Creation of a structured solar cell material dataset and performance prediction using large language models.

Materials scientists usually collect experimental data to summarize experiences and predict improved materials. However, a crucial issue is how to proficiently utilize unstructured data to update existing structured data, particularly in applied disciplines. This study introduces a new natural language processing (NLP) task called structured information inference (SII) to address this problem. We propose an end-to-end approach to summarize and organize the multi-layered device-level information from the literature into structured data. After comparing different methods, we fine-tuned LLaMA with an F1 score of 87.14% to update an existing perovskite solar cell dataset with articles published since its release, allowing its direct use in subsequent data analysis. Using structured information, we developed regression tasks to predict the electrical performance of solar cells. Our results demonstrate comparable performance to traditional machine-learning methods without feature selection and highlight the potential of large language models for scientific knowledge acquisition and material development.

Quantifying the impact of wildfire smoke on solar photovoltaic generation in Australia.

The 2019-20 Australian wildfires caused extreme haze events across New South Wales (NSW), which reduced photovoltaic (PV) power output. We analyze 30-min energy data from 160 geographically separated residential PV systems in NSW with a total capacity of 312 kW from 6 Nov 2019-15 Jan 2020. The observed mean power reduction rate for PV energy generation as a function of the fine particulate matter (PM2.5) concentration is 13 ± 2% per 100 µg/m3 of PM2.5. The resulting energy loss for residential and utility PV systems is estimated at 175 ± 35 GWh, equating to a worst-case financial loss of 19 ± 4 million USD. We found the relative impact to be most significant in the mornings and evenings, which may necessitate the installation of additional energy storage. As PV systems are sensitive to smoke and become ubiquitous, we propose employing them to support wildfire detection and monitoring.

Automatically Generated Datasets: Present and Potential Self-Cleaning Coating Materials.

The rise of urbanization coupled with pollution has highlighted the importance of outdoor self-cleaning coatings. These revolutionary coatings contribute to the longevity of various surfaces and reduce maintenance costs for a wide range of applications. Despite ongoing research to develop efficient and durable self-cleaning coatings, adopting systematic research methodologies could accelerate these advancements. In this work, we use Natural Language Processing (NLP) strategies to generate open- and traceable-sourced datasets about self-cleaning coating materials from 39,011 multi-disciplinary papers. The data are from function-based and property-based corpora for self-cleaning purposes. These datasets are presented in four different formats for diverse uses or combined uses: material frequency statistics, material dictionary, measurement value datasets for self-cleaning-related properties and optical properties, and sentiment statistics of material stability and durability. This provides a literature-based data resource for the development of self-cleaning coatings and also offers potential pathways for material discovery and prediction by machine learning.

Opinion Mining by Convolutional Neural Networks for Maximizing Discoverability of Nanomaterials.

The scientific literature contains valuable information that can be used for future applications, but manual analysis presents challenges due to its size and disciplinary boundaries. The prevailing solution involves natural language processing (NLP) techniques such as information retrieval. Nonetheless, existing automated systems primarily provide either statistically based shallow information or deep information without traceability, thereby falling short of delivering high-quality and reliable insights. To address this, we propose an innovative approach of leveraging sentiment information embedded within the literature to track the opinions toward materials. In this study, we integrated material knowledge into text representation and constructed opinion data sets to hierarchically train deep learning models, named as Scientific Sentiment Network (SSNet). SSNet can effectively extract knowledge from the energy material literature and accurately categorize expert opinions into challenges and opportunities (94% and 92% accuracy, respectively). By incorporating sentiment features determined by SSNet, we can predict the ranking of emerging thermoelectric materials with a 70% correlation to experimental outcomes. Furthermore, our model achieves a commendable 68% accuracy in predicting suitable nanomaterials for atomic layer deposition (ALD) over time. These promising results offer a practical framework to extract and synthesize knowledge from the scientific literature, thereby accelerating research in the field of nanomaterials.

Bifacial and Semitransparent Sb2 (S,Se)3 Solar Cells for Single-Junction and Tandem Photovoltaic Applications.

Thin-film solar cells are expected to play a significant role in the space industry, building integrated photovoltaic (BIPV), indoor applications, and tandem solar cells, where bifaciality and semitransparency are highly desired. Sb2 (S,Se)3 has emerged as a promising new photovoltaic (PV) material for its high absorption coefficient, tunable bandgap, and nontoxic and earth-abundant constituents. However, high-efficiency Sb2 (S,Se)3 solar cells exclusively employ monofacial architectures, leaving a considerable gap toward large-scale application in aforementioned fields. Here, a bifacial and semitransparent Sb2 (S,Se)3 solar cell and its extended application in tandem solar cells are reported. The transparent conductive oxides (TCOs) and the ultrathin inner n-i-p structure provide high long-wavelength transmittance. Despite the MnS/ITO Schottky junction, power conversion efficiencies (PCEs) of 7.41% and 6.36% are achieved with front and rear illumination, respectively, contributing to a great bifaciality of 0.86. Consequently, the reported device gains great enhancement in PV performance by exploiting albedo of surroundings and shows exceptional capability in absorbing tilt incident light. Moreover, an Sb2 (S,Se)3 /Si tandem solar cell with a PCE of 11.66% is achieved in preliminary trials. These exciting findings imply that bifacial and semitransparent Sb2 (S,Se)3 solar cells possess tremendous potential in practical applications based on their unique characteristics.

10.3% Efficient Green Cd-Free Cu2 ZnSnS4 Solar Cells Enabled by Liquid-Phase Promoted Grain Growth.

Small grain size and near-horizontal grain boundaries are known to be detrimental to the carrier collection efficiency and device performance of pure-sulfide Cu2 ZnSnS4 (CZTS) solar cells. However, forming large grains spanning the absorber layer while maintaining high electronic quality is challenging particularly for pure sulfide CZTS. Herein, a liquid-phase-assisted grain growth (LGG) model that enables the formation of large grains spanning across the CZTS absorber without compromising the electronic quality is demonstrated. By introducing a Ge-alloyed CZTS nanoparticle layer at the bottom of the sputtered precursor, a Cu-rich and Sn-rich liquid phase forms at the high temperature sulfurization stage, which can effectively remove the detrimental near-horizontal grain boundaries and promote grain growth, thus greatly improving the carrier collection efficiency and reducing nonradiative recombination. The remaining liquid phase layer at the rear interface shows a high work function, acting as an effective hole transport layer. The modified morphology greatly increases the short-circuit current density and fill factor, enabling 10.3% efficient green Cd-free CZTS devices. This work unlocks a grain growth mechanism, advancing the morphology control of sulfide-based kesterite solar cells.

Designing 3d metal oxides: selecting optimal density functionals for strongly correlated materials.

Transition metal oxides (TMOs) have remarkable physicochemical properties, are non-toxic, and have low cost and high annual production, thus they are commonly studied for various technological applications. Density functional theory (DFT) can help to optimize TMO materials by providing insights into their electronic, optical and thermodynamic properties, and hence into their structure-performance relationships, over a wide range of solid-state structures and compositions. However, this is underpinned by the choice of the exchange-correlation (XC) functional, which is critical to accurately describe the highly localized and correlated 3d-electrons of the transition metals in TMOs. This tutorial review presents a benchmark study of density functionals (DFs), ranging from generalized gradient approximation (GGA) to range-separated hybrids (RSH), with the all-electron def2-TZVP basis set, comparing magneto-electro-optical properties of 3d TMOs against experimental observations. The performance of the DFs is assessed by analyzing the band structure, density of states, magnetic moment, structural static and dynamic parameters, optical properties, spin contamination and computational cost. The results disclose the strengths and weaknesses of the XC functionals, in terms of accuracy, and computational efficiency, suggesting the unprecedented PBE0-1/5 as the best candidate. The findings of this work contribute to necessary developments of XC functionals for periodic systems, and materials science modelling studies, particularly informing how to select the optimal XC functional to obtain the most trustworthy description of the ground-state electron structure of 3d TMOs.

Converting Sewage Water into H2 Fuel Gas Using Cu/CuO Nanoporous Photocatalytic Electrodes.

This work reports on H2 fuel generation from sewage water using Cu/CuO nanoporous (NP) electrodes. This is a novel concept for converting contaminated water into H2 fuel. The preparation of Cu/CuO NP was achieved using a simple thermal combustion process of Cu metallic foil at 550 °C for 1 h. The Cu/CuO surface consists of island-like structures, with an inter-distance of 100 nm. Each island has a highly porous surface with a pore diameter of about 250 nm. X-ray diffraction (XRD) confirmed the formation of monoclinic Cu/CuO NP material with a crystallite size of 89 nm. The prepared Cu/CuO photoelectrode was applied for H2 generation from sewage water achieving an incident to photon conversion efficiency (IPCE) of 14.6%. Further, the effects of light intensity and wavelength on the photoelectrode performance were assessed. The current density (Jph) value increased from 2.17 to 4.7 mA·cm-2 upon raising the light power density from 50 to 100 mW·cm-2. Moreover, the enthalpy (ΔH*) and entropy (ΔS*) values of Cu/CuO electrode were determined as 9.519 KJ mol-1 and 180.4 JK-1·mol-1, respectively. The results obtained in the present study are very promising for solving the problem of energy in far regions by converting sewage water to H2 fuel.

Large volume tomography using plasma FIB-SEM: A comprehensive case study on black silicon.

The xenon plasma focused ion beam and scanning electron microscopy (PFIB-SEM) system is a promising tool for 3D tomography of nano-scale materials, including nanotextured black silicon (BSi), whose topography is difficult to measure with conventional microscopy techniques. Advantages of PFIB-SEM include high material removal rates, precise control of milling parameters and automated slice-and-view procedures. However, there is no universal sample preparation procedure nor is there an established ideal workflow for the PFIB-SEM slice-and-view process. This work demonstrates that specimen preparation, including the orientation of the volume of interest, is critical for the quality of the final reconstructed 3D model. It thoroughly explores three unique configurations incrementally optimized for higher total throughput. All three sampling configurations are applied to a resin-embedded BSi sample to determine the most favourable workflow and highlight each approach's advantages and disadvantages. The reconstructed 3D models of the BSi surface obtained are shown to be qualitatively closer to the topography measured directly by SEM. The height distribution data extracted from the rendered 3D models reveal a higher structure depth compared to that obtained from an atomic force microscopy measurement. Furthermore, the work demonstrates how samples with different rigidity react to long-term ion-beam interaction, as both amorphous (resin) and crystalline (Si) material is present in the tested specimen. This study improves the understanding of sample-beam interaction and broadens the utility of the 3D PFIB-SEM for more complicated sample structures.

Impact of Pregrown SiOx on the Carrier Selectivity and Thermal Stability of Molybdenum-Oxide-Passivated Contact for Si Solar Cells.

Thin SiOx interlayers are often formed naturally during the deposition of transition metal oxides on silicon surfaces due to interfacial reaction. The SiOx layer, often only several atomic layers thick, becomes the interface between the Si and deposited metal oxide and can therefore influence the electrical properties and thermal stability of the deposited stack. This work explores the potential benefits of controlling the properties of the SiOx interlayer by the introduction of pregrown high-quality SiOx which also inhibits the formation of low-quality SiOx from the metal-oxide deposition process. This work demonstrates that a high-quality pregrown SiOx can reduce the interfacial reaction and results in a more stoichiometric MoOx with improved surface passivation and thermal stability linked to its lower Dit. Detailed experimental data on carrier selectivity, carrier transport efficiency, annealing stability up to 250 °C, and in-depth material analysis are presented.

3D characterisation using plasma FIB-SEM: A large-area tomography technique for complex surfaces like black silicon.

This paper demonstrates an improved method to accurately extract the surface morphology of black silicon (BSi). The method is based on an automated Xe+ plasma focused ion beam (PFIB) and scanning electron microscope (SEM) tomography technique. A comprehensive new sample preparation method is described and shown to minimize the PFIB artifacts induced by both the top surface sample-PFIB interaction and the non-uniform material density. An optimized post-image processing procedure is also described that ensures the accuracy of the reconstructed 3D surface model. The application of these new methods is demonstrated by applying them to extract the surface topography of BSi formed by reactive ion etching (RIE) consisting of 2 µm tall needles. An area of 320 µm2 is investigated with a controlled slice thickness of 10 nm. The reconstructed 3D model allows the extraction of critical roughness characteristics, such as height distribution, correlation length, and surface enhancement ratio. Furthermore, it is demonstrated that the particular surface studied contains regions in which under-etching has resulted in overhanging structures, which would not have been identified with other surface topography techniques. Such overhanging structures can be present in a broad range of BSi surfaces, including BSi surfaces formed by RIE and metal catalyst chemical etching (MCCE). Without proper measurement, the un-detected overhangs would result in the underestimation of many critical surface characteristics, such as absolute surface area, electrochemical reactivity and light-trapping.

Bilayer MoOX/CrOX Passivating Contact Targeting Highly Stable Silicon Heterojunction Solar Cells.

Molybdenum oxide (MoOX, X < 3) has been successfully demonstrated as an efficient passivating hole-selective contact in crystalline Si (c-Si) heterojunction solar cells because of its large bandgap (∼3.2 eV) and work function (∼6.9 eV). However, the severe performance degradation coming from the instability of the MoOX and its interfaces has not been well addressed. In this work, we started with a c-Si(p)/MoOX heterojunction solar cell that yielded a power conversion efficiency (PCE) of 15.86%, in which the MoOX film was synthesized by industry-compatible atomic layer deposition (ALD). The initial PCE dropped to 10.20% after 2 days because of severe migration of O and Ag at the MoOX/Ag interface. We solved this by the insertion of a CrOX layer between the MoOX layer and the Ag electrode. The solar cell was found to be stable for more than 8 months in air because of the suppression of interface degradation. Our work demonstrates an effective way of improving the stability of silicon solar cells with transition metal oxide carrier selective contacts.

Efficient light harvesting in hybrid quantum dot-interdigitated back contact solar cells via resonant energy transfer and luminescent downshifting.

In this paper, we propose a hybrid quantum dot (QD)/solar cell configuration to improve performance of interdigitated back contact (IBC) silicon solar cells, resulting in 39.5% relative boost in the short-circuit current (JSC) through efficient utilisation of resonant energy transfer (RET) and luminescent downshifting (LDS). A uniform layer of CdSe1-xSx/ZnS quantum dots is deposited onto the AlOx surface passivation layer of the IBC solar cell. QD hybridization is found to cause a broadband improvement in the solar cell external quantum efficiency. Enhancement over the QD absorption wavelength range is shown to result from LDS. This is confirmed by significant boosts in the solar cell internal quantum efficiency (IQE) due to the presence of QDs. Enhancement over the red and near-infrared spectral range is shown to result from the anti-reflection properties of the QD layer coating. A study on the effect of QD layer thickness on solar cell performance was performed and an optimised QD layer thickness was determined. Time-resolved photoluminescence (TRPL) spectroscopy was used to investigate the photoluminescence dynamics of the QD layer as a function of AlOx spacer layer thickness. RET can be evoked between the QD and Si layers for very thin AlOx spacer layers, with RET efficiencies of up to 15%. In the conventional LDS architecture, down-converters are deposited on the surface of an optimised anti-reflection layer, providing relatively narrowband enhancement, whereas the QDs in our hybrid architecture provide optical enhancement over the broadband wavelength range, by simultaneously utilising LDS, RET-mediated carrier injection, and antireflection effects, resulting in up to 40% improvement in the power conversion efficiency (PCE). Low-cost synthesis of QDs and simple device integration provide a cost-effective solution for boosting solar cell performance.

15% Efficiency Ultrathin Silicon Solar Cells with Fluorine-Doped Titanium Oxide and Chemically Tailored Poly(3,4-ethylenedioxythiophene):Poly(styrenesulfonate) as Asymmetric Heterocontact.

In order to achieve a high performance-to-cost ratio to photovoltaic devices, the development of crystalline silicon (c-Si) solar cells with thinner substrates and simpler fabrication routes is an important step. Thin-film heterojunction solar cells (HSCs) with dopant-free and carrier-selective configurations look like ideal candidates in this respect. Here, we investigated the application of n-type silicon/poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) HSCs on periodic nanopyramid textured, ultrathin c-Si (∼25 µm) substrates. A fluorine-doped titanium oxide film was used as an electron-selective passivating layer showing excellent interfacial passivation (surface recombination velocity ∼10 cm/s) and contact property (contact resistivity ∼20 mΩ/cm2). A high efficiency of 15.10% was finally realized by optimizing the interfacial recombination and series resistance at both the front and rear sides, showing a promising strategy to fabricate high-performance ultrathin c-Si HSCs with a simple and low-temperature procedure.