Aluminothermic reduction of CeO₂: mechanism of an economical route to aluminum–cerium alloys

Abstract

Cerium oxide is a low-value byproduct of rare-earth mining yet constitutes the largest fraction of the rare earth elements. The reduction of cerium oxide by liquid aluminum is proposed as an energy- and cost-efficient route to produce high-strength Al–Ce alloys. This work investigated the mechanism of a multi-step reduction reaction to facilitate the industrial adaptation of the process. Differential scanning calorimetry in combination with time-resolved synchrotron diffraction data uncovered the rate-limiting reaction step as the origin of the reported temperature dependence of reduction efficiency. This is the first in situ study of a metallothermic reaction mechanism and will serve as guidance for cost- and energy efficient industrial process control.

---

New concepts

Aluminothermic reduction of cerium oxide can significantly reduce the cost and energy-consumption associated with Al–Ce alloy production compared to conventional processes. The direct use of cerium compounds accrued during rare earth element (REE) mining makes lightweight REE a value-added co-product instead of a waste product. In this study we demonstrate a novel approach to produce Al–Ce alloys by reducing CeO₂ in liquid aluminum at 95% reduction efficiency. This work provides the first ever mechanistic insight into the reaction kinetics and intermediate products of an aluminothermic reaction, which reveal a multi-step reaction mechanism with complex temperature dependency, details that are necessary for scaling on an industrial level. This work provides the basis of an economic route to produce high strength Al–Ce alloys with enhanced dispersion strengthening from embedded Al₂O₃ particles.

Nearly 200 years ago, H. C. Oersted revolutionized metallurgy, as he prepared metallic aluminum through the reduction of AlCl₃ by potassium amalgam, inventing the metallothermic reaction.¹ Since then, many elements have been isolated for the first time by reaction of their compounds with a more reactive metal to form a more stable compound and the sought-after elemental metal.² Metallothermic reactions remain a crucial technology applied in the iron thermite reaction for welding of train tracks and the industrial production of metals such as Be, Ti, Ta and the rare earth elements through the reduction of their halides by Mg, Na or Ca.^3,4

Rare earth elements tend to be co-located in ore deposits and are treated together in extraction. Cerium is the majority element in most deposits, and the growing need for Nd, Pr and the heavy lanthanides in permanent magnets and other energy transition technologies results in costly stockpiling of cerium oxide that has low demand.^5–7 Aluminum alloys with up to 10 wt% cerium have been developed in the last decade as a new class of light-weight, high-temperature materials, exploiting formation of the high-melting intermetallic compound Ce₃Al₁₁ in the microstructure.^6,8,9 As such, Al–Ce alloys are a competitive high strength alloy class that creates a high-value demand for excess cerium to help stabilize the rare earth market.^10,11

Al–Ce alloys are produced by alloying Al melts with metallic Ce, which itself is produced by calciothermic reduction of CeCl₃ in a prior energy-intensive step. This step can potentially be avoided by leveraging the following reaction (1), with a standard reaction enthalpy of −30.1 kJ mol⁻¹.

The direct reduction of CeO₂ in liquid aluminum shows promise as an economical route to produce Al–Ce alloys, circumventing the production of Ce metal, which will reduce the environmental footprint and facilitate commercial adaptation.

Mechanistic understanding of metallothermic reactions supports the effective parameter optimization in industry but is still in its infancy due to the harsh reaction conditions of these highly exothermic reactions. Aluminothermic reactions with ZrO₂,¹² MoO₃,^13,14 NaVO₃,¹⁵ MgO,¹⁶ and Cr₂O₃,¹⁷ have been studied for industrial alloy production, while reduction of ZnO,^18,19 was investigated by preparing Al/Al₂O₃ matrix composites.

Early investigations of the aluminothermic reduction of CeO₂,²⁰ and cerium carbonate,²¹ reported a strong temperature dependence of the conversion efficiency. No study has reported data on intermediate reaction products or mechanistic details.

To further fundamental understanding and accelerate industrial adaptation, we have investigated the reaction kinetics and mechanism of the aluminothermic reduction of CeO₂ for the preparation of Al–Ce alloys, by means of differential scanning calorimetry (DSC), metallography, time-resolved synchrotron X-ray diffraction, and thermodynamic calculations.

The reaction between liquid aluminum and CeO₂ particles was investigated via isothermal holds (800 °C ≤ T_hold ≤ 950 °C) in DSC on pressed pellets of aluminum–CeO₂ powder mixtures (Fig. 1A, experimental details in ESI†). The onset time of the exotherms (shaded areas under curves in Fig. 1A) was reduced with increasing reaction temperature from 15 min to less than 1 minute for the reactions at 825 °C and 950 °C, respectively.

Fig. 1 Calorimetry of Al–CeO₂ mixtures with color legend in the top panel. (A) Detailed view of isothermal holds where the shaded areas reflect the heat of reaction. Curves are offset for clarity. (B) Corresponding kinetic curves of degree of conversion vs. reaction time for aluminothermic reduction reactions, derived from peak integration in (A). Inset to (B): Arrhenius-plot with fit (dashed line) to the data from 825 °C to 925 °C provides E_a.

The kinetics of the reaction were further examined using the isothermal kinetic curves, i.e. normalized integral curves of the integrated peak areas, as the integral curves were found less sensitive to uncertainties in reaction onset time and baseline.^22,23 The kinetic curves (Fig. 1B) revealed an accelerated reaction rate with increasing temperature up to 925 °C, while the rate at 950 °C did not increase further. The reaction rate determined at half conversion (α = 0.5) confirmed the increase of the reaction rate with temperature and, using Friedman's isoconversional formulation of the Arrhenius equation (inset to Fig. 1B),^22,24,25 allowed estimation of an effective activation energy E_a = 161 kJ mol⁻¹. A second broad exothermic effect was observed with onset about 20, 13, 4 and 2 minutes after the completion of the initial reaction at 825, 850 °C, 875 °C and 900 °C, respectively (see Fig. 1A and Fig. S1, ESI†). This process of unclear origin appeared accelerated at higher temperatures.

The reaction products after the DSC measurement at 850 °C were identified by powder X-ray diffraction (PXRD, see Fig. S2, ESI†) as Al (Cu structure, Fm3̄m, a(300 K) = 4.045(2) Å), CeO_2−x‡(CaF₂ structure, Fm3̄m, a(300 K) = 5.4054(2) Å), Al₂O₃ (Corundum structure, R3̄c, a(300 K) = 4.753(3) Å, c(300 K) = 12.984(4) Å) and Ce₃Al₁₁ (La₃Al₁₁ structure, Immm, a(300 K) = 4.3879(2) Å, b(300 K) = 10.0513(6) Å, c(300 K) = 13.0079(7) Å) confirming successful reduction of Ce^IVO_2−x to Ce⁰₃Al₁₁ (arguendo, we assume oxidation state Ce⁰ in the intermetallic compounds). After reaction at 950 °C, Al, Al₂O₃ and Ce₃Al₁₁ but no residual CeO_2−x were observed, and the relative amounts of Ce₃Al₁₁ and Al₂O₃ doubled and tripled, respectively, compared to the reaction at 850 °C (see Fig. S3 and Table S1 for details, ESI†).

Scanning electron micrographs of the pellet after reaction at 950 °C (Fig. 2) showed large agglomerates of Ce₃Al₁₁ grains with 30 µm to 50 µm diameter in an aluminum matrix, as identified by energy-dispersive X-ray spectroscopy (EDS) (Table S2 and Fig. S4, ESI†). Higher magnification revealed small crystallites of Al₂O₃ with a narrow size distribution around 5 µm embedded in the Ce₃Al₁₁ and Al phases (red arrows in Fig. 2B–E). These are significantly larger than the Al₂O₃ particles with 200 nm diameter observed in composite materials produced by reduction of Ce₂(CO₃)₃.²¹

Fig. 2 Scanning electron microscopy on Al–CeO₂ sample after holding at 950 °C for 20 minutes. (A) Secondary electron image showing large grains of Ce₃Al₁₁ (bright phase) in an Al matrix (dark). (B) High-magnification backscatter-electron image reveals small Al₂O₃ crystals (red arrows) embedded between Ce₃Al₁₁ and Al phases, as well as unreacted oxides. (C)–(E) EDS elemental maps of Al, O, and Ce distribution in the microstructure.

After reaction at 850 °C, the micrographs (Fig. S5, ESI†) showed significant amounts of unreacted CeO_2−x particles as well as Al and Ce₃Al₁₁ grains. Detailed investigation of the micrographs (Fig. S5B–E, ESI†) revealed that the Ce₃Al₁₁ grains appear to crystallize only from within the aluminum matrix.

To better understand the kinetics and mechanism of the aluminothermic reduction of CeO₂ as well as the origin of the observed second exotherm in DSC, time-resolved synchrotron diffraction data were recorded on pressed pellets of aluminum–CeO₂ powder mixtures contained in glass capillaries (12.2.2, Advanced Light Source/LBNL, USA). Diffraction data were recorded while ramping the sample temperature to the isothermal hold temperature (T_hold = 850 °C, 900 °C and 950 °C) to monitor the formation and decomposition of crystalline phases by means of the peak areas of selected peaks (details in ESI†). At selected times, the relative phase fractions were determined by full pattern Rietveld refinement of the diffraction data (insets to Fig. 3).

Fig. 3 Phase formation (lower panels) observed by time-resolved synchrotron diffraction for the reaction of Al with CeO₂ at (A) 850 °C and (B) 950 °C, together with temperature programs (upper panels). Curves for respective phases are normalized using results from Rietveld refinements to approximately reflect the evolution of relative phase amounts over time. Insets: Relative amounts of Ce-oxide phases determined at selected times, normalized by Ce content.

As soon as melting of aluminum was complete, a rapid conversion of CeO_2−x (CaF₂ structure, Fm3̄m, lattice parameter a(300 K) = 5.4280(2) Å) to the defect oxide Ce₃O_5+x (Bixbyite structure, space group Ia3̄, a(1123 K) ≈ 11.37 Å) was observed at all temperatures.²⁷ At 850 °C (Fig. 3A), Ce₃O_5+x reacted then slowly to form the sesquioxide Ce₂O₃ (HT-La₂O₃ structure, Pm3̄, a(1123 K) ≈ 3.94 Å, c(1123 K) ≈ 6.17 Å). After 20 min of reaction time, a molar ratio of Ce₃O_5+x to Ce₂O₃, normalized by cerium content (i.e. CeO_1.67+0.5x to CeO_1.5), of about 1 : 1.7 was determined by Rietveld refinement (inset to Fig. 3A, Table S4, ESI†). Upon cooling of the reaction mixture, the signal for both phases, Ce₃O_5+x and Ce₂O₃, disappeared and together with the solidification of the excess aluminum, the crystallization of Ce₃Al₁₁ and formation of CeO_2−x was observed, in line with observations after DSC.

The reactions were strongly accelerated at 950 °C (Fig. 3B) as the melting of aluminum initiated the rapid conversion of CeO_2−x to Ce₃O_5+x which then decomposed rapidly to form Ce₂O₃.

Almost instantly, the ternary oxide CeAlO₃ (CaTiO₃ structure, Pm3̄m, a(1223 K) ≈ 3.82 Å) formed and the evolution of relative phase amounts suggests that CeAlO₃ was formed by consumption of Ce₂O₃. After 10 min, a ratio of Ce₂O₃ to CeAlO₃, normalized by cerium content (i.e. CeO_1.5 to CeAlO₃), of about 1 : 93 was observed (Inset to Fig. 3B, Table S4, ESI†). Upon cooling of the reaction mixture, the residual Ce₂O₃ signal dropped to zero, while CeAlO₃ remained constant, and Al and Ce₃Al₁₁ crystallized from the melt. The experiment at 900 °C (Fig. S7, ESI†), showed essentially a similar order of reactions as at 950 °C. After melting of Al and conversion of CeO_2−x to Ce₃O_5+x, the latter phase reacted quickly to form Ce₂O₃ and CeAlO₃. The relative amount of these phases, however, remained nearly constant over time at a 1 : 10 molar ratio (CeO_1.5 to CeAlO₃). The crystallization of small amounts of Ce₃Al₁₁ was observed prior to cooling down from 900 °C.

Thermal effects in DSC appeared delayed compared to the reactions observed in synchrotron data, most likely due to a difference in heating rate and sample mass. The first thermal effect coincides in time with the observed rapid conversion of CeO_2−x to Ce₃O_5+x and Ce₂O₃ for reactions at 850 °C and 900 °C. Synchrotron data at 900 °C revealed a broad peak in the signal for Ce₂O₃ (Fig. S1, ESI†), suggesting accelerated formation of Ce₂O₃, coinciding with the delayed second signal in the DSC curve. Above 900 °C, the first and second thermal effects seem to overlap in one peak. As the time-resolved synchrotron data indicate the continuous decrease of Ce₂O₃ amount over an extended duration, this suggests that the reduction of Ce₂O₃ to CeAlO₃ and subsequently metallic cerium is too slow to show a discernible DSC signal.

The reaction cascade taking place can then be summarized through eqn (2)–(5), omitting non-stoichiometry in the oxides or the dissolution of Ce−Al intermetallic phases in excess Al_(l), where the enthalpies were calculated using the thermodynamic database developed for this work. Due to enhanced diffusion in the liquid phase and the strongly negative mixing enthalpy²⁸ of Ce in Al_(l), the excess liquid aluminum present will enrich in Ce content by rapidly dissolving any formed Ce-containing intermetallic compounds. This will remove any formed metallic cerium quickly from the equilibrium reaction, up to the solubility limit (about 7 at% at 900 °C), and precludes the observation of CeAl₂ and CeAl₃ intermetallic compounds by synchrotron diffraction.

Ce₂O₃ + 3Al → CeAl₂ + CeAlO₃ − 123 kJ mol⁻¹ (4)

CeAlO₃ + 4Al → CeAl₃ + Al₂O₃ − 23.8 kJ mol⁻¹ (5)

The sequence of phase formation observed in synchrotron experiments can be rationalized by detailed analysis of the computed Al–Ce–O ternary phase diagram (Fig. 4A). The overall compositions (blue cross in Fig. 4A) of the samples are located in the three-phase equilibrium of Ce₃Al₁₁ with Al₂O₃ and (Al,Ce)_(l), the liquid solution of Ce in Al. The reaction between solid CeO_2−x particles and liquid aluminum Al_(l) is governed by the formation of product layers at the particle surface and a reaction front that penetrates towards the center of the particles.

Fig. 4 (A) Calculated isothermal section of the Al–Ce–O system (x_Ce ≤ 0.4) at 950 °C. Two- and three-phase equilibria are given as green and white areas (equilibrium phases indicated), respectively. Narrow two-phase equilibria have been omitted. Composition of synchrotron samples (blue cross) and connecting line (blue, dashed) from CeO_2−x to Al are indicated. (B) Phases in equilibrium at 950 °C for compositions along the join from CeO_2−x to Al (dashed line in panel A).

As a new phase is formed and expands into the particle from the particle–Al_(l) interface, the newly formed phase and excess Al_(l) at the particle surface will tend to form the next stable phases in the next local equilibrium. The relative phase amounts observed in a particle over time depend on the competing rates of phase formation (advance of reaction front into particle) and consumption of the phase by the subsequent reaction.

The observed evolution of phase fractions (Fig. 3) deviates significantly from the calculated phase amounts along the Al–CeO₂ vertical section (dashed line in Fig. 4A), which assume immediate and bulk equilibrium conditions (Fig. 4B).

During the first minutes, the rapid conversion of CeO_2−x to Ce₃O_5+x (eqn (2)) and then Ce₂O₃ (eqn (3)) was observed at all temperatures, in line with the thermodynamic predictions (Fig. 4B). Only the Ce₇O₁₂ phase was not observed experimentally, possibly due to the narrow stability window. The fast reactions between CeO_2−x, Ce₃O_5+x and Ce₂O₃ at all temperatures are in line with the high oxygen diffusion rate and the structural similarity of the cerium oxides.^29–31

The consequent reaction of Ce₂O₃ with Al to form CeAlO₃ (eqn (4)) shows markedly different dynamics depending on temperature, as outlined below. The observed strong temperature dependence of observed phase amounts of Ce₂O₃ and CeAlO₃ can be explained by the competition of the diffusion-limited reaction of Ce₂O₃ to CeAlO₃ and the consumption of CeAlO₃ to form Ce⁰ as liquid solution (Al,Ce)_(l), and solid Al₂O_3(s).

At 850 °C, Ce₃O_5+x reacts slowly to Ce₂O₃ (eqn (3)) but the consequent conversion to CeAlO₃ is extremely slow compared to its reduction to Ce⁰ (eqn (5)) and only small quantities of CeAlO₃ are therefore observed at any time. The Ce₂O₃ amount remained nearly constant during the observation period.

Rietveld phase fraction analysis of samples after DSC showed that about 63 at% of cerium was reduced to Ce⁰₃Al₁₁, about 7 at% was found as Ce^IIIAlO₃ and 30 at% as unreacted Ce^IVO₂.

At 900 °C, a steady state between conversion of Ce₂O₃ to CeAlO₃ and reduction to Ce⁰ is rapidly achieved, although at a much higher ratio of CeAlO₃ to Ce₂O₃ than at 850 °C, as diffusion through the product layer is accelerated. About 95 at% of cerium was reduced to Ce⁰₃Al₁₁, 2 at% was found as Ce^IIIAlO₃ and 3% as unreacted CeO₂.

At 950 °C, formation of CeAlO₃ from Ce₂O₃ and reduction to Ce⁰ progress at similar rates. Ce₂O₃ was mostly consumed over the observation period, but the amount of CeAlO₃ had not started to decrease by the time the reaction mixture was cooled down.

In the sample reacted during DSC, about 92 at% of cerium was reduced to Ce⁰₃Al₁₁ and 8 at% was found as Ce^IIIAlO₃.

Equilibrium calculations (Fig. 4) predict reduction of Ce^IVO_2−x to Ce⁰ in the form of compounds CeAl₂, CeAl₃, Ce₃Al₁₁ and finally the liquid solution (Al,Ce)_(l). Synchrotron diffraction data showed no evidence of crystalline CeAl₂ or CeAl₃ and only Ce₃Al₁₁ was crystallizing from the liquid solution (Al,Ce)_(l). The large excess of Al_(l), rapid diffusion in the liquid, and highly negative mixing enthalpy²⁸ of Ce in Al_(l) facilitated rapid removal of cerium from the particle surface into the bulk of the melt.

Conclusions

The results demonstrate the aluminothermic reduction of cerium as a viable, potentially economical, and direct route for the production of Al–Ce alloys, avoiding the costly and energy intensive isolation of cerium metal and providing a high-value use for excess cerium oxide from mine tailings. Material generated in the direct reduction process has a large amount of Al₂O₃ particles embedded in the aluminum matrix expected to contribute to dispersion strengthening in the final alloy. Al–Ce master alloys produced by this process can be used in the Al casting industry after addition of pure Al and other alloying elements (e.g. Mg, Si) to tune the desired composition (below 3 at% Ce).⁸ The rapid cooling rates observed in typical casting operations can then produce the high-strength eutectic microstructures of Ce₃Al₁₁ particles in the Al matrix.

Time-resolved monitoring of the intermediate products revealed a complex multi-step reaction mechanism and clarified the importance of temperature control to maximize reaction yield while minimizing heating costs, which will be crucial for industrial implementation of this process.§

Author contributions

Conceptualization by A. A., D. W., R. T. O, and S. K. M. Methodology, validation by A. A. E. E. M and H. B. H. Investigation and experimentation by A. A., E. E. M, J. S., M. K., S. K., N. H. and A. L. Writing – original draft by A. A. and E. E. M. Writing – review and editing – A. A., E. E. M, H. B. H., J. S., M. K., S. K., N. H., A. L., R. T. O., D. W., and S. K. M. Visualization by A. A.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was sponsored by the Critical Materials Institute, an Energy Innovation Hub funded by the U.S. Department of Energy (DOE), Office of Energy Efficiency and Renewable Energy and Advanced Manufacturing Office. Work performed at Ames National Laboratory under contract DE-AC02-07CH11358, LLNL under contract DE-AC52-07NA27344. Beamline 12.2.2 at the Advanced Light Source is a DOE Office of Science User Facility under contract no. DE-AC02-05CH11231.

Notes and references

1. H. C. Örsted, K. Dan. Vidensk. Selsk., 1825.
2. F. Wöhler, Ann. Phys., 1828, 89, 577–582.
3. A. Holleman, Band 2 Nebengruppenelemente, Lanthanoide, Actinoide, Transactinoide, De Gruyter, 2016.
4. F. W. Hall, Ullmann's Encyclopedia of Industrial Chemistry, John Wiley & Sons, Ltd, 2000.
5. K. Binnemans, P. T. Jones, K. Van Acker, B. Blanpain, B. Mishra and D. Apelian, JOM, 2013, 65, 846–848.
6. Z. C. Sims, M. S. Kesler, H. B. Henderson, E. Castillo, T. Fishman, D. Weiss, P. Singleton, R. Eggert, S. K. McCall and O. Rios, J. Sustainable Metall., 2022, 8, 1225–1234.
7. D. J. Cordier, U.S. Geological Survey (2023) Minerals yearbook 2023, can be found under https://www.usgs.gov/centers/national-minerals-information-center/rare-earths-statistics-and-information, 2023.
8. Z. C. Sims, O. R. Rios, D. Weiss, P. E. A. Turchi, A. Perron, J. R. I. Lee, T. T. Li, J. A. Hammons, M. Bagge-Hansen, T. M. Willey, K. An, Y. Chen, A. H. King and S. K. McCall, Mater. Horiz., 2017, 4, 1070–1078.
9. H. B. Henderson, D. Weiss, Z. C. Sims, M. J. Thompson, E. E. Moore, A. Perron, F. Meng, R. T. Ott and O. Rios, in Light Metals, ed. A. Tomsett, Springer International Publishing, Cham, 2020, pp. 227–232.
10. R. T. Nguyen, D. D. Imholte, O. R. Rios, D. Weiss, Z. Sims, E. Stromme and S. K. McCall, Resour., Conserv. Recycl., 2019, 144, 340–349.
11. T. Wu, A. Plotkowski, A. Shyam and D. C. Dunand, Mater. Sci. Eng., 2022, 833, 142551.
12. L. Chen, J. Yang, Y. Yang, Y. Zhang and Z. Wang, Mater. Today Commun., 2022, 31, 103714.
13. K. Sheybani, M. H. Paydar and M. H. Shariat, Int. J. Refract. Met. Hard Mater., 2019, 82, 245–254.
14. K. Sheybani, M. H. Paydar and M. H. Shariat, Trans. Indian Inst. Met., 2020, 73, 2875–2888.
15. Y. Zhang, X. Hu, F. Liu, J. Yang, L. Chen, W. Tao, A. Liu, Z. Shi and Z. Wang, J. Alloys Compd., 2023, 945, 169252.
16. J. Yang, M. Kuwabara, Z. Liu, T. Asano and M. Sano, Tetsu-to-Hagane, 2006, 92, 239–245.
17. K. Yoshitaka, J. Yang, Z. Liu and M. Kuwabara, J. High Temp. Soc., 2009, 34, 20–25.
18. A. Maleki, M. Panjepour, B. Niroumand and M. Meratian, J. Mater Sci., 2010, 45, 5574–5580.
19. A. Maleki, N. Hosseini and B. Niroumand, Ceram. Int., 2018, 44, 10–23.
20. J. S. Luna A, A. Flores V, R. Muñiz V, A. F. Fuentes, J. Torres, N. Rodríguez R, J. C. Ortiz and P. Orozco, J. Rare Earths, 2011, 29, 74–76.
21. H. Wang, G. Li, Y. Zhao and G. Chen, Mater. Sci. Eng., 2010, 527, 2881–2885.
22. S. Vyazovkin, Isoconversional Kinetics of Thermally Stimulated Processes, Springer International Publishing, Cham, 2015.
23. H. S. Ray and S. Ray, Kinetics of Metallurgical Processes, Springer, Singapore, 2018.
24. H. L. Friedman, J. Polym. Sci., Part C: Polym. Symp., 1964, 6, 183–195.
25. S. Vyazovkin and C. A. Wight, Int. Rev. Phys. Chem., 1998, 17, 407–433.
26. D. J. M. Bevan, J. Inorg. Nucl. Chem., 1955, 1, 49–59.
27. M. Zinkevich, D. Djurovic and F. Aldinger, Solid State Ionics, 2006, 177, 989–1001.
28. M. I. Ivanov, V. V. Berezutskii, M. A. Shevchenko, V. G. Kudin and V. S. Sudavtsova, Powder Metall. Met. Ceram., 2015, 54, 80–92.
29. S. Ackermann, J. R. Scheffe and A. Steinfeld, J. Phys. Chem. C, 2014, 118, 5216–5225.
30. N. V. Skorodumova, S. I. Simak, B. I. Lundqvist, I. A. Abrikosov and B. Johansson, Phys. Rev. Lett., 2002, 89, 166601.
31. H. Charlton, G. Baldinozzi and M. Patel, Front. Nucl. Eng., 2023, 1, 1096142.
32. M. Kunz, A. A. MacDowell, W. A. Caldwell, D. Cambie, R. S. Celestre, E. E. Domning, R. M. Duarte, A. E. Gleason, J. M. Glossinger, N. Kelez, D. W. Plate, T. Yu, J. M. Zaug, H. A. Padmore, R. Jeanloz, A. P. Alivisatos and S. M. Clark, J. Synchrotron Radiat., 2005, 12, 650–658.
33. A. Doran, L. Schlicker, C. M. Beavers, S. Bhat, M. F. Bekheet and A. Gurlo, Rev. Sci. Instrum., 2017, 88, 013903.
34. L. Schlicker, A. Doran, P. Schneppmüller, A. Gili, M. Czasny, S. Penner and A. Gurlo, Rev. Sci. Instrum., 2018, 89, 033904.
35. M. Basham, J. Filik, M. T. Wharmby, P. C. Y. Chang, B. El Kassaby, M. Gerring, J. Aishima, K. Levik, B. C. A. Pulford, I. Sikharulidze, D. Sneddon, M. Webber, S. S. Dhesi, F. Maccherozzi, O. Svensson, S. Brockhauser, G. Náray and A. W. Ashton, J. Synchrotron Radiat., 2015, 22, 853–858.
36. J. Filik, A. W. Ashton, P. C. Y. Chang, P. A. Chater, S. J. Day, M. Drakopoulos, M. W. Gerring, M. L. Hart, O. V. Magdysyuk, S. Michalik, A. Smith, C. C. Tang, N. J. Terrill, M. T. Wharmby and H. Wilhelm, J. Appl. Cryst., 2017, 50, 959–966.
37. B. H. Toby and R. B. Von Dreele, J. Appl. Cryst., 2013, 46, 544–549.
38. L. Kaufman and H. Bernstein, Computer Calculation of Phase Diagrams With Special Reference to Refractory Metals, Academic Press Inc., United States, 1970.
39. N. Saunders and A. P. Miodownik, CALPHAD (Calculation of Phase Diagrams): A Comprehensive Guide, Elsevier, 1998.
40. H. Lukas, S. G. Fries and B. Sundman, Computational Thermodynamics: The Calphad Method, Cambridge University Press, Cambridge, 2007.
41. M. C. Gao, N. Ünlü, G. J. Shiflet, M. Mihalkovic and M. Widom, Metall. Mater. Trans. A, 2005, 36, 3269–3279.
42. H. Mao, M. Selleby and O. Fabrichnaya, Calphad, 2008, 32, 399–412.
43. C. Guéneau, N. Dupin, L. Kjellqvist, E. Geiger, M. Kurata, S. Gossé, E. Corcoran, A. Quaini, R. Hania, A. L. Smith, M. H. A. Piro, T. Besmann, P. E. A. Turchi, J. C. Dumas, M. J. Welland, T. Ogata, B. O. Lee, J. R. Kennedy, C. Adkins, M. Bankhead and D. Costa, Calphad, 2021, 72, 102212.
44. W. T. Fu and D. J. W. Ijdo, J. Solid State Chem., 2006, 179, 2732–2738.
45. G. Grimvall, Thermophysical Properties of Materials, North-Holland, 1986.
46. O. Kubaschewski, Materials Thermochemistry, Pergamon Press, New York, 1993.

---

Footnotes

† Electronic supplementary information (ESI) available. Experimental methods as well as details on diffraction experiments, metallography, and DSC. The authors cite additional ref. 32–46. See DOI: https://doi.org/10.1039/d4mh00087k

‡ Significant homogeneity ranges were reported for the oxides CeO_2−x and Ce₃O_5+x.^26,27

§ Please refer to the ESI† for the Methods section, which includes details of sample preparation, DSC, PXRD, and electron micrograph and spectroscopy data collection, Rietveld refinement, and equilibrium calculations.^32–46

---