Constructing and Evaluating Machine-Learned Interatomic Potentials for Li-Based Disordered Rocksalts

Lithium-based disordered rocksalts (LDRs), which are an important class of positive electrode materials that can increase the energy density of current Li-ion batteries, represent a significantly complex chemical and configurational space for conventional density functional theory (DFT)-based high-throughput screening approaches. Notably, atom-centered machine-learned interatomic potentials (MLIPs) are a promising pathway to accurately model the potential energy surface of highly disordered chemical spaces, such as LDRs, where the performance of such MLIPs has not been rigorously explored yet. Here, we represent a comprehensive evaluation of the accuracy, transferability, and ease of training of five atom-centered MLIPs, including the artificial neural network potentials developed by the atomic energy network (AENET), the Gaussian approximation potential (GAP), the spectral neighbor analysis potential (SNAP) and its quadratic extension (qSNAP), and the moment tensor potential (MTP), in modeling a 11-component LDR chemical space. Specifically, we generate a DFT-calculated data set of 10,842 configurations of disordered LiTMO2 and TMO2 compositions, where TM = Sc, Ti, V, Cr, Mn, Fe, Co, Ni, and/or Cu. To provide a point-of-comparison on the performance of atom-centered MLIPs, we also trained the neural equivariant interatomic potential (NequIP) on a subset of our data. Importantly, we find AENET to be the best potential in terms of accuracy and transferability for energy predictions, while MTP is the best for atomic forces. While AENET is the fastest to train among the MLIPs considered at low number of epochs (300), the training time increases significantly as epochs increase (3300), with a corresponding reduction in training errors (∼60%). Note that AENET and GAP tend to overfit in small data sets, with the extent of overfitting reducing with larger data sets. Finally, we observe AENET to provide reasonable predictions of average Li-intercalation voltages in layered, single-TM LiTMO2 frameworks, compared to DFT (∼10% error on average). Our study should pave the way both for discovering novel disordered rocksalt electrodes and for modeling other configurationally complex systems, such as high-entropy ceramics and alloys.

A 3.2 V Binary Layered Oxide Cathode for Potassium-Ion Batteries

Potassium-ion batteries (KIBs) can offer high energy density, cyclability, and operational safety while being economical due to the natural abundance of potassium. Utilizing graphite as an anode, suitable cathodes can realize full cells. Searching for potential cathodes, this work introduces P3-type K0.5Ni1/3Mn2/3O2 layered oxide as a potential candidate synthesized by a simple solid-state method. The material works as a 3.2 V cathode combining Ni redox at high voltage and Mn redox at low voltage and exhibits highly reversible K+ ion (de)insertion at ambient and elevated (40-50 °C) temperatures. First-principles calculations suggest the ground state in-plane Mn-Ni ordering in the MO2 sheets is strongly correlated to the K-content in the framework, leading to an interwoven and alternative row ordering of Ni-Mn in K0.5Ni1/3Mn2/3O2. Postmortem and electrochemical titration reveal the occurrence of a solid solution mechanism during K+ (de)insertion. The findings suggest that the Ni addition can effectively tune the electronic and structural properties of the cathode, leading to improved electrochemical performance. This work provides new insights in the quest to develop potential low-cost Co-free KIB cathodes for practical applications in stationary energy storage.

Averting H+-Mediated Charge Storage Chemistry Stabilizes the High Output Voltage of LiMn2O4-Based Aqueous Battery

H+ co-intercalation chemistry of the cathode is perceived to have damaging consequences on the low-rate and long-term cycling of aqueous zinc batteries, which is a critical hindrance to their promise for stationary storage applications. Herein, the thermodynamically competitive H+ storage chemistry of an attractive high-voltage cathode LiMn2O4 is revealed by employing operando and ex-situ analytical techniques together with density functional theory-based calculations. The H+ electrochemistry leads to the previously unforeseen voltage decay with cycling, impacting the available energy density, particularly at lower currents. Based on an in-depth investigation of the effect of the Li+ to Zn2+ ratio in the electrolyte on the charge storage mechanism, a purely aqueous and low-salt concentration electrolyte with a tuned Li+/Zn2+ ratio is introduced to subdue the H+-mediated charge storage kinetically, resulting in a stable voltage output and improved cycling stability at both low and high cathode loadings. Synchrotron X-ray diffraction analysis reveals that repeated H+ intercalation triggers an irreversible phase transformation leading to voltage decay, which is averted by shutting down H+ storage. These findings unveiling the origin and impact of the deleterious H+-storage, coupled with the practical strategy for its inhibition, will inspire further work toward this under-explored realm of aqueous battery chemistry.

Evaluation of P3-Type Layered Oxides as K-Ion Battery Cathodes

Given the increasing energy storage demands and limited natural resources of Li, K-ion batteries (KIBs) could be promising next-generation systems having natural abundance, similar chemistry, and energy density. Here, we have investigated the P3-type K0.5TMO2 (where TM = Ti, V, Cr, Mn, Co, or Ni) systems using density functional theory calculations as potential positive intercalation electrodes (or cathodes) for KIBs. Specifically, we have identified ground-state configurations and calculated the average topotactic voltages, electronic structures, on-site magnetic moments, and thermodynamic stabilities of all P3-K0.5TMO2 compositions and their corresponding depotassiated P3-TMO2 frameworks. Additionally, we evaluated the dynamic stability and K-mobility in select P3 structures. We find that K adopts the honeycomb or zig-zag configuration within each K-layer of all P3 structures considered, irrespective of the transition-metal (TM). In terms of voltages, we find the Co- and Ti-based compositions to exhibit the highest (4.59 V vs. K) and lowest (2.24 V) voltages, respectively, with the TM contributing to the redox behavior upon K (de-)intercalation. We observe all P3-K0.5TMO2 to be (meta)stable and hence experimentally synthesizable according to our 0 K convex hull calculations, while all depotassiated P3-TMO2 configurations are unstable and may appear during electrochemical cycling. Also, we verified the stability of the prismatic coordination environment of K compared to octahedral coordination at the K0.5TMO2 compositions using Rouxel and cationic potential models. Finally, combining our voltage and stability calculations, we find P3-KxCoO2 to be the most promising cathode composition, while P3-KxNiO2 is worth exploring. We also find P3-KxMnO2 to be worth pursuing given its dynamic stability and facile migration of K+ at both potassiated and depotassiated compositions. Our work should contribute to the exploration of strategies and materials required to make practical KIBs.

Performance of the r2SCAN Functional in Transition Metal Oxides

We assess the accuracy and computational efficiency of the recently developed meta-generalized gradient approximation (metaGGA) functional, restored regularized strongly constrained and appropriately normed (r²SCAN), in transition metal oxide (TMO) systems and compare its performance against SCAN. Specifically, we benchmark the r²SCAN-calculated oxidation enthalpies, lattice parameters, on-site magnetic moments, and band gaps of binary 3d TMOs against the SCAN-calculated and experimental values. Additionally, we evaluate the optimal Hubbard U correction required for each transition metal (TM) to improve the accuracy of the r²SCAN functional, based on experimental oxidation enthalpies, and verify the transferability of the U values by comparing against experimental properties on other TM-containing oxides. Notably, including the U-correction with r²SCAN increases the lattice parameters, on-site magnetic moments, and band gaps of TMOs, apart from an improved description of the ground state electronic state in narrow band gap TMOs. The r²SCAN and r²SCAN+U calculated oxidation enthalpies follow the qualitative trends of SCAN and SCAN+U, with r²SCAN and r²SCAN+U predicting marginally larger lattice parameters, smaller magnetic moments, and lower band gaps compared to SCAN and SCAN+U, respectively. We observe the overall computational time (i.e., for all ionic+electronic steps) required for r²SCAN(+U) to be lower than SCAN(+U). Thus, the r²SCAN(+U) framework can offer a reasonably accurate description of the ground state properties of TMOs with better computational efficiency than SCAN(+U).

Study of pnictides for photovoltaic applications

For the transition into a sustainable mode of energy usage, it is important to develop photovoltaic materials that exhibit better solar-to-electricity conversion efficiencies, a direct optimal band gap, and are made of non-toxic, earth abundant elements compared to the state-of-the-art silicon photovoltaics. Here, we explore the non-redox-active pnictide chemical space, including binary A₃B₂, ternary AA'₂B₂, and quaternary AA'A''B₂ compounds (A, A', A'' = Ca, Sr, or Zn; B = N or P), as candidate beyond-Si photovoltaics using density functional theory calculations. Specifically, we evaluate the ground state configurations, band gaps, and 0 K thermodynamic stability for all 20 pnictide compositions considered, besides computing the formation energy of cation vacancies, anion vacancies, and cation anti-sites in a subset of candidate compounds. Importantly, we identify SrZn₂N₂, SrZn₂P₂, and CaZn₂P₂ to be promising candidates, exhibiting optimal (1.1-1.5 eV) hybrid-functional-calculated band gaps, stability at 0 K, and high resistance to point defects (formation energies >1 eV), while other possible candidates include ZnCa₂N₂ and ZnSr₂N₂, which may be susceptible to N-vacancy formation. We hope that our study will contribute to the practical development of pnictide semiconductors as beyond-silicon light absorbers.

Advancing MXene Electrocatalysts for Energy Conversion Reactions: Surface, Stoichiometry, and Stability

MXenes, due to their tailorable chemistry and favourable physical properties, have great promise in electrocatalytic energy conversion reactions. To exploit fully their enormous potential, further advances specific to electrocatalysis revolving around their performance, stability, compositional discovery and synthesis are required. The most recent advances in these aspects are discussed in detail: surface functional and stoichiometric modifications which can improve performance, Pourbaix stability related to their electrocatalytic operating conditions, density functional theory and advances in machine learning for their discovery, and prospects in large scale synthesis and solution processing techniques to produce membrane electrode assemblies and integrated electrodes. This Review provides a perspective that is complemented by new density functional theory calculations which show how these recent advances in MXene material design are paving the way for effective electrocatalysts required for the transition to integrated renewable energy systems.

First principles investigation of anionic redox in bisulfate lithium battery cathodes

The search for an alternative high-voltage polyanionic cathode material for Li-ion batteries is vital to improve the energy densities beyond the state-of-the-art, where sulfate frameworks form an important class of high-voltage cathode materials due to the strong inductive effect of the S⁶⁺ ion. Here, we have investigated the mechanism of cationic and/or anionic redox in Li_xM(SO₄)₂ frameworks (M = Mn, Fe, Co, and Ni and 0 ≤ x ≤ 2) using density functional calculations. Specifically, we have used a combination of Hubbard U corrected strongly constrained and appropriately normed (SCAN+U) and generalized gradient approximation (GGA+U) functionals to explore the thermodynamic (polymorph stability), electrochemical (intercalation voltage), geometric (bond lengths), and electronic (band gaps, magnetic moments, charge populations, etc.) properties of the bisulfate frameworks considered. Importantly, we find that the anionic (cationic) redox process is dominant throughout delithiation in the Ni (Mn) bisulfate, as verified using our calculated projected density of states, bond lengths, and on-site magnetic moments. On the other hand, in Fe and Co bisulfates, cationic redox dominates the initial delithiation (1 ≤ x ≤ 2), while anionic redox dominates subsequent delithiation (0 ≤ x ≤ 2). In addition, evaluation of the crystal overlap Hamilton population reveals insignificant bonding between oxidized O atoms throughout the delithiation process in the Ni bisulfate, indicating robust battery performance that is resistant to irreversible oxygen evolution. Finally, we observe that both GGA+U and SCAN+U predictions are in qualitative agreement for the various properties predicted. Our work should open new avenues for exploring lattice oxygen redox in novel high voltage polyanionic cathodes, especially using the SCAN+U functional.

Fundamental investigations on the sodium-ion transport properties of mixed polyanion solid-state battery electrolytes

Lithium and sodium (Na) mixed polyanion solid electrolytes for all-solid-state batteries display some of the highest ionic conductivities reported to date. However, the effect of polyanion mixing on the ion-transport properties is still not fully understood. Here, we focus on Na_1+xZr₂Si_xP_3−xO₁₂ (0 ≤ x ≤ 3) NASICON electrolyte to elucidate the role of polyanion mixing on the Na-ion transport properties. Although NASICON is a widely investigated system, transport properties derived from experiments or theory vary by orders of magnitude. We use more than 2000 distinct ab initio-based kinetic Monte Carlo simulations to map the compositional space of NASICON over various time ranges, spatial resolutions and temperatures. Via electrochemical impedance spectroscopy measurements on samples with different sodium content, we find that the highest ionic conductivity (i.e., about 0.165 S cm^–1 at 473 K) is experimentally achieved in Na_3.4Zr₂Si_2.4P_0.6O₁₂, in line with simulations (i.e., about 0.170 S cm^–1 at 473 K). The theoretical studies indicate that doped NASICON compounds (especially those with a silicon content x ≥ 2.4) can improve the Na-ion mobility compared to undoped NASICON compositions.

Battery solid-state electrolytes rely on mixed polyanion networks to attain high ionic conductivities. Here, the authors investigate the effect of polyanion mixing on the solid-state electrolyte ion conductivity via theoretical calculations and electrochemical measurements.

Assessing cathode property prediction via exchange-correlation functionals with and without long-range dispersion corrections

We benchmark calculated interlayer spacings, average topotactic voltages, thermodynamic stabilities, and band gaps in layered lithium transition-metal oxides (TMOs) and their de-lithiated counterparts, which are used in lithium-ion batteries as positive electrode materials, against available experimental data. Specifically, we examine the accuracy of properties calculated within density functional theory (DFT) using eight different treatments of electron exchange-correlation: the strongly constrained and appropriately normed (SCAN) and Perdew-Burke-Ernzerhof (PBE) density functionals, Hubbard-U-corrected SCAN and PBE (i.e., SCAN+U and PBE+U), and SCAN(+U) and PBE(+U) with added long-range dispersion (D) interactions (i.e., DFT(+U)+D). van der Waals interactions are included respectively via the revised Vydrov-Van Voorhis (rVV10) for SCAN(+U) and the DFT-D3 for PBE(+U). We find that SCAN-based functionals predict larger voltages due to an underestimation of stability of the MO₂ systems, while also predicting smaller interlayer spacings compared to their PBE-based counterparts. Furthermore, adding dispersion corrections to PBE has a greater effect on voltage predictions and interlayer spacings than with SCAN, indicating that DFT-SCAN - despite being a ground-state theory - fortuitously captures some short and medium-range dispersion interactions better than PBE. While SCAN-based and PBE-based functionals yield qualitatively similar band gap predictions, there is no significant quantitative improvement of SCAN-based functionals over the corresponding PBE-based versions. Finally, we expect SCAN-based functionals to yield more accurate property predictions than the respective PBE-based functionals for most TMOs, given SCAN's stronger theoretical underpinning and better predictions of systematic trends in interlayer spacings, intercalation voltages, and band gaps obtained in this work.

Factors Governing Oxygen Vacancy Formation in Oxide Perovskites

The control of oxygen vacancy (V_O) formation is critical to advancing multiple metal-oxide-perovskite-based technologies. We report the construction of a compact linear model for the neutral V_O formation energy in ABO₃ perovskites that reproduces, with reasonable fidelity, Hubbard-U-corrected density functional theory calculations based on the state-of-the-art, strongly constrained and appropriately normed exchange-correlation functional. We obtain a mean absolute error of 0.45 eV for perovskites stable at 298 K, an accuracy that holds across a large, electronically diverse set of ABO₃ perovskites. Our model considers perovskites containing alkaline-earth metals (Ca, Sr, and Ba) and lanthanides (La and Ce) on the A-site and 3d transition metals (Ti, V, Cr, Mn, Fe, Co, and Ni) on the B-site in six different crystal systems (cubic, tetragonal, orthorhombic, hexagonal, rhombohedral, and monoclinic) common to perovskites. Physically intuitive metrics easily extracted from existing experimental thermochemical data or via inexpensive quantum mechanical calculations, including crystal bond dissociation energies and (solid phase) reduction potentials, are key components of the model. Beyond validation of the model against known experimental trends in materials used in solid oxide fuel cells, the model yields new candidate perovskites not contained in our training data set, such as (Bi,Y)(Fe,Co)O₃, which we predict may have favorable thermochemical water-splitting properties. The confluence of sufficient accuracy, efficiency, and interpretability afforded by our model not only facilitates high-throughput computational screening for any application that requires the precise control of V_O concentrations but also provides a clear picture of the dominant physics governing V_O formation in metal-oxide perovskites.

Evaluation of Mg Compounds as Coating Materials in Mg Batteries

Mg batteries utilizing a Mg metal anode with a high-voltage intercalation cathode define a potential pathway toward energy storage with high energy density. However, the making of Mg batteries is plagued by the instability of existing electrolytes against the Mg-metal anode and high-voltage cathode materials. One viable solution to this problem is the identification of protective coating materials that could effectively separate the distinct chemistries of the metal-anode and the cathode materials from the electrolyte. Using first-principles calculations we mapped the electrochemical stability windows for non-redox-active Mg binary and ternary compounds in order to identify potential coating materials for Mg batteries. Our results identify Mg-halides and Mg(BH₄)₂ as promising anode coating materials based on their significant reductive stability. On the cathode side, we single out MgF₂, Mg(PO₃)₂, and MgP₄O₁₁ as effective passivating agents.

High magnesium mobility in ternary spinel chalcogenides

Magnesium batteries appear a viable alternative to overcome the safety and energy density limitations faced by current lithium-ion technology. The development of a competitive magnesium battery is plagued by the existing notion of poor magnesium mobility in solids. Here we demonstrate by using ab initio calculations, nuclear magnetic resonance, and impedance spectroscopy measurements that substantial magnesium ion mobility can indeed be achieved in close-packed frameworks (~ 0.01–0.1 mS cm^–1 at 298 K), specifically in the magnesium scandium selenide spinel. Our theoretical predictions also indicate that high magnesium ion mobility is possible in other chalcogenide spinels, opening the door for the realization of other magnesium solid ionic conductors and the eventual development of an all-solid-state magnesium battery.

Low magnesium mobility in solids represents a significant obstacle to the development of Mg intercalation batteries. Here the authors show that substantial magnesium ion mobility can be achieved in close-packed ternary selenide spinel materials.

Magnesium ion mobility in post-spinels accessible at ambient pressure

First-principles modeling of a new mixed V/Ti post-spinel phase reveals a possible route to ambient-pressure oxide cathodes exhibiting fast Mg diffusion.

Role of Structural H2O in Intercalation Electrodes: The Case of Mg in Nanocrystalline Xerogel-V2O5

Cointercalation is a potential approach to influence the voltage and mobility with which cations insert in electrodes for energy storage devices. Combining a robust thermodynamic model with first-principles calculations, we present a detailed investigation revealing the important role of H2O during ion intercalation in nanomaterials. We examine the scenario of Mg(2+) and H2O cointercalation in nanocrystalline Xerogel-V2O5, a potential cathode material to achieve energy density greater than Li-ion batteries. Water cointercalation in cathode materials could broadly impact an electrochemical system by influencing its voltages or causing passivation at the anode. The analysis of the stable phases of Mg-Xerogel V2O5 and voltages at different electrolytic conditions reveals a range of concentrations for Mg in the Xerogel and H2O in the electrolyte where there is no thermodynamic driving force for H2O to shuttle with Mg during electrochemical cycling. Also, we demonstrate that H2O shuttling with the Mg(2+) ions in wet electrolytes yields higher voltages than in dry electrolytes. The thermodynamic framework used to study water and Mg(2+) cointercalation in this work opens the door for studying the general phenomenon of solvent cointercalation observed in other complex solvent-electrode pairs used in the Li- and Na-ion chemical spaces.

First-principles evaluation of multi-valent cation insertion into orthorhombic V2O5

A detailed exploration and systematic evaluation of V₂O₅ as a cathode material for multi-valent batteries using first-principles.