Abstract
Hydrogen induced corrosion of uranium, which leads to the formation of toxic and pyrophoric UH3, raises significant safety concerns for long-term storage of nuclear materials. Previous work suggests hydrogen diffuses through the grain boundaries (GBs) of the passivating oxide layer to initiate hydriding reactions. However, the atomistic mechanisms underlying this phenomenon and the structural factors that control its initiation are not well understood. To address this knowledge gap, here we use a high-throughput density functional theory (DFT) workflow to investigate the adsorption of H and H2 in the defective bulk UO2. Specifically, we have exhaustively investigated the adsorption of H (107 sites) and H2 (26 sites) in three different coincident site lattice (CSL) GBs: Σ3, Σ5, and Σ9. Compared to the binding energies in pristine UO2, we observe significantly stronger hydrogen adsorption at these GB sites. Interestingly, we find that the trends in H and H2 adsorption vary considerably across the three GB models. In particular, while a small number of sites in Σ5 and Σ9 show exothermic adsorption of H and H2, respectively, no such sites are found in Σ3. These results provide fundamental atomistic insights that could guide the development of future corrosion mitigation strategies for the storage of nuclear materials.


Introduction
Uranium is not only an important nuclear fuel but also a legacy waste that requires special considerations for long-term storage. Its high reactivity with atmospheric gases such as oxygen, − water, − and hydrogen − has been well documented. Pure uranium reacts with oxygen and moisture , to form a thin passivating layer of UO2 that slows down reactivity with other gases. However, hydrogen, due to its small kinetic diameter, can diffuse readily through this oxide layer, eventually leading to the formation of UH3 via the reaction 2U + 3H2 → 2UH3. This hydriding reaction raises several safety concerns , as UH3 is highly toxic and pyrophoric. , Thus, over the past decade, significant efforts have focused on understanding the properties of U, − UO2, − and UH3 − and their interactions with common atmospheric gases.
Recent literature suggests that nucleation of UH3 initiates at the grain boundaries (GBs) within the metal. For example, using electron backscatter diffraction (EBSD), Banos and Scott showed that majority of hydriding sites are located at medium or high angle grain boundaries. Similar results were observed by Darnbrough et al., where the authors suggest that uranium corrosion preferentially occurs perpendicular to the grain boundary compared to along the interface. While these experimental results clearly demonstrate the relevance of UO2 GBs, the atomistic mechanisms associated with hydrogen diffusion and subsequent UH3 formation remain unclear. Although previous computational works have investigated various properties of UO2 GB using atomistic methods, ,− these studies have largely been limited to pristine non-defective uranium oxides. Therefore, to the best of our knowledge, systematic investigation of the interactions of H and H2 with UO2 GBs remain lacking.
To address this knowledge gap, here we use extensive high-throughput density functional theory (DFT) calculations to investigate hydrogen adsorption across three dominant grain boundaries (GBs) that have been previously observed in UO2. Our results reveal that all three GBs exhibit stronger hydrogen binding energies compared to the pristine oxide. However, interestingly, we observe significant variations in the interactions of these GBs with hydrogen. As strong hydrogen adsorption is likely correlated with higher rates of UH3 nucleation, we anticipate that our findings could inform the rational design of corrosion mitigation strategies in the future.
Methods
Bulk Structures
UO2 is modeled as a fluorite structure (cubic Fm3̅m) with Wycoff positions 4b (i.e., ) for uranium and 8c (i.e., ) for oxygen atoms. While its ground state electronic structure consists of a transverse 3k antiferro-magnetic ordering, , here, we use a simpler collinear anti-ferromagnetic ordering in the z-direction. This strategy is computationally more efficient and has been previously shown to reproduce the UO2 band gap.
Grain Boundaries
Grain boundaries are often classified as general GBs or coincident site lattice (CSL) GBs. The CSL GBs are formed by rotation of one lattice so as to coincide with the atomic positions of the other. While general GBs lack symmetric order and are predominant in UO2, CSL GB models are advantageous as their inherent symmetry results in relatively small unit cells that are amenable for DFT calculations. We included Σ9, Σ5, and Σ3 GBs in our analysis as these GBs are frequently observed in experiments. Table summarizes the key characteristics of the 3 GB used in this study.
1. Overview of the Three Most Frequently Observed UO2 CSL GBs Considered in This Study.
| GB Type | Rotation Axis | GB Plane | Misorientation Angle (deg) | Total Atoms In Supercell |
|---|---|---|---|---|
| Σ3 | [110] | [11̅1̅] | 70.5 | 432 |
| Σ9 | [1̅14̅] | [2̅21] | 180 | 408 |
| Σ5 | [001] | [310] | 36.8 | 228 |
The atomistic structures of the three CSL GBs were generated using pymatgen following the procedure outlined by Grimmer. Overlapping atoms at the interface were removed if they were closer than 64, 90, or 64% of their pure crystal bond lengths for Σ3, Σ5, and Σ9, respectively. These thresholds were chosen to maintain the stoichiometry of UO2. Sufficiently large unit cells were used to avoid interactions of adsorbed hydrogen with its periodic images. Similar approaches have been used by Williams et al. to study oxygen diffusion in UO2 CSL GBs.
Identification of Binding Sites
We have used a Wigner–Seitz cell construction to search for potential binding sites for hydrogen, with a distance cutoff of 0.9 A from the nearest neighbor. We have thus created an exhaustive list of candidate hydrogen adsorption sites across three GBs. Duplicates were avoided by visual inspection. Similar protocols as above were used to generate structures for H2 adsorption. Additional details are presented in the Supporting Information (SI).
DFT+U Calculations
The Vienna ab initio simulation package (VASP) − was used for all DFT calculations. We use the PBESol exchange–correlation functional because of its small errors in reproducing hydrogen chemisorption in UO2. Additional dispersion corrections were not included since bonded interactions are expected to dominate. Our calculations have used a plane wave energy cutoff of 500 eV. The reciprocal space was sampled using 1 × 1 × 1, 2 × 2 × 1, 1 × 1 × 1 grids generated using the Monkhorst–Pack method for Σ3, Σ5, and Σ9, respectively. These computational settings were chosen after extensive benchmarking as shown in Figure S1.
The Hubbard U approach was used to reduce self-interaction errors using a U eff of 3.6 eV. This value was obtained using a self-consistent linear response approach (Figure S2). The +U corrections are critical as standard DFT calculations result in an incorrect metallic ground state. Spin orbit coupling was not included in our calculations in order to reduce the computational costs.
The conjugate gradient algorithm was used for geometry optimization, with the residual minimization method with direct inversion in the iterative subspace (RMM-DIIS) , for converging the electron density. The geometric optimizations were terminated when the forces on all atoms were below 0.05 eV/Å and total energy change was ≤10–5 eV. A sequential optimization protocol was used to accelerate these calculations. Specifically, we used 40 optimization steps at lower plane wave energy cutoffs (i.e., 300 eV), that were followed by a full optimization at 500 eV. The cell volume was allowed to relax in pristine GB systems (i.e., without any adsorbates) but was held constant during H and H2 adsorption. Additionally, only the atoms at and near the GB interface were allowed to relax during the adsorption energy calculations (represented by bright atoms in Figure ).
1.

UO2 CSL grain boundaries investigated in our study, with (a) Σ9, (b) Σ5, (c) and Σ3. The oxygen atoms are colored red and the uranium atoms are yellow, excluding those that reside in the GB, which are colored green. The bright colored atoms indicate those that were allowed to move during hydrogen adsorption optimization.
The adsorption (binding) energies E bind of H and H2 are calculated using the following equations
| 1 |
| 2 |
In this case, E H+GB and E H2+GB are the total DFT computed energies of the GB with H or H2, respectively, E GB is the GB total energy, and E H2 is the total energy of molecular hydrogen. Owing to the complexity of the underlying electronic structure, we note that H adsorption sometimes resulted in final magnetic moments that differed significantly from the ground state AFM UO2 structure. These cases were excluded from our analysis; a few examples are presented in Figure S4 for the interested reader. All adsorption energies are reported using a 500 eV plane wave cutoff. Detailed results are presented in Tables S1 and S2.
Results and Discussion
Figure shows the structure of the three CSL GBs used in this work. The solid green and yellow atoms represent the uranium atoms located at and near the interface, respectively. While a one-atom thick interface is observed for Σ3, we observe distinct interface patterns in our our Σ5 and Σ9 GB models, referred to as kites. , While the remainder of this work will focus on investigating hydrogen adsorption in bulk GBs, we note that we also briefly considered the role of surface GBs using UO2 slab models. However, these calculations resulted in a metallic ground state (i.e., similar to Si), and are thus not discussed further.
We first discuss the relative stability of the three GBs explored in our study. The GB surface formation energies (E GB ) have been calculated using the equation
| 3 |
Here, E UO2 is the UO2 energy per formula unit, and N is the number of formula units in the GB system. The results are summarized in Table . Our calculations show that Σ3 GB is the most stable interface followed by Σ9 and Σ5, which is the least stable.
2. Details of GB System Sizes, E GB , E bind(H), and E bind(H).
| GB | Supercell Size (atoms) | EGB (ev/Å2) | Initial H Sites (Wigner–Seitz algorithm) | Duplicate H Sites Removed | Final H Binding Sites | Initial H2 Sites | H2 Sites Added | Final H2 Sites |
|---|---|---|---|---|---|---|---|---|
| Σ3 | 432 | 0.075 | 129 | 13 | 116 | 8 | 1 | 9 |
| Σ9 | 408 | 0.100 | 28 | 6 | 22 | 9 | 2 | 11 |
| Σ5 | 228 | 0.121 | 9 | 0 | 9 | 7 | 3 | 10 |
As these GB defects are the preferred sites for hydrogen adsorption, we have developed an automated high-throughput workflow (Figure ) that exhaustively samples all unique binding sites. Specifically, using a Wigner–Seitz cell construction, we identified 166 candidate sites for H binding across all three GBs. Of these, 19 duplicate sites were removed. The selected H sites were also used to guide the search for potential sites for H2 adsorption. Specifically, to sample a diverse range of possible binding environments, we selected five of the most exothermic and five of the most endothermic H binding sites for each GB in this study, which were visually inspected to ensure proper sampling of the GB interface. However, six sites converged outside the region of interest and were therefore excluded. In addition, six additional H2 sites total were included to comprehensively sample the GB interfaces. We note that our algorithm for H2 insertion is used only to obtain initial guess sites for H2 adsorption, and that other approaches for sampling the GB binding sites are also valid. The resulting set of H and H2 sites (Table ) was used in our DFT calculations. Although not explored here, we note that adsorption of additional H atoms resulted in unexpected changes to the magnetic ordering of the UO2 GB. Exploration of this phenomenon is the subject of future work.
2.
Schematic illustrating the workflow used to generate and test various binding sites across the GB systems. More information can be found in the Figure S4.
We have observed that these simulations are computationally intensive due to the size of the unit cell at average of 400 atoms and the complexity of the electronic structure of UO2. For instance, a typical calculation of H binding energy required ca. 56 h (wall-clock time) using two NVIDIA H100 GPUs; further details are available in Figure S5. Thus, to facilitate access to our optimized structures, we have provided our results as an Atomic Simulation Environment (ASE) database in the SI.
The sites identified above were used for adsorption energy calculations at a PBESol+U level of theory. Figure shows the distribution of H binding energies for the three CSL GBs. As expected, our DFT calculations show that H adsorption in GB is significantly more favorable compared to the pristine oxide. Indeed, a majority of the GB sites show binding energies ranging from 0.5–1.5 eV with a few outliers at lower and higher values. Specifically, the strongest H binding (−0.17 eV) is observed for the Σ5 GB. The most favorable sites for Σ3 (+0.37 eV) and Σ9 (+0.45 eV) are less stable, suggesting preferential H adsorption within Σ5. In contrast, similar analysis for H2 adsorption reveals that the Σ9 GB is preferred. Specifically, we find several exothermic H2 binding sites (ca. −0.27 eV) for Σ9 that are more favorable than the most stable H2 sites in Σ5 (+0.74 eV) or Σ3 (+0.82 eV). Tables S1 and S2 summarizes the adsorption energies of all sites considered in this study.
3.
Scatter plot showing (a) H and (b) H2 binding energies, with the most stable octahedral binding site in pristine UO2 (at 1.56 eV) is shown as a black dashed line for reference.
To rationalize the above adsorption energy trends, we now investigate the structure of preferred GB binding sites in more detail. As a reference, Figure (a),(b) shows the DFT-optimized structure of an adsorbed H atom in the distorted octahedral (more stable, 1.56 eV) and tetrahedral site (2.48 eV) of pristine UO2. H adsorption in the octahedral site is preferred by 0.92 eV, and likely arises due to the proximity of several U atoms in the coordination sphere and the absence of any neighboring oxygen atoms.
4.
Visualizations of several adsorption structures created using VESTA, namely the UO2 (a) octahedral and (b) tetrahedral binding sites, (c) the most favorable H Σ3 site at +0.37 eV, (d) the exothermic site observed for H in Σ5, and (e) an H2 exothermic binding site in a Σ9 kite. All shown GB adsorption sites occur with no oxygen nearest neighbors to H or H2, and where the adsorbate is coordinated by four or less uranium atoms. The numbers in insets are the uranium distances in Å from the closest hydrogen.
In contrast to the high-symmetry sites observed in the pristine oxide, the most stable H-bound sites for Σ3 and Σ5 UO2 GBs show completely different local coordination environments. While the H atom in the Σ3 site is surrounded by four uranium atoms (avg distance of 2.45 Å), the most stable Σ5 site features only 3 nearby U atoms (avg. U–H distance of 2.33 Å). Additionally, we find that the H atom in Σ5 is located closer to one of the U atoms (shortest U–H = 2.17 Å), while it is located more centrally for Σ3 (shortest U–H = 2.31). In both cases, and in contrast from the tetrahedral site of the pristine oxide, no neighboring oxygen atoms are observed.
Inspection of the most stable H2 bound sites in Σ9 shows two different types of preferred binding locations. These sites, as shown in Figure (e) (E bind(H2) = −0.27 eV) and Figure S7 (E bind(H2) = −0.11 eV) show an absence of oxygen atoms with shortest U–H distances of 2.57 and 2.59 Å, respectively. These results can help explain why H2 adsorption is favored in Σ9, which provides larger interstitial cavities compared to Σ3 and Σ5. Furthermore, a comparison of E bind for H and H2 for similar sites is discussed in SI.
Motivated by qualitative insights into the factors that impact H (and H2) adsorption in UO2 GBs for the most stable sites, we have also attempted to develop predictive models for H binding energies using our DFT data set. Our results using linear regression (Figure S9) and Gaussian processes regression (GPR, in Figure S10) achieved mean absolute errors of 0.27 and 0.18 eV, respectively. However, detailed analysis using K-fold cross validation suggested that the GPR was not generalizable to different subsets of the data (Figure S11) Although this can perhaps be resolved by using larger DFT data sets, we have not pursued this direction in this work as our primary interest lies in identifying the favorable H binding sites. Details of the fitting procedure and it is limitations are provided in the SI.
Conclusions
In this work, we have utilized the DFT+U method to investigate hydrogen adsorption in three different grain boundaries (GBs) in bulk UO2. This was achieved by developing a high-throughput DFT workflow that systematically samples 133 plausible H and H2 binding across Σ3, Σ5, and Σ9 GBs. Although the presence of the GB defect results in stronger hydrogen adsorption (compared to the pristine oxide), we find that the trends in binding energy vary considerably across the three GBs. Specifically, we find a few sites in Σ5 and Σ9 GBs that show favorable (i.e., exothermic) H adsorption, as well as several exothermic H2 binding sites in Σ9. A careful geometric analysis of the binding configurations suggests that hydrogen prefers to occupy sites that are located close to U atoms. Our analysis concludes that uranium-rich GB sites in defective bulk UO2 are favored for H adsorption and thus, should be further investigated to develop corrosion mitigation strategies in the future.
Supplementary Material
Acknowledgments
This work was funded by a subcontract to the University of California, Davis from Lawrence Livermore National Laboratory (LLNL) Master Agreement No. B654505. Computational resources at LLNL institutional computing facility were utilized and authors specifically thank Don Frederick (LLNL) for supporting this resource. Other computational resources were sponsored by the Department of Energy’s Office of Energy Efficiency and Renewable Energy and located at the National Renewable Energy Laboratory, allocation CHE230058 from the Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support (ACCESS) program, which is supported by National Science Foundation, and the high performance computing facility at University of California, Davis. The authors thank Professor Jeremy Mason and Professor Louise Berben at UC Davis for insightful discussions and feedback on the research direction. VMD , was used to create all GB visualizations, and matplotlib, seaborn and pandas were used to create various plots. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344.
Atomic simulation environment database files (ase-db) consisting of DFT optimized structures are available on GitHub https://github.com/kul-group/kul-repository/tree/main/repo_2025_hyd_in_UO2_GB.
The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jpcc.5c01120.
Benchmarking analysis for plane-wave energy cutoff and k-point mesh with respect to unit cell energy and volume; k-point mesh for GBs was converged and shown in above subplots; wall time for three different GB systems with hydrogen, and final H binding energies sorted in ascending order (PDF)
The authors declare no competing financial interest.
References
- Cubicciotti D.. The reaction between uranium and oxygen. J. Am. Chem. Soc. 1952;74:1079–1081. doi: 10.1021/ja01124a505. [DOI] [Google Scholar]
- Guéneau C., Baichi M., Labroche D., Chatillon C., Sundman B.. Thermodynamic assessment of the uranium-oxygen system. J. Nucl. Mater. 2002;304:161–175. doi: 10.1016/S0022-3115(02)00878-4. [DOI] [Google Scholar]
- Ritchie A., Greenwood R., Randles S.. The kinetics of the uranium-oxygen-water vapour reaction between 40 and 100° C. J. Nucl. Mater. 1986;139:121–136. doi: 10.1016/0022-3115(86)90030-9. [DOI] [Google Scholar]
- Halpern J., Smith J.. Kinetics of the oxidation of uranium (IV) by molecular oxygen in aqueous perchloric acid solution. Can. J. Chem. 1956;34:1419–1427. doi: 10.1139/v56-182. [DOI] [Google Scholar]
- Ritchie A.. A review of the rates of reaction of uranium with oxygen and water vapour at temperatures up to 300 C. J. Nucl. Mater. 1981;102:170–182. doi: 10.1016/0022-3115(81)90557-2. [DOI] [Google Scholar]
- Martin T. L., Coe C., Bagot P., Morrall P., Smith G. W., Scott T., Moody M.. Atomic-scale studies of uranium oxidation and corrosion by water vapour. Sci. Rep. 2016;6:25618. doi: 10.1038/srep25618. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Banos A., Hallam K., Scott T.. Corrosion of uranium in liquid water under vacuum contained conditions. Part 1: The initial binary U+ H2O (l) system. Corros. Sci. 2019;152:249–260. doi: 10.1016/j.corsci.2019.02.017. [DOI] [Google Scholar]
- Sinkov, S. I. ; Delegard, C. H. ; Schmidt, A. J. . Mitigation of Hydrogen Gas Generation from the Reaction of Water with Uranium Metal in K Basins Sludge; U.S. Department of Energy, 2010. [Google Scholar]
- Burke J. E., Smith C. S.. The formation of uranium hydride1. J. Am. Chem. Soc. 1947;69:2500–2502. doi: 10.1021/ja01202a073. [DOI] [PubMed] [Google Scholar]
- Harker R. M.. The influence of oxide thickness on the early stages of the massive Uranium-hydrogen reaction. J. Alloys Compd. 2006;426:106–117. doi: 10.1016/j.jallcom.2006.02.014. [DOI] [Google Scholar]
- Bloch J., Simca F., Kroup M., Stern A., Shmariahu D., Mintz M., Hadari Z.. The initial kinetics of uranium hydride formation studied by a hot-stage microscope technique. J. Less Common Met. 1984;103:163–171. doi: 10.1016/0022-5088(84)90375-8. [DOI] [Google Scholar]
- Banos A., Scott T. B.. Statistical analysis of UH3 initiation using electron back-scattered diffraction (EBSD) Solid State Ionics. 2016;296:137–145. doi: 10.1016/j.ssi.2016.09.018. [DOI] [Google Scholar]
- Owen L., Scudamore R.. A microscope study of the initiation of the hydrogen-uranium reaction. Corros. Sci. 1966;6:461–468. doi: 10.1016/S0010-938X(66)80063-X. [DOI] [Google Scholar]
- Harker R. M.. The influence of oxide thickness on the early stages of the Massive uranium-hydrogen reaction. J. Alloys Compd. 2006;426:106–117. doi: 10.1016/j.jallcom.2006.02.014. [DOI] [Google Scholar]
- Banos A., Harker N., Scott T.. A review of uranium corrosion by hydrogen and the formation of uranium hydride. Corros. Sci. 2018;136:129–147. doi: 10.1016/j.corsci.2018.03.002. [DOI] [Google Scholar]
- Stitt C. A., Harker N., Hallam K., Paraskevoulakos C., Banos A., Rennie S., Jowsey J., Scott T.. An investigation on the persistence of uranium hydride during storage of simulant nuclear waste packages. PLoS One. 2015;10:e0132284. doi: 10.1371/journal.pone.0132284. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gibb, T. R. P., Jr. Hydrides And Metal-Hydrogen Systems; Fairchild Engine and Airplane Corporation, 1951. [Google Scholar]
- Solbrig, C. W. ; Krsul, J. R. ; Olsen, D. N. . Pyrophoricity of Uranium in Long-Term Storage Environments; U.S. Ajpniic Energy Commission, 1994. [Google Scholar]
- Taylor C. D., Lookman T., Lillard R. S.. Ab initio calculations of the uranium-hydrogen system: Thermodynamics, hydrogen saturation of α-U and phase-transformation to UH3. Acta Mater. 2010;58:1045–1055. doi: 10.1016/j.actamat.2009.10.021. [DOI] [Google Scholar]
- Li H., Wang S., Li G., Wu H., Pu Z., Bin R., Meng D., Luo W.. UH3 produced at the initial stage of U-H2O corrosion. Corros. Sci. 2023;217:111143. doi: 10.1016/j.corsci.2023.111143. [DOI] [Google Scholar]
- Jones C. P., Scott T. B., Petherbridge J. R., Glascott J.. A surface science study of the initial stages of hydrogen corrosion on uranium metal and the role played by grain microstructure. Solid State Ionics. 2013;231:81–86. doi: 10.1016/j.ssi.2012.11.018. [DOI] [Google Scholar]
- Sunder S., Miller N., Shoesmith D.. Corrosion of uranium dioxide in hydrogen peroxide solutions. Corros. Sci. 2004;46:1095–1111. doi: 10.1016/j.corsci.2003.09.005. [DOI] [Google Scholar]
- Mohun R., Middleburgh S., Thomas P., Corkhill C.. Enhanced radiation damage tolerance in Zr-doped UO2. Materialia. 2024;38:102226. doi: 10.1016/j.mtla.2024.102226. [DOI] [Google Scholar]
- Broczkowski M., Noël J., Shoesmith D.. The inhibiting effects of hydrogen on the corrosion of uranium dioxide under nuclear waste disposal conditions. J. Nucl. Mater. 2005;346:16–23. doi: 10.1016/j.jnucmat.2005.04.070. [DOI] [Google Scholar]; Long-term Evolution of Spent Nuclear Fuel - A selection of papers from the Spent Fuel Workshop 2004
- Zheng G., Wang Q., Chen J., Qiu R., Zhu M.. Quantifying Local Atomic Distortions in UO2 Grain Boundaries: Correlation with Energetic and Electronic Properties. Inorg. Chem. 2023;62:16047–16058. doi: 10.1021/acs.inorgchem.3c02179. [DOI] [PubMed] [Google Scholar]
- Zhang Y., Wang B., Lu Y., Yang Y., Zhang P.. Electronic, mechanical and thermodynamic properties of α-UH3: A comparative study by using the LDA and LDA+ U approaches. J. Nucl. Mater. 2012;430:137–141. doi: 10.1016/j.jnucmat.2012.07.002. [DOI] [Google Scholar]
- Soshnikov B. A., Kulkarni A., Goldman N.. Elucidating the initial steps in α-uranium hydriding using first-principles calculations. Langmuir. 2022;38:9335–9346. doi: 10.1021/acs.langmuir.2c01170. [DOI] [PubMed] [Google Scholar]
- Soshnikov A., Lindsey R., Kulkarni A., Goldman N.. A reactive molecular dynamics model for uranium/hydrogen containing systems. J. Chem. Phys. 2024;160:094117. doi: 10.1063/5.0183610. [DOI] [PubMed] [Google Scholar]
- Harker R. M.. The influence of oxide thickness on the early stages of the massive uranium-hydrogen reaction. J. Alloys Compd. 2006;426:106–117. doi: 10.1016/j.jallcom.2006.02.014. [DOI] [Google Scholar]
- Darnbrough J., Harker R., Griffiths I., Wermeille D., Lander G., Springell R.. Interaction between U/UO2 bilayers and hydrogen studied by in-situ X-ray diffraction. J. Nucl. Mater. 2018;502:9–19. doi: 10.1016/j.jnucmat.2018.01.031. [DOI] [Google Scholar]
- Nerikar P. V., Rudman K., Desai T. G., Byler D., Unal C., McClellan K. J., Phillpot S. R., Sinnott S. B., Peralta P., Uberuaga B. P., Stanek C. R.. Grain boundaries in uranium dioxide: scanning electron microscopy experiments and atomistic simulations. J. Am. Ceram. Soc. 2011;94:1893–1900. doi: 10.1111/j.1551-2916.2010.04295.x. [DOI] [Google Scholar]
- Bourasseau E., Onofri C., Ksibi A., Iltis X., Belin R. C., Lapertot G.. Atomic structure of grain boundaries in UO2 Bicrystals: A coupled high resolution transmission electron Microscopy/Atomistic simulation approach. Scr. Mater. 2022;206:114191. doi: 10.1016/j.scriptamat.2021.114191. [DOI] [Google Scholar]
- Yang L., Wirth B.. Tilt grain boundary stability in uranium dioxide and effect on xenon segregation. J. Nucl. Mater. 2023;577:154302. doi: 10.1016/j.jnucmat.2023.154302. [DOI] [Google Scholar]
- Zhang L., Wang L.-F., Sun B., Liu H., Li G., Yu H.-L., Zhang Q., Song H.. Theoretical study on the hydrogen capture and damage mechanisms of PuO2 nanograin boundary. J. Solid State Chem. 2022;313:123314. doi: 10.1016/j.jssc.2022.123314. [DOI] [Google Scholar]
- Willis B.. Structures of UO2, UO2. x andU4O9 by neutron diffraction. J. Phys. 1964;25:431–439. doi: 10.1051/jphys:01964002505043100. [DOI] [Google Scholar]
- Pegg J. T., Shields A. E., Storr M. T., Scanlon D. O., De Leeuw N. H.. Noncollinear relativistic DFT+ U calculations of actinide dioxide surfaces. J. Phys. Chem. C. 2019;123:356–366. doi: 10.1021/acs.jpcc.8b07823. [DOI] [Google Scholar]
- Blackburn E., Caciuffo R., Magnani N., Santini P., Brown P., Enderle M., Lander G.. Spherical neutron spin polarimetry of anisotropic magnetic fluctuations in UO 2. Phys. Rev. B. 2005;72:184411. doi: 10.1103/PhysRevB.72.184411. [DOI] [Google Scholar]
- Pegg J. T., Shields A. E., Storr M. T., Wills A. S., Scanlon D. O., De Leeuw N. H.. Magnetic structure of UO 2 and NpO 2 by first-principle methods. Phys. Chem. Chem. Phys. 2019;21:760–771. doi: 10.1039/C8CP03581D. [DOI] [PubMed] [Google Scholar]
- Ong S. P., Richards W. D., Jain A., Hautier G., Kocher M., Cholia S., Gunter D., Chevrier V. L., Persson K. A., Ceder G.. Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysis. Comput. Mater. Sci. 2013;68:314–319. doi: 10.1016/j.commatsci.2012.10.028. [DOI] [Google Scholar]
- Grimmer H.. The generating function for coincidence site lattices in the cubic system. Acta Crystallogr., Sect. A: Found. Crystallogr. 1984;40:108–112. doi: 10.1107/S0108767384000246. [DOI] [Google Scholar]
- Williams N. R., Molinari M., Parker S. C., Storr M. T.. Atomistic investigation of the structure and transport properties of tilt grain boundaries of UO2. J. Nucl. Mater. 2015;458:45–55. doi: 10.1016/j.jnucmat.2014.11.120. [DOI] [Google Scholar]
- Kresse G., Hafner J.. Ab initio molecular dynamics for liquid metals. Phys. Rev. B. 1993;47:558. doi: 10.1103/PhysRevB.47.558. [DOI] [PubMed] [Google Scholar]
- Kresse G., Furthmüller J.. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 1996;6:15–50. doi: 10.1016/0927-0256(96)00008-0. [DOI] [PubMed] [Google Scholar]
- Kresse G., Furthmüller J.. Efficient iterative schemes for ab initio total-energy calculations using a plane-wave basis set. Phys. Rev. B. 1996;54:11169. doi: 10.1103/PhysRevB.54.11169. [DOI] [PubMed] [Google Scholar]
- Perdew J. P., Ruzsinszky A., Csonka G. I., Vydrov O. A., Scuseria G. E., Constantin L. A., Zhou X., Burke K.. Restoring the density-gradient expansion for exchange in solids and surfaces. Phys. Rev. Lett. 2008;100:136406. doi: 10.1103/PhysRevLett.100.136406. [DOI] [PubMed] [Google Scholar]
- Pegg J. T., Shields A. E., Storr M. T., Scanlon D. O., de Leeuw N. H.. Interaction of hydrogen with actinide dioxide (011) surfaces. J. Chem. Phys. 2020;153:014705. doi: 10.1063/5.0010200. [DOI] [PubMed] [Google Scholar]
- Monkhorst H. J., Pack J. D.. Special points for Brillouin-zone integrations. Phys. Rev. B. 1976;13:5188. doi: 10.1103/PhysRevB.13.5188. [DOI] [Google Scholar]
- Liechtenstein A. I., Anisimov V. I., Zaanen J.. Density-functional theory and strong interactions: Orbital ordering in Mott-Hubbard insulators. Phys. Rev. B. 1995;52:R5467. doi: 10.1103/PhysRevB.52.R5467. [DOI] [PubMed] [Google Scholar]
- Kulik H. J., Marzari N.. Systematic study of first-row transition-metal diatomic molecules: A self-consistent DFT+U approach. J. Chem. Phys. 2010;133:114103. doi: 10.1063/1.3489110. [DOI] [PubMed] [Google Scholar]
- Press, W. H. Numerical Recipes 3rd Ed.: The Art of Scientific Somputing; Cambridge University Press, 2007. [Google Scholar]
- Wood D. M., Zunger A.. A new method for diagonalising large matrices. J. Phys. A: Math. Gen. 1985;18:1343. doi: 10.1088/0305-4470/18/9/018. [DOI] [Google Scholar]
- Pulay P.. Convergence acceleration of iterative sequences. the case of scf iteration. Chem. Phys. Lett. 1980;73:393–398. doi: 10.1016/0009-2614(80)80396-4. [DOI] [Google Scholar]
- Sun C., Goel R., Kulkarni A. R.. Developing Cheap but Useful Machine Learning-Based Models for Investigating High-Entropy Alloy Catalysts. Langmuir. 2024;40:3691–3701. doi: 10.1021/acs.langmuir.3c03401. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Frolov T., Olmsted D. L., Asta M., Mishin Y.. Structural phase transformations in metallic grain boundaries. Nat. Commun. 2013;4:1899. doi: 10.1038/ncomms2919. [DOI] [PubMed] [Google Scholar]
- Zhu Q., Samanta A., Li B., Rudd R. E., Frolov T.. Predicting phase behavior of grain boundaries with evolutionary search and machine learning. Nat. Commun. 2018;9:467. doi: 10.1038/s41467-018-02937-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sagisaka K., Nara J., Bowler D.. Importance of bulk states for the electronic structure of semiconductor surfaces: implications for finite slabs. J. Phys.: Condens. Matter. 2017;29:145502. doi: 10.1088/1361-648X/aa5f91. [DOI] [PubMed] [Google Scholar]
- Larsen A. H., Mortensen J. J., Blomqvist J., Castelli I. E., Christensen R., Dułak M., Friis J., Groves M. N., Hammer B., Hargus C.. et al. The atomic simulation environment–a Python library for working with atoms. J. Phys.: Condens. Matter. 2017;29:273002. doi: 10.1088/1361-648x/aa680e. [DOI] [PubMed] [Google Scholar]
- Momma K., Izumi F.. VESTA3 for three-dimensional visualization of crystal, volumetric and morphology data. J. Appl. Crystallogr. 2011;44:1272–1276. doi: 10.1107/S0021889811038970. [DOI] [Google Scholar]
- Boerner, T. J. ; Deems, S. ; Furlani, T. R. ; Knuth, S. L. ; Towns, J. . ACCESS: Advancing Innovation: NSF’s Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support.. In Practice and Experience in Advanced Research Computing 2023: Computing for the Common Good; ACM: New York, NY, USA, 2023; pp 173–176. [Google Scholar]
- Humphrey W., Dalke A., Schulten K.. VMD - Visual Molecular Dynamics. J. Mol. Graphics. 1996;14:33–38. doi: 10.1016/0263-7855(96)00018-5. [DOI] [PubMed] [Google Scholar]
- Stone, J. An Efficient Library for Parallel Ray Tracing and Animation, Ph.D. Thesis; University of Missouri-Rolla, 1998. [Google Scholar]
- Hunter J. D.. Matplotlib: A 2D graphics environment. Comput. Sci. Eng. 2007;9:90–95. doi: 10.1109/MCSE.2007.55. [DOI] [Google Scholar]
- Waskom M. L.. Seaborn: statistical data visualization. J. Open Source Software. 2021;6:3021. doi: 10.21105/joss.03021. [DOI] [Google Scholar]
- McKinney, W. Data Structures for Statistical Computing in Python, Proceedings of the 9th Python in Science Conference; SciPy, 2010; pp 56–61. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Atomic simulation environment database files (ase-db) consisting of DFT optimized structures are available on GitHub https://github.com/kul-group/kul-repository/tree/main/repo_2025_hyd_in_UO2_GB.



