Symposium EN02—Thin-Film Compound Semiconductor Photovoltaics

Copper and arsenic have been used as primary dopants in CdSeTe photovoltaic absorbers. However, both atomic species suffer from very low dopant activation. Arsenic concentration in these devices are on the order of 10¹⁶ cm⁻³, which makes it notoriously difficult to correlate nanoscale distributions to the local charge transport properties. To measure and correlate these properties, measurement techniques require high sensitivity to elemental concentration, large penetration depth, and operando compatibility. In this work, we use nanoscale X-ray microscopy to correlate chemical distribution – structure – electrical properties through X-ray Fluorescence, X-ray Absorption Near edge spectroscopy, and X-ray Beam Induced Current at a pixel-to-pixel level (50-120nm).
Our cross-section studies unveil the segregation of As and the transition of the maximum charge-collection efficiency from the back interface to the front interface as a function of processing conditions, which is correlated to the balance of acceptors and donors, and thus the activation of the dopants. We will show linear combination fitting of XANES and the presence of active species as well as inactive species through the absorber.

Arsenic doping is a significant topic of interest in CdTe solar cells. Carrier concentrations have far surpassed those achieved with copper, and stability has also improved. However, challenges remain in reducing the V_oc deficit in devices. Arsenic is known to self-compensate in CdTe, which can lead to low dopant activation levels and associated voltage fluctuations that contribute to losses. Additionally, SIMS has shown significant As accumulation near the front interface, likely leading to reduced activation levels in the Se-rich region. Our modeling has demonstrated that deep levels near the interface can drastically reduce device performance. Furthermore it suggested that reduced doping near the interface can improve V_oc and thus a quasi- p-i-n structure was proposed where the intrinsic layer would be only a few hundred nm thick in the Se-rich region. In practice, an intrinsic layer is difficult to achieve when dopant diffusion is performed afterwards. However, reduction of the As accumulation would be ideal. In this work, we explore methods of accomplishing this through anneals on evaporated Cd(Se,Te) prior to vapor transport deposited (VTD) CdTe:As.
Our absorber preparation involves the evaporation of an undoped Cd(Se,Te) layer, followed by a thicker VTD layer of CdTe:As. Arsenic is introduced in situ and diffuses throughout the absorber. Since As readily diffuses along grain boundaries, we believe the small Cd(Se,Te) grains allow significant As accumulation in the evaporated layer. While CdCl₂ is known to increase grain size, its formation of a eutectic phase also can massively increase As diffusion rates. We have demonstrated that anneals over CdTe powder help remove voids and create more compact VTD films. We hypothesized that performing similar anneals on the undoped layer may reduce two potential pathways for As diffusion. First, it can reduce the prevalence of grain boundaries and inhibit excessive As diffusion near the interface. Second, Se vacancies, which can contribute to diffusion and recombination, may be filled in. If these are able to sufficiently reduce As pileup, it should yield a higher V_oc due to less near-interface compensation.
Anneals were performed with various temperatures and ambients, as well as with both CdTe and Cd(Se,Te) powder, the latter as an effort to inhibit possible Se loss in the film. The effects of hydrogen and oxygen incorporation were explored, with hydrogen expected to reduce undesired oxidation. Following these anneals, films were CdCl₂ annealed to obtain measurable lifetime via time resolved photoluminescence (TRPL). The highest lifetime was observed with CdTe powder in a nitrogen-only ambient at 575°C. Microscopy on annealed films showed substantive grain growth prior to CdCl₂ treatments, with higher temperatures yielding larger grains.
In addition to material studies of the undoped evaporated material, equivalent films were processed into devices, including VTD CdTe:As and subsequent densification and CdCl₂ anneals. J-V measurements showed slightly increased V_oc for devices with annealed Cd(Se,Te) layers compared to those without the anneal. SIMS measurements showed that As content decreased approximately one order of magnitude within ~200 nm of the front interface, a 2-3 order of magnitude decrease compared to films without the pre-VTD anneal.
Combining material characterization with device performance, we see improved V_oc correlated with lower near-interface As concentration and larger Cd(Se,Te) grains. The pre-VTD anneals therefore had the desired result. Ongoing work involves strategies to remove unintentional oxidation during this anneal, as well as further reducing the As content near the interface to increase the observed effect on V_oc. Hydrogen anneals and thicker undoped layers, respectively, are being explored for these purposes. Through this work, we hope to provide a method to reduce near-interface compensation in As-doped CdTe devices.

The drive for 25% CdSeTe solar cells depends on effective acceptor doping above 1x10¹⁶ cm^-3 which can be achieved using group V doping. However, the incorporation of a Group V dopant onto the Group VI site is highly compensated with associated deep level defects. Until recently this has resulted in no discernable improvement in V_oc compared with much lower acceptor concentrations achieved with Cu doping. The increase in acceptor concentration leads to a reduction of the depletion width from over 1 μm to 0.2 μm. Recombination within the depletion region leads to a reduction in V_oc. It has become clear that high acceptor concentration CdTe and CdSeTe solar cells are less tolerant towards defects close to the front emitter. In this talk the Group V doping of CdTe will be reviewed and recent results presented on the influence of metal saturated conditions on doping with As. Using metal organic chemical vapour deposition (MOCVD) it has been possible to create a wide range of different thermodynamic conditions. In this study we have explored metal saturated growth that has led to an improvement in V_oc close to 900 mV. The role of the front interface will be considered as this becomes more influential in achieving high V_oc for high acceptor concentrations. We have used a combinatorial approach to track V_oc with the Group II and Group VI precursor ratio and acceptor concentration. Approaches to unlock the true potential of Group V doped CdSeTe solar cells using doped emitters with V_oc>1 V will be discussed.

Deep level transient spectroscopy (DLTS) is capable of resolving electrically active defect levels within semiconductors. The technique works by temperature dependant analyis of capacitance transients which occur as trap states within the band gap are depopulated. DLTS allows the energetic poisition of defect levels within the bandgap to be determine, alongside estimates for the capture cross section and density of such defect level. It evolved from analysis single crystal materials such as silicon and gallium arseninde but is also able to resolve defect levels present within polycrystalline thin films. For thin film solar cells this can provide useful information on native and extrinsic defect and/or doping levels aiding the development of new processes. This presentation will demonstrate the capability of the techqiue to determine the defect content inorganic thin film materials such as in emerging antimony chalcogenide materials Sb₂Se₃ and Sb₂S_3, as well as the impact of alloying and group V doping in CdTe via analysis of CdSe_1-xTe_x and CdTe:As solar cells. We will also demonstrate the links between device performance and deep level defect content and show how this measurment approach can be used to inform device process development.

CdTe-based PV devices are currently the only commercially viable alternative to Si. Significant mysteries remain as to the effects of grain boundaries on device performance, and the local environment of dopants and how it relates to activation. In this talk, we will discuss characterization-informed first principles and atomistic modeling of grain boundaries and impurities in CdTe and related systems. The use of x-ray absorption near edge spectra (XANES) and scanning transmission electron microscopy (STEM), in particular, gives unprecedented atomic-level information and the interpretation of XANES and STEM data benefits from AI/ML and modeling.

References:
1. S. Rojsatien, A. K. Mannodi Kanakkithodi, T. Walker, T. Nietzold, E. Colegrove, B. Lai, Z. Cai, M. Holt, M. K. Y. Chan, Mariana I. Bertoni, Quantitative analysis of Cu XANES spectra using linear combination fitting of binary mixtures simulated by FEFF9, Radiation Physics and Chemistry. DOI:10.1016/j.radphyschem.2022.110548.
2. M. Polak, R. Jacobs, A. K. Mannodi Kanakkithodi, M. K. Y. Chan, D. Morgan, Machine Learning for Impurity Charge-State Transition Levels in Semiconductors from Elemental Properties Using Multi-Fidelity Datasets, Journal of Chemical Physics 156, 114110 (2022). DOI:10.1063/5.0083877.
3. A. Mannodi Kanakkithodi, X. Xiang, L. Jacoby, R. Biegaj, S. T. Dunham, D. R. Gamelin, M. K. Y. Chan, Universal Machine Learning Framework for Impurity Level Prediction in Group IV, III-V and II-VI Semiconductors, Patterns 3(3), 100450 (2022). DOI:10.1016/j.patter.2022.100450.
4. F. G. Sen, A. K. Mannodi-Kannakithodi, T. Paulauskas, J. Guo, L. Wang, A. Rockett, M. J. Kim, R. F. Klie, M. K. Y. Chan, Computational Design of Passivants for CdTe Grain Boundaries, Solar Energy and Solar Cells 232, 111279 (2021). DOI:10.1016/j.solmat.2021.111279.
5. A. Mannodi-Kanakkithodi, M. Toriyama, F. G. Sen, M. J. Davis, R. F. Klie, and M. K.Y. Chan, Machine-learned impurity level prediction for semiconductors: the example of Cd-based chalcogenides, npj Computational Materials 6, 39 (2020). DOI:10.1038/s41524-020-0296-7.
6. J. Guo, A. Mannodi-Kanakkithodi, F. G. Sen, E. Schwenker, E. Barnard, A. Munshi, W. Sampath, M. K. Y. Chan, R. F. Klie, Effect of selenium and chlorine co-passivation in polycrystalline CdSeTe devices, Applied Physics Letters 115, 153901 (2019). DOI:10.1063/1.5123169.

There is progressive interest toward fundamentally understanding the roles attributed to passivated back contact layers in thin-film semiconductor PV. Within the cadmium telluride PV community, special emphasis is given to a back passivation layer capable of reducing back interface recombination losses while ensuring carrier selectivity and the possibility for bifaciality. Transparent thin-film oxide layers such as NiO_x or TeO_x may offer a viable solution for back passivation layers in CdTe PV. However, atomic scale effects for NiO_x and TeO_x in relation to possible defect interactions during copper and/or chloride-based treatments as well as energy band alignment of the absorber/passivating oxide interfaces are not well understood. The current work uses a first-principles computational approach based on density functional theory to investigate the electronic structures of NiO and TeO₂ bulk layers in the presence of relevant defects along with the formation of pristine interfacial morphologies when either oxide layer is in contact with the absorber layer in CdTe PV. For the bulk defect studies, defect pair interactions are evaluated to determine avenues of dopability and/or compensation that may affect bulk carrier transport within the oxide thin-film layers. Additionally, an atomic-scale perspective of the pristine absorber/oxide interface structures is given with respect to plane orientation, termination layer, and associated surface reconstruction to give insight on how NiO and TeO₂ differ in band alignment features influencing charge transport at the back interface. Details from the first-principles-based computational study provide further insight on the mechanisms inherent to passivating oxide layers that could potentially lead to benefits in CdTe thin-film PV device performance.

Low-dimensional trigonal selenium (Se) was the first material to exhibit the photovoltaic effect and has recently experienced a renaissance in research interest as a candidate solar cell absorber, due to its desirable properties (suitable band gap, high earth abundance, low temperature processing), potential implementation in silicon tandem cells and ‘simple’ elemental chemistry. Though cell efficiencies have improved much since its initial application in 1883, Se still lags behind leading technologies such as silicon and lead-halide perovskites, with the fundamental limitations (doping density, defect-mediated recombination, intrinsic Urbach tailing…) remaining unclear.

In this work, we use hybrid density functional theory to investigate the electronic properties, band alignment and defect chemistry of Se, in order to shed light on the key limiting factors for solar cells. We find the crystal dimensionality to dictate the energies of intrinsic defects (vacancies and interstitials) in this system and their impact on electronic properties. In doing so, we identify key deleterious impurities for performance and important considerations for the experimental fabrication of high-efficiency selenium solar cells, as well as establishing an outlook for the attainable performance of Se solar cells based off the fundamental intrinsic limitations.

References:
1 M. Zhu, G. Niu and J. Tang, J. Mater. Chem. C, 2019, 7, 2199–2206.
2 D. M. Chapin, C. S. Fuller and G. L. Pearson, J. Appl. Phys., 1954, 25, 676–677.
3 A. Goetzberger, J. Luther and G. Willeke, Sol. Energy Mater., 2002, 11.
4 T. K. Todorov, S. Singh, D. M. Bishop, O. Gunawan, Y. S. Lee, T. S. Gershon, K. W. Brew, P. D. Antunez and R. Haight, Nat. Commun., 2017, 8, 682.
5 T. H. Youngman, R. Nielsen, A. Crovetto, B. Seger, O. Hansen, I. Chorkendorff and P. C. K. Vesborg, Sol. RRL, 2021, 5, 2100111.
6 S. R. Kavanagh, A. E. Samli, A. Walsh and D. O. Scanlon, Submitted

By coupling together a junction-based photovoltaic system and a bulk photovoltaic effect-based material in a nanocomposite thin film device, the well-known limitations^1–3 of both technologies could be overcome. This idea relies on the proven ability of ferroelectrics to influence coupled materials, such as photocatalysts and organic photovoltaics. By using the recently-developed Electronic Lattice Strain procedure,⁴ we identified the BaTiO₃/hematite interface as a promising candidate for the proposed device. Screening was performed in terms of epitaxially-compatible interfaces (to minimize defects) and appropriate band alignment (to minimize charge transfer). To gain insights into the geometry, polarization and electronic properties of the aforementioned system, Density Functional Theory (DFT) modelling is employed. To this purpose, we assume the BaTiO₃ (110) surface as substrate on which the hematite (100) surface grows epitaxially strained to it. Within the supercell approach, we are required to employ 5×5 surface unit cells of BaTiO₃ (110) and 4×2 units of hematite (100) to reproduce the minimally strained interface (~2.6%).

Our preliminary tests show that GGA-PBE functional with a Hubbard-type correction to the Ti 3d orbitals, where U_eff = 2.6 eV, can reproduce the spontaneous polarization in BaTiO_3. However, the bandgap is still poorly described: ~30% off with respect to experimental value. In the case of hematite, where U_eff = 3.5 eV is applied to the Fe 3d orbitals, the bandgap is in excellent agreement with experimental reports. Even though bandgaps can be accurately described by DFT via hybrid functionals, such as HSE06, their application for the above heterojunction is impractical due to the size of the interface model which consists of 855 atoms.

In order to estimate a proper offset from the BaTiO₃/hematite heterojunction, we will employ bulk HSE06 values to correct the band alignments of this interface, which is computed with the PBE+U method. Our scheme is tested with heterojunctions that can be tackled with the HSE06 hybrid functional, ranging from covalent to ionic solids. This scheme might allow us to model not only this complex epitaxially-compatibles interface, but also other heterojunction for novel devices.

REFERENCES
1 A. Zenkevich, Y. Matveyev, K. Maksimova, R. Gaynutdinov, A. Tolstikhina and V. Fridkin, Phys. Rev. B - Condens. Matter Mater. Phys., 2014, 90, 161409.
2 K. T. Butler, J. M. Frost and A. Walsh, Energy Environ. Sci., 2015, 8, 838–848.
3 J. E. Spanier, V. M. Fridkin, A. M. Rappe, A. R. Akbashev, A. Polemi, Y. Qi, Z. Gu, S. M. Young, C. J. Hawley, D. Imbrenda, G. Xiao, A. L. Bennett-Jackson and C. L. Johnson, Nat. Photonics, 2016, 10, 611–616.
4 K. T. Butler, Y. Kumagai, F. Oba and A. Walsh, J. Mater. Chem. C, 2016, 4, 1149–1158.

Exploiting physical and electrical properties of polar materials is currently an area of substantial research interest, with the scope to create multifunctional materials that aim to solve the challenges of the next decades in the fields of renewable energy, transportation, and quantum computing to name a few. Among these, ferroelectric materials whose spontaneous polarisation can be switched by an applied electric field, are known to generate currents without the need of a heterojunction as required for traditional semiconductors. The phenomenon is known as the bulk photovoltaic effect (BPVE), and it results in large photovoltages well above the bandgap which have the potential to surpass the Shockley-Quessier (S-Q) limit for single junction devices^1,2. However, the high photovoltage is limited by the conductivity which combined with the intrinsic wide band gap results in unremarkable efficiencies below 1%^3,4. Different approaches have been adopted to overcome this limitation like ionic doping, heterostructures or localized surface plasmons resonances (LSPR)⁵. Here we report the use of a two-phase nanocomposite approach coupling a ferroelectric with a narrow band gap material that allows us to exploit the built-in voltage of the former with the high absorption and conductivity of the latter. Epitaxial nanocomposites thin films of BaTiO₃:Sm₂O₃ and BaTiO₃:MgO were fabricated via pulsed laser deposition (PLD) and the secondary phase was etched away and replaced with α-Fe₂O₃ phase. The combination of atomic force microscopy allowed us to assess the electrical properties of the films, the presence of BPVE and how these are affected by strain in vertically aligned nanocomposites (VANs).
References
1. Lopez-Varo, P. et al. Physical aspects of ferroelectric semiconductors for photovoltaic solar energy conversion. Phys Rep 653, 1–40 (2016).
2. Spanier, J. E. et al. Power conversion efficiency exceeding the Shockley-Queisser limit in a ferroelectric insulator. Nature Photonics vol. 10 (2016).
3. Wu, L., Podpirka, A., Spanier, J. E. & Davies, P. K. Ferroelectric, Optical, and Photovoltaic Properties of Morphotropic Phase Boundary Compositions in the PbTiO3-BiFeO3-Bi(Ni1/2Ti1/2)O3 System. Chemistry of Materials 31, 4184–4194 (2019).
4. Li, C. et al. Enhanced photovoltaic response of lead-free ferroelectric solar cells based on (K,Bi)(Nb,Yb)O3 films. Physical Chemistry Chemical Physics 22, 3691–3701 (2020).
5. Han, X., Ji, Y. & Yang, Y. Ferroelectric Photovoltaic Materials and Devices. Advanced Functional Materials vol. 32 Preprint at https://doi.org/10.1002/adfm.202109625 (2022).

We will present recent progress of photovoltaic devices using chalcogen-based compounds: kesterites and chalcopyrites. Kesterite Cu₂ZnSn(S,Se)₄ (CZTSSe), promising earth-abundant materials for thin film solar cells, has been widely studied and active research efforts raise the efficiency of flexible kesterite solar cells close to 12%. To fabricate the robust and highly efficient flexible thin film solar cells, understanding the changes in device properties under mechanical stress is crucial. In this study, transport of photogenerated carriers with mechanical stress was characterized. CZTSSe samples with and without Na deposited on flexible Mo foil were prepared and solar cell parameters after the mechanical bending were measured. Degradation of open-circuit voltage (V_OC) was varied by the direction of mechanical bending and the existence of Na elements. Photo-assisted Kelvin probe force microscopy was utilized to characterize the transport of photogenerated carriers, and it was found that the magnitude of surface photovoltage changed under mechanical bending state. It refers the degradation of local V_OC and transport of the photogenerated carriers was limited with mechanical stress resulting in decrease of V_OC of devices. In addition, CZTSSe sample with Na doping showed the larger SPV under bending than sample without Na which suggest the passivation of defects and prevention of damage from bending. To quantify the disorder in energy band structure under mechanical bending, Urbach energy was obtained from absorption coefficient function. Attribution of Na in transport of photogenerated carriers will be investigated to clarify the robust carrier transport under mechanical stress compared to the CZTSSe without Na.

The well-known Shockley-Queisser limit for single junction solar cells can be overcome using multiple solar cells with different band gaps on top of each other. In the case of a tandem solar cell, the ideal band gap of the bottom cell is about 1 eV, of the top cell about 1.7 eV and efficiencies of 42% can be achieved theoretically.[1] As bottom cell often Si is used, which has a band gap of 1.12 eV. For a full thin film tandem structure, which could be suitable for flexible or light weight applications, Cu(In,Ga)Se₂ (CIGS) can also be used. The band gap of CIGS varies between 1.0 and 1.6 eV depending on the Ga/In ratio. Optimal performance of CIGS solar cells is achieved with about 30% Ga and using Ga gradients at the front and the back. The band gap is then around 1.2 eV, thus above the ideal value for a bottom cell. Without Ga in the absorber layer the band gap will be reduced to 1 eV, but also the quality of the absorber layer and its interfaces are less optimal without the Ga gradients. When low band gap CIGS (low or without Ga) will be used as bottom cell, other methods to increase the absorber quality and its interface with the buffer layer need to be investigated. For this purpose, we explore incorporation of small amount of Ag in the absorber layer (ACIS) to improve the bulk properties, and stoichiometric growth in combination with InSe on the surface to improve the buffer/absorber interface.
CIS layers were grown by a 3-stage process on Mo with or without a 10 nm Ag layer. When Ag is added in the CIS absorber, the efficiency of the ACIS solar cell increases about 2% absolute, mainly due to an increased fill factor. Depositing InSe on top of a stoichiometric CIS layer, increases the Voc by 50-60 mV. The efficiency increases about 1% absolute compared to the copper poor CIS. When depositing InSe on a stoichiometric ACIS absorber layer, the solar cell performance deteriorates. The cell is completely shunted, and Voc, FF and Jsc are reduced.
The various solar cells are then further analysed using bias-dependent admittance spectroscopy. The data of these measurements is presented in a so-called CVf-map, and with SCAPS modelling the CVf-maps may be analysed.[2, 3] At first the difference between ACIS and CIS is analysed. Minor differences in the CVf-maps are observed between ACIS and CIS. The CVF-map of ACIS looks slightly cleaner in a larger bias and frequency range, but in general they are very similar, indicating that there are no large differences in the absorber or at the interfaces. The ‘cleaner’ CVf-map for ACIS, corroborates with the higher FF measured for ACIS. When InSe is deposited on a stoichiometric CIS sample, the apparent doping is much higher, which could explain the increase in Voc measured for this kind of structure. When InSe is deposited on a stoichiometric ACIS sample, the apparent doping still increases but a defect signature is seen at 0V bias voltage which seems to be responsible for the deterioration of the performance.
Temperature dependent CVf measurements are further performed to allow for disentanglement of the various features that are seen in the CVf maps. Using SCAPS modelling, CVf maps and low-temperature data, a model is proposed on why the ACIS absorber layer has such a different response when InSe is deposited and how this can be improved.

[1] De Vos, 1980 J. Phys. D: Appl. Phys. 13 839
[2] G. Brammertz et. al . 2020, IEEE, Journal of Photovoltaics, vol. 10, 1102-1111
[3] M. Burgelman et al, ‘SCAPS: a Solar Cell Capacitance Simulator’, https://scaps.elis.ugent.be/, Accessed: October 2022.

Monolithic integration, the process by which solid-state devices are made by sequentially depositing layers of materials on top of each other, is used in all commercial thin film-based technology. This process is so foundational nowadays that it is difficult to imagine any other way to create high-efficiency solid-state devices. In the context of photovoltaics (PV), monolithically-integrated multi-junction (MJ) cells have demonstrated their capability in overcoming the thermodynamic efficiency limits of single-junction (SJ) devices, as well as being a practical solution to create devices with sufficient photovoltage for direct solar-to-hydrogen (STH) conversion.Despite its wide acceptance, however, monolithic integration presents major limitations in process compatibility, restricting materials selection and synthesis to a sub-set of compatible systems while limiting the adoption of emerging promising candidates.
The strategy we present in this communication, further referred to as semi-monolithic integration, relies on the exfoliation and bonding at room temperature of independently processed, substrate-grown SJ devices to create a whole thin film-based MJ device. By design, this scheme allows successive integration of thermally, chemically and/or mechanically incompatible material classes onto a single substrate without compromising state-of-the art SJ fabrication processes. First, we report on whole chalcopyrite-based semi-monolithic MJ devices comprised of 1.85 eV ordered vacancy compound (OVC) CuGa₃Se₅ and 1.13 eV Cu(In,Ga)Se₂ sub-cells. Our process takes advantage of the weak out of plane van-der Waals forces in the MoSe₂ layer which naturally forms at the Cu(In,Ga)Se₂/Mo interface. Both chalcopyrite cells were successively transferred and bonded onto a single rigid FTO host substrate at room temperature using transparent conductive composites (TCC). A champion CuGa₃Se₅/CIGSe MJ device showed power conversion efficiency of 5.04% with open-circuit voltage (Voc), short-circuit current density and fill factor of 1.24 V, 7.19 mA/cm², and 56.7%, respectively. Following this successful proof-of-concept demonstration, we constructed the world’s first whole-chalcopyrite triple junction MJ device comprising 1.13 eV and 1.44 eV Cu(In,Ga)Se₂ and 1.85 eV CuGa₃Se₅ sub-cells. The device exhibited a V_OC of 1.85 V and was capable to splitting water with an STH efficiency of 3% in a PV-electrolysis configuration.

Tandem solar cells (TSCs) are one approach to exceeding the efficiency limits of single-junction solar cells by making better use of the solar spectrum with multiple junctions. Among various types of perovskite-based TSCs. All-perovskite TSCs are of particularly attractive for building- and vehicle-integrated photovoltaics, or space energy areas as they can be fabricated on flexible and lightweight substrates with a very high power-to-weight ratio. However, the efficiency of flexible all-perovskite tandems is lagging far behind their rigid counterparts primarily due to the challenges in developing efficient wide-bandgap (WBG) perovskite solar cells (PSCs) on the flexible substrates as well as their low open-circuit voltage (V_OC). Firstly, the relatively rougher surface of the flexible substrate poses a certain difficulty in the deposition of a uniform layers through solution processing. Moreover, the WBG PSCs suffer from a larger V_OC-deficit (E_g-V_OC) when compared with narrow-bandgap (NBG) PSCs, which has been constantly reported to be >550 mV. To construct a high-performance flexible all-perovskite TSC, there is an urgent need to develop high-quality WBG PSC based on flexible substrates with a reduced V_OC-deficit.
Here, we first developed a modified spin-coating protocol of [2-(9H-carbazol-9-yl)ethyl] phosphonic acid (2PACz) on flexible ITO substrate. Long resting time together with multiple spin-coating ensure the deposition of a uniform 2PACz layer, which serves as a lossless hole selective contact and allows the subsequent uniform growth of a 1.77 eV WBG perovskite. This translates to ~40 mV gain in V_OC of the devices compared to those based on poly-triarylamine. To further reduce the V_OC-deficit, we employed a post-deposition treatment (PDT) on the perovskite surface with 2-thiopheneethylammonium chloride (TEACl), which suppresses both the bulk and interfacial recombination, boosting the V_OC of the flexible NIR-transparent WBG PSC by ~100 mV. We identified an accumulation of TEA⁺ on the surface of perovskite, which widens the surface band structure and reduces the interfacial recombination by realizing a better band alignment between perovskite and electron transport layer. In addition, we found that Cl^- tends to diffuse more into the bulk perovskite, which could be helpful for reducing the non-radiative in the bulk. Based on this, we achieved a high V_OC of 1.29 V and 15.1%-efficient NIR-transparent WBG (1.77 eV) PSCs grown on flexible substrates. The high V_OC corresponds to a record low V_OC-deficit of 480 mV for perovskite with a bandgap around 1.80 eV. In conjunction with flexible NBG (1.24 eV) PSCs, a 23.8% monolithic flexible all-perovskite TSC (0.09 cm²) with a superior V_OC of 2.1 V is achieved, which is on par with the V_OC reported on the 28% all-perovskite tandems grown on the rigid substrate. Moreover, a 23.6% efficiency was also realized on a larger area device (1.04 cm²), which is close to that of the small device and shows a great upscaling potential of the flexible monolithic TSC.
In summary, our work provides a holistic and effective approach to improve the performance (V_OC in particular) of flexible NIR-transparent WBG PSC and all-perovskite monolithic TSC. The understanding of the passivation mechanism of TEACl PDT sheds new light on the future work in targeted choice of passivation reagent for the perovskite solar cells. Further reducing the V_OC-deficit within the WBG and NBG subcells will be the key for reaching a 30% efficiency in flexible monolithic all-perovskite TSCs.

Ordered-vacancy wide band gap CuGa₃Se₅ is a promising candidate as absorber material for the top cell in tandem solar cells as well as in photoelectrochemical water-splitting devices. For wide band gap absorber materials, it is increasingly important to also use a wide band gap buffer layer to avoid absorption losses. Instead of the traditionally used CdS buffer layer, a conductive oxide Mg_xZn_1-xO (MZO) was thus chosen in this study, deposited by RF sputtering on the absorber. To further improve the interface properties, a Cd²⁺ partial electrolyte treatment (CdPE) can be applied to the CuGa₃Se₅ surface prior to MZO layer deposition. Further optimization of the interface requires a more detailed understanding of the chemical and electronic properties at the MZO/absorber interface, such as the band edge positions, band gaps at the interface and, in particular, the band alignment.

In this contribution, we investigate the CuGa₃Se₅ absorber after CdPE, as well as its interface with the MZO layer using a toolchest of electron spectroscopies, including laboratory-based x-ray and UV photoelectron spectroscopy (XPS and UPS), x-ray-excited Auger electron spectroscopy (XAES), and inverse photoemission spectroscopy (IPES). With this combination of techniques, a detailed electronic and chemical picture of the interface is painted and will be discussed in view of its impact on the performance of CuGa₃Se₅ solar devices.

Photovoltaics (PV) based on crystalline silicon wafers has developed into a mature technology for the conversion of sunlight into electrical energy. Thin film PV is an emerging alternative technology because of short energy payback time and minimum use of high purity materials, addressing the urgent need for cost-competitive renewable energy technologies.
Compound semiconductors with a high absorption coefficient are the most advanced and most efficient absorber materials in thin film PV technologies. Highly efficient devices are based on absorber layers out of ternary or quaternary chalcogenides or hybrid halide perovskites.
Record efficiencies are reached with solar cells based on Cu(In,Ga)Se₂ (CIGSe) Cu₂ZnSn(S,Se)₄ (CZTSSe) and (Cs,MA,FA)Pb(I,Br)₃ absorbers (power conversion efficiency of 23.3%, 13.0% and 25.5% respectively [1]). CIGSe crystallizes in the chalcopyrite-type structure, whereas CZTSSe adopts the kesterite-type structure. Both compounds (and silicon as well) belong to the adamantine family [2] which is characterized by a 3D network of corner sharing tetrahedra. (Cs,MA,FA)Pb(I,Br)₃ (MA-methyl-ammonium, FA-formamidinium) crystallizes in the perovskite-type structure, characterized by a 3D network of corner sharing octahedra. In this sense chalcogenide materials are based on common building blocks: an A₂B₂X-tetrahedron in ternary A^IB^IIIX₂ and an A₂BCX-tetrahedron in quaternary A^I₂B^IIC^IVX₄ chalcogenide semiconductors. The parent structure of the adamantines is the diamond-type. In binary, ternary and quaternary adamantines the metals are ordered on defined cation sites which led to typical superstructures with lowered symmetry. The parent structure (aristotype) of ABX₃ perovskites is the cubic perovskite-type structure in which the B cations are octahedrally coordinated by the anions. Thus, the PbX₆-octahedra are the building blocks of hybrid halide perovskites. Tilting and distortions of the octahedra led to structures with lowered symmetry (Bärnighausen tree [3]), but all these crystal structures are named perovskite structure which leads to confusion sometimes [4]. IN the chalcogenide compounds it is a tilting and distortion of the tetrahedra which led to crystal structures with lower symmetry (e. g. wurtz-stannite structure).
The success of CIGSe, CZTSSe and (Cs,MA,FA)Pb(I,Br)₃ as PV material has to do with their overall structural flexibility. In the tetrahedrally coordinated chalcogenides this flexibility originates from the propensity of the crystal structure to stabilize intrinsic point defects as vacancies, anti-sites, and interstitials. Deviations from the stoichiometric composition led to the formation of such intrinsic point defects without dramatic structural changes, but with significant influence on the electrical and optical properties of the material [5,6]. On the other hand, hybrid halide perovskites were shown to have a high defect tolerance. Here the flexibility of the crystal structure gives remarkable positional freedom of the molecular cation and ionic movement.
The presentation will give an overview of the basic building block principle of these compound semiconductors, the flexibility of the crystal structure and the resulting effect on the optoelectronic materials properties.
[1] NREL Efficiency Chart, www.nrel.gov/pv/cell-efficiency.html
[2] B. Pamplin, Progr. Crystal Growth Charact. 3 (1981) 179
[3] H. Bärnighausen, Commun. Math. Chem. 9 (1980) 139
[4] J. Breternitz, S. Schorr, Adv. Energy Mater 8 (2018) 1802366
[5] S. Schorr, G. Gurieva (p.123); A. Lafond et al. (p.93) in: Crystallography in Materials Science, ed. by S. Schorr and C. Weidenthaler, De Gruyter, 2021
[6] S. Schorr, Sol. En. Mat. Sol. Cells 249 (2022) 112044

The bifacial cells and the bifacial modules have been the major products in the photovoltaic (PV) industry. In case of the bifacial cells, the market share is over than 60% in 2022. One of the main reasons is the higher power due to the rear side absorption as well as the front side absorption. And also the bifacial cell has the cost reduction benefit because the consumption amount of the rear side paste (Aluminum paste) is lower by over than 70%. Therefore, the market share is expected to be approximately 85% in 2032. In addition to that, the bifacial modules have more reliability properties by using the glass at the rear side instead of the backsheet. The market share of this bifacial module product will be over 60% in 2032 (ITRPV 2022).

The major parameter of the bifacial module is the bifaciality and the module power from the rear side absorption. The bifaciality (ratio of the module power of the rear and front sides ×100) of PERC in PV industry is approximately 70% in 2022 and will be 75% in 2032. This bifaciality is important for reducing the levelized cost of electricity (LCOE) because this high bifaciality value can contribute to the higher power. Thus, some researches have reported some major parameters for increasing the bifaciality, such as the frame design, the junction box design, the rear glass properties including the white grid design.

In this study, the effect of the current mismatched cells has been investigated and presented for the rear module power and the bifaciality.

The passivated emitter and rear cell (PERC) with M4 size wafer was used and the bifacial module structure follows the conventional structure in the PV market such as 2.0 mm glasses for the front and the rear side, the transparent encapsulant with 500 mm for the front and rear sides. All cells are half cut cells and total cell numbers are 144 piece in one module. 115 module were produced and the module power on the front/rear side at STC were measured and all module electrical luminescence (EL) also measured. The bifaciality for all modules were calculated.

Some modules with the lower rear module power and bifaciality were found to have some dark EL cells in modules. These dark EL cells with lower cell efficiency seem to cause the current mismatch during the rear module power measurement because the only fil factor (FF) of the modules dropped in the electrical parameters which also supported this current mismatch phenomenon. Thus, some modules were intentionally designed in this study to account for the dark EL effect on the bifacial module and its bifaciality value to verify the occurrence of this phenomenon. One of 6 strings is composed of this lower cell efficiency in a module and then the front/rear module power and module EL image were checked. And finally, these cells with lower cell efficiency can cause the more severe current mismatch at the rear module power and the bifaciality.

The several studies on the current mismatch effect on the monofacial module have been already reported, but the rear module power loss and the bifaciality were investigated in the view of the cell’s rear current mismatch effect in this study for the first time. This was more critical than that case of the monofacial module. There was more power loss at the rear module power and lower bifaciality due to the rear current mismatch of the cell.

This study indicates that the bifacial module needs to more management on the cell properties, especially rear short circuit current distribution control to avoid the rear module power loss and the bifacial loss, and then finally lower energy yield in the field.

Thin film CdTe solar cells have been fabricated through RF magnetron sputtering on soda lime glass substrates coated with transparent conducting oxide layers. The performance of such devices is expected to be sensitive to the CdTe absorber layer deposition parameters, such as the target power, which controls the flux of depositing species, gas pressure, which controls the energetic bombardment of the growing film, and deposition temperature, which controls the surface diffusion of the depositing species. Recent studies of the Urbach absorption tails of both sputtered and close space sublimated CdTe applied in solar cells have suggested that the broader band tails found in the sputtered form of CdTe provides a route for photogenerated carrier recombination [1]. Such broader band tails appear to be associated with smaller grains and residual stresses that exist even after an optimized CdCl₂ thermal treatment. These results motivates the study of sputtered CdTe solar cells prepared with absorbers fabricated at elevated temperatures in order to explore the role of deposition temperature in controlling the grain size, Urbach tail breadth, and band tail recombination for both as-deposited and CdCl₂ treated materials. The CdCl₂ treatment may play the dominant role, however, compared to deposition temperature, in controlling the CdTe grain size and the resulting optoelectronic properties. Although the effect of the CdCl₂ treatment on the quality of CdTe films and devices has been studied extensively, few studies have investigated the role of the as-deposited material properties which in the case of low temperature sputtered CdTe are more strongly modified by the treatment [1]. In this work, we have fabricated CdS/CdTe solar cell devices on 15 cm x 15 cm TEC-15/HRT glass substrates in the superstrate configuration with the CdTe deposition temperatures varied from 150 to 330 °C. The structural and the optical characteristics of the as-deposited and treated CdTe materials are investigated using X-ray diffraction, scanning electron microscopy, photothermal deflection spectroscopy, and spectroscopic ellipsometry (SE) in order to compare the impact of deposition temperature and the CdCl₂ treatment on the CdTe film structure and optoelectronic properties. Mapping spectroscopic ellipsometry (M-SE) in conjunction with specialized calibration procedures have been applied to optimize the window layer and back contact layer thicknesses independently for each deposition temperature [2]. In this process, it has been found that at the low sputtering pressures an appreciable thickness of the CdS window layer evaporates prior to the CdTe sputter deposition at the elevated temperature, and this possibly unrecognized effect has a significant influence on solar cell process optimization at elevated temperatures.

[1] J. J. Andrews, M. Beaudoin, S. K. O’Leary, P. Koirala, B. Ramanujam, X. Tan, M. A. Razooqi Alaani, P. Pradhan, N. J. Podraza, and R. W. Collins, Journal of Applied Physics 129, 165302 (2021).
[2] M. A. Razooqi Alaani, P. Koirala, P. Pradhan, A. B. Phillips, N. J. Podraza, M. J. Heben, and R. W. Collins, Solar Energy Materials and Solar Cells 221, 110907 (2021).

Polycrystalline “mosaic” semiconductors as used in current state of the art solar cells have grain boundaries that primarily extend throughout the cell’s thickness. Extensive work done in the impact of grain boundaries on the photovoltaic properties of cadmium telluride (CdTe) and other mosaic cells suggests that grain boundaries dominate the recombination of photocarriers, and thus largely determine the open-circuit voltage of such cells. Further, CdTe solar cells, one of the important commercial cells, have now achieved laboratory power conversion efficiencies above 20%. A better understanding of how this was achieved may guide research to improve efficiencies in this class of polycrystalline cells.
The work reported here illustrates a potential route to enhance the open-circuit voltage of polycrystalline solar cells by eliminating a photocarrier recombination channel that has not, to our knowledge, been identified in earlier work. We used the semiconductor modeling software Sesame to do two-dimensional numerical calculations of the optoelectronic properties of an idealized n-CdS/p-CdTe solar cell assuming that the recombination of the carrier occurs primary at the grain boundary. We find that the electron carriers photogenerated closer to the CdS interface, but relatively distant from the grain boundary, follow a shortcut through the front CdS emitter to recombine with a hole in the CdTe absorber layer. The recombination event occurs at a “hot spot” near the intersection of the grain boundary with the CdS layer. This recombination channel reduces the open-circuit voltage in the models by about 0.05 V. Blocking the shortcut recombination channel could raise the efficiency of champion thin film CdTe cells from 21%, where it has been for several years, to 22%.

Funding Sources: This research was partially supported by the National Science Foundation though Grants No. CBET-1336147 and CBET-1336134.

To improve the performance of Cu(In,Ga)Se₂ (CIGS) thin-film photovoltaic devices, a robust understanding of the dominant diffusion pathways of the alloy species In and Ga is needed. For example, because the bandgap of CIGS varies significantly between pure CuInSe₂ (CIS) and pure CuGaSe₂ (CGS), knowledge and control of In/Ga interdiffusion in the system is necessary to optimize the spatial bandgap tuning. This in turn is necessary to produce the highest-quality devices. Here, we identify the most probable defects and mechanisms for mediating In and Ga diffusion by reviewing previous experimental and computational literature. We determine that defect complexes comprised of In_Cu and Ga_Cu antisites and V_Cu vacancies are the primary drivers of diffusion. With the aid of density functional theory software packages, we calculate the binding energies and migration barriers of these complexes in bulk CIS and CGS. We find that the binding between In_Cu, Ga_Cu, and V_Cu defects can be modeled as simple electrostatic interactions between point charges. We also find that the migration barriers for Ga_Cu are much higher than those for In_Cu, explained by a larger lattice distortion arising during Ga_Cu migration than during In_Cu migration. Using these results, we develop rate expressions for the different atomistic processes and incorporate them into analytic models and kinetic lattice Monte Carlo (KLMC) simulations, which are employed to predict the diffusivity of In and Ga in CIS under variations in composition and temperature. The activation energy of In diffusion in CIS from our KLMC simulations is 1.24 eV, and for Ga it is 1.54 eV. We show that In diffusion is dominated by correlated hops, where most In_Cu-V_Cu exchanges quickly reverse. Consequently, very few hops contribute to net long-range diffusion. In contrast, Ga diffusion is dominated by uncorrelated hops, which show little to no memory of the previous state and move in a random walk. We develop an analytic model for In and Ga diffusion, suitable for use in a continuum simulation, and show that it predicts complex formation and diffusivity as seen in our KLMC simulations with great accuracy. We find our models produce results that match well with experiment. The methods used here are highly transferable and can be applied to understand diffusion of both native and extrinsic species in many different materials.

Over the years, luminescence spectroscopy has established itself as one of the fundamental methods for analyzing the photophysical properties of a variety of samples, ranging from simple organic molecules to semiconductor light-emitting materials and photovoltaic (PV) devices. The commonly used steady-state methods (i.e. excitation and emission spectroscopy) provide valuable insights into the photophysics of samples. However, such results give only a partial view of sample`s behavior after photoexcitation.
A further piece of the puzzle is often revealed by performing time-resolved luminescence spectroscopy, as it provides deeper insights into the photophysical processes occurring in the sample under investigation. It is worth to mention, that the luminescence lifetime is an intrinsic characteristic of emitting species. It indicates how long species under consideration will remain in electronically excited states before returning to ground state. Each emitting species has a characteristic luminescence lifetime, that can be influenced by its environment.
A series of spectroscopic and microscopic methods based on luminescence lifetime have been developed and provide further information, which are not accessible by steady-state experiments. Acquiring time-resolved spectroscopic data at regions of interest (ROI) of the sample can help in inferring structural-to-photophysical relationships in different materials and can give information about important photophysical processes as well as changes in the local environment of emitting species. For example, fluorescence lifetime imaging microscopy (FLIM) is a very well established imaging method in life sciences where the lifetime information is combined with spatial localization in the sample, allowing investigating biochemical or physical processes, or probing the local environment of the fluorophore. As processes commonly investigated in materials science are mostly not classical fluorescence processes, in general the term time-resolved photoluminescence (TRPL) imaging is more adequate. In materials science, TRPL imaging can be used for the characterization of key parameters like e.g. charge carrier dynamics and mobility in semiconductors. It is worth to mention, that very often these processes occur on timescales ranging from tens of picoseconds to several hundreds of nanoseconds.
Here we will demonstrate the performance of a spectrometer-microscope assembly for characterization and analysis of different materials in terms of lifetime, spectral and spatial resolution. Using a laser driver with burst capabilities enables measurements of long luminescence decays in the range of µs to ms. The benefitsof this multi-dimensional approach will be demonstrated with a series of examples reflecting a broad range of applications in materials science research.

Over the years, luminescence spectroscopy has become one of the fundamental methods for analyzing the photophysical properties of a variety of samples, ranging from organic molecules to semiconductor materials and photovoltaic (PV) devices. It is worth emphasizing that detection sensitivity is a key parameter to meet today's demands for handling weak luminescent samples and forshort measurement times in the optical evaluation of PV devices. The introduction of single-photon counting based data acquisition has proven to yield a major sensitivity increase and very high dynamic range – it is the ideal method for measuring weak photoluminescence (PL).
The commonly used steady-state luminescence spectroscopy methods provide valuable insights into the photophysics of a sample. However, such results give only a partial view of the sample’s behavior after photoexcitation. A further piece of the puzzle is often revealed by performing time-resolved luminescence spectroscopy, as it provides deeper insights into the photophysical processes occurring in the sample under investigation. An even more comprehensive picture is gained by including spatial information. Acquiring time-resolved spectroscopic data at regions of interest (ROI) in the sample can help in inferring structural-to-photophysical relationships in PV materials. Gathering such information is an important step toward the optimization of structure as well as preparation process of such materials in order to increase the performance of PV devices.
In terms of detector performance, superconducting nanowire single-photon detectors (SNSPDs) stand out due to achievable close-to-unity detection efficiency in NIR spectral range, picosecond time resolution (down to < 20 ps (FWHM)), and low dark noise (< 100 counts/s)^1,2. SNSPDs are a rather new technology that is being used in multiple applications in the field of quantum optics, luminescence lifetime measurements, and singlet oxygen detection. The high quantum efficiency is especially important for material science applications in the NIR-range beyond 1000 nm, where other available single photon detectors have low sensitivity, high dark noise and slow time response.
In this report we demonstrate the combination of the MicroTime 100 upright confocal fluorescence lifetime microscope with the Single Quantum EOS SNSPDs as a powerful tool for photophysical research on CIGS (Cu(InGa)Se₂) devices, yielding spatial and temporal information on semiconductor samples studied through PL emission.
While one of the used SNSPDs had a classical single-mode fiber coupling to guide the light onto the sensor, the other detector used an internal multi-mode fiber instead². We detected a significant increase in photoluminescence sensitivity of both designs compared to a standard NIR-PMT (H10330-45, Hamamatsu) as well as a several times higher sensitivity of the multi-mode fiber coupled nanowire compared to the singe-mode fiber one, in spite of comparable photon quantum efficiencies in this wavelength range for the sensor only. The increased sensitivity combined with the lower dark count rate resulted in an increase of the signal-to-noise ratio by more than 3 orders of magnitude compared to the NIR-PMT. Moreover, the very high sensitivity of the SNSPD as well as high temporal resolution (instrument response function of the overall system was below 100 ps) reveals clearly the differences in the PL decay profiles of the defect and unquenched areas of the weakly luminescent thin-film CIGS sample³ even at low illumination levels, which is extemely difficult to resolve when using the NIR-PMT detector.

¹https://doi.org/10.1063/5.0045990
²https://doi.org/10.1364/AO.58.009803
³https://doi.org/10.1021/acsaem.9b01370

Our previously reported 17.8 % efficiency InP nanowire solar cell1 showed a short-circuit current Isc of 29.3 mA/cm², which is not far from the theoretical maximum Isc = 34.6 mA/cm², but the loss in the open circuit voltage with respect to the radiative limit still amounted to 272 mV. To avoid this loss and reach the radiative limit we have to increase both the internal radiative efficiency η_int PL and the photon escape probability Ρ_esc towards unity. We report top-down etched InP nanowires intended to both optimize the amount of light outcoupling as well as the directionality of the emitted light. We are able to tune the tapering angle of the nanowire from 3.8° to the optimal angle of 1.8°to maximize the light outcoupling.
Top-down etched nanowires allow to implement both a front surface and a back surface field to increase both current collection and η_int PL. The photon entropy loss is governed by the (ε_in)/(ε_out) term, which is responsible for a 300 mV loss in the open circuit voltage. To circumvent this loss, we need to redirect all the emitted photoluminescence from the cell back to the sun ε_in=ε_out, For this purpose, we have fabricated PMMA microlenses by using a reflow process, which can be precisely positioned with respect to the InP nanowires; after placing the lenses all the main solar cell parameter improved by 15%.

1)https://doi.org/10.1063/1.5028049
2)http://dx.doi.org/10.1117/12.2608084

The research on Kesterite-based solar cells is reaching a turning point. After a strong effort and record efficiencies in the early 2010s, this promising thin film photovoltaic absorber, free of toxic and critical raw materials, has seen a decade of relative stalling in terms of performance despite the continuous progress of competing technologies. Recent advances in the understanding of the bottlenecks hampering Kesterite solar cells have permitted new efficiency records within the past two years, finally breaking the 13% threshold. Device modelling is bound to be instrumental going forward and in spite of an extensive usage throughout the past decade, no realistic and quantitative representation of a Kesterite solar cell has been reported. This often leads to models relying on a partial knowledge of material and device parameters extracted from different sources, and this lack of consistency severely limits the accuracy of those models both quantitatively and qualitatively while rendering results’ comparison between different laboratories almost impossible. In the meantime, low-cost tandem devices combining either inorganic thin film, perovskite and crystalline silicon cells have emerged as the next frontier for future efficient large scale photovoltaic deployment, and accurate quantitative device modelling combining electrical and optical simulation is more than ever needed to assess existing devices and predict possible improvements in future architectures.
This work reports on the modelling of Kesterite solar cells by combining SCAPS for the electrical part and the transfer matrix method for the optical part, using an in-lab developed modelling program. The objective is twofold. Firstly, accurate and quantitative baselines based on the systematic characterization of Kesterite solar cells fabricated at our group using a consistent experimental process are proposed. Our focus is on narrow bandgap Kesterite Cu₂ZnSnSe₄ (CZTSe) and wide bandgap Kesterite Cu₂ZnSnS₄ (CZTS). Specifically, we demonstrate in both cases the simultaneous reproduction of all four photovoltaic figures of merit (voltage, current density, fill factor and efficiency) with less than 1% of relative discrepancy between the model and the experimental devices.
Secondly, and using the aforementioned accurate modelling baselines and transfer matrix optical modelling, a case study on the potential of Kesterite solar cells in tandem devices is performed, namely for a narrow bandgap CZTSe subcell in tandem with a Perovskite top cell, and for a wide bandgap CZTS top cell in tandem with a crystalline silicon bottom cell. In each case, several material and optical improvements to the Kesterite cell are proposed and evaluated by the model with a high degree of quantitative accuracy, including bulk defect density, interface defects, bandgap engineering, doping gradient and selective contacts. It is found that a moderate reduction in the density of Sn_Zn antisite defect permits to significantly enhance the efficiency of CZTSe solar cells up to 16%, which in combination with a 21% efficient Perovskite top cell allows to reach a total tandem efficiency of 30%. In contrast, even ambitious optimizations to the wide bandgap CZTS solar cell do not permit to overcome the 26% efficiency threshold in tandem with a silicon subcell. The results from our simulation are compared with experimental values reported in the literature using Kesterite in tandem devices, allowing to assess the excellent robustness of the baseline parameters determined in this work. The complete set of material, optical and device input parameters are openly shared, which not only allows various research groups working on Kesterite solar cells to assess their validity, but also permits for the community to use common baselines for the numerical modelling of devices for more consistent comparisons between research groups.

Transparent indium tin oxide (ITO) electrode has been widely used for optoelectronic devices such as solar cells, semiconductor lasers, and light-emitting diodes (LEDs) due to its low electrical resistivity and high optical transparency in visible and near infrared wavelength range. Large bandgap Al_0.5In_0.5P (~2.3 eV), which is also transparent to the visible light, is often employed in the GaAs based solar cells as a window layer.
In this study, ITO was investigated as an ohmic contact material for the cathode electrode for the GaAs solar cell employing Al_0.5In_0.5P as a window layer to avoid shadowing effect of the metal electrodes. A 130-nm-thick ITO was deposited on n-type Al_0.5In_0.5P layer (3x10¹⁸ cm^-3) using radio-frequency sputtering method at 300°C and an annealing process was subsequently carried out.
Ohmic contact has been successfully formed and confirmed by the circular transfer length method (CTLM). The lowest specific contact resistivity exhibits 1.8x10^-2 Ω-cm² after annealing at 450°C for 12 min_. This value is 100 times lower than previously reported research. While showing low specific contact resistivity, high optical transmittance was also obtained in visible wavelength range. It confirms that ITO can be employed for the transparent electrode on n-type Al_0.5In_0.5P.

Mono- and multicrystalline silicon as absorbing materials in solar cells comprise about 95% of the global photovoltaics market and a total worldwide installed PV capacity of close to 3TW was reached in 2020. The benchmark efficiency of 19% for industrial modules reached in 2018 was enabled by the widespread adoption of Passivated Emitter and Rear Cell (PERC) device architecture. Typically, these cells feature 160µm-thick absorbing layers produced via the Czoralski method, which expunges up to 40% high-quality silicon as kerf loss. It has been inferred that the PERC baseline is not able to achieve the cumulative installed PV capacity target set by the IPCC of 7-10 TW by 2030 and further reduction of capital expenditure (capex) is needed. Reducing the thickness of Si-wafers for solar cells to 50µm could potentially reduce the manufacturing capex by 48%, module cost by 28%, and LCOE by 24% [1]. Thin c-silicon suffers from poor light absorption already at 100µm thickness necessitating light trapping to reach competitive efficiencies. High-performance light-trapping structures with sub-micron features, however, usually involve slow and complex manufacturing processes hindering their adoption by industry.
We present hyperuniform disordered (HUD) light-trapping structures applied to ultra-thin solar cells via substrate-conformal imprint lithography. This approach enables rapid patterning of large areas (m²) at the nanoscale and can be performed on virtually any substrate and for any type of functional layer, such as the absorber, back-reflector, anti-reflection coating or carrier transporting layers. Correlated-disorder structures were shown to outperform periodic as well as random light trapping approaches [2] and the hyperuniform platform in addition offers engineered light scattering [3]. We developed a coupled-mode theory (CMT) approach for estimating absorption with HUD patterns, which reduces the parameter space at low computational effort and is used as a starting point for numerical optimizations. Furthermore, we show experimentally measured absorption in ultra-thin, free-standing, hyperuniform-patterned silicon slabs of thicknesses ranging from 1-30µm and compare with simulations. Though silicon is our main candidate, also other PV materials can benefit from our approach and some of these examples will be presented.
Finally, we show the performance of ultra-thin Si solar cells before and after the application of HUD light trapping structures.
In conclusion, our work aims to further expand the success of earth-abundant silicon to ultra-thin, flexible and semi-transparent PV devices, which can be produced with kerfless bottom-up technologies at significantly reduced capex and accompanying CO₂ emissions while maintaining high power conversion efficiencies – A type of device that could literally pave all roads, roofs, walls, and windows by 2050!

Publication
N. Tavakoli et al., “Over 65% Sunlight Absorption in a 1 μm Si Slab with Hyperuniform Texture”, ACS Photonics 2022, 9, 4, 1206–1217, DOI: 10.1021/acsphotonics.1c01668

References
[1] Liu, Z. et al (2020). Revisiting thin silicon for photovoltaics (..). Energy & Environmental Science, 13(1), 12-23.
[2] Bozzola et al. (2014). Broadband light trapping with disordered photonic (..). Prog. Photovolt: Res. Appl., 22, 1237– 1245.
[3] Florescu et al. (2009). Designer disordered materials with large, complete photonic band gaps. PNAS, 106(49), 20658-20663.

Transparent Photovoltaic (TPV) devices open novel and unexplored pathways for the development of alternative photovoltaic technologies with a clear aim towards distributed applications, focusing on on-site generation while reducing visual impact as well as other inconveniences found in conventional PV applications. Implementation of TPV architectures can potentially allow for a wide array of applications, for example, smart windows in façades and other glazing elements such as vehicle windows, or canopies. Other examples can be the use of TPV to ubiquitously power up sensors and future IoT devices with a low-power draw. In order to deploy these applications into everyday life, it is crucial to develop the technology to be cost-effective. To do so, the pursuit of functional devices that rely on cheap, earth-abundant, stable, and non-toxic materials is of paramount importance if lab research is to be translated to industry. Additionally, the use of industrially scalable techniques is also an important requirement.
In this work, we report the development of wide-bandgap inorganic-based TPV devices that are fabricated using an ultrathin a-Si:H as a highly transparent absorber and oxide layers that are acting as carrier selective contacts and transparent electrical contacts. The basic device structure is FTO/AZO/a-Si:H/MoO₃/ITO, where FTO and ITO serve as back and front contacts while AZO is also used as electron transport layer. We will also report results on structures including V₂O₅ as hole transport material and AZO as top contact, comparing them against the reference HTL/front-contact stack, MoO₃/ITO. We will present results showing that the introduction of an intrinsic ZnO nanometric layer in-between the AZO ETL and a-Si:H leads to an overall performance increase, with the most notable improvement coming from an increased V_oc as compared to reference devices, as well as an increase in FF. Preliminary results show that there is an optimal thickness value for the ZnO interfacial film, and beyond this value there is a worsening of the device performance, which has been attributed to an increase in series resistance leading to poor transport in the TPV devices. This increase in PCE directly translates into an increased Light Utilization Efficiency (LUE), as the introduction of the ZnO interfacial film does not reduce Average Visible Transmittance (AVT) or Color Rendering Index (CRI) of the devices.
At the present time, spectrophotometry and J-V measurements performed in dark and under AM1.5G illumination (from the front and back sides) as well as Spectral Response measurements (under no bias as well as under light and electrical bias) have been carried out. As a preliminary result, we present devices with PCE>1% and AVT=40% (LUE>0.4%), overcoming the barrier of PCE ≤0.5% that has been observed in literature for state-of-the-art oxide-based devices. Record device efficiency achieved with these devices in our lab is PCE = 2%. We provide robust statistics showing that the process is reproducible thanks to the use of PE-CVD and ALD techniques, which allow for chemical composition tailoring and high film conformality even at the nanometric scale. These techniques are industrially scalable and could thus allow for a rapid TRL transition towards higher tiers. Ongoing characterization with XPS is also being performed. Ongoing work is exploring further optimization of the ETL via introduction of novel dipolar materials based on organic polymers: (PEI) and exploring different generations of dendrimeric molecules (PAMAM).
The work presented is of high relevance for the development of the field of inorganic thin film Transparent PV technologies, especially those working on extremely thin absorbers. It might also spark interest in thin film technologies in general that rely on advanced ETL/Active layer/HTL architectures.

This work has received funding from the European Union H2020 Framework Programme under Grant Agreement no. 826002 (Tech4Win).

An InGaP compound semiconductor together with In_xGa_1-xAs related layers on a Ge substrate is one of the most popular active materials in the research of multijunction (MJ) solar cells formed by serially connecting multiple p-n junctions for its applications for power generation and space satellites. According to their energy band gaps, every p-n junction in the MJ solar cell forms an individual subcell being responsible for their own light spectrum independently. Then, an overall open-circuit voltage (V_OC) of the MJ solar cell is equal to the sum of V_OC of the individual subcell. However, an overall short-circuit current density (J_SC) of the MJ solar cell is determined by the subcell having the minimal J_SC. Thus, it is of particular importance to avoid any one of the subcells failing and producing a small J_SC. The conversion efficiency of the MJ solar cell cannot be improved without so-called current match. Here, a current mismatching ratio (CMMR) for the InGaP-InGaAs-Ge based solar cell is defined as:: CMRR=(|J_SCT-J_SCM|)/(J_SCT+J_SCM), where J_SCT and J_SCM are the short-circuit current densities of the InGaP subcell and the InGaAs subcell, respectively. Besides, an average value of the J_SCT and the J_SCM represents a possibly optimized overall J_SC in the triple-junction (3J) solar cell.
In this work, a concept of J_SC-modulation effects was applied to optimize layers’ parameters of every subcell in a MJ solar cell so that a fabricated MJ can be easily verified and theoretically simulated to achieve a high-efficiency MJ solar cell. An appropriate light spot with an adjustable power was used to change and to compensate the J_SC of one of the subcells in the MJ solar cell. For example, the appropriate light spot is with a wavelength of 405 nm. The one of the subcells is the InGaP subcell which is proposed to have the minimal J_SC. In the case, the measured overall J_SC will then be increased with increasing 405 nm light-spot power until it no more dominate the overall J_SC. Both J_SCT and J_SCM are then extracted. In experimental, the structure used to fabricate InGaP-InGaAs-Ge MJ solar cells consists of an InGaP subcell, an InGaAs subcell, and a Ge subcell. The 3J solar cell was fabricated to with a cell area of 0.25 cm² and a shielding ratio of 5%. To fabricate double junction (2J) solar cells having the same absorption spectrum, only 5% area of InGaP-based layers was removed. Thus, 95% area of an InGaP-based layer remains upon the 2J solar cells to work as a short wavelength filter. Various compensated lights were employed to determine key CMMR of InGaP-InGaAs-Ge related 3J and 2J solar cells. Experimental results of using a 405 nm compensated light reveal that J_SCs of 9.37 mA/cm² and 10.28 mA/cm² were determined for the InGaP subcell and InGaAs subcell, respectively. Thus, the CMMR for the as-fabricated 3J solar cell is about 4.44%. The possibly best design for the J_SCT and/or J_SCM in the 3J solar cell is considered to be 9.814 mA/cm². An optimal conversion efficiency of 18.47% is expected. Thus, remind that a 4.44% CMRR will lead to 0.82% (18.47% - 17.65%) reduction in the conversion efficiency. Excellent agreement in evaluated properties was obtained when a 532 nm, a 638 nm, and 808 nm compensated lights were used. Another 3J solar cell fabricated with an anti-reflected coating was also evaluated. Measured results reveal that an overall J_SC of 13.5 mA/cm² is still limited by the InGaP subcell, resulting in a conversion efficiency of 27%. Together with determined J_SC of 15.5 mA/cm² for the InGaAs subcell, a possible optimum conversion efficiency of 29.11% is expected.
Finally, J_SCs for our 3J and 2J solar cells under multiple suns were also nondestructively evaluated. Important conclusions concerning CMMR variation and improvement in conversion efficiency will be reported.

As more efficient and cost-effective photovoltaic (PV) architectures are developed, solar becomes an ever more competitive and viable replacement for fossil fuels. Full grid electrification necessitates the development of efficient, reliable, cost-effective technologies - and there is room for many different kinds of PV in this expanding market. The practical challenges and constraints of terawatt PV production have brought scalability and durability into sharp scientific focus. From a materials perspective, there are commonalities in the materials questions and challenges across different PV technologies. Whereas most PV technology is referred to by the absorber layer - e.g. silicon, or perovskite solar cells, other layers in the device are equally important to performance and durability. These layers are often composed of metal oxides, and are common across device technologies - for example, interfacial layers (such as charge transport layers, CTLs), and transparent conducting oxides (TCOs) used as electrodes.This work addresses materials oxide characterization and its relationship to materials and device performance and degradation across PV technologies.

Among the most promising PVs to date are two technologies with different levels of thin film incorporation: silicon heterojunction (SHJ) and perovskite PV. SHJ cells are part of the industrial Si PV portfolio, and perovskite cells are starting to be commercially produced after a decade of intensive research. However, there are well-known stability and cost limitations associated with each technology. Understanding the thin film materials science in these devices, and using that understanding to enhance device performance and stability is key to more reliable and cost effective electricity production.

Under practical aging conditions, careful materials-level characterization is required to understand the degradation mechanisms of these materials and the complex effects of aging on the multilayer system. This work details the effects of practical degradation challenges such as damp heat (DH) exposure and encapsulation degradation on the stability of inorganic metal oxides in both the SHJ and perovskite thin film photovoltaic architectures. For perovskite devices, MAPbI₃ films were deposited by spin coating onto a range of substrates and CTLs; substrates of glass and indium tin oxide (ITO) were paired with metal oxides (MOs) including MoO_X, NiO_X, and ZnO. SE and SEM were used to characterize the surface and bulk properties of the perovskite films. Results demonstrate that the underlying layers affect the rate of absorber degradation when exposed to heat and moisture.

Unencapsulated SHJ cells (a subset of which were exposed to low concentrations of acetic acid under an applied voltage) were aged under DH 85°C/85% relative humidity conditions. The contact-ITO interface was examined via optical profilometry (OP), spectroscopic ellipsometry (SE), and scanning electron microscopy (SEM). SHJ cells have interfaces unique to their architecture, namely the c-Si/a-Si:H and a-Si:H/ITO interfaces at the top of the device. Examining the degradation of unencapsulated SHJ cells and comparing those results to historical degradation profiles of encapsulated SHJ cells in addition to the simulated effects of acetic acid exposure will help to decouple the effects of encapsulation on ITO stability in this architecture. It is well known that ethylene vinyl acetate (EVA) encapsulation degrades and produces acetic acid upon exposure to heat and humidity. When under an applied voltage, even small concentrations of acetic acid can quickly corrode the contact-ITO interface, leading to decreases in efficiency and increases in series resistance of the cell. Here, XPS was used to monitor the changes in the front contact of the SHJ cells during DH and acetic acid exposure. Changes to the materials will be correlated to changes in device performance under the same aging conditions.

The efficiency of thin film solar technologies continues to increase; most notably, that of copper indium gallium selenide (CIGS) which has demonstrated lab-scale conversion efficiencies comparable with crystalline silicon-based solar cells.(1) Marketing such technology also provides an alternative to cadmium telluride thin films which have unavoidably toxic properties. The same hazards can be avoided via the deployment of zinc-based buffer layer materials, in replacement of the popular cadmium sulfide. Alloys of zinc oxide and zinc sulfide, Zn(O,S), exhibit a tuneable bandgap such to facilitate alignment with respective solar absorbers. This tunability requires control of Zn/S stoichiometry whilst also exercising high uniformity during fabrication, and given the intricate nature of thin film architectures, controllable and scalable fabrication methods must be considered. In this vein, chemical vapour deposition (CVD) is a viable technique for thin film deposition as single-source precursors can afford stoichiometric control. However, given their thermal sensitivity, single-source metal chalcogenide precursors are ill-suited for conventional CVD methods requiring volatile precursors. Instead, aerosol-assisted chemical vapour deposition (AACVD) presents an attractive deposition process which circumvents this criterion by atomizing a solution of the appropriate precursor.
In this body of work, a novel single-source precursor design is explored as a basis for low-temperature deposition of metal chalcogenides via AACVD. A series of zinc thioamidate complexes is presented as potential precursors for the deposition of ZnS. Viability studies were conducted using thermogravimetric and NMR experiments. Viable precursor candidates give precedent for ZnO/ZnS co-deposition via AACVD for solar cell technology, and, more broadly, clean deposition of metal chalcogenide materials.

(1) M.K. Sobayel, M.S. Chowdhury, T. Hossain, H.I. Alkhammash, S. Islam, M. Shahiduzzaman, Md. Akhtaruzzaman, K. Techato, M.J. Rashid, Solar Energy, 2021, 224, 271-278.

Orthorhombic tin sulfide (SnS) is a promising absorber material for thin-film solar cells (TFSCs) due to its ideal optical band gap of ~1.3 eV, non-toxic and relatively earth-abundant constituents, high absorption coefficient (≥ 10^-5), and the high theoretical limit of ~32%. To date, the cell efficiency has mostly remained below 4% for SnS TFSCs, which is fairly low compared to its theoretical limit. The improvement in the efficiency of SnS is mostly hindered because of its low heterojunction quality with the CdS buffer layer.
This work reports the study on improving the heterojunction quality at vapor-transport-deposited (VTD) SnS absorber/CdS interface via an annealing process. The highest efficiency of ~4.3% with good long-term stability was achieved by optimizing the annealing conditions for the heterojunction interface. This improved performance was primarily attributed to the reduction in interface defects of the SnS/CdS heterojunction, which can occur either during the deposition of transparent electrodes (Al-doped ZnO) or direct annealing of SnS/CdS heterojunction at 300 °C. The enhanced heterojunction interface quality is well supported by the reduced reverse saturation current density and shunt conductance of the fabricated devices measured under dark conditions. Although the SnS/CdS device exhibits an efficiency of over 4%, significant short-circuit current loss, mainly due to recombination, was revealed by quantum efficiency and optical analysis. Admittance spectroscopy analysis shows the presence of numerous defect densities of Sn and S vacancies (〉10¹⁷ cm^-3) in the VTD-grown SnS absorber. Here, a detailed analysis of the device’s performance will be presented.

Transparent conducting electrodes (TCEs) are essential for the carrier extraction in photovoltaic devices. The use of transparent metal oxide layers, in particular indium tin oxide (ITO), is the most common used approach. However, ITO comes with numerous drawbacks such as the relatively rarity of indium and hence high cost, strong UV absorption and relative high sheet resistance. Moreover, the brittleness and fragility of ITO limits its application into flexible thin-film photovoltaic devices. An exciting alternative TCE is the use of interconnected metallic nanowires which are highly flexible, where their sub-wavelength cross-sections makes them transparent to visible light, where especially silver nanowires are a good candidate because of their high conductivity.

In this work, we show a promising cost-effective method called selective-area electrochemical deposition (SAEC) for the fabrication of these metal nanowires. SAEC is based on a through-the-mask electrochemical deposition method, where the mask is made using substrate conformal imprint lithography. Next to the inherent low fabrication cost of the electrochemical deposition, it has numerous advantages such as, scalability, operate at ambient conditions, bottom-up grow, and yields high purity deposits. While submicron sized features have been already demonstrated by SAEC, obtaining homogeneous void free filling of features below 100 nm using direct plating is a major challenge because of poor control of nucleation of the seed layer and the terminal effect.

Here, we have explored the SAEC fabrication of large area silver nanowire grids on ITO substrates and on TOPCon solar cells to demonstrate the feasibility of this fabrication method. We show that the silver nucleation density, which determines the coalescence thickness, is the key parameter for the successful homogenous void-free filling of the trenches. We control the nucleation density of the silver nanoparticles by making use of the double pulse method. Here the nucleation density is controlled with a high overpotential pulse. After nucleation, we merge the silver nuclei into nanowires as determined by the shape of the mask, by applying a low overpotential growth pulse. The SAEC-fabricated TCEs show high transmission (>96%) and extreme low resistance (as low as 3 Ω/sq), resulting into a superior figure of merit (FoM). Due to the bottom-up nature of this technique, arbitrary high aspect ratio nanowires can be achieved and therefore the relationship of decreasing transmission with decreasing sheet resistance is broken.

As a proof of concept, we have also incorporated the silver nanowire network in TOPCon cells and compared the performance to similar cells based on ITO. This work represents a low-cost and scalable strategy to fabricate sub 100 nm predefined metal structures, enabling the commercial viable fabrication of nanostructure based TCEs for optoelectronic applications, including thin-film flexible photovoltaic devices.

The remarkable efficiency improvements of Cu(In,Ga)Se₂-based chalcopyrite thin film solar cells to more than 23 % in the past few years have been largely attributed to post deposition treatments (PDTs) of the absorbers with heavy alkali metal compounds such as KF, RbF and CsF ^[1]. Therefore, understanding the effect of alkali PDTs on the electronic properties and defect physics of the chalcopyrite absorber surfaces are of utmost importance in view of the p/n junction formation and device optimization of such solar cells. In the present work, we adopt a combined analytical approach using scanning tunneling spectroscopy (STS) with a lateral resolution in the nanometer-regime and X-ray photoelectron spectroscopy (XPS) to compare the defect electronic properties and chemical composition of RbF-treated and non-treated Cu(In,Ga)Se₂ absorber surfaces ^[2]. Our STS results show that RbF PDT is effective in passivating the electronic defect levels at the absorber surface, as indicated by the low differential conductance within the bandgap. This is achieved by preventing surface oxidation and the formation of In, Ga and Se oxides on RbF-treated samples, as opposed to bare Cu(In,Ga)Se₂ surfaces. The passivating effect of RbF PDT is confirmed by ab-initio density functional theory (DFT) calculations which show that the chalcopyrite surfaces after the PDT are significantly less prone to oxidation and the incorporation of oxygen into such surfaces is energetically less favored. Furthermore, our XPS data indicate the presence of chemisorbed Rb species on the surface with a bonding configuration similar to a RbInSe₂ bulk compound, which is the likely cause of the observed surface passivation. A quantitative analysis of the XPS data shows that the Rb coverage is only in the sub-monolayer regime.
During device fabrication, the chalcopyrite absorber usually undergoes several heating steps. Hence, to understand the effect of heat on RbF-treated samples, they were investigated before and after annealing under ultra-high vacuum conditions up to 320 °C. Our measurements indicate that there is a homogeneous distribution of Rb on the surface both before and after annealing, albeit with an increased concentration at the surface after the heat treatment. This suggests that the heat treatment leads to elemental diffusion of Rb from the bulk to the absorber surface. Additional depth-resolved magnetic sector secondary ion mass spectroscopy (SIMS) measurements highlighted that this diffusion in the bulk occurs predominantly along the grain boundaries. Annealing the samples to temperatures higher than 100 °C is found to lead to the formation of metallic Rb species on the surface, which results in a significant increase in the electronic defect levels and/or surface dipole formation as shown by STS and XPS measurements. Thus, heat-induced diffusion of Rb towards the surface effectively degrades the surface and will likely cause the deterioration of the absorber-window interface due to increased recombination losses at the p/n-junction. Our results strongly suggest that RbF PDT is a double-edged sword, where in addition to its advantageous surface passivation effect, a potential detrimental effect needs to be considered, especially during further device fabrication steps (e.g. the sputter deposition of ZnO on RbF-treated Cu(In,Ga)Se₂ thin films) at elevated temperatures.

[1] Nakamura, M. et al., IEEE Journal of Photovoltaics, 9(6), 1863–1867 (2019).
[2] Elizabeth, A. et al., ACS Applied Materials and Interfaces, 14(29), 34101–34112 (2022).

The power conversion efficiency of thin film photovoltaic (PV) devices has improved considerably over the past few decades, but thermodynamic limits and practical limits determined by device simulations suggest that there is yet significant room for improvement. In addition, the typically reported efficiency of champion devices represents the tail end of a wide distribution; narrowing that distribution is critical to the continued success of the thin film PV. An ongoing challenge is determining exactly where in the devices the greatest losses are occurring. Even in crystalline devices, identifying and quantifying the various loss pathways requires multiple test structures and characterization techniques. The difficulty is further exacerbated by the polycrystalline nature of thin film PV due to the presence of grain boundaries, incommensurate interfaces, a wide array of bulk deep defects, large band tails, various crystallite facets, nonuniformities, and uncertain dopant activation energies.

The goal of this work is to explore the possibility of developing a universal method for independently quantifying V_oc losses due to front interface, back interface, and bulk recombination mechanisms. If a universal approach is not feasible, then best practices and general guidelines will be developed. From a practical perspective, the scope of work considers determination of three key parameters: i) the effective front surface recombination velocity; ii) effective back surface recombination velocity; and iii) effective bulk lifetime. These parameters, while not exhaustive, are critical for diagnosing the regions of most significant performance losses and providing guidance for further improvement. Our approach is based on considering combinations of characterization techniques, test structures, and numerical simulations to minimize the coupling between the key parameters. Hypothesis testing of such decoupling will be done by employing numerical simulation of characterization techniques such as current-voltage as a function of temperature and intensity (JVTi), quantum efficiency (QE), capacitance-voltage (CV), time resolved photoluminescence (TRPL), and others. The role of band tails in V_oc loss, such as Urbach tails, will also be considered through analysis of sub-band gap features in photoluminescence (PL) data. Numerical simulations are conducted using the finite element method to solve to the semiconductor equations with time and temperature dependence and customized photogeneration scenarios. Cd(Se,Te) PV devices will be used as specific case study.

One approach considered so far employs a set of test structures, TRPL, and simulations in a sequential procedure to minimize the number of fitting parameters in each step. Three test structures are required: i) well-passivated double heterostructures (DHs) of various thicknesses (e.g. Al₂O₃/Cd(Se,Te)/Al₂O₃); ii) DHs of various thicknesses with one contact emulating the device back contact (e.g. Al₂O₃/Cd(Se,Te)/b.c); and iii) a complete device. Initial results from a thickness series of DHs quantified the front and back surface recombination velocities and the bulk lifetime. A challenge to this approach is maintaining consistent absorber properties while varying the DH thickness and between DHs and devices. Modifications to this approach and additional methods will be explored to spatially quantify recombination and identify regions of V_oc loss.

Bandgap gradient has been successfully applied in efficient Cu(In,Ga)Se₂ and Cu(Zn,Sn)Se₂ solar cells. However, this approach has not been realized in Cd(Se,Te) thin-film solar cells. Here, a facile strategy was proposed and demonstrated to incorporate the bandgap gradient in CdTe solar cells. The key to this achievement is proper oxygen management during the device fabrication, which is compatible to state-of-the-art CdTe production. Through oxygen management, a Cd(O,S,Se,Te) region with a gradient bandgap near the front junction can form without introducing any detrimental interface. With this bandgap gradient, the devices show significantly reduced nonradiative recombination and thereby demonstrated a record efficiency of 20.03%.

CdTe/Si four-terminal tandem solar cells are a techno-economically advantageous option for high efficiency, low-cost photovoltaic cells. The tandem design reduces thermalization losses and can increase efficiency above commercial Si cells, and both junctions can be mass produced at relatively low cost. A four-terminal configuration provides flexibility in the electrical design of the tandem module, as the top and bottom cells avoid the current matching requirement that can limit performance. However, this configuration introduces significant optical challenges, as three high-transparency conductive layers and an insulating interlayer are required that all lead to increased parasitic absorption and reflection losses. This study uses a combination of experiments and simulations to characterize and understand CdTe roughness induced optical losses and provide strategies to increase optical coupling in the CdTe/Si four-terminal module.

The CdTe device stack is comprised of a planar front transparent conductive oxide (TCO) deposited on plate glass with a CdSe_xTe_1-x absorber layer. A conformally grown back TCO serves as the rear electrode and adopts the roughness characteristics of the polycrystalline CdSe_xTe_1-x absorber layer. Transmission measurements of the devices measured with UV-Vis spectrophotometry demonstrated that the sub-bandgap CdSe_xTe_1-x module transmission was enhanced with reduced CdSe_xTe_1-x roughness.

Surface height data from atomic force microscopy (AFM) measurements were incorporated into finite-different time-domain (FDTD) simulations of the CdSe_xTe_1-x cell to capture the influence of roughness on the optical properties. The CdSe_xTe_1-x transmission was modeled into air and into ethylene-vinyl acetate (EVA). The reduced index contrast by encapsulating the back TCO with EVA increases the transmission and through the CdSe_xTe_1-x top cell by as much as 33%.

The dominant loss mechanisms preventing sub-bandgap CdSe_xTe_1-x transmission were backscattering, reflection, front TCO absorption, and back TCO absorption. The optical losses are mitigated through decreased roughness and inclusion of the interlayer. The reflection and front TCO absorption are both controlled by the increased backscattering from the rougher back surface of the CdSe_xTe_1-x. Spatially resolved maps of the back TCO absorption show a non-uniform absorption profile with regions of light localization and absorption enhancement controlled by the surface topography. Mesoscale sized surface features create a focusing effect from constructive interference of the scattered and diffracted optical plane waves. The roughness induced regions of local absorption enhancement leads to the global enhancement of the back TCO absorption. The reduced index contrast with inclusion of the EVA does not eliminate the focusing effect but reduces the magnitude of the intensity enhancement. Simulations on an ideally flat surface indicated that planarizing eliminates the light localization and parasitic absorption enhancement.

Higher index optical coatings were also simulated to increase optical coupling in the tandem device. 1.7 and 1.9 refractive index optical coatings with finite thickness did not significantly increase the CdSe_xTe_1-x transmission compared to just the EVA interlayer, but combining a 1.9 and 1.7 index film to create a graded index coating increased the sub-bandgap transmission by about 2% between 900 and 1000 nm.

This study describes how increased roughness and different interlayer materials influences the reflection, TCO absorption, and optical coupling in a CdTe/Si four-terminal tandem device, and demonstrates methods to mitigate losses and promote CdSe_xTe_1-x transmission to the Si cell.

Thin film solar cells of the 2nd and 3th generation with CIGS, CdTe, CZTS and perovskite based absorber layers are very promising solar cell technologies. One of the main advantages comparing to silicon technologies are the limited materials consumption and the flexibility to deposit them on a variety of surfaces including flexible substrates. Both silicon, CIGS and perovskite solar cells have nowadays already a descent efficiency. To further improve the efficiency of solar cells scientists are often investigating their potential in tandem structures. This can be tandem structures with a thin film solar cell on top of silicon but also a combination of thin film technologies.
For monolithic solar cells not only the optical quality of the junction between both cells but also electrical losses need to be minimized. Traditionally in tandems based on III-V semiconductors this is done with tunnel junctions. However for the materials used as carrier selective contact in thin film solar cells creating heavily doped layers to enhance the tunnel probability is still questionable. Therefore recombination junctions are good candidates to fulfill the task of forming a good optical and electrical contact between the two solar cells. Nonetheless traditionally in modeling potential thin film tandem solar cells the cells are modeled individually and combined as a monolithic cell by connecting both IV curves in series and adjusting the irradiation spectrum of the bottom cell. This often results in an overestimate of the performance by not including the electrical losses in the connection in between. Many TCAD simulation packages are not including the possibility to add an interface defect that allows a direct coupling between the electron and hole currents from both sides via this recombination.
In this work we developed a theoretical model for the current voltage characteristics (IV) of a recombination junction in between solar cells. Based on this IV curve the voltage loss of a tandem solar cell with known IV curves for the individual solar cells is evaluated.
We implemented the equations for such a recombination defect in DeVSim TCAD software, which allows python scripting. These TCAD simulations allow to solve the continuity equations for holes and electrons for the whole tandem structure. This software package is used to calculate the efficiency of a 1D - Perovskite CIGS solar cell. These calculations show a good correspondence with our theoretical model. Based on these TCAD simulations the impact of the recombination junction and its properties on the performance of the a CIGS – Perovskite tandem solar cell is evaluated.

N-doped ZnTe is a common p-type material with the valence band matched to CdTe absorber. The ZnSe1-xTex semiconductor alloys offer the potential to tune the valence band position to match other absorbers, however p-type doping if these materials is less explored. In addition p-type doping of these alloys would be beneficial for wide-gap absorber applications in photoelectrochemical CO2 reduction reactions for sustainable fuels.
This presentation will discuss N-doped ZnSe1-xTex thin films grown by RF sputtering from compound targets with N2 gas as a N source. Crystalline films form across a large range of growth conditions, with a p-type carrier density of p = 1e19 cm-3 and a carrier mobility of mu = 0.05 cm2/Vs determined by Hall effect, and verified by Seebeck coefficient measurements.
Surprisingly, some N-doped films there is a transformation from the usual zinc blende structure to wurtzite. Depending on the temperature and N2 flow rate during growth, wurtzite phase can be synthesized across most of the composition range, ranging from Te-rich (x ~ 0.7) to nearly pure ZnSe (x ~ 0.1). Density functional theory calculations suggest that dilute N concentrations help stabilize the wurtzite polymorph of ZnSe0.5Te0.5.
Ongoing progress towards controlling the p-type doping towards N:ZnSe1-xTex p-type contact applicaiton, and improving the charge trasnport properties of the ZnSe1-xTex alloys towards photoelectrochemical solar energy conversion, would be discussed in this presentation.

Nitrogen stabilizes the wurtzite polymorph in ZnSe1-xTe x thin films. Journal of Materials Chemistry C 2022. DOI: 10.1039/d2tc02716j

Ternary pnictides offer a vast landscape of promising optoelectronic materials, many of which are composed of non-toxic, earth-abundant elements. Though many of these materials provide unique combinations of band gap and lattice constant compared to traditional semiconductors, making them of interest for multi-junction solar cells and other applications, their potential utility is magnified by the possibility of optical property tuning with cation ordering. This tunability could enable homoepitaxial stacks in which layers of the same material with differing degrees of ordering provide different optical properties without introducing performance-reducing crystalline defects.
One paradigm for examining cation ordering is “Short Range Order” (SRO), which examines the local bonding of the anion. For instance, in a system with the anion nearest-neighbor bonded to four cations, it is possible for those cations to be 2 of metal A and 2 of metal B, 3 A and 1 B, etc. Only some of these arrangements, or motifs, conserve the local octet rule, and the presence of octet-rule-violating motifs can strongly affect electronic and optical behavior of the material.
One system for which the SRO behavior has been examined is ZnSn(O,N). ZnSnN₂ has potential for application in photovoltaics, but nitride growth is often contaminated by oxygen, leading to a wurtzite ZnSnN₂-ZnO alloy. Computationally, it is predicted that (ZnSnN₂)_0.75ZnO_0.25 is capable of perfect SRO and that the ordered alloy would have a band gap ~0.3 eV above that of the disordered material at the same composition. [1]
This work expands upon previous study of ZnSnN₂-ZnO alloy thin films. These films were grown via RF co-sputtering and then in-situ annealed. Prior X-ray Absorption Near-Edge Structure (XANES) measurements of these films indicate differing degrees of SRO with differing in-situ anneal treatments. Prior optical measurements also show a shift in band gap similar to the shift predicted computationally. [2] In this work, cross sections of these films are examined using Scanning Transmission Electron Microscopy Energy-Dispersive Electron Spectroscopy (STEM EELS), specifically focusing on the near-edge fine structure of the oxygen and nitrogen peaks. Similarly to XANES, the near-edge structure of EELS is sensitive to bonding environment and can be related to the presence of various SRO motifs. In this work, peak shape and location is correlated to SRO.
STEM EELS provides several orders of magnitude improvement in spatial resolution over XANES, allowing for the study of SRO motif density across the film growth direction as well as across grain boundaries and other features. Here, variations in SRO among samples, as measured by EELS, are discussed and compared to XANES results, and spatial variation in SRO within each sample is assessed. Changes in anion bonding environment at grain boundaries are illustrated. This work also discusses the extraction of spatially-resolved bandgap from low-loss EELS and correlating bandgap and local SRO.
Microscopy research conducted as part of a user project at the Center for Nanophase Materials Sciences (CNMS), which is a US Department of Energy, Office of Science User Facility at Oak Ridge National Laboratory.
[1] Pan, J., et al. Perfect Short-Range Ordered Alloy with Line-Compound-like Properties in the ZnSnN₂:ZnO System. npj Comput Mater 2020, 6 (1), 63.
[2] Melamed, C. L., et al. Short-Range Order Tunes Optical Properties in Long-Range Disordered ZnSnN₂ –ZnO Alloy. Chem. Mater. 2022, 34 (9), 3910–3919.

Adamantine-type compounds, including kesterites, are currently the most promising material for fully inorganic thin-film photovoltaic technology that is free of critical raw materials and thus offers sustainable solutions. To avoid indium and to find new compounds for absorber layers as well as window materials in thin film solar cells, the A^IB^IIIX^VI₂ chalcopyrite compound family is extended by chemical substitution to quaternary "defect adamantines" such as A^I-■-B^III-C^IV-X^VI₄ compounds. With band gap energies ranging from 1.42eV for CuGaSnSe₄ [1] to 2.34eV for CuAlGeSe₄ [1], the defect adamantines are interesting materials for photovoltaic applications.

Quaternary compounds of the Adamantine family are of type A^I₂-B^II-C^IV-X^VI₄ and A^I-■-B^III-C^IV-X^VI₄. Replacing one monovalent and one trivalent cation by one IV-valent cation in the chalcopyrite crystal structure and keeping the tetrahedral coordination, the quaternary compound A^I-■-B^III-C^IV-X^VI₄ with cation vacancies can be derived, the so called “defect adamantine”. These quaternary A^I-■-B^III-C^IV-X^VI₄ compounds crystallize in the chalcopyrite-type crystal structure (space group I-42d) or the cubic spinel-type structure (space group Fd-3m). The common feature of the cubic spinel structure and the tetragonal chalcopyrite structure is the tetrahedral coordination of the chalcogen, which is arranged in an (almost) cubic close spherical packing. In spinel, 1/8 of the tetrahedral voids are occupied by A^I cations and half of the octahedral voids are occupied by B^III and C^IV cations.
Starting from the ternary chalcopyrite CuAlS₂ and CuAlSe₂, we studied the group of defect adamantine with the general formula CuAlC^IIIX^VI₄ with C=Ge, Sn and X= S, Se.
Here we report on the growth of single crystals of these quaternary compounds such as CuAlGeSe₄, and CuAlSnS₄ and their structural as well as optoelectronic properties. The single crystals were grown by chemical vapor transport using iodine as a transport agent.
The evolved material and the crystals grown were characterized by X-ray diffraction (XRD) at 298K and LeBail analysis of the diffraction pattern. For CuAlGeS₄ and CuAlSnS₄, we determined the crystal structure to be the chalcopyrite-type (space group I-42d) and the cubic spinel structure (Fd-3m), respectively. The chemical composition of the crystals was determined by X-ray fluorescence.
The band gap energy was determined from the diffuse reflectance measured by UV-VIS spectroscopy.

[1] J.C. Woolley etal. Jpn.J.Appl.Phys. 19 (1980) Suppl. 19-3, 145-148

Ternary chalcogenides are gaining increasing attention for photovoltaics, and we recently showed NaBiS₂ to have intriguing properties, including >10⁵ cm^-1 absorption coefficient at its pseudo-direct bandgap of 1.4 eV¹. As a result, the spectroscopic-limited maximum efficiency (SLME) of this material can reach 26% with an absorber thickness of only 30 nm. This demonstration was made with nanocrystals (NCs) coordinated with long-chain organic ligands, which allowed us to probe the bulk properties, but severely limited intra-NC transport. Furthermore, we found that the inhomogeneous distribution of Na⁺ and Bi³⁺ cations, which both occupy the same crystallographic lattice site, results in the formation of S p states within the bandgap of NaBiS₂ and facilitates the formation of small hole polarons that intrinsically limit the bulk mobility of this material.

In this work, we examined whether the intra-NC and inter-NC charge-carrier transport could be improved through ligand engineering. We showed through FTIR measurements that the long-chain organic ligands used during synthesis were completely substituted by short LiI, NaI, KI, or TMAI (tetramethylammonium iodide) ligands dissolved in methanol. NaBiS₂ NC films treated by these iodides all showed surface photovoltage (SPV) under illumination, with enhanced sum mobility from OPTP (optical-pump terahertz probe) and TRMC (time-resolved microwave conductivity) measurements. Together, these measurements suggest an enhancement in intra-NC mobility following ligand exchange. However, OPTP measurements showed that the enhanced mobility in iodide-treated films still decayed by ~80% within 1 ps, suggesting that carrier localisation cannot be suppressed through ligand exchange treatment. Nevertheless, from TRMC measurements, we noticed that the residual mobility after localisation was higher than the non-treated samples, meaning that more carriers could be extracted. Solar cells based on these iodide-treated NaBiS₂ NC films have reached external quantum efficiencies exceeding 50%, but their power conversion efficiencies were still below 1%. We showed that, apart from the low photocurrents limited by carrier localisation, ion migration is another hidden limitation for NaBiS₂ devices. We finish with a discussion of possible routes forward to overcome these limiting factors.

1. Huang, Y. T. et al. Strong absorption and ultrafast localisation in NaBiS₂ nanocrystals with slow charge-carrier recombination. Nat. Commun. 13, (2022).

Certain phosphorus-containing III-V semiconductors (GaP, InP and related alloys) are among the best-performing PV absorbers. Yet, there is hardly any other phosphide material that has received extensive attention for applications in PV or optoelectronics in general. Exciting progress has been reported within the family of Zn-based phosphide compounds (Zn₃P₂, ZnGeP₂, ZnSnP₂ etc.), but high PV efficiencies are yet to be demonstrated.

In this contribution, I will discuss a radically different class of semiconductors that is quite unique to phosphides. They can be referred to as “P-rich phosphides”. In stark contrast to almost any other compound semiconductor previously investigated for PV applications, P-rich phosphides contain bonds between nonmetallic atoms (in their specific case, phosphorus-phosphorus bonds).

Due to their distinct bonding mechanisms, phosphorus-rich phosphides give us an opportunity to find new materials design routes for high-efficiency PV, potentially overcoming some of the design limitations of “standard” semiconductor featuring only metal-nonmetal bonds.
I will present the first successful thin-film synthesis [1] of any polycrystalline P-rich phosphide. The synthesized material is CuP₂, which is a 1.5 eV band gap semiconductor with strong optical absorption and native p-type doping in an attractive range for thin-film heterojunction solar cells.

Finally, I will highlight a recent computational screening study [2], in which five P-rich phosphides were identified as highly promising materials for PV. These new materials are still awaiting thin-film synthesis.

[1] Crovetto et al. Crystallize It before It Diffuses: Kinetic Stabilization of Thin-Film Phosphorus-Rich Semiconductor CuP₂. J. Am. Chem. Soc. 2022, 144, 13334–13343.
[2] Kangsabanik et al. Indirect Band Gap Semiconductors for Thin-Film Photovoltaics: High-Throughput Calculation of Phonon-Assisted Absorption. J. Am. Chem. Soc. 2022. Available online. https://doi.org/10.1021/jacs.2c07567.

In this talk, I will present a brief review 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 with established CdTe and CIGS thin film PV [1]. We found that the performance of these emerging materials are limited by Voc deficit which arises from non-radiative recombination. The defects in emerging chalcogenide are generally attributed to the anisotropy in binary and ternary compounds or due to the structural complexity of quaternary compounds. Here, I will also share our strategy to facilitate directional growth of the higher mobility [hk1] planes of Sb₂(S,Se)₃ resulting in power conversion efficiency of > 9% [2]. I will also show the effect of Zn substitution in CXTS (where X is Mn,Mg,Sr,Ba,Ni,Co,Fe) on the structural, optoelectronic and photovoltaic properties [3]. We found that cation substitute must have d-orbitals to replace Zn and form a stable quaternary compound. The most promising Zn substitute is Cd as it passivates the Sn deep defects and resulted in an efficiency of ~8% for Cu₂CdSnS₄ device [4].

[1] S. Hadke, L.H. Wong* et al, Chemical Reviews, 2021, https://doi.org/10.1021/acs.chemrev.1c00301
[2] X. Jin, L. H. Wong* et al, Advanced Material, 2021, 33, 2002887.
[3] S. Lie, L.H Wong* et al, J. Mater. Chem. A, 2022,10, 9137-9149
[4] S.Hadke, L.H. Wong* et al, Advanced Energy Materials, 2019, 9 (45), 1902509.

CdTe has fundamental advantages for photovoltaics that include its near-optimal band gap for the solar spectrum and the ability to keep its defect level small in thin-film applications. At the same time panels with CdTe as an absorber can be manufactured on a large scale at low cost and demonstrated long-term stability. These advantages have translated into major commercial success in the U.S. and to some extent in China as well. Recent areas of progress have been the effective use of a CdSeTe alloy layer, enhancement of the transparent front contact, higher voltage with an organic PTAA rear layer, high luminescence efficiency of the cell structure, development of semi-transparent panels, and progress towards bifacial performance. However, several significant challenges remain, and the combination of these is limiting the voltage of the cells. Most fundamental is the large valence-band energy of CdTe, which makes it difficult to contact the typical p-type material. In addition, doping to a desired hole concentration while maintaining the low defect level has been difficult, and the passivation of interfaces and grain boundaries in a completed cell has been only partially successful. Each of these issues has been showing improvement, but the overall challenge is to find the right combination of materials and processes that can be collectively optimized. Since this challenge may be too large for an individual laboratory, researchers at several locations have formed a consortium to combine facilities and experience to jointly increase the overall progress of the CdTe technology.

We employed first principles methods based on density functional theory and hybrid theory to study Zn_xCd_1-xTe, , alloys in the zinc blende (B3) crystal structure. Cluster Expansion and Monte Carlo formalisms were used to provide a phase diagram determining consolute temperature of 387 K at 40% Zn concentration. Opto-electronic properties are calculated with the hybrid HSE06 functional for disordered Zn_xCd_1-xTe alloys, which were modelled using special quasi-random structures. Bowing effects in the bandgap and effective carrier masses were observed in alloying which can be attributed to local geometrical distortions as portrayed by bond length distributions. Downward bowing in the electronic bandgap beneficial in photovoltaic applications due to increased net photocurrent. Absorption coefficients display robust absorption in Zn_xCd_1-xTe as showed by substantial optical absorption throughout all Zn concentrations, augmented by low reflectivity that causes a high absorption. Presence of short-range order in Zn_0.25Cd_0.75Te was observed via clustering and anti-clustering among different cationic species of Zn and Cd. It has a magnitude of 0.11 at 1000 K compared to -0.79 at 400 K. The bandgap of Zn_0.25Cd_0.75Te at 1000 K was obtained to be slightly higher, 0.05 eV, than at 400 K, consistent with the value of short-range order. Thus, short-range order and possibly composition can be utilized in the design and synthesis of the appropriate solar cell, thus improving the performance in this system. Our value of calculated Urbach energy of 18.94 meV closely matches with the experimental value of 15 meV for CdTe. All the values of exciton binding energy calculated for Zn_xCd_1-xTe alloys are below 100 meV which is considered optimal for solar cells. Charge carrier mobilities are also calculated for Zn_xCd_1-xTe alloys.

Acknowledgement
We thank the Ohio Supercomputer Center (OSC) for computational resources. This material is based on research sponsored by Air Force Research Laboratory under agreement number FA9453–19–C–1002. We also thank The National Science Foundation Division of Civil, Mechanical, and Manufacturing Innovation through grant 1629239. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation thereon.

Cadmium telluride (CdTe) photovoltaic is the most cost-competitive thin-film photovoltaic technology due to its low manufacturing cost and high module efficiency. In the past decade, the power conversion efficiency (PCE) has been significantly improved. For most efficient CdTe solar cells and modules, doping with Cu or a group-V elements such as As is essential. Introducing the appropriate amount of dopants at the appropriate location is critical for achieving high PCEs. In this talk, we will discuss the situation of Cu and group-V doping via an ex-situ process, i.e., diffusion of dopants into CdCl₂-treated CdTe thin films. For Cu doping, we fabricated CdTe thin film solar cells using a metallic Cu source and ionic Cu sources (solution processed CuCl and CuCl₂). We found that the ionic Cu precursors offer much better control of Cu diffusion than the metallic Cu precursor. Therefore, we also used group V high ionic materials (i.e., group V chlorides, VCl₃ such as PCl₃, AsCl₃, SbCl_3, and BiCl₃) as dopant precursors in a solution method for ex-situ group V doping of CdTe films. Our ex-situ group V doped CdTe solar cells achieved V_OCs as high as that of Cu doped CdTe solar cells. We found that a low dopant density increases the carrier lifetime, while excessive dopants reduces the carrier lifetime. A higher dopant density near the back interface is beneficial for maximizing the fill factor of CdTe solar cells. For thinner CdTe absorbers, it is necessary to tune the doping process to optimize the cell performance.

Intrinsic zinc oxide (i-ZnO) is systematically used in combination with a buffer layer (generally cadmium sulfide) and a transparent conductive oxide (TCO) for the fabrication of the n-type side of thin film chalcogenide solar cells. However, despite this near ubiquitous empirical use, its role remains debatable. Moreover, a particularly intriguing factor regarding the performances is the role of oxygen content in the ZnO. Its concentration modifies mainly the oxide stoichiometry, conductivity and transparency. Oxygen is commonly used during the preparation of the transparent oxides, and allows improving the transmission to the detriment of the resistivity. While incorporating a little amount of oxygen during the TCO deposition is generally used for this reason , the role of oxygen in i-ZnO remains unclear but appears necessary to obtain a high resistivity layer, by preventing oxygen vacancies or zinc interstitials responsible of the n-type conductivity. Moreover, adding oxygen during the i-ZnO deposition could also affect the CdS buffer surface in a similar way as sputtered oxygenated CdS (CdS:O) commonly used in CdTe heterojunction solar cells, and could also reduce the film crystallinity, increase the optical bandgap, change the composition at the surface of the CdS and/or increase the defects density. As a result of these multiple interplays between oxygen content and film properties, the solar cells’ performance are significantly impacted.
The work here presented brings insights on the role of the i-ZnO layer in the solar cell structure by investigating the effect of the oxygen incorporation during its preparation and its influence on the performance of the devices. By varying the oxygen ratio (from 0 to 15% of oxygen in the total gas mixture with Ar) during the i-ZnO deposition, several batches of solar cells (using both CZTS and CIGS absorbers) were prepared and characterized. Limited differences were observed in their optoelectronic properties, with all the oxygen ratios leading to comparable values under illumination. However, a clear difference was observed in the dark behaviour before illumination, with devices with high oxygen content requiring longer light soaking times to reduce instabilities. To highlight this instability, a series of J-V measurements in (1) dark, (2) red light, (3) white light and (4) again dark conditions was performed in CZTSe devices. These measurements were made after keeping the samples in dark for several days to avoid any possible effects from light-sensitive defects states. A first J-V measurement in dark (dark1) was recorded followed by a red illuminated J-V curve (using a 550 nm long pass filter with a 1 sun 1.5 AM spectrum white light), then a 1 sun 1.5 AM illuminated curve and finally another dark J-V curve (dark2). All the measurements are shown in the figure 1, and put in evidence a charging mechanism with white light illumination, that is required to stabilize the devices. It will be shown during the presentation that this relates to ionized acceptor defects at both the CdS and CdS/i-ZnO interfaces, which are neutralized when photogenerated holes are created with UV absorbed photons. The potential role of oxygen vacancies in this behaviour will be discussed. These results demonstrate the importance of controlling the defects density at interfaces to reduce device instabilities. More interestingly, this photo-induced mechanism, which shows reversibility after dark discharging, could be further exploited into specific light-based memory devices, which will be the developed in another presentation of MRS Spring 2023.

Remarkable progress has been achieved in CdTe photovoltaics (PVs) to improve cell performance while reducing manufacturing costs. At the best efficiency of 22 %, researchers have demonstrated the short-circuit current (J_sc) and fill factor (FF) close to the limit of their maximum values via band-gap engineering and front-contact optimization. Considerable efforts have been devoted to increasing the carrier lifetime and the doping density (N_A > 10¹⁶ cm^-3) to improve the open-circuit voltage (V_oc) of CdTe-based solar cells. Recent studies suggested that the Voc improvement in polycrystalline CdTe PVs requires well-passivated back contact. One possible strategy is to engineer the band-gap offset (e.g., CdMgTe; △E_CB ≈ 0.2 eV) to reflect minority carrier electrons, thereby decreasing surface recombination. Another approach is to utilize a stable Al₂O₃ layer as an electron reflector; experimental works confirmed the improved performance with a conformal Al₂O₃ layer. In this configuration, the precise control of Al₂O₃ (≈ 1 nm) over the entire CdTe layer is essential because the J_sc greatly depends on the tunneling current. While promising, both CdMgTe and fully covered Al₂O₃ passivation on CdTe introduce unfavorable valence band offset that blocks the hole transport.

Here, we describe the design and fabrication of PERC-like (passivated emitter and rear contact) CdTe solar cells. Conventional lithography and wet-etching processes were used to pattern variable hole arrays on 20 nm-thick Al₂O₃ coated CdSe_(1-x)Te_x (CST) solar cells. The CST absorber materials were doped with traditional copper elements or group-V dopants (e.g., P, As, Sb; GrV). We performed quantitative and qualitative PV analysis for a series of Cu-doped and GrV doped PERC devices. In contrast to the similar behaviors of J_sc and FF, we show a significant difference in the V_oc behaviors for Cu-doped and GrV-doped CdTe PERC solar cells. The V_oc of GrV-doped CdTe PERC remains constant, whereas the V_oc of the Cu-doped PERC shows a radical decrease with an increase in the exposed CdTe area. The defect chemistry of GrV-doped CdTe can be significantly different from that of traditional Cu-doped CdTe absorber materials. We will discuss the possible mechanisms responsible for these Voc changes of the CdTe PERC devices.

The path towards the next generation of inexpensive and high efficiency solar cells is based on the utilization of earth-abundant, environmentally-friendly, and chemically stable materials. One promising candidate is zinc-phosphide (Zn₃P₂), with a direct bandgap of 1.5 eV, high absorption coefficient ( > 10⁵ cm^-1), and excellent air stability. This work will present an overview of recent key findings related to the engineering of Zn₃P₂ absorbers for incorporation into high-efficiency solar cell devices.

First part of the talk will focus on three different approaches for synthesizing high crystal quality Zn₃P₂ thin films. These will include monocrystalline layer growth using molecular beam epitaxy [1], selective area epitaxy [2] and Van der Waals epitaxial growth on graphene substrates [3].

The second part of the talk will showcase the tunability of zinc-phosphide functional properties achieved by variation in the compositional stoichiometry [4,5]. In this case, we will show how electron and X-ray diffraction and Raman spectroscopy, along with density functional theory calculations, point to the favorable creation of P interstitial defects over Zn vacancies in P-rich and Zn-poor compositional regions. Photoluminescence and absorption measurements show that these defects create additional energy levels at about 180 meV above the valence band. Furthermore, they lead to the narrowing of the bandgap, due to the creation of band tails in the region of around 10–20 meV above the valence and below the conduction band. These results are also corroborated with the transport measurements on Zn₃P₂ devices.

Finally, we will discuss the device architecture for a polycrystalline Zn3P2 thin film solar cell on an InP substrate with a 4.4 % conversion efficiency. We will show the dominant recombination mechanisms in the device using different techniques and identify the key factors that limit the device efficiency. Additionally, we will provide a perspective on the next-generation Zn₃P₂-based solar cells.

[1] M. Zamani et al. J. Phys. Energy 3 034011 (2021)
[2] S. Escobar Steinvall et al., Nanoscale Adv., 3, 326-332 (2021)
[3] R. Paul et al., Cryst. Growth Des., 20, 6, 3816–3825 (2020)
[4] E. Z. Stutz et al. Faraday Discuss., 239, 202-218 (2022)
[5] M. Dimitrievska et al. Adv. Funct. Mater. 31, 2105426 (2021)

In recent years quaternary chalcogenides have gained a lot of attention, especially the kesterite-type semiconductor compounds Cu₂ZnSn(S,Se)₄ (CZTSSe) which consist mostly of earth abundant and non-toxic elements. These compounds are a promising low-cost alternative absorber material for thin film solar cells due to their suitable criteria for photovoltaic applications: p-type semiconductor behavior, direct band-gap between 1.0-1.5 eV and absorption coefficient>10⁴ cm^-1[1]. The record conversion efficiency of 13.5% reported for a CZTSSe based thin film solar cell was reached when the polycrystalline absorber layer exhibits an off-stoichiometric composition [2]. Deviations from stoichiometry cause intrinsic point defects (vacancies, anti-sites, interstitials), which determine the electronic properties of the semiconductor significantly [3]. 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 (Cu_Zn and Zn_Cu anti-sites in Cu-Zn planes at z=¼ and ¾), which is always present in these compounds [4,5], is discussed as a possible reason for band tailing. Conventional structural characterization is done with X-ray diffraction, but in the case of isoelectronic cations, like Cu⁺ and Zn²⁺, they are difficult to distinguish. Nevertheless, the neutron scattering lengths of Cu and Zn are different; by neutron diffraction we can distinguish between Cu⁺ and Zn²⁺ site occupation in the crystal structure [4,5]. Kesterite-type based thin film solar cell technologies are mainly based on polycrystalline absorber layers, which makes it quite hard to correlate the crystallographic structure (determined via neutron diffraction) to the photovoltaic performance of these materials. A promising low-cost alternative technology uses CZTSSe monograins (single crystals of 50-100 μm size) which are fixed in a polymer matrix to form a flexible solar cell [6].
An in-depth analysis of neutron diffraction data provides information on the cation distribution in the crystal structure allowing the determination of the type and concentration of intrinsic point defects including a distinction between Cu and Zn [5]. On the other hand, neutron diffraction requires large sample volumes, thus kesterite monograins offer the unique possibility to correlate structural disorder in kesterite-type absorbers with device performance parameters.
We will present a detailed structural investigation of CZTSSe monograins with S/(S+Se)=0.8 based on neutron powder diffraction experiments, examining the influence of small changes in the chemical composition on the Cu/Zn disorder. We will show a correlation between all of the obtained knowledge on the chemical composition, the presence and types of secondary phases, the Cu/Zn disorder, the intrinsic point defects, the optical bandgap obtained from diffuse reflectance with the stability and efficiency of the respective devices. We will compare the “optimal” composition area of monograins with S/(S+Se)=0.8 to previously studied S/(S+Se)=0.6.
This work has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement no.952982. A portion of this research used resources at the Spallation Neutron Source, a DOE Office of Science User Facility operated by the ORNL. The authors gratefully acknowledge the PSI and the SINQ for providing us beamtime at the HRPT diffractometer and Australia's Nuclear Science and Technology Organisation (ANSTO) for providing us beamtime at the ECHIDNA diffractometer.
[1] S. Delbos, EPJ Photovoltaics 2012; 3: 35004.
[2] J. Zhou et al., Nano Energy 89 (2021) 106405.
[3] S. Schorr et al., in: Crystallography in Materials Science, ed. by S. Schorr and C. Weidenthaler, De Gruyter, 2021.
[4] S. Schorr, Solar Energy Materials & Solar Cells 2011; 95: 1482-1488
[5] G. Gurieva et. al., J Appl Phys 2018; 123: 161519/1-12
[6] www.crystalsol.com

CZTSSe kesterite materials constitute a promising candidate for thin-film photovoltaics, based on abundant and non-toxic elements. Despite their high absorption coefficient and tunable bandgap for either single or tandem junctions, they suffer from significant V_oc losses impeding their efficiency to level with competing technologies such as CIGS, CdTe and Perovskites. Two main culprits for this large V_oc deficit are electrostatic fluctuations and high density of bulk defects, highly dependent on the absorber composition typically considered as ideal within well-established Cu-poor and Zn-rich ranges. An increasingly popular solution to tackle these two loss mechanisms and boost V_oc is the alloying of Ge to substitute Sn, leading here to Cu₂Zn(Sn_1-x,Ge_x)Se₄ (CZTGSe) compounds, the bandgap of which is tunable through conduction band level following the x=Ge/Ge+Sn ratio. This work focuses precisely on the opto-electrical characterization of Sn-Ge bandgap-graded CZTGSe solar cells at both the absorber and complete device levels.

A sequential process is implemented and optimized, involving first the physical evaporation of a CZTG metallic precursor stack, then pre-annealed in N₂ and finally selenized. We obtain CZTGSe absorbers with low surface roughness and acceptable grain size within a Cu-poor and Zn-rich composition range. They exhibit an average x ratio around 45%, with a clear Sn-Ge segregation appearing naturally during the annealing step that leads to x=70% at the back surface and x=20% at the front surface. This corresponds to a back bandgap gradient which should simultaneously improve carrier collection and limit interface recombination, similarly to state-of-the-art CIGS. Yet, the low carrier lifetime is likely limited by poor top surface quality with observed surface defects and ZnSe secondary phases.

Advanced electrical characterization is performed on complete CZTGSe solar cell samples to gauge the impact of this Sn-Ge back bandgap grading approach and detect potential loss mechanisms. An efficiency of 7.2% is attained, along with values of 484 mV, 34.5 mA/cm² and 43.3 % for V_oc, J_sc and FF, respectively. Even though the performance is lower than record kesterite devices, the J_sc value is close to the Shockley-Queisser (SQ) limit for a confirmed minimum bandgap of 1.18eV, probably due to the largely enhanced carrier collection by our back bandgap gradient design. However, our devices are affected by important V_oc and FF deficits, i.e. about 50% of their SQ limits, for which the analysis of dark and light IV curves suggest are mainly the consequence of high-density deep defects as well as dominant electrical parasitics. This is investigated more deeply via temperature-dependent admittance spectroscopy, enabling the identification of a relatively deep defect level at the CZTGSe/CdS interface with activation energy of 260 meV and capture cross section of the order of 10^-16 cm^-2. Minority carrier recombination via this defect state is likely the explanation for the poor V_oc and FF values observed. SCAPS-1D simulations are well in agreement with these experimental findings and allow to highlight this interface loss mechanism as the main limiting factor in our samples.

Eventually, Ge inclusion into kesterites effectively permits bandgap gradient to boost carrier collection and approach SQ limits in terms of J_sc. However, our study reveals that performance is still largely restrained by V_oc and FF losses at the CZTGSe/CdS interface which should be here the first area of improvement. This is preliminarily attempted via front surface sulfurization, but further investigations are required whether it concerns the Sn-Ge bandgap gradient design or the top interface quality using chemical treatments or alternative buffer materials.

This work has received funding from the European Union's Horizon 2020 research and innovation program under grant agreement 952982 (CUSTOM-ART project).

Since the origin of thin-film (TF) solar cells, there has been an ongoing comparison between silicon (Si) and TF solar cells. It is a fact that there is much to compare. However, we believe that the vast history of Si solar cells taught us a lot and will continue to teach, as it is currently the most efficient sibling of photovoltaic (PV) technologies. The passivation (P) concept came into play for silicon solar cells around the 1980s, and then really slowly, the P concept, i.e., PESC, PERC, PERL, became the standard production design for Si PV and boosted the efficiency from ~16% to 25% [1]. These remarkable achievements, of course, took the attention of the younger sibling, TF PV technology. In 2013, Vermang et al. integrated the rear surface passivation concept into CIGS solar cells and improved the solar cell performance compared to the unpassivated ones [2]. This advancement led many researchers to look further into the passivation integration to the CIGS solar cells, including me. Here, we will present the industrially viable approaches to applying a rear surface passivation layer in ultrathin-film CIGS solar cells. As a passivation layer, first, we tried aluminum oxide (AlO_x) since it has been proven to be an excellent passivation layer for Si and CIGS solar cells. One challenge, though, is that this layer blocks the current; hence there should be contact openings in the passivation layer to allow current flow to the back contact or the applied layer must be thin enough to allow tunneling. To create the openings in the passivation layer, we simply added alkali halide salt, which is proven to be beneficial for CIGS solar cells [3], via spin-coating before CIGS growth [4]. During the deposition, depending on the type of salt, we achieved nano to micron-sized openings in the dielectric layers up to 30nm thick [5]. The succession of the AlO_x passivation layer leads us to investigate another potential dielectric layer candidate, the hafnium oxide (HfO_x). As a result, with the same simple contacting approach, the solar cells' efficiency is considerably enhanced [6]. The fact that via spin-coating, randomly distributed openings have been created in the passivation layer; we investigated whether it is an ideal way to apply the salt solution or not. For this, we benefitted from hyperspectral imaging and checked the homogeneity of the process and its effect on the solar cell performance [7]. Despite the addition of various passivation techniques, the efficiency of the CIGS solar cells is still not in line with the sector leader, Si solar cells. However, it took ~20 years for Si technology to accept the passivation concept's success and integrate it into the production process; as a younger sibling, we still have close to 15 years to learn, improve and adapt the passivation approach for CIGS solar cells. Furthermore, Si solar cells require a lot of energy during their production phase due to the wafer production process. Hence, its carbon footprint, i.e., CO₂ production, is remarkably high. The TF PV technologies, e.g., CIGS, perovskite or organic, on the other hand, do not require high deposition temperature, i.e., they have a rather low carbon footprint. Considering that we need to reduce our carbon footprint for a sustainable future, we believe that TF technology has a brighter future than other siblings.

[1] Green, M. et al., https://doi.org/10.1088/0268-1242/8/1/001
[2] Vermang, B. et al., http://dx.doi.org/10.1016/j.solmat.2013.07.025
[3] Wang, Y. Et al., https://doi.org/10.1016/j.jmst.2020.07.050
[4] Birant, G. et al., https://doi.org/10.1016/j.solener.2020.07.038
[5] Birant, G. et al., https://doi.org/10.1109/PVSC43889.2021.9518593
[6] Birant, G. et al., https://doi.org/10.1051/epjpv/2020007
[7] Birant, G. et al., https://doi.org/10.1002/pssa.202100830

Although the impact of alkali doping on the properties of CIGS solar cells has been a subject of ongoing debate for many years now, most research still focuses simply on the comparison between alkali-rich and alkali-free devices, without looking into the impact of the different alkali concentrations. We are convinced that in order to better understand the pathways of change and the specific factors impacting it, one has to study them over a broad range of different samples with controlled properties.

In this work, we present a systematic study of CIGS solar cells and corresponding thin films doped with a controlled concentration of sodium and potassium spanning three orders of magnitude. In case of thin films conductivity measurements were made, and on solar cells Deep Level Capacitance Profiling (DLCP) was used to estimate the doping level. Characterization of both cells and thin films together allows for further insight into the specific pathways of alkali impact – i.e. through distinguishing between the effects taking place on the interface and in the absorber – be it in its bulk or at the grain boundaries. Moreover, as the alkalis are known to segregate primarily on grain boundaries, a series of samples prepared with different evaporation methods, resulting in different grain sizes, were studied. Among all of these sample series, the dependence of conductivity, barrier height and free hole concentration on the alkali concentration (Na or K) follow a similar trendline, suggesting a common cause. One of effects that could lead to such result is the alkali-induced passivation of defects at the grain boundaries. As such, the values of energy barriers from conductivity measurements and the hole concentrations obtained from DLCP were then confronted with the results of 1D SCAPS simulations. The influence of donor defects at the grain boundaries on the barrier height and the average concentration of holes were studied. The simulation results were consistent not only qualitatively but also with good quantitative accuracy with the conductivity barrier heights and the free hole concentration values measured by DLCP.

All of those results point to the beneficial alkali effect being primarily linked to the passivation of donor defects and subsequent reduction of the energy barriers at the grain boundaries.

Chalcopyrite-type thin-film solar cells with bandgap energy (E_g) above 1.5 eV are ideal candidates for application as stable inorganic top cells in tandem devices. In recent years, the power conversion efficiency (PCE) could be increased for sulfide-based Cu(In,Ga)S₂ to 16.0% [1] and for pure CuGaSe₂ to 11.9% [2]. The implementation of a state-of-the-art post-deposition treatment (PDT) with heavy alkali elements like Rb or Cs resulted in an increased open-circuit voltage (V_OC) on wide-bandgap Cu(In,Ga)Se₂ (CIGS) with Ga/(Ga+In) (GGI) ratios above 0.7 [3]. Furthermore, the addition of Ag to wide-bandgap CIGS was demonstrated as a measure to reach higher V_OC values [4]. All these results were achieved with laboratory deposition coaters in a static mode. However, from an industrial point of view a fast and scalable in-line deposition process would be preferable for the growth of the wide-bandgap CIGS absorber by co-evaporation.
In this study, we present our results for wide-bandgap selenide-based CIGS solar cells deposited with an industrially relevant 30 × 30 cm² multi-stage co-evaporation system. So far, best cells with E_g > 1.5 eV achieved efficiencies up to 6% without RbF-PDT. After implementation of an in-line RbF-PDT the typical boost in V_OC and power conversion efficiency (PCE) could be obtained and best cells resulted in PCEs above 9%.
The solar cell performance strongly depends on the substrate temperature during CIGS growth and the choice of subsequent buffer and high-resistive layers. By exchanging the standard CdS/i-ZnO buffer layers with CBD-Zn(O,S)/(Zn,Mg)O, which is a good option for CIGS [5] and Cu(In,Ga)(S,Se)₂ [6] with E_g in the range of 1.0 – 1.2 eV, the V_OC gap between CdS- and Zn(O,S)-buffered wide-bandgap solar cells increased even further. Nevertheless, (Zn,Mg)O is a good partner for CdS and with this buffer combination we achieved a high V_OC of 899 mV for a CIGS cell with E_g = 1.54 eV (GGI = 0.83), resulting in an open-circuit voltage deficit (ΔV_OC) of 640 mV.
Moreover, we observe a gain in PCE as a result of slight Ag addition with low Ag/(Ag+Cu) ratios < 0.1 for CIGS absorbers with GGI > 0.6. Ag was incorporated by an additional evaporation source and these wide-bandgap (Ag,Cu)(In,Ga)Se₂ (ACIGS) solar cells exhibit increased FF values compared to the Ag-free CIGS reference samples. This observation was also made for our ACIGS cells compared to CIGS with standard GGI = 0.3, i.e. higher FF for Ag-containing samples. For high GGI values above 0.8, the Ag incorporation shifted E_g to lower values which resulted in an increased short-circuit current density and finally in an increased PCE which was accompanied by a slight increase in FF. We observe an increasing Na content in the bulk of the absorber with increasing AAC at GGI = 0.8, possibly due to altered absorber growth during deposition when more Ag is provided. Our best fully in-line prepared cell with a wide-bandgap ACIGS (E_g ≈ 1.5 eV) achieved a PCE of 11.4% (without ARC). The corresponding V_OC and ΔV_OC values were 829 mV and 680 mV, respectively.
A minor reduction of E_g of about 0.2 eV to 1.34 eV (corresponding to GGI = 0.61), i.e. near the 2^nd theoretical PCE maximum, resulted in a Ag-free CIGS cell with 18.2% PCE (with ARC) and a good V_OC of 856 mV with ΔV_OC = 480 mV.

[1] N. Barreau et al., EPJ Photovolt. 13 (2022) 17
[2] F. Larsson et al., Prog. Photovolt. Res. Appl. 25 (2017) 755
[3] S. Zahedi-Azad et al., Thin Solid Films 669 (2019) 629
[4] J. Keller et al., Sol. RRL 5 (2021) 2100403
[5] D. Hariskos et al., IEEE J. Photovolt. 6 (2016) 1321
[6] M. Nakamura et al., IEEE J. Photovolt. 9 (2019) 1863

The introduction of alkali post-deposition treatments (PDTs) for Cu(In,Ga)Se₂ (CIGSe)-based thin-film solar cells has paved the way for efficiencies above 23 % on a laboratory scale.¹ For the next big step in solar-cell performance, new approaches need to be pursued. In recent studies, the incorporation of silver in CIGSe, i.e., forming (Ag,Cu)(In,Ga)Se₂ (ACIGSe), has been identified as a promising route.² The addition of Ag increases the absorber band gap, which could increase the open circuit Voltage (V_OC).³ However, the impact of Ag on the chemical and electronic structure of the absorber surface, in particular with a PDT, is not yet understood.
In this contribution, we investigate in-line deposited and industrially relevant RbF-PDT ACIGSe absorbers and their interfaces with the molybdenum back contact as a function of the Ag/(Ag+Cu) (AAC)-ratio. For this purpose, the deeply-buried ACIGSe/Mo interface was accessed using sample cleaving and studied by synchrotron-based soft and hard x-ray photoelectron spectroscopy (PES and HAXPES, respectively), as well as laboratory-based x-ray and UV PES (XPS and UPS, respectively) and inverse photoemission (IPES). A detailed picture of the chemical and electronic structure of the ACIGSe surface and the ACIGSe/Mo interface as a function of the AAC-ratio is painted and will be discussed in view of the device performance.

(1) Nakamura, M.; Yamaguchi, K.; Kimoto, Y.; Yasaki, Y.; Kato, T.; Sugimoto, H. Cd-Free Cu(In,Ga)(Se,S)₂ Thin-Film Solar Cell With Record Efficiency of 23.35%. IEEE J. of Photovoltaics 2019, 9 (6), 1863–1867. https://doi.org/10.1109/JPHOTOV.2019.2937218.
(2) Keller, J.; Sopiha, K. V.; Stolt, O.; Stolt, L.; Persson, C.; Scragg, J. J. S.; Törndahl, T.; Edoff, M. Wide-Gap (Ag,Cu)(In,Ga)Se₂ Solar Cells with Different Buffer Materials—A Path to a Better Heterojunction. Prog. Photovoltaics 2020, 28 (4), 237–250. https://doi.org/10.1002/pip.3232.
(3) Edoff, M.; Jarmar, T.; Nilsson, N. S.; Wallin, E.; Högström, D.; Stolt, O.; Lundberg, O.; Shafarman, W.; Stolt, L. High V_oc in (Cu,Ag)(In,Ga)Se₂ Solar Cells. IEEE J. of Photovoltaics 2017, 7 (6), 1789–1794. https://doi.org/10.1109/JPHOTOV.2017.2756058.

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 chalcopyrite 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, (Ag,Cu)InSe₂ and Ag(In,Ga)Se₂ 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 [3]. The average Ag-Se and Cu-Se bond lengths determined by EXAFS clearly increase with increasing Ag content in both (Ag,Cu)GaSe₂ and (Ag,Cu)InSe₂, whereas the Ga-Se and In-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 and (Ag,Cu)InSe₂ shows the same behavior. In contrast, Ag(In,Ga)Se₂ exhibits a bond length dependence similar to that previously observed for Cu(In,Ga)Se₂. This clearly demonstrates that the peculiar behavior of a decreasing bond length with increasing crystal lattice is related to the mixing of the group-I lattice site, independent of the nature of the group-III atom. Furthermore, the element-specific bond lengths determined by EXAFS enable us to model the anion positions of the different local motifs present in the alloys and to estimate their contribution to the band gap bowing. Our study thus provides unique and systematic insight into the local structure of the (Ag,Cu)(In,Ga)Se₂ alloy system, including its impact on a key material property.

[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).

Cu(In,Ga)Se₂ (CIGS) and its derivative materials are versatile, as well as resource and cost-efficient and thus, attractive for photovoltaic energy conversion device applications. To date, we have developed lightweight and flexible Cu(In,Ga)Se₂ (CIGS) minimodules using a monolithically interconnected structure with independently certified photovoltaic efficiency values greater than 18% (number of cells: 17, designated area: 68 cm², without Ag or S alloying). This accomplishment was the culmination of techniques for controlling alkali-metal doping, the activation of metastable acceptors, and a minimodule fabrication process developed on flexible substrates. Modification of surface and interface, and bulk crystal quality by alloying with Ag or S, or other elements is expected to be a promising approach for further enhancements of the photovoltaic efficiency value. In addition, it has been suggested that mechanically scribed cell edges (equivalent to the side wall of P3 in the module fabrication process) can be an important factor in the degradation of device performance; therefore, there is room for improvement with proper passivation/termination treatments in the cell separation process. In fact, significant increases in the shunt resistance, fill factor (FF), and thus, photovoltaic efficiency values, and suppression of photovoltaic performance deterioration under low irradiance conditions were observed for CIGS solar cells fabricated using a photolithography-based cell separation process in lieu of a conventional mechanical scribing process.
Wide-bandgap CIGS-based photoabsorber materials are promising candidates for photocathode applications for the splitting of water into hydrogen as well as top-cell applications in tandem solar cells. We have demonstrated an over 8% half-cell solar-to-hydrogen (HC-STH) efficiency using ternary CuGaSe₂ (CGS)-based thin-films. This HC-STH value was demonstrated for a photocathode that underwent surface modification of CGS layer with RbF (q.e. not a postdeposition treatment). It has also been suggested that such surface modification of CGS thin-films with a Cu-deficient phase layer (CDL) may be effective in enhancing both the HC-STH and the open circuit voltage (V_OC) values of corresponding CGS solar cell devices. A CGS/CdS solar cell fabricated with a CDL demonstrated an independently certified photovoltaic efficiency value of 11.05% with a V_OC of 0.960 V, a short circuit current density of 15.9 mA/cm², and a FF of 72.4% (designated area: 0.497 cm², without RbF). These results suggest that improving wide-bandgap CIGS-based device performance requires different approaches from those used for conventional narrow-bandgap CIGS devices.

With the aim to reduce recombination losses in the kesterite based solar cells, the Ag alloying was performed on Cu₂ZnSn(S,Se)₄ monograins, leading to the formation of (Ag_1-xCu_x)₂ZnSn(S,Se)₄ (ACZTSSe) solid solution. Ag is known to partially replace Cu in ACZTSSe, reducing the influence of Cu-Zn disordering and improving the open circuit voltage (V_OC) of the solar cells. Controversial results on the influence of the substitution of Cu by Ag in kesterite thin films have been reported in the literature. However, Ag alloyed kesterite thin film solar cells have been shown to have improved performance due to the reduced defect concentrations compared to the pure Cu₂ZnSnSe₄ and Cu₂ZnSnS₄ absorbers. Recently, Yuancai Gong et al. reported 13 % efficient thin film solar cells based on Ag alloyed kesterite CZTSSe^[1].
We have previously studied the influence of Ag alloying on the properties of Cu₂(Zn,Cd)SnS₄ monograin powders and on the performance of the corresponding monograin layer solar cells^[2]. The incorporation of Ag into the Cu_1.85(Zn_0.8Cd_0.2)_1.1SnS₄ monograin absorber material improved the efficiency of monograin layer solar cells from 6.62% (x = 0) to 8.73% (x = 0.01) resulting from the changes in the defect structure, which led to reduced non-radiative recombination losses in the absorber.
Here, we report the influence of Ag incorporation into the Cu₂ZnSn(S,Se)₄ monograins on the structural, optical and electrical properties of the absorber material as well as on the performance of the monograin layer solar cells. Various ratios of [Ag]/ ([Cu] + [Ag]) were used in the synthesis process and losses of outcome Ag concentrations in the resulting material compared to the input were observed. The formation of ACZTSSe solid solutions is determined by X-ray diffraction as well as Raman spectroscopy. Changes in the defect structure and related radiative recombination mechanisms were studied by temperature dependent photoluminescence spectroscopy and correlated to the defects detected by admittance spectroscopy of the corresponding solar cells. Reduced recombination of photoexcited charge carriers as well as improved device performance of the ACZTSSe compared to the CZTSSe monograin layer solar cell devices is reported.

References:
[1] Yuancai Gong et al. Nat Energy 7, 966–977 (2022). https://doi.org/10.1038/s41560-022-01132-4
[2] Kristi Timmo et al. J. Mater. Chem. A, 2019, 7, 24281–24291, DOI: 10.1039/c9ta07768e

Cu₂ZnSn(S, Se)₄ compounds, widely known as CZTSSe or kesterte referring to their crystal structure, are a promising material for thin film solar cells due to their composition out of non-toxic and earth-abundant elements and suitable optical and electronic properties . Yet, the performance of these solar cells is limited by defect-driven recombination, with the highest-ever reported power conversion efficiency being 13.6% so far. The properties and suitability for PV application of thin-film kesterite absorbers are improved upon alloying with light alkali elements such as Li and Na, owing to defect passivation and increased carrier concentration. Moreover, Li alloying can be used to tune the absorber bandgap. However, the presence of light alkali elements, either added intentionally or diffused from the substrate, interacts with kesterite phase synthesis during annealing at high temperatures. As a consequence, Li-alloying in (Li, Cu)₂ZnSn(S, Se)₄ kesterite absorbers is limited to only a few percent of Li/(Li+Cu), before degraded morphology deteriorates the solar cell performances.
In this work we present an alternative strategy to incorporate controlled amounts of Li into crystallized kesterite absorbers, bypassing the alkali concentration limitations related to absorber synthesis. The Mo back contact of the spin coated and annealed CZTSSe absorbers is electrically connected to metallic Li and both are placed in a liquid electrolyte yielding a so-called half-cell configuration. Implementing this battery-inspired approach, we demonstrate the electrochemical treatment of CZTSSe and the successful transfer of Li into the absorber layer. We show that the treatment results in the incorporation of Li into the kesterite crystal grains, yielding alloyed Li_xCu_2-xZnSn(S,Se)₄ absorbers with modified chemical composition. Using liquid electrolyte allows us to disassemble the half-cell after the treatment and deposit subsequent layers on top of the absorber to fabricate a working solar cell device with decent performance.
We present the benefits and limitations of our novel electrochemical Li alloying strategy based on morphological, structural, electrical, and optical characterization of treated absorbers layers and devices. We further discuss how this innovative approach could be combined with conventional alkali doping and alloying strategies in order to circumvent deteriorated solar cell performances related to the absorber synthesis in presence of high Li concentrations. Furthermore, the presented electrochemical approach opens new research directions, such as the investigation of ion migration in CZTSSe absorbers.

Renewable energy production comes hand-to-hand with the development of suitable and future photovoltaic technologies, where large scale production depends on new concept thin film materials and device architectures, based on low cost processes and critical raw materials free, where Cu₂ZnSnSe₄ presents as a good candidate. In this context, it is a necessity to jump from small-scale laboratory sized cells to wider adaptable sizes, to achieve a plethora of different end applications, not only extensive (solar farms) but also for delocalized production (IoT, ViPV…), where the technological flexibility of the thin film technologies for the development of cells and modules with customized design is strongly relevant. The problems and difficulties of upscaling to a larger surface reveal themselves in an increased lateral inhomogeneity due to the increased ratio of area to cover vs thickness of the layers. The devices herein explored are based on a thin film structure, for the samples to perform properly, the layers need to present high quality and maintain their properties laterally, which is indeed a challenge, especially if the production cost has to be kept low. This challenge comes as a binary problem: the understanding of the specific physical or chemical phenomena, and the measurement of the effects of these variations in the final output of the PV devices. Usually, subtle variations, like thickness oscillations or compositional shifts, are unobserved at a small scale, because the process fluctuations can happen in areas larger than the measured cell, and finally define or limit the output of a device.
In this research a set of 3 samples (5x5 cm²) are studied as the initial step for large-scale production of Cu₂ZnSnSe₄ devices. Each sample is isolated in 3x3 mm² individual cells, which permits to achieve a complete data library that aids to obtain a considerable amount of cell by cell information to implement statistical analysis. In those devices (3x3 mm²), several characterization techniques are employed comprising: Raman spectroscopy (RS) with λ_ex = 442 and 532 nm; photoluminescence (PL) with λ_ex = 325, 442 and 532 nm, temperature dependent PL (PL-T) with λ_ex = 442 nm; current voltage characteristics (J-V); and image analysis. With the combination of those techniques, we can extract how the physicochemical properties of the front interface layers CZTSe/CdS/ZnO/ITO in the device architecture affect the final output of the device or influence each other. The data library is then composed of 3 x 196 cells that are analyzed as one complete dataset, where statistical correlations and small variations arise between the cells with some insightful results. Fluctuations are found in the CdS buffer layer as small thickness variations, bandgap changes and non-uniform defect energies. These variations of the CdS layer turn out to be strongly correlated to the final efficiency of the individual cells and define the observed patterns in optoelectronic performance along the surface of the 5x5 cm² samples, which also reflects in the visual appearance of the sample. These small changes, intrinsic to the processes, otherwise unseen in a discrete set of small samples or a set with strong variations in one of the fabrication steps (i.e. different samples with different absorber synthesis temperatures), imply strong modifications in the final efficiency of the individual devices, which are usually not perceived due to the low statistical significance of the data sets, this could potentially shadow or misguide the results of experiments and thus hamper progress in the development of these complex structures. In addition, the correlations found can be used as a quality control parameter in the fabrication of CZTSe modules, with simple approaches like visual inspection, applied in-line in the process.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 952982 (Custom-Art)

Increasing the power conversion efficiency (PCE) of kesterite Cu₂ZnSn(S,Se)₄ (CZTSSe) solar cells has remained challenging over the last decade, in part due to V_oc-limiting defect states at the absorber/buffer interface. Previously, we found that substituting the conventional CdS buffer layer with In₂S₃ in CZTSSe devices fabricated from nanoparticle inks produced an increase in the apparent doping density of the CZTSSe film and a higher built-in voltage arising from a more favourable energy band alignment at the absorber/buffer interface. However, any associated gain in V_oc was negated by the introduction of photo-active defects at the interface (Campbell et al., J. Appl. Phys 127, 205305, 2020). This present study incorporates a hybrid Cd/In dual buffer in CZTSSe devices which demonstrate an average relative increase of 11.5% in PCE compared to CZTSSe devices with standard CdS buffer, with the best performing dual buffer device (CZTSSe/CdS/In₂S₃) achieving 7.8% PCE (compared to 5.9% PCE in the CZTSSe/In₂S₃/CdS hybrid buffer device).
Spectral response measurements of hybrid buffer devices confirmed the presence of photo-active interface defects when the In₂S₃ buffer is adjacent to the CZTSSe absorber. Current density-voltage analysis using a double diode model revealed the presence of i) a large recombination current in the quasi-neutral region (QNR) of the CZTSSe absorber in the standard CdS-based device, ii) a large recombination current in the space charge region (SCR) of the hybrid buffer CZTSSe/In₂S₃/CdS device and iii) reduced recombination currents in both the QNR and SCR of the CZTSSe/CdS/In₂S₃ device. This accounts for the notable 9.0% average increase in J_sc observed in the CZTSSe/CdS/In₂S₃ in comparison to the CdS-only CZTSSe solar cells.
Energy dispersive X-ray (EDX), secondary ion mass spectroscopy (SIMS) and grazing incidence X-ray diffraction (GIXRD) compositional analysis of the CZTSSe layer in the three types of kesterite solar cells suggests there is diffusion of elemental In into the absorbers with a hybrid buffer. In incorporation into the CZTSSe absorber increases the doping density of the SCR in the absorber. In diffusion arises from the post-deposition heat treatment (PDHT) of the hybrid buffer layers following their successive chemical bath depositions. When the In₂S₃ buffer is adjacent to the absorber of the hybrid buffer devices it is subjected to a double PDHT which greatly increases In incorporation into the absorber. As a result, the doping density in this absorber reaches 10¹⁷ cm^-3 which negatively affects device performance. However, in the CZTSSe/CdS/In₂S₃ device, the single PDHT of the In₂S₃ layer is sufficient to increase the doping density of the SCR region of the CZTSSe without adversely affecting overall device performance.
It is expected that optimisation of the thickness of the hybrid buffer layers will lead to further improvements in device performance. These hybrid In₂S₃/CdS buffers provide a promising way to boost the efficiency of kesterite CZTSSe solar cells.

Tandem photovoltaic devices based on CIGS and perovskite solar cell (PSC) sub-cells offer the implementation of all-thin-film tandems with all the advantages of thin-film technology, together with the possibility to achieve high efficiencies above the thermodynamic limit of single-junction solar cells. More precisely, the CIGS/perovskite tandem solar cell technology allows for low cost, light-weight, large area roll-to-roll fabrication and lower integration and installation costs, finally promising lower energy costs compared to current technologies. To this date, record power conversion efficiencies of over 24% and 27% have been achieved for monolithic two-terminal (2T) and stacked four-terminal (4T) tandems, respectively.
The crucial step for coupling CIGS and PSC devices in one tandem cell lies in the optimization of the layer stack with respect to light management and current collection, which include the growth of the transparent electrode layers. These requirements become even more stringent in monolithic 2T tandems, where cell performance requires current matching between top and bottom cells. This means that in particular for the CIGS cell we need to tailor the photoactive band gap to make optimum use of the partial illumination in the spectral region above ~750 nm wavelength, which is transmitted through the perovskite cell stack. One means to realize a beneficial trade-off between gains in the open circuit voltage and losses due to reduced photoabsorption is by the design of band gap grading, i.e. Ga in-depth distribution in the CIGS absorber film.
Hence, our approach to achieve high performance monolithic interconnected perovskite/CIGS 2T tandem solar cells rests on four major building blocks. (i) The adaptation of the CI(G)S bottom cell including control over the surface roughness, (ii) growth and optimization of the PSC layer stack on top, notably with focus on the hole transport layer, (iii) the design of the tunneling/recombination junction connecting the two sub-cells, and (iv) the implementation of light management structures for improved current matching.
All these tasks involve a dedicated focus on the control and analysis of the interfaces in each sub-cell and between the two cells, which will be the focus on my talk. I will lay out our methodology to assess the interface chemistry and the effect on its optoelectronic properties and functionality in the device. This point is particularly relevant for the optimized PSC and significantly affects the overall device performance and stability.

By stacking two or more solar cells in a tandem configuration, the solar cell can more efficiently utilize the solar spectrum, and power conversion efficiency (PCE) is much higher than the Shockley-Quessier limit of single junction solar cells can be obtained. Theoretical calculations on tandem solar cells have predicted an ideal combination of ~1.1 eV and ~1.7 eV as suitable bottom and top cells, respectively. Wide bandgap perovskite and Cu(In, Ga)Se₂ (CIGS) are ideal materials for tandem solar cells, which carried out the theoretical calculations assuming that both the efficiencies of top and bottom cells reach the Shockley-Quessier limit. However, in reality, deviation from the Shockley-Quessier limit is more significant for a wider bandgap perovskite. Therefore, we adopted perovskite with a narrower bandgap than the ideal value (~1.7 eV) for tandem solar cells with CIGS. As the bandgap of perovskite was reduced, a CIGS bottom cell's bandgap must also be reduced for better current matching. Ga content can lower the overall bandgap of CIGS by precise control during deposition. Typical high-performing CIGS includes the so-called double grading bandgap profiles in which the bandgap first reduces with the thickness and opens up towards the front interface with n-type CdS buffer layer; however, it has not been confirmed whether the double-grading, which was optimized for single-junction CIGS solar cells, is still the best option for tandem solar cells. In this study, we controlled the bandgap profile of CIGS and compared single- and double-grading configurations of CIGS with a similar overall bandgap for applications to tandem solar cells with perovskite. Our investigation on the CIGS bottom cell under red light-filtered (i.e., wavelength smaller than 825 nm) illumination to mimic actual conditions during the operation of perovskite/CIGS solar cells showed that single-graded CIGS have a higher fill factor value than the double-graded counterpart. With further optimization of the monolithic tandem devices, we achieved PCE of 25.1% using perovskite of 1.65 eV and CIGS of 1.04 eV. Our study highlights the importance of bandgap tuning perovskite/CIGS tandem solar cells.

Metal-halide perovskites are of tremendous interests for both single-junction and multi-junction photovoltaic devices due to their easy bandgap tunability via composition engineering. Perovskites possess long carrier diffusion lengths and high defect tolerance, enabling low-cost and solution-processed solar cells. To date, the power conversion efficiency (PCE) of single-junction perovskite solar cells (PSCs) has reached a certified value of 25.7%, which approaches that of crystalline silicon solar cells.
Monolithic all-perovskite tandem solar cells have the promise to deliver higher PCEs than do single-junction PSCs while maintaining the benefits of low-cost solution processing. Wide-bandgap (WBG, ≈1.8 eV) perovskite is a crucial component to pair with narrow-bandgap perovskite in low-cost monolithic all-perovskite tandem solar cells. However, the stability and efficiency of WBG PSCs are constrained by the light-induced halide segregation and by the large photovoltage deficit.
In this presentation, I will discuss our recent progress on a steric engineering that obtains high-quality and photostable WBG perovskites (≈1.8 eV) suitable for all-perovskite tandems. By alloying dimethylammonium and chloride into the mixed-cation mixed-halide perovskites, wide bandgaps are obtained with much lower bromide contents while the lattice strain and trap densities are simultaneously minimized. The WBG PSCs, with a considerably improved PCE of 17.7% and an increased Voc of 1.26 V, exhibited superior operational stability and maintain 90% of their initial PCEs after 1000 h operation at the maximum power point (MPP). With the triple-cation/triple-halide WBG perovskites enabled by steric engineering, a stabilized power conversion efficiency of 26.0% in all-perovskite tandem solar cells is further obtained. The strategy provides an avenue to fabricate efficient and stable WBG subcells for multijunction photovoltaic devices.

Quasi one-dimensional (Q1-D) structures based on van der Waals materials such as (Sb,Bi)₂(S,Se)₃ have demonstrated impressive progresses over the past 5 years. These types of materials typically form nano/micro-ribbons with strong covalent bonding in one direction, and at the same time, weak van der Waals bonding in the others two. This fingerprint feature confers unusual optoelectronic properties to these materials when properly oriented, and has allowed fast progresses in terms of conversion efficiency of solar cell devices, with a current record exceeding 10%. Beyond Sb and Bi-chalcogenide compounds, there is a limited body of knowledge on other Q1-D systems such as mixed chalco-halide van der Waals compounds, that combine chalcogens (S,Se) with halogens (Br,I) in the same structure. In fact, recent theoretical studies suggests that mixed chalco-halides can simultaneously achieve the robustness and stability of chalcogenide materials, along with the excellent optical and electrical properties of halides, and in particular showing high defect tolerance similarly to halide-perovskites.

This presentation will introduce a novel family of materials based on mixed Sb and Bi chalco-halides [(Sb,Bi)(S,Se)(Br,I)]. The first part will be devoted to reviewing the most relevant results reported so far for the few examples available in the literature, that have reached encouraging conversion efficiencies around 5% within a very limited timeframe, with a meso-porous solar cell configuration. In addition, the fundamental properties of these compounds obtained by DFT modelling will be discussed. The calculation of the bandgaps, band structures, optical and transport properties will be presented, discussing the trends of these properties depending on the chalcogen or the halogen introduced into the structure. The results will show that all these compounds can be relevant for different thin film photovoltaic applications, with bandgaps ranging from 1.6 eV for SbSeBr up to 2.3 eV for BiSBr.

In the second part of the presentation, the complexity of the synthesis of mixed chalco-halides will be discussed, and a new methodology for their synthesis, developed by the authors of this work, and based on the combination of co-evaporation of chalcogenides and high-pressure reactive annealing under halogen atmosphere, will be presented. This methodology allowed for the first time the demonstration of working solar cells with absorbers produced by vapor deposition techniques with these compounds. The tunability of the Q1-D structures by changing the synthesis temperature and pressure will be demonstrated. It will be shown that the synthesis of bromine-based compounds requires higher temperatures and pressures than iodine ones, due to the thermodynamics associated with the incorporation of Br and I in the chalcogenide phase. A detailed analysis of the mechanisms behind the formation of the different compounds will be presented by implementing interrupted-synthesis processes, and supported by the combination of thermo-gravimetric analysis, differential scanning calorimetry, and advanced structural/compositional characterizations.

The last part of the presentation will be devoted to the challenges and possible technological solutions for the fabrication of planar-heterojunction solar cell devices with these innovative photovoltaic absorbers. The use of different electrons and holes transport layers will be discussed, demonstrating for the first-time conversion efficiencies with these architectures between 1-5% and with very encouraging Voc values above 600 mV in some cases. Finally, the perspective of these materials and the possible advantages with respect to current chalcogenide and halide technologies, will be presented and discussed.

Although halide perovskites have taken the photovoltaics field by storm, concerns over the toxic and easily accessible lead content have prompted a search for lead-free alternatives. Bismuth-based compounds have particularly gained attention, owing to their electronic similarities to lead, but advantages of being free from the same toxicity limitations [1]. This talk focuses on our recent work on one of these materials: BiOI.

BiOI is not a perovskite, but replicates important features of the electronic structure of lead-halide perovskites that are believed to be conducive to defect tolerance. We show through computations and experiment that BiOI indeed tolerates the dominant vacancy and anti-site defects. Whilst the bandgap of BiOI (1.9 eV) is too wide for single-junction photovoltaics, we show that it is optimal for indoor light harvesting for sustainably powering Internet of Things (IoT) devices [2], and also suitable for solar energy harvesting as a photocathode for water splitting [3]. In the latter work, we develop a device structure that improves the operational stability of BiOI photocathodes from minutes to months, and combine these devices with BiVO₄ photoanodes to realise the first demonstration of all-oxide-based bias-free syngas production [3]. We also show this device structure to be beneficial for other novel halide photocathodes, such as lead-halide perovskites [4]. We finish off with a discussion of the key factors limiting charge-carrier transport and Urbach energies in BiOI, and how these could be overcome by tuning the composition and structure of this material.


[1] Ganose, Scanlon, Walsh, Hoye*, Nat. Commun., 2022, 13, 4715
[2] Peng, Huq, Mei, … Hoye*, Pecunia*, Adv. Energy Mater., 2020, 11, 2002761
[3] Andrei, Jagt, …, Driscoll*, Hoye*, Reisner*, Nat. Mater., 2022, 21, 864
[4] Andrei, et al., Nature, 2022, 608, 518

Bismuth vanadate (BiVO₄) is an outstanding candidate for visible-light driven water splitting due to its low production cost, high stability, low toxicity, and reasonable band gap with well-placed conduction and valence band edges relative to both hydrogen and oxygen evolution potentials.¹ In this presentation, the development and application of new and novel precursor systems tailor-made for the dual-source aerosol-assisted chemical vapour deposition (AACVD) of highly nanostructured, phase-pure monoclinic-scheelite BiVO₄ thin films is outlined. Having identified the ‘best in show’ precursors for AACVD application, the BiVO₄ thin film deposition conditions are optimised for the most suitable precursor ratios and deposition temperatures for high quality films, with variations in nanostructure as a function of deposition time also being investigated in parallel by scanning electron microscopy (SEM). Light-driven water splitting performance is then evaluated using standard techniques including UV/Vis spectroscopy, linear sweep voltammetry under simulated sunlight irradiation, incident photon-to-current efficiency (IPCE) measurements, chronoamperometry, and electrochemical impedance spectroscopy (EIS). Pristine BiVO₄ films yielded photocurrent densities of 1.23 mAcm^-2 at 1.23 V vs RHE, the highest photocurrent density of any CVD-grown BiVO₄ film to date.

The potential of AACVD for semiconductor doping and hybridisation studies are discussed, most importantly the compatibility of new metal precursors with existing ones. This is exemplified by the facile deposition of W-doped BiVO₄ without adding complexity or steps to the process. This work provides a roadmap in which an effective ‘toolbox’ of cross-compatible metal precursors can be rapidly deployed with AACVD to fabricate a range of doped and/or mixed-metal materials with high photocatalytic performance.


References
1 C. Jiang, S. J. A. Moniz, A. Wang, T. Zhang, J. Tang, Chem. Soc. Rev., 2017, 46, 4645–4660.

I-V-VI₂ ternary chalcogenides have recently attracted growing attention as earth-abundant, nontoxic, and air-stable absorbers for photovoltaic applications.^1,2 In particular, our recent work on the NaBiS₂ & AgBiS₂ members of this family has revealed ultra-strong optical absorption for these compounds – the highest of all current PV materials.³ A key benefit of such intense light absorption is that it allows for ultrathin (<100 nm) solar cell devices, dramatically reducing material consumption, weight and manufacturing demand, directly lowering the cost and facilitating applications in space for example, in addition to benefitting quantum efficiency and photovoltaic (PV) performance.

Our collaborative work on AgBiS₂³ showed the crucial importance of controlling cation distribution and disorder in these materials, yielding record-breaking efficiencies >9% – the highest of any Bi-based solar absorber.³ However, the impact of disorder on the charge-carrier properties in these materials is remains poorly understood. Herein, we investigate the key properties which dictate the relationship between disorder on the cation sublattice and carrier transport in these materials. We find the band-edge orbital character to be a crucial factor in the sensitivity of carrier localisation (and thus solar cell efficiency) to cation disorder, resulting in ultra-fast carrier trapping, despite slow carrier recombination, in NaBiS₂.⁴ We extend this analysis by explicitly calculating the phononic properties of these compounds, alongside Inelastic Neutron Scattering (INS) measurements, to probe the carrier de-population process in these compounds.

This work reveals the critical role of cation disorder in the photovoltaic performance of these disordered inorganic PV compounds, alongside key considerations for future research in this area.

(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) Schnepf, R. R.; Cordell, J. J.; Tellekamp, M. B.; Melamed, C. L.; Greenaway, A. L.; Mis, A.; Brennecka, G. L.; Christensen, S.; Tucker, G. J.; Toberer, E. S.; Lany, S.; Tamboli, A. C. Utilizing Site Disorder in the Development of New Energy-Relevant Semiconductors. ACS Energy Lett. 2020, 5 (6), 2027–2041. https://doi.org/10.1021/acsenergylett.0c00576.
(3) Wang, Y. ^‡ & Kavanagh, S. R. ^‡; Burgués-Ceballos, I.; Walsh, A.; Scanlon, D.; Konstantatos, G. Cation Disorder Engineering Yields AgBiS2 Nanocrystals with Enhanced Optical Absorption for Efficient Ultrathin Solar Cells. Nat. Photon. 2022, 16 (3), 235–241. https://doi.org/10.1038/s41566-021-00950-4.
(4) Huang, Y.-T. ^‡ & Kavanagh, S. R.^‡; Righetto, M.; Rusu, M.; Levine, I.; Unold, T.; Zelewski, S. J.; Sneyd, A. J.; Zhang, K.; Dai, L.; Britton, A. J.; Ye, J.; Julin, J.; Napari, M.; Zhang, Z.; Xiao, J.; Laitinen, M.; Torrente-Murciano, L.; Stranks, S. D.; Rao, A.; Herz, L. M.; Scanlon, D. O.; Walsh, A.; Hoye, R. L. Z. Strong Absorption and Ultrafast Localisation in NaBiS2 Nanocrystals with Slow Charge-Carrier Recombination. Nat Commun 2022, 13 (1), 4960. https://doi.org/10.1038/s41467-022-32669-3.

The recent success of the van der Waals Sb₂(S,Se)₃ system has put low dimensional semiconductors in the spotlight to develop high efficiency photovoltaics (PV). Indeed, their performance has increased consistently, triplicating in less than a decade, and they also possess many benefits compared to other PV thin film systems, including earth-abundancy, low toxicity, thermal stability, and anisotropic crystal structure – resulting in unique electrical properties when the material is correctly oriented. However, van der Waals semiconductors beyond Sb₂(S,Se)₃ have been scarcely investigated, constituting an open field for research. Amongst this family of largely overlooked materials, antimony chalcohalides (Sb[S,Se]X, X = Br, I) are highly attractive due to their wide bandgaps in the 1.6-2.0 eV range (ideal for tandem and semi-transparent applications), customizable optical properties by using different halides or chalcogenides while retaining their quasi-1D structure, defect-tolerance, and ferroelectric properties, which might offer new opportunities (an alternative to perovskites) to improve electron-hole separation and achieve high photovoltages.

The first attempts to synthesize these materials for PV implementation have been essentially focused on low temperature (< 300^oC) and atmospheric pressure solution-based chemical routes, although PV performance of the resulting devices has been overall low. Aiming towards a more controlled, reproducible and versatile system, in this work, SbSeI has been synthesized by a novel procedure based on the selective iodination of co-evaporated Sb₂Se₃ films at pressures above 1 atm, developing an extremely uniform morphology consisting in single-crystal micro-columnar structures, which can reach lengths up to tens of microns, and which height and density can be finely tuned by adjusting the annealing temperature and pressure. Structure and crystalline quality have been characterized by advanced microscopy imaging techniques, Raman spectroscopy and X-ray diffraction. Also, Kelvin probe microscopy has been used to determine the local work function of these micro-columnar structures, revealing a work function variation among the exposed crystalline facets.

In addition to the excellent crystalline quality and low dimensional structure, SbSeI crystals show an overwhelmingly high photoluminescence (PL) signal, significantly larger than that of Sb₂Se₃ thin films, suggesting that they might have the capability of developing a remarkable PV effect. Indeed, electrical and optoelectronic measurements have been performed on micro-scale devices by transferring micro-columnar crystals to a SiO₂/Si wafer, and connecting them to Pt and Au contacts, yielding a clear diode response with the Au-Au and Au-Pt electrodes, which is greatly enhanced under illumination. Memory effect, capacitance-voltage and ferroelectric measurements have also been performed. Finally, solar cell prototypes have been prepared using a substrate configuration, resulting in high open-circuit voltages around to 600 mV.

Overall, this work presents a general study of synthesis and characterization of novel chalcohalide-based micro-scale solar cells, showing their compatibility with scalable physical vapor deposition techniques. We report the formation of highly crystalline SbSeI micro-columnar structures, showing excellent PL response, bandgap of 1.7 eV, and diode performance of micro-scale devices. These results demonstrate the potential of low dimensional chalcohalides to be implemented into innovative architectures and advanced functionalities (such as for enhanced carrier collection and light trapping), pointing the way to develop defect-tolerant and ferroelectric absorbers with anisotropic electrical properties for the next-gen PV, photo-electrocatalysis, piezo- and thermo-electric applications.

Reducing our dependence on fossil fuels for energy is important to eliminate greenhouse gas emissions. First-generation solar cell production is not sustainable, as manufacturing crystalline silicon is resource intensive. Anatase TiO₂ is a wide band-gap semiconductor oxide that has applications in third-generation solar cells and photoelectrocatalysis for water splitting and may become part of the portfolio of materials for clean energy production. One approach to improve devices based on TiO₂ thin films is to use crystallographic anisotropy to induce space-charge separation of photogenerated charge carriers. Exploiting the space-charge anisotropy intrinsic to the crystallographic directions and facets of anatase may improve both the reactivity and charge transport for photocatalytic and charge transport applications. It has been found that the (001) and (101) anatase facets are preferentially oxidizing and reducing, respectively. In addition, charge transport has lower resistance along the [001] direction. Therefore, anatase thin films for PEC should be designed with the c-axis normal to the conducting substrate and (001) facets at the surface to improve both charge transport and reactivity, respectively. Previously, we reported a synthesis for anatase thin films with [001] texture (Ichimura 2012). In this study, we examine thin film growth systematically to determine the conditions that maximize the [001] orientation and (001) facet expression of anatase thin films.

This work aims to optimize a one-pot hydrothermal synthesis of anatase TiO₂ thin films as a function of reagent concentration, pH, and temperature for photoelectrochemical applications. In this research, TiF₄ is the titanium source and HF is a structure-directing agent used to promote (001) facet growth. The TiO₂ thin films are synthesized in batches to reliably control concentration and pH. Preliminary syntheses have revealed that pH is an important factor to control grain size while temperature has a large effect on the growth of the (001) facets. The thin films are characterized by field emission scanning electron microscopy (FESEM), atomic force microscopy (AFM), and grazing incidence X-ray diffraction (GIXRD). The TiO₂ film parameters (grain size and statistics, film thickness, and extent of orientation) will be correlated with reagent concentration, pH, and temperature. Preliminary PEC measurements that employ a Ni(II) based catalyst for the oxygen evolution reaction will also be reported. Additionally, ¹⁹F NMR will be used to correlate film structure and solution composition.

The major challenges in lead chalcogenide-based optoelectronic devices processed with chemical bath deposition (CBD) are uncontrollable electronic properties of thin films. The morphologies and the chemical functions of surfaces strongly influence adhesion at interfaces within the deposited layer and surface-modified supporting substrate as well as the physical properties of the synthesized thin film. Surface treatment modifies these properties at the surface and can improve adhesion. However, the outcome depends upon the specific parameters of the processing.
In this work, we employed the surface modification strategy onto silicon-based substrates to explore how surface treatment parameters will affect the growth of lead selenide thin films in CBD growth and their opto-electrical characteristics, applicable for developing low-cost mid-wavelength IR photodetectors. As a model system, we have chosen CBD-grown PbSe thin films on oxide-free and oxidized Si substrates. To control the interfacial stability between PbSe layers and Si substrates, we utilized a surface plasma treatment of Si substrates and systematically investigated the effects of surface conditions onto the physical properties of as-grown PbSe thin films. To understand the underlying growth process of the integrated PbSe thin films, we characterized the optical and electrical properties using a conventional PL, an IR-PL, and I-V measurements. We also investigated the structural and surface morphology of the PbSe films by means of XRD and SEM/EDS.

Despite the successful history of CdS/Sb₂Se₃ heterojunction devices, cadmium toxicity, parasitic absorption from the relatively narrow CdS band gap (2.4 eV) and multiple reports of inter-diffusion at the interface forming Cd(S,Se) and Sb₂(S,Se)₃ phases present significant limitations to this device architecture. Among the options for alternative partner layers, the wide band gap (3.0-3.2 eV), non-toxic titanium dioxide (TiO₂) has demonstrated the most promise. FTO/TiO₂/Sb₂Se₃/P3HT/Au devices have achieved 7.3% PCE, compared with 10.6% for an FTO/CdS/Sb₂Se₃/spiro-OMeTAD/Au device structure. TiO₂/Sb₂Se₃ device structures deliver competitive J_sc and V_oc with these record PCE devices, however the devices are severely hindered by comparatively low FF (~52% and ~%68 respectively).

It’s typically assumed that the meta-stable anatase phase of the polymorphic TiO₂ is attained for processing temperatures <500°C, since the anatase-rutile phase transition is commonly reported to be >600°C, however there is generally a lack of rigorous analysis of how the phase of TiO₂ may influence the device performance. In this study we correlate TiO₂ phase differences attained via low-temperature processed (500°C) RF magnetron sputtered and solution spin-cast TiO₂ window layers with ‘S-Shape’ ‘kinks’ in current-voltage measurements of TiO₂/Sb₂Se₃ solar cell device structures, which act to severely reduce FF in these devices. The S-Shape current-voltage curve under illumination is attributed to an interfacial charge barrier, highlighted by calculated J_sc discrepancy between JV and EQE measurement.

This work reports approaches to distinguish between TiO₂ polymorphic phases using both surface and bulk characterisation methods. We also demonstrate the impact that phase changes may have on both the front contact/TCO and antimony chalcogenide band alignment. A device fabricated with a rutile-TiO₂ window layer (FTO/r-TiO₂/Sb₂Se₃/P3HT/Au) achieved a PCE of 6.88% and near-record 32.44 mA cm^-2 J_sc, which is comparable to established solution based TiO₂ fabrication methods which produced a highly anatase partner layer with 6.91% PCE. S-shape current-voltage curves are attributed to multi-phase TiO₂ films with a pure rutile surface and anatase bulk which generates a valley in conduction band minima alignment and drastically impedes electron extraction.

This result has implications for groups attempting to transition either from CdS partner layers to TiO₂, or even for transition from one TiO₂ deposition method to another. The work presented here shows that TiO₂ phase control is reliant on more factors than simply sintering temperature, which is significant considering TiO₂ phase is not always confirmed in Sb₂Se₃ device literature, especially in the context of our result which indicates TiO₂ phase differences alone can reduce PCE from ~7% to ~2% as a result of severe fill-factor losses in the face of this interfacial charge barrier. This is perhaps one of the reasons comparatively few groups are reporting high efficiency TiO₂/Sb₂Se₃ heterojunction devices, despite the clear limitations of CdS/Sb₂Se₃ structures.

Metal chalcohalides constitute an extended family of semiconductor materials encompassing different compositions, structural and optoelectronic properties. Indeed, since the 1960s, several chalcohalide materials have been reported to possess photoconductive, electro-optical and ferroelectric effects, and their wide variety of compositions and structures could lead to a broad bandgap range, potentially suitable for photovoltaics (PV). However, these semiconductors have been largely overlooked among the PV field, likely shadowed by the prevalence of Si in the solar cell market, and the success of other thin film technologies such as Cu(In,Ga)Se₂ and halide perovskites. Nevertheless, recent studies have reported excellent performance of several chalcohalide materials such as SbSI and SbSeI, reaching efficiencies up to 5% in a very short period of time, benefited by their quasi-1D structure, which favors a high mobility and excellent transport properties. But the chalcohalide family includes other members which have properties suitable for PV, as well as being constituted by low-toxicity and earth-abundant elements.

Among these, silver chalcohalides stand out for their crystalline structure analogous to the successful perovskite, suggesting that they could possess great tolerance to defects. Indeed, their structure corresponds to that of a perovskite, switching cation sites by anions, and anion sites by cations; hence denominated an anti-perovskite. Also, theoretical calculations have reported that silver chalcohalides have bandgaps in the 0.9-2.0 eV range, making them ideal for single-junction or tandem PV devices. Despite these promising properties, there is no published information on their implementation in solar cells, and so far, they have only been synthesized by solid-state reactions (powder) and laser ablation at high temperatures (ultra-thin films).

In this work, we present a procedure to prepare Ag₃SBr and Ag₃SI by low-temperature solution-based methodologies, using the amine-thiol solvent to dissolve Ag₂S and AgX (Br,I) precursors, followed by solution deposition by blade or spin-coating to obtain polycrystalline thin films on Mo and Al₂O₃-coated glass substrates. The structural properties of these films have been characterized by X-ray diffraction, confirming the formation of the anti-perovskite phase. Also, a series of samples have been subjected to different annealing treatments to determine the optimal synthesis conditions, whereby it has been observed that the anti-perovskite phase starts to form at 200^oC coexisting with Ag₂S (which decreases upon increasing temperature). Likewise, formation of AgBr has been detected at temperatures above 250^oC, bounding the range of optimum growth. Ag₃SBr and Ag₃SI films have also been characterized by scanning electron microscope imaging, Raman and transmission spectroscopies measurements.

Thus, this work demonstrates for the first time the viability of solution-processing methods to prepare Ag chalcohalide anti-perovskite thin films, using the amine-thiol (“alkahest”) solvent. Importantly, this methodology has reduced the synthesis temperature to 200^oC, opening the door to its implementation with different substrates and the manufacture of solar cells. In addition, this versatile chemical system tolerates different cation and anion substitutions, offering a viable approach for bandgap tuning. Strategies to incorporate Cu₂S to the precursor solutions to prepare (Cu,Ag)₃SX films and replace S by Se will also be presented and discussed.

The interest in low-dimensional materials has been regained due to the re-discovery of the high potential of these materials as absorbers for high-efficiency photovoltaic conversion. In particular, recent progresses in research on the quasi-1D material Sb₂(S,Se)₃, whose efficiency has been steadily increasing over the last decade, has attracted the interest of the scientific community to this class of materials. Among other interesting properties, these materials have the advantage of being constituted by abundant elements of low or no toxicity. Theoretical modelling has shown the very interesting properties of a new class of semiconductors based on semi-metal chalcohalides ([Bi,Sb][S,Se]X, with X = Br, I). Among them, they present unique electrical properties due to their anisotropic crystalline structure, potential defect tolerance and ferroelectricity similarly to the perovskite family. The potential applications of these materials range their implementation in tandem, semitransparent PV technologies and photocatalysis due to their bandgaps ranging from 1.6 to 2.3 eV.
In this work the properties of SbSeBr synthesized by co-evaporation of Sb₂Se₃, followed by a reactive annealing under halide atmosphere at different pressures above 1 atm that allows a better control on the incorporation of the halogen element into the crystal structure. A wide range of synthesis conditions are explored by modifying the reactive annealing pressure, temperature and duration. The first findings reveal that both the pressure and the temperature have a great effect on the size of the small irregular grains. Compared to analogous material such as SbSeI, this material requires relatively long process times to achieve a complete transformation of Sb₂Se₃ to SbSeBr. It is interesting to note that, so far, no reports on the synthesis of this particular material by means of the versatile, effective and easily scalable chemical routes are published.
The composition and crystal structure are characterized by combining X-Ray Fluoresence, X-Ray diffraction, Raman spectroscopy, and Energy-dispersive X-ray spectroscopy. The first results show that by the method developed in our lab it is possible to obtain a single-phase material with a very high degree of both purity and crystallinity. An identification of the main Raman modes will also be presented.
The absorber layers have been used to fabricate solar cells, reporting the first working devices for this material with efficiencies above 0.5% and with a very encouraging Voc (>600 mV). The main current limiting factor for the efficiency of the devices is the complex morphology of the SbSeBr layers, suggesting that a proper optimization of the novel synthesis process and an appropriate choice of Hole Transport Layer and Electron Transport Layer can boost the efficiency of these promising materials.
Finally, once the processing conditions are fully optimized, solar cell devices will be completed in both substrate and superstrate configurations, opening promising pathways for the development of highly efficient solar cells based on quasi-1D materials by a proper selection of device configuration, transport layers and material synthesis conditions.

Organic solar cells (OSCs) have emerged as promising photovoltaic technology because of low-cost, lightweight, flexibility, and transparency. Recently, the power conversion efficiencies (PCEs) of OSCs have exceeded 18% with the development of non-fullerene acceptors (NFAs). Reported Y6 which has an A-DA’D-A structure, various Y-series derivatives have been designed. However, a trade-off between an open-circuit voltage (V_OC) and a short-circuit current density (J_SC) remains a problem even though the PCEs of OSCs based on the Y-series derivatives have increased rapidly. Hence, it is crucial to finely tune the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels of photoactive materials for balance between V_OC and J_SC. In this work, we report efficient and stable OSCs based on a new asymmetric NFA, IPC-BEH-IC2F. This asymmetric NFA is comprised of a weak electron-donating core dithienothiophen[3,2-b]-pyrrolobenzothiadiazole (BEH) and two kinds of end groups having strong electron-accepting ability 9H-indeno[1,2-b]pyrazine-2,3-dicarbonitrile (IPC) with tricyclic fused ring and 2-(5,6-difluoro-3-oxo-2,3-dihydro-1H-inden-1-ylidene)malononitrile (IC2F) which used as an end group in Y6. The IC2F-BEH-IC2F (or Y6) and IPC-BEH-IPC having symmetric structures were also characterized to compare with asymmetric IPC-BEH-IC2F. The IPC moiety significantly affects the optical properties and electronic structures of the NFAs. The asymmetric IPC-BEH-IC2F shows the highest extinction coefficient among the three NFAs due to its strong dipole moment and highly crystalline feature. Its HOMO and LUMO energy levels are successfully tuned between those of the IC2F-BEH-IC2F and IPC-BEH-IPC through the asymmetric structure. Besides, the IPC moiety promotes strong intermolecular π-π stacking due to their tricyclic fused system, which is beneficial for charge transport. Inverted type OSCs based on polymer:NFA blends were fabricated. Among all these polymer:NFA based OSCs, the asymmetric IPC-BEH-IC2F based OSC blended with PBDB-T showed a balance between V_OC and J_SC, leading to the highest PCE of 12.70% with a high V_OC of 0.85 V and a J_SC of 22.29 mA cm^-2. In addition, the PBDB-T:IPC-BEH-IC2F based OSC also exhibited the best long-term stability under ambient conditions due to the strongly interacting IPC moiety, which forms the densely packed morphology and prevents penetration of Ag atom into the photoactive layer. These results demonstrate that using asymmetric NFAs can be useful for improving the photovoltaic performance and stability of OSCs.

To successfully develop a regioregular polymer, poly[4,8-bis(5-(2-hexyldecyl)thiophen-2-yl)benzo[1,2-b:4,5-b’]dithiophene][5,5’-bis(7-(4-(2-butyloctyl)thiophen-2-yl)-6-fluorobenzo[c][1,2,5]thiadiazol-4-yl)-2,2’-bithiophene] (PDBD-FBT), symmetric monomer synthesized in high yield by tin homo-coupling reactions. PDBD-FBT is suitable as donor material in organic photovoltaics (OPVs) because it shows high crystallinity and strong face-on packing properties. These properties were amplified by thermal annealing (TA). It causes power conversion efficiency (PCE) enhancement in PDBD-FBT-based OPVs. Using PDBD-FBT as polymer donor and 2,2'-((2Z,2'Z)-((12,13-bis(2-heptylundecyl)-3,9-diundecyl-12,13-dihydro-[1,2,5]thiadiazolo[3,4-e]thieno[2",3’':4’,5']thieno[2',3':4,5]pyrrolo[3,2-g]thieno[2',3':4,5]thieno[3,2-b]indole-2,10-diyl)bis(methanylylidene))bis(5,6-difluoro-3-oxo-2,3-dihydro-1H-indene-2,1-diylidene))dimalononitrile as electron acceptor, PCE of 7.91% was achieved without any additive and TA treatment at optimized film thickness approximately 100nm. After TA treatment, PCE of 12.53% was achieved with 58% increase compared with the reference devices. Owing to the strong crystallinities, trap-assisted recombination occurs by excessively formed grain boundaries; however, enhanced exciton dissociation probability covers these drawbacks. Even thick film condition approximately 340nm, this tendency is more pronounced (73% of PCE enhancement by TA treatment is observed). Although thick film’s OPVs are difficult to achieve high efficiency, it shows similar efficiency as the device with optimized film thickness. Our study demonstrates that it is possible to manufacture thickness-insensitive OPVs based on regioregular polymers with strong crystallinity and face-on characteristics, thereby providing a solution to the thickness variation of large-area organic solar cell modules.

In Organic solar cells, Y6 derivatives have a good candidate due to dominant face-on orientation, high electron mobility, and low energy loss with high electroluminescent quantum efficiency. However, Y6 derivatives have a long and complicated synthetic procedure to synthesize dithieno[2'',3'':4',5']thieno[2',3':4,5]pyrrolo[3,2-e:2',3'-g] [2,1,3]benzothiadiazole (BTP) core. So, we designed two new Y6 derivatives (YBO-2O/-FO) with a simplified synthetic procedure by moving solubilizing alkyl substituents on the BTP to the terminal (3-oxo-2,3-dihydro-1H-inden-1-ylidene)malononitrile (INCN) groups. We investigated the intermolecular packing characteristics when the bulky alkyl side chain on the core of Y6 is removed through molecular dynamics (MD) simulation. YBO-2O/-FO are more formed in the core-core and terminal-terminal (CC-TT) with tighter packing structure than Y6 which coincides with GIWAXS (tighter packing than Y6) and SCLC (faster electron mobility than Y6). The HOMO and LUMO energy level was upshifted by the electron-donating terminal group (INCNO2). YBO-FO has an adequate HOMO energy level with that of the polymer donor PM6 which is a good hole transfer rate at the interface between donor and acceptor. In blend films, the hole transfer from YBO-2O/-FO to PM6 is investigated by transient absorption spectroscopy (TAS), demonstrating efficient hole transfer from YBO-FO to PM6. Despite the simplified synthesis, YBO-FO shows a power conversion efficiency of 15.01% similar to Y6.

The so-called supramolecular chirality is the transfer of the chirality arising from a single chiral object to higher-order supramolecular arrangements. This is particularly interesting in the preparation of organic photovoltaic devices, since thin films show an intimate dependence between efficiencies and their supramolecular arrangement. This gives a predominant role to the correlation between (chiral ) supramolecular arrangement and the physical parameters of the devices (FF, Voc, Isc, etc).
Moreover, chirality enables the possibility to use Circular Dichroism (CD) spectroscopy which is the preferred technique to study optical activity, with the advantage that CD spectra are very sensitive to changes in supramolecular structure compared with other spectroscopic techniques. For this reason, this branch of spectroscopy may give unique information about conformational changes in natural and synthetic materials. Our project deals with the synthesis of new enantiomerically pure DPP dyes with high absorbance and great photostability that find application as active layer in photovoltaic devices in blend with different acceptors
The knowledge of the correlation between the supramolecular structure and the electron transfer efficiency of DPPs is one of the main goals of the project. In this direction, our work was enriched by CD studies performed in the solid state at the Diamond Light Source Synchrotron at the B23 beamline. Here thin films of the chiral materials are analyzed to obtain a Mueller Matrix Polarimeter (MMP) which allows the possibility to obtain the real CD spectrum in the solid state without linear anisotropies, like linear dichroism or linear birefringence that may mislead the interpretations. Moreover, optical microscopy and AFM analysis were performed with the thin films prepared and modified under a variety of conditions (like spin coating, blade coating, and annealing at different temperatures).
In conclusion, the fabrication of photovoltaic devices has been performed taking into account the knowledge obtained through solid state CD and from the Muller Matrix.

The power conversion efficiencies (PCEs) of small molecule acceptors (SMAs)-based organic solar cells (OSCs) have remarkably increased in recent years, but their thermal and long-term stabilities should be significantly enhanced for the commercialization. In addition, relatively lower open-circuit voltages (V_oc) of OSCs compared to other types of solar cells (e.g., perovskite solar cells) should be addressed to further improve their PCEs. Here, we demonstrate that dimerizing a SMA with a properly-designed conjugated linker resolves the critical issues associated with the stability and V_oc of OSCs. The dimerized SMA (DYBO) connected by a benzodithiophene (BDT) conjugated linker affords OSCs with excellent PCEs (> 18%), which outperform those of the reference OSCs (PCE = ~17%) based on a monomer-type SMA (MYBO) with the identical SMA building block. The incorporation of the electron-donating BDT linker in DYBO effectively upshifts the lowest unoccupied molecular orbital energy level and reduces the voltage loss, synergistically increasing the V_oc of the devices. Importantly, the DYBO-based OSCs have excellent thermal and photo stabilities. For example, the DYBO-based OSCs retain more than 80% of the initial PCE even after 6000 hr of thermal exposure at 100 °C, whereas the MYBO-based OSCs sharply degrade their PCE to ~80% of the initial value after only 36 hr exposure. The main origins of the improved stabilities in the DYBO-based OSCs are (1) the significantly increased glass transition temperature (T_g) from 80 to 179 °C by the dimerization, which stabilizes the blend morphology during thermal stresses; (2) the improved miscibility of DYBO with a BDT-based polymer donor. Thus, we highlight the importance of the molecular design of dimerized SMAs for achieving OSCs with excellent PCEs and stabilities.

Perovskite microcells have a great potential to be applied to diverse types of optoelectronic devices including light-emitting diodes, photodetectors, and solar cells. Although several perovskite fabrication methods have been researched, perovskite microcells without a significant efficiency drop during the patterning and fabrication process could not be developed yet. We herein report the fabrication of high-efficiency perovskite microcells using swelling-induced crack propagation and the application of the microcells to colored solar windows. The key procedure is a swelling-induced lift-off process that leads to patterned perovskite films with high-quality interfaces. Thus, a power conversion efficiency (PCE) of 20.1 % could be achieved with the perovskite microcell, which is nearly same as the PCE of our unpatterned perovskite photovoltaic device (PV). The semi-transparent PV based on microcells exhibited a light utilization efficiency of 4.67 and a color rendering index of 97.5 %. The metal–insulator–metal structure deposited on the semi-transparent PV enabled to fabricate solar windows with vivid colors and high color purity.

On account of the excellent photovoltaic properties (e.g. tunable bandgap, long carrier diffusion length, and high absorption coefficient) and solution processability of the perovskite, a number of studies have been conducted on the perovskite as a next-generation solar cell. The rapid increase of the power conversion efficiency (PCE) of the perovskite solar cells (PSCs) has been achieved in the last decade, however, it has been implemented only in the conventional (n-i-p) structure of the PSCs. The inverted (p-i-n) structure, which has low-temperature processability, negligible hysteresis effects, and better device stability than n-i-p structure, is in the spotlight. Furthermore, p-i-n structure is suitable for fabricating tandem solar cells that can overcome the Shockley-Queisser limit (S-Q limit) of single junction solar cell.
In order to catch up with the efficiency of the n-i-p structure, numerous efforts have been made in the p-i-n structure. Due to the ionic nature of perovskite, which generates defect states, a typical method is to use an additive to passivate the defects in the perovskite layer. The perovskite crystal can be passivated during crystal growth process with incorporated additive, or prevent halide ion migration which leads to decrease efficiency of solar cells and device stability with placing additive material between perovskite and electron transporting layer (ETL).
A new family of 2D materials, called MXene, has emerged as a great candidate for additives in PSCs replacing other materials due to its outstanding electrical conductivity, tunable optical properties, and robust physical, chemical properties. Since MXene’s surface has negatively charged groups (e.g. -OH, -O, -F, and -Cl), it is reasonable to expect that interaction of MXene with Pb²⁺ ion in perovskite not only passivates the defect sites of the perovskite but also affects to the crystal growth of perovskite. Moreover, MXene would decrease the charge transfer resistance in PSCs by its high electrical conductivity, increase the thermal stability by its good thermal conductivity. Lastly, incorporation of MXene in the perovskite cause the change of work function of perovskite to improve the device performances.
Herein, we demonstrate the affect of MXene as an additive in the p-i-n PSCs by engineering two different incorporating methods. We compared the difference in the mechanism of each crystal growth when MXene is added by two different methods, following compare the differences in morphology of perovskite and device performance according to each method. Consequently, the short-circuit current density (Jsc) and open-circuit voltage (Voc) are greatly improved, so that obtain over 23% of PCE.

Organic-inorganic hybrid perovskite solar cells (OIHPs) have attracted the PV community due to low cost and high power conversion efficiency, but still have suffered from poor stability. In order to shield perovskite solar cells (PSCs) from extrinsic degradation factors such as oxygen and moisture and assure long-term stability, effective encapsulation technology is indispensable. Here we designed a facile process to build glass-glass encapsulated semitransparent PSCs by laminating two half cells where a perovskite layer is formed on a hole transport layer (HTL, p-type) and an electron transport layer (ETL, n-type), respectively, with a transparent conducting oxide (TCO) glass substrate. By laminating the half cells through a thermocompression bonding process, the device is self-encapsulated during manufacturing. Soft mechanical characteristics and atomic migration nature of the perovskites enable the lamination between the perovskite layers with minimizing voids and good adhesion properties. During the lamination process, surface of the perovskite transformed into bulk with enlarged grains, and the interfaces has much smoother and denser morphology. Through surface iodine extraction experiments and TRPL analysis, it was confirmed that defects and trap density decreased in the laminated perovskite. The thermcompression increases hydrophobicity on interfaces of the perovskite. Therefore, the laminated perovskite has higher stability against moisture. The self-encapsulated semitransparent PSCs with a wide bandgap perovskite (E_g ~ 1.67eV) is performed with the PCE of 17.24 %, and thermal stability at 85 °C is secured for over 3,000 hours.

The binary chalcogenide-based thin-film solar cells (TFSCs) have attracted great attention due to their low-cost and nontoxic constituents, simple structure, and ability to exhibit high power-conversion efficiency (PCE). The orthorhombic tin selenide (a-SnSe), a binary chalcogenide, is a potential candidate for the low-cost TFSCs due to its appropriate optical bandgap, high optical absorption coefficient (α〉10⁵ cm^-1) in the visible region, intrinsic p-type conductivity, and high hole mobility (200 cm²V^-1s^-1). However, the reported PCE for SnSe TFSCs is significantly low (less than 1%) in most cases due to poor crystallinity, small grains, and a large number of defects. Herein, we report the highest cell efficiency of 2.51% and a high short-circuit current density of 28.07 mA cm^-2 for a-SnSe TFSCs grown via vapor-transport-deposition (VTD). The effect of evaporation temperature and duration was studied on the performance of SnSe TFSCs. The grain size and surface roughness of the SnSe thin films were mainly affected by the evaporation temperature and duration and were found to be the main parameters to influence the shunt properties of the device. Significantly large shunt losses are detected in the case of both small and extremely large grains. The shunt losses for SnSe thin film with small grains are associated with high grain-boundary scattering. The presence of extremely large grains results in increased surface roughness of the SnSe thin film, which causes nonuniform deposition of the CdS buffer layer and, consequently, higher shunt losses. The SnSe thin film with moderate-sized grains and inferior surface roughness exhibits improved shunt properties owing to uniform deposition of the CdS buffer layer and subsequent layers and significantly improved the device performance. Detailed studies on the morphological and structural properties of the orthorhombic VTD-SnSe thin films and their impact on the photovoltaic device performance will be discussed.

We fabricated a hole transport layer (HTL) using transfer printing to fabricate high efficiency PbS quantum dot solar cell with minimized hysteresis. We applied PbS-EDT(1,2-ethanedithiol) as an HTL and controlled the adhesive properties between the donor substrate and the HTL through the UVO process in order to effectively transfer the HTL to the PbS layer, which enabled multilayer transfer printing as a result.
During the transfer printing process, a densly packed gradient solid-solution layer without pinholes was created at the interface with the existing PbS layer to reduce the hysteresis phenomenon, and to improve the power conversion efficiency to about 10% according to the HTL thickness optimization.
In addition, the use of the EDT solvent can be remarkably reduced by using the transfer printing when forming the PbS-EDT layer. This can reduce the amount of organic solvent consumed in the process and improve the potential of environmentally friendly processes.

Chalcogenide semiconductors, especially 2D transition metal dichalcogenides, are extensively studied owing to their intriguing physical properties for electronic and photonic applications. However, in the past few years, ternary chalcogenide semiconductors such as chalcogenides perovskites have gained substantial interest as a new class of semiconductors due to their tunable chemistry, structure, and opto-electronic properties. For instance, BaZrS₃ (BZS) (band gap = 1.9 eV), a prototypical chalcogenide perovskite, is promising as an ultra-thin absorber in tandem solar cells^1-3, whereas BaTiS₃ (BTS) (band gap = 0.27 eV) exhibits a giant birefringence and large optical anisotropy in the infrared region⁴. Growth of thin-film chalcogenide perovskites is a critical step to enable investigations into their fundamental properties and also device applications. But the thin-film growth of chalcogenide perovskites is challenging due to a large mismatch in the vapor pressure of the cations and chalcogens, corrosive and reactive nature of most chalcogen precursors, the propensity to oxidize easily in the presence of oxygen at high temperatures and the lack of suitable non-reactive substrate surfaces for epitaxial growth.

Recently, we have demonstrated direct, single-step, epitaxial growth of BZS¹ and BTS⁵ thin films on oxide substrates by pulsed laser deposition (PLD) using Argon-H₂S background gas. H₂S is a toxic, hazardous, and flammable gas and reacts with many materials, and the ionized Argon and/or H₂S lead to significant degradation of the crystallinity of both the surface and film at the high growth temperatures. Here, we report an alternative hybrid PLD approach using organo-sulfur precursors as sulfurizing agents to grow chalcogenide semiconductors. To demonstrate the efficacy of this approach, we have demonstrated epitaxial growth of 3D binary chalcogenides such as BaS and SrS (wide band gap) and 2D binary chalcogenides such as TiS₂ (metallic) and ZrS₂ (semiconducting). The potential of these binary sulfides as suitable candidates for transparent and conducting layers in chalcogenide-based photovoltaic devices will be discussed. We have also realized epitaxial BZS and BTS films using hybrid PLD and a comparison of this approach to PLD using Ar-H₂S will be provided. Further, we will also discuss our future growth efforts and the applications of this novel growth method.

References:

1. M. Surendran, H. Chen, B. Zhao, A. S. Thind, S. Singh, T. Orvis, H. Zhao, J.-K. Han, H. Htoon, M. Kawasaki, R. Mishra and J. Ravichandran, Chemistry of Materials 33 (18), 7457-7464 (2021).
2. Y. Nishigaki, T. Nagai, M. Nishiwaki, T. Aizawa, M. Kozawa, K. Hanzawa, Y. Kato, H. Sai, H. Hiramatsu, H. Hosono and H. Fujiwara, Solar RRL 4 (5) (2020).
3. S. Niu, H. Huyan, Y. Liu, M. Yeung, K. Ye, L. Blankemeier, T. Orvis, D. Sarkar, D. J. Singh, R. Kapadia and J. Ravichandran, Adv Mater 29 (9) (2017).
4. S. Niu, G. Joe, H. Zhao, Y. Zhou, T. Orvis, H. Huyan, J. Salman, K. Mahalingam, B. Urwin, J. Wu, Y. Liu, T. E. Tiwald, S. B. Cronin, B. M. Howe, M. Mecklenburg, R. Haiges, D. J. Singh, H. Wang, M. A. Kats and J. Ravichandran, Nature Photonics 12 (7), 392-396 (2018).
5. M. Surendran, B. Zhao, G. Ren, S. Singh, A. Avishai, H. Chen, J.-K. Han, M. Kawasaki, R. Mishra and J. Ravichandran, Journal of Materials Research 37 (21), 3481-3490 (2022).

Current-voltage (J-V) measurements under illumination are essential for studying solar cells. They directly provides fundamental parameters of the solar cell, as the short-circuit current (J_SC), the open-circuit voltage (V_OC), and the fill-factor (FF); but a deeper investigation of transport phenomena along the fourth quadrant is hampered by the tight linear scale, limited between zero to ~1 V. In this work, we present transient measurements carried out in an organic solar cell (ITO/PEDOT:PSS/PTB7-Th:PC₇₁BM/Ca/Al), which allows to analyze J-V curves, by extending the scale from a linear to a logarithm scale. For this, we used a circuit similar to that used in transient measurements, by replacing the voltage source by a variable load resistance R_L (from 50Ω to 1MΩ). A voltage transient DV(t) and its time decay t are measured over each R_L, and t-R_L curves were obtained, keeping the device under different intensities of sunlight. Each t-R_L curve is divided in three distinct zones: one in which the extraction dominates the photocurrent; the second where a competition between extraction and non-geminate recombination is established; and the third, dominated by recombination and diffusion. An equivalent circuit is used to analyze the t-R_L curves, in which the diode resistance plays a relevant role, especially in the region close to V_OC. Not only extraction and recombination mechanisms can be better analyzed through this new approach, but also the distinction between drift and diffusion transports.

A highly and evenly transparent conductive electrode is fabricated on glass substrate thanks to the ability to produce stable continuous ultra-thin silver films, the computational global search of optical properties on ITO/Ag/ITO (IAI) sandwich structures using the transfer matrix method (TMM), and annealing process in the air at high temperature. First, ion beam treatment and aluminum cap layer enable the depositing of stable ultra-thin silver films at 6 and 7 nm. Particularly, ion beam treatment helps to prepare a surface with better wettability toward silver, and the aluminum cap layer help to protect silver film by impeding the movement of silver surface atoms. Secondly, the film thicknesses are then used as the inputs for the simulation of the ITO/Ag/ITO sandwich structure and we found the optimum design of ITO (45 nm)/ Ag (6-7 nm)/ITO (50 nm)/ Glass. Finally, the stable silver film allows further processing of annealing at high temperatures (200C) in the vacuum and in the air, which is not possible for a stand-alone silver film. The resulting 7 nm film on glass substrate has a sheet resistance of 10 Ohm/ sq. an average transmittance in the range of 400-800 nm of 89%. This process can be implemented in other substrates as well as other transparent conductive oxides such as SnO₂, AZO, TTO, and TiO₂. Therefore, a couple of TCOs were taken into consideration for a similar sandwich structure to compare the optical performance showing potential replacements for ITO in optoelectronic applications in the circumstance that indium is getting costly.

In the pursuit of lowering the cost and increasing the penetration of solar energy technologies, a family of inorganic thin-film photovoltaic materials is emerging. The oldest member of this family, selenium, is experiencing renewed interest due to its high bandgap of 1.95 eV in its trigonal phase, as well as its irresistible monoatomic simplicity. The high bandgap makes selenium a potential partner in tandem photovoltaic devices featuring e.g., silicon as the lower bandgap photoabsorber, but high-efficiency selenium thin-film solar cells have yet to be realized.
We present selenium thin-film solar cells with power conversion efficiencies exceeding 5% and following a novel optimization of the carrier-selective contacts, we demonstrate a record open-circuit voltage of 0.99V and highly encouraging fill-factors beyond 60%. These state-of-the-art devices are investigated in a combined experimental and first-principles study to reveal the origin of the still-significant photovoltaic losses [1]. In view of these results, we set forth strategies for continued improvements to the photovoltaic performance by means of non-equilibrium growth and defect engineering.

[1] Nielsen et al., J. Mater. Chem. A, (under review)

The emergence of new PV applications in the society requires the design of materials and devices with a different set of properties. At this scale, for a new photovoltaic (PV) technology is not sufficient to be only competitive with the Si and CdTe technologies in efficiency and reliability but one should also rely on green, environmentally friendly, and earth-abundant materials. An emerging class of highly promising PV materials currently under widespread investigation in the PV community are the inorganic antimony-and bismuth-based chalcogenide compounds. The excellent intrinsic material properties of these compounds allowed to rapidly approach 8-10% conversion efficiency, opening a new chapter in PV research, full of new possibilities but also scientific challenges. This talk will discuss the latest achievements in bismuth- and antimony-based thin film solar cell technology with the main emphasis on Sb₂Se₃ and Bi₂S₃ compounds. I will summarize the knowns and unknowns of their defect chemistry, including highlights of unique optoelectronic characteristics that are not yet fully explained. The discussion will highlight the progress achieved in our group, toward growth of high-quality Sb₂Se₃ and Bi₂S₃ absorber films and solar cells by rapid, high-volume, and in-line close-spaced sublimation and vapor transport deposition techniques. Special emphasis will be put on the key processing strategies to optimize absorber material properties (doping and alloying), understanding of buried interfaces and push the boundaries of understanding and performance. A thorough look at the formation chemistry and shed light on some of the synthesis and doping challenges, potential doping elements, the impact this has on carrier density and the impact on solar cell performance.

The exploration of novel absorber materials composed by earth-abundant and low-toxicity elements is important for thin-film solar cells. Among these materials, CuSbSe₂ is promising because of its suitable optical and electrical properties such as ≈ 1.1 eV bandgap and high absorption coefficient. However, the practical application of the reported hydrazine solution process is limited by the hazardous nature of hydrazine, and more detailed investigations into charge-carrier transport properties of CuSbSe₂ is needed. In this study, a novel thiol-amine solution processing route is used to fabricate CuSbSe₂ thin films. The fundamental properties of CuSbSe₂ thin films are investigated, including phase purity, absorption coefficient and bandgap, to demonstrate the feasibility of the thiol-amine route. Next, the charge-carrier kinetics are studied by transient absorption (TA) and optical-pump terahertz-probe (OPTP) spectroscopy, which reveal that, in contrast to other heavy pnictogen-chalcogenide compounds studied recently [1], CuSbSe₂ avoids carrier localisation. To further study the carrier-phonon coupling and carrier transport properties, the temperature-dependence of the mobility is determined through Hall-effect measurements, and compared with DFT calculations to show that Fröhlich coupling dominates in CuSbSe₂. The coupling to acoustic phonons is weak, owing to low deformation potentials. The cause of the low deformation potential is studied by investigating how the lattice distorts along the dominant phonon modes. The combined experimental-computational studies shown in this work lead to new insights into how materials avoiding carrier localisation could be designed in the future.

References
[1] Huang, Kavanagh, et al., Nat. Commun., 13, 4960 (2022)

Coupling a short wavelength infrared (SWIR) PV device with a bandgap of 0.6-0.7 eV to tandem PV technologies can improve overall power conversion efficiency (PCE) by 5-11%. Alternatively, a SWIR device can be used in a thermophotovoltaic system and (ideally) furnish a PCE > 50%. However, there is a limited variety of low-cost light absorbers for the SWIR region. Antimony trichalcogenides, such as Sb₂Se₃, exhibit a large absorption coefficient (~10⁵) and a long diffusion length (~1.7 µm). Made of earth-abundant materials and by rapid and low-cost deposition methods, Sb₂Se₃ solar cells with PCE > 10% were recently reported. Three other qualities make Antimony trichalcogenides promising SWIR-PV materials: i) Due to the unique quasi-one-dimensional crystal structure, certain grain boundaries in Sb₂Se₃ thin films are expected to be electronically benign. ii) due to the antibonding states at the top of the valence band, Sb₂Se₃ is expected to be (point) defect tolerant. iii) The bandgap of Sb₂Se₃ can be tuned to the SWIR region by alloying with Bi. Here we report a method for making thin films and solar cells of Bi-alloyed PV. By using closed space sublimation, we successfully incorporated Bi in Sb₂Se₃ up to 40%At. without forming secondary phases. The absorption edge extended from 1050 to 1370 nm. Proof-of-concept solar cells will also be presented.

Antimony sulphide (Sb₂S₃) is an emerging semiconductor material with excellent promise for future photovoltaic (PV) technologies owing to its superior optoelectronic properties coupled with earth-abundant and environmentally friendly elements. The relatively wide bandgap of Sb₂S₃ (1.7 eV) makes it wiser choice for application in semi-transparent PV and tandem solar cells. Although the material has potential even for single junction cells with a Shockley-Queisser efficiency limit of ≈ 28 % but the open circuit voltage deficiencies arising from defects and/or self-trapping of photogenerated carriers by lattice deformation limit the maximum efficiency to ≈ 16 %. Therefore, despite numerous attempts and various device architectures, the efficiency of Sb₂S₃ solar cells is still considerably low with an achieved maximum of 8 %, even for an absorber thickness ranging from 200 nm to over 2 µm. And, most of the high efficiency devices (7-8 %), which utilize an absorber thickness of at least 200 nm, are fabricated by either spin coating, chemical bath deposition (CBD) or vacuum deposition techniques. While CBD has the promise for large area upscaling, its throughput is relatively low and requires longer deposition time (≈ 4 hours). In this context, the study herein demonstrates 6.2 % efficient Sb₂S₃ solar cells from an absorber thickness of ≈ 100 nm deposited using industrially benign and scalable ultrasonic spray pyrolysis (USP) with high throughput (deposition time ≈ 20 min). The solar cells are apt for semi-transparent applications with an average visible transmittance of 27 % (400-800 nm) for the device stack with P3HT as hole transport material without the back metal contact. To the best of authors’ knowledge, the efficiency of 6.2 % is a reasonable record achieved for pristine 100 nm Sb₂S₃ absorbers devoid of any doping or post deposition treatment. The individual layers and solar cells are comprehensively characterized using advanced characterization techniques like photoelectron emission spectroscopy, photoluminescence, TRPL, and DLTS to understand the band energetics, device physics, defects and interfaces. The USP process parameters towards obtaining high quality absorbers leading to record efficiencies will be discussed in detail while emphasizing the role of defects on the performance of Sb₂S₃ solar cells. In addition, the roadmap towards complete semi-transparent Sb₂S₃ solar cells by replacing back metal contact, which is being pursued, will be presented.

Understanding the spatial dynamics of nanoscale exciton energy transport beyond the temporal decay is at the core of photosynthesis and essential to provide a better framework for further improvements of nanostructured optoelectronic devices, such as solar cells. The diffusion coefficient (D) of photovoltaic films has so far only been determined indirectly, from transient singlet-singlet annihilation (SSA) experiments.
Here, we present the full picture of the exciton distribution dynamics, adding the spatial domain to the temporal one, by spatio-temporally resolved photo-luminescence microscopy. In this way, we directly track diffusion and we are able to decouple the real spatial broadening from its overestimation given by SSA. We measured the diffusion coefficient, D = 0.017 ± 0.002 cm²/s, of the non-fullerene electron acceptor Y6, which combined with an exciton lifetime of τ = 840 ps, gives a Y6 film diffusion length of L = (Dτ)^1/2 ≈ 35 nm. Thus, we provide an essential tool that enables a direct and free-of-artifacts determination of diffusion coefficients, which we expect to be at the core of further methodical studies on exciton dynamics in energy materials.

Singlet fission (SF) is a spin-allowed carrier multiplication process in organic systems where a photo-generated spin-0 singlet exciton is converted into two low energy spin-1 triplet excitons, allowing quantum efficiencies up to 200% and offering the potential to surpass the Shockley-Queisser limit in conventional photovoltaics [1]. Though recent years have witnessed considerable efforts in understanding the fundamental principles and mechanisms of SF, most studies have been based on acene systems having poor photochemical stability and low extinction coefficients [2].

Here, I will introduce a new family of organic chromophores called Pechmann dyes designed based on the Baird's rule of aromaticity [3], undergoing SF. I will discuss their molecular photophysics in thin films prepared by thermal evaporation and present the sub-100 picosecond triplet exciton formation evidenced from ultrafast transient absorption and electron spin resonance spectroscopic studies [4]. Furthermore, using Pechmann dyes as a model system, we theoretically explore the role of excited state aromaticity in singlet fission and present a comprehensive design strategy for potential SF-capable systems. This work shed light into the relationship between structure, aromaticity, and singlet-triplet energy gap as a benchmark for designing novel, stable, and efficient SF candidates with better structural tunability for a new generation of photovoltaics.

References:
[1] A. Rao, R. H. Friend, Nat. Rev. Mater. 2, 1–12 (2017).
[2] P. J. Budden et al., Nat. Commun. 12, 1527 (2021).
[3] K. J. Fallon et al., J. Am. Chem. Soc. 141, 13867–13876 (2019).
[4] A. V. Girija et al., manuscript in submission (2022).

The replacement of fullerenes with non-fullerene acceptors (NFAs) in the fabrication of organic solar cell (OSC) devices in recent years has paved the way to achieve power conversion efficiencies approaching 20% for solution processed OSCs. This has been enabled by a few factors, key among which are stronger and spectrally varied absorption from NFAs compared to their fullerene counterparts and a significant reduction in the voltage losses of these devices. Despite the clear importance of NFAs in the future of OSCs, there are few studies on vacuum thermally evaporated (VTE) NFA OSCs. The VTE manufacturing process has already been commercially tested and is widely used in many established industries, notably the OLED industry which is responsible for most smartphone displays today. NFAs which can be used in VTE OSCs therefore represent a key milestone to successful commercialisation of the OSC field.

In this study we present results showcasing the suitability of the NFA BTIC-H. This NFA was first used in solution processed polymer system, but its size and structure make it suitable for VTE. We showcase device results, including current density (JV), sensitive external quantum efficiency (sEQE), and transfer matrix modelling (TMM), of various combinations of this NFA with popular small molecule donors. By using spectroscopic ellipsometry (SE) and ambient photoemission spectroscopy (APS) we see how the choice of donor molecule affects the optoelectronic environment within the devices. We use both planar heterojunction (PHJ) and bulk heterojunction (BHJ) architectures to better characterise the photophysical behaviour of these donor acceptor combinations. The ease of use of both these architectures is unique to the VTE process, as constructing planar heterojunctions with solution processing methods is difficult. Optical models calculated from SE are also compared with grazing incidence wide angle X-ray scattering (GIWAXS) to better understand the impact of morphology on the behaviour of these devices. Results are compared and contrasted with equivalent fullerene devices to highlight what has been improved on, and what still needs to be improved on compared to fullerene acceptors. Key results from this study show that when used with the donor molecule DCV5T-Me(3,3), BTIC-H devices have a similar Voc and Jsc to equivalent fullerene devices, but only when comparing PHJ architectures. Upon blending, the fill factor is reduced as charge collection issues arise in the DCV5T-Me(3,3):BTIC-H BHJ. GIWAXS results show clear and distinct out-of-plane features for BTIC-H neat films, which is comparable to the layer in a PHJ device, but these are strongly suppressed in a blend film, which is comparable to the layer in a BHJ device, suggesting that charge collection pathways are not properly formed in the BTIC-H: donor blend. Methods to encourage a more favourable microstructure, such as substrate heating during deposition, are discussed.

This study also presents a framework to appropriately test new NFA molecules for VTE OSCs, such as an appropriate screening criteria, as well as reveals some insight into additional design criteria for VTE NFAs. It is our hope that this work will contribute to the current understanding of the behaviour and role of NFAs in OSCs as well as allow for the fabrication of more efficient VTE OSCs.

Perovskite solar cells (PSCs) have not yet been commercialized due to their low stability against external factors such as light, moisture, and heat. The migration of iodide ions out of the perovskite is also one of the critical mechanisms of the poor stability of PSCs. In order to inhibit the degradation of PSCs due to external factors and ion migration, it is necessary to form an inorganic buffer layer by surface treatment on the surface of the perovskite. We introduced an inorganic PbS buffer layer by spin coating of Na₂S solution on surface of the perovskite. A PbS buffer layer prevent direct contact of moisture to perovskite and diffusion of the iodide ions. As a result, the PSCs with the PbS buffer layer maintained higher efficiency than those without PbS buffer layer under light and in the moist air and 85 °C heat test. In addition, PbS buffer layer improves charge transfer at the Perovskite/Spiro-OMeTAD interface, so increases power conversion efficiency of PSCs.

Metal oxides (MO) are sturdy materials; therefore, MO selective contacts can substantially improve PSCs' stability. However, MO surfaces are reactive, and thus they tend to react at the interface with HaPs and harm both stability and reproducibility of HaP-based electronic devices. ZnO is an extreme case of a reactive MO due to its highly alkaline surface, which readily reacts with HaP. We used steady-state measurements to understand the passivation of ZnO-HaP interface by Aluminum oxynitride (AlON) sub 2 nm layer deposited by atomic layer deposition (ALD). The effect of AlON passivation on the diffusion length of HaP under steady-state conditions, mimic real-life operation conditions for most solar cells. Yet, steady-state measurements require as high as possible photoconductivity (σ_ph), which is typically restricted to the absorber. To enable a steady-state charge transport measurement, we used ALD to grow a conformal, ultra-thin (∼4 nm) ZnO electron transport layer that is laterally insulating due to its thickness. The formation of ultra-thin continuous inorganic films enables steady-state measurements on stacks of layers. We show that the presence of the AlON layer prevents HaP degradation caused by the interaction with the ZnO layer, improves the HaP σ_ph, and doubles the HaP carrier diffusion lengths.

Copper-based oxides (Cu_xBi_yO_z) with a p-type character comprise wide band gap. However, these compounds suffer from the phenomenon of photocorrosion (leading to self-oxidation or self-reduction) and insufficient light absorption, which limits their usage and long-term durability for different applications. To address these limitation, copper-based chalcogenide has shown a prospective alternative. Copper-bismuth chalcogenides, Cu-Bi-S, have been explored as a promising material for solar energy conversion. Recently, devices based on wittichenite-type (Cu₃BiS₃) compound has spurred interest for photovoltaics and water splitting application^[1,2], due to its favorable optoelectronic properties, such as p-type conductivity and direct band gap (Eg ≈ 1.10–1.86 eV) with high optical absorption coefficient (>10⁵ cm) in the visible region^[3]. The aforementioned properties imply that the Cu_xBi_yS_z could be suitable for different applications, however, CBS is still largely unexplored as photo absorber in photovoltaic devices (maximum 1.281% PEC)^[4] and as a semiconductor/photocatalyst. This is largely due to the lack of control over different Cu-Bi-S phases, such as Cu₃BiS₃ and CuBi₃S₅, for which the stability domains are quite narrow in the phase diagram^[5]. In this work, we obtain CBS phases by sulfurization of Cu_xBi_yO_z (CBO) precursor films, with varying Cu:Bi stoichiometry (1:3, 3:1) in the optimized temperature range (350-425^oC). The films were characterized by SEM, RAMAN, XRD, optical absorption spectroscopy. XPS studies were performed in order to identify the stoichiometry and oxidation states of the 3:1 and 1:3 (Cu:Bi) precursor sulfurized at the best temperature condition. The results confirmed the total conversion of CBO to CBS. Although mixed phases were identified for both Cu:Bi ratio at different sulfurization temperature, 3:1 (Cu:Bi) precursor film sulfurized at 350^oC yields Cu₃BiS₃ single phase, which is corroborated by Raman peaks at 279 and 470 cm^-1 corresponding to Cu3BiS3 phase. Sulfurization of 1:3 (Cu:Bi) precursor film at 425^oC leads to CuBi₃S₅ phase; while at 350^oC both Cu₃BiS₃ and CuBi₃S₅ phases are observed, however, Cu₃BiS₃ is more evident, which is in agreement with Raman. Optical measurements showed that 1:3 and 3:1 (Cu:Bi) precursor sulfurized at 350^oC presented similar Eg (1.61 and 1.54 eV, respectively); while 1:3 sulfurized at 425^oC has low Eg (1.0 eV), indicating that the presence of CuBi₃S₅ phase decreases the band gap energy. Band-edges from XPS confirmed that CBS is promising for different applications.

References

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, 132004, doi:10.1088/1361-6528/abcf6d.
2. Yakushev, M.V.; Maiello, P.; Raadik, T.; Shaw, M.J.; Edwards, P.R.; Krustok, J.; Mudryi, A.V.; Forbes, I.; Martin, R.W. Electronic and structural characterisation of Cu₃BiS₃ thin films for the absorber layer of sustainable photovoltaics. Thin Solid Films 2014, 562, 195-199, doi:https://doi.org/10.1016/j.tsf.2014.04.057.
3. Cates, N.; Bernechea, M. Research Update: Bismuth based materials for photovoltaics. APL Materials 2018, 6, 084503, doi:10.1063/1.5026541.
4. Yin, J.; Jia, J. Synthesis of Cu₃BiS₃ nanosheet films on TiO₂ nanorod arrays by a solvothermal route and their photoelectrochemical characteristics. CrystEngComm 2014, 16, 2795-2801, doi:10.1039/C3CE41958D.
5. Chamorro, W.; Mesa, F.; Hurtado-Morales, M.; Gordillo, G. Study of Structural and Morphological Properties of ZnS Films Deposited on Cu₃BiS₃. In Proceedings of the 5th European photovoltaic solar energy conference and exhibition, Spain, 2010; pp. 575-579.

Chalcogenide perovskites have been recently under the researchers’ spotlight as novel absorber materials for photovoltaic applications. BaZrS₃, the most investigated compound of this family, shows high absorption coefficient, a bandgap of around 1.8 eV, and excellent environmental and thermal stability.[1- 3] Despite the high temperature synthesis that has been traditionally employed to prepare this material, milder conditions have been successfully applied recently, opening the possibility for its deposition in thin film and device integration.[4] In addition to the 3D perovskite BaZrS₃ the Ba-Zr-S compositional space contains 2-D Ruddlesden-Popper phases Ba_x+1Zr_xS_3x+1 (with x= 1, 2, 3) which have recently been reported. [5, 6] It is important to note that, differently from hybrid lead halide perovskites in which RP phases are achieved through the addition of a bigger cation, these lower dimensional phases can be obtained only through stoichiometric variation of the 3D elements. Consequently, 3D-2D mixtures, with different optoelectronic properties compared to the pure phases, can be easily created, requiring a deep understanding and control of the mechanisms to avoid the formation of secondary phases.
In this paper we report solid state synthesis and characterization of Ba-Zr-S phases. In order to avoid the use of the easily oxidised metals the starting materials used were BaS and ZrS₂. We report the protocols for synthesis at 900°C in quartz capsules, changing the precursor ratio BaS: ZrS₂ and revealing the formation of a mixture of 3D and 2D chalcogenide perovskite phases. To carefully resolve the composition, we complemented the XRD, SEM and EDS characterization of the synthetic products with the analysis of the Raman-IR spectra. For this purpose, we calculated the phonon mode, deriving the Raman and IR spectra, for the 3D and 2D Ba-Zr-S chalcogenide perovskites and for the binary precursors. This thorough characterization will show how complex can be to distinguish between different phases in the Ba-Zr-S compositional space, stressing the importance of the use of multiple techniques to soundly resolve the composition of these materials.

1. Nishigaki Y. et al., Solar RRL, 2020, 4, 1-8
2. Perera, S. et al., Nano Energy, 2016, 22, 129-135
3. Tiwari, D. et al., Journal of Physics: Energy, 2021, 3, 034010
4. Comparotto, C. et al., ACS Applied Energy Materials, 2022, 5, 6335-6343
5. Li, W. et al., Physical Review Materials, 2019, 3
6. Niu, S. et al., Journal of Materials Research, 2019, 34, 3819-3826

Chalcogenide perovskites are emerging photovoltaic materials with attractive optoelectronic properties including a high absorption coefficient, tunable bandgap, and high dielectric constant. In contrast to their halide counterparts, these materials have been shown to be stable, and earth-abundant with nontoxic constituents. However, the current rudimentary synthesis techniques have limited the growth of these materials. Traditional solid-state and vacuum processing methods demand temperatures in excess of 800-1000 °C bringing up difficulties in identifying suitable substrates and subsequent device fabrication. Hence, gaining insights into the growth mechanism of these materials becomes integral for engineering robust fabrication methodologies.
To address this, we systematically studied the growth mechanism of the BaZrS₃ perovskite through our earlier reported hybrid precursor route utilizing barium thiolate and zirconium hydride precursors. Preliminary findings demonstrate that the high-temperature requirements are related to mass transfer barriers between precursors rather than thermodynamic limitations. We also identified an accessible liquid phase capable of overcoming these diffusional barriers and driving the reaction towards lower temperatures as well as assisting in grain growth.
In conclusion, this work is focused on understanding the growth mechanism of the BaZrS₃ perovskite and identifying the potential liquid phase to drive low-temperature synthesis and support subsequent thin film fabrication.

Chalcogenide perovskites have recently garnered great interest for photovoltaic applications due to their predicted and experimentally verified high stabilities compared with halide perovskites while retaining their excellent optoelectronic properties. One of the greatest challenges with using chalcogenide perovskites for applications in solar cells is that they have historically required high synthesis temperatures, making them incompatible with the glass substrates and the conductive rear-contact layers required to create a photovoltaic device.
Our group has recently published two routes of synthesizing BaZrS₃ and BaHfS₃ using solution-deposited routes utilizing inks with solid suspensions. However, both synthesis routes resulted in island growth of these materials. In this work, we explore methods to create continuous films of BaZrS₃ and BaHfS₃ using inks containing fully dissolved molecular precursors consisting of metal dithiocarbamates and metal dithiocarboxylates. By adjusting the barium to transition metal ratios in the inks, this solution processed synthesis route was also extended to the synthesis of Ba-Zr-S and Ba-Hf-S Ruddlesden-Popper phases, a class of materials that has not been extensively studied. Analytical characterization of the molecular precursors and optoelectronic characterization of the ternary phase materials were then carried out to determine their properties.

For III-V compound semiconductor solar cells to achieve scalability for very large-scale photovoltaics, we need to radically reimagine how they are fabricated. We outline an approach for scalable synthesis of GaAs thin film solar cells that bypasses epitaxial growth and eliminates or minimizes vacuum processing. Starting with thin film absorbers that are formed by mechanical exfoliation from GaAs bulk ingots, we demonstrate that high efficiency cells are possible using methods adopted from silicon and perovskite thin film photovoltaics, such as use of diffused junctions formed at ambient pressures, and solution processing of ohmic contacts and passivation layers using earth-abundant materials. We demonstrate diffused junction GaAs solar cells with 1-Sun AM1.5G efficiencies as high as 23.5%.

Thin-film GaAs photovoltaic pn junction structures were synthesized from bulk GaAs crystals by diffusion doping and spalling to mechanically exfoliate 3-micron thick films of GaAs. The p/n junction is fabricated with a simple zinc-diffusion doping technique, which has yielded open circuit voltages > 960 mV. The films are spalled onto electroplated nickel stressor layers that also serve as the mechanical support for the film. Using 110 oriented GaAs crystals results in large-area mirror-smooth spalled films with 0.1 nm RMS roughness. Preliminary fabrication of thin film diodes yielded implied V_oc values over 830 mV, with ideality factor of two, and an absence of parasitic shunts. The diodes are supported on a Kapton tape, demonstrating their potential application for lightweight photovoltaic manufacturing and deployment. These diodes represent the first steps toward high-efficiency low-cost, epitaxy-free, III-V photovoltaics. Finally, we also performed a technoeconomic analysis indicating a scalable pathway for epitaxy/vacuum-free III-V photovoltaics to achieve LCOE parity with future silicon photovoltaics, with a capital cost of 8cent/Watt for module manufacturing.

The increased energy yield of bifacial PV reduces $/kWh costs and accelerates solar PV adoption. Successful interface passivation at c-Si contacts has enabled excellent bifacial performance for that technology. In contrast, poor electronic and chemical passivation at the back contact of thin film solar cells have thus far prohibited commercially viable bifacial CIGS and CdTe PV technologies, which utilize a p-type polycrystalline absorber. In these devices, the product of electron density and interface recombination velocity yields high recombination current density at the back contact. We review recent progress towards improved experimental and computational understanding and control of the CdTe back contact properties using viable transparent interface materials. As just one example of many, we have studied CdCl₂-treated CdTe devices subjected to low-temperature annealing following deposition of copper and chromium nitrate solutions intended to form Cu_xCrO_y as a p-type buffer layer. The approach has demonstrated backside illuminated devices with high current density and high fill factor, with performance that persists even for thicker absorber layers, showing efficient charge collection even without a fully-depleted device. Our modeling efforts aim to delineate whether passivation consists of reduced downward band-bending (electronic passivation), reduced back-contact defect density (chemical passivation) and back surface recombination velocity (BSRV), or both types of passivation. Devices with 8% back-illuminated efficiency and >50% bifaciality have been demonstrated, but significant progress remains in order to demonstrate high-efficiency state-of-the-art devices with beneficial energy yield. Future bifacial performance gains rely on improved understanding and control of the back contact interface, remediating defects to reduce BSRV, and engineering band-bending which propels electrons away from the back contact region. We will also discuss the implications for bifacial CdTe devices of transitioning from Cu-doped to As-doped absorbers.

<p>Cu-based chalcogenide thin films such as chalcopyrite, kesterite, and etc. have drawn significant attentions as a photovoltaic absorber since the Cu(In,Ga)Se₂ (CIGSe) record cell achieved the power conversion efficiency of 23.35% on a glass substrate. In recent years, there is high demand on lightweight and flexible photovoltaic devices to realize BIPV and VIPV applications. As a result, flexible solar cells have been studied using various flexible substrates such as stainless steel (STS), polyimide (PI), and etc. Furthermore, ultra-thin glass (UTG) is being considered as an emerging potential for flexible solar cells. Despite the successful achievement of high-efficiency flexible CIGSe solar cells (> more than 20%), each flexible material has advantages and disadvantages. The STS substrate contains detrimental Fe element that leads to formation of deep-level defects. Besides, they have a rough surface compared to flat glass substrate. On the other hand, the PI substrate is vulnerable to high-temperature processes so that it is hard to adopt commonly applied conditions (ex. high-temperature process of >550C) for glass substrate to produce high-quality CIGSe films. Presumably, the performance gap between rigid and flexible cells originates from limitations on the material properties and/or high-temperature process. In order to overcome these drawbacks, UTG is also attempted as a flexible substrate of CIGSe solar cells. In this work, we have developed compatible deposition process for high-efficiency CIGSe solar cells on various flexible substrates: STS and PI as well as UTG. Their efficiencies of PV devices on STS, PI, and UTG yielded more than ~19% which was higher than CIGSe solar cells on the rigid glass substrate (~17%). It means that UTG could be also suggested as a novel flexible substrate for CIGSe solar cells. In addition to flexible solar cells, the CIGSe PV devices can be used as a bottom cell in tandem configuration combined with a highly efficient peroveskite-based top cell. To maximize use of the full spectrum of incident light, we fabricated monolithic perovskite/CIGSe tandem solar cells. The best tandem cells showed higher than 20%. Consequently, the obtained results indicate that chalcogenide solar cells still possess versatile potential to demonstrate flexible and tandem applications</p>

Thin film solar cells based on polycrystalline Cu(In,Ga)Se₂ (CIGS) have reached efficiencies of 23.35% on glass and 22.2% on flexible substrates to date. It becomes more and more appealing to develop advanced device architectures like bifacial or semi-transparent devices. However, transparent electrical back contacts (TBC) required in these applications typically suffer from unfavorable transport properties due to the formation of detrimental GaO_x at the CIGS/TBC interface after high temperature CIGS deposition process. The combination of short diffusion length of carriers and high rear interface recombination results in low performance under rear illumination. As a result, efficiencies of devices with TBC remain below that of Mo back contact counterparts, with the highest reported efficiency of 16.1%. The highest efficiency of bifacial devices measured under front and rear illuminations are 9.0% and 7.1%, respectively. Therefore, further development of TCO-based devices including bifacial, bifacial tandem, semi-transparent, and ultra-thin rear-back-contact (RBC) devices remained stagnant.

To overcome those challenges, we investigated prospects of reducing the absorber deposition temperature by adding a small amount of Ag (equivalent of few nanometers) in CIGS to get rid of the detrimental GaO_x interface layer while maintaining the high quality of the absorbers. In this work, absorbers were grown by co-evaporation method on Ag-coated ITO/glass substrates at nominal temperatures ranging from 300°C to 450°C. We identify an optimal process temperature of around 350°C, which gives rise to GaO_x-free interface, high absorber quality, high transmittance of ITO back contact and suitable GGI back gradings. A bifacial CIGS solar cell was obtained with efficiencies of 19.77% and 10.89% under front and rear 1-sun illumination, as independently certified by Fraunhofer ISE. To the best of our knowledge, both values are the highest efficiencies reported for a chalcopyrite-based TBC structure.

Finally, we demonstrate high performance 4-terminal bifacial perovskite/CIGS tandem solar cells. We achieved a power generation density of 28.0 mW/cm² BiFi₃₀₀ with a gain of 8.9 mW/cm² as compared to the performance of the stand-alone CIGS cell. Further improved performance will be presented. Another promising application is all-thin-film 2-terminal bifacial perovskite/CIGS tandem solar devices. Due to higher J_SC in the bottom cell resulting from the additional albedo light, the current matching condition allows perovskite top cell with a reduced bandgap, which generally has less stability issue from halide segregation. Therefore, anticipated high performance and improved stability of bifacial monolithic perovskite/CIGS tandem solar cells could feature a prominent place in future photovoltaics markets.

Our study provides insights into the key limiting factors for bifacial CIGS thin film solar cells. It paves the way for the development of not only TBC-based chalcopyrite solar cell architectures but also next generation of PV technologies toward high-energy yield.

Thin dielectric layers are widely introduced as passivation, tunneling or carrier-transport layers in advanced photovoltaic (PV) cells, implemented in Si as well as more recently other thin-film semiconductor materials.

Specific characterization techniques and physical models need to be considered to identify and assess their electrical properties of importance for the performance of the PV cells.

The main concepts and their significance will be reviewed based on study cases incorporating different reference or innovative materials.

1) The electrical passivation performance (i.e. fixed oxide charges, interface traps and carrier lifetimes) of different Al₂O₃/SiO₂ thin dielectric stacks realized on monocrystalline Si substrates have been extracted from ac measurements of capacitance (C) and conductance (G) over wide ranges of frequencies and temperatures, with an analytical model comprising a 6-element equivalent circuit [1, 2]. The dc leakage current (I) at low voltage was furthermore related to the trap-assisted tunneling. The effective carrier lifetimes were measured by the contactless photoconductance decay method [3]. Lifetimes of almost 1 ms were obtained at low excess carrier density for the dielectric stack featuring the lowest values of oxide charges, interface states and trap-assisted tunneling current. Bias temperature instability was lately investigated on MOS capacitors, before and after annealing, towards optimizing the process conditions for long-term reliability [4].

2) Such extensive electrical characterization methodology, supported by modelling via numerical simulations (SCAPS, Silvaco Atlas), has been extended to Al₂O₃, HfO₂ and SiO_x layers implemented as a rear passivation layer for ultra-thin (UT) CIGS PV cells [5-7]. Extraction of intrinsic cell parameters from the dark current and ac measurements can be correlated with the PV performance under light [8]. Subsequently, the measured parameters have been introduced in a 2D model of rear-passivated UT-CIGS cells in ATLAS. The model has been exploited to understand and optimize the trade-off between passivation quality and series resistance (related to the ratio of passivated area versus the sizes of the rear opening contacts) that impacts gains or losses of cell performances, confirming experimental trends [9].

3) A third category of materials include TiO₂ or SnO₂ metal oxides that we have studied as electron-transport layers on Si, CIGS or perovskite substrates [10-11]. For such doped oxides, the C-V characteristics often show large variations and unexpected behaviors. Based on simulations that fairly reproduce our main observations and provide new physical insight, the results have been related to full or partial depletion effects, depending on the oxide doping level and thickness.

References : [1] Yan Y. et al, Microelectronic Engineering (2022); [2] Yan Y. et al, Solid-State Electronics (2022); [3] Yan Y. et al, Proc. of the IEEE Latin America Electron Devices Conference (LAEDC, 2021); [4] Yan Y. et al, submitted (2022); [5] Kotipalli R. et al, AIP Advances (2015); [6] Cunha J. et al, Solar RRL (2021); [7] Scaffidi R. et al, Physica Status Solidi A (2021); [8] Lontchi J. et al, E-MRS Spring Meeting Symp. A (2020, canceled), replaced by Virtual Chalcogenide PV Conference 2020; [9] Lontchi J. et al, IEEE Journal of Photovoltaics (2020); [10] Lontchi J. et al, Applied Physics Express (2021); [11] Cunha J. et al, Advanced Materials Interfaces (2021)

Zinc tin diphosphide (ZTP) is an alternative material for commercialized thin-film photovoltaics due to of the non-toxic and earth-abundant constituent elements, the suitable bandgap, and the high absorption coefficient^[1,2]. Our group has reported the ZTP bulk crystal-based solar cell with the record conversion efficiency of 3.87% which is still lower than the theoretical limit^[3]. In particular, the short circuit current density of 12.1 mA cm^-2 is lower compared with that of CIGS solar cells, which is probably due to the really short minority carrier lifetime of sub-nanosecond^[4].
We prepared ZTP bulk crystals by solution growth method using Sn as flux. In our previous work, the starting composition was on the line of the Sn–ZnP₂ pseudo–binary system^[5], while it is well known that the formation behavior of intrinsic defects which could be carrier traps depends on the chemical potentials of constituent elements. In this study, we thus attempted to enhance the minority carrier lifetime of ZTP crystals by the control of composition, namely chemical potentials in the system. We actually performed the crystal growth in the compositions with 1.3 and 4 mol% of Zn (Zn-rich and P-rich) in addition to the conventional composition, and the other conditions were the same in the previous report^[5].
We characterized the ZTP crystal grown under the conventional condition by deep-level transient spectroscopy (DLTS) in which Schottky diodes with Ag (Schottky) /ZTP / Cu₃P / Cu (ohmic) were used. DLTS measurement showed that ZTP had two traps for electrons and five traps for holes within the bandgap. Particularly, the nearest electron trap level to conduction band minimum (CBM), E1, had extremely high capture cross section of ~10^-11 cm² compared with other traps. Furthermore the time constant for electron capture from CBM to E1 was estimated to be ~10^-4 ns, which was equivalent to the reported minority carrier lifetime^[4], indicating that the trap E1 might be the recombination center of ZTP.
We also evaluated the influence of growth condition on the properties of recombination. The fluorescence lifetimes of minority carriers were characterized by time-resolved photoluminescence (TRPL) at the photon energy of 1.69 eV under room temperature. The net fluorescence lifetime of samples of Zn-rich, conventional, and P-rich conditions were estimated to be 0.39, 0.066, and 0.065 ns, respectively. The lifetime of Zn-rich sample is higher than others and the results in the literature^[4]. Steady–state PL spectra were also measured at 77K and found to be contained several emissions with energy of around 1.28, 1.52, and 1.77 eV. The emission at 1.77 eV might come from Band-to-Band transition and its intensity in the sample of Zn-rich condition was higher than that of others. This indicates that the concentration of the active recombination center E1 is low in Zn-rich sample, which is consistent with the results of TRPL.
In spite of the change of defect properties by growth conditions, the compositions of all the obtained ZTP were stoichiometric. Therefore, it was implied that the compositional difference of a few mole could affect the formation of intrinsic defects, especially related to E1. We consider two contributions of this phenomena from the viewpoint of thermodynamics: entropy and enthalpy. Both of them depend on chemical potentials of constituent elements in the system during crystal growth, where ZTP and liquid are in thermodynamic equilibrium. Hence, we evaluated the chemical potential of elements in liquid phase using sub-regular solution model, and clarified the precipitation temperature is lower in Zn-rich condition. In the presentation, we will discuss more quantitatively the effects of the entropy and the enthalpy on the formation of defects in ZTP.
[1] S. Nakatsuka and Y. Nose, JPCC (2017). [2] T. Yokoyama et al., APEX (2013). [3] T. Kuwano et al., SOLMAT (2021). [4] S. Nakatsuka et al., PSSa (2017). [5] S. Nakatsuka et al., PSSc (2015).

Sb₂S₃ has become popular in several research fields [1,2] and most prominently among the PV community, achieving efficiency values close to 8% in short time. One of the key aspects of this compound is its high structural anisotropy, where the atoms are bonded by covalent bonds along one crystallographic axis forming ribbons, which are attached together by van der Waals forces along the other two directions [3]. This translates into highly anisotropic properties that are, in fact, an important parameter to control when implementing Sb₂S₃ in optoelectronic devices, where, for example, increased mobility is achieved along the covalently bonded axis. Likewise, this anisotropy is expected to have a strong impact on the vibrational properties of Sb₂S₃.
Being a material that attracted significant interest in the last years, the vibrational properties of this anisotropic compound are still not properly understood in the literature. In order to establish a clearer understanding of the discrepancies between Raman spectra of Sb₂S₃ in existing articles, an investigation of the fundamental vibrational properties of the Sb₂S₃ compound by Multiwavelength Raman spectroscopy was performed. As a starting point, in order to fully identify the features present in the Raman spectra of Sb₂S₃, a single crystal grains powder sample was analyzed in order to minimize the possible influence of the crystallographic orientation, with a multiwavelength approach using eight excitation wavelengths in the range from 325 up to 1064 nm. Resonance and pre-resonance effect yielded changes in the intensity of several Raman peaks, highlighting them in dependence on the excitation laser employed. Furthermore, low temperature (~20 K) measurements were also carried out with λ_ex = 632.8 nm excitation on the powder sample, reducing the thermal agitation of the material and reducing the width of the peaks, which allowed to better resolve the low intensity peaks. Finally, polarization dependent Raman scattering measurements have been also performed on a Sb₂S₃ single crystal mineral, which not only enrich the knowledge about the specific peaks, but also give rise to some changes in peak intensities in the different configurations. As a result, the deconvolution of the obtained Raman spectra for each wavelength, polarization and temperature was performed, thus enabling the generation of a complete fitting function using a combination of Lorentzian curves. This permitted us to establish the position of up to 19 Raman active optical modes from the 30 theoretically predicted, which serves as an important stepping stone for the correct interpretation of the experiments through all the research fields interested in the material.
The study is completed by a deeper analysis of the Sb₂S₃ single crystal mineral. X-ray diffraction analysis allowed to identify the specific crystallographic orientation of these facets, and Raman scattering measurements showed a clear difference in the spectra measured at different fundamental crystallographic planes. The found variations allow the implementation of Raman spectroscopy, a non-invasive technique with lateral micrometric precision, to be used as a quality control tool for the in-situ and in-sample crystallographic orientation assessment of Sb₂S₃ thin films that are proposed as emerging PV technologies with a relevant cost-efficient potential. Finally, application of Raman scattering in device grade layers and finished solar cells corroborates the potential of the technique for the identification of the potential presence of secondary phases and the crystalline assessment of the layers, which provides with a powerful non-destructive characterization tool for device optimization.

This work is part of the R+D+i MaterOne project Ref. PID 2020-116719RB-C42 funded by MCIN/AEI/10.13039/5011000110033.

[1] R. Kondrotas et al., Joule 2 (2018) 857
[2] H. Liu et al., Small 17 (2021) 2008133
[3] Z. Cai et al., Sol. RRL 4 (2020) 1900503

Silver Antimony Sulfur Selenide, AgSb(S_xSe_1-x)₃ thin-film solar cells have been emerging as third-generation photovoltaic devices with promising properties of the tunable bandgap (0.7 - 1.9 eV), good doping concentration (10¹⁶ cm^-3), high absorption coefficient (>10⁴ cm^-1), and simplified processing conditions with high vapor pressure and low melting point (550 C). There have been several studies about AgSb(S_xSe_1-x)₃ thin-film absorbers by depositing Sb₂(S_xSe_1-x)₃, Ag, and Se thin-film precursors, followed by the heat treatment. With the favorable optical bandgap of AgSb(S_xSe_1-x)₃ thin-films (x=0.53, 0.58, and 0.61), the efficiency has been improved with >2.77 %. In another approaches, AgSbSe₃ thin-film absorber devices demonstrate <0.1 %, whereas AgSbS₃ thin-films devices show approximately 0.27 %. Since Antimony Sulfur Selenide, Sb(S_xSe_1-x)₃ thin-films have demonstrated good optical and electronic properties as an absorber layer, further optimization of thin-film absorber layers could be achieved by utilizing both Sb(S_xSe_1-x)₃ and AgSb(S_xSe_1-x)₃ thin-films. As we reported earlier, substituting Ag in Sb(S_xSe_1-x)₃ thin-films tends to increase the bandgap of the absorber layer by lowering the valence band based on studies of other thin-film absorber layers such as Copper Indium Gallium Diselenide (CIGS) and Copper Zinc Tin Sulfur Selenide (CZTSSe) thin-films.

In this study, we have conducted a theoretical study on thin-film absorbers of AgSb(S_xSe_1-x)₃, Sb(S_xSe_1-x)₃, and the combination of AgSb(S_xSe_1-x)₃ and Sb(S_xSe_1-x)₃ thin-films from the electronic band structure perspective. To fully utilize the promising properties of both Sb₂(S_xSe_1-x)₃ and AgSb(S_xSe_1-x)₃ films, we investigated different compositions and concentrations of Sulfur and Selenium with proposed empirical equations for electron affinity and bandgap energy. Furthermore, we varied the doping of thin-film absorbers to search for the optimal conditions for the band alignment to Molybdenum metal back contact films. For our analytical simulation, we have used in-house MATLAB modeling suites that have been developed in our group.

Four different structures of thin-film absorbers have been modeled above Molybdenum metal thin-films. The workfunction of Molybdenum metal thin-films is 4.9 eV and an inevitable MoS_xSe_1-x interlayer was inserted during our simulation throughout this study. For both AgSb(S_xSe_1-x)₃ and Sb(S_xSe_1-x)₃ thin-films, the electron affinity and bandgap energy increase as Sulfur (x) composition increases. However, the increased bandgap is not directly translated into improved solar cells efficiency due to the alignment of thin-film electronic structures. For example, the best efficiency was achieved with 2 um AgSb(S_0.4Se_0.6)₃ thin-film devices (18.4 %) at sulfur concentration, x = 0.4. However, once we combine two AgSb(S_xSe_1-x)₃ and Sb(S_xSe_1-x)₃ thin-films while keeping a total thickness, 2 um (1 um/1 um), an interface between AgSb(S_0.4Se_0.6)₃/ Sb₂(S_0.4Se_0.6)₃ and Molybdenum metal thin-films is preferably formed due to reduced effective Schottky hole barrier. If we assume the same amount of defect states at the interface, the improved effective Schottky hole barrier is 128 mV due to the favorable band alignment, which is approximate 4.3 times better than a AgSb(S_0.4Se_0.6)₃ thin-film structure. With a bi-layer AgSb(S_0.4Se_0.6)_3/Sb₂(S_0.4Se_0.6)₃ thin-film absorber, we studied various doping concentrations impact on device efficiency based on the modified electronic band structure of each thin-film. The doping concentration of a AgSb(S_0.4Se_0.6)₃ thin-film mainly increases the photogenerated current while a Sb₂(S_0.4Se_0.6)₃ thin-film improves open circuit voltage.

Zinc oxide has attracted attention for decades due to its important optical, electrical, chemical and microstructural properties, interesting for optoelectronic applications in light emitting devices, transparent conduction oxide (TCO) in solar cells, gas sensors, flat screens and touch panel displays, in vacuum fluorescent displays, in photovoltaics or thermal mirrors. But the n-type conductivity of non-doped ZnO films is not high enough to be used in devices. Group–III atoms (Al, Ga, In) are very important donors for ZnO film–based technology. However, the free electron density in doped ZnO films was revealed to be limited to 10¹⁹ - 10²¹ cm^-3. due to the self-compensating effect, related to generation of the acceptor-type defects to counteract of donor doping. The nature of these acceptor type defects and the factors that stimulate the self-compensating effect still remain to be investigated.
It was supposed that one reason for the self – compensating effect can be connected with a high level of elastic stresses in the doped ZnO films with high donor contents. In the present work, the variation of the optical, structural and electrical parameters of the ZnO nanocrystal (NC) films doped with Ga and In atoms has been investigated. ZnO films of 2 groups were grown by ultrasonic spray pyrolysis with the In contents 1.0 at% (1) or 2.0 at% (2) and the different concentrations of Ga from the range 0.5-3.0 at %. Using co-doping by Ga and In atoms with lower (Ga) and higher (In) ionic radii compared to the Zn ions, it is expected to lower the stresses in the films and gain insight into the factors favored to the self-compensating effect.
The ZnO films have been studied using scanning electron microscopy (SEM), energy dispersive X ray spectroscopy (EDS), X ray diffraction (XRD), photoluminescence (PL) and X-ray photo electronic spectroscopy (XPS). Non-monotonic changes in the morphology of films were revealed that correlates with non-monotonic changes of oxygen and indium contents in ZnO films with variation of Ga contents. Compressive stresses arising due to In doping prevent effective oxidation at the film crystallization. The Ga doping up to 1.5 at% allows to compensate the compressive stresses in the films that is favored to the effective ZnO oxidation and the dissolution of In ions in the thermal annealing. High donor doping was shown to be accompanied by the appearance of a new near band edge (NBE) emission band, related to carrier recombination in the shallow donor-acceptor pairs (DAPs), which is a “fingerprint” of a beginning of the self-compensating process. Simultaneously, the effective charges of the Ga (In) ions decrease and the rate of electrical resistivity reduction slows down. The nature of shallow donors and acceptors in DAPs, as well as the impact of co-doping by Ga and In atoms on the self-compensating effect have been discussed.

Chalcogenide perovskite is an emerging class of semiconductor materials with potential applications in electronics and optoelectronics. SrHfS₃ in the distorted perovskite structure is shown to be a direct gap semiconductor with strong green photoluminescence. Together with its high stability defect tolerance, it many find applications such as LEDs. In this work we report the synthesis of SrHfS₃ and SrZrS₃ thin films by magnetron sputtering using metal targets, followed by CS₂ sulfurization. Salts was used improve the film crystallinity. Sample growth on conductive substrates was also attempted, with an eye for device fabrications. Our work demonstrates a new way of fabricating SrHfS₃ and SrZrS₃ thin films with controllable composition.

We present developments in surface photovoltaic spectroscopy (SPS), aimed at characterizing passivation strategies for CdTe cell interfaces, particularly the rear contact. Effective rear contact passivation has been predicted to increase device efficiency by as much as 3.5% when coupled with a hole selective contact. We demonstrate how SPS is a critical tool in evaluating novel solutions to eliminate voltage loss at interfaces. We examine the effect of known wet etchant strategies on SPS signal, correlating and contrasting the information gained with that available via TRPL. We also explore passivating interface layers in CdTe and Si based devices. SPS signals are evaluated via SCAPS 1D modeling and analytical modeling and the effects of surface recombination velocity, band bending, and recombination in bulk and SCR are extracted.

This presentation summarizes our latest research on wide-gap, silver-alloyed Cu(In,Ga)Se₂ (CIGS) solar cells (i.e. forming ACIGS). The processed absorber layers had band gap values (E_G) in the range of 1.40 to 1.65 eV, making them attractive for application in top cells of tandem devices. To reach this high E_G level, the [Ga]/([Ga]+[In]) (GGI) ratio needs to be at least 0.65. For pure CIGS solar cells, such high Ga concentrations not only lead to a large deficit in open-circuit voltage (V_OC), but also to a distinct deficit in short-circuit current density (J_SC) and fill factor (FF). The main problem seems to be a deteriorating absorber quality with increasing GGI. However, since the application of alternative buffer layers can mitigate the V_OC loss, wide-gap CIGS devices with standard CdS buffers must be partly limited by interface recombination as well.

Alloying silver into wide-gap CIGS has been explored by several groups before [1-3]. Although the performance gap to low-E_G devices could not be closed completely, the results highlight the potential for solar cells with E_G > 1.4 eV.

In our recent works [4-8], we first focused on identifying the optimum compositional window for high-performance ACIGS solar cells, while keeping the GGI level above 0.65. Calculations based on density functional theory show that the conduction band shifts downwards with Ag addition. A beneficial positive conduction band offset with CdS is predicted for an [Ag]/([Ag]+[Cu]) (AAC) value of at least 0.3 (increasing with increasing GGI). Our experimental results, with highest efficiencies for AAC = 0.5 at a band gap of 1.45 eV (GGI = 0.72), confirmed this trend. The champion device with a (Zn,Sn)O buffer exhibited an efficiency of 15.1% without anti-reflection coating (ARC) [4]. However, for higher GGI and AAC values (i.e. E_G > 1.5 eV), the efficiency starts to decline significantly.

To investigate the limitations for ACIGS solar cells with higher E_G, the effect of the absorber stoichiometry was studied for a constant GGI = 0.85 and AAC = 0.8, yielding E_G = 1.62 eV (i.e. sweet-spot for tandem applications) [5]. It was found that already a minor group-I deficiency leads to a substantial formation of ordered vacancy compounds (OVCs). For too off-stoichiometric ACIGS, OVC patches form at the back interface, leading to a significant FF drop. The highest efficiency of 10.2% (no ARC) is reached for a reverse voltage sweep (hysteresis found).

Next, the AAC and GGI ratios were simultaneously reduced again to about 0.60 and 0.75, respectively. Although the OVC formation was mitigated, our analysis of a large sample set proved a clear stoichiometry effect on the solar cell characteristics [6]. It was revealed that near-stoichiometric absorbers approach full depletion, whereas off-stoichiometric materials have doping densities “common” for CIGS absorbers. Due to the generally low diffusion length, this results in a low carrier collection (but higher V_OC) for off-stoichiometric, and perfect carrier collection (but lower V_OC) for stoichiometric samples. Consequently, the highest V_OC of 916 mV was measured for an off-stoichiometric sample (E_G = 1.46 eV, CdS buffer). By applying an RbF post-deposition treatment, the V_OC could be further increased to 926 mV and an efficiency of 16.3% (no ARC) was achieved [7].

Finally, the Mo back contact was exchanged by an In₂O₃:H film, with almost complete near-infrared transparency [8]. After fine-tuning the sodium supply, an efficiency of 12% (no ARC) was reached, using a stoichiometric ACIGS absorber with E_G = 1.44 eV.

[1] T. Nakada et al., MRS Proc. 2005, 865, F11.1.
[2] G. M. Hanket et al., IEEE Photovolt. Spec. Conf. 2009, 001240.
[3] K. Kim et al., Nano Energy 2018, 48, 345.
[4] J. Keller et al., Prog. Photovoltaics Res. Appl. 2020, 28, 237.
[5] J. Keller et al., Sol. RRL 2020, 4, 2000508.
[6] J. Keller et al., Sol. RRL 2021, 5, 2100403.
[7] J. Keller et al., Sol. RRL. 2022, 6, 2200044.
[8] J. Keller et al., Sol. RRL. 2022, 6, 2200401.

Transitioning into global carbon neutrality by 2050 is vital to ensure the health and future of Earth. The fabrication of a low-cost, high efficiency thin film photovoltaic (PV) material for building-integrated PV will help fulfil the need for renewable energy sources moving forward. Cu₂ZnSn(S,Se)₄ (CZTSSe) is an earth-abundant, non-toxic thin film solar absorber with a theoretical power conversion efficiency (PCE) of 20%, [1] and can help fulfil this PV requirement. Severe voltage deficits due to deep-level defects and interfacial recombination have plagued the performance of a CZTSSe/CdS device, where little is known about the surface electronic landscape of the key absorber-buffer interface. This study replaces the use of toxic CdS with ZnSnO (ZTO) as the buffer material, deposited using atomic layer deposition, due to ZTO having a larger, tuneable bandgap between 3.3 to 3.7 eV. The electronic structure of the CZTSSe/ZTO interface will be probed using energy-filtered photoemission electron microscopy (EF-PEEM) to identify possible harmful features hindering performance. Furthermore, band alignment studies, using ultraviolet photoelectron spectroscopy, will investigate the suitability of ZTO as a buffer layer.

[1]: S. Kim, J. A. Márquez, T. Unold and A. Walsh, Energy Environ. Sci., 2020, 13, 1481-1491

Inorganic solar cells based on Cu₂ZnSn(S_xSe_1-x)₄ thin-film absorbers have reached power conversion values of 11.5% ¹ and 13.0 % ² in the sulfide and sulfo-selenide forms, respectively. Although these performance levels are some of the most promising in In-free substrate device configuration, they remain far below their theoretical limit.³ It is widely recognised that cell voltage is the key limiting factor in these devices, and a variety of strategies have been investigated including doping and alloying the absorber layer.^4-6

The complexity in the composition of these compound semiconductors offers key advantages such as tuneable band energetics, but also brings about significant challenges in terms of secondary phases and point defects which can be detrimental to device performance.⁷ For instance, voltage loss has been linked Cu/Zn antisites as well as Sn defect.^3,7,8 In this contribution, we investigate the position and distribution of defect states in the bulk and at the surface of Cu₂ZnSnS₄ (CZTS) thin films prepared by solution based methods, employing variable temperature photoluminescence (PL) spectroscopy and high-resolution energy-filtered photoemission of electron microscopy (EF-PEEM), a technique which can unravel valuable information on variations of the surface electronic properties of these complex materials.^9,10,11

Variable temperature PL spectra are dominated by emission from quasi-donor acceptor pairs (QDAP) with activation energy values strongly affected by the introduction Sb and Na:Sb as dopants.¹² Indeed, we will show that incorporation of Sb in the range 0.05% wt leads to a clear transition from localized to non-localized QDAP emission. EF-PEEM also reveals significant changes in the surface electronic landscape upon doping, with mean local effective work function decreasing by more than 0.2 eV upon Sb doping. PL and EF-PEEM clearly show that Na can regulate Sb doping levels, keeping the same PL characteristics of CZTS without any added dopants but with a smoother surface electronic landscape as probed by EF-PEEM.^10,12 We will also correlate variable temperature PL of the thin-films with variable temperature admittance spectroscopy of full devices. This analysis allows linking spectroscopic signatures associated with Cu/Zn and Sn defect states with admittance responses of the device. We will show that dopants such as Na and Sb can affect not only carrier density, but also decrease the density of defects associated with elemental disorder in the film as well as the relative position of the band edges, which can have a profound defect in device performance.

References:
1. Yan et al. Nature Energy 3 (2018) 764
2. Y. Gong et al. Nature Energy 7 (2022) 966-977
3. A. Walsh Faraday Discuss. (2022) DOI: 10.1039/d2fd00135g
4. Y. Romanyuk et al. J. Phys. Energy 1 (2019) 044004
5. S. Lie et al. J. Mater. Chem. A 10 (2022) 9137
6. D. Tiwari et al. Chem. Mater. 28 (2016) 4991
7. D.B. Mitzi and Y. Kim Faraday Discuss. (2022) DOI: 10.1039/d2fd00132b
8. Kattan et al. Nanoscale 8 (2016) 14369
9. D. Tiwari et al. iScience 9 (2018) 36
10. D. Tiwari et al. ACS Energy Lett. 3 (2018) 2977
11. M.C. Naylor et al. Faraday Discuss. (2022) DOI: 10.1039/d2fd00132b
12. D. Tiwari et al. ACS Appl. Energy Mater. 5 (2022) 3933

It has been shown for many solar cell absorbers that Voc loss correlates with the Urbach energy, that describes the decay of tail states into the band gap.
In chalcopyrites it was furthermore observed that heavy alkali postdeposition treatments reduce both: Urbach energy and Voc loss. This behaviour has been attributed to passivation of grain boundaries [1]. However, recently we have shown, that alkali treatment in single crystalline films without grain boundaries also leads to a reduction of Voc loss and Urbach energies [2]. This behaviour can be traced back to a reduced compensation by increasing the doping level [3]. It is to be expected that a similar mechanism plays a role in polycrystalline films.
Part of the dependence of the Voc loss on Urbach energy can be attributed to radiative and non-radiative recombination in and through tail states [4]. However, the dependence is stronger than would be expected from the increased recombination. We now understand that this is due to a simultaneity: increasing doping has a simultaneous effect, increasing Voc and decreasing tail states.
Thus, we propose, that one effect of heavy alkali treatment is due to the observed increase of Na doping inside the grains, because it increases the net-doping, reduces tail states and thereby reduces radiative and non-radiative recombination. All this increases Voc.

[1] Siebentritt et al. Adv. Energy Mat. 10, 1903752 (2020).
[2] Ramírez et al. Solar RRL 2021, 2000727 (2021).
[3] Ramírez et al. submitted (2022).
[4] Wolter et al. Progr. Photovolt. 30, 702 (2022).

Understanding of electronic band structure and charge carrier dynamics of halide perovskites and their interface is critical for the advancement of perovskite photovoltaics[1,2]. Inorganic Cesium lead bromide (CsPbBr₃) perovskite nanocrystals (NCs) with a bandgap of 2.3 eV are regarded as one of the potential candidates as an active layer in solar cells devices, thanks to significant progress made recently in terms of improving its intrinsic properties and stability under environmental conditions. However, stability under UV and X-ray exposure remains an open question[3-6]. The stability of spin-coated CsPbBr₃ NCs deposited on gold (Au) substrates is investigated here under the impact of the UV laser light and X-rays. Under the UV laser pulse, the NCs exhibit distinctly different behavior depending upon the energy of the laser pulse, its fluence and time NCs exposed to it. When the NCs are exposed to UV laser pulses with the photon energy of 3.6 eV (greater than the bandgap), the NCs are modified irreversibly. Both metallic lead and ionic lead with a new component of Pb4f are detected in the Pb4f core-level spectra. Similarly, when NCs exposed exclusively to X-ray with high flux, same components of Pb4f were found in Pb4f spectrum. The modification is more or less similar in either case; however, under UV laser exposure, a large chemical shift of 0.69 eV towards higher binding energy observed after the exposure, which is not the case with X-ray exposure alone. The chemical shift could be attributed to the accumulation of charge carriers at the interface under UV laser and getting trapped by the metallic lead (donor-type surface states) [6]. Therefore, depending on the type of radiation exposure either UV laser or X-rays, the CsPbBr₃ NCs show different behavior and thus the energy level alignments. We show that the chemical properties of NCs are sensitive to UV laser and X-rays irradiation, and investigation of these changes under the type of radiation exposure might help us understand the underlying phenomenon that governs these changes.