Browsing by Author "Mitzi, David B"

Multinary chalcogenide semiconductors have long been a mainstay within optoelectronics industries. Chalcogenide materials consisting of at least four elements—i.e., quaternaries—have tunable structural, optical, and electronic properties, allowing the semiconductors to be tailored for specific applications. Recently, the I2-II-IV-X4 (I = Li, Cu, Ag; II = Ba, Sr, Pb, Eu; IV = Si, Ge, Sn; X = S, Se) family of materials has emerged as a source of promising semiconductors with applications in nonlinear optics and optoelectronics. A major challenge for these complex compounds is maintaining the ability to predictably control the desired properties. In this dissertation, solid-state chemistry methods are used to tackle three major goals: to investigate known I2-II-IV-X4-type compounds that have not been thoroughly explored; to develop predictable property trends within the wider family of materials; and to predict and make brand new semiconductors.

The study of the quaternary semiconductor Cu2BaGeSe4 and the mixing of Ge and Sn within this compound (to make Cu2BaGe1-xSnxSe4) serves as the platform for branching into new compounds in the I2-II-IV-X4 family. Using the structural analysis established in the work on Cu2BaGe1-xSnxSe4, a structural tolerance factor is developed to predict the probable crystal structure of hypothetical compounds that fit into the family of the I2-II-IV-X4-type materials. Four new semiconductors (Cu2PbGeS4, Cu2SrSiS4, Ag2SrSiS4, and Ag2SrGeS4) were made and found to conform to the anticipated crystal structures based on the structural tolerance factor. The newly synthesized Cu2SrSiS4, Ag2SrSiS4, and Ag2SrGeS4 are potential nonlinear optical materials, while each of the four semiconductors may be used as buffer or n-type layers in thin film solar cells. In the pursuit of new I2-II-IV-X4-type compounds, a family of cubic compounds is found to either co-exist and compete with the synthesized materials (Ag2Sr3Si2S8 & Ag2Sr3Ge2S8) or to be more stable than the hypothetical I2-II-IV-X4 materials (Ag2Pb3Si2S8 & Ag2Sr3Sn2S8) with the same elemental makeup. Of these four cubic semiconductors, Ag2Sr3Si2S8 and Ag2Pb3Si2S8 have not been reported by others. The compounds within this family have predictable trends of optical properties and have applications as nonlinear optical materials if large single crystals can be synthesized.

Finally, the solvothermal synthesis and properties of Ag2(NH4)AsS4 are explored, extending the scope of this dissertation from the I2-II-IV-X4 materials to those of the form Ag2-I’-V-X4 (I’ = NH4, K, Rb, Cs; V = As, Sb, Nb, Ta, V, P). Similar to the other studied Ag-based materials, Ag2(NH4)AsS4 has applications as a nonlinear optical material and as a buffer layer in solar cells. Understanding the Ag2(NH4)AsS4 synthesis technique allows future researchers to synthesize new Ag2-I’-V-X4-type semiconductors with the same methods or apply these principles to the fabrication of Ag2(NH4)AsS4 thin films. The work presented in this dissertation furthers the understanding of the synthesis and prediction of quaternary chalcogenide semiconductors and lays the foundation for future device and thin film studies using the semiconductors studied here.

Halide semiconductors have recently emerged as a class of materials that unite outstanding optoelectronic properties with the ability to process device-quality thin films at low or even room temperature. Successful adoption of halide semiconductor-based technologies will, however, be contingent on the development of device architectures and film processing approaches that enable efficient, low-cost devices with stable performance and rigorous study of these materials’ photophysical properties. Herein, the goals are twofold: first, to develop low-cost device processing methods that deliver efficient solar cell performance while managing sources of instability; second, to extend existing thin film processing techniques to novel materials, enabling investigation of their optoelectronic properties.

After general introduction to halide perovskite materials, films and devices in Chapters 1 and 2, Chapter 3 confronts the first device challenge—i.e., reducing solar cell cost—by investigating cheap electron and hole transport layers (ETL and HTL) for halide perovskite solar cells. Efficient CH3NH3PbI3 perovskite solar cells are constructed using earth-abundant ETL CdS and HTL CuCrO2 that are deposited at low temperature (<100 °C). Although CuCrO2 appears to yield an inert interface with CH3NH3PbI3, X-ray photoelectron spectroscopy reveals that CdS can easily release Cd into the CH3NH3PbI3 film. X-ray diffraction (XRD) measurements show that excessive amounts of Cd cause phase segregation of insulating compounds in the perovskite, compromising solar cell performance. Nevertheless, careful optimization of device fabrication avoids this detrimental interaction, leading to solar cells with power conversion efficiency of over 15%. In addition to demonstrating efficient devices using low-cost materials, this work emphasizes the importance of managing interfacial as well as bulk stability.

Chapter 4 focuses on the second device challenge—i.e., managing instability—by developing inherently robust architectures via lamination and hot pressing. This technique circumvents the intrinsic thermal instability of perovskite thin films during processing and forms a self-encapsulating device architecture. Annealing MAPbI3 films under pressure in a specially-constructed tool allows significant grain growth at temperatures that would ordinarily decompose them rapidly to PbI2. However, these temperatures can also activate unexpected reactions with carrier transport materials previously thought to be inert, such as nickel oxide. Applying this knowledge, techniques are developed that avoid reactivity-related problems and recover the targeted solar cell performance.

Chapters 5 and 6 of this dissertation focus on developing deposition methods for new halide semiconductor films, with emphasis in this case on exploration of fundamental physical properties rather than device fabrication. In Chapter 5, resonant infrared matrix-assisted pulsed laser evaporation (RIR-MAPLE) is first used to deposit films of the archetypal halide perovskite CH3NH3PbI3, which possesses properties comparable to those prepared by more conventional methods such as spin coating, as determined by XRD, electron microscopy and optical spectroscopy. CH3NH3PbI3 solar cells fabricated using RIR-MAPLE reach power conversion efficiency of over 12%. RIR-MAPLE is then extended to the deposition of layered lead halide perovskite films incorporating oligothiophene-derived molecular cations, which cannot be controllably deposited by other methods. By varying the number of rings in the thiophene chain and the halide component of the inorganic layers, the photoluminescence emission from these films can be tuned to originate from either the inorganic or the organic component or be quenched altogether, supporting prior computational predictions of the tunable quantum well nature of these types of perovskite structures. Carrier transfer between the inorganic and organic moieties can synergistically populate triplet states in the organic, showcasing the unique physical properties attainable in complex-organic perovskites.

Chapter 6 focuses on the halide semiconductor indium(I) iodide, which possesses elements of its electronic structure like those of halide perovskites, that are often invoked as an explanation for these materials’ remarkable defect tolerance. Indium(I) iodide is prepared in thin-film form by thermal evaporation. A photovoltaic effect is demonstrated for the first time in this material, with solar cells demonstrating ~0.4% power conversion efficiency. Overall, the results advance our scientific understanding of halide semiconductors, and provide crucial pathways by which they can be made more technologically effective, and be studied in greater depth.

Emerging multinary chalcogenide semiconductors have long been in the spotlight for addressing challenges related to the scarcity and toxicity of commercially available solar energy converting materials, CdTe and Cu(In,Ga)(S,Se)2. Kesterite-based Cu2ZnSn(S,Se)4 (CZTSSe) materials have emerged as alternatives to address these issues for environmentally benign, low-cost photovoltaic (PV) and photoelectrochemical (PEC) applications. However, the significant degree of detrimental defect formation associated with near identical coordination and atomic size for all the Cu/Zn/Sn cations involved in the structure prevent the PCE values for these materials from reaching the level comparable to already commercialized thin-film solar energy converting technologies. The I2-II-IV-VI4 (I = Li, Cu, Ag; II = Ba, Sr; IV = Si, Ge, Sn; VI = S, Se) family of materials offer tunable structural, optical, and electronic properties that can potentially enable their pervasive application in the solar energy conversion industry, in addition to promise of Earth-abundance and nontoxicity through their individual components. In addition, this family of materials is expected to be disorder-resistant because the constituent atoms share dissimilar chemistry, ionic size, and coordination. As a first example from this family of compounds, experimental and computational analyses for Cu2BaSn(S,Se)4 (CBTSSe) agree with the assertion that the large size and distinct coordination for Ba relative to Cu/Sn inhibits III and III antisite disordering and related band tailing. However, while the CBTSSe system seems to meet the expectation of antisite defect resistance, the broader implications of atomic size/coordination discrimination as a defect control mechanism and the solar energy converting potential of this material remain to be resolved. In this dissertation, the goals, therefore, are fourfold: to develop low-cost thin-film processing methods that deliver efficient solar absorber device performance, to further study disorder within CBTSSe solution-processed films using different measurement techniques, to examine the validity of the constituent element size/coordination discrepancy strategy to control disorder/band tailing in multinary chalcogenide semiconductors, to reveal factors that limit PV performance within CBTSSe devices for future high-performance Earth-abundant photovoltaic technologies. After giving scientific motivations of this dissertation in Chapter 1 and a general introduction to Earth-abundant defect-resistant chalcogenide materials, their depositions, and device applications in Chapter 2, Chapter 3 confronts the first solar absorber device application challenge, i.e., reducing solar cell cost, by investigating solution-based synthetic pathways for thin-film deposition of CBTSSe material. Efficient PEC devices are constructed using CBTSSe thin films deposited from molecular solution and ball milling approaches to provide a preliminary measure of the solution-processed film quality relative to existing co-sputtered CBTSSe analogs. The molecular solution path involves the dissolution of Ba(NO3)2, Cu(CO2CH3)2, and SnI2 in low-toxicity solvent DMSO, followed by spin coating, sulfurization and selenization steps at high temperatures. The ball milling approach, on the other hand, involves dispersion of ball-milled Cu2S, BaS, Sn, and S particles in low-toxicity ethanol, followed by sulfur and selenium treatments at high temperatures. After addressing the challenge related to the formation of a BaSO4 secondary phase (i.e., modifying pre-bake conditions toward high temperatures—e.g., 540 °C—and supplying sulfur during the pre-baking step), a micron thick, single phase CBTSSe absorber layer with grains as large as 4.5 µm and band gap (E_g) of 1.68 eV was grown from the molecular solution approach. The films prepared from molecular solution and ball milling approaches (micron thick and with band gap E_g = 1.56 eV) were employed in a Pt/TiO2/CdS/CBTSSe photocathode structure, exhibiting ~10 mA/cm2 and 5.54 mA/cm2 photocurrent densities, respectively, at 0 V reversible hydrogen electrode (VRHE). The performance levels of these first solution-deposited CBTSSe PEC devices are parallel to the performance of analogous vacuum-deposited CBTSSe films, which offered ~12 mA/cm2 current density at 0 VRHE. The high quality CBTSSe films deposited from the molecular solution approach mentioned in this chapter enabled us to evaluate CBTSSe material defects, optoelectronic properties and PV performance parameters, and how they connect with processing variables and device geometries in Chapters 4-6. These chapters report power conversion efficiencies (PCE) enhancing from 2.9% to a world record value of ~6.5% for this new absorber. Chapter 4 explores intrinsic properties of compositionally stoichiometric large E_g material CBTS (~2 eV) and alloyed CBTSSe with ~1.6 eV E_g using several spectroscopic techniques, including low-temperature photoluminescence (LT-PL). In addition, for alloyed CBTSSe, we diagnose PV device performance limiting factors for a 2.9% PCE device. LT-PL measurements of CBTSSe show a defect PL emission line at ~1.5 eV and deep defect feature at 1.15 eV. This measurement shows no apparent excitonic luminescence features (unlike CBTS). Tracing the unique photoluminescence signatures of these emission lines under various excitation intensities and temperatures helps conclude that the 1.5 eV emission line belongs to shallow quasi donor-acceptor pair (QDAP) recombination, whereas the 1.15 eV emission belongs to deep QDAP transitions. On the other hand, the LT-PL spectrum for CBTS consists of several weak near-band-edge PL emission lines and strong emissions at 2.08 eV and 2.11 eV, and broad/weak defect bands at 1.95/1.6 eV. Based on findings from excitation and temperature dependent PL analysis, PL lines are attributed to bound exciton (at 2.08 eV), free exciton (at 2.11 eV), donor-acceptor pair (1.95 eV), and free to bound transitions (1.6 eV). LT-PL analyses point out that the potential fluctuations are more substantial in Zn-based CZTSSe than CBTSSe, which in turn has stronger fluctuations than CBTS. LT-PL, time-resolved terahertz (THz) and photo electron spectroscopy as well as capacitance voltage analyses on CBTSSe films point out that the presence of deep defects within the band gap, short bare film surface minority carrier lifetimes (~50 ps), non-ideal band alignment at the CdS/CBTSSe interface (i.e., cliff-like 0.63 eV conduction band offset measured by ultraviolet photoemission spectroscopy) and low density of charge carriers are the primary issues that limit the maximum achievable (1.31 V) VOC to 0.47 V for the 1.59 eV E_g material, resulting in 2.9% PCE devices. The physical measurements provided on the stoichiometric solution-processed CBTSSe absorber bring up critical directions for future performance improvement of the devices based on this Earth-abundant I2-II-IV-VI4 family. Chapter 5 focuses on the first VOC enhancement strategy, i.e., improving intrinsic material quality (e.g., minimizing deep defects and increasing charge carriers) by process optimization through Cu stoichiometry tuning. This chapter targets to increase our understanding of the phase stability and optoelectronic property sensitivity for CBTSSe material. According to XRD and SEM analyses, phase purity is sustained throughout a Cu content range (as measured by [Cu] / [Ba+ Sn]) of nominally 0.94 to 1.01, and grain enlargement correlates with Cu content. THz spectroscopy and Hall effect reveal that the lifetime, carrier mobility, p-type conductivity and hole density (~1013 cm-3) are nominally independent of Cu content. The champion PCE exceeds 4.7% for all copper compositions in the phase pure region, with a record value of 5.1%, representing the highest performance level achieved to this point for a solution-processed device using the indium tin oxide (ITO)/intrinsic zinc oxide (i-ZnO)/CdS/ CBTSSe device architecture. This PCE is equivalent to that for the best vacuum-deposited devices using the same device structure and represents 30% relative change in VOC (0.605 V) compared to the devices with 2.9% PCE. These findings suggest that CBTSSe films and solar cells may be less sensitive to Cu stoichiometry compared to kesterite materials and therefore that they may provide a more stable material platform to prepare high quality thin-film solar cells. Finally, Chapter 6 focuses on the second VOC enhancement strategy, i.e., improving interface quality by developing a device architecture that minimizes non-idealities related to band alignment. Chapter 6 demonstrates a >25% improvement in VOC (from ~0.60 V to ~0.76 V) and corresponding enhancement in PCE (~5.1 % to ~6.2 % without anti-reflection coating; ~6.5 % with MgF2 anti-reflection coating, representing the world record PV PCE for the I2-II-IV-IV4 family of materials) for solution-processed CBTSSe solar cells, by introducing an alternative buffer/window stack with a lower electron affinity relative to the conventional CdS/i-ZnO/ITO stack, more in accord with the low electron affinity of the CBTSSe layer. The front stack used for this study consists of a successive ionic layer adsorption and reaction-deposited Zn1-xCdxS buffer combined with sputtered Zn1-xMgxO/Al-doped ZnO window/top contact layers. We have investigated the impact of the front buffer/window by examining the device properties using a combined experimental and device simulation (SCAPS-1D) approach. We show the importance of considering both the buffer and window layer band positions relative to the absorber in optimizing CBTSSe solar cell performance. These chapters help to examine the general validity of the proposition that atomic size/coordination discrimination can be used to target needed defect resistance within the broader family of complex multinary chalcogenide films by comprehensive screening of semiconductor prospects for CBTSSe films obtained via a solution processing approach and targeted towards light-absorbing applications, particularly in PV and PEC solar. The processing-property understanding gained for solution-processed CBTSSe in these chapters can be translated to other I2-II-IV-VI4 family members towards highly efficient solar devices.

Kesterite Cu2ZnSnS4-xSex (CZTSSe) has once gained wide attention as a potential alternative to the CdTe and Cu(In,Ga)(S,Se)2 (CIGSSe) photovoltaic (PV) technologies, which are currently facing challenges in terms of scalability due to the use of the scarcity (Te, and In) and toxicity (Cd) of the elements. However, the similarities between the Cu and Zn atoms in terms of cation size and coordination environment result in the formation of a high density of Cu- and Zn-related anti-site defects and defect clusters in the CZTSSe lattice and limit the open-circuit voltage and efficiency of the CZTSSe solar cells. To target suppressing the formation of anti-site defects and related defect clusters, Cu2-II-IV-X4 (II = Sr, Ba; IV = Ge, Sn; X = S, Se) compounds, which have a significantly larger and chemically more differentiated group-2 element (i.e., Ba, and Sr) instead of Zn, have been introduced. Among Cu2-II-IV-X4 compounds, Cu2BaSnS4 (CBTS) and Cu2BaSnS4-xSex (CBTSSe) with trigonal structure (P31 space group) have been the first materials to have gained attention, and their thin-film deposition using both solution- and vacuum-based techniques, as well as their PV devices, have already been demonstrated. On the other hand, there are only a handful of studies on Cu2BaGeSe4 (CBGSe) and Sn-alloyed Cu2BaGe1-xSnxSe4 (CBGTSe) systems, which have the same crystal structure as CBTS and CBTSSe. In this dissertation, we explore film growth, material properties, property engineering methods, and PV application of this relatively unexplored system, CBGTSe, to get a better understanding of this compound as a potential PV material. To achieve this goal, the following studies are conducted and presented throughout this dissertation: (1) development of a deposition process for high-quality CBGTSe films, which yield functioning solar cells, and examination of the associated solar cell properties; (2) characterization of important optoelectronic properties of CBGSe and CBGTSe films and identification of major bottleneck for their solar cell performance; and (3) demonstration of two different film modification strategies (i.e., alloying and doping) for CBGTSe films and investigation of associated changes in the film properties.The first study demonstrates a high-quality CBGTSe film growth via sequential vacuum deposition (i.e., sputtering, and evaporation) of elemental layers (i.e., Cu, Ba, Ge, and Sn) followed by a selenization step to convert the metallic layer of Cu–Ba–Ge–Sn into CBGTSe compound. This work investigates film growth mechanisms via ex-situ analysis and reveals the critical process parameters (i.e., pre-annealing temperature, and Cu-content) for high-quality films. Additionally, functioning solar cell devices based on CBGTSe as light absorber are demonstrated for the first time. Second, the optoelectronic properties of CBGSe (Sn-free) films are also investigated in detail and compared with its isostructural CBTS using various analysis techniques, including temperature-dependent photoluminescence (PL), Hall effect, photoelectron spectroscopies, optical-pumped terahertz probe spectroscopy (OPTP), and open-cell time-resolved microwave conductivity (oc-TRMC), which reveal possible bottlenecks for the solar cell performance and possible directions for the improvement. Next, two different modification approaches—i.e., 1) alloying with Ag, and 2) doping with group-1 (alkali metals) elements—, which have been used for the existing CIGSSe and CZTSSe technologies, are demonstrated to modify overall optoelectronic properties of the CBGTSe films. First, the partial substitution of Cu by Ag is examined as a potential film property modification strategy. The study reveal how much Cu can be substituted with Ag while maintaining its original trigonal crystal structure and how phase purity, morphology, charge carrier properties, band positions, and recombination properties, which are all critical for the PV and optoelectronic applications, change as a function of Ag-content. The intrinsic background carrier densities for CBGTSe films are relatively low (p = ~1012 cm-3) compared to other related chalcogenides (p = 1015–1017 cm-3 for CIGSSe, and CZTSSe), which can limit its applications as photovoltaic, thermoelectronic, and optoelectronic devices. Therefore, as prospective dopants for the CBGTSe films, alkali elements (Li, Na, K, and Rb) are evaluated to address the low hole carrier density and potentially to allow for property tunability. The study demonstrates orders of magnitude enhancement in hole carrier density via alkali-doping. The changes in other film properties (i.e., film morphology, carrier mobility, and minority carrier lifetime) with alkali-element doping are also examined. Additionally, to address inappropriate band alignment within solar cells based on the Cu2-II-IV-X4 family (e.g., Cu2BaGe1-xSnxSe4, Cu2BaSnS4-xSex), which typically has shown noticeably lower electron affinity (EA) than conventional CdS/i-ZnO/ITO buffer/window stacks, we introduce Zn1-xCdxS/Zn1-xMgxO/ZnO:Al as an alternative low-EA buffer/window stack. The low-EA buffer and window layers contribute to improvement in the properties of CBTSSe solar cells and yield a maximum PCE of 6.5% (with MgF2 anti-reflection coating), which represents the current record PCE for CBTSSe-based solar cells. The study reveals that the alternative buffer/window stack improves overall recombination properties for the CBTSSe solar cells from band offset estimation using photoelectron spectroscopy, and recombination property analysis. In addition, device modeling and simulation results provide directions for further improvement of device performance. The works presented in this dissertation provide baseline understanding and knowledge on film synthesis, material properties, property engineering, and associated solar cells for the CBGTSe compound as well as for the relevant Cu2-II-IV-X4 multinary chalcogenide compounds.