Symposium S.CT05 : Defects, Order and Disorder in Structural and Functional Fluorite-Related Compounds

In this presentation, we will compare and contrast the structures of oxide compounds containing lanthanum (La) and either tungsten (W) or uranium (U). We will interrogate the structures of La-tungstates (LWO) versus La-uranates (LUO) using a geometric model based on layer stacking sequences of pseudo close-packed layers. The individual atomic layers in the stacking model considered here have hexagonal symmetry and are based on triangular atom nets or hexanets. The layers in the LWO and LUO compounds can be described based on geometric subdivisions of a triangular atom net, subdivisions originally described by Shuichi Iida.^§ One unusual layered atom structure involves the LWO compound, La₇W₁O_13.5. This compound has a stoichiometry very close to La₆W₁O₁₂, which is the stoichiometry associated with 6-1-12 delta (δ) phase oxides (actually, the 6-1-12 compound is not observed in LWO). However, the structure of the 7-1-13.5 stoichiometry is very different compared to the δ-phase. The layer stacking model we develop for the 7-1-13.5 compound is closer to that of pyrochlore, rather than δ-phase. Interestingly, in LUO, the 6-1-12 phase is observed and we can describe the layer stacking in this compound as consistent with a δ-phase structure. We will perform a detailed comparison between layered atom arrangements in La₇W₁O_13.5 and La₆U₁O₁₂ in an effort to rationalize why the structures of these two compounds, both containing hexavalent metal cations (W⁶⁺ and U⁶⁺), are so different.
^§ S. Iida, "Layer Structures of Magnetic Oxides," J. Phys. Soc. Japan 12 (3) (1957) 222-233.

Structural disorder plays a critical role in many technological applications to enhance specific functionalities, such as increasing conductivity via chemical doping in semiconductors or improving flux pinning in high-temperature superconductors. Recent results from neutron total scattering experiments have shown that the atomic arrangements of many disordered crystalline materials are not random nor are they represented by the long-range structure observed from diffraction experiments. This structural heterogeneity at different length-scales appears to be a general characteristic of disordered crystalline materials, but the underlying mechanisms are not well understood. This presentation reviews neutron total scattering experiments on a range of fluorite-related compounds that show complex disordering behavior across different length scales. Structural information from pair distribution functions with sensitivity to both cation and anion sublattices were utilized to investigate in detail the disordering process of A₂B₂O₇ pyrochlore oxides. The order-disorder transformation can be understood as a rearrangement of atomic-scale building blocks. The same concept can be used to describe disordering in weberite-type oxides (A₃BO₇). This knowledge has implications for tailoring the structural response of complex oxides to extreme environments. For example, by choosing a material with specific structural building blocks the resistance to ion irradiation can be greatly improved.

Uranium dioxide is the main constituent of most current nuclear fuels that is doped with fission products during its use in nuclear power plants. This industrial topic motivated many studies of UO₂ and doped UO₂. UO₂ crystalline structure is fluorite and it is considered to form ideal solid solutions with most of the cationic dopants generated as fission products during its operation. However, this historical description is now challenged by new experimental results that enlighten the existence of disorder and non-ideal solid solution behaviour. In this presentation, we will discuss some of these results and their implication on some key features regarding nuclear fuel behaviour will be analysed.
Firstly, the disorder at oxygen atom crystallographic site will be reviewed using neutron diffraction data on UO₂ as a function of temperature. Its implications on UO₂ properties will be discussed regarding specific heat and UO₂ irradiation resistance.
Secondly, the oxidation mechanism of UO₂ will be considered. Neutron diffraction data evidenced the existence of oxygen clusters, named cuboctahedra that accumulate in the pristine UO₂ lattice with several crystallographic arrangements. The morphological evolution of spent UO₂ fuel submitted to air atmosphere at temperature above 200°C, which is representative of an accidental scenario during air storage of spent nuclear fuel, will be analysed as a function of these crystallographic changes.
Finally, fission product incorporation in UO₂ will be presented and we will discuss some new research challenges this issue brings up.

The basic fluorite structure allows for both ordering and disorder, with multiple options in each such as those that are regular integer repeats of the unit cell, such as with pyrochlore, or those that repeat with a different repeat length. The degree of order/disorder within the atomic structure can often lead to modification of properties, such as ionic conducitivity, or how they respond to induced radiation damage. Enhanced understanding of the degree of short/long range order greatly improves the design and development of new materials. The determination of the degree or order/disorder can be chalenging given the cubic nature of the fluorite structure. To get round such challenges a combination of experiemtnal techniques can be applied, however each technique in itself has different sensitiviites and aability to proivde structural information. This presentation provide an overview of mutliple options and how they can be applied to the determination of order/disorder in fluorites, and how the derived information can be used to develop new mateiral options.

Crystalline, mesoporous NiO/GDC thin films with thicknesses ranging from 50 nm to 250 nm were synthesized through templated sol–gel chemistry coupled with the dip-coating process and heat-treatment in air. The thin films' microstructure is composed of two interpenetrated networks made of mesopores and inorganic components. Efficient coupling between the temperature and the NiO volume fraction (vol%) allows tuning of both the pore size with dimensions ranging from macro- to meso-size and the NiO or GDC crystallite size with a diameter below 10 nm. X-ray diffraction and impedance spectroscopy performed in 10% H₂ in Ar allowed the in situ study of the reduction process of NiO to metallic Ni. Coalescence of Ni particles generated by the reduction step and percolation phenomena controls the resulting conductivity of the final materials. Thin films with an initial content of 50 vol% of NiO exhibit lower electrical properties compared to those with an initial content of 70 vol% NiO. Electrical properties in a reducing atmosphere were also studied as a function of microstructure such as the pore dimension and the thickness of the pore wall. Excellent electrical properties are obtained for Ni70/GDC30-porous thin films synthesized with the block-co-polymer PS40-PEO36 that have a final conductivity of 9 x 10⁴ S m⁻¹ at 500 °C. These mesostructured nanocomposite thin films exhibit a connected pore network that ensures good gas diffusion and good particle–particle contact for GDC and Ni, which gives satisfactory electrical properties. These films have all the attributes to be used as anodes in micro-SOFCs. This approaches was also extended to mesostructured composite La_0.7Sr_0.3Co_0.2Fe_0.8O₃x/yGd–CeO₂ (LSCFx/yGDC) (with x and y representing the volumetric % of LSCF and GDC in the composite films) thin films. This technique allows us to define the heat-treatment, the LSCF content in the composite film and the thickness of the film to achieve good electrical conductivity. Electrical conductivity of 250 S cm⁻¹ was achieved for the pure, mesoporous LSCF film at 700 °C.

References:
Baldinozzi et al., J. Phys. Chem. C., 2012, 116, 7658-7663.
Muller et al., J. Mater. Chem., 2012,22, 9368-9373.
Muller et al., J. Mater. Chem. A, 2013,1, 10753-10761.
Muller et al., J. Mater. Chem. A, 2014,2, 6448-6455.

It is becoming ever more apparent that, in complex compounds such as pyrochlores, the detailed arrangement of the cations drives functionality. For example, both radiation tolerance and ionic conductivity have been linked to how easily cations can mix across sublattices. This is due to the fact that mass transport in these materials is a strong function of the cation distributions. However, the actual relationship between the cation, or chemical, structure of these compounds and the rates of transport are still not well established. While some reports find enhanced ionic conductivity in disordered materials, others find higher conductivity in ordered phases. Many of these studies use chemistry to influence the disorder, essentially changing multiple variables at once. However, the degree to which cations mix can be finely controlled using radiation damage without changing chemistry. Here, we use radiation damage as a tool to induce changes in the cation structure of thin-film model pyrochlores. We then characterize the extent to which those changes impact mass transport. We combine these experimental efforts with state-of-the-art simulation methodologies to understand how atomic scale mechanisms dictating mass transport change when the cation structure is modified. We have found that even small changes in cation structure can lead to large changes in the transport characteristics of the material. Our results provide new insight into mass transport in materials that exhibit chemical complexity well beyond the model systems studied where chemical disorder dictates the fundamental behavior of the material.

Complex oxides with fluorite derived structures play a key role in the context of the nuclear fuel cycle. We enhanced wet-chemical synthesis routes, e.g. co-precipitation or internal gelation, to fabricate tailor-made precursors for advanced nuclear fuel forms as well as potential nuclear waste forms with the defect fluorite or pyrochlore crystal structure. The internal gelation route enabled the fabrication of high-performance UO₂ fuel candidates using 1) dopants, such as Cr and Mn, to improve the fission gas retention, and 2) a second phase, here Mo, to enhance the thermal conductivity of UO₂-based nuclear fuel. Characterization of the thermal properties of enhanced UO₂-Mo fuel candidates fabricated via different approaches revealed significant differences in the thermal conductivity of the final fuel candidates. An in-depth understanding about the correlation between the synthesis approaches and the materials properties is underway. Fabrication of Pu-doped Nd₂Zr₂O₇ pyrochlores as potential nuclear waste forms revealed the enormous flexibility of the pyrochlore crystal structure to accommodate Pu(IV) at the usually trivalent A-site. We confirmed our experimental findings by atomistic simulations which resulted in the lowest solution energy for Pu(IV) occupying the pyrochlore's A-site. Charge neutrality is achieved by surplus oxygen at formerly oxygen vacant sites. The immobilization of actinides in a zirconate based pyrochlore is known to result in an order/disorder transition from a pyrochlore to a defect fluorite structure, a consequence to self-irradiation. For this order/disorder transition of Nd₂Zr₂O₇ a combined experimental approach of oxide melt solution calorimetry with ab initio thermodynamic modeling revealed a transition enthalpy of ~30 kJ/mol. This transition enthalpy corresponds to an entropy of disordering of ~16 kJ/mol. This presentation will summarize recent activities about developing suitable fabrication routes of fluorite-structure derived ceramics in the nuclear context, how the fabrication avenues affect their properties as well as the effect of radiation induced structural order/disorder transitions towards the materials properties and performance.

Aliovalent cation-doped fluorite oxide structures are ideal candidates to understand phase stability and their transformation pathways under radiation. These systems are relevant models for nuclear fuels and waste forms. Sc₄Hf₃O₁₂ is a delta phase crystallising in a trigonal symmetry. This and other A₄B₃O₁₂-type systems were observed to transform into a long-range defect fluorite phase upon low energy ion irradiations. On one instance Sc₄Hf₃O₁₂ was also observed to partially turn into a bixbyite phase, however, the reasons behind this were never clearly understood. Thus, in the present study, swift heavy ion irradiated Sc₄Hf₃O₁₂ was investigated as a function of fluence using x-ray diffraction. Structural parameters obtained from quantitative Rietveld analysis reveal that, while the long-range average structure develops into a defect fluorite, the local structure tends to convert into a bixbyite-like arrangement. Similarities in the polyhedra suggest this bixbyite-related local phase is a metastable system that effectively provides radiation resistance by creating large numbers of bixbyite-like replicas within an aristotype fluorite mesocrystal.

Phase formation in multicomponent rare-earth oxides is defined by a competition between the fluorite structure typical of rare-earth dioxides and the hexagonal, monoclinic, and bixbyite structures typical of sesquioxides. This competition is determined by a combination of composition, sintering atmosphere, and cooling rate. Polycrystalline ceramics comprising various combinations of Ce, Gd, La, Nd, Pr, Sm, and Y oxides in equiatomic proportions were synthesized by solid-state sintering. This synthesis method allows a rapid evaluation of the effects of composition (type and number of cations), sintering atmosphere (oxidizing, inert, and reducing), and cooling rate on phase formation. Single fluorite, bixbyite, or monoclinic phase compositions were obtained with a slow cooling of 3.3 C/min, indicating that rare-earth oxides follow a different phase stabilization process than that of transition metal high-entropy oxides. In an oxidizing atmosphere, both Ce and Pr induce the formation of fluorite or bixbyite phases, while only Ce plays that role in an inert or reducing atmosphere. Samples without Ce or Pr develop a single monoclinic phase. The phases formed at initial synthesis may be converted to a different one when the ceramics are annealed in an atmosphere different than the original sintering atmosphere. Additionally, phase evolution of a five-cation composition was studied as a function of sintering temperature. The binary oxides used as raw materials completely dissolve into a single bixbyite structure at 1450 C in air.

U_1-xNp_xO₂ and U_1-xPu_xO₂ mixed dioxides with the fluorite Fm-3m crystalline structures are considered as fuels and targets for the transmutation of the minor actinides in fast neutron reactors. We present an atomic scale structural analysis on a series of U_1−xAn_xO₂ synthesized by the sol−gel external gelation method, for which longer range structural analysis indicates that the process yields solid solutions. The atomic scale structure is probed using high resolution ¹⁷O solid-state NMR in a unique manner. Indeed, this technique allows a particular insight into the local structure of these solid-solutions as different oxygen signals attributed to the O(Np)_y(U)_4−y (with 0 ≤ y ≤ 4) units can be identified by increasing Np content. We thus show how the use of solid-state high resolution NMR opens new routes for the fine characterization of local order in these radioactive materials.

Ternary oxides with pyrochlore structure (A₂B₂O₇) exhibit a range of interesting properties, including ionic conduction, ferromagnetism, superconductivity, high dielectric character and colossal magnetoresistance.¹ Unlike the majority of the tin-based pyrochlores, Bi₂Sn₂O₇ exists in a number of polymorphic forms, with an α-Bi₂Sn₂O₇ to β-Bi₂Sn₂O₇ transition occurring at ≈400 K and β-Bi₂Sn₂O₇ to γ-Bi₂Sn₂O₇ transition occurring above 900 K.² The structural model for the highest temperature γ-Bi₂Sn₂O₇ is undisputed (Fd-3m) but there has been a controversy over the structural models for its lower temperature polymorphs. Previous studies by Evans et al. suggested the room temperature polymorph belongs to space group 7 (P1C1) with 352 atoms per cell.³ Recently, a simpler model for α-Bi₂Sn₂O₇ belonging to space group 9 (C1c1) with 88 atoms per cell, and the first ever detailed structural model for β-Bi₂Sn₂O₇ belonging to space group 41 (Aba2) has been reported.⁴ We perform first principles lattice dynamics calculation using Phonopy⁵ with gradient corrected density functional theory, starting from the highest symmetry γ phase, and map out the potential energy surfaces⁶ spanned out by imaginary mode eigenvectors, with the aim of elucidating the lowest energy structure. This approach successfully takes us from the highest temperature structure to the new structural model suggested for α. It also demonstrates that β is a thermal average of a lower symmetry structure separated by a very small energetic barrier. The success of the method highlights the strength of ab-intio lattice dynamics in predicting the dynamically stable structural model of a compound that undergoes displacive phase transitions and can speed up the exploration of different structures for solid-state applications.

(1) Walsh et al., Chem. Mater., 2007, 19, 5158-5164.
(2) Shannon et al., J. Phys. Chem. Solids, 1980, 41, 117–122.
(3) Evans et al., J. Mater. Chem., 2003, 13, 2098–2103.
(4) Lewis et al., J. Am. Chem. Soc., 2016, 138, 8031–8042.
(5) Togo et al., Phys. Rev. B, 2008, 78, 134106.
(6) Skelton et al., Phys. Rev. Lett., 2016, 117, 075502.

Since the stable valences of U in its oxides are IV, V, and VI, the formation of reduced species in UO₂ is problematic. This is demonstrated in XAFS measurements of UO₂ irradiated with protons and helium nuclei. The addition of O occurs by clustering that results in UO₂-U₄O₉ phase separation, facilitated by the virtually identical U sublattices of these two species. Ion beam irradiation that displaces O ions could give the same behavior, but whereas disorder is observed the signatures of U₄O₉ are not. This is most likely because, whereas the O-enriched regions with mixed valence U(IV,V) would be stable, the U(III) formed in the O-depleted regions is not. An different mechanism is activated. This could also be important in the chemistry of the fuel and the O potential formed in response to the temperature gradient.

Ceramic oxides are used for a wide variety of technologically relevant applications from electrochemical devices, novel resistive switching devices, and oxygen sensors. Applications such as these typically rely upon the ability of oxides to conduct ions efficiently through the lattice. However, the grain boundaries (GBs) can be orders of magnitude more resistive than their bulk counterparts. While methods developed for bulk-doping have been successful, they have provided little insight when optimizing the GB ionic conductivity in polycrystalline ceramic oxides such as CeO₂. In recent years, nanoscale compositional characterization of GB composition has revealed different nominal concentrations of solutes at the GBs which could result in orders of magnitude increase in GB ionic conductivity relative to the undoped samples. This study focuses on the impact which concentration plays in modulating the potential energy landscape for two potentially promising solutes, Ca and Ba, which may increase the ionic conductivity across the GB. Computational modeling is employed using density functional theory to optimize the GB interfacial structure for one GB misorientation in CeO₂. This study further develops our understanding of high solute GB composition enabling the development of methods such as selective doping to improve macroscopic ionic conductivity for both the grain and GB.

To understand the key factors which influence GB properties, spin-polarized density functional theory calculations were performed using VASP[1] within the generalized gradient approximation (GGA)[2,3] with the Perdew-Burke-Ernzerhof exchange correlation functional. The strong correlation effects were treated with the Hubbard U correction (GGA+U) formulated by Dudarev et. al.[4]. The GB structures are doped with a local GB solute concentration of 0\%, 20\%, and 40\%. The local bond strain, coordination, electronic structure, and thermodynamic energies are computed for each distinct substitutional solute location at the GB core. The oxygen vacancy binding energy along a few select [0,0,1] paths are computed to assess how changes in the local GB composition impacts the potential energy landscape for oxygen migration. This study further develops our understanding of the interplay between nanoscale GB composition and structure with high solute concentrations enabling the development of methods such as selective doping to improve macroscopic ionic conductivity for both the grain and GB.

References
1. G. Kresse, J. H. Ab initio molecular dynamics for liquid metals.Phys. Rev. B1993,47, 558.
2. J. P. Perdew, M. E., K. BurkePhys. Rev. Lett.1996,77, 3865.
3. J. P. Perdew, M. E., K. BurkePhys. Rev. Lett.1997,78, 1396.
4. S. L. Dudarev, S. Y. S. C. J. H. A. P. S., G. A. BottonPhys. Rev. B1998,57, 1505.
5. We gratefully acknowledge ASU’s HPC staff for support and assistance with computing resources along with the Extreme Science and Engineering Discovery Environment (XSEDE). We also acknowledge the National Science Foundation grant DMR-1308085 for funding.

Pyrochlore oxides (A₂B₂O₇) – and other fluorite-derived complex oxides – are interesting for not only their wide variety of desirable properties, such as ionic conduction and radiation tolerance, but also for their complex defect formation and disordering mechanisms that often give rise to their properties. We have used neutron total scattering to study in detail all structural aspects of the disordering of pyrochlore which involves randomization of both cation and anion sublattices. By means of diffraction and pair distribution function analysis of a Ho₂Ti_2-xZr_xO₇ (x = 0.0-2.0) solid solution series, the disordering mechanism was studied simultaneously over multiple material length scales with novel insight into the local atomic arrangements. With increasing Zr-content, the series exhibits an order-disorder transformation from pyrochlore (space group: Fd-3m) to defective fluorite (space group: Fm-3m) across a narrow compositional range (x = 1.0-1.5) over the long length-scale, while the local atomic arrangement changes gradually to a weberite-type structure (space group: C222₁) over the whole compositional range. This distinct disordering scheme can be explained by the movement of a 48f oxygen to a vacant 8a site, creating 7-coordinated Zr⁴⁺ sites that produce local weberite-type building blocks. These building blocks accumulate until a critical Zr-content (x ~ 1.2) is reached which triggers the rearrangement of weberite-type units into long-range structural defective fluorite.

Defects are known to play an important role in several thermophysical properties of actinide oxides, from ionic self-diffusion to heat transport.
Point defects are also related to high-temperature superionic transitions observed with many single actinide oxides and mixed oxides (MOX).

Recent advances in potential development have enabled accurate molecular dynamics (MD) simulations of a broad range of actinide oxide compositions, and have been shown to reproduce defects properties well.
This has been used in this work to investigate the effect of defects populations on thermophysical properties such as elastic constants and vibrational properties of several mixed oxides containing U, Pu, and Th.
Both the low-temperature fluorite structure and the high-temperature superionic phase have been considered.
These results are important to improve our understanding of the behaviour and high-temperature thermodynamics of actinide oxides beyond the superionic transition.

Transparent Conducting Oxides are fundamental components of modern optoelectronic devices, bringing together the unusual combination of optical transparency and electrical conductivity. Presently, the champion material is Sn-doped In₂O₃ (ITO), capable of reaching resistivity as low as 10^-5 Ω cm and mobility around 40 cm² V^-1 s^-1. However, it has emerged that the use of alternative dopants returns greater mobility on the order of 100 cm² V^-1 s^-1, and in turn can match or even surpass the levels of conductivity reached by ITO.[1] Higher mobility affords a lower carrier concentration, allowing for improvements in NIR transparency due to a lower resonant plasmon frequency, ultimately leading to increased efficiencies in solar cells.[2]

Here Ce-doped In₂O₃ (ICO) films grown by rapid plasma deposition demonstrate electron mobility upward of 130 cm² V^-1 s^-1.[3] Hybrid DFT calculations show that Ce 4f states lie well above the conduction band minimum, thereby not affecting its dispersion. A low electron effective mass is maintained at the band edge and therefore high electron mobility is achieved, a feature shared with other successful resonant donors in In₂O₃.

Defect formation energies place the transition level of the Ce donor from 1+ (active) to neutral (inactive) around 0.1 eV above the CBM – this suggests that only a small shift in the Fermi level will cause Ce (III) to begin to occupy In sites rather than Ce (IV). This is observed in Hard X-ray Photoelectron Spectroscopy, where Ce (III) states begin to emerge as a function of increasing carrier concentration. This potentially places an upper limit on Ce-doping of In₂O₃ but does not prevent films achieving equal levels of conductivity and surpassing electron mobility of commercial ITO films, with feasible applications in concentrated multi-layer solar cells and beyond.

References:
[1] Swallow, J. E. N. et al, Mater. Horiz, 2019, DOI: 10.1039/C9MH01014A
[2] Aydin, E. et al, Adv. Funct. Mater., 2019, 29, 1901741 DOI: 10.1002/adfm.201901741
[3] Koida, T. et al, Phys. Status Solidi A, 2018, 215, 1700506, DOI: 10.1002/pssa.201700506

High-temperature thermal-barrier coatings are an essential component of modern gas-turbine engines. Despite remaining the standard industrial thermal barrier coating, yttria-stabilised zirconia (YSZ) top-coat suffers profound performance issues at very high-temperatures, principally resulting from high-temperature phase instability limiting the maximum operating temperature.¹ As a result, alternative materials with superior insulation properties and stability are essential for higher operating temperatures and improve engine efficiencies. To this end, novel ceramic materials, rare-earth zirconates pyrochlores including La₂Zr₂O₇, have been looked to due to their low lattice thermal conductivity, phase stability across a wide temperature range, high melting point and low thermal expansion coefficient.²

Recent work by Voneshen et al. ascribed the exceptionally low thermal conductivity in La₂Zr₂O₇ to a strongly anharmonic phonon mode associated with the kagome planes in the material, observed from inelastic neutron scattering (INS) data.³ In this work we calculate the thermal conductivity and phonon lifetimes of La₂Zr₂O₇ using ab initio lattice dynamics calculations using both the linearised Boltzmann transport theory (LBTE) within the single mode relaxation time approximation (RTA) and the temperature dependent effective potential (TDEP) methods.^4,5 These results are supported by new experimental data measuring the phonon-phonon scattering rates of the individual modes in La₂Zr₂O₇.


1. Witz, G.; Shklover, V.; Steurer, W. J. Am. Ceram. Soc. 2008, 90, 2935-2940
2. Guo, X.; Li, L,; Park, H-M.; Knapp, J.; Jung, Y-G.; Zhang, J. J. Therm. Spray. Tech. 2018, 27, 581-590
3. Voneshen, D. J.; Hatnean, M. C.; Perring, T. G.; Walker, H. C.; Refson, K.; Balakrishnan, G.; Goff, J. P. 2019, arXiv:1809.06265
4. Togo, A.; Chaput, L.; Tanaka, I. Phys. Rev. B 2015, 91, 094306
5. Hellman, O.; Steneteg, P.; Abrikosov, I. A.; Simak, S. I. Phys. Rev. B 2013, 87, 104111

Radiation stability is often a key limiting factor in performance of fluorite-structured materials and determining their suitability for use in energy-related applications. In an effort to mitigate the effects of radiation, nanostructured materials are of interest as they incorporate high defect sink strengths [Rose et al., Nanostructured Materials (1995), Nita et al., Journal of Nuclear Materials (2004)]. Recently, it has been shown that the response of CeO₂, ThO₂, and UO₃ to highly ionizing radiation is strongly dependent on the material’s redox response [Tracy et al., Nature Communications (2015)]. When exposed to swift heavy ions, cations in the material are subject to changes in valence which drives swelling and microstrain as irradiation-induced defects accumulate. In this work, we present new insights into how crystallite size affects irradiation-induced redox response and defect accumulation in fluorite-structured simple oxides. Using 946 MeV Au ions at the UNILAC accelerator of the GSI Helmholtz Center, we irradiated microcrystalline and nanocrystalline materials of different compositions containing cations known to reduce (CeO₂), remain univalent (ThO₂), and oxidize (UO₂) under ionizing conditions. Irradiated samples were characterized by synchrotron X-ray diffraction/absorption, neutron total scattering with pair distribution function (PDF) analysis, transmission electron microscopy, and Raman spectroscopy. Each composition exhibits a distinct response between microcrystalline and nanocrystalline forms, such as magnitude of volumetric swelling and secondary phase formation, driven mainly by redox processes. PDF analysis reveals small peroxide-like defects in CeO₂ and mono- and di-interstitial clusters in UO₂. Our findings imply that nanocrystallinity has negative effects on a material’s response to highly ionizing radiation. These results shed more light onto the interplay of particle size and cation redox behavior and their effect on defect production in an important class of materials, an insight that is essential in developing advanced materials for energy-related applications.

Reducible oxides can be employed in solid oxide fuel cell anodes and cathodes largely because their redox properties enable fuel oxidation and oxygen exchange [1]. These processes involve creation and annihilation of oxygen vacancies at the surface and transport through the bulk. Here, we show that electron microscopy can be employed to identify the most active sites for vacancy creation/annihilation on the surface of CeO₂ (ceria). Under reducing conditions, ceria releases oxygen through the formation of an oxygen vacancy coupled with a cation transition from Ce⁴⁺ to Ce³⁺. Moreover, a lattice expansion (~0.2-0.3 Å) occurs once oxygen vacancies are created on CeO₂ surfaces [2]. The relative ease with which oxygen vacancies are created/annihilated on CeO₂ is strongly dependent on particle surface structure, and theoretical calculations have predicted significant variations in activation energies for lattice oxygen removal at different surface sites, which is likely to be associated with orders of magnitude differences in local oxygen exchange rates [3]. We are interested in probing and correlating local exchange activity with atomic surface structure and have used time-resolved aberration-corrected transmission electron microscopy (AC-TEM) to observe and analyze dynamic atomic-level cation displacements associated with oxygen vacancy creation and annihilation.

A FEI Titan AC-TEM equipped with a Gatan K2 IS direct detection camera (with high detective quantum efficiency) was used to image CeO₂ nanocubes at 40 frames/second and 5000 e^-/(Ųs). Ce atomic column displacements were analyzed with picometer precision over time-resolved image sequences using MATLAB codes. By monitoring the frequency of cation displacements at a surface, a local indicator of activation energy for oxygen vacancy creation was obtained. A considerable degree of diversity and heterogeneity in the type of surface sites that show high activity was observed, with low coordination number sites such as steps and edges, as well as locally strained sites, exhibiting the highest displacement frequency and therefore enhanced oxygen exchange activity. Our approach directly links atomic surface structure to vacancy creation, indicating that the ability to quantitatively characterize nanoparticle surfaces to locate active sites can contribute to understanding and tailoring their exchange properties.
Acknowledgements
We acknowledge support of NSF grant DMR-1308085, the use of ASU’s John M. Cowley Center for High Resolution Electron Microscopy and use of the K2 IS camera courtesy of Gatan.

References
[1] Gorte, R.J., AIChE Journal, 56 (2010), p. 1126-1135.
[2] Muhich, C.L., The Journal of Physical Chemistry C, 121 (2017), p. 8052-8059.
[3] Migani, A., et al, Journal of Materials Chemistry, 20 (2010), p. 10535-10546.

Spent nuclear fuel (SNF) of light water reactors (LWR) is constituted of a matrix of unfissioned UO₂ containing a small fraction of fission products (FP) and minor actinides (MA). These MA are mainly α-emitters with very long half-lives: therefore, SNF will keep α-self-irradiating for millennia after discharge from the reactor, resulting in widespread changes of the thermophysical properties of the material. In order to properly coordinate and license any disposal strategy for SNF, the long-term effect of α-self-irradiation has to be known and anticipated.
Due to the lack of real old SNF, a proxy system to study the effect of α-irradiation on the SNF matrix is provided by the synthesis of UO₂ doped with short-lived α-emitters. In this way, a significant amount of radiation damage can be stockpiled over a laboratory timescale in a simpler surrogate system that does not combine radiation damage with chemical or density gradients (built-in in real SNF).
In the present work, UO₂ doped with ²³⁸Pu was produced to study the effect of α-self-irradiation on the crystalline disorder and lattice swelling. The composition of the samples was carefully chosen, based on the dopant specific activity, in order to reach saturation of the lattice parameter swelling within the 3 years timespan of the project. Samples were periodically characterized by means of XRD and Raman spectroscopy up to 0.4 dpa, equivalent to a spent nuclear fuel with a 40 GWd/t UO₂ burnup stored during 300 years, or representative of a 65 GWd/t spent MOX fuel (45 % Pu) after 25 years of storage.
Lattice swelling as a function of dpa was assessed with very good accuracy and benchmarked against literature data: saturation was reached at a value of 0.3 % around 0.4 dpa. For the first time, microstrain was also monitored in (U,Pu)O₂ as a function of self-irradiation.
Periodic Raman spectroscopy acquisitions on (U,Pu)O₂ as a function of the dose represent an innovative probe, and they showed fast and progressive degradation of the structural order up to 0.1 dpa, and a slow but persistent increase up to 0.4.
SEM characterizations were additionally performed and highlighted that no loss of structural integrity is associated with this microstructural evolution, at least within the measured dpa range.

Fluorite-based nanostructured oxides are promising materials for wide-ranging applications. In these applications, due to the reduced dimensions, surfaces and interfaces play a critical role in influencing novel functionalities. In nanostructured oxides, defects are ubiquitous at surfaces, grain boundaries, and heterointerfaces. The interaction of defects with such structural anomalies are often found to dictate the atomic-scale structure and correlated properties of oxides. Based on atomistic simulations, we elucidate the influence of defects on structure and properties in three different cases – (i) Misfit dislocations in CeO₂/MgO heterostructures: We find that dopant–defect complexes at misfit dislocations have extended stability revealing that they would influence ionic transport at heterointerfaces of fluorite-structured thin film electrolytes.(ii) Grain boundaries in nanocrystalline doped CeO₂: We find that segregation energies, availability of favorable sites, and overall stronger binding of dopant–defect complexes would influence ionic transport across grain boundaries in nanocrystalline doped ceria. (iii) Steps and trenches at low-index surfaces of pyrochlores oxides: We find that one fundamental mechanism to eliminate the surface dipole in pyrochlores is the formation of structural defects such as steps and trenches, which explain the enhanced reactivity and extraordinary surface-driven properties of pyrochlores. Overall, our results shed light on the complex interplay between defects and structural anomalies, and assist in disentangling the multifaceted role of defects in fluorite-based nanostructured oxides.

The surface strain of oxide nanoparticles plays a crucial role in tuning a materials’ properties [1]. For example, surface strain can regulate and control surface diffusion processes and can change the ability of the surface atoms to bond with adsorbates and thus modify the reactivity of a material [2]. In reducible oxides such as CeO₂, the degree of strain can be tuned by different means such as particle size, shape, non-stoichiometry, structural deformations (e.g. oxygen vacancies) etc. A critical step in understanding the properties of nanoparticle surfaces is to measure the surface strain on different crystal facets and in the vicinity of surface defects. We have developed a method to determine the surface strain in oxide nanoparticles with atomic resolution using high resolution transmission electron microscopy (HRTEM). In a typical TEM image, the signal from the heavier Ce atomic columns is much stronger than the signal from the lighter oxygen columns. Consequently, more precise measurements can be made on the cation sublattice in order to map the strain on or near the nanoparticle surface.
CeO₂ nanoparticles were synthesized by the hydrothermal method [4] and imaged using negative Cs imaging in a FEI Titan AC-ETEM with a single-electron-detection K2 camera operated in the counting mode. The K2 camera allowed high quality electron imaging to be performed. (111), (110), and (100) CeO₂ nanoparticle surfaces were imaged in a [110] projection at 5000 e^-Å^-2s^-1 with Ce and O atomic columns visible at the surface. Custom written MATLAB codes are used to identify and determine the positions of atomic columns. Atomic resolution strain maps were created to visualize cation sublattice deformations at different locations in the nanoparticle. The bulk of the nanoparticle is relatively strain free, but the surfaces show varying degrees of compressive and tensile strain along different crystallographic directions of the nanoparticle. The surface strain field varies in a complicated fashion with the highest tensile and compressive strains along different directions occurring at different points on the surface. The highest degree of strain is associated with defects such as step sites. These local strain fields on CeO₂ nanoparticle surfaces are correlated with local activity for oxygen vacancy creation and annihilation.
References:
[1] Madsen, Jacob, et al. "Accuracy of surface strain measurements from transmission electron microscopy images of nanoparticles." Advanced structural and chemical imaging 3.1 (2017): 14.
[2] Mavrikakis, Manos, Bjørk Hammer, and Jens Kehlet Nørskov. "Effect of strain on the reactivity of metal surfaces." Physical Review Letters 81.13 (1998): 2819.
[3] Gopal, Chirranjeevi Balaji, et al. "Equilibrium oxygen storage capacity of ultrathin CeO_2-δ depends non-monotonically on large biaxial strain." Nature communications 8 (2017): 15360.
[4] Yang, Zhiqiang, et al. "Single-crystalline ceria nanocubes: size-controlled synthesis, characterization and redox property." Nanotechnology 18.18 (2007): 185606.
[5] We gratefully acknowledge support of NSF grant DMR-1308085, the use of ASU’s John M. Cowley Center for High Resolution Electron Microscopy and use of the K2 IS camera courtesy of Gatan.

Yttria partially stabilized zirconia (ZrO₂)_x(Y₂O₃)_½-x and amorphous zirconia have been investigated to understand accommodation of excess oxygen into its structure. Yttria stabilized zirconia was investigated experimentally and through atomic scale simulation methods. A new Raman peak was observed after treatment at 840 cm⁻¹, consistent with previous reports of solid state peroxide ions (O₂²⁻). This was corroborated using atomic scale simulation based on density functional theory; these also highlighted the near-zero solution enthalpy for excess oxygen in the monoclinic structure via the formation of a peroxide ion defect.

Amorphous zirconia (a-ZrO2) was simulated using a combination of state-of-the-art atomic scale methods. This combination has enabled the complex chemistry of the amorphous system to be efficiently investigated. Notably, the a-ZrO₂ system was observed to accommodate excess oxygen readily – through the formation of neutral peroxide (O₂²⁻) defects - the same defect observed in the yttria stabilized zirconia system. This has implications not only in the a-ZrO₂ system, but also in other systems employing network formers, intermediates and modifiers.

Oxygen transfer is a critical functionality in many technologies involved in automotive exhaust control and clean energy conversion. In these applications, catalytically active metal nanoparticles (e.g., Pt) are typically dispersed on reducible oxides (e.g., fluorite CeO₂), since reducible oxides can transfer their lattice oxygen to reactive adsorbates at the metal-support interface, the so-called three-phase boundary [1,2]. Attaining an atomic-level understanding of the catalytically-driven oxygen transfer process is of great interest to the scientific community and is required to design more efficient catalysts. This work employs aberration-corrected environmental transmission electron microscopy (ETEM) to investigate the atomic-scale structural dynamics that occur at the Pt/CeO₂ three-phase boundary under reaction conditions (i.e., in situ) and during catalysis (i.e., operando).

Nanostructured CeO₂ cubes and rods were synthesized, loaded with 17 wt.% Pt nanoparticles (NPs), and used as model systems. An ISRI RIG-150 microreactor coupled to a Varian 450 gas chromatograph was used to characterize the CO oxidation performance of the catalysts. An image-corrected FEI Titan ETEM was used to visualize the atomic structures that form under reaction conditions. Modified specimen preparation techniques and in situ electron energy-loss spectroscopy (EELS) was implemented to track the gas composition during catalysis [3,4]. In this way it became possible to correlate active site dynamics with catalytic turnover frequency (TOF) for CO oxidation. It was observed that as the TOF increases, frequent oxygen transfer disrupts the bonds that anchor the ~2 nm Pt nanoparticles, leading to dynamic reconfigurations of the entire nanoparticles and their translational motion ~2-5 Å across the support. Interestingly, proximal Ce surface structures up to ~1 nm away from the Pt/CeO₂ interface undergo concurrent dynamics, which is consistent with a Mars-van Krevelen mediated interfacial oxygen transfer process.

An additional investigation was done by coupling a high frame-rate direct electron detector to the ETEM, which enabled the transient dynamic behavior to be observed at higher time resolution (25 milliseconds). During this experiment a Pt NP on the (111) surface of a CeO₂ rod was observed to restructure dynamically in a 10 mTorr atmosphere of CO and O₂, although the same particle was stable in 0.5 mTorr of inert N₂. Interestingly, an examination of the time- series in CO and O₂ shows the Pt NP undergoes a sequence of structural reconfigurations that correlate with dynamics at interfacial Ce sites. This paper will present results from both experiments [5].

[1] Cargnello, M., et al., Science 2013, 341, p. 771
[2] Vayssilov, G. N., et al., Nature Materials 2011, 10(310).
[3] Miller, B., Crozier, P., Microsc. Microanal. 2014, 20(815).
[4] Miller, B. K., Barker, T. M., Crozier, P. A. Ultramicroscopy 2015, 156, 18.
[5] We gratefully acknowledge support of NSF grant CBET-1604971 and the use of facilities at Arizona State University’s John M. Cowley Center for High Resolution Electron Microscopy.

Systems with strong local correlations but looser long-range cationic order are a challenging curiosity both from an academic and technological perspective. Describing structural features at different length scales is crucial for understanding the properties these systems can exhibit. Metal-oxide frameworks like many fluorite-derived structures seem capable to produce flexible long-range structures, allowing more freedom for tuning properties than most strictly periodic systems. Diffuse scattering (eventually condensing as weak superlattice spots) in such materials has been often discounted as intricate and burdensome but it is attracting greater attention recently, with advances in techniques and more flexible model paradigms. This experimental information provides invaluable feedback for modelling. We review some structural features encountered in pristine and irradiated oxygen-deficient fluorite-related structures that have promising potential for tackling the problems of frustration between nanoscale and long-range correlations.

Diffusion in UO₂ nuclear fuel is important for fuel performance, because it connects to key phenomena such as release of fission gas to the plenum, swelling and fuel creep. Fission gas retention and release also impact nuclear fuel performance in indirect ways by, for example, influencing the fuel thermal conductivity. Diffusion coefficients in fluorite UO₂, which here refer to U self-diffusion and fission gas diffusion, are among the properties with highest uncertainty, especially in irradiation environments. This directly translates to fission gas release, swelling and creep predictions. In this talk, we will highlight the development of new mechanistic models for U and Xe diffusion within the grains of UO₂, both under out-of-pile (no irradiation) and in-pile (irradiation) conditions. The models are applied to standard UO₂ as well as to UO₂ fuel doped with Cr. The point defect and Xe-vacancy cluster properties determining the defect concentrations and diffusion rates are calculated using a combination of density functional theory and empirical potential methods. With this data as input, the interaction of point defects under irradiation, leading to the formation of highly mobile Xe-vacancy clusters, is modeled using a methodology labeled free energy cluster dynamics (FECD). As suggested by the name, this method is based on traditional cluster dynamics but with an explicit connection to the free energy representing the driving force of defect reactions. In addition, this model accounts for the non-stoichiometry in UO_2±x, as governed by the O chemical potential, which strongly influences the diffusion rates by changing the the equilibrium concentration of U and O vacancies. In order to build confidence in the modeling approach, our predictions are first validated against experiments for U self-diffusion. Good agreement is found for intrinsic diffusion, but some discrepancies emerge under irradiation, which we speculate may be related to sample non-stoichiometry. Application of the methodology to simulations of Xe diffusion shows that at high temperature, also equivalent to intrinsic (no irradiation) conditions, transport is dominated by the Xe_U2O cluster (a Xe atom occupying a trap site formed by two U and one O vacancy) and at intermediate temperatures the Xe_U4O3 cluster takes over, followed by atomic mixing due to irradiation damage at the lowest temperatures. The predicted Xe diffusion rates as function of temperature and fission rate are in good agreement with experimental data. Next, this model is adapted to Cr-doped UO₂ by changing the O chemical potential and then used to predict the Xe diffusion rate. The Xe diffusivity increases compared to standard UO₂, both at high and intermediate temperatures, while there is no effect at low temperatures where atomic mixing dominates. Finally, the importance of interstitial U and O clusters will be discussed in the context of rapid diffusion mechanisms, recombination rates and dislocation loop formation.

Zinc-blende-related chalcogenide semiconductors CdTe and Cu(In,Ga)(S,Se)₂ (CIGS) currently represent the fastest growing commercial thin-film photovoltaic (PV) technologies. To address prospective scalability issues related to elemental scarcity (Te, In) and/or heavy-metal toxicity (Cd), Cu₂ZnSn(S,Se)₄ (CZTS) has also been vigorously pursued as a prospective drop-in replacement for CIGS, but efficiency improvement has been hindered by adverse defect characteristics (band tailing and deep defects), in part due to the similarity in chemistry among component metals and associated anti-site disordering. This talk will discuss promising emerging alternative multinary chalcogenides based on earth-abundant Cu₂BaSn(S,Se)₄ (CBTS) as an example, which offer similarities to CZTS in terms of electronic structure, but with introduced atomic size and coordination preference differences that reduce likelihood of atomic disordering—i.e., the much larger Ba ion, occupying a site that has 8-fold coordination rather than 4-fold (as for Cu, Zn and Sn in CZTS), reduces the probability of anti-site disordering and associated defects [1,2]. Simple solution- and vacuum-based film deposition processes enable fabrication of absorber layers with initial PV device sunlight-to-electricity power conversion efficiencies exceeding 5% for CBTS [3,4] and analogous photoelectrochemical (PEC) cells yielding a stable (over 10 hr) 12 mA/cm² photocurrent at 0 V/RHE [5]. This talk will discuss aspects of defect engineering within this family, as well thoughts on design strategies for expanding the known members of the I₂–II–IV–VI₄ family beyond CBTS (e.g., [6]). If desirable electronic structure tunability associated with a multi-element stoichiometry can be combined with earth-abundant components and control over defect formation, multinary chalcogenides may provide a bright path forward in the quest for high-performance, low-cost and scalable PV and PEC devices.

References:
[1] D. Shin, B. Saparov, T. Zhu, W. P. Huhn, V. Blum, D. B. Mitzi, Chem. Mater. 28, 4771 (2016).
[2] D. Shin, B. Saparov, D. B. Mitzi, Adv. Energy Mater. 7, 1602366 (2017).
[3] D. Shin, T. Zhu, X. Huang, O.Gunawan, V. Blum, D. B. Mitzi, Adv. Mater. 29, 1606945 (2017).
[4] B. Teymur, Y. Zhou, E. Ngaboyamahina, J. T. Glass, D. B. Mitzi, Chem. Mater. 30, 6116 (2018).
[5] Y. Zhou, D. Shin, E. Ngaboyamahina, Q. Han, C. Parker, D. B. Mitzi, J. T. Glass, ACS Energy Lett. 3, 177 (2018).
[6] T. Zhu, W. P. Huhn, G. C. Wessler, D. Shin, B. Saparov, D. B. Mitzi, V. Blum, Chem. Mater. 29, 7868 (2017).