Thin film chalcogenide photovoltaic materials

Resume : First-principles density functional theory (DFT) is the state-of-the-art theoretical framework for computing materials properties. For decades, DFT-based calculations have generated reliable results for photovoltaic materials. By virtue of powerful computers, high-throughput materials screening based on DFT calculations has been employed for photovoltaic materials design and development. The crucial disadvantage of DFT-based materials modelling is high computational costs, hindering the application of this method to large-scale calculations. An emerging approach in atomistic materials modelling that can be of interest for the photovoltaic research is computational simulations based on machine learning interatomic potentials (MLIPs). It has been shown that MLIPs can potentially enable the performance of large-scale simulations with quantum accuracy orders of magnitude faster than current methods based on DFT. It is very likely that in the coming years MLIPs will become an integrated part of the photovoltaic modelling toolkit. A key component for developing MLIPs is building a robust and representative reference dataset for model training. Very often, high-throughput calculations required for building reference datasets are done manually. This requires performing different types of calculations and handling a large amount of data. An important task in modern computational materials science is to accomplish these tasks automatically. In this regard, we have developed a framework, PyFLAME, for the systematic construction of MLIPs with minimal human intervention. To demonstrate the performance of MLIPs generated with our approach, we compare the results of molecular statics and molecular dynamics calculations using the constructed potentials to reference DFT results.

Resume : Cu(In,Ga)Se2 (CIGS) solar cells have reached conversion efficiencies above 23% for CIGS in polycrystalline form. There has been much debate regarding the role of grain boundaries (GBs) in CIGS thin-film solar cells, but how exactly the grain boundaries affect the efficiency of the cells remains controversial. In this study, we performed three-dimensional numerical simulations of Cu-poor and Cu-rich CIGS solar cells to investigate the impact of grain boundaries on the solar-cell performance. Depending on the Cu-poor or Cu-rich composition of CIGS, grains and grain-boundaries with compositional variations and different energy gaps, different defect density and defect energy levels have been simulated. The effect of band edge shifts at GBs on the cell performance have been also studied. Simulation results show that GBs in Cu-rich cell can be quite detrimental for the cell performance, while GBs in Cu-poor cell can benefit from features such as type inversion or passivation due to a wide bandgap at the GB that help reduce the impact of grain boundaries on the device performance.

Resume : A key point to optimize the performance of Cu(In,Ga)Se2 (CIGSe) solar cells is the back contact barrier. A typical back contact is an opaque metal such as molybdenum (Mo). Basically, the contact of Mo and CIGSe, as a metal-semiconductor junction, should be considered as Schottky-like contact. Despite different fabrication and measurement procedures in experimental research, most of them confirm that the Mo/CIGSe interface shows Ohmic or quasi-Ohmic behavior [1-3]. However, there are contradictions about the mechanisms converting the Schottky contact to a quasi-Ohmic/Ohmic one. They can stem from different work functions for Mo, the existence of a thin layer between Mo and CIGSe, or interface states due to dangling bonds or surface contamination [4-5]. Therefore, for an improved understanding, it is of great importance to investigate the effect of interface states on the behavior of the back contact barrier. In this work, we aim at modeling a two-dimensional CIGSe solar cell stack in COMSOL Multiphysics by considering the impact of interface states on the back barrier. According to the dependency of the barrier height for holes at the p-CIGSe/Mo junction on the interface state densities (Dit), the interface state parameter (γ) is defined [5]. For Dit→∞, γ=0, and the Fermi level is pinned. Consequently, the barrier height is totally insensitive to the work function of Mo and the contact becomes Ohmic. For Dit→0, γ=1, and the barrier height is directly related to the work function of Mo, leading to the ideal Schottky model. Our simulations reveal that the increase of interface state density from 1010 cm-2 to 1014 cm-2 enhances the short-circuit current density (JSC) and the open-circuit voltage (VOC) by 10.9% and 40%, respectively. Therefore, a 6% relative growth in the efficiency of the solar cell occurs. Furthermore, a comparison of band diagrams for different interface state densities indicates that an increase to 1013 cm-2 and beyond leads to a quasi-ohmic behavior of the CIGSe/Mo junction, resulting from Fermi-level pinning. In contrast, for lower interface state densities, the Schottky contact appears. Overall, the interface state density is one of the crucial mechanisms to elaborate the behavior of the CIGSe/Mo interface, and this work provides guidelines in which range a Schottky- or an Ohmic-like contact may be formed. [1] Abou-Ras, D., Nikolaeva, A., Caicedo Dávila, S., Krause, M., Guthrey, H., Al-Jassim, M., Morawski, M. and Scheer, R. Sol. RRL 2019, 3, 1900095. [2] Hsiao, K.-J., Jing-Da, L., Hsing-Hua, H. and Ting-Shiuan, J. Phys. Chem. Chem. Phys. 2013, 15, 18174. [3] Nicoara, N., Manaligod, R., Jackson, P., Hariskos, D., Witte, W., Sozzi, G., Menozzi, R. and Sadewasser S. Nat. Commun. 2019, 10, 3980. [4] Michaelson, H. B. J. Appl. Phys. 1977, 48, 4729. [5] Schroder, D. K. and Meier, D. IEEE Trans. Electron Devices 1984, 5, 637.

Resume : Cu(In,Ga)(S,Se)2 is the most efficient type of thin-film solar cells at the moment. The band gap can be engineered by changing the Ga and S content, but substitution of Cu with Ag is another promising route towards an optimized absorber material. In this work we examine the influence of this substitution on the electronic and structural properties of CIGS solar cells absorber materials. We present a cluster expansion model that describes the energetics of the system, calculate the phase diagram of the system, determine the band gap variation with Ag content and investigate the diffusion of Ag and Cu in the boundary phases, as well as the solid solution. The presented results will provide new insights into the thermodynamics of Ag containing CIGS absorber, which can be utilized to further improve the performance of thin-film solar cells.

Resume : The introduction of post deposition treatments (PDT) of the absorber and the increase of band gaps of the different layers in the cell stack led to a considerable increase of CIGSSe small laboratory cell efficiencies beyond 23% in the past years. The development on mid-size monolithically integrated modules involves an increase by more than two orders of magnitude in device area and the use of industrially compatible processes and equipment. At the AVANCIS’ R&D Center in Munich we have reached an externally certified power conversion efficiency of 19.8% (independently certified by NREL) which represents a new world record for an encapsulated CIGSSe thin-film module with integrated serial connection of size 30x30cm2. The champion efficiency was achieved by combining a Na post deposition treated wider band gap CIGSSe absorber with a simplified Zn(O,S) / ZnO:Al all-sputter buffer and TCO stack. In the present contribution, we demonstrate how device simulation, materials and device characterization were used to increase the CIGSSe module efficiency to the verge of 20% and what loss mechanisms have to be overcome for further efficiency development of the single junction device. Major contributions to the efficiency increase of the modules stem from the reduction of recombination losses in the bulk of the absorber as well as at the interface towards the buffer and the back contact by adjustment of the Ga- and S-profiles and from bulk and interface passivation with alkali PDT. The band alignment at the absorber to buffer interface had to be controlled by the surface-S of the CIGSSe absorber and the S/(S+O) ratio of the Zn(O,S) buffer layer. The new device stack also pays off in terms of an improved temperature coefficient and does not compromise on the long-term stability of the electrical parameters of the modules as compared to the previous baseline products.

Resume : The defects occurring in a pristine commercial Cu(In,Ga)Se2 (CIGS) module and small area laboratory CIGS cells (0.35 cm2) were investigated to gain insight into their origin and their impact on device performance. Defects in the commercial module were localized by electroluminescence (EL) imaging of the entire module, after which lab-scale samples of approximately 3.5 cm in diameter (‘cores’) were drilled out at relevant locations. The front glass and encapsulant were mechanically removed from the cores to characterize the solar cell stack with mapping and microscopy techniques. Cores and laboratory cells were investigated with photoluminescence (PL), lock-in thermography (ILIT), scanning electron microscopy and energy-dispersive X-ray spectroscopy to determine the electrical behaviour, microscopic structure and chemical composition of the defects. The defect-free cores yielded efficiencies in agreement with the nameplate of the module. After initial characterization, some laboratory cells were exposed to damp heat (85°C, 85% relative humidity) for up to 120 hours to determine the evolution of the defects. Three types of shunting defects were found in the module: 1. Interconnection defects caused the largest shunts in EL images of the entire module. PL and ILIT analysis of cores with an interconnection defect showed that they caused partial shunting of two cells on either side of the interconnect. Optical microscopy at the showed discoloration in the active area next to the shunt location, while confocal microscopy showed an apparent increase in surface roughness. These features could either be root-cause of shunting or a result from the shunting. 2. Shunting protrusion defects with a diameter in the range of 60-600 μm and a height ranging from 25-400 μm. The largest protrusion defect found consisted mainly of selenium, suggesting that such defects can form during the selenization step in CIGS production. 3. Exposed molybdenum (Mo) defects were the only defects found in both the module and the laboratory cells. They were typically ~100 μm to 1 mm in size and occurred as both shunted and non-shunted defects. In laboratory cells, shunted exposed Mo defects resulted in up to 10% relative efficiency loss. Our best hypothesis for their formation in modules is that the CIGS layer delaminates at locations where the Mo layer is contaminated (e.g. with dust particles) and the adhesion between CIGS and Mo is insufficient. We propose that the shunting is due to a short-circuit between the TCO and Mo, due to conductive material bridging the depression. Damp heat treatment of the cells with exposed Mo defects partially “healed” the shunting. This phenomenon is likely the result of oxidation of the conductive materials that were initially short circuiting the device.

Resume : For humanity to go 100% renewable it is estimated that a few percent of our land surface must be covered in photovoltaic panels, which is an enormous area. To make this more socially acceptable, we and others suggest aesthetically appealing PV modules covered with colour images to fit in the local environment. Existing methods for producing colourful solar cells use additional layers on top of the devices. Recently Cho et al. showed that the colour of Cu(In,Ga)Se2 (CIGS) solar cells can be changed using the interference phenomenon by only varying the thicknesses of the internal sputtered buffer and window layers of the device[1]. However, sputtered layers are uniform in thickness and hence only single colours can be achieved, whereas inkjet printing through the ability to change droplet spacing, allows a single layer to have different thicknesses in different spatial locations giving us the option of creating images. Here we complete single and two colour solar cells based on CIGS absorbers with an inkjet printed Zn(O,S) buffer layer[2], as well as sputtered i-ZnO and In:SnO2 (ITO) window layers. The optical constants of these layers were derived by ellipsometry and then used to simulate the CIGS device stacks visual appearance upon varying the buffer and window layer thicknesses using the transfer matrix method. Using these results as thickness inputs, we completed devices with 30 to 70 nm Zn(O,S), i-ZnO of 50 nm, and ITO of 150 and 350 nm. Initially, this resulted in dull-looking devices with highly diffuse reflection. The brightness of the solar cells was improved by reducing the surface roughness of the CIGS from 200 nm to 100 nm RMS using Br2 etching. SEM revealed that KCN etching and a 120 s UV-ozone treatment of CIGS absorber were required for the Zn(O,S) ink droplets to wet the freshly Br2 etched surface. We achieved purple, magenta, blue, green, yellow, and grey coloured solar cells. The single coloured solar cells reported efficiencies of ~0.60 to 1.09 times the reference cell. We also successfully fabricated single solar cells with two hues having their efficiency values midway between those of the individually coloured solar cells. The average short-circuit current for the different colour solar cells is >30 mA/cm2, which suggests low losses due to optical reflections. We will simulate the efficiency expected with state-of-the-art CIGS absorber layers yielding the same colours (with varying Zn(O,S) and ITO thicknesses) using the SCAPS simulation tool. While we are a still long way from printed solar cell art adorning our buildings, we do show a proof of principle that we can print solar cell layers in a way that we can have different colours in the same solar cell. [1] Cho et al. 28, 798– 807, Prog Photovolt Res Appl. 2020 [2] Chu et al. 13, 13009−13021, ACS Appl. Mater. Interfaces, 2021

Resume : While the atom probe tomography (APT) technique has been around for more than 50 years, only recently has there been substantial efforts toward application in photovoltaics. Historically the three-dimensional nanoscale compositional technique was applied primarily to metals but advances in the areas of hardware and software have improved the application space through the use of various wavelength laser pulses, increased field-of-view (FOV), and mass-spectral-quality. In the current work, we will briefly introduce the technology of APT including electrostatics, optics, and laser pulsing, and then give a brief overview of some applications of APT in photovoltaics including chalcogenides. We will use the applications to demonstrate the different ways that the 3D data can be interrogated to give new knowledge about the materials that would be difficult or impossible with any other microscopy. We will then present the two new atom probes which comprise CAMECA’s 6000 product line and briefly introduce the benefits of each over previous generations. The LEAP 6000 XR system introduces deep ultraviolet (DUV) laser pulsing to commercial atom probe systems for the first time, improving success rates and data collection rates on key applications. It is completely automated and compatible with the existing local electrode and microtip workflow. It also has a hybrid pulsing mode using both voltage and laser pulses simultaneously providing a higher effective pulse fraction that dramatically improves the signal-to-noise ratio. The Invizo 6000 also incorporates the benefits of DUV laser pulsing, but it does so with a novel counter electrode design, allowing high-angle dual laser beam illumination that is thermally coincident and symmetric providing improved data quality and reconstruction fidelity. Ions leave the specimen surface and are electrostatically directed to the detector in a nominally straight flight path using a double Einzel lens that captures nearly the entire specimen volume, as much as a doubling of the previously achieved field-of-view. This opens new applications that require a large FOV and make site-specific specimen preparation easier. The large FOV also allows more careful control of the evaporation as well as more information about the specimen, further improving reconstruction fidelity.

Resume : This paper reports the effect of Na doping Cu2ZnSnS4 powders on the structural, morphological and optical properties. CZTS powders were synthesized by direct melting method of the constituent elements with different sodium doping concentrations from 0.5 % to 2 %. The resulting CZTS:Na powders were characterized by X-ray diffraction, Raman spectroscopy, scanning electron microscopy, scanning transmission electron microscopic and UV-visible-NIR spectrophotometer. X-ray diffraction and scanning transmission electron microscopic analysis confirmed the formation of CZTS:Na with kesterite structure and preferential orientation along (112) plane. SEM results shown the microstructure with large grains with increasing the Na doping concentration. Optical measurements revealed that the band gap of CZTS decreases from 1.66 to 1.53 eV by increasing Na-doping. In addition, electrical investigations indicated that all ingots exhibit p-type conductivity. Keywords: Cu2ZnSnS4; Direct fusion method; Semiconducting materials; Na-doping; Characterization

Resume : Quaternary chalcogenide semiconductors’ family, particularly Cu2ZnSnS4 (CZTS) material acquired an enormous research interest thanks to its low-cost, non-toxic and abundant components in the earth’s crust. Furthermore, its direct band gap around 1.5 eV, a high absorption coefficient (105cm-1 ) a reasonable efficiency rate 12.6% and a p-type conductivity make him a suitable candidate as absorber layer for solar cells application. Alkali metal doping represents one of the main ways to achieve high-efficiency CZTS solar cells. Na was incorporated into CZTS bulk crystals compound from elemental Cu, Zn, Sn, S using the direct melting method. Crushed powders for 10 min of these ingots were used as raw materials for the thermal evaporation. Thin films of undoped and Na-doped CZTS were deposited onto glass substrates at 100°C temperature by thermal evaporation technique. After the deposition, all CZTS thin films were annealed in a furnace in sulfur atmosphere at a temperature of 400 °C during 30 min. Keywords: Cu2ZnSnS4; Na-doped; Thermal vacuum evaporation; Thin films; Absorber layer; Photovoltaic

Resume : We have successfully produced Cu2CoSnS4 (CCTS), Cu2NiSnS4 (CNTS), Cu2ZnSnS4 (CZTS) and Cu2FeSnS4 (CFTS) materials by direct melting. Constituent elements taken in stoichiometric compositions were used as starting points. After that, Cu2XSnS4 (M=Co, Ni, Zn and Fe) thin films were growth using a single source vacuum thermal evaporation method. The resulting Cu2XSnS4 films were characterized by X-ray diffraction, Raman spectroscopy, scanning electron microscopy, energy dispersive X-ray analysis and optical spectrophotometry to investigate the crystal structure, composition, morphology, absorption coefficient and optical band gap. Hot probe method shows that Cu2XSnS4 (M=Co, Ni, Zn and Fe) ingots and fabricated samples exhibit p-type semiconductor. This work demonstrates that Cu2XSnS4 (M=Co, Ni, Zn and Fe) thin films are promising absorber materials for energy-conversion applications. Keywords: Cu2XSnS4; Thin films, Thermal evaporation; Structural properties; Optical Properties.

Resume : Research on quaternary Cu-based chalcogenide semiconductors has caught a large interest for photovoltaic applications, because these materials consist of non-toxic and earth abundant elements. While being environmentally friendly and low cost, materials like Cu2MnGeS4 or Cu2MnSnS4 are very promising candidates for use in tandem solar cells, because by alloying they can cover a wide bandgap energy range of 1.52-1.72 eV [1]. This study presents new insight into the crystal structure of the solid solution Cu2Mn(GexSn1-x)S4. Cu2Mn(GexSn1-x)S4 has already been studied earlier [2], where the structural analysis of the solid solution based on X-ray powder diffraction (XRD). Because Cu, Ge and Mn are electronic similar elements, they are difficult to distinguish when analyzing conventional X-ray diffraction data. But their neutron scattering lengths are different, that is why we apply neutron diffraction to analyze the crystal structure of Cu2Mn(GexSn1-x)S4 mixed crystals. The endmembers of the Cu2Mn(GexSn1-x)S4 solid solution crystallize in different structures: Cu2MnSnS4 crystallizes in the tetragonal stannite type structure (space group I4 ̅2m), whereas Cu2MnGeS4 adopts the orthorhombic wurtz-stannite type structure (space group Pmn21). Thus, within the solid solution a structural transition from the tetragonal to the orthorhombic crystal structure can be expected. In the presented study powder samples of Cu2Mn(GexSn1-x)S4 mixed crystals were synthesized by solid state reaction of pure elements in evacuated silica tubes at temperatures of 800°C. The basis of our investigations is a careful determination of their chemical composition as well as the presence of secondary phases by WDX spectroscopy using an electron microprobe system. This is important, because it was shown for other quaternary chalcogenides (e.g., Cu2ZnSnS4 and Cu2ZnSnSe4) that even small deviations in composition (off-stoichiometry) can have a critical influence on the cation distribution in the crystal structure and the band gap energy of the material [3]. The lattice parameters of the mixed crystals were obtained by LeBail refinement of powder XRD data. Rietveld analysis of neutron diffraction data has shown that Sn-rich mixed crystals (0 ≤ x ≤ 0.2) adopt the stannite type structure, whereas Ge-rich mixed crystals (0.8 ≤ x ≤ 1) of this series adopt the wurtz-stannite type structure. The results of the chemical composition study in combination with structural characterization and optical bandgap evaluation from diffuse reflectance of Cu2Mn(GexSn1-x)S4 mixed crystals will be presented providing special attention to the phase transition from the stannite to the wurtz-stannite crystal structure. [1] Beraich et al. (2020) Journal of Alloys and Compounds, 845, 156216 [2] Bernert & Pfitzner (2005) Crystalline Materials, 220(11), 968. [3] Schorr & Weidenthaler (2021) Crystallography in Materials Science, deGruyter.

Resume : Many efforts have been made word-wide to obtain earth-abundant variants of CuInxGa(1-x)Se2 (CIGS) for photovoltaic application: a lot of metals and semimetals have been tested to substitute zinc in Cu2ZnSnS4 (CZTS). This chalcogenide is the been more investigate compound in other to increase the environmental sustainability of this class of thin film PV devices. An potential alternative to CZTS is Cu2FeSnS4 (CFTS). CFTS, which crystallizes into a stannite structure (space group: I-42m), shows high absorption coefficient (104 cm-1) and direct band gap suitable for PV applications. Iron is way cheaper (0.05$/kg vs. 2,7$/kg) and has higher occurrence in the Earth crust than zinc (6.3% mean vs. 65ppm mean) leading to a more economical and environmental sustainability of CFTS in respect of CZTS. In literature, few works are available about CFTS thin films and just some of them report PV performance, that are quite interesting and need to better investigate. In this work CFTS thin films were grown by an high-throughput two-steps deposition process: metallic precursors were deposited separately by sputtering technique in a stack, and subsequently treated at high-temperature in sulphur vapours in order to induce the synthesis of the quaternary chalcogenide. CFTS thin films have been deposited both on Molybdenum-coated soda-lime glass to finalize preliminary solar device and on bare glass substrates to measure Eg via UV-Vis absorption spectra; on the Mo-SLG samples different thin films have been deposited to finalize the solar forming the standard structure SLG/Mo/CFTS/CdS/ZnO/AZO/Al grid. This work aims to correlate the effect of the main growth parameters with the chemical-physical properties of CFTS thin film that have been deposited. Final composition (relative metallic precursor ratio), precursors order in the stack and thermal treatment profile, are the experimental parameters under investigation. Scanning Electron Microscopy (SEM) and Energy Dispersive X-ray (EDX) analysis have been used to check the morphology (of front and Mo-CFTS interface) and chemical composition. For many of experimental parameters tested, CFTS thin films compact on front and void-rich on back has been obtained with composition ranging from [Cu]/[Fe]+[Sn] ≈ 0,7-0,85, [Fe]/[Sn] ≈ 1.2-1.3, [Sn]/[Cu] ≈1.6-2). µ-Raman spectroscopy and X-ray diffraction (XRD) measurements carried on pristine samples have been used to identify the main crystalline phase and the formation of by-product (Es:Rhodo-stannite, Chalcopyrite, Pyrite) as a function of metallic precursor ratio. Preliminary XPS data have been collected to better interpret the other results in order to interpret the effect of the different growth parameters tested in this work. Preliminary CFTS-based thin films have finalized to validate chemical bath deposition for CdS coating in order to create p-n junction to test PV performance of our samples.

Resume : As the pursuit for low-cost thin film solar cells is going on, Cu2ZnSn(S,Se)4 compounds, widely known as CZTS and kesterite referring to their crystal structure, are attracting considerable interest as light absorbers. These absorbers are composed of non-toxic and earth-abundant elements, but the resulting solar cells suffer from high Voc-deficit as a consequence of inherently high defect density. Alkali doping is a promising approach to improve the photovoltaic performance of Cu2ZnSn(S,Se)4 kesterite absorber-based solar cells via defect suppression and/or passivation. Due to its similar ionic radius to copper, lithium can occupy VCu in the CZTS lattice resulting in crystallinity and conductivity improvement, a reduction in grain boundary recombination and a tunability of the band gap. On the other hand, sodium enhances grain growth leading to an improved absorber morphology, and additionally passivates defects at grain boundaries. There have been many recent studies, which thoroughly investigated these two alkali dopants individually but not together. In this work, we spin coat the precursor solution onto a Mo-coated soda lime glass substrate followed by an annealing step under Se-enriched atmosphere. A SiOx diffusion barrier between the Mo layer and the substrate helps to control alkali ion inter-diffusion between absorber and substrate. Various amounts of Li- and Na-chlorides are added to the precursor solution simultaneously. This intentional addition of alkalis to the absorber layer allows us to examine the properties of CZTS absorbers under the presence of multiple alkali elements. Our goal is to understand, how Li and Na together influence the synthesis and the electronic properties of solution-processed CZTS, aiming at overcoming the current limitations of kesterite absorbers in terms of Voc – deficit. We found a mutual dependency of the final Li- and Na-concentrations in the CZTS absorber and we were able to fabricate kesterite solar cells with efficiencies around 10 % using various combination of Li- and Na-concentrations. We will present the influence and benefits of combined Li and Na doping on the electrical and electronic properties based on morphological, J-V, admittance and optical characterization. Our results highlight the importance of alkali element management during the synthesis of kesterite absorbers, and point towards beneficial concentration combinations of the herein investigated alkali elements Li and Na in the precursor solution.

Resume : In this work, the quaternary chalcogenide Cu2NiSnS4 (CNTS) films were grown on soda-lime glass by spray pyrolysis technique. The purpose of this study is to find the best route to obtain single-phase CNTS films. Thus, the effect of deposition temperatures (Td = 250, 300, and 350 °C), and sulfurization temperatures (Ts = 450, 500, and 550 °C) is investigated by using X-ray diffraction, Raman spectroscopy, energy dispersive spectrometer, Scanning electron microscopy, and conventional optical spectroscopies. The X-ray diffraction and the Raman spectroscopy measurements indicate the formation of the CNTS phase with cubic structure, as well as the presence in different amounts of secondary phases. The full width at half maximum values of the most intense CNTS diffraction lines shows that a high sulfurization temperature is required. The bandgap values inferred from diffuse reflectance data are discussed with respect to stoichiometry and phase composition. All authors acknowledge financial support from the Moroccan Ministry of Higher Education and Research through the PPR/37/2015 project, and from the Romanian Government through the Core Program PN19-03 (contract no. 21 N/08.02.2019) and PN-III-P4-ID-PCE-2020-0827 (Contract no. PCE74 09/02/2021) project.

Resume : In the current study, Cu2CoSnS4 (CCTS) thin films were grown by evaporation of powdered CCTS on various glass substrate temperature. As-grown thin films were annealed in sulfur atmosphere at 400°C for 2 hours. We have systematically studied the structural and optical properties of post-sulfurized CCTS thin films by means of X-ray diffraction, Raman Scattering and UV-vis spectroscopy. CCTS films exhibit polycrystalline structure with stannite phase and (112) preferred orientation. Profound analysis has been thoroughly made to investigate the variation of the optical and dielectric properties and other related parameters as a function of substrate temperature. Optical constants of post-sulfurized CCTS thin films such as absorption coefficient, Urbach energy, extinction coefficient and refractive index were investigated by optical measurement from transmittance and reflectance spectra in the UV-VIS-NIR range. We have discussed the dispersion of the refractive index in terms of the Wemple–DiDomenico single oscillator model from which optical dispersion parameters such as the single oscillator energy (Eo) and dispersion energy (Ed) were calculated in the investigated wavelength range. We have calculated the optical conductivity and the dissipation factor of the studied films. The electric free carrier susceptibility and the carrier concentration on the effective mass ratio were determined by using the model of Spitzer and Fan. Key words: CCTS, Thin films, Optical dispersion parameters, Dielectric parameters

Resume : Kesterite (Cu2ZnSn4 / S4) is a non-toxic and readily available quaternary compound. This material is a promising light absorber for solar cells, being a good alternative to crystalline silicon. The Shockley-Queisser (SQ) limit of this absorber is about 32.8%, achieving solar cells with high certified efficiencies up to 12.5%. The standard CZTSe solar cell fabrication process incorporates Molybdenum as Hole transport layer (HTL) and Cadmium Sulfide as electron transport layer (ETL). However, CdS material is a well-known toxic and non-environmental friendly compound being interesting replace it as ETL from the PV fabrication chain. With this idea in mind, transition metal oxides could be a good choice to replace CdS in CZTSe solar cells. In this work, very thin atomic layer deposited (ALD) Al2O3/TiO2 stack as ETL on kesterite solar cells has been explored. The aforementioned TMO-based stack is capped with a thermal evaporated glycine-based amino acid film as intermediate layer between the ETL and the top transparent conductive oxide (TCO). The main function of the glycine film is to act as a strong dipole reducing the high work function of the top TCO-based electrode in order to guarantee a proper electron extraction from the absorber. Fabricated solar cells use a ALD V2O5 thin layer (~ 10 nm thick) as HTL at the rear side deposited on a FTO/glass substrate, reaching a bifacial solar cell structure with photocurrent densities and open circuit voltages up to 27 mA·cm2 and 250 mV, respectively. Moreover, intermediate dipole thin-film layers based on amino acid (Glycine) and conjugate polyelectrolyte (PEI) have also been used to fine-tune the Fermi levels between the metal electrode and the selective layer. The influence of thermal anneals on solar cell performance will be presented and discussed at the conference.

Resume : Recently, 2-dimensional transition metal dichalcogenides (2D-TMDs) have attracted great attention from the research community due to their unique physical and optoelectronic properties. However, their fabrication techniques still have some limitations. Mechanical exfoliation (ME) gives limited-size monolayers that are not suitable for practical devices. Also, the thickness of the exfoliated materials is random, making it difficult to obtain monolayers consistently. Meanwhile, these shortcomings of ME methods can be solved by chemical vapor deposition (CVD) methods. However, the CVD approach often produces non-uniform and much lower-quality materials compared with the ME one. The non-uniformity is caused by a high density of nucleation sites whereas the lower quality is caused by defects such as antisite and vacancies. We successfully synthesize a high-quality large-area MoS2 monolayer employing a combinational phase precursor (CPP) CVD approach. First, we optimize the MoS2 monolayer growth for the solid precursor of molybdenum trioxide (MoO3(s)) by controlling the temperature and pressure at 700, 720, 750oC and 0.025, 0.05, 0.075 MPa, respectively. The MoO3-xSy formation is found across the SiO2/Si substrate (over 1 cm2) at low pressure (0.025 MPa) and temperature (700oC), and confirmed by Raman at 229.796 cm-1, 603.268 cm-1 and 642.855 cm-1. In contrast, saturated nucleation and multilayers are found at the highest temperature (750oC) and pressure (0.075 MPa). Thus, 720oC and 0.05 MPa are our system's optimal growth temperature and pressure to grow MoS2 monolayers (10-15 µm) with the solid precursor MoO3(s). Next, sodium nitrate catalyst is demonstrated to assist the reaction equilibrium of the solid precursor CVD process, leading to an increased density and size of MoS2 monolayer flakes (~120 µm). The metal oxyanion salt formation at an intermediate phase reduces the energy activation, thus improving the growth size and the chemical structure of the monolayer crystal. However, the monolayers’ photoluminescence (PL) intensity is significantly reduced due to excess residues. Therefore, a suspension solution-based precursor is tested using the optimized temperature, pressure, and catalyst from the solid precursor case. It is found to give a high density of uniform triangles with an average size of ~80 µm. Finally, combining both precursor phases (CPP) yields the largest monolayer flakes with an average size of ~200 µm, and the highest quality with PL intensities one order of magnitude higher than of a standard ME monolayer. The results are confirmed by various techniques, including PL and Raman spectroscopy, TRPL, AFM, and HR-TEM. This work opens a door for growing large-area monolayers with much higher quality than the pristine ME one without any chemical treatments, making them promising candidates for high-performance 2D material-based optoelectronic devices such as solar cells, photodetectors, light-emitting diodes, and phototransistors.

Resume : Transition metal dichalcogenides (TMDCs) are versatile two-dimensional (2D) materials that have unique optical and electronic properties. Amongst these, molybdenum disulfide (MoS2) has attracted attention due to its potential applications in high performance transistors as a result of its semiconductor nature and high current on/off ratio (108) [1]. Remarkably, MoS2 possesses a thickness-dependent bandgap that ranges from 1.2 eV for bulk (indirect bandgap) to 1.9 eV for monolayer (direct bandgap). This spans the near infrared to a visible range of the electromagnetic spectrum and potentially makes it suitable for photodetectors and photovoltaics [2]. Due to its layered structure, MoS2 exists in three known polytypes (1T, 2H, 3R). The most stable and comprehensively studied being the 2H configuration which is particularly abundant in the earth’s crust [3]. Ranging from material synthesis to property exploration, both bulk and 2D MoS2 have been extensively investigated. Various deposition techniques such as, physical vapour deposition (PVD) and chemical vapour deposition (CVD) have been proven be successful growth methods of monolayer MoS2 [4]. In addition, solution based methods, such as hydrothermal synthesis and solvothermal synthesis have also shown great promise [5]. However, there are still great challenges on scalable production in terms of large-area uniform growth. Moreover, there are further challenges on controlling the phases of the layers as different phases may coexist within a same flake which can result in a vast discrepancy in optical and electrical properties. Here we demonstrate the epitaxial growth by an Aerosol-Assisted Chemical Vapor Deposition (AACVD) method [6] results in large area coverage (~1 cm2) of MoS2 on both glass and Silicon dioxide (SiO2) substrates. The morphology and thickness can be tuned by precursor concentration and/or growth temperature allowing us to engineer different MoS2 structures, such as nanorods, snowflake-like structures and 2D layers. The as-grown MoS2 structures are characterized using X-ray diffraction (XRD), Raman and electron dispersive X-ray spectroscopies, which confirmed the presence of 2H-MoS2. The XRD results show that the MoS2 structures grown on SiO2 substrates are under strain compared to the growth obtained on glass substrates, which also confirmed by the Raman shift observed in E12g and A1g modes. However, it is apparent that this substrate induced strain does not significantly affect the growth of different morphologies. The successful growth of 2H-MoS2 with different morphologies offer the prospect for large-area device fabrication that exploit distinctive optical and electronic properties. References [1] B. Radisavljevic et al., Nat. Nanotechnol. 6, 147–50 (2011). [2] J. Zhu et al., RSC Adv. 6, 110604–9 (2016). [3] I. Song et al., RSC Adv. 5, 7495–514 (2015). [4] Y. Xie et al., Nanotechnology 28, 1-11 (2017). [5] F. Wang et al., RSC Adv. 8, 38945–54 (2018). [6] A. A. Tedstone et al., Chem. Mater. 29, 3858–62 (2017).

Resume : High-performance, lightweight solar cells with low-cost and non-toxic substituents are a major target in the field of solar photovoltaics.1?4 Ultrathin solar cells can reduce material consumption, weight and manufacturing demand, directly lowering the cost, in addition to benefitting quantum efficiency and photovoltaic (PV) performance. To enable ultrathin architectures, however, strong optical absorption is an absolute necessity. Typically viewed as a fixed material parameter, efforts to enhance light trapping and realize thinner solar absorbers have focused on optical architectures that add complexity, cost and additional loss mechanisms.5 Here, we demonstrate that engineering the cation distribution and disorder in an emerging ternary chalcogenide absorber can lead to major absorption enhancement, enabling ultrathin PV devices with short circuit current density of 27 mA·cm-2 and a record power conversion efficiency of >9%,6 at a film thickness of only 30 nm.6 We rationalise this behaviour through in-depth modelling of the thermodynamics of cation interactions in this disordered system, alongside simulations of the electronic structure and optical properties as functions of cation distribution. Our calculations show that optical transition moments can be maximised through control of the cation disorder in the material, enabling the absorption and efficiency enhancements realized experimentally in this work. Our findings provide insights and guidance to researchers working in the emerging field of disordered semiconductor materials, both in terms of electronic and optical tuning strategies, and accurate simulation procedures for predicting the impact of disorder on material parameters. (1) Huang, Y.-T.; Kavanagh, S. R.; Scanlon, D. O.; Walsh, A.; Hoye, R. L. Z. Perovskite-Inspired Materials for Photovoltaics and beyond?from Design to Devices. Nanotechnology 2021, 32 (13), 132004. https://doi.org/10.1088/1361-6528/abcf6d. (2) Buchanan, M. Perovskite Fever. Nature Physics 2020, 16 (10), 996?996. https://doi.org/10.1038/s41567-020-01060-8. (3) Li, Z.; Kavanagh, S. R.; Napari, M.; Palgrave, R. G.; Abdi-Jalebi, M.; Andaji-Garmaroudi, Z.; Davies, D. W.; Laitinen, M.; Julin, J.; Isaacs, M. A.; Friend, R. H.; Scanlon, D. O.; Walsh, A.; Hoye, R. L. Z. Bandgap Lowering in Mixed Alloys of Cs2Ag(SbxBi1-x)Br6 Double Perovskite Thin Films. J. Mater. Chem. A 2020, 8 (41), 21780?21788. https://doi.org/10.1039/D0TA07145E. (4) Krajewska, C. J.; Kavanagh, S. R.; Zhang, L.; Kubicki, D. J.; Dey, K.; Ga?kowski, K.; Grey, C. P.; Stranks, S. D.; Walsh, A.; Scanlon, D. O.; Palgrave, R. G. Enhanced Visible Light Absorption in Layered Cs 3 Bi 2 Br 9 through Mixed-Valence Sn( II )/Sn( IV ) Doping. Chem. Sci. 2021, 12 (44), 14686?14699. https://doi.org/10.1039/D1SC03775G. (5) Massiot, I.; Cattoni, A.; Collin, S. Progress and Prospects for Ultrathin Solar Cells. Nat Energy 2020, 5 (12), 959?972. https://doi.org/10.1038/s41560-020-00714-4. (6) Yongjie Wang & Seán R. Kavanagh et al. Cation disorder engineering yields AgBiS2 nanocrystals with enhanced optical absorption for efficient ultrathin solar cells. Nat. Photon. 16, 235–241 (2022)

Resume : In recent years, antimony sulfoselenides (Sb2(S,Se)3-x) have emerged as a promising absorber material in solar cells owing to their high absorption coefficient (>10-5) and tunable bandgap (1-1.8 eV), low melting points. Further, relatively non-toxic, less expensive and earth abundant elements present in this compound add their technological importance in photovoltaics. In the present work Sb2(S,Se)3-x , we prepare Sb2(S,Se)3-x thin films by sulfur replacement in chemically deposited Sb2S3 films using Se nanocolloids obtained by laser irradiation of Se powder. A comparative analysis was done on the crystalline structure, morphology and optoelectronic properties of Sb2(S,Se)3-x films and pure Sb2S3 films. We observed an increment in the photocurrent values after Se incorporation compared to the pure compound. Bandgap values were reduced for Sb2(S,Se)3-x compared to Sb2S3 films. The complete study of structural, morphological, and opto-electonic properties of the Sb2(S,Se)3-x films will be presented in the conference.

Resume : Antimony selenide is one of the emerging absorbers with promising photovoltaic properties such as high absorption coefficient (105/cm) and direct bandgap of 1.2 eV. Being low cost and nontoxic material, it has big upper hand above CdTe and CIGS absorbers in the race towards development of sustainable environmentally friendly energy sources. Different groups discovered that the physical methods, such as close spaced sublimation (CSS) and vapour transport deposition (VTD), with high deposition rate, allows deposition of Sb2Se3 thin films with columnar ribbon like structure ensuring efficient charge carrier transport in the solar cell device. The most commonly used buffer layers for Sb2Se3 thin film solar cells are titanium dioxide (TiO2) and cadmium sulphide (CdS). For the latter one, remarkable solar cell efficiencies, between 7-9% were achieved by several groups, employing close-spaced sublimation and VTD techniques, however, there is still little research on the development and optimization of VTD Sb2Se3 in combination with TiO2 heterojunction partner layer. In particular, there is no yet consensus on deliverability of the optimal VTD processing conditions for fabrication of Sb2Se3 in relation with TiO2 buffer layer. Considering these aspects, this work provides systematic study of effect of VTD deposition conditions, including substrate temperature, source to substrate distance and deposition rate) on the properties of Sb2Se3 thin films grown onto TiO2. The analysis of the thin film properties was done by XRD, SEM, and UV-Vis. From XRD and SEM analysis the structural and morphological properties of the films were identified. The films have desired columnar structure with predominant (221) crystallite orientation. The impact of VTD Sb2Se3 deposition conditions on the performance of superstrate configuration glass/FTO/TiO2/Sb2Se3/Au solar cell was analysed by current voltage and quantum efficiency. A maximum efficiency of 3.9% was achieved for antimony selenide solar cells with source to substrate distance of 10 cm, 20°C/min ramping rate, 5 min deposition time, 500°C source temperature and 450°C substrate temperature.

Resume : The search for new semiconductor materials and the improvement of the existing materials represent a major focus for emerging photovoltaic device integration and innovative applications. Doping is a commonly used approach for modifying and widening physical and chemical semiconductor properties, while alloying provides simultaneous adjustment of both the band gap and lattice constant, allowing for an increase in radiant efficiency at a broader range of wavelengths. In this study, a novel approach of deposition by close-spaced sublimation of Bismuth-based thin film alloys is investigated. Three systems are analyzed: Bi2S3-SnS, Bi2S3- Sb2S3 and Bi2S3-CuS, obtained by in-situ addition of SnS, Sb2S3 and CuS (with concentrations between 0.1 and 10 at%) into Bi2S3, respectively. Bi2S3 thin films exhibited Eg~1.25 eV, n-type conductivity, and unexpected high electron concertation of ~1017-1018 cm-3. SnS introduced p-type conductivity and adjust the carrier density in Bi2S3 whereas no significant changes in morphology, structure, and band gap compared to Bi2S3 films were observed. Bi2S3-Sb2S3 films show basic orthorhombic structure but have a larger lattice parameter due to the substitution of Bi for Sb in Sb2S3 lattice while keeping high resistivity (≥105 Ωcm) behavior as characteristic for Sb2S3 films. For Bi2S3-CuS system, CuS with Eg~2.5 eV introduces p-type conductivity and widens the band gap of Bi2S3. Variation of the band gap, between 1.1-1.6 eV is observed for the investigated alloys, opening new opportunities for the development of a novel absorber material for photovoltaic applications.

Resume : There is always a demand and a big interest in the development of low-cost solar cells based on technology that requires only earth-abundant and non-toxic materials. One such material is antimony sulphide (Sb2S3) which is suitable for semi-transparent solar cell applications. Sb2S3 has plenty of advantages that make it a promising material for usage as the absorber in thin-film solar cells, like low-cost, earth-abundance and environmental-friendliness, suitable direct band gap of around 1.7 eV, high light absorption coefficient (around 1.8 x 105 cm-1 in the visible region at 450 nm). The main focus of the current study is to investigate the photovoltaic performance of the devices by varying the Sb2S3 absorber layer, and the TiO2 window layer thickness and to find the most optimal conditions for efficient solar cells. TiO2 and Sb2S3 layers were deposited using ultrasonic spray pyrolysis in the air which is a low-cost, simple and scalable thin film deposition method. The TiO2 and Sb2S3 layer were deposited onto prefabricated Glass/FTO (fluorine-doped tin oxide). The TiO2 layer was deposited by USP at 340°C and annealed in air at 450°C for 30 minutes. The absorber was deposited by USP from a precursor solution (SbCl3/SC(NH2)2 = 1:3) at 195°C in air and followed by annealing in a vacuum at 175°C for 5 minutes. The hole-transport material – P3HT was deposited using sping coating in ambient conditions. The studied solar cells have the general structure of Glass/FTO/TiO2/Sb2S3/P3HT/Au and are assembled in superstrate configuration. The properties of the separate layers, like crystal structure, morphology and optical properties were studied using different analysis tools, such as UV-Vis, Raman and XRD measurements. The photovoltaic performance of the assembled solar cell devices was analyzed by current-voltage(I-V) and external quantum efficiency(EQE) measurements. It was shown, that increasing the Sb2S3 deposition times from 30 min to 45 min leads to an increase in thickness from ca 60nm to ca 90nm. Increasing the TiO2 layer deposition times from 36 min to 46 min, resulting in thickness change from 55 nm to 85 nm. Results have shown, that increasing the thickness of the absorber and proportionally also the thickness of the window layer increases the current output and efficiency up to 5% (with the contact area of 7.064mm2) of the solar cell. The resulting Glass/FTO/TiO2/Sb2S3/ structure has an average visible transmittance value higher than 20% which qualifies them for use in semi-transparent applications. Corresponding results will be discussed.

Resume : The use of impurities, introduced into the semiconductor material in a controlled manner, with the intention changing the resistivity by several orders of magnitude, forms the absolute basis for the fabrication of any semiconductor device from integrated circuits to semiconductor lasers and solar cells. The main reason for the increased interest in ZnTe semiconductor is its wide application in light-emitting diodes, solar cells, infrared detectors, electroluminescence devices, photocatalysis and THz emitters, etc. Despite the long-term scientific research, the problem of studying energy parameters of defects for ZnTe is still relevant and today. Motivated by this, we investigate how the copper doping from solution affects the structural and photoluminescence properties of ZnTe thin films obtained by close space sublimation method deposited on glass substrates. The deposited films were immersed in solution with different concentrations, then heated in vacuum. The structure and composition are studied using X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS). XRD analysis of undoped and Cu-doped ZnTe thin films exhibited polycrystalline behavior with highest sharp peak corresponding to (111) direction as a preferred orientation. The resistivity of the film, immersed for 30 min, was reduced by four orders of magnitudes. The XPS analysis confirmed the presence of Cu2+ oxidation state and the formation of Cu2Te phase. Also, we explore Cu electronic states in ZnTe using photoluminescence as the main investigative method. Furthermore, the effect of Cu-doped ZnTe layer have been optimized for producing ZnO/CdS/ZnTe solar cell with maximum working parameters such as: open-circuit voltage (Voc), short-circuit current density (Jsc), fill factor (FF), photovoltaic conversion efficiency (η).

Resume : Gallium Sulphide thin films are highly perspective for being used in photodetector applications. Ga2S3 thin films have been prepared by using of Close-Spaced Sublimation (CSS) method on n-Si (100), 4-8 Ohm·cm and p-Si, 10 Ohm·cm, as-synthesized Ga2S3 as source compound. The substrate temperature in the range of 500-700°C has been used, while the source temperature was 800°C. The GI-XRD (CuKa) investigations have been used for structural characterizations. (including Rietveld refinement). A detailed phase analysis has been carried out for source material during sublimation cycles as well as for as-prepared thin films. It has been shown that after a certain number of depositions Ga2S3 thin films showed the presence of GaS phase along with Ga2S3 (initial source material didn’t’ show any structural modification). Crystallites’ size and film strain analysis has been carried out involving Williamson-Hall method and preferential crystallites growth orientation (March-Dollase factor). These parameters are correlated to substrate temperature. The morphology of the samples surface has been investigated by both SEM and AFM (and STM) revealing crystallites conglomerates and even flakes (with growth steps) having linear dimensions in the micrometer range. The investigations revealed a direct relation between the physical properties of the investigated Ga2S3 layers to the deposition technology. Acknowledgements: NARD (ANCD) projects: 20.80009.5007.12, 20.80009.5007.02

Resume : Two dimensional-transition metal dichalcogenides (2D-TMDs), such as MoS2, MoSe2, etc., are attractive quantum materials with extraordinary light absorption at visible wavelengths and tunable band gap in the range of 1-2 eV [1]. Theoretical predictions suggest that solar cell devices employing TMDs exhibit the highest power conversion efficiency per weight among all solar cell technologies, along with exceptional mechanical flexibility. Moreover, optoelectronic devices based on 2D-TMDs can attain open circuit (Voc) values in the range of 1.1-1.3 V, particularly for MoS2 and WS2. Lastly, the introduction of foreign atoms in TMDs allows a controllable modulation of optical and electronic properties as a promising route to unlock the full potential of TMDs for optoelectronic applications. Indeed, semiconducting TMDs can be substitutionally doped with Group V (e.g, Nb) and Group VII (e.g., Re) elements for p- and n-type doping, respectively [2]. To enable the large-scale implementation of TMDs, it is essential to develop controllable synthesis approaches to engineer the composition and phase of continuous TMD films. In this work, we will present our study on mono- and multilayer MoS2 prepared by sulfurization of epitaxial oxides grown by Pulsed Laser Deposition (PLD) [3]. We will show that PLD can allow for a great tunability of oxygen vacancy content in the transition metal oxides and this allow a more facile conversion from oxides to TMDs. Next, we will discuss two different approaches for doping of TMDs by PLD through mixing of dopant in the precursor oxide films or post-synthesis implantation technique using PLD. We will use combined X-ray photoelectron spectroscopy (XPS) and ultraviolet photoelectron spectroscopy (UPS) to determine the effect of doping on the band structure and band alignment of van der Walls homojunction TMDs. Raman and photoluminescence spectra of pristine and doped MoS2 will be used to unambiguously determine the effect of doping on the exciton emission in TMDs. Finally, we will discuss the performance of the TMD photovoltaic devices, as well as the relation between the photocurrent generation in n-MoS2/p-MoS2 homojunction and the number of TMD layers. References 1. K. F. Mak, C. Lee, J. Hone, J. Shan, and T. F. Heinz, Physical Review Letters 105, 136805 (2010). 2. L. Loh, Z. Zhang, M. Bosman, and G. Eda, Nano Research 14, 1668 (2021). 3. F. Bertoldo, R. R. Unocic, Y.-C. Lin, X. Sang, A. A. Puretzky, Y. Yu, D. Miakota, C. M. Rouleau, J. Schou, K. S. Thygesen, D. B. Geohegan, and S. Canulescu, ACS Nano 15, 2858 (2021).

Resume : Kesterite materials based on Cu2ZnSnS4 (CZTS), Cu2ZnSnSe4 (CZTSe) and alloys thereof (CZTSSe) have band gap energies ranging between 0.9 eV and 1.5 eV. Larger band gap energies would be needed for applications such as tandem or multijunction solar cells and several alloying options exist that can provide band gap widening. One option is exchange of Sn with Ge, for which band gap energies can be increased up to around 1.5 eV for Cu2ZnGeSe4 (CZGSe) and to 2.2 eV for Cu2ZnGeS4 (CZGS). Several studies have shown improved crystallization and device performance from addition of small amounts of Ge to CZTSe. There are less studies showing good device performance for the higher Ge or S contents that are needed to reach above 1.7 eV in band gap energy. In our work, we use compound co-sputtering from sulfide or selenide binary targets for precursor deposition, followed by annealing in sulfur or selenium atmosphere for absorber formation. Some devices were made with addition of low amounts of Ge to CZTSe and CZTSSe. In this case device performance and open circuit voltage (VOC) deficit was improved. For Cu2ZnGexSn1-x(SySe1-y)4, a challenge for characterization is accurate determination of composition, i.e. x and y, and interpretation of XRD and Raman peak shift from the double alloying. For CZTS, alloying with Ge was investigated using both GeS target and metallic Ge target [1]. With the GeS target, recrystallisation during annealing was found to be inhibited. A possible reason is the formation of wurtzite phase material in the more sulfur rich precursor sputtering. From variation of the Ge content in CZTS, slightly improved grain growth was seen for Ge/Ge+Sn of 0.2 but not for higher Ge contents. Device performance was not improved for CZTS with Ge alloying, and relatively low VOC was obtained for Cu2ZnGeS4 (CZGS) and alloyed CZGTS using standard device structure including CdS buffer layer. Since the band gap increase with Ge alloying is expected to mainly occur in the conduction band, the poor energy band alignment between CZTS and CdS is expected to become even worse. In order to improve this, the alternative buffer layer material Zn1-xSnxOy (ZTO), deposited by atomic layer deposition was studied [2]. The VOC was substantially increased with ZTO as compared to CdS, reaching up to 1.1 V, but the VOC deficit for the best CZGS/ZTO device was still higher than for our record CZTS/CdS or CZTS/ZTO devices. Peeling of CZGS during etching or wet chemical processing was an issue, and CZGS/ZTO devices with varying ZTO deposition conditions, and thereby ZTO band gap energies, were fabricated without KCN etching. For this series, the VOC was relatively constant, most likely due to an oxide interlayer remaining between absorber and buffer. Based on these results, possibilities for improved performance of CZGTSSe devices above 1.7 eV band gap energy will be discussed. [1] N. Saini, J. K. Larsen, K Sopiha, J. Keller, N. Ross and C. Platzer-Björkman, Physica Status Solidi A (2019) 1900492 [2] N. Saini, N. Martin, J. K. Larsen, A. Hultqvist, T. Törndahl and C. Platzer Björkman, Solar RRL (2021) 2100837

Resume : As the demand for renewable energy sources increases ahead of net carbon-zero by 2050 [1], the commercial availability of a high-efficiency, low-cost thin film solar material becomes increasing important. Cu2ZnSn(S,Se)4 (CZTSSe) is an earth-abundant, non-toxic thin film solar absorber with a tuneable direct band gap between 1.0 eV and 1.5 eV, and a theoretical power conversion efficiency (PCE) of 20 % under AM1.5G illumination [2]. However, CZTSSe devices are currently limited by significant voltage (Voc) losses which have been linked to structural disorder including CuZn and SnZn antisite defects, secondary phases (including Cu2S, SnS and ZnS), and high concentration of point defects clusters, such as (2CuZn and SnZn clusters) [3]. These defects can cause bandgap fluctuations and reduce the minority carrier lifetime through charge carrier recombination. Consequently, controlling elemental disorder, particularly in Sn occupied sites is key for pushing the performance of CZTSSe devices. A recent study has shown that replacing Sn(II) precursors by Sn(IV) complexes in solution based processed CZTS can lead to a substantial increase in device performance [4]. SnCl2 is a well-established Sn precursor in solution processed methods, however, this precursor may cause the formation of Sn secondary phases and deep-level SnZn point defects and clusters, which have been recognised as the main cause of Voc deficits (Voc,def). Building on our previous work [5,6], in this study, energy-filtered photoemission electron microscopy (EF-PEEM) has been used to provide high resolution images of the surface electronic landscape of the CZTSSe absorber layer upon fabricating films using Sn(II) and Sn(IV) precursor sources. The use of a Sn(IV) precursor source may decrease Sn defects and Sn(II) secondary phases, suppressing Voc,def and improving overall performance. EF-PEEM can probe local fluctuations in the work function (WF) of CZTSSe films on the sub-micron scale, allowing for the mapping of WF tailing, where a narrower, even distribution indicates a more uniform WF across the surface. [1]: United Nations, “Paris Agreement”, in United Nations Treaty Collection, 2015 [2]: S. Kim, J. A. Márquez, T. Unold and A. Walsh, Energy Environ. Sci., 2020, 13, 1481-1491 [3]: N. Kattan, B. Hou, D. J. Fermin, D. Cherns, Applied Materials Today, 2015, 1, 52-59 [4]: Y. Gong, Y. Zhang, E. Jedlicka, R. Giridharagopal, J. A. Clark, W. Yan, C. Niu, R. Qiu, J. Jiang, S. Yu, S. Wu, H. W. Hillhouse, D. S. Ginger, W. Huang, H. Xin, Sci China Mater., 2021, 64, 52-60 [5]: D. Tiwari, M. Cattelan, R. L. Harniman, A. Sarua, A. Abbas, J. W. Bowers, N. A. Fox, D. J. Fermin, iScience, 2018, 9, 36-46 [6]: D. Tiwari, M. Cattelan, R. L. Harniman, A. Sarua, N. A. Fox, T. Koehler, R. Klenk, D. J. Fermin, ACS Energy Lett., 2018, 3, 2977-2982

Resume : Quaternary chalcogenides have gained a lot of attention, especially kesterite-type semiconductors Cu2ZnSn(S,Se)4 (CZTSSe) which consist of earth abundant and non-toxic elements. These compounds are promising low-cost alternative absorbers for thin film solar cells due to their suitable criteria for photovoltaic applications, a record conversion efficiency of 13.5 % was reported recently [1]. In such high efficient devices the polycrystalline absorber layer shows an off-stoichiometric composition. Deviations from stoichiometry cause intrinsic point defects in the material (vacancies, anti-sites, interstitials), determining the electronic properties of the semiconductor significantly. It is agreed in literature that large band tailing observed in Cu-based kesterite-type semiconductors causes voltage losses limiting the efficiency of kesterite-based devices. The Cu/Zn disorder, which is always present in these compounds [2], is discussed as a possible reason for band tailing. Structural characterization is conventionally based on X-ray diffraction, but isoelectronic cations, like Cu and Zn2 , are difficult to distinguish when analyzing diffraction data. But their neutron scattering lengths are different, thus their site occupations in the crystal structure can be determined by neutron diffraction. The experimental determination of the order parameter Q, which is a quantitative measure of the degree of Cu/Zn disorder [3], requires a differentiation between these cations as well. Kesterite-type based thin film solar cell technologies are mainly based on thin absorber layers, which makes it quite hard to correlate the crystallographic structure (determined via neutron diffraction) to a photovoltaic performance of these materials. A promising alternative technology uses CZTSSe monograins (single crystals, 50-100 μm size) fixed in a polymer matrix to form a flexible solar cell. We will present a detailed structural investigation of CZTSSe monograins based on neutron powder diffraction, examining the influence of small changes in their chemical composition on Cu/Zn disorder as well as determining the type and concentration of intrinsic point defects present. We present correlating trends between chemical composition, presence of secondary phases, Cu/Zn disorder, occuring intrinsic point defects as well as the optical bandgap (obtained by diffuse reflectance) of the CZTSSe monograins with the stability and efficiency of respective devices. This work has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement no. 952982 (Custom-Art). Part of this research used resources at the SNS, a DOE Office of Science User Facility operated by the ORNL. Authors gratefully acknowledge the PSI and SINQ for providing beamtime at the HRPT diffractometer. [1] X. Xu et al., Adv. Energy Mater. (2021) 2102298 [2] S. Schorr et al., J. Phys.: Energy 2 (2020) 012002 [3] D. Toebbens et al., Phys. Stat. Sol. B 253 (10) (2016) 1890

Resume : One of the paramount goals in the field of photovoltaic device is fabrication of highly efficient flexible thin film solar cells. Kesterite Cu2ZnSn(S,Se)4 (CZTSSe) has attracting attention as promising earth-abundant material for low-cost thin film solar cells and wide research on CZTSSe solar cells raise its efficiency to over 11% [1]. However, little is reported about changes in its characteristics occurred by mechanical bending. In this study, carrier transport of CZTSSe under mechanical bending was investigated. CZTSSe samples deposited on flexible Mo foils were prepared with and without Na doping. Efficiency of Na doped CZTSSe sample was 11.19% and over 93% and 90% of initial efficiency was maintained after the 1000 cycles of concave and convex bending, respectively. Degradation of open-circuit voltage (VOC) was the largest among other parameters therefore, photo-assisted Kelvin probe force microscopy was utilized to investigate the local VOC loss. Under the mechanical bending state, a change in surface voltage degraded due to the limited photogenerated carrier transport resulting in decrease of Fermi level splitting and local VOC. Interestingly, CZTSSe sample with Na showed the lower degradation of local VOC compared to the sample without Na suggesting that the existence of Na element hindered the recombination of photogenerated carriers under mechanical stress. In addition, through low-temperature photoluminescence measurement and ultraviolet photoelectron spectroscopy, changes in electric band structure under mechanical bending was observed. Role of Na in transport of photogenerated carriers under mechanical bending will be investigated further to suggest the optimal Na incorporation for robust absorber.

Resume : Semi-transparent solar cells make the extension of photovoltaics into our daily life possible. In the last years, kesterite-type material has attracted a special attention to be used as a sustainable absorber in thin-film solar cells because of its low toxicity and earth abundant elements. The substitution of Sn with Ge increases the band gap energy Eg of the semiconductor material. The full potential of the total substitution of Sn with Ge in kesterite-based material has been demonstrated by achieving an efficiency of 8.5 % and Voc values of 65 % of the theoretical limit for Cu2ZnGeSe4 (CZGSe)-based solar cells [1]. In this work, 900 nm thick CZGSe thin films are grown by co-evaporation and subsequent annealing at maximum temperatures of 480 or 525 ºC onto semi-transparent Mo/V2O5/FTO back contacts. The higher annealing temperature has led to CZGSe layers with bigger grain size, a reduction of GeSe2 secondary phase and a thicker MoSe2 at the back interface of the solar cell. This strategy of increasing annealing temperature allows for a reduced cross-over between the illuminated and dark J-V curves and higher EQE that results in an improved device performance from 4.1 to 5.8 % when using 480 and 525 ºC, respectively. Both absorber layers present an optical band gap energy of 1.47 eV. A total transmittance higher than 50 % in the near-infrared spectral range is obtained for the CZGSe layer annealed at 525 ºC, which can be explained by the higher diffuse transmittance related to the larger grain size [2]. In order to further increase the transparency, kesterite thin films with higher Eg are fabricated by sulfurization of co-evaporated CZGSe, producing Cu2ZnGe(S,Se)4 (CZGSSe) thin films. The influence of transparent back contacts is investigated on CZGSSe-based PV devices. For that, different back contact configurations are studied, Mo/V2O5/FTO or Mo/FTO varying the thickness of the Mo layer. Efficiencies of 3.1, 2.7 and 1.5 % with Eg = 1.73, 1.85 and 1.91 eV respectively, are achieved. CZGSSe thin films of only 400 nm thickness are deposited to increase the transparency, obtaining straight-through transmittance near 40 % at around 1150 nm, Eg = 2 eV and device performance of 0.8 %. The transparent back contact affect not only the PV parameters, but also the optical properties and the in-depth elements distribution through the absorber layer. Further optimization on the annealing treatment to develop the proper [S]/([S]+[Se]) band gap grading together with the back transparent interface is under investigation to meet the challenge of increasing performance and transparency at the same time. These results show a great potential for Building Integrated Photovoltaic applications among others. [1] L. Choubrac et al, Appl. Energy Mater. 3 (2020) 5830. [2] A. Ruiz-Perona et al, J. Phys. Mater. 4 (2021) 034009.

Resume : Semi-transparent photovoltaic technology combines the conversion of light into electricity with visible light transparency, which allows its integration as windows and skylights within energy-sustainable buildings [1]. Different types of solar cells have emerged with potential for semi-transparent applications [1], including thin-film Cu(In,Ga)Se2 (CIGS). Solar cells using CIGS as an absorber material achieved a record of 23.4% power conversion efficiency [2]. To fabricate a semi-transparent CIGS solar cell, the back contact needs to be replaced by a transparent back contact (TBC). To use a TCO as a TBC, there are some requirements, i.e. stability against thermal stress during absorber formation, formation of an ohmic contact to the absorber, high transparency and low sheet resistance [3]. Hydrogen-doped In2O3 (IO:H) has been investigated due to its promising high mobility when compared to other TCOs (ZnO:Al, SnO:F and In2O3:Sn) [4]. Usually, IO:H thin films are deposited by radio-frequency (RF) magnetron sputtering, with water vapour or hydrogen as dopant source [4]. In this work, a RF sputtering deposition process for IO:H is investigated, where oxygen was pulsed during the entire deposition and H was supplied from a Ar/H (5%) mixed gas. To optimize the optoelectronic properties, sputtering conditions were varied, such as O2 partial pressure, Ar/H2 partial pressure and duration. The O2 pulses consisted in providing oxygen for one minute in intervals of 2 minutes (1’ ON 2’ OFF). After deposition, the films were post-annealed in vacuum at 200 °C for one hour. The film structural properties were evaluated by X-ray diffraction (XRD), electrical properties by 4-point probe and, the optical properties by a UV-Vis spectrophotometer. The oxygen plays an important role in optoelectronic properties, where a high content of oxygen allows higher transparency but also increases the film sheet resistance. Therefore, a balance between electrical and optical properties is needed. An IO:H film with average visible transparency over 80% and sheet resistance after annealing below 20 ohms/sq was obtained. [1] J. Sun and J. J. Jasieniak, “Semi-transparent solar cells,” J. Phys. D. Appl. Phys., vol. 50, no. 9, p. 093001, Mar. 2017. [2] National Renewable Energy Laboratory, “Best Research Cell Efficiencies.” https://www.nrel.gov/pv/cell-efficiency.html, 2021. [3] J. Keller et al., “Using hydrogen-doped In2O3 films as a transparent back contact in (Ag,Cu)(In,Ga)Se2 solar cells,” Prog. Photovoltaics Res. Appl., vol. 26, no. 3, pp. 159–170, Mar. 2018. [4] M. L. Addonizio, A. Spadoni, A. Antonaia, I. Usatii, and E. Bobeico, “Hydrogen-doped In2O3 for silicon heterojunction solar cells: Identification of a critical threshold for water content and rf sputtering power,” Sol. Energy Mater. Sol. Cells, vol. 220, no. October 2020, p. 110844, 2021.

Resume : Ag alloying in Cu(In,Ga)Se2 (CIGS) has been found to significantly lower down the growth temperature and enlarge the grain size with better crystallinity. In addition to the manufacturing change in temperature, the substitution of Cu may also bring different defect properties and post-processing reactions. To further investigate the Ag alloying effect, (Ag,Cu)(In, Ga)(S, Se)2 (ACIGSSe) by surflization after selnization (SAS) process with Ag/(Ag Cu) ratio ranged from 0, 0.01, 0.05 to 0.1 were designed to evaluate the photovoltaic performance and clarify metastable characteristics. The composition analysis and elemental depth profile were under energy-dispersive X-ray spectroscopy(EDS) and X-ray photoelectron spectroscopy (XPS), respectively. SunVoc, Bias external quantum efficiency (bias-EQE), and temperature-dependent JV (JVT) were used to analyze the interface recombination at the pn junction. Various stress (light soaking and dark-aging)-induced or light-induced metastable defects were examined by the Admittance spectroscopy (AS) under dark and intentional UV light illumination. Our results demonstrate that Ag alloying effectively impedes S in-depth diffusion during the sulfurization process, and less S accumulation at the back contact. The higher Ag/(Ag Cu) ratio, the higher the light-induced defect density, while the activation energy of the trap was not proportional to the Ag incorporation, which may relate to the selenium–copper divacancy complex (VSe –VCu).

Resume : High sulfur content accumulated at the backside of the Cu (In, Ga) (S, Se)2 (CIGSSe) absorber is often observed under the precursor with sulfurization after the selenization (SAS) process. And this raised value of S/(S+Se) ratio at the back side results in distinct rollover at the first quadrant of the current density-voltage (J-V) curve, which may come from the presence of double diode formation. High S content in CIGSSe could lower valence band energy then form a Schottky diode to Mo or even Mo(S,Se)2/Mo contact. The reversely second diode repels the carrier transport, then deteriorates the fill factor and further impacts the efficiency. After heat soaking (HS), the distortion in illuminated J-V curve under high forward bias of CIGSSe solar cells was reduced and the cell performance was significantly improved with increased fill factor. In the work, the CIGSSe with heat soaking was investigated, the secondary ion mass spectrometry (SIMS) measurement was used to reveal the composition variation of S/(S+Se) ratio before and after heat soaking. The actual Schottky barrier height was extracted by current-voltage temperature dependent (JVT). Our results show a pronounced decrease in Schottky barrier height after heat soaking.

Resume : Cadmium sulfide (CdS) is one of n-type II-IV family semiconductors, which is the most common buffer layer material with favorable band alignment both to Cu(In, Ga)Se2 (CIGS) and ZnO. Multiple approaches are able to deposit CdS buffer layer on CIGS absorber layer, such as chemical bath deposition (CBD) and physical vapor deposition (PVD) methods. While CBD process is regarded as a universal method to obtain homogeneous film without pin-hole at the CdS/CIGS interface. The CBD solution also simultaneously removes the surface oxides and promotes Cd diffusion within CIGS that generated a buried junction. On the other hand, some impurities such as Cu from CIGS absorber was also found to dissolve out into the CdS during CBD process, indicating the buffer chemistry indeed interacts with CIGS surface circumstance. But the impact from other impurities such as light alkali (Na) and heavy alkali (Cs) metal ions are less discussed in the CIGS solar cells system. In the work, the buffer chemistry of chemical bath deposition-CdS thin film on the Cs-treated CIGS absorbers is investigated by scanning surface potential microscopy (SPoM). An intentional alkali metal incorporation (Na and Cs) in the CBD solution was also used to further confirm the alkali metal effect on the point of view of the CdS thin film to CIGS photovoltaic characteristics.

Resume : Transparent conductive oxide (TCO) front electrical contact for Cu(In,Ga)Se2 (CIGS) thin film solar cells is generally deposited by magnetron sputtering, which utilises a (Ar) plasma ignited via direct current (DC) or radio frequency (RF). The ions generated in the plasma bombard and thereby liberate particles from the target, which are then transferred to the substrate. The exposure of the substrate to plasma emission, ions, atoms, particles and heat can harm sensitive substrates (buffer/CIGS stack) by generating defects or structural changes, thereby compromise device performance. "Sputter damage" is frequently suggested as an explanation for non-ideal device behaviour, albeit direct investigations of sputter damage on CIGS are rare. This work investigates the influence of rf co-sputtering of ZnO and MgO from ceramic targets on bare CIGSe absorbers as well as on CIGS absorbers with either CBD-CdS or ALD ZnMgO buffer layers. Carrier lifetimes in CIGS have been measured before and after the film deposition to quantify its impact. The possible interactions of the plasma with the absorber are discussed and related to device performance via TRPL and JV measurements. Additionally, aging and recovery effects of the absorber related to storage in air, N2 and UHV are investigated. Measurements and analysis of different samples show that exposure to plasma decreases carri-er lifetimes in CIGS, with strong impact of target power and exposure time. The main mechanism of degradations seems to be related to O- ions originating from the ceramic targets. The decrease in carrier lifetime coincides with a decrease in Voc of the completed devices. SCAPS simulations are used to support the hypothesis of recombinatory defects at the absorber front interface. In conclusion, this work implements a metric to quantify the impact of sputter and plasma damage on CIGS absorbers treated and stored under several conditions, and provides design rules to mitigate said damage by optimization of process conditions and buffer choice.

Resume : Semi-transparent solar cells have attracted massive interest in the photovoltaic industry as they allow the integration of photovoltaics in buildings, e.g. as windows. Cu(In,Ga)Se2 (CIGSe) photovoltaics are a promising option for such applications. There are two different approaches to achieve transparent CIGSe solar cells. One involves the fabrication of ultra-thin CIGSe absorbers on see-through transparent back contacts. This method is, however, limited by the formation of GaOx parasitic layers at the back contact interface [1], which leads to significant efficiency losses. An alternative approach consists of micro-patterning the solar cell in different shapes, giving it the transparency characteristic. In this work, a micro-structuring approach is investigated, where an opaque CIGSe solar cell is spatially segmented into many micro-sized line-shaped solar. First, a photolithography step defines a pattern of micro-lines. By varying the line’s width and spacing, it is possible to control the window average visual transparencies (AVT). Next, a bromine solution is used to etch the solar cell areas unprotected by the photoresist. The acidic nature of bromine solutions makes it a challenge to etch the CIGSe layer without damaging the photoresist. Here, two different bromine solutions with different concentrations were compared: bromine: methanol and aqueous bromine. For lower concentrations of bromine: methanol solutions (c=0.02 mol/l), the etch rate was very low, and the solution lost etching efficiency with time, due to the highly volatile nature of bromine. Increasing the concentration enabled a higher etching rate. However, the photoresist was etched, meaning that the bromine: methanol solution is not viable to perform the desired etch. For the aqueous bromine solutions, similar results were obtained for lower concentrations of 0.02 mol/l, where the etching rate was too low. However, for higher concentrations (c=0.1 mol/l) different results were obtained. The solution did not damage the photoresist and provided a high etching rate. Therefore, it was possible to etch the CIGSe layer completely, while maintaining the desired micro pattern. After the bromine etching, the sample was dipped in sodium hypochlorite (commercial bleach) to remove the exposed molybdenum and create a molybdenum undercut. The aqueous bromine solution is also capable of etching the buffer and window layers, thus with two etching steps, it is possible to completely etch the CIGSe stack. This work established a baseline process leading to the development of semi-transparent CIGSe photovoltaics. Here, we achieved a solar cell efficiency of 7.9% for an AVT of 50.7.%. References: [1] S. Muhammad et al., Solar Energy Materials and Applications 178, 29-37 (2018)

Resume : Absorber layers for high-efficiency Cu(In,Ga)Se2 (CIGS) solar cells are typically deposited by co-evaporation of the individual elements or by sputter deposition of Cu, In, and Ga and subsequent annealing in selenium atmosphere. The typical 3-stage co-evaporation process starts with In-Ga-Se deposition, followed by Cu-Se deposition until a Cu-rich composition is reached, and finishes with an In-Ga-Se deposition, to achieve overall Cu-poor composition. The Cu-rich stage leads to larger grain sizes. For sputter deposition, a variety of approaches have been reported, and the most common process is based on the deposition of a Cu-In-Ga precursor layer, followed by annealing in selenium atmosphere. Other approaches use reactive sputtering in H2Se atmosphere or sputtering from a single CIGS target with or without subsequent annealing. However, most of these sputtering processes do not provide the benefits of Cu-poor and Cu rich processing steps as achieved by the multi-stage evaporation processes. Here, we report on the influence of the Cu-content of Cu-In-Ga targets on the crystallite size of CIGS absorber layers deposited by a novel pulsed hybrid reactive magnetron sputtering (PHRMS) process [1] and the impact on solar cell performance. We compare depositions from Cu-poor and Cu-rich targets. Se was evaporated simultaneously from a valved cracker source, with pulses of Se flux of 100 ms at 1 s repetition rate. The grain size of the CIGS absorbers layers is larger for higher deposition temperature and for the Cu-rich layers, as confirmed by qualitative (electron microscopy) and quantitative (X-ray diffraction) analysis. Remarkably, the crystallite sizes observed in our work are up to 5-10 times larger than data from the literature. We attribute the large crystallite sizes to the PHRMS process, in agreement with previous results on thermoelectric selenide materials [2]. Furthermore, solar cell devices show a significantly higher efficiency for CIGS absorbers deposited from the Cu-rich target, mainly originating from an improvement in Voc, while Jsc and fill factor are only slightly modified. We also present a two-stage growth process, where after deposition of a thin Cu-rich CIGS layer, a final In-Se layer was deposited, thus resembling the second and third stages in 3-stage co-evaporation. These solar cells exhibit a notable improvement in the best cell efficiency of 6.7 %, compared to 4.4 % for the Cu-rich deposited CIGS with KCN etching. Our results contribute to the further industrialization of CIGS solar cells fabricated by cost-effective sputtering technology and eliminate the need for toxic KCN etching. [1] D. Fuster et al., Solar Energy 198, 490 (2020). [2] J.A. Perez-Taborda et al., Adv. Energy Mater. 8, 1702024 (2017).

Resume : CdTe has become the leading thin-film photovoltaic technology and has overcome several efficiency limitations, especially within the absorber with a CdSeTe layer and at the front interface with an appropriate buffer layer. Simultaneous passivation and Fermi-level matching at the back contact, however, has been more difficult to optimize. The incorporation of a Te layer, which has previously been shown to be effective, is extended here to Se/Te alloys and Te oxides, each of which has shown a positive impact on cell performance.

Resume : The efficiency of a solar cell is often limited by electron-hole recombination mediated by defect states within the band gap of the photovoltaic (PV) semiconductor.1-5 The Shockley-Read-Hall (SRH) model considers a static trap that can successively capture electrons and holes. In reality however, true trap levels vary with both the defect charge state and local structure. Here we consider the role of metastable structural configurations in capturing electrons and holes, taking the tellurium interstitial in CdTe as an illustrative example.6 Consideration of the defect dynamics, and symmetry-breaking, changes the qualitative behaviour and activates new pathways for carrier capture. Our results reveal the potential importance of metastable defect structures in non-radiative recombination, in particular for semiconductors with anharmonic/ioniccovalent bonding, multinary compositions, low crystal symmetries or highly-mobile defects. 1. Y.-T. Huang, S. R. Kavanagh, D. O. Scanlon, A. Walsh and R. L. Z. Hoye, Nanotechnology, 2021, 32, 132004. 2. U. Rau, B. Blank, T. C. M. Müller and T. Kirchartz, Phys. Rev. Applied, 2017, 7, 044016. 3. C. Freysoldt, B. Grabowski, T. Hickel, J. Neugebauer, G. Kresse, A. Janotti and C. G. Van de Walle, Rev. Mod. Phys., 2014, 86, 253–305. 4. S. R. Kavanagh, A. Walsh and D. O. Scanlon, ACS Energy Lett., 2021, 6, 1392–1398. 5. I. Mosquera-Lois and S. R. Kavanagh, Matter, 2021, 4, 2602–2605. 6. S. R. Kavanagh, D. O. Scanlon, A. Walsh, C. Freysoldt - arXiv preprint arXiv:2202.12212, 2022

Resume : Cubic zincblende Cd1 xZnxTe is promising in view of optoelectronic application due to its tunable direct optical band gap in the visible (1.45 – 2.26 eV) as a function of x. As electron-phonon scattering adversely affects optoelectronic properties, it is important, before any practical use, to elucidate the phonon behavior near the Brillouin zone center where photocarriers are generated. Far-infrared (IR) absorption and Raman scattering are well-suited for doing so because they naturally operate near Γ – due to the quasi vertical dispersion of light. The Raman and IR properties of Cd1 xZnxTe have been studied extensively over the past mid-century, mostly at small to moderate Zn content. The known Raman spectra, all taken in resonant conditions favoring the polar LO modes, point towards a basic Modified Random-Element-Isodisplacement (MREI) 1-bond→1-mode behavior, justified by non overlapping of the CdTe (147 – 169 cm 1) and ZnTe (182 – 210 cm 1) phonon optical bands [1]. A problem with the LO modes is that they couple easily via their long-wave electric field due to the ionicity of the chemical bonding in a zincblende crystal, which screens any Raman fine structure. In fact, a recent IR study by Kozyrev [2] into the purely-mechanical (PM) (non-polar, deprived of electric field) transverse optical (TO) modes of Cd1 xZnxTe, which hardly couple, revealed a bimodal Zn-Te pattern, presumably falling into the scope of our so-called percolation model (1-bond→2-mode) [3]. In this work, we present a detailed Cd1 xZnxTe Raman study at large Zn content, nearly unexplored by Raman scattering. First, we focus on the (non-polar) PM-TO’s detected out of resonance conditions in the classical (reflection-like) backscattering geometry, to test directly, from the Raman side, the percolation scenario lately advanced by Kozyrev. The experiment is supported by ab initio phonon calculations (Siesta code) done in the Zn-dilute limit using prototypal percolation-type impurity motifs [3]. Next, we take advantage of the large optical band gap at large Zn content to address the phonon-polaritons (polar TO’s) – remained unexplored so far – by near-forward (transmission-like) Raman scattering. Last, we discuss persisting LO anomalies at small Zn content: that the Zn-impurity mode shows up as strongly as the CdTe-matrix one and that the CdTe-like signal vanishes immediately on departing from the Zn-dilute limit. The entire discussion is supported by contour modeling of the PM-TO, PP and LO Raman lineshapes within the linear dielectric approach [4]. [1] D.N. Talwar et al., Phys. Rev. B 48, 17064 (1993) [2] S.P. Kozyrev, Semiconductors 49, 885 (2015) [3] O. Pagès et al., Phys. Rev. B 86, 045201 (2012) [4] D.T. Hon and W.L. Faust, Appl. Phys. 1, 241 (1973)

Resume : One of the factors determining the formation of the CdTe surface in CdS/CdTe heterojunction and thus influencing the performance of the solar cells is the CdTe deposition technology and the surface conditioning before the preparation of an ohmic contact. Electronic properties of the CdTe thin films deposited by Close-Spaced Sublimation (CSS) have been investigated by Kelvin Probe (KP) and Photoelectron Yield Spectroscopy (PYS) in device configuration. The CdTe thin films have been deposited by CSS onto CdS/TCO (AZO, SnO2)/glass substrates, varying the substrate temperature between 324-453°C and keeping the evaporation temperature at 550°C. The substrate temperature increase reveals a decrease trend of the ionization energy for CdTe thin films deposited on CdS/SnO2 stacks. In addition, the obtained results demonstrate an influence of the substrate temperature for CdTe deposition on the Fermi level position and thus on the charge carrier concentration within the PYS information depth. The Fermi level position changes between 105 meV and 144 meV below the conduction band edge (CBM) for the CdTe prepared on CdS/SnO2 substrates. We conclude therefore that the surface of the CdTe films is of an n-type conductivity, in contrast to that of the p-type CdTe single crystal used as CSS source material The KP and PYS results are correlated with the thin-film composition and structure. A comparison between CSS CdTe thin film and CdTe stoichiometric single crystal grown by Bridgman method is given as well. Additional information is brought considering thermal annealing in presence of CdCl2 as well as bromine surface etching. Acknowledgements: NARD project: 20.80009.5007.12

Resume : In this work, we report on the preparation and characterization of nanostructured cadmium telluride (CdTe) / copper (Cu) nanowires (NWs) interfaces with great potential to be used within "substrate"-type photovoltaic cells based on AII-BVI heterojunctions. In particular, the multi-step preparation protocol involved an electrochemical synthesis procedure within a supported anodic aluminum oxide (AAO) nanoporous template for generating first a homogeneous array of vertically-aligned Cu NWs, to be subsequently embedded within a compact CdTe thin film. In a second stage, we engaged three deposition methods (i.e., vacuum thermal evaporation – VTE, radio-frequency magnetron sputtering – RF-MS and electrochemical deposition – ECD) for obtaining CdTe layers potentially capable of consistently penetrating the previously prepared Cu NWs array. The comparative analysis was performed in terms of critically evaluating the morphological, optical and structural properties of the deposited CdTe films. The presented results demonstrated that under optimized processing conditions, the ECD approach could allow the cost-effective fabrication of absorber layer/collecting electrode CdTe/Cu nanostructured interfaces that could improve the charge collection mechanisms, and hence could allow the fabrication of more efficient solar cells based on AII-BVI semiconducting compounds. Keywords: Supported nanoporous alumina (AAO) templates; Copper (Cu) nanowire arrays; Cadmium Telluride (CdTe) thin films; Electrochemical deposition (ECD); Absorber layer/collecting electrode CdTe/Cu nanostructured interfaces. Acknowledgements: This research was funded by the Executive Unit for Financing Higher Education, Research, Development and Innovation – UEFISCDI (Romania) through the grants: 115/2020 (PN-III-P1-1.1-TE-2019-0868), 25/2020 (PN-III-P1-1.1-TE-2019-0846), 195/2020 (PN-III-P1-1.1-PD-2019-0466), as well as by the Fonds de la Recherche Scientifique – FNRS (Belgium).

Resume : Electronic defect levels in semiconductor devices are often investigated by thermal admittance spectroscopy, a method in which the device capacity is measured as a function of frequency and temperature. When applied to CIGS-based thin film solar cells, these commonly show a characteristic signature which has an activation energy of roughly 0.1 eV and has been termed the “N1” signature. However, even though this feature has been observed for at least two decades, the origin of the N1 signature is still under quite some debate in the community. Explanations include a defect level in the absorber bulk, at the heterointerface, in the buffer layer or at the back contact, as well as mobility freeze-out or an injection barrier at the buffer/i-layer interface. In order to contribute to an answer on this issue, in this study we fabricated several Cu(In,Ga)(S,Se)2 devices in which the individual layers of the cell stack were systematically varied. Between two variations only one layer was different. This approach allowed us to test different hypotheses on the origin of the N1 signature by measuring admittance spectroscopy on the variations and checking whether a change in a specific layer led to a change in the N1 signature. Additionally, we investigated the role of the N1 signature in device performance. To this end, we compared changes in the N1 signature to changes in recombination losses. The latter were quantified on the basis of open-circuit voltages, absorber band-gaps and doping densities, which were derived from current-voltage, quantum efficiency and capacitance-voltage measurements, respectively.

Resume : The prospective Cd1-xBexTe semiconductor mixed crystal is attractive for studies because its band gap can be tuned in a wide range (1.45 – 2.7 eV). The Be-Te bonding exhibits high covalency (as small ionicity as fi=0,22), in contrast with Cd-Te (fi=0.74) [1]. The incorporation of the covalent/stiff Be-based bond into soft/ionic CdTe-based materials is thus expected to reinforce the lattice. A large contrast in Cd-Te/Be-Te bond stiffness presumably exacerbates the vibrational properties, giving a chance to reassess the phonon behavior of semiconductor mixed crystals. As such, Cd1-xBexTe can be considered as a benchmark system to further test our percolation model for vibrations in terms of a 1-bond→2-mode behavior reflecting a sensitivity of bond vibrations to the local environment, beyond the historical MREI (modified-random-element isodisplacement) 1-bond→1-mode one [2]. The insight is gained by Raman scattering, with a support from ab initio phonon calculations (Siesta code). The studied Cd1-xBexTe bulk single crystals (x≤0.7) are grown by the Bridgman method with Be content up to the largest achievable value. Accordingly to X-ray diffraction, all crystals retain the zincblende structure as the parent compounds. The x value is derived by assuming a linear variation of the lattice constant. An insight into the nature of the Be-Cd substitution as to whether it is ideally random or not, is gained by solid state 125Te nuclear magnetic resonance measurements. The optical band gap and the dispersion of the refractive index in the visible (where operates the Raman scattering) – potentially useful to address the phonon-polaritons by forward near-Raman scattering – are measured by spectroscopic ellipsometry. The Cd1-xBexTe Raman spectra show the well-separated alloy-related optical phonons at small Be content (x=0 and x~0.01, ~0.07) in Cd-Te (~ 150 cm 1) and Be-Te (~ 400 cm 1) ranges. The Be-impurity Raman mode (x~0.01) is detected at ~385 cm 1, i.e., near the value observed by Sennet et al. by applying infrared absorption to Be-doped CdTe [3], independently supported by calculations of the impurity mode within the mass defect approximation by the cited authors, and by our own ab initio phonon calculations. As Be content grows the Be-impurity mode develops into a proper transverse-longitudinal optic (TO-LO) band. However, the Be-Te Raman signal remains partly screened by the second-order Cd-Te Raman signal, so that it is not possible, yet, to decide about its unimodal (MREI) or bimodal (percolation) character by experimental means. Hence, a basic MREI description is adopted at this stage, useful to fix ideas. The contour modeling of the TO and LO Raman signals in their x-dependence is achieved accordingly within the linear dielectric approach. [1] N. E. Christensen et al., Phys. Rev. B 36, 1032 (1987). [2] O. Pagès et. al., Phys. Rev. B 86, 045201 (2012). [3] C.T. Sennett et al., J. Phys. C 2, 1137 (1969).

Resume : Cu(In,Ga)Se2 (CIGS) solar cells are one of the top performing thin-film photovoltaic technologies. The respective absorber is a quaternary compound whose complex properties may hinder further performance improvement. Therefore, simulation studies of complete solar cells play a vital role in further unveiling the impact of their electric, electronic, and optical properties. Solar Cell Capacitor Simulator (SCAPS) one dimensional software has dedicated features for the electrical simulations of CIGS and CdTe thin-film solar cells. However, the baseline model was only updated in two major contributions, in 2003 and 2011. Since then, many improvements were implemented within the CIGS technology to boost its performance to the current cell record value, such as the implementation of the post-deposition treatment (PDT) with impact on the CIGS properties. Therefore, this work is presented to the scientific community as an updated baseline model that allows to analyse the properties of state-of-the-art CIGS solar cells, being supported for the first time with Finite-Difference Time-Domain based simulations for accurate optical data and validated by champion cells with the same architecture. The proposed model will show the importance of considering theoretical and experimental CIGS defects with low formation energy, namely VCu, as well as InCu and CuIn antisite defects, together with grain boundary (GB) defects. The novel defect approach allows to implement the CIGS PDT effects on such a SCAPS baseline model, such as GB passivation and increase of the net free carrier concentration. Other interests of the CIGS community have been the fabrication of ultrathin and bifacial solar cells, with manufacturing cost reduction and environmental benefits, as well as the added potential for increased power production. Hence, the developed baseline model for champion solar cells with 3 µm of thickness value is also adapted to high-efficiency ultrathin 490 nm CIGS cells, paving the way for an ultrathin baseline model. Our baseline points toward ultrathin CIGS champion cells with at least 21.0 % of power conversion efficiency after implementing strategies to overcome the inherent limitations of ultrathin devices, with open-circuit voltage values even higher than current thin-film champion cells, 822 vs 734 mV, respectively. Therefore, to achieve such high performance, the SCAPS model suggests the implementation of a high reflecting passivated layer at the rear contact, an optimized anti-reflecting layer at the front contact, and the reduction of the CIGS defect density through PDT. The performance challenges of bifacial CIGS solar cells and the respective impact of rear passivation and absorber thickness will be discussed through the SCAPS model. In this work it is presented to the CIGS community a powerful working tool that can be used to discuss and identify performance limitations in standard and novel CIGS solar cells architectures.

Resume : Engineering the interface of thin film chalcogenide solar cells has become as critical to extending conversion efficiency to >20% as improving properties of the absorber material. For CdTe thin film technology, various buffer and interfacial layers, such as CdS, ZnO, Zn1–xSnxO (ZTO) or MgxZn1-xO, (MZO) have been successfully implemented for the interface engineering, but there is yet significant space for R&D in this direction. In this paper, interface engineering via sputtering of CdO nanolayer at the CdS-buffer/CdTe-absorber interface is demonstrated as an efficient approach to improve the performance of solar cell device. The process steps for fabrication of CdTe thin-film solar cells with superstrate configuration included, formation of a CdS/CdTe stack by close spaced sublimation (CSS) over the TCO/glass substrate, CdCl2 air treatment, cleaning of the CdTe film surface and formation of a back-contact. Herein, two different SnO2/CdS/CdTe/Ni and SnO2/CdS/CdO/CdTe/Ni heterostructures (HJs) prepared by close space sublimation (CSS) method are discussed. These two types differ from each other in that at the interface of one of them prior to CdTe CSS deposition, i-CdO interfacial layer with various thicknesses of 5 nm…30 nm was deposited by dc magnetron sputtering. A comparative study on CdS/CdTe and CdS/CdO/CdTe interfaces have been conducted by current-voltage (J-V), capacitance-voltage (C-V) and admittance spectroscopy measurements. The J-V characteristics of the devices with an area of 0.5 cm2 under 100 mW/cm2 illumination showed that i-CdO layer with thickness of about 6 nm increases both short circuit current density (Jsc) and open circuit voltage (Voc) by 5% and 25%, respectively. Analysis of the J-V and C-V dependencies indicate the presence of a conduction band secondary barrier at the absorber-metal contact interface for both types of HJs. The interface state density and interfacial recombination centers responsible for the collection losses of photogenerated carriers in both devices depend on their concentration states and length of the depletion region. Furthermore, the C-V-T and admittance measurements are correlated with the device performances in order to guide the future improvement.

Resume : The CdTe/CdS heterojunction-based photovoltaic devices proved to be a suitable candidate for silicon technology but further studies are required in order to enhance their performances. One of the issues that should be addressed is related to the transparent conductive electrode (TCE) for the superstrate configuration. By this study, we propose the analysis of the photovoltaic performances of fabricated structures by evaluating the specific parameters such as short-circuit current density, open circuit voltage, fill factor, and maximum output power, for a CdTe/CdS heterojunction and by using three different TCEs; i.e. ITO/ZnO, ITO/ZnO:Al:ZnO, and ZnO:Al:ZnO. In order to complete the photovoltaic architectures, a Cu:Au back electrode was deposited by thermal evaporation. The TCEs, the CdS window layer, and the CdTe active layer were grown by radio-frequency magnetron sputtering. The physical properties of ITO/ZnO, ITO/ZnO:Al:ZnO, and ZnO:Al:ZnO architectures were investigated by optical spectroscopy, X-ray diffraction, atomic force microscopy, and current-voltage characteristics. In order to improve the overall quality of fabricated TCEs thin films, they were subjected to a thermal treatment at 400°C for 30 minutes, in air. Keywords: CdTe/CdS heterojunction, ITO, ZnO, Al:ZnO, photovoltaic structures Acknowledgments: This work was financially supported by The Romanian National Authority for Scientific Research, UEFISCDI, under projects no. 25TE/2020 and project 115TE/2020.

Resume : Development of many-body systems theory is one of the central problem of condensed matter physics. The Feynman diagram technique of summation of perturbation theory series and the method of equations of motion are two main approaches in the study of quasiparticles bound states. A statistical two-particle problem of the excitons bound states, by taking into account the effects of polarization and screening from first principles within the framework of the finite temperature Green functions method, fluctuation-disipation theorem and Heisenberg equations with infinitesimal perturbations is considered. It is demonstrated that integral equations with vortex function account determine the spectrum of bound states by taking into account the strong correlation effects and screening without using the weak-coupling approximation. It has been shown that statistical description defines more accurately many-particle effects, since statistics allows to calculate the experimentally observed physical quantities within the framework of the density matrix method and mixed states with relative accuracy up to 10^-11. It has been demonstrated, that below the Motts transition of excitons to the electron-hole plasma the constancy of the radiation wavelength with increasing excitation level is a consequence of the compensation of the exchange interaction contributions. It was shown that the CdTe PL radiation wavelength related to the exciton recombination weakly depends on the plasma concentration, since the contributions to bandgap and to the exciton binding energy at low momenta are canceled. The band gap and the exciton binding energy temperature dependence is due to the direct Coulomb interaction, which depends on the volume of the unit cell. The radiation frequency at exciton recombination in this case follows bandgap vs temperature dependence. Acknowledgements: NARD (ANCD) projects: 20.80009.5007.12

Resume : The beneficial impact of alkali-fluoride post-deposition treatments (PDT) on Cu(In,Ga)Se2 (CIGSe) thin-film solar cells was already demonstrated by a series of efficiency improvements over the last few years [1,2], and sustained efforts were dedicated to understand their effects on the optoelectronic properties of CIGSe. In this contribution we highlight how advanced scanning probe microscopy methods can contribute to determine a correlation between structural and electronic properties of CIGSe. Kelvin probe force microscopy (KPFM) was used to investigate the local electronic properties of grain boundaries (GBs) for CIGSe absorbers with different alkali-fluoride (KF, RbF and CsF) PDTs. Interest in GB electronic properties is triggered by compositional studies indicating alkali accumulation. A statistical analysis of more than 240 GBs shows a strong difference of the potential variation across the GBs for samples with different alkali fluorides and allowed a direct correlation to the open-circuit voltage of reference solar cell devices. We attribute a beneficial effect on the efficiency to the lower potential barriers and the more homogenous potential variation at the GBs found for the RbF-treated samples. Our results suggest that heavier alkali elements might lead to better passivation by reducing the density of charged defects at GBs [3]. As KPFM probes only the surface and surface-near region, the above results regard only information related to the properties close to the CIGSe surface. Nevertheless, the CIGSe bulk properties are also relevant to understand device performance, and are known to be modified by alkali-fluoride PDTs. Therefore, to assess CIGSe bulk properties we use a newly emerging 3D SPM method. A highly-doped diamond-coated tip with an applied force of several µN indents on the surface, leading to the removal of a thin layer of material with every scan frame. The combination of tip-induced material removal and sensing capability of conductive AFM enables a 3D mapping of the current. Electrical measurements using this conductive-AFM tomography on RbF CIGSe indicate a depth-dependence of the current revealing insights into the conductive paths in relation to grains and GBs. The observed nanoscale conductivity fluctuations are discussed in relation to the solar cell performance. [1] P. Jackson et al. Phys. Status Solidi - Rapid Res. Lett. 10 (2016) 583 [2] Solar Frontier, Efficiency of 23.35%; http://www.solar-frontier.com /eng/news/2019/0117_press.html. (2019) [3] N. Nicoara et al., Nature Commun. 10 (2019) 3980

Resume : The incorporation of Ag into Cu(In,Ga)Se2 absorbers enables a reduced deposition temperature, which benefits the growth of thin film solar cells. Furthermore, Ag alloying leads to a wider optical band gap and improves the open-circuit voltage and thus the photovoltaic conversion efficiency. Contrary to other semiconductor alloys, however, the band gap increase occurs even though the crystal lattice expands with increasing Ag content. Moreover, the Ga-Se bond length of (Ag,Cu)GaSe2 is predicted by theoretical calculations to decrease with increasing Ag content [1]. This prediction is counterintuitive, since in other chalcogenide alloys, for example Cu(In,Ga)Se2, all bond lengths increase as the lattice expands [2]. To unravel this mystery, we studied the atomic-scale structure, especially the element-specific bond lengths, of (Ag,Cu)GaSe2 alloys using extended X-ray absorption fine structure spectroscopy (EXAFS). The thin films were grown on Mo-coated soda lime glass by a single stage co-evaporation process at a substrate temperature of 550°C [3]. The average Ag-Se and Cu-Se bond lengths determined by EXAFS clearly increase with increasing Ag content whereas the Ga-Se bond length decreases despite the expansion of the chalcopyrite lattice. Thus, (Ag,Cu)GaSe2 thin film alloys indeed exhibit the counterintuitive bond length dependence predicted by theoretical calculations. Correlating this peculiar structural behavior with the electronic band structure will contribute to a deeper understanding of (Ag,Cu)GaSe2 alloys and their key material properties. [1] S. Chen et al., Phys. Rev. B 75, 205209 (2007). [2] C. S. Schnohr et al., Phys. Rev. B 85, 245204 (2012). [3] J. H. Boyle et al., J. Appl. Phys. 115, 223504 (2014).

Resume : We apply the admittance loss map technique [1] to thin film CIGS and Kesterite solar cell devices. Bias and frequency-dependent admittance measurements are a very powerful tool to identify different electronic defects in these thin film devices, but also show a vast complexity of data. By graphically representing the data as loss maps and by simulating the loss maps of different defects in these devices using the SCAPS software, we try to get insight into the origin of the defects. We will then extend this technique to low temperature measurements. As usual, low temperature measurements of these loss maps allow us to derive the activation energy of the corresponding defects. We will discuss the different advantages and difficulties/limitations of this analysis method and will discuss similarities and differences observed between CIGS and Kesterite devices fabricated using a two-step process involving metal deposition and selenization/sulfurization in an H2Se/H2S environment. [1] Brammertz et al., IEEE Journal of Photovoltaics 10, Issue: 4, July 2020, Pages 1102 – 1111.

Resume : The efficiency of photovoltaic devices based on chalcopyrite (CIGS) thin films have recently been boosted to >23%. The latest efficiency improvement was achieved by a post-deposition treatment (PDT) with alkali salts. RbF was demonstrated as efficient alkali salt for the PDT of co-evaporated CIGS-based solar cells. The RbF-PDT treatment leads to the formation of different secondary surface phases such as RbInSe2 and In2Se3 in the presence of Na, which moves to the film surface by diffusion from the applied soda-lime glass (SLG) substrates. To clarify the chemical and electronic properties of these secondary phases and the effect of Na, individual RbInSe2 and In2Se3 layers were prepared on Mo/SLG substrates with and without Na barrier and their properties were studied by bulk and surface sensitive methods. The integral bulk composition of the films was investigated by X-ray Fluorescence (XRF) analysis. The surface composition was studied on as-prepared samples by X-ray Photoelectron Spectroscopy (XPS) in ultra-high vacuum. Kelvin Probe (KP) and Photoelectron Yield Spectroscopy (PYS) methods were applied to determine the work function (WF), ionization energy (IE) and the valence band edge (VBM) from measurements in N2 ambience. The XRF measurements confirm the 2:3 and 1:1:2 integral stoichiometry of In2Se3 and RbInSe2 films, respectively. The surface composition however is In-rich for all samples, while the Rb content on the RbInSe2 surface changes from Rb-poor on Mo/barrier/SLG to that of the bulk stoichiometry on Mo/SLG substrates. No sodium is detected on the surface of films deposited on Mo/barrier/SLG, while an amount of 0.4% and 1.7% of Na-containing species is detected on In2Se3 and RbInSe2 layers deposited on Mo/SLG substrates, respectively. That is because Rb, as a heavy alkali element, triggers the diffusion of Na in the absence of an O- and H-containing ambient and its absence explains the lack of Na on the In2Se3 surface. The surface photovoltage data derived from KP measurements indicates an n-type conductivity for all In2Se3 and RbInSe2 films. The WF in the dark varies between 4.59-4.73 eV for In2Se3 and 3.74-4.17 eV for RbInSe2 on no barrier and Na-diffusion barrier containing substrates, respectively. The variations of IE correlate with the changes in surface composition. When Mo/barrier/SLG is used as a substrate, ionization energies of 5.55 eV and 5.31 eV are found for In2.4Se3 and Rb0.5In2.0Se2, respectively, while for layers on Mo/SLG substrate the IE values decrease to 5.43 eV and 5.23 eV for In2.3Se3 and Rb1.0In1.5Se2. The VBM position derived from flat band conditions is determined as 1.00 eV and 1.03 eV below the Fermi level for In2Se3 on Mo/SLG and Mo/barrier/SLG substrates, respectively. The VBM position of RbInSe2 films is strongly influenced by the presence of Na and changes from 1.31 eV for films without Na with a Rb0.5In2.0Se2 surface composition to 1.55 eV for a Na containing surface with a Rb1.0In1.5Se2 composition.

Resume : CdTe and CuInGa(S,Se)2 have been widely regarded as the most successful among the thin film solar cell technology. In recent years, there has been growing interest to investigate a newer class of chalcogenides which consist of less toxic and more abundant elements. In this talk, I will present brief summary of emerging binary (i.e.Sb2S3, Sb2(S,Se)3 etc), ternary (i.e CuSbS2, AgBiS2, etc), and quaternary (Cu2ZnSn(S,Se)4 and its emerging derivatives) chalcogenides, focusing especially on the comparative analysis of their optoelectronic performance metrics, electronic band structure, and point defect characteristics [1]. We found that while most of these emerging materials have small effective mass and high carrier mobility, their performance is limited by non-radiative recombination which leads to Voc deficit. The complicated defect properties in emerging chalcogenides is either due to the anisotropy in binary and ternary quasi-1D and quasi-2D compounds or due to the structural complexity of quaternary compounds. In this talk, I will also present our strategies to suppress the deep defects in Cu2ZnSn(S,Se)4 by cation substitution [2] and to facilitate directional growth of the higher mobility [hk1] planes of Sb2(S,Se)3, resulting in power conversion efficiency of >9% [3]. These advancement in improving the intrinsic properties of emerging chalcogenide thin film absorbers also inspire their application as photocathode to generate clean hydrogen by solar water splitting [4]. [1] S. Hadke, L.H. Wong* et al, Chemical Reviews, 2021, https://doi.org/10.1021/acs.chemrev.1c00301 [2] S.Hadke, L.H. Wong* et al, Advanced Energy Materials, 2019, 9 (45), 1902509. [3] X. Jin, L. H. Wong* et al, Advanced Material, 2021, 33, 2002887. [4] YF Tay, LH Wong* et al, Journal of Materials Chemistry A, 2020, 8 (18), 8862-8867

Resume : Silver bismuth sulfide nanocrystals (AgBiS2 NCs) are extremely promising absorber materials for low-cost, solution-processable and environmentally friendly solar cells. In 2016, a power conversion efficiency Over 6% has been demonstrated in ultrathin (35 nm) AgBiS2 NCs solar cells. However, calculating from the absorption coefficient, high current density (> 25 mA/cm2) can only be achieved with absorber thickness over 200 nm. Yet, due to short diffusion length, incomplete charge extraction was observed in ultrathin devices. The conflict between the need of thick absorber and the short diffusion length severely limits the device performance. Here, we propose that, instead of directly increasing the diffusion length in AgBiS2 NC films, by enhancing the film absorption coefficients, high absorption and near-unity extraction efficiency can be achieved simultaneously with ultrathin (~30 nm) films. We combine theoretical DFT calculations with experimental results to demonstrate that enhanced optical transition moment matrix and absorption coefficients can be obtained by cation disorder homogenization in AgBiS2 NCs. An absorption coefficient that is higher than any other commonly used photovoltaic materials was obtained. Finally, implementing the ultra-absorbing films into devices with careful choose of hole transport layers, a high Jsc of 27 mA/cm2 and a record efficiency up to 9.17% (8.85% certified) were obtained with 30 nm AgBiS2 NCs absorber. The devices also showed superior shelf and photo-stability under ambient conditions. Our work not only establishes the extraordinary potential of solution-processable, RoHS-compliant, ultrathin AgBiS2 NC solar cells, but also demonstrates the importance and power of cation disorder engineering in multinary systems.

Resume : Vanadium doped indium sulfide, In2S3:V, is studied as a potential absorber material for intermediate band solar cells. Based on electronic considerations, it is usually assumed that V occupies octahedrally coordinated In sites, although geometrical considerations would favor tetrahedral In sites. In this study, we have therefore combined experimental X-ray diffraction and X-ray absorption spectroscopy with ab initio theoretical calculations of both α and β phase to elucidate the incorporation of V in In2S3:V thin films grown with different V content and different growth temperatures. Both, experiment and theory, show that the lattice constant decreases with increasing V content. Furthermore, the crystallinity of the material diminishes with increasing V content and decreasing growth temperature. Most importantly, though, (i) comparing the shape and position of the measured and calculated V X-ray absorption edge, (ii) comparing experimentally determined and calculated V-S bond lengths, and (iii) evaluating the calculated heat of solution of V on different lattice sites all indicate that V is incorporated on octahedral rather than tetrahedral lattice sites in the In2S3 matrix. For this material system, the electronic benefit of octahedral coordination thus indeed outweighs the mechanical stress induced by the significant local lattice relaxation associated with incorporation of V on octahedral In sites.

Resume : Atomic layer deposition (ALD) stands unparalleled in the growth of binary compounds but has been relatively less explored in multi-element films due to complications such as, non-ideal growth reactions and film inhomogeneity. Herein is presented the ALD of ternary Cu2SnS3 (CTS) thin films within a reasonably wide temperature window of 150 ‒ 190 oC using a super-cycle growth strategy. The use of rationally designed deposition schemes, which involved matching the diffusion length of cations to the sub-layer thickness in each supercycle resulted in homogenous monoclinic CTS films. The optoelectronic quality of the films was manifested by the presence of a double absorption edge, which is scarcely observed with other deposition techniques. Further composition dependent characterization revealed that tangible excursions from idea stoichiometry had minimal impact on optical properties whereas electrical properties were significantly affected, with holes concentration varying by orders of magnitude. On the other hand, post-deposition heat treatments initially aimed at reducing recombination-active grain boundaries, strongly affected both optical and electrical properties. This was identified to be the result of cation disorder induced during heat treatment, which triggered a progressive phase transformation from monoclinic to cubic CTS. The first order effect of this transformation was a decrease in photo-absorptive ability and the creation of intra-bandgap states leading to electronic disorder. In addition, the heat treatment resulted in notable alterations in hole concentrations. From the perspective of solar cell performance, the results suggested that deviation from stoichiometry and the formation of secondary phases in near stoichiometric CTS films would strongly affect fill factors, while open circuit voltage (Voc) and short circuit current (Jsc) are less affected. Conversely, cation disorder associated with phase transformation during heat treatment would have a more direct impact on Voc and Jsc. Lastly, the photovoltaic viability of the ALD CTS films was demonstrated, with the best cell obtained after heat-treatment yielding a power conversion efficiency of 1.75 %, which although encouraging, represented a compromise between degraded bulk optoelectronic quality and reduced recombination-active grain boundaries.

Resume : By means of photoelectron spectroscopy with in-situ surface preparation and interface formation, it is shown that the Fermi energy at the surface of vacuum cleaved n-type SnS single crystals can be shifted almost through the whole energy gap by evaporation of MoO3. This indicates that there is no intrinsic Fermi level pinning mechanism, which can limit the photovoltage of SnS solar cells.

Resume : Thin films 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. However, finding the optimum composition in MgxZn1-xO and CdSeyTe1-y is complicated as Mg in MgxZn1-xO and Se in CdSeyTe1-y both change the conduction band alignment and interface defect density, which can have significant impacts on interface recombination, barrier heights, and VOC. This problem may be possible to address by ppplying a co-optimization approach to absorber and contact layers in these inorganic chalcogenide thin film solar cells. This presentation will report on a high-throughput approach to co-optimize 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. The main result is 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/cm2) 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 presentation will report on the optimal compositions of the commercially-relevant MgxZn1-xO / CdSeyTe1-y solar cells, and demonstrates a general approach to co-designing performance of alloyed thin film solar cells and other optoelectronic devices.

Resume : CIGS-based solar cells achieve highest efficiencies up to 23.35% [1] with a solution-grown Zn(O,S) buffer layer. Usually, the deposition of the Zn(O,S) layer is performed by a similar route as the traditional CdS buffer layer per reaction of a Zn salt with thiourea in an ammonia solution. However, in contrast to CdS, the deposition of Zn(O,S) evolves with a low deposition rate, which makes the process less attractive for a production line. One of the approaches to accelerate the Zn(O,S) deposition is the use of additives [2, 3, 4] in order to increase the decomposition rate of the thiourea and thus the precipitation rate of Zn(O,S). In this work we tested the impact of the oxidative agent sodium peroxodisulfate Na2S2O8 and the reductive agent hydroxylamine NH2OH on the deposition of Zn(O,S) as buffer in CIGS-based solar cells. The growth rate was increased to a similar level as the CdS and the material consumption could be drastically reduced. The composition of the layers was analyzed by ToFSIMS measurements and the SSO ratio was found to vary between 0.6 to 0.9 depending on the thiourea concentration. For cell preparation, we used in-line deposited CIGS layers with a RbF post-deposition treatment. Conversion efficiencies up to 19% comparable to reference cells with CdS could be demonstrated. So far, the reaction mechanism of the oxidative or reductive additives is not well understood. For instance, it is a contradiction to explain a faster decomposition of thiourea by the use of a peroxide as additive as it is known that this reaction leads to the formation of a thiourea dioxide and not to a thiourea decomposition under sulfide release. In this paper, we propose a reaction mechanism for both, the hydroxylamine as well as the peroxide route. We show that the increased sulfide release in both routes is a consequence of N-hydroxyguanidine intermediates formation. Literature 1) M. Nakamura, K. Yamaguchi, Y. Kimoto, Y. Yasaki, T. Kato, and H. Sugimoto, Cd-free Cu(In,Ga)(Se,S)2 thin-film solar cell with record efficiency of 23.35%, IEEE Journal of Photovoltaics 9, 1863-1867 (2019). 2) M. Buffière, S. Harel, L. Arzel, C. Deudon, N. Barreau, J. Kessler, Fast chemical bath deposition of Zn(O,S) buffer layers for Cu(In,Ga)Se2 solar cells, Thin Solid Films 519, 7575-7578 (2011). 3) T. Hildebrandt, N. Loones, M. Bouttemy, J. Vigneron, A. Etcheberry, D. Lincot, and N. Naghavi, Toward a better understanding of the use of additives in Zn(S,O) deposition bath for high-efficiency Cu(In,Ga)Se2-based solar cells, IEEE Journal of Photovoltaics, 5, 1821-1826 (2015). 4) T. Hildebrandt, M. Kozolinsky, N. Loones , M. Bouttemy, J. Vigneron, A. Etcheberry, D. Lincot, F. Donsanti , and N. Naghavi, Fast chemical bath deposition process at room temperature of ZnS-based materials for buffer application in high-efficiency Cu(In,Ga)Se2-based solar cells, IEEE Journal of Photovoltaics 8, 1862-1867 (2018).

Resume : Cu(In,Ga)Se2 (CIGS) have been considered as promising candidate materials for wide bandgap thin film solar cells, because they have an extremely high absorption coefficient with a bandgap that can be tuned in the range 1-1.7 eV by varying the material composition. However, it is commonly recognized that the use of a high-Ga content in CIGS may increase both the bandgap and the number of defects as well as induce a buffer/CIGS conduction band offset (CBO) issue and leads to a degradation in cell performances. To address these issues, in this study the bandgaps of CIGS absorbers were varied from 1.18-1.62 eV by changing the material composition from the Ga/(Ga+In) (GGI) ratio of 0.38 to 0.92 while adjusting the Cu/(Ga+In) (CGI) ratio around 0.78-0.92 by reducing the Indium content in the absorber during the 3-stage process. The electronic and compositional properties of the films were studied using different methods such as XRF, GD-OES, XRD and PL to better understand the limitations on device performance. GD-OES profiles showed a progressive loss of the Ga-grading that degrades its well-known positive effect in the carrier collection while PL spectra confirmed the increase of the bandgap energies. Moreover, the standard CdS buffer layer not only significantly limits the current output from the cell but also does not present a good bandgap alignment with wide bandgap CIGS absorber layer leading to a CBO at the interface. The impact of two wide bandgap Cd-free buffer layers: CBD-Zn(O,S) and ALD-ZnSnO as a function of GGI content in CIGS absorbers were studied and compared. Due to their higher bandgap, these two buffer layers could form a spike that is a positive CBO with CIGS absorber layer leading theoretically to a better efficiency compared to CdS. However, without any post-treatments of the absorber we observed similar efficiencies whatever the buffer layer used. In all the cell structures, the short circuit current (Jsc) decreases monotonously with the gap mainly due to the reduction of the absorption range as confirmed by the quantum efficiency (EQE) spectra and probably due to the reduction of the collected photogenerated carriers, as indicated by the decrease of the EQE at longer wavelength. This might also be due to a reduction of the Ga-grading that otherwise helps the carrier collection. The increase of the open-circuit voltage (Voc), expected with the increase of the band gap, is monotonous for the cells with Zn(O,S) but not for CdS buffer*. We achieved a higher Voc with Zn(O,S). To improve the performances, the impact of deposition conditions and post-treatment of CIGS absorbers, the thickness of buffer layer and their composition were implemented and evaluated using dark and light IV and CV analyses. This work will serve as a guideline to improve the performances of wide bandgap CIGS absorbers and to match the best buffer layer to the best CIGS composition. *The same absorbers are being completed in cells with ZnSnO buffer layer.