![](/logos/buttonlogos/cross.png)
![](/logos/buttonlogos/cross.png)
research papers
Web-CONEXS: an inroad to theoretical X-ray absorption spectroscopy
aDiamond Light Source, Harwell Science and Innovation Park, Didcot, Oxfordshire OX11 8UQ, United Kingdom, and bChemistry – School of Natural and Environmental Science, Newcastle University, Newcastle Upon-Tyne NE1 7RU, United Kingdom
*Correspondence e-mail: joshua.elliott@diamond.ac.uk, tom.penfold@newcastle.ac.uk, sofia.diaz-moreno@diamond.ac.uk
Accurate analysis of the rich information contained within X-ray spectra usually calls for detailed electronic structure theory simulations. However, density functional theory (DFT), time-dependent DFT and many-body perturbation theory calculations increasingly require the use of advanced codes running on high-performance computing (HPC) facilities. Consequently, many researchers who would like to augment their experimental work with such simulations are hampered by the compounding of nontrivial knowledge requirements, specialist training and significant time investment. To this end, we present Web-CONEXS, an intuitive graphical web application for democratizing electronic structure theory simulations. Web-CONEXS generates and submits simulation workflows for theoretical X-ray absorption and X-ray emission spectroscopy to a remote computing cluster. In the present form, Web-CONEXS interfaces with three software packages: ORCA, FDMNES and Quantum ESPRESSO, and an extensive materials database courtesy of the Materials Project API. These software packages have been selected to model diverse materials and properties. Web-CONEXS has been conceived with the novice user in mind; job submission is limited to a subset of simulation parameters. This ensures that much of the simulation complexity is lifted and preliminary theoretical results are generated faster. Web-CONEXS can be leveraged to support beam time proposals and serve as a platform for preliminary analysis of experimental data.
Keywords: Web-CONEXS; X-ray absorption spectroscopy; XAS; first-principles simulation; density functional theory; DFT; X-ray absorption near-edge structure; XANES.
1. Introduction
X-ray spectroscopic capabilities at third- and fourth-generation light sources such as synchrotrons and X-ray free-electron lasers (XFELs) are increasing at a rapid rate due to the development and deployment of high-brilliance light sources and flexible modular beamlines, which permit highly detailed, complex, time-resolved investigations in situ and under operando conditions (Dent et al., 2009; Diaz-Moreno et al., 2009
; Mosselmans et al., 2009
; Taylor et al., 2013
; Figueroa et al., 2018
; Diaz-Moreno et al., 2018
; Hayama et al., 2021
). X-ray spectroscopic data obtained on these beamlines can unravel both the structural and the electronic properties of technologically relevant systems across the chemical (Bartlett et al., 2019
; Boada et al., 2022
; Ma et al., 2022
), physical (Arias-Egido et al., 2021
; Nicholls et al., 2022
; Das et al., 2022
) and materials sciences (García et al., 2021
; Alkhalifah et al., 2022
; Foster et al., 2022
). These massive advances in technological capability increase the focus on the far-from-trivial task of analysing and interpreting the wealth of information into the electronic, geometric and spin structure of matter that X-ray spectra contain.
This challenge can be offset through theoretical materials modelling and simulations based on electronic structure theory (Timrov et al., 2020; Kellett et al., 2020
; Groot et al., 2021
; Begum et al., 2021
; Rankine & Penfold, 2021
). Theoretical support for comes in a range of different flavours which differ primarily in the balance between computational intensity and accuracy they offer: methods include atomic multiplet theory (Haverkort et al., 2012
), multiple scattering theory (MST), single-particle approaches based on ab initio electronic structure such as density functional theory (DFT) (Besley, 2021
) and Hartree–Fock (HF) (Norman & Dreuw, 2018
), and methods that investigate many-body phenomena such as time-dependent DFT [TD-DFT (Norman & Dreuw, 2018
; Besley, 2021
; Bussy & Hutter, 2021
)], many-body perturbation theory [MBPT (Vinson et al., 2011
; Gilmore et al., 2015
; Vorwerk et al., 2017
, 2019
)] as well as those based on correlated wavefunctions (Maganas et al., 2017
, 2018
; Norman & Dreuw, 2018
; Chantzis et al., 2018
; Pinjari et al., 2016
).
The development and curation of such a wide range of available theoretical approaches have ensured that theoretical
has kept pace with experimental advances. However, for non-experts using as a characterization tool, this diversity in theoretical approaches can make it difficult to know where to start. To illustrate this point, consider a scenario in which X-ray absorption near-edge structure (XANES) spectra have been collected at a synchrotron facility, and electronic structure theory simulations could be employed to help interpret the data. To perform first-principles simulations, users must make several informed decisions which could serve as severe obstacles to those without experience or access to an experienced collaborator. In particular, for any given material, the user must select an appropriate method for the simulation. The method is one of the different flavours of theory outlined above. It is possible to draw on the literature to help with this selection, yet without an intuitive feel for (i) computational viability, (ii) the validity of underpinning physical approximations and (iii) subtle differences between similar systems that can lead to vastly different requirements, then it can be challenging to evaluate the transferability of previous work.The selection of the method is only the first barrier; other constraints play a role in the limited uptake of theoretical modelling. The next choice is the specific calculator for the simulation. The calculator is the software or computer code that implements the physical equations and materials relationships. Again, usually, some direction can be taken from the literature; however, as in the case of point (iii), the user must also recognize that the nature of the material will also dictate the choice of the calculator as well as the availability of features across different software packages. For example, crystalline materials with well defined repeat unit cells lend themselves to calculators that implement periodic boundary conditions and basis sets based on plane waves such as Quantum ESPRESSO (QE), (Giannozzi et al., 2009, 2017
, 2020
) CASTEP (Clark et al., 2005
) and Exciting (Gulans et al., 2014
). On the other hand, finite systems such as molecules and clusters of atoms can instead leverage a calculator that is based on localized (Gaussian) basis functions and open boundary conditions such as ORCA (Neese, 2012
, 2018
, 2022
), NWChem (Valiev et al., 2010
; Apra et al., 2020
) and QChem (Epifanovsky et al., 2021
). In addition, the choice of calculator requires the user to fully understand the physical equations, including key approximations, which are implemented. This forces the user to develop a working knowledge of the mapping between the model and materials relationship and the set of code-specific input parameters. Usually, each calculator has its own set of input parameters and syntax, meaning the user must invest time to learn the unique calculator language. Finally, the user will likely need access to an advanced or high-performance computing (HPC) facility to conduct theoretical simulations. However, note that this is only sometimes the case; software such as FEFF (Rehr et al., 2010
), MXAN (Benfatto et al., 2021
) and FDMNES (Bunău & Joly, 2009
) (in Green's function mode) that employ MST, and Quanty/Crispy which leverage semi-empirical atomic multiplet theory (Haverkort et al., 2012
), all circumvent HPC requirements. Yet software implementing higher-precision approaches like full potential DFT and beyond-DFT simulations demands HPC to go beyond crude small-scale models. Although gaining access to such a facility is another obstacle, the user must also understand and conform to the HPC syntax for running simulations within a batch job environment. This is, in fact, a larger issue than it appears at first glance because non-conformity may lead to improper and disruptive use of computing resources.
To summarize, a non-expert user faces four critical obstacles before running ab initio simulations: choice of model, choice of calculator, familiarity with calculator language and access and familiarity with HPC. To address these obstacles, in this work, we introduce Web-CONEXS, an intuitive and user-friendly web application deployed on the Diamond Light Source (DLS) ISPyB platform. Web-CONEXS mitigates many difficulties associated with ab initio simulations by directly interfacing with calculators and the HPC facility. It tackles challenges with the choice of model and calculator by limiting the number of available options, thereby reducing the number of decisions the user has to make. The integrated calculators have been tailored so that they can be selected according to molecular or crystalline materials, with a third option for large systems. Web-CONEXS provides a viable minimal set of input parameters for any given material and calculator, which can be used as a simulation starting point. Finally, Web-CONEXS automates job submissions and returns the results to the user, ultimately providing the foundation for further computation.
This manuscript is organized as follows: in the next section, we outline the Web-CONEXS workflow from the user perspective and based on the machinery hidden from the user. We then provide technical details on the implementation and discuss each interfaced calculator. In the final section, we highlight a case study for each calculator and describe how to build on the Web-CONEXS example to achieve production-level simulation results.
2. Implementation
Web-CONEXS has been conceived and implemented with the non-specialist user in mind. Its target is to provide support for users with limited or no experience of simulation and HPC to gain experience which can subsequently be transferred to independent use of HPC. It is not expected to be used by researchers with experience in the simulation of materials using electronic structure-based approaches or for detailed analysis of data. Web-CONEXS can be accessed by past and present DLS synchrotron users through the ISPyB experiment portal.
2.1. Web-CONEXS workflow
This section outlines the workflow for an Web-CONEXS platform. Fig. 1 depicts a flowchart for the process. Steps that relate directly to the user are connected with dashed arrows and steps hidden from the user are connected with solid arrows.
![]() | Figure 1 Schematic of the Web-CONEXS user simulation workflow. Dashed arrows denote the steps taken by a Web-CONEXS user, solid arrows denote steps hidden from the user. Blue and red dashed boxes group together steps hidden from the user: in blue are the steps taken to build the input file and in red are the steps taken to execute the calculation. |
The first step for the user is selecting one of the available calculators. These calculators are external software packages that implement the physical equations and materials relationships. Currently, Web-CONEXS supports the following calculators: ORCA (Neese, 2012, 2018
, 2022
) for DFT and TD-DFT on molecular systems and cluster type calculations, FDMNES (Joly, 2001
; Joly et al., 2009
; Bunău et al., 2021
) for finite differences and Green's functions approaches, and QE (Giannozzi et al., 2009
, 2017
, 2020
) for DFT on periodic systems. Further details on the supported features for each calculator are provided in the following section.
On the Web-CONEXS platform, a user case can be fully described using atomic coordinates and unit-cell vectors for crystalline systems. The user is asked to provide the atomic coordinates atom-by-atom, by reading an existing file in either .xyz or .cif format, or through a calculator-specific input file. In some instances, atomic coordinates can be provided by a reference to an external database of simulation data: Web-CONEXS currently interfaces to the Materials Project API (Jain et al., 2013; Ong et al., 2015
), which allows for the definition of a vast catalogue of experimental and theoretical materials structures based solely on a materials project identification code. Based on the input structure, Web-CONEXS returns a sensible, minimal set of simulation parameters for the user to verify. The permitted input options for each calculator have been limited to avoid overly complex and excessively long calculations, which are outside the remit of Web-CONEXS. A detailed list of the allowed parameters is provided in the supporting information. In addition to the calculator parameters, there are also compute options related to the batch job scheduler. These options control the calculations' size and the workload distribution across compute cores working in parallel. Web-CONEXS sets these options automatically to decrease the demand on the user knowledge requirement of specific HPC architecture and syntax.
The user can then submit the calculation to a computing facility for execution. During the simulation, the user can query the batch job status and stop the simulation at any point through the Web-CONEXS user interface. On completion, Web-CONEXS processes the raw output and returns a text file containing energy versus intensity via email to the user.
In the background, the user workflow is augmented by several steps, denoted by solid arrows and grouped by the blue and red dashed boxes in Fig. 1. Firstly, on inputting the atomic coordinates, Web-CONEXS takes several steps to provide the set of minimal input parameters. For QE calculations, the user may want to provide a reference to their material from the Materials Project database (Jain et al., 2013
; Ong et al., 2015
). In this case, not only can the coordinates be derived from the database reference, but also a guess of the calculator (QE) input parameters, which can be derived from the first-principles simulations also stored in the database. These externally sourced parameters can be translated to the QE namespace before being presented to the user. In practice, this leads to overly complex input files and simulations that exceed the compute resources currently available to Web-CONEXS. In the present implementation, Web-CONEXS derives the system geometry, lattice vectors and the k-point mesh used to sample the first from the Materials Project database entries. Conversely, Web-CONEXS uses system information to determine the minimal calculator parameters if the user supplies coordinates.
When the user submits a compute job, the job is scheduled to run on resources that are part of the IRIS infrastructure. IRIS is a collaborative project that aims to provide computing infrastructure to DLS by allocating CPU cores, storage and GPUs. The Web-CONEXS project has a resource allocation of 690 CPUs and 2.9 TB of memory. The computing infrastructure is hosted by the STFC, which provides resources using an Openstack based Cloud. These consist of virtual machines (VMs), object storage and block storage. The IRIS infrastructure is virtualized and suitable for high-throughput computing (HTC), but it does not provide high-performance/low-latency storage or networks and, hence, is not an HPC. To facilitate these workloads inside DLS, a workload scheduler different from the Gridengine based system is used. This system is called HTCondor.
HTCondor has semantics similar to Gridengine, plus a few extra features that make it ideal in a Cloud environment. For example, it has the notion of jobs (which can be started and stopped), input files that HTCondor can transfer to the location of the running job and output files which HTCondor can return to the user. HTCondor submits jobs to a submitter via ssh or Python API. The submitter matches the job to resources advertised by the worker nodes or `executors' in HTCondor terminology. The HTCondor workers run on IRIS VMs, whilst the submitter runs inside DLS.
Before the job is finalized and submitted to the HTCondor submitter, the Web-CONEXS backend generates any files needed for the calculator's execution and for reporting the results. These include input files for a chosen calculator, instructions to run a containerized calculator(s), and transfer input and output files. Containerization carries the benefit that each simulation carried out on the Web-CONEXS platform is packaged, fully portable and reproducible across any computing architecture at any point in time. On completion, output files from the calculator are transferred back by HTCondor and then returned to the user as an email.
2.2. Calculators
2.2.1. ORCA
ORCA is a quantum chemistry software package that performs simulations in open boundary conditions (Neese, 2012, 2018
, 2022
). This makes it an ideal calculator for molecules, clusters and nanoparticles. By default, ORCA employs Gaussian basis sets for expanding the electronic wavefunctions and can treat materials using an all-electron approach or with pseudopotentials. To calculate the spectrum at the K- and L-edges, ORCA implements both TD-DFT and correlated wavefunction-based methods, including singles (CIS), complete active space self-consistent field (CASSCF) and coupled cluster theory, including singles and doubles (CCSD) (Maganas et al., 2017
, 2018
; Chantzis et al., 2018
).
On the Web-CONEXS platform, it is possible to compute the and spectra with ORCA using TD-DFT and DFT, respectively; the workflow for such a calculation is straightforward. In one step, the spectrum can be calculated by specifying a TD-DFT or DFT simulation and fixing the excitation range from one of the core orbitals to the virtual orbitals. To do this the user is only required to provide structural information and then select from appropriate options on the interface for the system they will investigate. This structure may be loaded from an .xyz or .gmnt file, or inserted manually as Cartesian coordinates. For an ORCA simulation, the user should specify the total charge and of the system, and the orbitals expected to participate in the X-ray transitions. In other words, the orbital range that describes the initial and final states for absorption.
Additional and more specialist options that the user is requested to provide are the basis set, exchange-correlation potential and solvent environment. To aid in this selection, Web-CONEXS limits the choice to three basis sets and three exchange-correlation potentials. These have been chosen to provide a reasonable first guess at the XAS/XES spectrum in most cases, in particular with the option to select the triple zeta def2-TZVP basis set for the simulation of higher-Z elements in the periodic table. Further details can be found in the supporting information. In mode, the acceptor core orbitals are set to the lowest-energy orbitals, spin channels to α and β and with interactions are turned on. Note that in some cases the transitions of interest could be to higher energy orbitals. In this instance, the user must manually modify the core orbital within the input and perform an additional calculation outside the platform.
2.2.2. Quantum ESPRESSO
Web-CONEXS provides an interface to the pw.x and xspectra.x codes of the QE software suite for the simulation of ordered crystalline systems (Giannozzi et al., 2009, 2017
, 2020
). QE is a collaborative open-source project implementing Kohn–Sham DFT, expanding electronic wavefunctions using a plane wave basis set and treating the highly oscillatory core level states within the pseudopotential approximation. The XSpectra package (Taillefumier et al., 2002
; Gougoussis et al., 2009a
,b
; Bunău & Calandra, 2013
), which is utilized to simulate core-level excitations, computes the spectrum at the K- and L-edges by solving Fermi's Golden rule with the recursive Lanczos–Haydock extended fraction approach. For the core levels that act as the initial states for the excitation, XSpectra leverages Blochl's projector augmented wave (PAW) method to recover all-electron wavefunctions (Blöchl, 1994
).
Obtaining the QE is a multistep process; in the simplest case, the first step is to perform a single-point self-consistent field calculation to compute the electronic density of the ground state. This is done with the pw.x code. The second step is to use the Lanczos–Haydock algorithm to calculate the within the dipole approximation, for a given atom in the system using the xspectra.x code. The QE approach also offers additional complexity by including a (fractional) core-hole potential (Bunău & Calandra, 2013), anisotropic polarization of the transition matrix operator (Taillefumier et al., 2002
), the calculation of quadrupole contributions to the spectrum (Gougoussis et al., 2009
b) and several different Hubbard correction schemes (Gougoussis et al., 2009
b; Timrov et al., 2020
). Though Web-CONEXS can undoubtedly be an entry point for these more complex simulations, they remain outside the scope of the platform.
The minimum requirements for calculating the QE software suite are a description of the structure, a cutoff that defines the size of the plane wave basis set and a set of Gauge-including PAW (GIPAW) pseudopotentials (Pickard & Mauri, 2001) for describing the core levels. The user must provide the for their material in Cartesian coordinates; this may be provided on an atom-by-atom basis or by reading in a .xyz file. The user must also provide the set of vectors. The Web-CONEXS QE interface also provides access to the Materials Project database through the Materials API (MAPI); structural information may be imported directly to Web-CONEXS by providing a Materials Project ID number. The size of the basis set and the description of the core electrons are fixed in Web-CONEXS, which includes the complete set of PBESol pseudopotentials from A. Dal Corso's PSlibrary (version 1.0.0; Dal Corso, 2014
) modified to include GIPAW reconstructions for each element. Furthermore, the user can select from a set of predefined input parameters depending on whether the material of interest is a metal, semiconductor or insulator; more details relating to these parameters are provided in the supporting information. For the calculations, the user can then select the absorbing species and edge to be computed based on the input structure.
2.2.3. FDMNES
For the simulation of large molecular, crystalline or low-dimensional systems, Web-CONEXS has an interface to the FDMNES code (Joly, 2001; Joly et al., 2009
; Bunău et al., 2021
). FDMNES can be described as user-friendly software for the computation of the XANES spectrum through the solution of Fermi's Golden rule up to and including electronic octupole and magnetic dipole transitions, implementing the fully relativistic local approximation (Joly, 2001
). FDMNES performs these calculations in real space, building clusters (of user-defined size) around absorbing atoms in the molecule or Of particular importance, FDMNES offers two methods for the computation of the spectrum: first, with MST, which provides a way to rapidly compute (at low precision) the spectrum with Green's functions; and second, with a finite difference method (FDM) approach. Though computationally efficient and not requiring access to HPC, MST calculations rely on the Green's function formalism on top of a muffin-tin (MT) potential that can result in limited agreement with experimental spectra. Instead, the FDM overcomes limitations associated with the MT approximation, allowing for an unconstrained potential to lead to better agreement with experimental measurements at the cost of computational complexity and the requirement for HPC.
A typical simulation requires as input: the structure of the system to be investigated, the type of calculation to be performed (either MST or FDM), the size of the cluster around the absorbing atoms, and the absorbing element and edge that will be computed. Many other input parameters may be added or modified to systematically improve the simulation spectrum with respect to the experimental one (Joly et al., 2009). Within Web-CONEXS, the user must provide information on the structure of their system. Geometric data may be provided from an existing file; currently supported file formats are .cif and .pdb. Alternatively, the atomic coordinates (Cartesian) and simulation cell data may be entered manually. Additionally, the user must specify either the MST or the FDM simulation type, absorbing element and edge. Web-CONEXS then fixes several parameters automatically to ensure the feasibility of the simulation on the IRIS compute cluster: the cluster sizes surrounding the absorbing atom are fixed at 6 Å; moreover, the spectrum is computed in the energy range −10 eV ≤ E ≤ 50 eV surrounding the expanding the electronic transition operator up to the quadrupole approximation.
3. Case studies
3.1. Pre-edge features in spin-crossover complexes
The ORCA software package can be utilized for the investigation of molecular systems; we used the Web-CONEXS ORCA interface to simulate the pre-edge features of the Fe K-edge in the organometallic spin crossover complex [Fe(bpy)3]2+. The geometry of the [Fe(bpy)3]2+ complex is depicted in Fig. 2(a), the Fe2+ ion is octahedrally coordinated, with the five 3d orbitals split by crystal field effects into the eg and t2g sets. Fe2+ has the valence 3s23p63d64s0; in the low-spin (LS) configuration, the t2g set is filled while the eg set is empty. This means that the pre-edge features in the Fe K-edge can arise from 1s → 3d transitions to the empty eg set and (MLCT) excitations where there is sufficient spatial overlap of the Fe and ligand π-orbitals.
![]() | Figure 2 Results of the case studies conducted using the Web-CONEXS platform. (a) Plot comparing the theoretical and experimental Fe K-edge for the organometallic complex [Fe(bpy)3]2+ in the low-spin case. (b) Plot comparing the Zn L3-edge of ZnO measured experimentally computed with Web-CONEXS and with fully converged parameters. (c) Plot of the Ag K-edge XANES spectra of Ag nanoparticles measured experimentally and computed using the FDMNES code. In all cases, experimental spectra are plotted with a dashed black line, the Web-CONEXS result is plotted with solid orange lines and an `accurate' calculation is plotted with a solid grey line. |
Fig. 2(a) reports the high-energy-resolution fluorescence-detected (HERFD) pre-edge spectrum with a dashed black line (Capano et al., 2013
). Three features are shown: firstly, a peak at 7113.5 eV, which is attributed to a Fe 1s → 3d transition; and a second peak and shoulder at 7114.5 and 7115.5 eV, respectively, that arise from MLCT excitations (Capano et al., 2013
).
We initialized the Web-CONEXS calculation, first selecting to perform an calculation with the B3LYP hybrid exchange-correlation functional and def-SVP basis set. The total charge for the calculation was set to 2e, while the multiplicity (2S + 1, with S equal to the total spin) in the LS configuration is 1. The geometry, previously determined by is imported directly from an .xyz file. We selected ABSQ to obtain the which means that the computed spectrum will include the electric dipole and quadrupole transitions, with an 0.5 eV Gaussian broadening in the relevant energy range to the Fe K-edge. In Fig. 2(a), the theoretical pre-edge, computed based on Web-CONEXS parameters, is plotted with a solid orange line. We have applied a rigid shift to the spectrum energy a posteriori, which accounts for a systematic underestimation of the core-electron binding energy. To help guide the eye we have chosen the magnitude of the shift to align the second peak of the theoretical and experimental spectra since the relative energy separation of the two MLCT excitations is well reproduced. Web-CONEXS reproduces the two peaks and shoulder features observed in the experimental measurement; in particular, the relative position of the MLCT peak and shoulder are well described. The peak corresponding to the 1s → 3d transition is red-shifted by approximately 1 eV compared with the experiment, due to neglect of in Hartree–Fock theory which over-stabilizes the 3d(eg) orbitals (Reiher et al., 2001
) as discussed by Capano et al. (2013
).
Systematic improvements, building on the input parameters generated by Web-CONEXS, lead to better agreement with the experiment. These improvements may be carried out (independently from the Web-CONEXS platform) by the user; this requires access to a computing cluster and resubmission of the (updated) input files using a job script. The grey solid line plotted in Fig. 2(a) is the result of a simulation with improved (but more computationally demanding) input parameters, which we define as production level. In this case, the energetic alignment of all three of the pre-edge features is found to agree with the experimental spectrum. For the computation of this spectrum, we first increased the size of the basis set, opting to use TZVP. Next, following Capano et al. (2013
), we tuned the fraction of exact exchange contained in the B3LYP functional. A Hartree–Fock exchange of 12.5% was found to reproduce the peak-splitting observed in the experiment. Additionally, we computed the first 50 core excitations to ensure convergence in the edge region. Finally, we applied a Gaussian broadening with a full width at half-maximum equal to 0.5 eV. These changes are reflected in the orca.inp file provided in the supporting information.
3.2. L-edge spectra in crystalline powders
To demonstrate the capabilities of the Web-CONEXS QE interface, we computed the Zn L3-edge of ZnO. This well defined crystalline material adopts the hexagonal wurtzite structure with the P63mc. Within the 3d lattice, Zn2+ ions are ionically bonded to O2−, which leads to them having a complete 3d shell and the valence 3s23p63d104s0. Transitions at the Zn L3 edge correspond to Zn 2p → Zn 4s and Zn 2p → Zn 4d excitations. Fig. 2(b) plots the Zn L3 edge for ZnO powder (Primo et al., 2022
) in dashed black lines. The main is a broad feature with at least two peaks centred at 1026.5 eV and 1028 eV. A second broad absorption feature is observed at higher energy, with a maximum at 1032.5 eV. Finally, there is a shoulder in the rising edge located at 1022.5 eV. The low-energy shoulder is assigned to Zn 2p → Zn 4s transitions, while the higher energy peaks are considered to be due to Zn 2p → Zn 4d transitions (Primo et al., 2022
).
To set up the ZnO system within Web-CONEXS, we first imported the structure using the Materials Project identification number MP-2133 (Jain et al., 2013; Ong et al., 2015
), opting for the L3 and selecting Zn as the absorbing species. In Fig. 2
(b), the Web-CONEXS computed spectrum is plotted with a solid orange line. The theoretical line shape broadly agrees with the experimental measurement, demonstrating how the software can be utilized for preliminary modelling of the spectrum. The edge is reproduced with an approximate 1.5 eV difference between the two absorption peaks, close to the observed experimental value. The second broad signal is present but red-shifted to 1031 eV and is significantly narrower than in the experimental data.
In all instances, the input and output provided by Web-CONEXS for QE should be a starting point for further production-level calculations of the XANES spectrum. To highlight this, a fully converged simulation of the Zn L3-edge is shown in grey in Fig. 2(b). Careful consideration of the various input parameters leads to a theoretical spectral line shape that qualitatively agrees with the experimental measurement over the entire energy range. As with the Web-CONEXS result, the energy-splitting of the peaks in the is well reproduced (1.5 eV). Moreover, a shoulder at 1028.3 eV emerges as an additional feature beyond the resolution of the reference experimental data. The energetic alignment of the second broad feature is also greatly improved, with the simulated spectrum suggesting the presence of two transitions at 1031 eV and 1032.5 eV. The Zn 2p → Zn 4s low-energy shoulder is not captured in these simulations, possibly due to the approximate treatment of many-body excitonic effects via a static core-hole potential within the underlying theoretical framework.
To obtain this fully converged simulation, starting from the Web-CONEXS input, we carried out convergence tests on the cutoff of the plane-wave basis set and the density of the k-point mesh used to sample the first during the SCF and XANES calculation steps. We found that a 70 Ry cutoff on plane waves and a 6 × 6 × 6 k-point grid sufficient to converge simulations of the To approximate the many-body effects due to the presence of a core hole, we constructed a Zn pseudopotential with a hole in the 2p level and added a total charge of +1.0 e to the simulation cell. Artificial interactions between periodic replicas of the core-hole were removed using a 3 × 3 × 3 with a k-point grid scaled to 3 × 3 × 3 points. To construct the powder spectrum, we repeated the XANES calculation, averaging over the three polarization directions. A constant Gaussian broadening of 0.4 eV was applied to the spectral range. All files related to these simulations are provided in the supporting information.
3.3. K-edge spectra for Ag nanoparticles
As an example using the FDMNES interface, we have computed the Ag K-edge for bulk Ag, which can exist within Ag nanoparticles (Godfrey et al., 2020). The of Ag is presented in Fig. 2
(c); it adopts an f.c.c. structure with the Fm3m. Each Ag0 has the valence 4s24p64d105s1, leading to a K-edge spectrum characterized by 1s → 5p transitions. Fig. 2
(c) reports the experimental Ag K-edge XANES spectrum for Ag foil with a black dashed line (Godfrey et al., 2020
). The spectrum has a shoulder in the rising edge at 25517 eV. The main peak is broad, with a maximum at 25527 eV and there is a second, broad and more intense peak at 25551 eV.
We set up the Ag crystal on the Web-CONEXS FDMNES interface using a .cif obtained from entry mp-124 of the Materials Project database (Jain et al., 2013; Ong et al., 2015
). We then selected to compute the K-edge for Ag, opting to treat the material as a crystal. The Web-CONEXS result is plotted with a solid orange line in Fig. 2
(c). The simulated spectrum is in good agreement with the experimental measurement. In particular, the energetic alignment (25516 eV, 25528 eV and 25549 eV), energy range, and intensity of the shoulder and two peaks all match their experimental counterparts.
Even though, in this case, the Web-CONEXS result satisfactorily reproduces the experimental spectrum, we performed a series of further simulations with the aim of demonstrating how the basic input parameters provided by the platform may be improved. The resulting simulation, which is plotted with a solid grey line in Fig. 2(c), similar to the Web-CONEXS result, reproduces all of the main features of the spectrum over a larger energy range. To obtain the final simulated spectrum, we included spin-orbit interactions and set the electronic occupancy of the spin-up and spin-down valence channels to 4d105s0.5. Furthermore, we increased the radius of the cluster where the final state calculation is performed to 8 Å and increased the energy range for the calculation to −20 eV ≤ EF ≤ 60 eV. All files related to these simulations are provided in the supporting information.
4. Conclusions
Computational core-hole spectroscopy can be instrumental in characterizing the electronic transitions observed in an experimental X-ray spectra. For non-specialist users, theoretical X-ray simulations based on electronic structure theory require a significant (and prohibitive) investment in time and resources. Here, we introduced Web-CONEXS, a web application hosted on ISPyB at DLS, which provides a platform for modelling the spectra based on minimal user input.
Web-CONEXS interfaces the three different electronic structure theory codes, each with well defined uses. For molecular systems and cluster-type simulations, Web-CONEXS interfaces to the ORCA package; for periodic and crystalline systems, Web-CONEXS interfaces to the QE software suite; and for large computationally demanding systems, Web-CONEXS interfaces to FDMNES.
Web-CONEXS offsets complexities associated with creating code-specific input files by generating minimal viable input that serves as a starting point for the user based on their material of interest. In addition, Web-CONEXS handles the submission of batch jobs to the High-Throughput Compute facility IRIS, and the retrieval and post-processing of simulation output, further reducing the burden on the user.
To demonstrate the efficacy of the approach, we applied Web-CONEXS to three different user cases: (i) the pre-edge features of an organometallic spin-crossover complex in the hard X-ray regime, (ii) the L-edge powder spectrum of a crystalline transition metal oxide in the soft X-ray regime and (iii) the FDMNES example. In each example, we demonstrated how Web-CONEXS can be utilized as a relatively good starting point for modelling the spectrum and how to build on the generated input to obtain production-level simulations in agreement with the experimental measurements.
Supporting information
Simulation parameters and software input files. DOI: https://doi.org/10.1107/S1600577524005630/ok5117sup1.pdf
Funding information
Funding for this work was provided by the Engineering and Physical Sciences Research Council (EPSRC) via CONEXS (grant Nos. EP/S021493/1 awarded to SDM; EP/S022058 awarded to TJP). We acknowledge computational resources from ARCHER2 UK National Computing Service, which was granted via HPC-CONEXS, the UK High-End Computing Consortium (EPSRC grant No. EP/X035514/1 awarded to TJP, JDE and SDM). Furthermore, computing resources have been provided by the STFC Scientific Computing Department's SCARF cluster and IRIS digital research infrastructure.
References
Alkhalifah, M. A., Howchen, B., Staddon, J., Celorrio, V., Tiwari, D. & Fermin, D. J. (2022). J. Am. Chem. Soc. 144, 4439–4447. CrossRef CAS PubMed Google Scholar
Apra, E., Bylaska, E. J., de Jong, W. A., Govind, N., Kowalski, K., Straatsma, T. P., Valiev, M., van Dam, H. J. J., Alexeev, Y., Anchell, J., Anisimov, V., Aquino, F. W., Atta-Fynn, R., Autschbach, J., Bauman, N. P., Becca, J. C., Bernholdt, D. E., Bhaskaran-Nair, K., Bogatko, S., Borowski, P., Boschen, J., Brabec, J., Bruner, A., Cauët, E., Chen, Y., Chuev, G. N., Cramer, C. J., Daily, J., Deegan, M. J. O., Dunning, T. H. Jr., Dupuis, M., Dyall, K. G., Fann, G. I., Fischer, S. A., Fonari, A., Früchtl, H., Gagliardi, L., Garza, J., Gawande, N., Ghosh, S., Glaesemann, K., Götz, A. W., Hammond, J., Helms, V., Hermes, E. D., Hirao, K., Hirata, S., Jacquelin, M., Jensen, L., Johnson, B. G., Jónsson, H., Kendall, R. A., Klemm, M., Kobayashi, R., Konkov, V., Krishnamoorthy, S., Krishnan, M., Lin, Z., Lins, R. D., Littlefield, R. J., Logsdail, A. J., Lopata, K., Ma, W., Marenich, A. V., Martin del Campo, J., Mejia-Rodriguez, D., Moore, J. E., Mullin, J. M., Nakajima, T., Nascimento, D. R., Nichols, J. A., Nichols, P. J., Nieplocha, J., Otero-de-la-Roza, A., Palmer, B., Panyala, A., Pirojsirikul, T., Peng, B., Peverati, R., Pittner, J., Pollack, L., Richard, R. M., Sadayappan, P., Schatz, G. C., Shelton, W. A., Silverstein, D. W., Smith, D. M. A., Soares, T. A., Song, D., Swart, M., Taylor, H. L., Thomas, G. S., Tipparaju, V., Truhlar, D. G., Tsemekhman, K., Van Voorhis, T., Vázquez-Mayagoitia, Verma, P., Villa, O., Vishnu, A., Vogiatzis, K. D., Wang, D., Weare, J. H., Williamson, M. J., Windus, T. L., Woliński, K., Wong, A. T., Wu, Q., Yang, C., Yu, Q., Zacharias, M., Zhang, Z., Zhao, Y. & Harrison, R. J. (2020). J. Chem. Phys. 152, 184102. Google Scholar
Arias-Egido, E., Laguna-Marco, M. A., Piquer, C., Jiménez-Cavero, P., Lucas, I., Morellón, L., Gallego, F., Rivera-Calzada, A., Cabero-Piris, M., Santamaria, J., Fabbris, G., Haskel, D., Boada, R. & Díaz-Moreno, S. (2021). Nanoscale, 13, 17125–17135. Web of Science CAS PubMed Google Scholar
Bartlett, S. A., Besley, N. A., Dent, A. J., Diaz-Moreno, S., Evans, J., Hamilton, M. L., Hanson-Heine, M. W. D., Horvath, R., Manici, V., Sun, X.-Z., Towrie, M., Wu, L., Zhang, X. & George, M. W. (2019). J. Am. Chem. Soc. 141, 11471–11480. CrossRef CAS PubMed Google Scholar
Begum, V., Gruner, M. E., Vorwerk, C., Draxl, C. & Pentcheva, R. (2021). Phys. Rev. B, 103, 195128. CrossRef Google Scholar
Benfatto, M., Della Longa, S., Pace, E., Chillemi, G., Padrin, C., Natoli, C. R. & Sanna, N. (2021). Comput. Phys. Commun. 265, 107992. Web of Science CrossRef Google Scholar
Besley, N. A. (2021). WIREs Comput. Mol. Sci. 11, e1527. Google Scholar
Blöchl, P. E. (1994). Phys. Rev. B, 50, 17953–17979. CrossRef Web of Science Google Scholar
Boada, R., Chaboy, J., Hayama, S., Keenan, L. L., Freeman, A. A., Amboage, M. & Díaz-Moreno, S. (2022). J. Phys. Chem. C, 126, 5935–5943. CrossRef CAS Google Scholar
Bunău, O. & Calandra, M. (2013). Phys. Rev. B, 87, 205105. Google Scholar
Bunău, O. & Joly, Y. (2009). J. Phys. Condens. Matter, 21, 345501. Web of Science PubMed Google Scholar
Bunău, O., Ramos, A. Y. & Joly, Y. (2021). The FDMNES code, https://doi.org/10.1107/S1574870720003304. Google Scholar
Bussy, A. & Hutter, J. (2021). Phys. Chem. Chem. Phys. 23, 4736–4746. CrossRef CAS PubMed Google Scholar
Capano, G., Penfold, T. J., Besley, N. A., Milne, C. J., Reinhard, M., Rittmann-Frank, H., Glatzel, P., Abela, R., Rothlisberger, U., Chergui, M. & Tavernelli, I. (2013). Chem. Phys. Lett. 580, 179–184. CrossRef CAS Google Scholar
Chantzis, A., Kowalska, J. K., Maganas, D., DeBeer, S. & Neese, F. (2018). J. Chem. Theory Comput. 14, 3686–3702. CrossRef CAS PubMed Google Scholar
Clark, S. J., Segall, M. D., Pickard, C. J., Hasnip, P. J., Probert, M. I. J., Refson, K. & Payne, M. C. (2005). Z. Kristallogr. 220, 567–570. Web of Science CrossRef CAS Google Scholar
Dal Corso, A. (2014). Comput. Mater. Sci. 95, 337–350. Web of Science CrossRef CAS Google Scholar
Das, M., Bhowal, S., Sannigrahi, J., Bandyopadhyay, A., Banerjee, A., Cibin, G., Khalyavin, D., Banerjee, N., Adroja, D., Dasgupta, I. & Majumdar, S. (2022). Phys. Rev. B, 105, 134421. CrossRef Google Scholar
Dent, A. J., Cibin, G., Ramos, S., Smith, A. D., Scott, S. M., Varandas, L., Pearson, M. R., Krumpa, N. A., Jones, C. P. & Robbins, P. E. (2009). J. Phys. Conf. Ser. 190, 012039. CrossRef Google Scholar
Diaz-Moreno, S., Amboage, M., Basham, M., Boada, R., Bricknell, N. E., Cibin, G., Cobb, T. M., Filik, J., Freeman, A., Geraki, K., Gianolio, D., Hayama, S., Ignatyev, K., Keenan, L., Mikulska, I., Mosselmans, J. F. W., Mudd, J. J. & Parry, S. A. (2018). J. Synchrotron Rad. 25, 998–1009. Web of Science CrossRef IUCr Journals Google Scholar
Diaz-Moreno, S., Hayama, S., Amboage, M., Freeman, A., Sutter, J. & Duller, G. (2009). J. Phys. Conf. Ser. 190, 012038. Google Scholar
Epifanovsky, E., Gilbert, A. T. B., Feng, X., Lee, J., Mao, Y., Mardirossian, N., Pokhilko, P., White, A. F., Coons, M. P., Dempwolff, A. L., Gan, Z., Hait, D., Horn, P. R., Jacobson, L. D., Kaliman, I., Kussmann, J., Lange, A. W., Lao, K. U., Levine, D. S., Liu, J., McKenzie, S. C., Morrison, A. F., Nanda, K. D., Plasser, F., Rehn, D. R., Vidal, M. L., You, Z.., Zhu, Y., Alam, B., Albrecht, B. J., Aldossary, A., Alguire, E., Andersen, J. H., Athavale, V., Barton, D., Begam, K., Behn, A., Bellonzi, N., Bernard, Y. A., Berquist, E. J., Burton, H. G. A., Carreras, A., Carter-Fenk, K., Chakraborty, R., Chien, A. D., Closser, K. D., Cofer-Shabica, V., Dasgupta, S., de Wergifosse, M., Deng, J., Diedenhofen, M., Do, H., Ehlert, S., Fang, P.., Fatehi, S., Feng, Q., Friedhoff, T., Gayvert, J., Ge, Q., Gidofalvi, G., Goldey, M., Gomes, J., González-Espinoza, C. E., Gulania, S., Gunina, A. O., Hanson-Heine, M. W. D., Harbach, P. H. P., Hauser, A., Herbst, M. F., Hernández Vera, M., Hodecker, M., Holden, Z. C., Houck, S., Huang, X., Hui, K., Huynh, B. C., Ivanov, M., Jász, Ji, H., Jiang, H., Kaduk, B., Kähler, S., Khistyaev, K., Kim, J., Kis, G., Klunzinger, P., Koczor-Benda, Z., Koh, J. H., Kosenkov, D., Koulias, L., Kowalczyk, T., Krauter, C. M., Kue, K., Kunitsa, A., Kus, T., Ladjánszki, I., Landau, A., Lawler, K. V., Lefrancois, D., Lehtola, S., Li, R. R., Li, Y.., Liang, J., Liebenthal, M., Lin, H.., Lin, Y.., Liu, F., Liu, K.., Loipersberger, M., Luenser, A., Manjanath, A., Manohar, P., Mansoor, E., Manzer, S. F., Mao, S.., Marenich, A. V., Markovich, T., Mason, S., Maurer, S. A., McLaughlin, P. F., Menger, M. F. S. J., Mewes, J.., Mewes, S. A., Morgante, P., Mullinax, J. W., Oosterbaan, K. J., Paran, G., Paul, A. C., Paul, S. K., Pavošević, F., Pei, Z., Prager, S., Proynov, E. I., Rák, Ramos-Cordoba, E., Rana, B., Rask, A. E., Rettig, A., Richard, R. M., Rob, F., Rossomme, E., Scheele, T., Scheurer, M., Schneider, M., Sergueev, N., Sharada, S. M., Skomorowski, W., Small, D. W., Stein, C. J., Su, Y.., Sundstrom, E. J., Tao, Z., Thirman, J., Tornai, G. J., Tsuchimochi, T., Tubman, N. M., Veccham, S. P., Vydrov, O., Wenzel, J., Witte, J., Yamada, A., Yao, K., Yeganeh, S., Yost, S. R., Zech, A., Zhang, I. Y., Zhang, X., Zhang, Y., Zuev, D., Aspuru-Guzik, A., Bell, A. T., Besley, N. A., Bravaya, K. B., Brooks, B. R., Casanova, D., Chai, J.., Coriani, S., Cramer, C. J., Cserey, G., DePrince, A. E. III, DiStasio, R. A. Jr., Dreuw, A., Dunietz, B. D., Furlani, T. R., Goddard, W. A. III, Hammes-Schiffer, S., Head-Gordon, T., Hehre, W. J., Hsu, C.., Jagau, T.., Jung, Y., Klamt, A., Kong, J., Lambrecht, D. S., Liang, W., Mayhall, N. J., McCurdy, C. W., Neaton, J. B., Ochsenfeld, C., Parkhill, J. A., Peverati, R., Rassolov, V. A., Shao, Y., Slipchenko, L. V., Stauch, T., Steele, R. P., Subotnik, J. E., Thom, A. J. W., Tkatchenko, A., Truhlar, D. G., Van Voorhis, T., Wesolowski, T. A., Whaley, K. B., Woodcock, H. L. III, Zimmerman, P. M., Faraji, S., Gill, P. M. W., Head-Gordon, M., Herbert, J. M. & Krylov, A. I. (2021). J. Chem. Phys. 155, 084801. CrossRef PubMed Google Scholar
Figueroa, S. J. A., Beniz, D. B., Mauricio, J. C., Piton, J. R., Parry, S. A. & Cibin, G. (2018). J. Synchrotron Rad. 25, 953–959. Web of Science CrossRef CAS IUCr Journals Google Scholar
Foster, C., Shaw, S., Neill, T. S., Bryan, N., Sherriff, N., Natrajan, L. S., Wilson, H., Lopez-Odriozola, L., Rigby, B., Haigh, S. J., Zou, Y.-C., Harrison, R. & Morris, K. (2022). Langmuir, 38, 2576–2589. CrossRef CAS PubMed Google Scholar
García, Á., Haynes, T., Retuerto, M., Ferrer, P., Pascual, L., Peña, M. A., Abdel Salam, M., Mokhtar, M., Gianolio, D. & Rojas, S. (2021). Ind. Eng. Chem. Res. 60, 18759–18769. PubMed Google Scholar
Giannozzi, P., Andreussi, O., Brumme, T., Bunau, O., Buongiorno Nardelli, M., Calandra, M., Car, R., Cavazzoni, C., Ceresoli, D., Cococcioni, M., Colonna, N., Carnimeo, I., Dal Corso, A., de Gironcoli, S., Delugas, P., DiStasio, R. A., Ferretti, A., Floris, A., Fratesi, G., Fugallo, G., Gebauer, R., Gerstmann, U., Giustino, F., Gorni, T., Jia, J., Kawamura, M., Ko, H.-Y., Kokalj, A., Küçükbenli, E., Lazzeri, M., Marsili, M., Marzari, N., Mauri, F., Nguyen, N. L., Nguyen, H.-V., Otero-de-la-Roza, A., Paulatto, L., Poncé, S., Rocca, D., Sabatini, R., Santra, B., Schlipf, M., Seitsonen, A. P., Smogunov, A., Timrov, I., Thonhauser, T., Umari, P., Vast, N., Wu, X. & Baroni, S. (2017). J. Phys. Condens. Matter, 29, 465901. Web of Science CrossRef PubMed Google Scholar
Giannozzi, P., Baroni, S., Bonini, N., Calandra, M., Car, R., Cavazzoni, C., Ceresoli, D., Chiarotti, G. L., Cococcioni, M., Dabo, I., Dal Corso, A., de Gironcoli, S., Fabris, S., Fratesi, G., Gebauer, R., Gerstmann, U., Gougoussis, C., Kokalj, A., Lazzeri, M., Martin-Samos, L., Marzari, N., Mauri, F., Mazzarello, R., Paolini, S., Pasquarello, A., Paulatto, L., Sbraccia, C., Scandolo, S., Sclauzero, G., Seitsonen, A. P., Smogunov, A., Umari, P. & Wentzcovitch, R. M. (2009). J. Phys. Condens. Matter, 21, 395502. Web of Science CrossRef PubMed Google Scholar
Giannozzi, P., Baseggio, O., Bonfà, P., Brunato, D., Car, R., Carnimeo, I., Cavazzoni, C., de Gironcoli, S., Delugas, P., Ferrari Ruffino, F., Ferretti, A., Marzari, N., Timrov, I., Urru, A. & Baroni, S. (2020). J. Chem. Phys. 152, 154105. Web of Science CrossRef PubMed Google Scholar
Gilmore, K., Vinson, J., Shirley, E. L., Prendergast, D., Pemmaraju, C. D., Kas, J. J., Vila, F. D. & Rehr, J. J. (2015). Comput. Phys. Commun. 197, 109–117. Web of Science CrossRef CAS Google Scholar
Godfrey, I. J., Dent, A. J., Parkin, I. P., Maenosono, S. & Sankar, G. (2020). ACS Omega, 5, 13664–13671. CrossRef CAS PubMed Google Scholar
Gougoussis, C., Calandra, M., Seitsonen, A., Brouder, C., Shukla, A. & Mauri, F. (2009a). Phys. Rev. B, 79, 045118. CrossRef Google Scholar
Gougoussis, C., Calandra, M., Seitsonen, A. P. & Mauri, F. (2009b). Phys. Rev. B, 80, 075102. CrossRef Google Scholar
Groot, F. M. F. de, Elnaggar, H., Frati, F., Wang, R., Delgado-Jaime, M. U., van Veenendaal, M., Fernandez-Rodriguez, J., Haverkort, M. W., Green, R. J., van der Laan, G., Kvashnin, Y., Hariki, A., Ikeno, H., Ramanantoanina, H., Daul, C., Delley, B., Odelius, M., Lundberg, M., Kuhn, O., Bokarev, S. I., Shirley, E., Vinson, J., Gilmore, K., Stener, M., Fronzoni, G., Decleva, P., Kruger, P., Retegan, M., Joly, Y., Vorwerk, C., Draxl, C., Rehr, J. & Tanaka, A. (2021). J. Electron Spectrosc. Relat. Phenom. 249, 147061. Google Scholar
Gulans, A., Kontur, S., Meisenbichler, C., Nabok, D., Pavone, P., Rigamonti, S., Sagmeister, S., Werner, U. & Draxl, C. (2014). J. Phys. Condens. Matter, 26, 363202. Web of Science CrossRef PubMed Google Scholar
Haverkort, M. W., Zwierzycki, M. & Andersen, O. K. (2012). Phys. Rev. B, 85, 165113. Web of Science CrossRef Google Scholar
Hayama, S., Boada, R., Chaboy, J., Birt, A., Duller, G., Cahill, L., Freeman, A., Amboage, M., Keenan, L. & Diaz-Moreno, S. (2021). J. Phys. Condens. Matter, 33, 284003. Web of Science CrossRef Google Scholar
Jain, A., Ong, S. P., Hautier, G., Chen, W., Richards, W. D., Dacek, S., Cholia, S., Gunter, D., Skinner, D., Ceder, G. & Persson, K. A. (2013). APL Materials, 1, 011002. Google Scholar
Joly, Y. (2001). Phys. Rev. B, 63, 125120. Web of Science CrossRef Google Scholar
Joly, Y., Bunău, O., Lorenzo, J. E., Galéra, R. M., Grenier, S. & Thompson, B. (2009). J. Phys. Conf. Ser. 190, 012007. CrossRef Google Scholar
Kellett, C. W., Kennepohl, P. & Berlinguette, C. P. (2020). Nat. Commun. 11, 3310. CrossRef PubMed Google Scholar
Ma, Y., Lu, W., Han, X., Chen, Y., da Silva, I., Lee, D., Sheveleva, A. M., Wang, Z., Li, J., Li, W., Fan, M., Xu, S., Tuna, F., McInnes, E. J. L., Cheng, Y., Rudić, S., Manuel, P., Frogley, M. D., Ramirez-Cuesta, A. J., Schröder, M. & Yang, S. (2022). J. Am. Chem. Soc. 144, 8624–8632. CrossRef CAS PubMed Google Scholar
Maganas, D., DeBeer, S. & Neese, F. (2017). Inorg. Chem. 56, 11819–11836. Web of Science CrossRef CAS PubMed Google Scholar
Maganas, D., DeBeer, S. & Neese, F. (2018). J. Phys. Chem. A, 122, 1215–1227. CrossRef CAS PubMed Google Scholar
Mosselmans, J. F. W., Quinn, P. D., Dent, A. J., Cavill, S. A., Moreno, S. D., Peach, A., Leicester, P. J., Keylock, S. J., Gregory, S. R., Atkinson, K. D. & Rosell, J. R. (2009). J. Synchrotron Rad. 16, 818–824. Web of Science CrossRef CAS IUCr Journals Google Scholar
Neese, F. (2012). WIREs Comput. Mol. Sci. 2, 73–78. Web of Science CrossRef CAS Google Scholar
Neese, F. (2018). WIREs Comput. Mol. Sci. 8, e1327. Google Scholar
Neese, F. (2022). WIREs Comput. Mol. Sci. 12, e1606. Google Scholar
Nicholls, R. J., Diaz-Moreno, S., Iliffe, W., Linden, Y., Mousavi, T., Aramini, M., Danaie, M., Grovenor, C. R. M. & Speller, S. C. (2022). Commun. Mater. 3, 52. CrossRef Google Scholar
Norman, P. & Dreuw, A. (2018). Chem. Rev. 118, 7208–7248. Web of Science CrossRef CAS PubMed Google Scholar
Ong, S. P., Cholia, S., Jain, A., Brafman, M., Gunter, D., Ceder, G. & Persson, K. A. (2015). Comput. Mater. Sci. 97, 209–215. Web of Science CrossRef Google Scholar
Pickard, C. J. & Mauri, F. (2001). Phys. Rev. B, 63, 245101. Web of Science CrossRef Google Scholar
Pinjari, R. V., Delcey, M., Guo, M., Odelius, M. & Lundberg, M. (2016). J. Comput. Chem. 37, 477–486. CrossRef CAS PubMed Google Scholar
Primo, J. O., Correa, J. S., Horsth, D. F. L., Das, A., Zając, M., Umek, P., Wattiez, R., Anaissi, F. J., Onderwater, R. C. A. & Bittencourt, C. (2022). Nanomaterials, 12, 4345. CrossRef PubMed Google Scholar
Rankine, C. D. & Penfold, T. J. (2021). J. Phys. Chem. A, 125, 4276–4293. CrossRef CAS PubMed Google Scholar
Rehr, J. J., Kas, J. J., Vila, F. D., Prange, M. P. & Jorissen, K. (2010). Phys. Chem. Chem. Phys. 12, 5503. Web of Science CrossRef PubMed Google Scholar
Reiher, M., Salomon, O. & Artur Hess, B. (2001). Theor. Chem. Accounts, 107, 48–55. CrossRef CAS Google Scholar
Taillefumier, M., Cabaret, D., Flank, A.-M. & Mauri, F. (2002). Phys. Rev. B, 66, 195107. Web of Science CrossRef Google Scholar
Taylor, R. P., Finch, A. A., Mosselmans, J. F. W. & Quinn, P. D. (2013). J. Lumin. 134, 49–58. Web of Science CrossRef CAS Google Scholar
Timrov, I., Agrawal, P., Zhang, X., Erat, S., Liu, R., Braun, A., Cococcioni, M., Calandra, M., Marzari, N. & Passerone, D. (2020). Phys. Rev. Res. 2, 033265. CrossRef Google Scholar
Valiev, M., Bylaska, E. J., Govind, N., Kowalski, K., Straatsma, T. P., Van Dam, H. J. J., Wang, D., Nieplocha, J., Apra, E., Windus, T. L. & de Jong, W. A. (2010). Comput. Phys. Commun. 181, 1477–1489. Web of Science CrossRef CAS Google Scholar
Vinson, J., Rehr, J. J., Kas, J. J. & Shirley, E. L. (2011). Phys. Rev. B, 83, 115106. Web of Science CrossRef Google Scholar
Vorwerk, C., Aurich, B., Cocchi, C. & Draxl, C. (2019). Electron. Struct. 1, 037001. CrossRef Google Scholar
Vorwerk, C., Cocchi, C. & Draxl, C. (2017). Phys. Rev. B, 95, 155121. Web of Science CrossRef Google Scholar
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.