Browsing by Author "Sampath, W. S., committee member"
Now showing 1 - 8 of 8
Results Per Page
Sort Options
Item Open Access Characterization and modeling of CdCl2 treated CdTe/CdS thin-film solar cells(Colorado State University. Libraries, 2010) Maxwell, Graham Lane, author; Manivannan, Venkatesan, advisor; Sampath, W. S., committee member; Sites, James R., committee memberCdTe photovoltaic technology has the potential to become a leading energy producer in the coming decades. Its physical properties are well suited for photovoltaic energy conversion. A key processing step in the production of high efficiency CdTe/CdS solar cells is a post-CdTe deposition heat treatment with CdCl2, which can improve performance by promoting CdTe recrystallization, QE response, defect passivation and others. Understanding the effects of the CdCl2 treatment is crucial in order to optimize processing conditions and improve performance. This study investigates the effects of variations of CdCl2 treatment duration on CdTe/CdS solar cells manufactured at Colorado State University. In order to investigate the optimal time of CdCl2 treatment, sample solar cells were tested for microstructural and performance properties. Device microstructure was analyzed using X-ray diffraction (XRD), scanning electron microscopy (SEM), and atomic force microscopy (AFM). Device performance was analyzed using current density-voltage (J-V) measurements, time-resolved photoluminescence (TRPL), quantum efficiency (QE), and laser beam induced current (LBIC) measurements. Little change in microstructure was observed with extended CdCl2 treatment and is attributed to the high CdTe deposition temperatures used by heat pocket deposition (HPD). This deposition technique allows for large initial grains to be formed with low lattice strain energy which prevents recrystallization and grain growth that is often seen with other deposition techniques. The CdCl2 treatment initially improves performance significantly, but it was shown to that extending the CdCl2 treatment can reduce performance. Overall performance was reduced despite an increase in minority carrier lifetime values. The mechanism of reduced performance is suggested to be the formation of a low bandgap CdTe layer resulting from sulfur diffusion from the CdS layer. Sulfur diffusion primarily occurs during the CdCl2 treatment and also leads to thinning of the CdS layer. Solar cell modeling was employed to investigate possible mechanisms for performance degradation. Modeling was done with AMPS and SCAPS modeling software. Models were created to investigate the effects of minority carrier lifetime, CdS thickness, and a low bandgap CdTe layer. Modeling results showed that the formation of a low bandgap CdTe layer combined with CdS thinning reduces device performance. Further research is needed using a statistically significant number of samples to investigate other possible degradation mechanisms associated with extended CdCl2 treatment.Item Open Access Comparative analysis of Cu(InGa)Se2 solar cells(Colorado State University. Libraries, 2016) Counts, Kahl, author; Sites, James, advisor; Sampath, W. S., committee member; de la Venta, Jose, committee memberCu(InGa)Se2, often abbreviated CIGS, photovoltaics have proven to be a commercially viable solar-energy conversion technology. Diverse processes have been employed in the manufacture, with varying end products, most resulting in high efficiency. A collaborative project was undertaken with several CIGS labs and industrial partners to explore the different electrical and spatial characteristics of CIGS solar cells relative to one another. Characterization methods utilized include, current-voltage measurements, quantum efficiency, capacitance-frequency and capacitance-voltage, electroluminescence, light-beam-induced current and Auger profling. Specific parameters for each cell were extracted from the measurements. Together the methods used are a tool for understanding device performance and optimization. Efforts were made to identify strengths, similarities and differences and to connect processing details with observed characteristics.Item Open Access Design strategies for high-efficiency CdTe solar cells(Colorado State University. Libraries, 2017) Song, Tao, author; Sites, James R., advisor; Kanevce, Ana, committee member; Gelfand, Martin, committee member; Wu, Mingzhong, committee member; Sampath, W. S., committee memberWith continuous technology advances over the past years, CdTe solar cells have surged to be a leading contributor in thin-film photovoltaic (PV) field. While empirical material and device optimization has led to considerable progress, further device optimization requires accurate device models that are able to provide an in-depth understanding of CdTe device physics. Consequently, this thesis is intended to develop a comprehensive model system for high-efficiency CdTe devices through applying basic design principles of solar cells with numerical modeling and comparing results with experimental CdTe devices. Four key topics about high-efficiency CdTe cells are covered in this dissertation: (a) material optimization of CdTe absorber, (b) roles of emitter/absorber interface on carrier transport, (c) substrate choices for monocrystalline CdTe cells, and (d) back contact configurations for thin-film polycrystalline CdTe cells. Finally, comparisons between simulation and experiment are carried out to identify both beneficial and detrimental mechanisms for CdTe cell performance and to guide future cell optimization. The CdTe absorber is central to cell performance. Numerical simulation has shown the feasibility of high energy-conversion efficiency (open-circuit voltage VOC > 1000 mV, efficiency η > 25%), which requires both high carrier density (p >1016 cm-3) and long minority carrier lifetime (τn > 100 ns). As the minority carrier lifetime increases (τn > 10 ns), the carrier recombination at the back surface becomes a limitation for cell performance with absorber thickness < 3 µm. Hence, either a thicker absorber or an appropriate back-surface-field layer is a requisite for reducing the back-surface recombination. When integrating layers into devices, more careful design of interfaces is needed. One consideration is the emitter/absorber interface. It is shown that a positive conduction-band offset ΔEC ("spike") at the interface is beneficial to cell performance, since it can induce a large valence-band bending which suppresses the hole injection near the interface for the electron-hole recombination, but too large a spike is detrimental to photocurrent transport. In a heterojunction device with many defects at the emitter/absorber interface (high SIF), a thin and highly-doped emitter can induce strong absorber inversion and hence help maintain good cell performance. Performance losses from acceptor-type interface defects can be significant when interface defect states are located near mid-gap energies. In terms of specific emitter materials, the calculations suggest that the (Mg,Zn)O alloy with 20% Mg, or a similar type-I heterojunction partner with moderate ΔEC (e.g., Cd(S,O) or (Cd,Mg)Te with appropriate oxygen or magnesium ratios) should yield higher voltages and would therefore be better candidates for the CdTe-cell emitter. The CdTe/substrate interface is also of great importance, particularly in the growth of epitaxial monocrystalline CdTe cells. Several substrate materials (CdTe, Si, GaAs, and InSb) have been discussed and all have challenges. These have generally been addressed through the addition of intermediate layers between the substrate and CdTe absorber. InSb is an attractive substrate choice for CdTe devices, because it has a close lattice match with CdTe, it has low resistivity, and it is easy to contact. However, the valence-band alignment between InSb and p-type CdTe, which can both impede hole current and enhance forward electron current, is not favorable. Three strategies to address the band-offset problem are investigated by numerical simulation: (a) heavy doping of the back part of the CdTe layer, (b) incorporation of an intermediate CdMgTe or CdZnTe layer, and (c) formation of an InSb tunnel junction. Each of these strategies is predicted to be helpful for higher cell performance, but a combination of them should be most effective. In addition, the CdTe/back contact interface plays a significant role in carrier transport for conventional polycrystalline thin-film CdTe devices. A significant back-contact barrier φb caused by metallic contact with low work function can block hole transport and enhance the forward current and thus result in a reduced VOC, particularly with fully-depleted CdTe devices. A buffer contact layer between CdTe absorber and metallic contact is strongly needed to mitigate this detrimental impact. The simulation has shown that a thin tellurium (Te) buffer as well as a highly doped p-type CdTe layer can assume such a role by reducing the downward valence-band bending caused by large φb and hence enhancing the extraction of the charge carriers. Finally, experimental CdTe cells are discussed in parallel with the simulation results to identify limiting mechanisms and give guidance for future efficiency improvement. For the monocrystalline CdTe cells made at NREL, it is found that the sputter damage causing large numbers of defect states near the Cd(S,O)/CdTe interface plays an important role in limiting cell performance, particularly for cells with low oxygen Cd(S,O) (with a "cliff" band offset). Other effects, such as the large series resistance and reflection, also reduce the cell performance. A lattice-matched material with less deposition damage and with a type-I interface is suggested to introduce less interfacial recombination in future emitter growth on epitaxial CdTe absorbers. For polycrystalline CdTe solar cells made at CSU, it is demonstrated that an MZO emitter forms a spike at the MZO/CdTe interface and a Te buffer layer mitigates large back-contact barrier φb. Both play very important roles in achieving good cell performance (VOC ~ 860 mV, η ~ 18.3%). The simulation has also shown that the electron reflector would be an effective approach to further increase VOC even with a relative low CdTe carrier concentration (~1014 cm-3).Item Open Access Device characterization of cadmium telluride photovoltaics(Colorado State University. Libraries, 2014) Geisthardt, Russell M., author; Sites, James, advisor; Gelfand, Martin, committee member; Topič, Marko, committee member; Sampath, W. S., committee memberThin-film photovoltaics have the potential to make a large impact on the world energy supply. They can provide clean, affordable energy for the world. Understanding the device physics and behavior will enable increases in efficiency which will increase their impact. This work presents novel approaches for evaluating efficiency, as well as a set of tools for in-depth whole-cell and uniformity characterization. The understanding of efficiency losses is essential for reducing or eliminating the losses. The efficiency can be characterized by a breakdown into three categories: solar spectrum, optical, and electronic efficiency. For several record devices, there is little difference in the solar spectrum efficiency, modest difference in the optical efficiency, and large difference in the electronic efficiency. The losses within each category can also be further characterized. The losses due to the broad solar spectrum and finite temperature are well understood from a thermodynamic physics perspective. Optical losses can be fully characterized using quantum efficiency and optical measurements. Losses in fill factor can be quantified from series and shunt resistance, as well as the expected fill factor from the measured V oc and A. Open-circuit voltage losses are the most significant, but are also be the hardest to understand, as well as the most technology-dependent. Characterization of the whole cell helps to understand the behavior, performance, and properties of the cell. Several different tools can be used for whole-cell characterization, including current-voltage, quantum efficiency, and capacitance measurements. Each of these tools give specific information about the behavior of the cell. When combined, they can lead to a more complete understanding of the cell performance than when taken individually. These tools were applied to several specific CdTe experiments. They have helped to characterize the baseline performance of both the deposition tool and the measurement systems. Characterization of plasma-cleaned cells show an improvement in performance, even at thinner CdS layer thickness. Measurements of thinning CdTe samples reveal additional optical losses, likely caused by the increasing importance of the back diode. Characterization of Cd(S,O) devices show improved performance, both from improved optical properties and theorized improvement in band alignment properties. Uniformity can have an effect on whole-cell performance, but can also be an important parameter to characterize on its own. Light-beam-induced current is a powerful tool for characterizing uniformity. The LBIC tool was upgraded to improve its accuracy, functionality, and speed. The improved LBIC system aids in the collection of uniformity data. A number of parameters can be varied to provide in-depth uniformity information and help identify causes of nonuniformity. The wavelength can be varied to provide information on different layers. This can help identify variations in CdS thickness and local CdTe band gap. An applied voltage bias can be used to identify locations with weak diode properties. The resolution can also be varied to provide information on nonuniformities at different scales, from variations across the whole cell to variations on the size of several grains. LBIC can also be paired with electroluminescence to create a powerful nonuniformity characterization suite. The two can be paired with EL used as a screening tool to identify cells or areas which need further characterization from LBIC.Item Open Access Imaging as characterization techniques for thin-film cadmium telluride photovoltaics(Colorado State University. Libraries, 2014) Zaunbrecher, Katherine, author; Sites, James R., advisor; Gelfand, Martin, committee member; Buchanan, Kristen, committee member; Sampath, W. S., committee memberThe goal of increasing the efficiency of solar cell devices is a universal one. Increased photovoltaic (PV) performance means an increase in competition with other energy technologies. One way to improve PV technologies is to develop rapid, accurate characterization tools for quality control. Imaging techniques developed over the past decade are beginning to fill that role. Electroluminescence (EL), photoluminescence (PL), and lock-in thermography are three types of imaging implemented in this study to provide a multifaceted approach to studying imaging as applied to thin-film CdTe solar cells. Images provide spatial information about cell operation, which in turn can be used to identify defects that limit performance. This study began with developing EL, PL, and dark lock-in thermography (DLIT) for CdTe. Once imaging data were acquired, luminescence and thermography signatures of non-uniformities that disrupt the generation and collection of carriers were identified and cataloged. Additional data acquisition and analysis were used to determine luminescence response to varying operating conditions. This includes acquiring spectral data, varying excitation conditions, and correlating luminescence to device performance. EL measurements show variations in a cell's local voltage, which include inhomogeneities in the transparent-conductive oxide (TCO) front contact, CdS window layer, and CdTe absorber layer. EL signatures include large gradients, local reduction of luminescence, and local increases in luminescence on the interior of the device as well as bright spots located on the cell edges. The voltage bias and spectral response were analyzed to determine the response of these non-uniformities and surrounding areas. PL images of CdTe have not shown the same level of detail and features compared to their EL counterparts. Many of the signatures arise from reflections and severe inhomogeneities, but the technique is limited by the external illumination source used to excite carriers. Measurements on unfinished CdS and CdTe films reveal changes in signal after post-deposition processing treatments. DLIT images contained heat signatures arising from defect-related current crowding. Forward- and reverse-bias measurements revealed hot spots related to shunt and weak-diode defects. Modeling and previous studies done on Cu(In,Ga)Se2 thin-film solar cells aided in identifying the physical causes of these thermographic and luminescence signatures. Imaging data were also coupled with other characterization techniques to provide a more comprehensive examination of nonuniform features and their origins and effects on device performance. These techniques included light-beam-induced-current (LBIC) measurements, which provide spatial quantum efficiency maps of the cell at varying resolutions, as well as time-resolved photoluminescence and spectral PL mapping. Local drops in quantum efficiency seen in LBIC typically corresponded with reductions in EL signal while minority-carrier lifetime values acquired by time-resolved PL measurements correlate with PL intensity.Item Open Access Performance and metastability of CdTe solar cells with a Te back-contact buffer layer(Colorado State University. Libraries, 2017) Moore, Andrew, author; Sites, James, advisor; Krueger, David, committee member; de la Venta, Jose, committee member; Sampath, W. S., committee memberThin-film CdTe photovoltaics are quickly maturing into a viable clean-energy solution through demonstration of competitive costs and performance stability with existing energy sources. Over the last half decade, CdTe solar technology has achieved major gains in performance; however, there are still aspects that can be improved to progress toward their theoretical maximum efficiency. Perhaps equally valuable as high photovoltaic efficiency and a low levelized cost of energy, is device reliability. Understanding the root causes for changes in performance is essential for accomplishing long-term stability. One area for potential performance enhancement is the back contact of the CdTe device. This research incorporated a thin-film Te-buffer layer into the contact structure, between the CdTe and contact metal. The device performance and characteristics of many different back contact configurations were rigorously studied. CdTe solar cells fabricated with the Te-buffer contact showed short-circuit current densities and open-circuit voltages that were on par with the traditional back-contacts used at CSU. However, the Te-buffer contact typically produced ~2% larger fill-factors on average, leading to greater conversation efficiency. Furthermore, using the Te buffer allowed for incorporation of ~50% less Cu, which is used for p-type doping but is also known to decrease lifetime and stability. This resulted in an additional ~3% fill-factor gain with no change in other parameters compared to the standard-Cu treated device. In order to better understand the physical mechanisms of the Te-buffer contact, electrical and material properties of the Te layer were extracted and used to construct a simple energy band diagram. The Te layer was found to be highly p-type (>1018 cm-3) and possess a positive valence-band offset of 0.35-0.40 eV with CdTe. An existing simulation model incorporating the Te-layer properties was implemented and validated by comparing simulated results of CdTe device performance to experimental values. The Te layer improves performance is attributed to a reduction in the downward energy band bending between the CdTe and typical contact metals. The stability, or rather the metastability, of CdTe solar cells was also studied with a focus on the Te back contact. A metastable device has a series of quasi-stable local energy-minimuma which the device may transition among. This work primarily focused on changes, both beneficial and detrimental, caused by diffusion and drift of atoms in the CdTe lattice. As atoms moved and/or became ionized their defect states were shifted, which resulted in changes in the CdTe doping and recombination. Changes in performance for devices in equilibrium and under stress conditions were analyzed by electrical and material characterization. Mobile impurities and mechanisms responsible for the changes were identified---primarily the migration of interstitial Cu and Cl. The stability of CdTe solar cells with different back contacts were compared. It was found that any contact that included the Te layer was almost always more stable than the traditional contact used at CSU, most likely because of less sensitivity to the impurity profiles in the CdTe. Moreover, the Te contact configuration that introduced the least amount of Cu into the CdTe was discovered to be the most stable, both in storage and under stress conditions.Item Open Access Processing and characterization of thin cadmium telluride solar cells(Colorado State University. Libraries, 2017) Wojtowicz, Anna, author; Sites, James R., advisor; Sampath, W. S., committee member; de la Venta, José, committee memberCadmium telluride (CdTe) has the highest theoretical limit to conversion efficiency of single-junction photovoltaic (PV) technologies today. However, despite a maximum theoretical open-circuit voltage of 1.20 V, record devices have historically had voltages pinned around only 900 mV. Voltage losses due to high recombination rates remains to be the most complex hurdle to CdTe technology today, and the subject of on-going research in the physics PV group at Colorado State University. In this work, an ultrathin CdTe device architecture is proposed in an effort to reduce bulk recombination and boost voltages. By thinning the CdTe layer, a device's internal electric field extends fully towards the back contact. This quickly separates electrons-hole pairs throughout the bulk of the device and reduces overall recombination. Despite this advantage, very thin CdTe layers also present a unique set of optical and electrical challenges which result in performance losses not as prevalent in thicker devices. When fabricating CdTe solar cells, post-deposition treatments applied to the absorber layer are a critical step for achieving high efficiency devices. Exposure of the polycrystalline CdTe film to a chlorine species encourages the passivation of dangling bonds and larger grain formation, while copper-doping improves device uniformity and voltages. This work focuses on experiments conducted via close-space sublimation to optimize CdCl2 and CuCl treatments for thin CdTe solar cells. Sweeps of both exposure and anneal time were performed for both post-deposition treatments on CdTe devices with 1.0 μm absorber layers. The results demonstrate that thin CdTe devices require substantially less post-deposition processing than standard thicker devices as expected. Additionally, the effects of CdTe growth temperature on thin devices is briefly investigated. The results suggest that higher growth temperatures lead to both electrical and stoichiometric changes in CdTe closely associated with lower carrier lifetimes and poorer overall performance.Item Open Access Spectroscopic ellipsometry as a process control tool for manufacturing cadmium telluride thin film photovoltaic devices(Colorado State University. Libraries, 2013) Smith, Westcott P., author; Kirkpatrick, Allan T., advisor; James, Susan, advisor; Puttlitz, Christian, committee member; Sampath, W. S., committee member; Wu, Mingzhong, committee memberIn recent decades, there has been concern regarding the sustainability of fossil fuels. One of the more promising alternatives is Cadmium Telluride (CdTe) thin–film photovoltaic (PV) devices. Improved quality measurement techniques may aid in improving this existing technology. Spectroscopic ellipsometry (SE) is a common, non-destructive technique for measuring thin films in the silicon wafer industry. SE results have also been tied to properties believed to play a role in CdTe PV device efficiency. A study assessing the potential of SE for use as a quality measurement tool had not been previously reported. Samples of CdTe devices produced by both laboratory and industrial scale processes were measured by SE and Scanning Electron Microscopy (SEM). Mathematical models of the optical characteristics of the devices were developed and fit to SE data from multiple angles and locations on each sample. Basic statistical analysis was performed on results from the automated fits to provide an initial evaluation of SE as a quantitative quality measurement process. In all cases studied, automated SE models produced average stack thickness values within 10% of the values produced by SEM, and standard deviations for the top bulk layer thickness were less than 1% of the average values.