The open-circuit voltage (Voc) of Cd(Se,Te) solar cells has stalled near 900 mV, limiting further efficiency gains despite state-of-the-art devices achieving 23.1%. As a critical step in overcoming this barrier, we report here a physics-based device model that explicitly integrates defect interactions with potential fluctuation effects to explain this long-standing limitation. By coupling first-principles defect calculations with device-level simulations, we show that voltage deficits in As-doped Cd(Se, Te) are governed primarily by two mechanisms: bandgap fluctuations and electrostatic disorder induced by low dopant activation (2-4%). The combined effects of bandgap and electrostatic fluctuations contributes 90-130 mV Voc loss for associated Urbach energy of 27-30 meV, whereas the effect of spatial variations in defect transition energies is found to be small. These results identify viable pathways for breaking the 900 mV barrier, namely enhancing dopant activation and reducing alloy disorder. More broadly, the modeling framework established here provides the transferable approach to quantitatively connect defect interactions and real-space disorder with device performance, offering guidance for other polycrystalline thin-film semiconductors with band tails.

Cadmium selenide (CdSe), with a 1.7 eV bandgap, is a promising high-bandgap semiconductor for tandem solar cells, yet device efficiencies are hindered by rapid minority carrier recombination. Here, polycrystalline CdSe solar cells are investigated using radiative emission spectroscopy, time-resolved photoluminescence, and density functional theory, revealing fast (sub-nanosecond) minority carrier trapping by selenium vacancy-related defect states with densities of (5-50) x 1017 cm-3, limiting carrier mobility and increasing recombination. By reducing absorber thickness to approximate to 0.5 mu m, trapping effects are mitigated, achieving a record open-circuit voltage of 917 mV, a 165 mV improvement over prior reports. These findings clarify the role of Se vacancies in limiting CdSe solar cell performance and provide insights applicable to CdSe and CdSeTe thin-film photovoltaics. This work advances understanding of defect-mediated losses in II-VI semiconductors and suggests pathways for improving solar cell performance through defect control.

This study investigates the defect properties and doping limitations of group V elements in CdSe x Te1-x . Group V acceptor dopants are able to increase hole concentrations and thereby enhance solar cell performance. However, their doping efficiency is limited by the formation of compensating donor defects with concentrations that depend on alloy composition, processing temperatures, and Cd segregation into grain boundaries. We use density functional theory (DFT) and lattice Monte Carlo (LMC) to identify the lowest-energy Se/Te alloy configurations and to understand the impact of temperature and local alloy configuration on As/P defect formation. Continuum simulations were then employed based on the results of the DFT and LMC calculations to explore As/P dopability in CdSeTe under various growth temperatures, initial chemical potentials, and alloy compositions. Moreover, the segregation of Cd at grain boundaries was investigated to understand its impact on compensating defects. The results of our LMC simulations suggest that P should be a more effective p-type dopant than As in CdSeTe, while both dopants become less effective as Se content increases. Additionally, the continuum simulations highlight that both As and P doping can enhance p-type conductivity, and both of them can reach hole density on the order of 1016 cm-3 for 873 K initial growth temperature and 1017 cm-3 for 1173 K initial growth temperature. We find that managing chemical potentials and the formation of compensating defects is crucial for optimizing carrier density and dopant activation efficiency and ensuring they remain stable.

Inorganic halide perovskites have been considered as higher-stability alternatives to hybrid halide perovskites for optoelectronic applications. However, the formation of defects in these materials can significantly affect carrier density and recombination lifetime. Thus, understanding their properties is of great importance for achieving improved device performance. Through GGA+U calculations with spin-orbital coupling, we calculate the properties of native defects in CsPbI3/CsSnI3 and the defect-assisted non-radiative recombination rates for selected deep levels. Importantly, we discover that the iodine interstitials in CsPbI3 and tin interstitials in CsSnI3 have high non-radiative carrier capture rates. Besides, we also notice that complete defect structural relaxations are only achievable when large supercells are used. This work not only provides insights into potential non-radiative recombination pathways due to the formation of corresponding interstitials, but also highlights the importance of utilizing large supercells for defect calculations in perovskites.

Point defects directly impact solar cell device performance by limiting the carrier lifetime. In this work, density functional theory calculations are first used to determine the formation energy and diffusion energy barriers of dominant defects in Cu(In,Ga)Se-2. Next, continuum reaction-diffusion models are developed to analyze the redistribution of defects during manufacturing processes. We estimate defect capture cross sections using a first-principles-based approach. These cross sections are combined with our calculated defect profiles and trap energy levels to parameterize a Shockley-Read-Hall recombination model, which we implement into a device simulator to predict carrier lifetimes and device performance. In that way, a predictive technology computer aided design model is built to predict and optimize the performance of Cu(In,Ga)Se-2 solar cells.

Tin halide perovskites (THP) have attracted increasing interest as being a nontoxic alternative to lead halide perovskites (LHP) for applications in optoelectronics. Although having a more ideal bandgap of similar to 1.3 eV, the power conversion efficiency (PCE) for THP-based solar cell is still far from its theoretical maximum. As a result, understanding the defect formation and recombination processes in THP is of great importance. In this work, through density functional theory (DFT) calculations and careful treatment of finite-size corrections, we find that tin vacancies are the main source of self-p doping, and tin interstitials yield high nonradiative recombination (NR) rate. As a result, tin-moderate condition for chemical potentials should be implemented in order to maximize the efficiency. Our findings not only provide further insights into the nature of the NR pathway in THP, but also have implications for making high-performance THP-based solar cells.

GaN High Electron Mobility Transistors (HEMTs) plays a vital role in high-power and high-frequency electronics. Meeting the demanding performance requirements of these devices without compromising reliability is a challenging endeavor. Field Plates are employed to redistribute the electric field, minimizing the risk of device failure, especially in high-voltage operations. While machine learning is applied to GaN device design, its application to field plate structures, known for their geometric complexity, is limited. This study introduces a novel approach to streamlining the field plate design process. It transforms complex 2D field plate structures into a concise feature space, reducing data requirements. A machine learning-assisted design framework is proposed to optimize field plate structures and perform inverse design. This approach is not exclusive to the design of GaN HEMTs and can be extended to various semiconductor devices with field plate structures. The framework combines technology computer-aided design (TCAD), machine learning, and optimization, streamlining the design process. Meeting the demanding performance requirements of semiconductor devices without compromising reliability is a challenging endeavor. This study introduces a novel approach to streamlining the semiconductor design process. A machine learning-assisted design framework is proposed to optimize field plate structures and perform inverse design. The framework combines technology computer-aided design, machine learning, and optimization, streamlining the design process. image

CdTe and its alloy CdSeTe are widely used in optoelectronic devices, such as radiation detectors and solar cells, due to their superior electrical properties. However, the formation of point defects and defect complexes in these materials can significantly affect their performance. As a result, understanding the point defect formation and interaction processes in CdTe and CdSeTe alloy is of great importance. In recent years, density functional theory (DFT) calculations have emerged as a powerful tool for investigating the properties of defects in semiconductors. In this paper, we use the results of DFT+U calculations and continuum model to understand and optimize As and P dopability in CdSeTe alloy.

Quantum technologies based on quantum point defects in crystals require control over the defect charge state. Here we tune the charge state of shallow nitrogen-vacancy and silicon-vacancy centers by locally oxidizing a hydrogenated surface with moderate optical excitation and simultaneous spectral monitoring. The loss of conductivity and change in work function due to oxidation are measured in atmosphere using conductive atomic force microscopy (C-AFM) and Kelvin probe force microscopy (KPFM). We correlate these scanning probe measurements with optical spectroscopy of the nitrogen-vacancy and silicon-vacancy centers created via implantation and annealing 15-25 nm beneath the diamond surface. The observed charge state of the defects as a function of optical exposure demonstrates that laser oxidation provides a way to precisely tune the Fermi level over a range of at least 2.00 eV. We also observe a significantly larger oxidation rate for implanted surfaces compared to unimplanted surfaces under ambient conditions. Combined with knowledge of the electron affinity of a surface, these results suggest KPFM is a powerful, high-spatial resolution technique to advance surface Fermi level engineering for charge stabilization of quantum defects.

In this work, we present an exploration of deep learning models for predicting defect properties in cubic phase semiconductors. The nature of impurity energy levels strongly influences the performance of semiconductors in a wide range of applications, such as solar cells, field effect transistors, and qubits for quantum computing. In this work, we employ two types of deep learning models, a crystal defect graph neural network and a chemical environment-encoded artificial neural network, to predict defect properties. The models are trained on a data set of charge-dependent defect formation energies obtained from density functional theory computations and descriptors based on elemental properties, defect local environment, and relevant semiconductor properties. We assess the models' performance and showcase their capability in optimizing semiconductor devices, particularly when used in tandem with compositionally constrained thermodynamics and technology computer-aided design models.

CdTe and its alloy CdTeSe are widely used in optoelectronic devices, such as radiation detectors and solar cells, due to their superior electrical properties. However, the formation of defects and defect complexes in these materials can significantly affect their performance. As a result, understanding the defect formation and recombination processes in CdTe and CdTeSe alloy is of great importance. In recent years, density functional theory (DFT) calculations have emerged as a powerful tool for investigating the properties of defects in semiconductors. In this paper, we use DFT+U calculations to comprehensively study the properties of intrinsic defects as well as extrinsic defects induced by commonly used dopants, such as Cu and group V elements, in CdTe and CdTeSe alloy. This work provides insights into the effects of these defects on the electrical and optical properties of the material.

We extend the highly-parallelizable open-source electronic transport code TRANSEC (Feldmanet al2014Phys. Rev.B90035445;https://gitlab.com/computational-physics2/transec/) to perform real-space atomic-scale electronic transport calculations with periodic boundary conditions in the lateral dimensions. We demonstrate the use of TRANSEC in periodic Cu and Rh bulk structures and in large periodic Rh point contacts, in preparation to perform calculations of reflection probability across Rh grain boundaries.

To improve the performance of Cu(In,Ga)Se2 thin-film photovoltaic devices, a robust understanding of the dominant diffusion pathways of the alloy species In and Ga is needed. Here, the most probable defect complexes and mechanisms for In and Ga diffusion are identified with the aid of density functional theory. The binding energies and migration barriers for these complexes are calculated in bulk CuInSe2 and CuGaSe2. Analytic models and kinetic lattice Monte Carlo simulations are employed to predict the diffusivity of In and Ga under variations in composition and temperature. We find that a model based on coulombic interactions between group III antisites and vacancies on the Cu-sublattice produces results that match well with experiment.

Ytterbium-doped all-inorganic lead-halide perovskites (Yb3+:CsPb(Cl1-xBrx)(3)) generate near-infrared photoluminescence (PL) quantum yields exceeding 100% by quantum cutting. Experimental and computational studies have suggested complex dopant speciation in these materials arising from the formation of lattice defects needed to compensate for the excess charge of Yb3+ relative to Pb2+, but the relationship between quantum cutting and such speciation is still poorly understood. Here, we use cryogenic photoluminescence spectroscopy and density functional theory-assisted kinetic Monte Carlo simulations to investigate changes in Yb3+ speciation induced by anion (Cl- vs Br-) and trivalent-dopant (Yb3+ vs Gd3+) alloying in CsPb(Cl1-xBrx)(3) (0.00 <= x <= 1.00) perovskite nanocrystals (NCs). The experimental results reveal nonstatistical distributions of Yb-Cl and Yb-Br bonds in Yb3+:CsPb(Cl1-xBrx)(3) NCs. Monte Carlo simulations reproduce the experimental trends well and predict thermodynamic favorability for Yb3+ dopants to retain Cl- coordination, even in the presence of high lattice Br- concentrations. For a given lattice composition (e.g., CsPbCl3), low-temperature PL spectra reveal that the relative populations of three dominant Yb3+ species change substantially with Gd3+ codoping. These results further show that quantum cutting is largely insensitive to these differences in Yb3+ speciation, ruling out any "magic" configuration of Yb3+ ions and defects. These results are discussed in relation to the microscopic prerequisites for the concerted sensitization of two Yb3+ ions during quantum cutting. Overall, these findings highlight the mechanistic robustness of Yb3+:CsPb(Cl1-xBrx)(3) quantum cutting in lead-halide perovskites and provide deeper insight into the inner workings of this unique phenomenon.

Cadmium telluride (CdTe) and its alloy CdTeSe are widely used in optoelectronic devices, such as radiation detectors and solar cells, due to their superior electrical properties. However, the formation of defects and defect complexes in these materials can significantly affect their performance. As a result, understanding the defect formation and recombination processes in CdTe and CdTeSe alloy is of great importance. In recent years, density functional theory (DFT) calculations have emerged as a powerful tool for investigating the properties of defects in semiconductors. In this paper, we use DFT calculations to study the properties of arsenic extrinsic defects in CdTeSe alloy, providing insights into the effects of these defects on the electrical properties of the material. We found that AX centers formation is more favorable in the alloy, which can potentially reduce the dopability of As in CdTeSe alloy. Our findings not only provide new insights into the mechanisms of As extrinsic defects in CdTeSe but also have implications for the development of high-performance solar cells.

CsPb(Cl1-xBrx)3(0 <= x <= 1) nanocrystals and thinfilms doped with a series oftrivalent rare-earth ions (RE3+=Y3+,La3+,Ce3+,Gd3+,Er3+,Lu3+) have been prepared andstudied using variable-temperature and time-resolved photoluminescence spectroscopies. Wedemonstrate that aliovalent (trivalent) doping of this type universally generates a new and often-emissive defect state ca. 50 meV inside the perovskite band gap, independent of the specificRE3+dopant identity or of the perovskite form (nanocrystals vs thinfilms). Chloride-to-bromideanion exchange is used to demonstrate that this near-band-edge photoluminescence shifts withchanging band-gap energy to remain just below the excitonic luminescence for all compositionsof CsPb(Cl1-xBrx)3(0 <= x <= 1). Computations show that this shift stems from the effect of the changing lattice dielectric constantson a shallow defect-bound exciton. Microscopic descriptions of this dopant-induced near-band-edge state and its relation toquantum cutting in Yb3+-doped CsPb(Cl1-xBrx)3are discussed.

Progress in the application of machine learning techniques to the prediction of solid-state and molecular materials properties has been greatly facilitated by the development state-of-the-art feature representations and novel deep learning architectures. A large class of atomic structure representations based on expansions of smoothed atomic densities have been shown to correspond to specific choices of basis sets in an abstract many-body Hilbert space. Concurrently, tensor network structures, conventionally the purview of quantum many-body physics and quantum information, have been successfully applied in supervised and unsupervised learning tasks in computer vision and natural language processing. In this work, we argue that architectures based on tensor networks are well-suited to machine learning on Hilbert-space representations of atomic structures. This is demonstrated on supervised learning tasks involving widely available datasets of density functional theory calculations of metal and semiconductor alloys. In particular, we show that certain standard tensor network topologies exhibit strong generalizability even on small training datasets while being parametrically efficient. We further relate this generalizability to the presence of complex entanglement in the trained tensor networks. We also discuss connections to learning with generalized structural kernels and related strategies for compressing large input feature spaces.

Solid-state laser refrigeration of semiconductors remains an outstanding experimental challenge. In this work, we show that, following excitation with a laser wavelength of 532 nm, bulk diamond crystals doped with H3 centers both emit efficient up-conversion (anti-Stokes) photoluminescence and also show significantly reduced photothermal heating relative to crystals doped with nitrogen–vacancy (NV) centers. The H3 center in diamond is a highly photostable defect that avoids bleaching at high laser irradiances of 10–70 MW/cm2 and has been shown to exhibit laser action, tunable over the visible band of 500–600 nm. The observed reduction of photothermal heating arises due to a decrease in the concentration of absorbing point defects, including NV-centers. These results encourage future exploration of techniques for H3 enrichment in diamonds under high-pressure, high-temperature conditions for the simultaneous anti-Stokes fluorescence cooling and radiation balanced lasing in semiconductor materials. Reducing photothermal heating in diamond through the formation of H3 centers also opens up new possibilities in quantum sensing via optically detected magnetic resonance spectroscopy at ambient conditions.

We develop a framework powered by machine learning (ML) and high-throughput density functional theory (DFT) computations for the prediction and screening of functional impurities in groups IV, III-V, and II-VI zinc blende semiconductors. Elements spanning the length and breadth of the periodic table are considered as impurity atoms at the cation, anion, or interstitial sites in supercells of 34 candidate semiconductors, leading to a chemical space of approximately 12,000 points, 10% of which are used to generate a DFT dataset of charge dependent defect formation energies. Descriptors based on tabulated elemental properties, defect coordination environment, and relevant semiconductor properties are used to train ML regression models for the DFT computed neutral state formation energies and charge transition levels of impurities. Optimized kernel ridge, Gaussian process, random forest, and neural network regression models are applied to screen impurities with lower formation energy than dominant native defects in all compounds.