Dft Implementation Within FLAPW Method Enables Accurate Calculations and Validates $GW$ Results

Understanding the electronic structure of materials is fundamental to designing new technologies, and researchers continually refine the computational methods used to explore this realm. Wejdan Beida, Gustav Bihlmayer, and Christoph Friedrich, alongside colleagues at Forschungszentrum Jülich and RWTH Aachen University, have developed an advanced computational approach by implementing density-functional theory within the all-electron full-potential linearized augmented-plane-wave method. This work represents a significant step forward in accurately modelling materials, as it avoids approximations commonly used in these calculations and allows for direct comparison with experimental data and more sophisticated theoretical techniques. The resulting method promises to improve our understanding of material properties and accelerate the discovery of novel materials with tailored functionalities.

The research employs the full-potential linearized augmented-plane-wave FLAPW method, implemented in the FLEUR code, to investigate materials with strong electronic interactions. This approach, known as density functional theory plus U plus V DFT+U+V, extends standard density functional theory by accounting for the effects of electron correlation, crucial for accurately describing the behavior of many materials. The method incorporates both onsite Coulomb interaction, represented by U, and intersite Coulomb interaction, represented by V, offering a more complete picture of electronic behavior than traditional approaches.

DFT+U Validation for Semiconductors and Oxides

This research presents a thorough assessment of the DFT+U method for calculating the electronic structure and properties of semiconductors like silicon and germanium, and transition metal oxides like nickel oxide. Scientists aimed to validate the accuracy of DFT+U by comparing its results with experimental data and more advanced calculations. The study demonstrates that DFT+U significantly improves the prediction of band gaps in semiconductors compared to standard DFT, with accuracy dependent on selecting an appropriate value for the U parameter. For nickel oxide, DFT+U accurately describes its electronic structure, magnetic properties, and structural parameters, aligning well with data from photoemission spectroscopy and X-ray absorption spectroscopy.

First Principles Determination of U and V Parameters

Scientists have developed a method for accurately determining the U and V parameters, which quantify the strength of electron interactions, directly from the material’s electronic structure. Using the constrained random-phase approximation cRPA within the DFT+U+V framework and the FLEUR code, they calculated these parameters without relying on empirical inputs. The team investigated the impact of different mathematical representations of the electronic structure, specifically muffin-tin functions and maximally localized Wannier functions, on the calculated interaction parameters. This implementation allows for a more accurate description of materials where both onsite and intersite electronic interactions are important, extending beyond traditional models. The research demonstrates that including the intersite V term improves the description of materials across diverse classes, including graphene, silicon, germanium, and charge-transfer insulators like nickel oxide, leading to a more complete understanding of their electronic behavior.

DFT+U+V Improves Material Property Predictions

This work presents a new implementation of density functional theory combined with the DFT+U+V approach, performed using the all-electron full-potential linearized augmented-plane-wave method. Researchers successfully extended the standard DFT+U framework by incorporating an intersite Coulomb interaction, V, and implemented this within the FLEUR code, determining the interaction parameters using the constrained random phase approximation. The methodology was thoroughly tested on a diverse range of materials, including graphene, silicon, germanium, and nickel oxide, representing both covalent and ionic bonding, and both non-magnetic and magnetic systems. Results demonstrate that the inclusion of the V parameter generally improves the accuracy of calculated physical properties, such as lattice constants and electronic band gaps, bringing them into closer agreement with experimental values. This advancement provides a more accurate and reliable method for predicting the properties of complex materials and furthering our understanding of their behavior.