Predicted Band Gap Research Articles

Predicting band gaps of semiconductors with quantum chemistry

The following article gives a brief introduction to quantum chemistry and its application to the prediction of band gaps of inorganic and organic semiconductors. Two important quantum chemistry concepts —Density Functional Theory (DFT) and Coupled Cluster Theory (CC)— are shortly explained. These two concepts are used to calculate the optical and the transport band gap of a set of semiconductors modelled with an electrostatic embedding approach.

Electronic, optical and thermoelectric properties of boron-doped nitrogenated holey graphene.

We employ first principles calculations to investigate the electronic, optical, and thermoelectric properties of ten boron-doped nitrogenated holey graphene (NHG) monolayers. We find that most of the proposed structures remain stable during ab initio molecular dynamics simulations, in spite of their increased formation energies. Density functional theory calculations employing a hybrid functional predict band gaps ranging from 0.73 eV to 2.30 eV. In general, we find that boron doping shifts optical absorption towards the visible spectrum, and also reduces light reflection in this region. On the other hand, the magnitude of optical absorption coefficients are reduced. Regarding the thermoelectric properties, we predict that boron doping can enhance the figure of merit ZT of NHG by up to 55%. Our results indicate that boron-doped NHG monolayers may find application in solar cells and thermoelectric devices.

A progressive learning method for predicting the band gap of ABO3 perovskites using an instrumental variable

A progressive learning method with an instrumental variable and bond-valence vector sums was used to improve the bandgap prediction precision.

Nonempirical hybrid functionals for band gaps of inorganic metal-halide perovskites

Nonempirical hybrid functionals are investigated for band-gap predictions of inorganic metal-halide perovskites belonging to the class CsBX3, with B = Ge, Sn, Pb and X = Cl, Br, I. We consider both global and range-separated hybrid functionals and determine the parameters through two different schemes. The first scheme is based on the static screening response of the material and thus yields dielectric-dependent hybrid functionals. The second scheme defines the hybrid functionals through the enforcement of Koopmans' condition for localized defect states. We also carry out quasiparticle self-consistent GW calculations with vertex corrections to establish state-of-the-art references. For the investigated class of materials, dielectric-dependent functionals and those fulfilling Koopmans' condition yield band gaps of comparable accuracy (similar to 0.2 eV), but the former only require calculations for the primitive unit cell and are less subject to the specifics of the material.

Electronic and optical properties of Sr3X2 (X=N, P, As, Sb and Bi) compounds: first principles study

ABSTRACTDirect bandgap semiconductors are very essential to fulfil the demand for the advancement in optoelectronic devices. Therefore it is important to predict new potential candidates having such unique features. In current work, Sr3X2 (X=N, P, As, Sb and Bi) compounds have been reported for the first time by well trusted FP-APW+lo method. For the better prediction of the energy band gap, mBJ is used alongwith routine generalised gradient approximation (GGA). The results show small and direct energy band gaps at Γ-Γ symmetry points with magnitude in the range from 0.62 eV (Sr3P2) to zero energy band gap (Sr3Bi2). In partial density of state Sr-d state and X-p state are contributed in the band structure. The compounds show mostly covalent bonding nature. The frequecy dependent optical properties in the linear optical range are also investigated.

A Statistical Learning Framework for Accelerated Bandgap Prediction of Inorganic Compounds

This study deals with an application of machine learning (ML) techniques for electronic bandgap predictions of a host of entries in the open-source Materials Project (MP) database and inorganic perovskite compounds. Initially, a dataset of 4616 inorganic compounds having available experimental bandgap data was used to generate predictive ML models—support vector machine, k-nearest neighbors, random forest, kernel ridge regression (KRR), and artificial neural networks—requiring only compositional features based on simple elemental attributes. This was followed by identification of the most crucial features for the bandgap and an evaluation of various performance metrics against the dimensionality of the feature space. The trained KRR model having the highest accuracy was then regressed on more than 22,000 entries in the MP database, and the trends are elucidated. Subsequently, out-of-sample validation was performed on a couple of datasets containing several discovered halide perovskites, in conjunction with ab-initio investigations of the undiscovered ones. Finally, the optimal classification and regression models were employed to classify a dataset of 46,970 unknown inorganic halide perovskites into metals and nonmetals followed by bandgap predictions of the nonmetallic entries.

Ultranonlocality and accurate band gaps from a meta-generalized gradient approximation

Determining the electronic structure of solids and molecules from first principles computation is the task of Density Functional Theory. Systematically incorporating the derivative discontinuity into kinetic energy dependent functionals allows to accurately predict band gaps and to describe non-local charge transfer at semilocal computational cost.

First principles studies of CsLnCdTe3 (Ln = Gd−Tm) for green energy resources

The structural, electronic, optical and thermoelectric properties of alkali metal cadmium telluride CsLnCdTe3 (Ln = Gd–Tm) semiconductors are investigated using density functional theory. The structural parameters are calculated using generalized gradient approximation which show a close agreement with experimental values. The electronic properties are investigated using GGA + U approximation because the Hubbard potential (U) gives better results and are very reliable for band structure calculations. The predicted band gaps show direct band gap nature. The estimated band gap of CsTmCdTe3 is 1.93 eV which is in agreement with the experimental value of 2.01 eV. The narrow band gaps, high dielectric constants and high absorption ranges make these materials suitable hosts for solar energy conversion applications. BoltzTrap calculations are used to obtain thermoelectric parameters like seebeck coefficient, electronic and thermal conductivity. The high seebeck effect and large power factor values confirm the efficiency of these materials in thermoelectric converter technology.

Phillips-Inspired Machine Learning for Band Gap and Exciton Binding Energy Prediction.

In this work, inspired by Phillips's ionicity theory in solid-state physics, we directly sort out the critical factors of the band gap's feature correlations in the machine learning architected with the Lasso algorithm. Even based on a small 2D materials data set, we can fundamentally approach an accurate and rational model about the band gap and exciton binding energy with robust transferability to other databases. Our machine learning outputs can reveal the exact physics pictures behind the predicted quantity as well as the "secondary understanding" of the correlation between the approximated physics models in exciton. This work stresses the significant value of physics endorsement on the machine learning (ML) algorithm and provides a symbolic regression solution for the "few-shot" training scheme for ML technology in materials science. Moreover, physics-inspired secondary understanding could be an essential supplement for ML in scientific research fields.

Articial Neural Networks for Accurate Prediction and Analysis of Perovskite Bandgaps

Perovskites are a unique class of semiconductors with long carrier diffusion lengths which have shown remarkable promise for photovoltaic applications as their solar conversion efficiencies increased rapidly by nearly 20% over the past decade. Here we evaluate the efficacy of artificial neural networks for the analysis and prediction of 3D perovskite band gaps; which is the key parameter used in solar cell design. The set of input parameters for our artificial neural network includes descriptors such as s, p, d, and f orbital radii, electronegativity, octahedral factor, and formation energy. Furthermore, we apply a recursive feature correlation filter to eliminate features which highly correlate with other features in our model as well as features which have a low contribution with regards with regards to predicting bandgap. We find that splitting the dataset between training, testing, and validation sets in a 70:15:15 ratio provides a highly reliable model that yields a mean square error of less than 0.25 eV. We then validate the capability of artificial neural network by testing its predictability for bandgaps based on unseen inputs.

How Transition Metals Enable Electron Transfer through the SEI: Part I. Experiments and Butler-Volmer Modeling

Transition metal dissolution from high-voltage Li-ion battery cathodes disrupts the formation and performance of the solid-electrolyte interphase (SEI). SEI contamination by transition metals results in continual Li loss and severe capacity fade. Fundamental understanding of how metals undermine SEI passivation is necessary to mitigate this degradation. This two-part study interrogates the mechanisms by which transition metals facilitate through-film charge-transfer and SEI failure. Part I presents experimental results in which we intentionally contaminate SEIs with Mn, Ni, and Co. Rotating disk electrode voltammetry of a redox mediator quantifies how each metal impacts the charge-transfer characteristics of the SEI. A physics-based model finds that all three metals disrupt the electronic properties of the SEI more than the morphology. Surprisingly, the Butler-Volmer kinetics of charge-transfer through a Mn-contaminated SEI are an order of magnitude faster than for a Co-contaminated SEI, even with similar embedded metal concentrations. Such trends between metals are inconsistent with bandgap predictions from density functional theory, implying an alternative redox-cycling mechanism, which is mathematically developed and compared to experiment in Part II.

Assessment of the exact-exchange-only Kohn-Sham method for the calculation of band structures for transition metal oxide and metal halide perovskites

In the present paper we assess the performance of the exact-exchange-only (EXX) Kohn-Sham method to predict band structures of transition metal oxide and metal halide perovskites with the main emphasis on band gaps. For the considered set of prototypical systems with cubic structure, the performance of the EXX method for the prediction of band gaps is comparable with other popular methods for this purpose like density-functional calculations with the Heyd-Scuseria-Ernzerhof (HSE) hybrid functional and one-shot $GW$ (${G}_{0}{W}_{0}$). Comparison with experimental values suggests that in the case of perovskites the EXX method, like HSE and $GW$ methods, should be supplemented by a treatment of electron-phonon effects. The EXX method is computationally more efficient than ${G}_{0}{W}_{0}$ and also provides post-self-consistent access to one-electron wave functions and band structures within the complete Brillouin zone with much lower computational effort than both ${G}_{0}{W}_{0}$ and hybrid DFT methods. We confirm the importance of including the higher-lying semicore shells within the explicitly treated states for halide perovskites which was already noted by earlier studies. Calculations taking into account only the valence space without the semicore states, however, remain still a reasonable option in practice. Spin-orbit coupling effects seem to be slightly underestimated in DFT calculations both with respect to $GW$ and experiment. The source of this underestimation can be of both technical and formal nature, and this point requires more critical consideration. For ${\mathrm{CH}}_{3}{\mathrm{NH}}_{3}{\mathrm{PbI}}_{3}$, sizable Rashba spin splittings appear for the structure with the C-N bond of the methylammonium ion not lying along the long diagonal of the cubic unit cell.

Atom table convolutional neural networks for an accurate prediction of compounds properties

Machine learning techniques are widely used in materials science. However, most of the machine learning models require a lot of prior knowledge to manually construct feature vectors. Here, we develop an atom table convolutional neural networks that only requires the component information to directly learn the experimental properties from the features constructed by itself. For band gap and formation energy prediction, the accuracy of our model exceeds the standard DFT calculations. Besides, through data-enhanced technology, our model not only accurately predicts superconducting transition temperatures, but also distinguishes superconductors and non-superconductors. Utilizing the trained model, we have screened 20 compounds that are potential superconductors with high superconducting transition temperature from the existing database. In addition, from the learned features, we extract the properties of the elements and reproduce the chemical trends. This framework is valuable for high throughput screening and helpful to understand the underlying physics.

ThermoEPred-EL: Robust bandgap predictions of chalcogenides with diamond-like structure via feature cross-based stacked ensemble learning

Implementation on rapid and accurate bandgap prediction has great practical implications for a range of applications. While quantum mechanical computations are enormously computation-time intensive, using informatics-based statistical learning approaches can be a promising alternative due to its availability, power and relatively limited cost of high-performance computational equipment. Here we demonstrate a systematic ensemble learning model which integrates a novel feature-engineering approach and a robust learning framework for predicting bandgaps of one series of typical thermoelectric materials: chalcogenides with diamond-like structure. After combining a feature crossing technique with a feature selection method, the proposed optimal descriptor set is identified by searching the feature space of 23,454 descriptors stemmed from the elemental features. Stable statistic-based feature selection methods are applied to identify the most crucial and relevant descriptors. The stacked ensemble learning model, which integrates the advantages of three different level-0 models (LASSO, SVR, and AdaBoost) and one level-1 model (GBDT), obtains 90.48% prediction accuracy, thus improving model accuracy and robustness. The results demonstrate the interpretability and generalizability of the stacked ensemble model, which can be applied to bandgap predictions in other material systems, thereby accelerating the design and optimization process for discovering new functional materials.

Efficient band gap prediction of semiconductors and insulators from a semilocal exchange-correlation functional

A semilocal exchange-correlation functional is proposed with the efficient prediction of the solid-state band gap. The underlying construction of the exchange functional is based on the modeling of the exchange hole and constructing the exchange energy functional. Being a meta-generalized gradient approximation (meta-GGA) level functional, it holds the key feature of the derivative discontinuity through the generalized Kohn-Sham (gKS) formalism. We validate our construction by demonstrating the functional performance for solid-state band gaps and comparing it with the other meta-GGA and hybrid functionals. It is shown that for the semiconductors having narrow, and moderate band gaps, as well as for layered materials the present functional performs as accurately as (or comparable to) the expensive hybrid functional, while for wide-band gap solids, it outperforms the hybrid functional. This indicates that the present functional holds promise for the multiscale modeling of materials with a very low computational cost. Also, the underlying construction is practically very useful as it can be easily implemented in any density functional platform that supports the gKS formalism.

Homogenization Approach and Bloch-Floquet Theory for Band-Gap Prediction in 2D Locally Resonant Metamaterials

This paper provides a detailed comparison of the two-scale homogenization method and of the Bloch-Floquet theory for the determination of band-gaps in locally resonant metamaterials. A medium composed by a stiff matrix with soft inclusions with 2D periodicity is considered and the equivalent mass density of the homogenized medium is explicitly obtained both for in-plane and out-of-plane wave propagation through two-scale asymptotic expansion. The intervals of frequency where the effective mass is negative identify the band-gaps of the material. The Bloch-Floquet problem is then considered and, through an asymptotic analysis, its is shown that it leads to the same prediction of the band-gaps. The results are confirmed by some examples and the limits of the asymptotic approach are explicitly given and numerically verified.

Band Gap Prediction for Large Organic Crystal Structures with Machine Learning

AbstractMachine‐learning models are capable of capturing the structure–property relationship from a dataset of computationally demanding ab initio calculations. Over the past two years, the Organic Materials Database (OMDB) has hosted a growing number of calculated electronic properties of previously synthesized organic crystal structures. The complexity of the organic crystals contained within the OMDB, which have on average 82 atoms per unit cell, makes this database a challenging platform for machine learning applications. In this paper, the focus is on predicting the band gap which represents one of the basic properties of a crystalline material. With this aim, a consistent dataset of 12 500 crystal structures and their corresponding DFT band gap are released, freely available for download at https://omdb.mathub.io/dataset. An ensemble of two state‐of‐the‐art models reach a mean absolute error (MAE) of 0.388 eV, which corresponds to a percentage error of 13% for an average band gap of 3.05 eV. Finally, the trained models are employed to predict the band gap for 260 092 materials contained within the Crystallography Open Database (COD) and made available online so that the predictions can be obtained for any arbitrary crystal structure uploaded by a user.

Accurate Band Gap Predictions of Semiconductors in the Framework of the Similarity Transformed Equation of Motion Coupled Cluster Theory.

In this work, we present a detailed comparison between wave-function-based and particle/hole techniques for the prediction of band gap energies of semiconductors. We focus on the comparison of the back-transformed Pair Natural Orbital Similarity Transformed Equation of Motion Coupled-Cluster (bt-PNO-STEOM-CCSD) method with Time Dependent Density Functional Theory (TD-DFT) and Delta Self Consistent Field/DFT (Δ-SCF/DFT) that are employed to calculate the band gap energies in a test set of organic and inorganic semiconductors. Throughout, we have used cluster models for the calculations that were calibrated by comparing the results of the cluster calculations to periodic DFT calculations with the same functional. These calibrations were run with cluster models of increasing size until the results agreed closely with the periodic calculation. It is demonstrated that bt-PNO-STEOM-CC yields accurate results that are in better than 0.2 eV agreement with the experiment. This holds for both organic and inorganic semiconductors. The efficiency of the employed computational protocols is thoroughly discussed. Overall, we believe that this study is an important contribution that can aid future developments and applications of excited state coupled cluster methods in the field of solid-state chemistry and heterogeneous catalysis.

Acoustic bandgap characteristics of a duct with a cavity-backed and strip mass-attached membrane array mounted periodically

In this paper, acoustic bandgap analysis of a duct with a cavity-backed and strip mass-attached membrane (CMM) array periodically mounted on its sidewalls is investigated. An analytical model for the bandgap prediction of such periodic vibro-acoustic coupling system has been established via the energy principle in conjunction with transfer matrix method. Results indicate that the proposed CMM resonator can provide a total noise reflection in some frequency and periodic CMM resonators can give rise to attractive Bragg bandgaps. The location and width of the Bragg- and the resonant-gaps can be effectively and expediently adjusted by changing the mass position. Multiple CMM resonators with artificial imperfection via changing the mass position and weight can further strengthen the attenuation effects. The transmission loss spectrum is stretched by several remarkable peaks corresponding to the resonant frequencies of these resonators. It can be concluded that the attached mass can provide an effective and convenient means to enhance and adjust the noise absorption bandwidth of such periodic structure in low frequency.

Adjustable potential probes for band-gap predictions of extended systems through nonempirical hybrid functionals

We describe a nonempirical procedure for achieving accurate band gaps of extended systems through the insertion of suitably defined potential probes. By enforcing Koopmans' condition on the resulting localized electronic states, we determine the optimal fraction of Fock exchange to be used in the adopted hybrid functional. As potential probes, we consider point defects, the hydrogen interstitial, and various adjustable potentials that allow us to vary the energy level of the localized state in the band gap. By monitoring the delocalized screening charge, we achieve a measure of the degree of hybridization with the band states, which can be used to improve the band-gap estimate. Application of this methodology to AlP, C, and MgO yields band gaps differing by less than 0.2 eV from experiment.