
==== Front
Nat Commun
Nat Commun
Nature Communications
2041-1723
Nature Publishing Group UK London

39289372
52386
10.1038/s41467-024-52386-3
Article
in situ observation of reversible phase transitions in Gd-doped ceria driven by electron beam irradiation
http://orcid.org/0000-0002-9762-4586
Ran Ke ran@gfe.rwth-aachen.de

123
Zeng Fanlin 4
Jin Lei 2
http://orcid.org/0000-0002-7302-7103
Baumann Stefan 4
http://orcid.org/0000-0002-6398-2254
Meulenberg Wilhelm A. 45
Mayer Joachim 12
1 https://ror.org/04xfq0f34 grid.1957.a 0000 0001 0728 696X Central Facility for Electron Microscopy GFE, RWTH Aachen University, Aachen, Germany
2 grid.8385.6 0000 0001 2297 375X Ernst Ruska-Centre for Microscopy and Spectroscopy with Electrons ER-C, Forschungszentrum Jülich GmbH, Jülich, Germany
3 grid.461610.4 0000 0004 0450 8602 Advanced Microelectronic Center Aachen, AMO GmbH, Aachen, Germany
4 https://ror.org/02nv7yv05 grid.8385.6 0000 0001 2297 375X Institute of Energy and Climate Research IEK-1, Forschungszentrum Jülich GmbH, Jülich, Germany
5 https://ror.org/006hf6230 grid.6214.1 0000 0004 0399 8953 Faculty of Science and Technology, Inorganic Membranes, University of Twente, Enschede, AE The Netherlands
17 9 2024
17 9 2024
2024
15 815616 4 2024
4 9 2024
© The Author(s) 2024
2024
https://creativecommons.org/licenses/by/4.0/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
Ceria-based oxides are widely utilized in diverse energy-related applications, with attractive functionalities arising from a defective structure due to the formation of mobile oxygen vacancies (VO⋅⋅). Notwithstanding its significance, behaviors of the defective structure and VO⋅⋅ in response to external stimuli remain incompletely explored. Taking the Gd-doped ceria (Ce0.88Gd0.12O2-δ) as a model system and leveraging state-of-the-art transmission electron microscopy techniques, reversible phase transitions associated with massive VO⋅⋅ rearrangement are stimulated and visualized in situ with sub-Å resolution. Electron dose rate is identified as a pivotal factor in modulating the phase transition, and both the VO⋅⋅ concentration and the orientation of the newly formed phase can be altered via electron beam. Our results provide indispensable insights for understanding and refining the microscopic pathways of phase transition as well as defect engineering, and could be applied to other similar functional oxides.

Ceria-based oxides are used in diverse energy-related applications, with functionalities arising from a defective structure due to the formation of mobile oxygen vacancies. Here authors reveal sub-Ångström details of oxygen dynamics during the reversible phase transition of Gd-doped ceria (CGO).

Subject terms

Ceramics
Structural properties
Electronic devices
Fuel cells
https://doi.org/10.13039/501100001659 Deutsche Forschungsgemeinschaft (German Research Foundation) 387282673 Ran Ke https://doi.org/10.13039/501100002347 Bundesministerium für Bildung und Forschung (Federal Ministry of Education and Research) NEUROTEC, 16ME0399, and 16ME0398K Ran Ke issue-copyright-statement© Springer Nature Limited 2024
==== Body
pmcIntroduction

Owing to the flexile valence switching between Ce4+ and Ce3+ cations and the facile formation of oxygen vacancies (VO⋅⋅)1–4, ceria has been recognized as one of the best candidates for catalysts5–7 and solid electrolytes8,9. Ceria-based catalysts with stabilized and adequate active sites are attracting continuous attention10,11, and important advances have also been made towards the production of sustainable and clean energy12,13. Besides, memristors based on ceria are a critical component for next-generation nanoelectronics. The conductive species can be effectively controlled by external field to readily achieve an ON/OFF switching, which have been extensively explored for possible large-scale integrated circuits14–17. Regarding all these functional oxide materials and their associated devices, the formation and migration of VO⋅⋅ under external stimuli constitute the fundamental processes, which are often linked with modifications of the ceria structure and subsequent phase transitions. Typically, ceria crystallizes in a fluorite-type (F-type) cubic structure (space group Fm3¯m, a ≈ 5.42 Å), where the metal site (M) sitting at (0,0,0) is coordinated to eight O at (1/4,1/4,1/4)18. Transition from F-type to the so-called C-type (space group Ia3¯) takes place when enough VO⋅⋅ and MM′ are introduced to the system19–22. As a result, the cell parameter is doubled, the M site is six-coordinated to O, and the crystallographic positions are split23,24. The C-type structure is described as M1−x4+Mx3+O2−δ2−, where δ=x/2 is the concentration of VO⋅⋅ and Mx3+ is trivalent ions like Gd3+ and/or reduced Ce3+.

Benefiting from the cutting-edge transmission electron microscopy (TEM) techniques, significant insights into the transition process have been acquired, providing outstanding spatial, chemical and temporal resolution10,13,25,26. By in situ environmental TEM (ETEM), the redox process in ceria nanoparticles was studied at elevated temperature and in H2/O2/CO2 environment27–30. At ambient temperature, in situ electrical probe TEM was also able to drive reversible resistance switching by applying electrical field16,31. Nevertheless, considering the significance of phase transitions in the context of technological applications, the reported findings still lack elaborate investigations or adequate spatial/elementary resolution, primarily due to the challenges associated with real-time visualization of dynamic O atoms. Questions pertaining to transition details, including the selection of external stimuli, the behaviors of VO⋅⋅, the feasibility of fine tuning, and the underlying mechanism remain unsolved.

Herein, the Gd-doped ceria (CGO)32,33 with high oxygen ion conductivity and giant electrostriction under external electric field34,35 is chosen for our study (Ce0.88Gd0.12O2-δ, in Supplementary Note 1 and Fig. 1). Utilizing the incident electron beam (e-beam) as an external stimulus, the phase transition of CGO is in situ probed down to sub-Å scale by TEM36–39. Varying electron dose rate (EDR), the transitions can be accelerated, retarded, hold, and reversed. Negative spherical aberration imaging (NCSI)40–42 and integrated differential phase contrast (iDPC)43 imaging are employed, enabling high contrast for both light oxygen and heavy metal atoms, as well as facilitating the measurement of atomic positions with ultrahigh precision. Quantifying lattice distortions also allows a direct estimation of the local VO⋅⋅ concentration. Together with the proposed mechanism, our findings are crucial for the engineering and optimization of numerous energy-related applications relying on ceria and analogous oxides.Fig. 1 The reversible transition between F- and C-type.

a Schematic of the reversible transition: in TEM mode, e-beam with high and low EDR is used to stimulate the F-to-C and C-to-F transition, respectively. F- and C-type CGO30 models are viewed along < 001 > at the bottom of (a). Single unit cells of each type are outlined by the solid squares. b–c Enlarged models from the dashed rectangles in a. The three sets of parallel lines in (c) indicate split metal positions. Below the models are the simulated HRTEM images. Laterally averaged intensity profile and the estimated dO_v (virtually extended as the dotted lines) are plotted on the left and right side of the simulated images.

Results and Discussion

Reversible transition between F- and C-type CGO

Figure 1a sketches the experimental design for in situ transition study. In TEM mode, e-beam with high EDR is able to stimulate the F-to-C transition, while the C-to-F reverse transition is assisted with a low EDR. Both F- and C-type models are viewed along < 001 > in Fig. 1a, and having 30% of the M sites occupied by trivalent ions (CGO30). As mentioned earlier, comparing with F-type, the cell parameter of C-type is doubled (the solid squares in Fig. 1a). Furthermore, part of the M positions in the C-type (the pairs of lines in Fig. 1c) and all its O positions are splitting evidently. Based on the CGO30 models, high-resolution TEM (HRTEM) images are simulated44, in Figs. 1b–c. As plotted on the left and right side in Fig. 1b, identical peak intensities from the M positions and constant dO_v (the distance between two neighboring O along the vertical direction) are revealed from the F-type. In contrast, the split M positions in Fig. 1c leads to a higher intensity peak than the unsplit ones, and the measured dO_v varies in a breathing-like manner.

Similar features are observed by experiment as well. Figure 2a–c shows the iDPC results from the F-type along <001 > . As a F-to-C transition is usually unavoidable under the HRTEM imaging condition, iDPC technique is therefore employed for safely imaging the F-type structure (Supplementary Note 2 and Fig. 2). In Fig. 2a, both the M and O sites can be clearly resolved, showing a good agreement with the embedded F-type model and the simulated iDPC image44. Moreover, on the left side of Fig. 2a, the laterally averaged intensity profile shows constant M peaks, similar as in Fig. 1b. Each M and O position in Fig. 2a are then determined using two-dimensional Gaussian fitting45. In Figs. 2b, c, the mapped distances between neighboring M and O positions (dM and dO) are overlaid on the iDPC image. On the right side of Fig. 2c is the laterally averaged dM_v and dO_v as defined in Fig. 2a, where both values are almost constant. Figure 2d–f are the corresponding HRTEM results from the C-type. Consistent with the simulation in Fig. 1c, alternating M peaks are noticed in the intensity profile on the left side of Fig. 2d. The mapped dM_v in Fig. 2e are still constant, while the dO_v in Fig. 2f are regularly oscillating.Fig. 2 Experimental visualization of both F- and C-type.

a iDPC image of the F-type along < 001 > , together with a F-type model and the simulated iDPC image. On the left is the laterally averaged intensity profile. b–c The mapped distances between neighboring M and O positions (dM and dO) based on (a). On the right side of (c) are the laterally averaged dM_v and dO_v. d–f The corresponding HRTEM results from the C-type. g Diffraction patterns from a transition cycle. Circles are placed at the same position for all the patterns. h–i FFT patterns from (a and d).

The additional ordering in the C-type structure also causes extra spots in the diffraction pattern, as those in Fig. 2g, which were recorded during a transition cycle. Marked by the circles, extra spots only show up for the C-type, and are indexed as {010} following the F-type. Figures 2h and 2i are the FFT patterns from Figs. 2a, d, respectively. In agreement with the diffraction patterns, only C-type is associated with extra spots. It’s also noticed in Fig. 2i that there are only two {010} spots along the vertical direction, while in Fig. 2g the {010} spots are located along both the vertical and lateral direction. This difference can be explained by the two perpendicular orientations of the C-type, which will be discussed later.

Modulating the phase transition

As shown in Figs. 2g, i, the extra spots in either diffraction or FFT patterns enable a direct detection of the C-type during the phase transition. Given that the transition is a rather fast process, yet acquiring diffraction patterns is relatively time consuming, the FFT pattern based on TEM image acquired with short exposure time but sufficient spatial resolution is thus employed to in situ study the phase transition. Two cases are shown in Figs. 3a, b (F-to-C) and Figs. 3c, d (C-to-F). In Figs. 3a, e-beam with high EDR (2656 e · Å−2 · s-1, and noted as 0.69 relative to the highest EDR 3825 e · Å−2 · s-1 listed in Fig. 3, to which all the EDRs are normalized, Supplementary Table 1) is used for inducing the phase transition and imaging. A series of TEM images were recorded (Supplementary Fig. 3), and the corresponding FFT patterns at different time are listed in Fig. 3b. Weak {010} spots can already be recognized at ~10 s. As the irradiation continues, the {010} spots get continuously stronger. A much lower EDR (noted as 0.06) is used in Fig. 3c for the C-to-F reverse transition. Due to both the weak beam and low magnification, the FFT in Fig. 3d is rather noisy, when comparing with Fig. 3b. Nevertheless, the {010} spots are still visible at 0 s and start to fade away after ~26 s under continuous low EDR illumination, as revealed by the dashed circles. At 78 s, the extra spots are no longer detectable suggesting the recovery of the F-type structure.Fig. 3 Transitions with different EDRs.

a, b One F-to-C transition with EDR 0.69 (relative value, where the highest EDR 3825 e · Å-2 · s-1 listed in Fig. 3 is noted as EDR 1.0). A series of 40 TEM images were recorded with 0.04 s exposure time and 0.8 s interval. The image at the outset and the FFT patterns (absolute value) during the transition are shown. c–d One C-to-F transition with EDR 0.06. A series of 40 TEM images were recorded with 0.04 s exposure time and 2 s interval. The image at the outset and the FFT patterns (absolute value) during the transition are shown. e–g r (intensity ratio between the four {010} and the four {020} spots), rA, and rB (intensity ratio between the two {010} and the two {020} spots along A and B) as a function of time from each transition. Linear fitting is applied to r, and the estimated slopes are listed in (h).

Under the thin sample condition, the intensity of {010} spots are directly associated with the amount of C-type within the irradiated region. The intensity ratio between the four {010} and the four {020} spots, denoted r, can thus be used to evaluate the transition process. Several image time series were acquired with various EDRs. Figure 3e plots the r as a function of time for all of the F-to-C transitions. Each curve in Fig. 3e is divided into several regions for linear fitting (the solid lines), and the fitted slopes are listed in Fig. 3h, shedding light on the efficiency of each transition. Overall, the higher the EDR, the faster the F-to-C transition. Continuous irradiation can speed up the transition, as the slopes are noticed increasing with time for each EDR. One exception is EDR 1.0, where the fitted slope drops from 0.013 to 0.003 after ~14 s irradiation. Moreover, the intensity ratio rA and rB between the two {010} and the two corresponding {020} spots along the indicated A and B directions in Fig. 3b are also plotted in Fig. 3f, to study the anisotropy of the phase transition. Both rA and rB are growing at a similar pace at the beginning for all the transitions. After a certain point depending on the EDR, rA prevails, suggesting that the C-type structure favors orienting with its b axis (as defined in Fig. 1a) along A direction in the present case (Supplementary Note 3 and Fig. 4). Besides, the apparent deviations between rA and rB are found to be closely correlated with the change of slopes in Fig. 3e. For instance, an evident split of rA and rB is observed at ~14 s for EDR 1.0 in Fig. 3f, which coincides with the abrupt drop of the slope from 0.013 to 0.003 as shown in Fig. 3h. In contrast, the C-to-F reverse transitions are slower and show less dependence on the EDRs, as shown in Fig. 3g. The relatively large values of r are mainly resulting from the high noise level in the FFT patterns associated with low EDRs. The fitted slopes are plotted in Fig. 3h as well, where the values are almost identical for all the three transitions.Fig. 4 Pushing the VO⋅⋅ ordering.

A series of 50 HRTEM images was recorded with 0.5 s interval. Two of them are shown in (a). b The determined dO_v and dM_v from each labeled layer, as defined in (a). c AO_v and AM_v as a function of time, based on the HRTEM image time series. d AO_v and AM_v estimated from the simulated HRTEM images. The image simulations are based on C-type models with varying δ from other studies21,23,46,47. Solid lines suggest linear fits.

Modifying the C-type

After its formation, further modifying the C-type structure is also feasible. As already shown in Fig. 2f, the breathing-like oscillation of dO_v is a robust signature of the C-type, and the amplitude of the oscillation is reported to be dependent on the VO⋅⋅ concentration,δ 21,23,46,47. Thus, via monitoring the dO_v oscillation, δ variation under the e-beam can be revealed.

Figure 4a shows two images from an HRTEM image time series. The defined dO_v and dM_v are averaged within each labeled layer and plotted in Fig. 4b for both images. The dM_v is rather constant among the labeled layers and also close between the two images. The dO_v is oscillating as expected. However, the oscillation at 17 s is obviously broader than that at 4.5 s. In order to track the oscillation in a quantitative way, AO_v=△dO_v/dO_v¯, is calculated from each image in the series, where △dO_v and dO_v¯ are the deviation and mean value of all the measured dO_v from a single image. On the right of Fig. 4b,△dO_v (the double-headed arrows) and dO_v¯ (the dotted lines) determined from Fig. 4a are indicated. Similarly, AM_v for the M positions are determined as well. Both AO_v and AM_v are plotted as a function of time in Fig. 4c. Evidently, AO_v is increasing with prolongated irradiation, while AM_v shows ignorable changes with time.

The results in Fig. 4c is further compared with the C-type models with varying δ from other studies21,23,46,47. In order to enable a direct comparison, HRTEM images are first simulated based on these models (Supplementary Note 4 and Fig. 5). From these simulated images, AO_v and AM_v are estimated as well, and plotted as a function of δ in Fig. 4d. Similar to Fig. 4c, AM_v keeps constant within the whole range in Fig. 4d, while AO_v increases almost monotonously as a function of δ. Correlating the values of AO_v between Figs. 4c and 4d allows a direct determination of the δ within the irradiated region. Explicitly, the local δ can be driven up to 0.44 by e-beam in the present case.Fig. 5 Rotating the C-type structure.

a, b HRTEM images of the C-type structure before and after e-beam irradiation. c, d The same as in (a, b) with the brighter metal layers outlined by yellow/vertical and blue/parallel lines. e, f The enlarged images from the squared regions in (c, d). At the top is the vertically averaged intensity profiles, and atomic layers are labeled at the bottom. g–h The mapped distances between neighboring O positions, based on (e, f). Two regions (#1 and #2) are defined by the lines in magenta. From region #1, vertically averaged dO_l are plotted at the top. From region #2, vertically averaged dO_l and laterally averaged dO_v are plotted at the bottom and on the right, respectively.

In Fig. 3f, the C-type structure is observed orienting along two perpendicular directions, but not equally developed as a function of time (see divergence of rA and rB). The competition between the two orientations is also revealed by HRTEM image shown in Figs. 5a, b, where a massive rotation of 90° is manifest. As already shown in Fig. 2d, along the vertical direction, intensity of the neighboring M layers is alternating. Thus, the brighter M layers in Figs. 5a, b are marked by the yellow/vertical and blue/lateral lines in Figs. 5c, d, to highlight the two orientations. While the yellow lines are covering the whole image in Fig. 5c, up to half of Fig. 5d is occupied by the blue lines. The squared regions in Figs. 5c, d are further enlarged in Figs. 5e, f, with the intensity profiles plotted at the top. Consistent with the outlines in Fig. 5c, the M peaks are regularly oscillating in Fig. 5e. In contrast, such an oscillation breaks abruptly in Fig. 5f starting from Layer 12. Besides, distances between neighboring O positions, dO, are mapped in Fig. 5g-h. Two regions, #1 and #2, are defined by the lines in magenta. Within region #1, the vertically averaged dO_l (the distance between neighboring O positions along the lateral direction) as plotted at the top suggest similar oscillations in both images, only the gap is increasing from ~21 pm in Fig. 5g to ~31 pm in Fig. 5h. For region #2, the vertically averaged dO_l at the bottom shows an oscillation with a gap of ~25 pm in Fig. 5g but a rather flat line in Fig. 5h. On the contrary, a flat line is obtained on the right of Fig. 5g based on the laterally averaged dO_v, while an oscillation with a gap of ~31 pm shows up in Fig. 5h.

Chemical structure of the CGO

Figure 6a shows the iDPC image of the F-type CGO along <110> direction, where the M and its nearest O positions distanced by ~135 pm can be clearly resolved. The experimental image is also in a good agreement with the overlaid F-type model and the simulated iDPC image. Electron energy loss spectroscopy spectrum imaging (EELS SI) with atomic resolution was recorded, in Fig. 6b, suggesting a successful Gd substitution of Ce and a uniform distribution of O. Due to the reducing sample thickness from left to right, as shown by the t/λ profile at the lower-left corner in Fig. 6b (t: sample thickness and λ: electron inelastic mean free path), dropping intensity is noticed in both the ADF image and elemental maps. The laterally averaged intensity profile of Ce, Gd, and O are plotted on the right side of Fig. 6b, where Ce and Gd peaks show up simultaneously (the dotted line) while the O peak are between the Gd/Ce peaks (the solid line). Figure 6c further plots the fine structure of Ce M4,5 edge from three different locations (#1-#3, Supplementary Fig. 6). Among the three curves, different features are noticed, including the slight shift of the shoulders (the dashed lines), and the varying height difference between the M5 and M4 edges (the bars with round ends). Based on the M5/M4 intensity ratio in second derivative spectra, the Ce3+ ratios (Ce3+/(Ce3++Ce4+)) can be then determined as ~34%, ~2%, and ~18% for each curve respectively32,48,49.Fig. 6 Chemical structure of CGO.

a iDPC image of the CGO along <110> direction. A F-type model and the simulated iDPC image are overlaid. b EELS SI results: simultaneously acquired ADF image and elemental maps plotting the intensity from the Ce M4,5, Gd M4,5 and O K edge. At the lower-left corner is the vertically averaged t/λ profile, and on the right is the laterally averaged intensity profiles from the O, Gd and Ce map. c Fine structure of the Ce M4,5 edge from three different regions (#1 to #3).

The migration of VO∙∙ and CeM′

The release and acquisition of O atoms in ceria are often considered essential for the reversible transition between F- and C-type, which is also paired with the ordering of VO⋅⋅ and reduction of Ce50. To release the O atoms from ceria nanoparticles, heating treatment (>900 K), H2 environment, or very low O partial pressure (~10−26 Pa) are required during in situ ETEM27,28,30,51. Our experiment took place at room temperature under 7.5 × 10-6 Pa, and any heating effect by e-beam is rather ignorable (Supplementary Note 5 and Supplementary Table 1). Thus, the critical requirements by in situ ETEM can hardly be satisfied in our case. Intense e-beam could also remove the O atoms from the sample surface and generate VO⋅⋅s20,52. Usually, a threshold energy about tens of eV is required53. Here, the phase transitions were observed with different accelerating voltages: 60, 200, and 300 kV (Supplementary Fig. 7), and the maximum energy transfer during an elastic scattering process is ~8.7, ~32.8, and ~53.2 eV, respectively. Thus, in most cases, the transferred energy would be insufficient to knock the O out. Carefully examining the HRTEM image time series acquired at 300 kV also excludes any significant loss of atoms during imaging (Supplementary Fig. 8)54. Additionally, a redox process can be driven by electric field. By in situ biasing TEM16,31,37, reversible migration of VO⋅⋅ and the associated phase transition have been reproducibly achieved. Similarly, an electric field can be built up by e-beam irradiation55,56 and triggers the phase transition38,57. However, the built-up electric field under the TEM illumination will attract O2- to the irradiated region, opposite to the rising VO⋅⋅ concentration as observed. Thus, such a field should only have limited contribution to our observation. Moreover, the phase transition is only detected in TEM mode, but not in STEM mode. Considering the different illumination geometry, a parallel beam (> 100 s nm) in TEM mode and a focused probe (< 0.1 nm) in STEM mode, it is reasonable to interpret the phase transition as a collective rearrangement of a large number of VO⋅⋅s stimulated by the incident electrons. Obviously, the ultrafine STEM probe is unlikely to interact with many VO⋅⋅s at the same time, and cause detectable structural change.

The exact mechanism which accounts for the phase transition is still to be defined. Qualitatively, the phase transition consists of two competing processes: accumulating and ordering of the VO⋅⋅ towards the irradiated region (P1, the forward F-to-C transition), and dissipating and disordering of the VO⋅⋅ from the irradiated region (P2, the reversed F-to-C transition). Both P1 and P2 need to be activated, while the activation energy for P1 is usually higher than that of P2. Thus, the power of external stimulus (the EDR in our case) will decide which process prevails, and subsequently the direction and efficiency of the transition. Without external stimulus (e-beam blanked), the formed phase was found to be rather stable (Supplementary Fig. 9). In Fig. 5, an ultrahigh EDR 5.0 as well as an extended irradiation were used to push the F-to-C transition to its limit. Between Figs. 5g, h, continuous irradiation raised the gap in region #1 from ~21 pm to ~31 pm. In comparison, for region #2 with an initial gap of ~25 pm, the e-beam was soon no longer able to attract more VO⋅⋅ to this region. Instead, the already formed C-type structure was activated to rotate, and finally a gap of ~31 pm was detected in the perpendicular direction in Fig. 5h. This observation of C-type rotating instead of further forming could potentially account for the dropped slope in Fig. 3h for EDR 1.0. The highly dynamic diffusion of VO⋅⋅ causes not only the O lattice distortion, but also non-saturated chemical bonding. As plotted in Figs. 4c, d, the experimental AO_v goes up to ~0.04, corresponding roughly to a δ=0.44 and M0.124+M0.883+O2−0.442−. For our sample, only 12% of the M sites are occupied by Gd3+ (Supplementary Note 1 and Fig. 1), which is incapable of migration. Thus, within the irradiated region, considerable Ce4+ has to be reduced to Ce3+. As shown in Fig. 6c, the local Ce valences vary significantly, which could facilitate a substantial electron exchange between the Ce4+ and Ce3+ inside and outside the irradiated region.

Including all the observed features, a possible transition process is proposed in Fig. 7. Assuming an average composition of Ce0.674+Ce0.213+Gd0.123+O2−0.1652− with F-type structure (F-CGO33, 33% of the metal sites are occupied by M3+), Fig. 7a depicts the early stage of an F-to-C transition along < 001 > . One unit-cell of the F-type is outlined at the upper-left corner. Two regions, labeled as Aa∥ and Aa⊥, adopt the C-type structure (C-CGO40). The superscript ∥ and ⊥ refer to the two perpendicular orientations of the C-type structure. Due to charge compensation, the substrate changes accordingly to F-CGO32 as indicated at the lower-left corner. The rising r in Fig. 3e suggests an expansion of the C-type under e-beam irradiation. Thus, in Fig. 7b, two new regions, C-CGO50 and C-CGO40, are introduced as Bb∥ and Bb⊥. Besides, Figs. 4c, d suggest accumulating VO⋅⋅ and CeM′ under the e-beam irradiation. Therefore, the Aa∥ and Aa⊥ (both C-CGO40) evolve to Ab∥ (C-CGO80) and Ab⊥ (C-CGO60) respectively. As a result, the substrate is now F-CGO16. In Fig. 7c, the e-beam irradiation further pushes the VO⋅⋅ and CeM′ concentration, leading to Bc∥ (C-CGO60), Bc⊥ (C-CGO60), and Ac⊥ (C-CGO80). Moreover, instead of achieving even higher VO⋅⋅ and CeM′ concentration, the Ab∥ in Fig. 7b experiences a 90° rotation and integrates into Ac⊥ in Fig. 7c, similar to the observation in Fig. 5b. All these changes then leave a substrate with F-CGO9. Table 1 lists the compositions for the different regions. Corresponding to the outlined rectangles in Figs. 7a–c, Figs. 7d–f list the simulated HRTEM images. The experimentally observed features, including the broader oscillation of dO_x and 90° rotation of the C-type structure, are well embodied in the simulations.Fig. 7 Proposed F-to-C transition.

The C- and F-type are separated by the colored backgrounds, and one unit-cell of the F-type is outlined at the top-left corner. a Early stage of the transition: two small regions (Aa∥ and Aa⊥) take the C-type structure (C-CGO40, 40% of the metal sites are occupied by M3+) with a relative rotation of 90°. b Two more regions transfer to the C-type: Bb∥ (C-CGO50) and Bb⊥ (C-CGO40). Meanwhile, the A regions evolve to Ab∥ (C-CGO80) and Ab⊥ (C-CGO60). c Further irradiation raises the VO∙∙ concentration in the B and A⊥ region, where Bc∥, Bc⊥ and Ac⊥ become C-CGO60, C-CGO50, and C-CGO80. Besides, the Ab∥ region experiences a 90° rotation and are combined with A⊥. Composition of the F-type substrate changes accordingly, as indicated at the lower-left corners. d–f Simulated HRTEM images corresponding to the regions defined by the rectangles in (a–c).

Table 1 Compositions of the labeled regions in Figs. 7a–c

	A∥	A⊥	B∥	B⊥	substrate	F-type	C-type	
a	C-CGO40	C-CGO40	F-CGO32	F-CGO32	F-CGO30	85%	15%	
b	C-CGO80	C-CGO60	C-CGO50	C-CGO40	F-CGO16	57%	43%	
c	C-CGO80.r	C-CGO80	C-CGO60	C-CGO50	F-CGO9	57%	43%	

In summary, the reversible phase transitions of CGO were explored in situ with picometer precision, employing a combination of advanced TEM techniques. With proper EDRs, the transitions can be accelerated, retarded, or reversed. The simultaneous visualization of both M and O sites allows a precise determination of lattice distortions associated with the transition, and further sheds light on the local VO⋅⋅ concentration. Besides, the C-type structure was observed orienting along two perpendicular yet competing directions, suggesting that the e-beam can not only alter the VO⋅⋅ concentration but also manipulate the C-type orientation. In the end, a collective rearrangement of VO⋅⋅ and CeM′ stimulated by e-beam was proposed to account for the observed transitions, which calls for further numeric calculation to establish a solid theory. Our findings demonstrate a largely controllable phase transition of ceria, unravel the associated VO⋅⋅ dynamics with unprecedented resolution and showcase great promise for diverse applications. Boosting the performance of ceria-based catalysts and electrolytes would be possible by remarkably reducing the operating temperature, as the redox of ceria is realized at room temperature in our case. Benefitting from the convenient design of ceria with desired properties through adjusting the VO⋅⋅ concentration, ceria-based memristors are to be advanced with tunable on/off ratios and superior memory density. The same principle could also be extended to other functional oxides.

Methods

Pellet sample preparation

Dense pellet with a nominal composition of 85 wt.% Ce0.8Gd0.2O2-δ:15 wt.% FeCo2O4 (85CGO20-FCO) was prepared by solid state reactive sintering (SSRS) method58,59. The phase ratio shown in the nominal composition refers to the initial weight ratio of CGO20 to FCO in the powder precursors without considering any phase interactions after sintering. To prepare the powder precursors, commercially available powders of CGO20 (Treibacher Industrie AG, Austria) as well as Fe2O3 and Co3O4 (Sigma-Aldrich, Germany) were homogenized by ball milling in ethanol. Details of the ball milling procedures can be found in ref.58 After drying, the powder mixtures were pressed into pellets and sintered at 1200 °C for 10 h. During cooling, a slow rate of 0.5 °C/min was used between 900 °C and 800 °C in order to avoid cracking60, while for the other temperature ranges, a faster cooling rate of 3 °C/min was used. Based on the previous work32,33, the CGO grain interior in the composite is not affected by the FCO phase, and thus is chosen for this study.

TEM sample preparation

A pallet of 85CGO20-FCO was first embedded in resin (Kulzer). After gradual grinding and polishing, a bulk cross-sectional sample with a flat surface was prepared. TEM cross-sectional specimens were cut from the polished bulk sample by focused ion beam (FIB) milling using an FEI Strata 400 system with a Ga ion beam. Carbon coating was applied to reduce the possible charging problem during FIB preparation. Further thinning and cleaning were performed with an Ar ion beam in a Fischione Nanomill 1040 using 900 and 500 eV beam energy in sequence.

Characterization techniques

TEM bright field imaging and electron diffraction were performed by JEOL JEM F200 at 200 kV accelerating voltage. The image time series with various EDRs (in Fig. 3) were recorded in a consecutively way from the same region. The tilted sample was left inside the microscope for hours before imaging, to minimized potential sample drift. The diffraction patterns (in Fig. 2g) were recorded from a region with ~120 nm in diameter. HRTEM NCSI was recorded with 300 kV accelerating voltage by Spectra 300, equipped with a high-brightness X-FEG monochromated source, a piezo-enhanced CompuStage, and two Cs correction optics. The CETCOR below the objective lens can be used for high-resolution TEM imaging with a bottom-mounted, retractable, fast Ceta CMOS camera. HRTEM image simulation was performed using Dr. Probe under the experimental condition with a typical sample thickness around 5 nm. The structural models were visualized with VESTA61. High resolution HAADF and iDPC imaging, EDX chemical mapping and EELS SI with atomic resolution were conducted with 200 kV accelerating voltage (unless otherwise noted) in an FEI Titan G2 80-200 ChemiSTEM microscope equipped with an XFEG, a probe Cs corrector, a super-X EDX system, and a Gatan Enfinium ER (model 977) spectrometer with DUAL EELS acquisition capability62. The convergence semi-angle for STEM imaging and EDX chemical mapping was approximately 22 mrad, while the collection semi-angles were 80-200 mrad for HAADF imaging, around 10-60 mrad for iDPC imaging, and around 47 mrad for EELS spectrum imaging. HAADF and iDPC image simulation was performed using Dr. Probe under the experimental condition with a typical sample thickness around 8 nm. EELS spectrum images were recorded with 0.5 eV per channel energy dispersion and 0.1 s dwell time for each pixel. Multivariate statistical analysis (MSA) was performed to reduce the noise of the EEL spectra with weighted principle-component analysis (PCA).

Supplementary information

Supplementary Information

Peer Review File

Source data

Source data file

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-024-52386-3.

Acknowledgements

This work has been supported by the Deutsche Forschungsgemeinschaft (Project Number 387282673). K.R. acknowledges support by the Bundesministerium für Bildung und Forschung (NEUROTEC, 16ME0399, and 16ME0398K). The authors thanks Dr. Hongchu Du and Penghan Lu (Forschungszentrum Jülich) for valuable discussion.

Author contributions

K.R. designed and performed the TEM experiments and data analysis. L.J. and J.M. provide insightful discussion on the TEM results. K.R. and L.J. performed the simulation of TEM images. F.Z. synthesized the CGO under the supervision of S.B. and W.A.M. The manuscript was written by K.R. with contributions from all authors.

Peer review

Peer review information

Nature Communications thanks Abhaya Datye, who co-reviewed with Stephen Porter, and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Funding

Open Access funding enabled and organized by Projekt DEAL.

Data availability

The data that support the findings of this study are available within the paper and supplementary information files. Source data are provided in this paper.

Competing interests

The authors declare no competing interests.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
==== Refs
References

1. Skorodumova NV Simak SI Lundqvist BI Abrikosov IA Johansson B Quantum origin of the oxygen storage capability of ceria Phys. Rev. Lett. 2002 89 166601 10.1103/PhysRevLett.89.166601 12398742
Skorodumova, N. V., Simak, S. I., Lundqvist, B. I., Abrikosov, I. A. & Johansson, B. Quantum origin of the oxygen storage capability of ceria. Phys. Rev. Lett. 89, 166601 (2002).12398742
2. Chueh WC High-flux solar-driven thermochemical dissociation of CO2 and H2O using nonstoichiometric ceria Science 2010 330 1797 1801 10.1126/science.1197834 21205663
Chueh, W. C. et al. High-flux solar-driven thermochemical dissociation of CO2 and H2O using nonstoichiometric ceria. Science 330, 1797–1801 (2010).21205663
3. Chueh WC Hao Y Jung W Haile SM High electrochemical activity of the oxide phase in model ceria-Pt and ceria-Ni composite anodes Nat. Mater. 2012 11 155 161 10.1038/nmat3184
Chueh, W. C., Hao, Y., Jung, W. & Haile, S. M. High electrochemical activity of the oxide phase in model ceria-Pt and ceria-Ni composite anodes. Nat. Mater. 11, 155–161 (2012).
4. Lee JG Park JH Shul YG Tailoring gadolinium-doped ceria-based solid oxide fuel cells to achieve 2 W cm−2 at 550 °C Nat. Commun. 2014 5 4045 10.1038/ncomms5045 24893929
Lee, J. G., Park, J. H. & Shul, Y. G. Tailoring gadolinium-doped ceria-based solid oxide fuel cells to achieve 2 W cm−2 at 550 °C. Nat. Commun. 5, 4045 (2014).24893929
5. Nie L Activation of surface lattice oxygen in single-atom Pt/CeO2 for low-temperature CO oxidation Science 2017 358 1419 1423 10.1126/science.aao2109 29242344
Nie, L. et al. Activation of surface lattice oxygen in single-atom Pt/CeO2 for low-temperature CO oxidation. Science 358, 1419–1423 (2017).29242344
6. Fu Q Saltsburg H Flytzani-Stephanopoulos M Active nonmetallic Au and Pt species on ceria-based water-gas shift catalysts Science 2003 301 935 938 10.1126/science.1085721 12843399
Fu, Q., Saltsburg, H. & Flytzani-Stephanopoulos, M. Active nonmetallic Au and Pt species on ceria-based water-gas shift catalysts. Science 301, 935–938 (2003).12843399
7. Farmer JA Campbell CT Ceria maintains smaller metal catalyst particles by strong metal-support bonding Science 2010 329 933 936 10.1126/science.1191778 20724631
Farmer, J. A. & Campbell, C. T. Ceria maintains smaller metal catalyst particles by strong metal-support bonding. Science 329, 933–936 (2010).20724631
8. Farrauto R New material needs for hydrocarbon fuel processing: generating hydrogen for the PEM fuel cell Annu Rev. Mater. Res 2003 33 1 27 10.1146/annurev.matsci.33.022802.091348
Farrauto, R. et al. New material needs for hydrocarbon fuel processing: generating hydrogen for the PEM fuel cell. Annu Rev. Mater. Res 33, 1–27 (2003).
9. Murray EP Tsai T Barnett SA A direct-methane fuel cell with a ceria-based anode Nature 1999 400 649 651 10.1038/23220
Murray, E. P., Tsai, T. & Barnett, S. A. A direct-methane fuel cell with a ceria-based anode. Nature 400, 649–651 (1999).
10. Jones J Thermally stable single-atom platinum-on-ceria catalysts via atom trapping Science 2016 353 150 154 10.1126/science.aaf8800 27387946
Jones, J. et al. Thermally stable single-atom platinum-on-ceria catalysts via atom trapping. Science 353, 150–154 (2016).27387946
11. Liu HX Partially sintered copper-ceria as excellent catalyst for the high-temperature reverse water gas shift reaction Nat. Commun. 2022 13 867 10.1038/s41467-022-28476-5 35165303
Liu, H. X. et al. Partially sintered copper-ceria as excellent catalyst for the high-temperature reverse water gas shift reaction. Nat. Commun. 13, 867 (2022).35165303
12. Qiao Z Electrochemical and electrical properties of doped CeO2-ZnO composite for low-temperature solid oxide fuel cell applications J. Power Sources 2018 392 33 40 10.1016/j.jpowsour.2018.04.096
Qiao, Z. et al. Electrochemical and electrical properties of doped CeO2-ZnO composite for low-temperature solid oxide fuel cell applications. J. Power Sources 392, 33–40 (2018).
13. Wang BY Fast ionic conduction in semiconductor CeO2-δ electrolyte fuel cells Npg Asia Mater. 2019 11 51 10.1038/s41427-019-0152-8
Wang, B. Y. et al. Fast ionic conduction in semiconductor CeO2-δ electrolyte fuel cells. Npg Asia Mater. 11, 51 (2019).
14. Kim HJ Artificial synaptic characteristics with strong analog memristive switching in a Pt/CeO2/Pt structure Nanotechnology 2017 28 285203 10.1088/1361-6528/aa712c 28643696
Kim, H. J. et al. Artificial synaptic characteristics with strong analog memristive switching in a Pt/CeO2/Pt structure. Nanotechnology 28, 285203 (2017).28643696
15. Schmitt R Spring J Korobko R Rupp JLM Design of oxygen vacancy configuration for memristive systems Acs Nano 2017 11 8881 8891 10.1021/acsnano.7b03116 28850213
Schmitt, R., Spring, J., Korobko, R. & Rupp, J. L. M. Design of oxygen vacancy configuration for memristive systems. Acs Nano 11, 8881–8891 (2017).28850213
16. Gao P TEM studies of oxygen vacancy migration for electrically induced resistance change effect in cerium oxides Micron 2010 41 301 305 10.1016/j.micron.2009.11.010 20042340
Gao, P. et al. TEM studies of oxygen vacancy migration for electrically induced resistance change effect in cerium oxides. Micron 41, 301–305 (2010).20042340
17. Sun L Polaronic resistive switching in ceria-based memory devices Adv. Electron Mater. 2019 5 1900271 10.1002/aelm.201900271
Sun, L. et al. Polaronic resistive switching in ceria-based memory devices. Adv. Electron Mater. 5, 1900271 (2019).
18. Artini C Costa GA Pani M Lausi A Plaisier J Structural characterization of the CeO2/Gd2O3 mixed system by synchrotron X-ray diffraction J. Solid State Chem. 2012 190 24 28 10.1016/j.jssc.2012.01.056
Artini, C., Costa, G. A., Pani, M., Lausi, A. & Plaisier, J. Structural characterization of the CeO2/Gd2O3 mixed system by synchrotron X-ray diffraction. J. Solid State Chem. 190, 24–28 (2012).
19. Ou DR Oxygen-vacancy ordering in lanthanide-doped ceria: dopant-type dependence and structure model Phys. Rev. B 2008 77 024108 10.1103/PhysRevB.77.024108
Ou, D. R. et al. Oxygen-vacancy ordering in lanthanide-doped ceria: dopant-type dependence and structure model. Phys. Rev. B 77, 024108 (2008).
20. Johnston-Peck AC DuChene JS Roberts AD Wei WD Herzing AA Dose-rate-dependent damage of cerium dioxide in the scanning transmission electron microscope Ultramicroscopy 2016 170 1 9 10.1016/j.ultramic.2016.07.002 27469265
Johnston-Peck, A. C., DuChene, J. S., Roberts, A. D., Wei, W. D. & Herzing, A. A. Dose-rate-dependent damage of cerium dioxide in the scanning transmission electron microscope. Ultramicroscopy 170, 1–9 (2016).27469265
21. Scavini M Coduri M Allieta M Brune M Ferrero C Probing complex disorder in Ce1-xGdxO2-x/2 using the pair distribution function analysis Chem. Mater. 2012 24 1338 1345 10.1021/cm203819u
Scavini, M., Coduri, M., Allieta, M., Brune, M. & Ferrero, C. Probing complex disorder in Ce1-xGdxO2-x/2 using the pair distribution function analysis. Chem. Mater. 24, 1338–1345 (2012).
22. Li ZP Structural phase transformation through defect cluster growth in Gd-doped ceria Phys. Rev. B 2011 84 180201 10.1103/PhysRevB.84.180201
Li, Z. P. et al. Structural phase transformation through defect cluster growth in Gd-doped ceria. Phys. Rev. B 84, 180201 (2011).
23. Artini C Structural features of Sm- and Gd-doped ceria studied by synchrotron x-ray diffraction and μ-Raman spectroscopy Inorg. Chem. 2015 54 4126 4137 10.1021/acs.inorgchem.5b00395 25849073
Artini, C. et al. Structural features of Sm- and Gd-doped ceria studied by synchrotron x-ray diffraction and μ-Raman spectroscopy. Inorg. Chem. 54, 4126–4137 (2015).25849073
24. Scavini M Percolating hierarchical defect structures drive phase transformation in Ce1-xGdxO2-x/2: a total scattering study Iucrj 2015 2 511 522 10.1107/S2052252515011641 26306193
Scavini, M. et al. Percolating hierarchical defect structures drive phase transformation in Ce1-xGdxO2-x/2: a total scattering study. Iucrj 2, 511–522 (2015).26306193
25. Yoshida H Visualizing gas molecules interacting with supported nanoparticulate catalysts at reaction conditions Science 2012 335 317 319 10.1126/science.1213194 22267808
Yoshida, H. et al. Visualizing gas molecules interacting with supported nanoparticulate catalysts at reaction conditions. Science 335, 317–319 (2012).22267808
26. Hao XD Atomic-scale valence state distribution inside ultrafine CeO2 nanocubes and its size dependence Small 2018 14 1802915 10.1002/smll.201802915
Hao, X. D. et al. Atomic-scale valence state distribution inside ultrafine CeO2 nanocubes and its size dependence. Small 14, 1802915 (2018).
27. Sharma R Crozier PA Kang ZC Eyring L Observation of dynamic nanostructural and nanochemical changes in ceria-based catalysts during in situ reduction Philos. Mag. 2004 84 2731 2747 10.1080/14786430410001671467
Sharma, R., Crozier, P. A., Kang, Z. C. & Eyring, L. Observation of dynamic nanostructural and nanochemical changes in ceria-based catalysts during in situ reduction. Philos. Mag. 84, 2731–2747 (2004).
28. Crozier PA Wang RG Sharma R environmental TEM studies of dynamic changes in cerium-based oxides nanoparticles during redox processes Ultramicroscopy 2008 108 1432 1440 10.1016/j.ultramic.2008.05.015 18687527
Crozier, P. A., Wang, R. G. & Sharma, R. environmental TEM studies of dynamic changes in cerium-based oxides nanoparticles during redox processes. Ultramicroscopy 108, 1432–1440 (2008).18687527
29. Crozier PA Hansen TW in situ and operando transmission electron microscopy of catalytic materials Mrs Bull. 2015 40 38 45 10.1557/mrs.2014.304
Crozier, P. A. & Hansen, T. W. in situ and operando transmission electron microscopy of catalytic materials. Mrs Bull. 40, 38–45 (2015).
30. Bugnet M Overbury SH Wu ZL Epicier T Direct visualization and control of atomic mobility at {100} surfaces of ceria in the environmental transmission electron microscope Nano Lett. 2017 17 7652 7658 10.1021/acs.nanolett.7b03680 29166035
Bugnet, M., Overbury, S. H., Wu, Z. L. & Epicier, T. Direct visualization and control of atomic mobility at {100} surfaces of ceria in the environmental transmission electron microscope. Nano Lett. 17, 7652–7658 (2017).29166035
31. Gao P Electrically driven redox process in cerium oxides J. Am. Chem. Soc. 2010 132 4197 4201 10.1021/ja9086616 20218562
Gao, P. et al. Electrically driven redox process in cerium oxides. J. Am. Chem. Soc. 132, 4197–4201 (2010).20218562
32. Ran K Tuning the ceria interfaces inside the dual phase oxygen transport membranes Acta Mater. 2022 226 117603 10.1016/j.actamat.2021.117603
Ran, K. et al. Tuning the ceria interfaces inside the dual phase oxygen transport membranes. Acta Mater. 226, 117603 (2022).
33. Ran K The in situ generated emerging phase inside dual phase oxygen transport membranes Acta Mater. 2022 234 118034 10.1016/j.actamat.2022.118034
Ran, K. et al. The in situ generated emerging phase inside dual phase oxygen transport membranes. Acta Mater. 234, 118034 (2022).
34. Kossoy A Frenkel AI Wang Q Wachtel E Lubomirsky I Local structure and strain-induced distortion in Ce0.8Gd0.2O1.9 Adv. Mater. 2010 22 1659 1662 10.1002/adma.200902041 20496400
Kossoy, A., Frenkel, A. I., Wang, Q., Wachtel, E. & Lubomirsky, I. Local structure and strain-induced distortion in Ce0.8Gd0.2O1.9. Adv. Mater. 22, 1659–1662 (2010).20496400
35. Korobko R Giant electrostriction in Gd-doped ceria Adv. Mater. 2012 24 5857 5861 10.1002/adma.201202270 22933377
Korobko, R. et al. Giant electrostriction in Gd-doped ceria. Adv. Mater. 24, 5857–5861 (2012).22933377
36. Wei XK Jia CL Roleder K Dunin-Borkowski RE Mayer J In situ observation of point-defect-induced unit-cell-wise energy storage pathway in antiferroelectric PbZrO3 Adv. Funct. Mater. 2021 31 2008609 10.1002/adfm.202008609
Wei, X. K., Jia, C. L., Roleder, K., Dunin-Borkowski, R. E. & Mayer, J. In situ observation of point-defect-induced unit-cell-wise energy storage pathway in antiferroelectric PbZrO3. Adv. Funct. Mater. 31, 2008609 (2021).
37. Yao LD Electron-beam-induced perovskite-brownmillerite-perovskite structural phase transitions in epitaxial La2/3Sr1/3MnO3 Films Adv. Mater. 2014 26 2789 2793 10.1002/adma.201305656 24554587
Yao, L. D. et al. Electron-beam-induced perovskite-brownmillerite-perovskite structural phase transitions in epitaxial La2/3Sr1/3MnO3 Films. Adv. Mater. 26, 2789–2793 (2014).24554587
38. Tao J Reversible structure manipulation by tuning carrier concentration in metastable Cu2S Proc. Natl Acad. Sci. USA 2017 114 9832 9837 10.1073/pnas.1709163114 28855335
Tao, J. et al. Reversible structure manipulation by tuning carrier concentration in metastable Cu2S. Proc. Natl Acad. Sci. USA 114, 9832–9837 (2017).28855335
39. Zheng HM Observation of transient structural-transformation dynamics in a Cu2S nanorod Science 2011 333 206 209 10.1126/science.1204713 21737738
Zheng, H. M. et al. Observation of transient structural-transformation dynamics in a Cu2S nanorod. Science 333, 206–209 (2011).21737738
40. Jia CL Houben L Thust A Barthel J On the benefit of the negative-spherical-aberration imaging technique for quantitative HRTEM Ultramicroscopy 2010 110 500 505 10.1016/j.ultramic.2009.10.006
Jia, C. L., Houben, L., Thust, A. & Barthel, J. On the benefit of the negative-spherical-aberration imaging technique for quantitative HRTEM. Ultramicroscopy 110, 500–505 (2010).
41. Jia CL Determination of the 3D shape of a nanoscale crystal with atomic resolution from a single image Nat. Mater. 2014 13 1044 1049 10.1038/nmat4087 25242534
Jia, C. L. et al. Determination of the 3D shape of a nanoscale crystal with atomic resolution from a single image. Nat. Mater. 13, 1044–1049 (2014).25242534
42. Ge ZH Atomic-scale observation of off-centering rattlers in filled skutterudites Adv. Energy Mater. 2022 12 2103770 10.1002/aenm.202103770
Ge, Z. H. et al. Atomic-scale observation of off-centering rattlers in filled skutterudites. Adv. Energy Mater. 12, 2103770 (2022).
43. Yücelen E Lazic I Bosch EGT Phase contrast scanning transmission electron microscopy imaging of light and heavy atoms at the limit of contrast and resolution Sci. Rep.-Uk 2018 8 2676 10.1038/s41598-018-20377-2
Yücelen, E., Lazic, I. & Bosch, E. G. T. Phase contrast scanning transmission electron microscopy imaging of light and heavy atoms at the limit of contrast and resolution. Sci. Rep.-Uk 8, 2676 (2018).
44. Barthel J Dr. Probe: A software for high-resolution STEM image simulation Ultramicroscopy 2018 193 1 11 10.1016/j.ultramic.2018.06.003 29906518
Barthel, J. Dr. Probe: A software for high-resolution STEM image simulation. Ultramicroscopy 193, 1–11 (2018).29906518
45. Nord M Vullum PE MacLaren I Tybell T Holmestad R Atomap: a new software tool for the automated analysis of atomic resolution images using two-dimensional Gaussian fitting Adv. Struct. Chem. Imaging 2017 3 1 12 10.1186/s40679-017-0042-5 28261539
Nord, M., Vullum, P. E., MacLaren, I., Tybell, T. & Holmestad, R. Atomap: a new software tool for the automated analysis of atomic resolution images using two-dimensional Gaussian fitting. Adv. Struct. Chem. Imaging 3, 1–12 (2017).28261539
46. Artini C Pani M Lausi A Masini R Costa GA High temperature structural study of Gd-doped ceria by synchrotron X-ray diffraction (673 K ≤ T ≤ 1073 K) Inorg. Chem. 2014 53 10140 10149 10.1021/ic5011242 25192043
Artini, C., Pani, M., Lausi, A., Masini, R. & Costa, G. A. High temperature structural study of Gd-doped ceria by synchrotron X-ray diffraction (673 K ≤ T ≤ 1073 K). Inorg. Chem. 53, 10140–10149 (2014).25192043
47. Grover V Achary SN Tyagi AK Structural analysis of excess-anion C-type rare earth oxide:: a case study with Gd1-xCexO1.5+x/2 (x= 0.20 and 0.40) J. Appl Crystallogr 2003 36 1082 1084 10.1107/S0021889803010288
Grover, V., Achary, S. N. & Tyagi, A. K. Structural analysis of excess-anion C-type rare earth oxide:: a case study with Gd1-xCexO1.5+x/2 (x= 0.20 and 0.40). J. Appl Crystallogr 36, 1082–1084 (2003).
48. Fortner JA Buck EC The chemistry of the light rare-earth elements as determined by electron energy loss spectroscopy Appl Phys. Lett. 1996 68 3817 3819 10.1063/1.116627
Fortner, J. A. & Buck, E. C. The chemistry of the light rare-earth elements as determined by electron energy loss spectroscopy. Appl Phys. Lett. 68, 3817–3819 (1996).
49. Yan PF Microstructural and chemical characterization of ordered structure in yttrium doped ceria Microsc Microanal. 2013 19 102 110 10.1017/S1431927612013888 23308398
Yan, P. F. et al. Microstructural and chemical characterization of ordered structure in yttrium doped ceria. Microsc Microanal. 19, 102–110 (2013).23308398
50. Campbell CT Peden CHF Chemistry - oxygen vacancies and catalysis on ceria surfaces Science 2005 309 713 714 10.1126/science.1113955 16051777
Campbell, C. T. & Peden, C. H. F. Chemistry - oxygen vacancies and catalysis on ceria surfaces. Science 309, 713–714 (2005).16051777
51. Bevan DJM Kordis J Mixed oxides of the type MO2 (Fluorite)-M2O3-I. Oxygen dissociation pressures and phase relationships in the system CeO2-Ce2O3 at high temperatures J. Inorg. Nucl. Chem. 1964 26 1509 1523 10.1016/0022-1902(64)80038-5
Bevan, D. J. M. & Kordis, J. Mixed oxides of the type MO2 (Fluorite)-M2O3-I. Oxygen dissociation pressures and phase relationships in the system CeO2-Ce2O3 at high temperatures. J. Inorg. Nucl. Chem. 26, 1509–1523 (1964).
52. Zhu L Visualizing anisotropic oxygen diffusion in ceria under activated conditions Phys. Rev. Lett. 2020 124 056002 10.1103/PhysRevLett.124.056002 32083924
Zhu, L. et al. Visualizing anisotropic oxygen diffusion in ceria under activated conditions. Phys. Rev. Lett. 124, 056002 (2020).32083924
53. Egerton RF Radiation damage to organic and inorganic specimens in the TEM Micron 2019 119 72 87 10.1016/j.micron.2019.01.005 30684768
Egerton, R. F. Radiation damage to organic and inorganic specimens in the TEM. Micron 119, 72–87 (2019).30684768
54. Lin YY Wu ZL Wen JG Poeppelmeier KR Marks LD Imaging the Atomic Surface Structures of CeO2 Nanoparticles Nano Lett. 2014 14 191 196 10.1021/nl403713b 24295383
Lin, Y. Y., Wu, Z. L., Wen, J. G., Poeppelmeier, K. R. & Marks, L. D. Imaging the Atomic Surface Structures of CeO2 Nanoparticles. Nano Lett. 14, 191–196 (2014).24295383
55. Jiang N Electron beam damage in oxides: a review Rep. Prog. Phys. 2016 79 016501 10.1088/0034-4885/79/1/016501 26684361
Jiang, N. Electron beam damage in oxides: a review. Rep. Prog. Phys. 79, 016501 (2016).26684361
56. Jiang N Electron irradiation effects in transmission electron microscopy: Random displacements and collective migrations Micron 2023 171 103482 10.1016/j.micron.2023.103482 37167653
Jiang, N. Electron irradiation effects in transmission electron microscopy: Random displacements and collective migrations. Micron 171, 103482 (2023).37167653
57. Yang ZZ Guided anisotropic oxygen transport in vacancy ordered oxides Nat. Commun. 2023 14 6068 10.1038/s41467-023-40746-4 37770428
Yang, Z. Z. et al. Guided anisotropic oxygen transport in vacancy ordered oxides. Nat. Commun. 14, 6068 (2023).37770428
58. Zeng FL Optimization of sintering conditions for improved microstructural and mechanical properties of dense Ce0.8Gd0.2O2-δ-FeCo2O4 oxygen transport membranes J. Eur. Ceram. Soc. 2021 41 509 516 10.1016/j.jeurceramsoc.2020.09.009
Zeng, F. L. et al. Optimization of sintering conditions for improved microstructural and mechanical properties of dense Ce0.8Gd0.2O2-δ-FeCo2O4 oxygen transport membranes. J. Eur. Ceram. Soc. 41, 509–516 (2021).
59. Fischer L Phase formation and performance of solid state reactive sintered Ce0.8Gd0.2O2-δ-FeCo2O4 composites J. Mater. Chem. A 2022 10 2412 2420 10.1039/D1TA05695F
Fischer, L. et al. Phase formation and performance of solid state reactive sintered Ce0.8Gd0.2O2-δ-FeCo2O4 composites. J. Mater. Chem. A 10, 2412–2420 (2022).
60. Ramasamy M Structural and chemical stability of high performance Ce0.8Gd0.2O2-δ-FeCo2O4 dual phase oxygen transport membranes J. Membr. Sci. 2017 544 278 286 10.1016/j.memsci.2017.09.011
Ramasamy, M. et al. Structural and chemical stability of high performance Ce0.8Gd0.2O2-δ-FeCo2O4 dual phase oxygen transport membranes. J. Membr. Sci. 544, 278–286 (2017).
61. Momma K Izumi F VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data J. Appl Crystallogr 2011 44 1272 1276 10.1107/S0021889811038970
Momma, K. & Izumi, F. VESTA 3 for three-dimensional visualization of crystal, volumetric and morphology data. J. Appl Crystallogr 44, 1272–1276 (2011).
62. Kovács A Schierholz R Tillmann K FEI Titan G2 80−200 CREWLEY J. large-scale Res. facilities 2016 2 A43 10.17815/jlsrf-2-68
Kovács, A., Schierholz, R. & Tillmann, K. FEI Titan G2 80−200 CREWLEY. J. large-scale Res. facilities 2, A43 (2016).
