 at.%. In addition, appreciable Na accumulation is observed at the soda-lime glass/transparent conductive oxide interface in both degraded CdTe solar cells. The microstructure of the baseline CdTe mini-module with no stress was also investigated to demonstrate that the structure difference was caused by the light- and heat-induced degradation, instead of module fabrication. These results highlight the need to control the distribution of Na in fielded CdTe modules to sustain long-term high-power output.

An in situ nonequilibrium method to increase hole density in polycrystalline CdTe thin films to 1016 cm−3 using group V substitution on Te is presented. Single-phase CdTe films doped with P, As, and Sb were deposited at 550 °C at 100–200 nm/s onto moving cadmium sulfide/high resistance transparent buffer layer/transparent conductive oxide /glass superstrates by vapor transport deposition in Cd overpressure from high purity compound sources. Doping levels before and after activation were determined by capacitance-voltage analysis of diagnostic devices. Secondary ion mass spectrometry depth profiling confirmed dopant incorporation levels of 1017–1018 atoms/cm3 in as-deposited films. Electronic activation was carried out by post-deposition annealing in Cd or CdCl2 vapor with fast cooling, increasing acceptor concentrations to >1015 cm−3 for P and >1016 cm−3 for As and Sb, compared with mid −1014 cm−3 acceptor levels for undoped CdTe films. The activation methods are compatible with post-deposition processing presently used for high-efficiency CdTe solar cells. For the dopants As and Sb, the acceptor concentration increased by substitutional AsTe and SbTe formation, respectively, which was validated by cathodoluminescence spectroscopy.

We investigated potential-induced degradation (PID) in CuIn1-xGaxSe2 (CIGS) mini-modules stressed in the laboratory. Small cores were removed from the modules and were subjected to analysis. We completed a proof-of-concept correlative study relating cathodoluminescence to sodium content via time-of-flight secondary-ion mass spectrometry imaging. By comparing one-dimensional depth profile results and three-dimensional tomography results on stressed and unstressed CIGS mini-modules, we can see that PID in CIGS results from sodium migration through absorber, most likely via grain boundaries. Potassium concentration distributions show little change when adding a voltage bias to a temperature and humidity stress. This suggests doping with other large alkali ions, such as cesium and rubidium, rather than sodium can increase the PID resistance of CIGS modules.

We report on measurements of junction location in Cu(In,Ga)Se2 (CIGS) solar cells with different window-layer materials by nm-resolution electrical potential/field profiling across the junction using Kelvin probe force microscopy imaging on cross-section of the devices. The results illustrate that the device with a CdS window layer (CdS/CIGS) has a buried homojunction located inside the CIGS absorber with ∼40-nm junction depth, whereas the ZnOS/CIGS devices with and without partial electrolyte treatment prior to the window-layer deposition are similar, exhibiting a heterointerface junction. This junction location may contribute in part to the highest efficiency of the CdS/CIGS device among the three devices.

In this contribution, we probe spatial variations in charge-carrier recombination in CuInxGa1-x Se2 (CIGS) absorbers grown on soda–lime glass (SLG) and alkali-free sapphire substrates with NaF and KF postdeposition treatments (PDTs). Temperature-and illumination-dependent device measurements are used to track interface recombination and recombination in the quasi-neutral region. The analysis of these data reveals that the benefit of alkali PDTs depends on the substrate: interface recombination is reduced in devices grown on sapphire substrates, whereas recombination in the quasi-neutral regions is reduced in devices grown on SLG substrates. Cathodoluminescence (CL) spectrum imaging is used to study the spatial distribution of recombination with respect to the grain structure. The grain-boundary CL contrast is similar in films with no PDT, NaF PDT, or KF PDT. A reduced grain-boundary contrast is observed with a NaF + KF PDT; however, suggesting a reduced recombination strength at the grain boundaries (GBs) for combined NaF + KF treatment. CL spectra indicate band tailing, consistent with the fluctuating potential model. Fluctuating potentials are believed to reduce open-circuit voltage, but their spatial distribution has not been studied. Here, CL spectrum imaging data are used to generate maps of the root-mean-square value of the potential energy fluctuations—γ. These maps reveal a bimodal γ distribution for all samples: γ is generally in the range ∼15–50 meV or ∼100–180 meV. The higher γ range is more significantly affected by the PDTs; after the PDTs, it is strongly correlated with GBs. The lower γ range is correlated with higher emission intensity regions, typically grain interiors, and increases in area fraction after the PDTs. These results demonstrate how spatially resolved luminescence and device characterization measurements can be used to monitor changes in recombination in CIGS films and photovoltaic devices. Such measurements can complement empirical device optimization and help improve device performance.

Grain structure influences both transport and recombination in CdTe solar cells. Larger grains generally are obtained with higher deposition temperatures, but commercially it is important to avoid softening soda-lime glass. Furthermore, depositing at lower temperatures can enable different substrates and reduced cost in the future. We examine how initial deposition temperatures and morphology influence grain size and lifetime after CdCl2 recrystallization. Techniques are developed to estimate grain distribution quickly with low-cost optical microscopy, which compares well with electron backscatter diffraction data providing corroborative assessments of exposed CdTe grain structures. Average grain size increases as a function of CdCl2 temperature. For lower temperature close-spaced sublimation CdTe depositions, there can be more stress and grain segregation during recrystallization. However, the resulting lifetimes and grain sizes are similar to high-temperature CdTe depositions. The grain structures and lifetimes are largely independent of the presence and/or interdiffusion of Se at the interface, before and after the CdCl2 treatment.

We investigated potential-induced degradation (PID) in silicon mini-modules that were subjected to accelerated stressing to induce PID conditions. Shunted areas on the cells were identified with photoluminescence and dark lock-in thermography (DLIT) imaging. The identical shunted areas were then analyzed via time-of-flight secondary-ion mass spectrometry (TOF-SIMS) imaging, 3-D tomography, and high-resolution transmission electron microscopy. The TOF-SIMS imaging indicates a high concentration of sodium in the shunted areas, and 3-D tomography reveals that the sodium extends more than 2 μm from the surface below shunted regions. Transmission electron microscopy investigation reveals that a stacking fault is present at an area identified as shunted by DLIT imaging. After the removal of surface sodium, tomography reveals persistent sodium present around the junction depth of 300 nm and a drastic difference in sodium content at the junction when comparing shunted and nonshunted regions.

Advancing CdTe solar cell efficiency requires improving the open-circuit voltage $\rm{(V_{{\rm{OC}}})}$ above 900 mV. This requires long carrier lifetime, high hole density, and high-quality interfaces, where the interface recombination velocity is less than about 104 cm/s. Using CdTe single crystals as a model system, we report on CdTe/CdS electrical and structural interface properties in devices that produce open-circuit voltage exceeding 950 mV.

We present a 2-D numerical model simulating cathodoluminescence (CL) measurements on CdTe. The model is used to analyze the impact of material parameters on the measured CL intensity to establish when grain-boundary (GB) recombination velocity $S_{{\rm GB}}$ can be determined accurately from CL contrast. In addition to GB recombination, grain size and its ratio to the carrier diffusion length can impact CL measurements. Holding the grain interior and GB recombination rates constant, we find that as the grain size increases and exceeds the diffusion length, the observed CL contrast increases. For small-grain-size material, surface recombination lowers the overall intensity of the CL signal but does not significantly impact CL contrast. For large grains, high-surface recombination velocity can decrease the CL contrast. The model is combined with experimental results to quantify the $S_{{\rm GB}}$ in polycrystalline CdTe before and after the CdCl2 treatment and to predict the impact of GB recombination on device performance.

Kesterite Cu2ZnSnS4 (CZTS) thin films were grown by the sulfurization of stacked metal precursors deposited using radio-frequency magnetron sputtering on Mo-coated soda-lime glass substrates. In this paper, we report the role of the film chemical composition in the evolution of ${\hbox{Cu}}_{2-x}{\hbox{S}}$ phases and how to avoid their development through controlling the film composition. Furthermore, the effect of the elemental concentration on the structural and morphological properties of the final CZTS films has been investigated. The prepared CZTS films have a composition ratio M = Cu/(Zn + Sn) varying from 0.81 (Cu-poor) to 1.05 (Cu-rich). X-ray diffraction and Raman scattering studies revealed the presence of ${\hbox{Cu}}_{2-x}{\hbox{S}}$ phases in films with a Cu/(Zn + Sn) ratio higher than 1.00 and/or in films with a Sn/Cu ratio close to or less than the stoichiometric value of 0.50. However, ${\hbox{Cu}}_{2-x}{\hbox{S}}$ -phases-free CZTS films were achieved with Sn/Cu ratios sufficiently above 50% without regard to the Cu/(Zn + Sn) ratio. Plan and cross-sectional scanning electron microscopy showed compact films, in general. Electron back-scattered diffraction revealed randomly oriented CZTS films.

We report on scanning spreading resistance microscopy on cross sections of thin-film CdTe devices. The results show the capability of identifying the multiple layers, the depletion region, and the nonuniform doping. We observe carrier injection and depletion region movement by laser illumination or by electrically biasing the device, directly revealing the underlying physics of the solar cell junction in real space with resolutions of nanometer scale.

We used low-temperature cathodoluminescence (CL) spectrum imaging (CLSI) with nanoscale spatial resolution to examine charge-carrier recombination and defects at grain boundaries (GBs) and grain interiors (GIs) in as-deposited and CdCl 2 -treated CdTe thin films. Supporting time-resolved photoluminescence, T = 4 K photoluminescence, secondary ion mass spectrometry, and electron backscatter diffraction measurements were conducted on the same films. Color-coded maps of the luminescence transition energies (photon energy maps) were used to analyze the qualitative characteristics of the CLSI data. We applied an image analysis algorithm to the pixels in grayscale CL intensity images to compare the luminescence intensities and spectra at the GIs and GBs quantitatively and with statistical relevance. Our results show that GBs in as-deposited films are active recombination centers and are thus harmful to solar cell operation. CL GB defect contrast is quantifiably reduced for the CdCl 2 -treated film, which is direct evidence of passivation of deep GB core states resulting from the treatment. However, the CdCl 2 treatment is not a perfect fix for GB recombination, and GB recombination may still be limiting performance in CdCl 2 -treated devices.

A set of neighboring multicrystalline silicon wafers has been processed through different steps of solar cell manufacturing and then images were collected for characterization. The imaging techniques include band-to-band photoluminescence (PL), defect-band or subbandgap PL (subPL), and dark lock-in thermography (DLIT). Defect regions can be tracked from as-cut wafers throughout processing to the finished cells. The finished cell's defect regions detected by band-to-band PL imaging correlate well to diffusion length and quantum efficiency maps. The most detrimental defect regions, type A, also correlate well to reverse-bias breakdown areas as shown in DLIT images. These type A defect regions appear dark in band-to-band PL images, and have subPL emissions. The subPL of type A defects shows strong correlations to poor cell performance and high reverse breakdown at the starting wafer steps (as-cut and textured), but the subPL becomes relatively weak after antireflection coating (ARC) and on the finished cell. Type B defects are regions that have lower defect density but still show detrimental cell performance. After ARC, type B defects emit more intense subPL than type A regions; consequently, type B subPL also shows better correlation to cell performance at the starting wafer steps rather than at the ARC process step and in the finished cell.

We have investigated the microelectrical properties of CdTe thin films using scanning Kelvin probe force microscopy (SKPFM) and scanning spreading resistance microscopy (SSRM). Two films with the configurations of substrate and superstrate were subjected to the characterization studies. The electrical potential and resistance were properly mapped with the substrate film but not with the superstrate film because the underlying CdS/CdTe junction largely impacted the characterizations. The higher SKPFM potential on grain boundaries (GBs) of the substrate film than on the grain surface indicates positively charged GBs and upward band bending around the GB; therefore, the GBs are either depleted or inverted. The SSRM resistance mapping on this film shows nonuniformities and features that are associated with the grain structure and facets. However, the GBs do not exhibit distinct characteristic resistance. Comparing the low resistance channel along the GBs of high-performance CIGS films, the SSRM mapping of CdTe supports depletion of the GBs. In SSRM measurement, it is critical to adequately indent the probe to the film, and to apply a bias voltage larger than the onset voltage of the probe/film barrier, so that the contact resistance is minimized and that the local spreading resistance of CdTe film beneath the probe is measured.

At the CdTe/CdS interface, a significant Te-S interdiffusion has been found a few nanometers into the CdTe grain interiors with scanning transmission electron microscopy (STEM) and electron energy loss spectroscopy. This interdiffusion happens on both as-grown and CdCl2 -treated CdTe. S substitution at Te sites has been directly resolved in CdTe with STEM Z-contrast images, which further confirms the S diffusion into CdTe grain interiors. Moreover, when a sufficient amount of S substitutes for Te, a structural transformation from zinc-blende to wurtzite has been observed. In the CdCl2 treated CdTe, Cl segregation has also been found at the interface. STEM electron-beam-induced current shows that the p-n junction occurs a few namometers into the CdTe grains, which is consistent with the S diffusion range we observe. The shift of the p-n junction suggests a buried homojunction which would help reduce nonradiative recombination at the junction. Meanwhile, long-range S diffusion in CdTe grain boundaries (GBs) has been detected, as has Te and Cl diffusion in CdS GBs.