WorksPhysics and Astronomy
Ultrahigh-Entropy Compositionally Complex Ceramics: Fluorite-Pyrochlore Phase Stability and Order-Disorder Transitions
Abstract
Three groups of 32 ultrahigh-entropy compositionally complex ceramics (CCCs) containing 16-19 components were synthesized and characterized. The first group of 15 CCCs forms single ultrahigh-entropy fluorite or pyrochlore phases in 12 compositions, while two compositions exhibit fluorite-fluorite dual phases, observed for the first time. Compared with ternary A2B2O7 pyrochlore, these ultrahigh-entropy CCCs are more prone to ordering (pyrochlore formation), surprisingly even more at higher size disorder (more lattice distortion), with suppressed dual-phase formation relative to five-component high-entropy ceramics. Two new 19-component series, denoted as "19CCC-P|Fy" and "19CCC-P|Fz", form single ultrahigh-entropy fluorite or pyrochlore phases. The 19CCC-P|Fy series, maintaining a 2:2 ratio of 3+ and 4+ cations, exhibits a pyrochlore-to-fluorite order-disorder transition (ODT) at y ~ 0.84, consistent with the linear projection of the pyrochlore superstructure peak intensity as an order parameter. In contrast, the 19CCC-P|Fz series with non-2:2 ratios of 3+ and 4+ cations show an abrupt ODT at z ~ 0.41-0.44, despite a projected transition at z ~ 0.64 from the order parameter.
Introduction
High-entropy ceramics (HECs), commonly defined as multicomponent ceramic materials with an ideal configurational mixing entropy of per cation on at least one sublattice, have attracted considerable research attention since 2015 [1, 2]. In 2020, Luo and co-workers extended the concept of HECs to encompass a broader class of compositionally complex ceramics (CCCs), ranging from medium-entropy ceramics (MECs, per cation) to ultrahigh-entropy ceramics (UECs, ) [1, 3]. Within this expanded framework, non-equimolar design as well as long- and short-range cation ordering has been employed to tailor material properties. In some cases, medium-entropy compositions have been shown to outperform their higher-entropy counterparts, suggesting that maximizing configurational entropy alone is not necessarily the most effective strategy for further optimizing material properties. Nevertheless, UECs offer a vast compositional design space, within which new phase-stability and phase-transition phenomena have been observed in several studies, as discussed later. These emerging physical phenomena may provide new opportunities for discovering and designing CCCs with previously unexplored structures and properties. In addition to single-phase CCCs, recent studies have demonstrated that thermal properties can be tuned by tailoring phase fractions in dual-phase CCCs, leading to the proposal of a corresponding dual-phase CCC design strategy [4, 5].
Fluorite-based oxides demonstrate ultralow thermal conductivities () with significant applications in thermal barrier coatings (TBCs) [6, 7, 8]. Fluorite-based oxides encompass a broad family of crystal structures derived from the cubic fluorite phase ( or disordered defect fluorite , , No. 225). Among these, the ordered cubic pyrochlore phase (, , No. 227) has attracted the most attention. It can be viewed as a superstructure of the disordered defect fluorite basis structure, and the pyrochlore and fluorite structures are directly connected to each other through an order-disorder transition (ODT) of the cations and anions. In addition, another ordered derivative structure, the orthorhombic weberite phase (of superstructure), can also form in certain compositions.
Considerable efforts have been devoted to developing descriptors for predicting phase formation in fluorite-based oxides. The cation-radius ratio ratio has proven particularly effective. For ternary zirconates and hafnates, a critical cation-radius ratio of defines the transition threshold between fluorite and pyrochlore structures. When is below this threshold, the rare-earth cations and Zr or Hf share a single cation site, stabilizing the disordered fluorite structure. In contrast, when exceeds 1.46, the ordered pyrochlore structure is favored, with larger cations occupying the A sublattice and smaller cations occupying the B sublattice [9, 10].
The literature generally agrees that the averaged cation-radius ratio criterion remains broadly effective for distinguishing fluorite and pyrochlore phases in high-entropy CCCs [11, 12]. Another descriptor, the modified size disorder , originally introduced to predict reduced thermal conductivity in high-entropy pyrochlore CCCs, has also been introduced to forecast the phase formation in high-entropy fluorite-based CCCs [13]. The parameter was modified to account for the two distinct cation sublattices in the pyrochlore structure and is denoted by the subscript “P” to distinguish it from the size disorder parameter for the fluorite structure , which contains only a single cation sublattice and will be discussed later. Moreover, an additional fluorite-pyrochlore dual-phase region was observed when , % [11, 12, 14].
However, several questions remain. First, previous studies using the modified size disorder implicitly assume a pyrochlore structure, thereby providing only a partial view of phase stability. In contrast, the size-disorder parameter provides a more physically reasonable descriptor for predicting phase stability, because it does not presuppose cation sublattice splitting. Instead, it captures the onset of sublattice splitting when the degree of cation disorder or size mismatch exceeds a critical threshold. Second, it remains unclear whether these descriptors are still valid as the compositional complexity increases. Some studies have found that phase stability rules based on the averaged cation-radius ratio are already shifted in UECs and non-equimolar systems [15, 16, 17]. Moreover, UECs have been reported to exhibit a surprisingly strong capability for single-phase formation in some cases [17]. Whether this behavior is broadly applicable across other UEC systems, however, remains to be established.
Extensive research has explored the use of the pyrochlore–fluorite ODTs to tune thermal conductivities in CCCs, covering systems from simple ternaries to complex UECs. The notable drop in thermal conductivity observed during ODTs has been attributed to the formation of short-range ordered nanodomains [15, 16]. These studies have typically designed compositional series in either stoichiometric () or non-stoichiometric (). Recently, one study systematically investigated the ODT in a non-stoichiometric fluorite–pyrochlore–weberite series, series (). Linear projection of XRD relative peak intensities vs. compositional variables was employed to predict the critical point of phase transitions, revealing abrupt fluorite–pyrochlore and pyrochlore-weberite transitions accompanied by discontinuous changes in structural order parameters [17].
Despite these advances, a fundamental question remains unaddressed: How does non-stoichiometry influence the nature of the ODT in CCCs?
In this study, we systematically investigated phase stability and ODTs in ultrahigh-entropy CCCs by synthesizing and characterizing three groups of 32 compositions containing 16-19 constituent cation components. The first group comprises 15 ultrahigh-entropy CCCs, , equimolar mixture of 15 rare-earth elements; , , , , or their equimolar combinations), which predominantly form single fluorite and pyrochlore phases, along with two fluorite–fluorite dual phases observed here for the first time. This group exhibits an unexpected propensity for cation ordering compared with conventional ternary and five-component high-entropy systems reported in the literature, indicating a fundamentally altered phase stability landscape.
Two additional 19-component series further clarify the role of stoichiometry in governing pyrochlore–fluorite ODTs: a stoichiometric series, (; denoted as “19CCC-P|F”), and a non-stoichiometric series, (; denoted as “19CCC-P|F”). While both series’ critical points were predicted using linear projections, the stoichiometric series conforms to the prediction, exhibiting a continuous ODT, whereas the non-stoichiometric series deviates markedly, displaying an abrupt transition.
Together, these results establish a systematic understanding of phase stability and ODTs behavior in ultrahigh-entropy CCCs and highlight stoichiometry as a critical factor.
Experimental procedure
Composition design
We designed and fabricated 15 ultrahigh-entropy CCCs containing 16-19 components, ; , , , , or their equimolar combinations). In addition, we designed and fabricated two new 19-component series with nominal compositions:
Stoichiometric series “19CCC-P|F”: ().
Non-stoichiometric series “19CCC-P|F”: .
These series were derived by mixing the single-phase pyrochlore endmembers , with, a -fraction of the single-phase fluorite endmembers , and a -fraction of single-phase fluorite endmembers , respectively.
The independent compositional variables and are introduced to distinguish them from the previously defined compositional variable in a non-stoichiometric 20CCFBONb/Ta series , where denotes the fraction of , and to facilitate direct comparison and discussion between the studies [17].
The compositions were tuned as follows:
For the stoichiometric series “19CCC-P|F”, .
For the non-stoichiometric series “19CCC-P|F”, , where corresponds to .
Materials and synthesis
We first prepared three mixtures of , , and using commercially available high-purity binary oxide powders (LaO, PrO, NdO, SmO, EuO, GdO, TbO, DyO, YO, HoO, ErO, TmO, YbO, LuO, ScO, CeO, ZrO, HfO, and TiO; US Research Nanomaterials, TX, USA) with a particle size of µm and a purity . For each mixture, 10 g of powders were weighed in stoichiometric proportions and loaded into a 100 mL YSZ ball mill jar with a ball-to-powder ratio of 10:1. An appropriate amount of isopropyl alcohol (IPA) was added to form slurries. The powders were homogenized in a planetary ball mill (PQN04 gear-drive mill, Across International LLC, USA) at a rotation speed of 300 rpm for 24 h, followed by drying in an oven at 75 °C overnight.
To fabricate 15 ultrahigh-entropy CCCs and two new series CCCs 19CCC-P|F and 19CCC-P|F, the milled precursor mixtures were either weighed with CeO, ZrO, HfO, and TiO powders, or combined in 2 g batches according to the designed stoichiometry. Each batch were load into a poly (methyl methacrylate) high-energy ball mill (HEBM) vial with two tungsten carbide (WC) inserts and a Ø 5/16” ( mm) WC ball. Approximately 1 wt % stearic acid was added as a process control agent. The vials were then placed in a SPEX 8000D mill (SPEXCertPrep, NJ, USA) and HEBM for 100 min.
The as-milled powders were uniaxially pressed into green pellets under MPa using a hydraulic press. The green pellets were placed on a Pt foil in an alumina crucible and sintered in a high-temperature box furnace (SentroTech, OH, USA) at 1600 °C for 24 h with a ramping rate of 5 °C/min, followed by furnace cooling. The sintered pellets were subsequently grinded and polished for further characterization.
Characterization
X-ray diffraction (XRD) patterns were acquired using a Rigaku GIXRD diffractometer operating at 40 kV and 44 mA, in a continuous scan mode with Cu K radiation. The full XRD patterns were obtained with a step size of 0.02 ° and a scan speed of 2.3 °/min. Slow scans were performed to resolve characteristic peaks at specific degree ranges in a step scan mode, using a step size of 0.005 ° and a scan speed of 0.2 °/min. For compositions near the transition points, an even slower scan speed of 0.1 °/min was employed. The lattice parameters were determined via unit cell refinements using JADE software.
A Raman microscope (Renishaw inVia, England, UK) was used to further characterize ODTs from the selected polished specimens of the 19CCC-P|F and 19CCC-P|F series. A 633 nm He–Ne laser was employed with an estimated focused spot size of µm using a 20× objective.
Microstructural and compositional analyses were performed using scanning electron microscopy (SEM) and energy-dispersive X-ray spectroscopy (EDS) on a ThermoFisher (formerly FEI) Scios Dual Beam FIB/SEM equipped with an EDX detector at 20 kV.
Results and discussion
Formation of ultrahigh-entropy CCCs
The designed compositions and phases; the averaged cation-radius ratio ; size-disorder parameters ( and , with also listed for comparison with ); measured and relative densities; and the ideal configurational entropies of mixing per cation () for the first group of 15 ultrahigh-entropy CCCs with nominal formula were characterized, calculated, and summarized in Table 1. Phase formation was determined by XRD (Figure 1), revealing five single-phase fluorite compositions, seven single-phase pyrochlore compositions, and three dual- or multiphase compositions. Two fluorite–fluorite dual-phase compositions, and (denoted as “F+F”), were further examined using slow step-scan XRD in the range of 32–38°. The resulting patterns (Figure 1(d)) clearly reveal the coexistence of two distinct fluorite (200) peaks, with no detectable reflection near , where the characteristic pyrochlore (331) superstructure peak would normally appear. To our knowledge, this is the first report of fluorite-fluorite high-entropy dual phases.
Table 1. Designed compositions of the first group of 15 ultrahigh-entropy CCCs with the nominal formula , where denotes an equimolar mixture of 15 rare-earth elements (Sc, Y, La, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu), and TM = Ce, Zr, Hf, Ti, or their equimolar combinations. The table summarizes the observed phase, the averaged cation-radius ratio , and the size-disorder parameters for all 15 compositions. Here, “F” denotes a single-phase fluorite structure, “P” denotes a pyrochlore phase, “F+F” indicates dual-fluorite phases, and “F+P+” represents a three-phase mixture. Two size-disorder parameters were calculated for each composition (regardless the actual stable phase) for comparison: was calculated assuming a disordered fluorite structure, while was calculated assuming an ordered pyrochlore structure (following the definition of Wright et al. [18]). The ideal configurational entropies of mixing per cation are also calculated for all single-phase specimens based on the actual stable phases observed.
| Composition | Phase | (%) | (%) | (%) | Measured Density (g/cm) | Relative Density (%) | (k/cation) Fluorite | Pyrochlore A-site | Pyrochlore B-Site | Pyrochlore Mean | |
| F | 1.19 | 5.78 | 6.56 | 4.64 | 6.98 | 94.6 | 2.05 | ||||
| F | 1.44 | 11.59 | 6.56 | 4.64 | 6.75 | 96.7 | 2.05 | ||||
| F + F | 1.46 | 12.13 | 6.56 | 4.64 | 8.68 | - | |||||
| P | 1.71 | 17.47 | 6.56 | 4.64 | 5.96 | 93.9 | 2.71 | 0 | 1.35 | ||
| F | 1.30 | 9.58 | 11.49 | 8.12 | 6.81 | 94.9 | 2.05 | ||||
| F + F | 1.45 | 11.87 | 6.59 | 4.66 | 7.65 | - | |||||
| F | 1.31 | 9.97 | 12.06 | 8.53 | 7.71 | 94.6 | 2.05 | ||||
| F + P + | 1.40 | 13.80 | 19.13 | 13.52 | 6.68 | - | |||||
| P | 1.56 | 14.93 | 10.88 | 7.69 | 6.45 | 98.2 | 2.71 | 0.69 | 1.70 | ||
| P | 1.57 | 15.12 | 10.33 | 7.31 | 7.49 | 97.9 | 2.71 | 0.69 | 1.70 | ||
| F | 1.35 | 10.62 | 11.58 | 8.19 | 7.42 | 95.3 | 2.60 | ||||
| P | 1.41 | 13.12 | 16.21 | 11.47 | 6.75 | 99.4 | 2.71 | 1.10 | 1.90 | ||
| P | 1.42 | 13.29 | 16.33 | 11.55 | 7.38 | 97.6 | 2.71 | 1.10 | 1.90 | ||
| P | 1.53 | 14.07 | 10.09 | 7.13 | 7.20 | 98.2 | 2.71 | 1.10 | 1.90 | ||
| P | 1.43 | 12.89 | 14.56 | 10.30 | 7.18 | 98.2 | 2.71 | 1.39 | 2.05 |

Figure 1. XRD patterns of the first group of 15 ultrahigh-entropy CCCs. The diffraction patterns are shown separately for (a) five single-phase fluorite compositions, (b) seven pyrochlore compositions, and (c) three dual- or multiphase compositions. (d) Slow step-scan XRD patterns in the range of – for the two fluorite–fluorite dual-phase compositions, and (denoted as “F+F”), clearly showing the coexistence of two distinct fluorite peaks and the absence of any detectable reflection near , which corresponds to the characteristic pyrochlore superstructure peak.
Figure S1 shows representative cross-sectional SEM backscattered electron (BSE) images of multiphase specimen and the fluorite-fluorite dual-phase and specimens. Distinct dark and bright contrasts are observed in the BSE images, where the dark regions correspond to phases enriched in lighter (lower-) elements, while the bright regions correspond to phases enriched in heavier (higher-) elements. The bright and dark regions are uniformly distributed throughout the specimens.
The EDS elemental mappings (Figure S2-S4) were conducted on the three dual-phase and multiphase specimens to examine elemental distributions within each phase. For the fluorite-fluorite dual-phase specimens and , the EDS elemental maps (Figure S2(a) and S3(a)) indicates that each fluorite phase is compositionally homogeneous. No significant compositional differences between the two fluorite phases are detected within the noise level of EDS measurements. This conclusion is further supported by EDS point analysis acquired from the dark and bright regions (Figure S2(b) and S3(b)), which shows that the average compositions of the two fluorite phases are essentially identical within measurement errors.
For the multiphase specimen (Figure S4), the EDS elemental maps reveal clear segregation, with Ce, Sm, Nd, and Ti enriched in regions of distinct contrast. However, unambiguous identification of the three constituent phases (fluorite, pyrochlore, and rhombohedral) in this composition based solely on SEM BSE images and EDS elemental maps is challenging and is not the focus of this work.
Figure 2 (a, Figure 2b) presents plots of the averaged cation-radius ratio vs. the modified size disorder , assuming an pyrochlore structure, and vs. size disorder , assuming a disordered fluorite structure, for two groups of stoichiometric ultrahigh-entropy CCCs (A:B = 2:2): (1) 15 ultrahigh-entropy CCCs, and (2) the 19CCC-PF series, (), which is further discussed in § 3.2.

Figure 2. Phase stability of the first group of 15 ultrahigh-entropy CCCs and the second 19CCC-P|F series, (), both with the ideal 2:2 ratio of A-type (larger 3+ rare-earth cations) and B-type (smaller 4+ cations), plotted in (a) vs. and (b) vs. spaces. The typical fluorite–pyrochlore transition threshold () for ternary systems is indicated by the black lines. The gray-shaded region in (a) represents the fluorite–pyrochlore dual-phase stability range (, ) previously proposed for high-entropy compositions [refs]. The vertical dashed line in (b) represents a threshold used as an alternative criterion for the fluorite–pyrochlore transition.
The modified size disorder follows a previously established definition based on the pyrochlore structure, where larger 3+ rare-earth cations occupy the eightfold-coordinated A sites and smaller 4+ cations occupy the sixfold-coordinated B sites, in the absence of anti-site defects [18].
To provide a more comprehensive evaluation of phase stability, a new parameter, size disorder , is introduced:
This parameter assumes a disordered fluorite structure in which all cations are randomly mixed on a single lattice, and both 3+ and 4+ cations adopt their eightfold coordination radii.
As shown in Figure 2 (a,Figure 2b) blue circular symbols represent single-phase fluorites; half–dark-blue and half–light-blue circular symbols represent dual-phase fluorites; red square symbols represent single-phase pyrochlores, and star-shaped symbols represent multiphase composition. In Figure 2(a), the typical fluorite–pyrochlore transition threshold () for ternary systems are indicated by the black lines. The gray-shaded region marks the previously proposed fluorite–pyrochlore dual-phase stability window for high-entropy compositions (, ) [11, 12, 14].
As shown in Figure 2(a), at smaller value of , the fluorite phase remain stable until , as evidenced by the fluorite-fluorite dual phase specimens , and . However, at larger values (), an ultrahigh-entropy pyrochlore phase emerges at (corresponding to the specimen in the 19CCC-P|F series). These observations indicate a stronger ordering tendency in ultrahigh-entropy CCCs than in ternary and high-entropy CCCs.
Shifted ODT thresholds have been reported across different ultrahigh-entropy fluorite-based CCCs series: an 12-component CCCs series , ) exhibits an ODT at a higher threshold of , while a 10-component CCCs series , ) undergoes ODT at a much lower threshold of . More recently, a 20-component CCC series , ) was found to undergo ODT at . These observations collectively suggest that the averaged cation-radius ratio is no longer a fully reliable phase-stability criterion in ultrahigh-entropy systems.
When increases further, a multi-phase (fluorite, pyrochlore, and rhombohedral phase) forms in the specimen at and . This is not unexpected, as greater size mismatches correspond to increased lattice strain, which can promote phase separation.
However, somewhat surprisingly, although the value in these ultrahigh-entropy CCCs are generally larger than those reported for high-entropy CCCs [11, 12, 14], the formation of fluorite-pyrochlore dual phases is noticeably suppressed within the conventional dual-phase stability window (, ).
When the averaged cation-radius ratio is plotted against the size disorder (Figure 2(b)), it is evident that provides a more direct and reliable criterion, than either or , for defining the fluorite-pyrochlore transition threshold in ultrahigh-entropy CCCs. The fluorite-to-pyrochlore transition occurs at (indicated by the vertical dashed line), with fluorite phases remaining stable below this value and pyrochlore or multiphase phases stabilizing above it.
Pyrochlore-fluorite ODT
Pyrochlore-fluorite ODT in stoichiometric vs. non-stoichiometric CCCs series
Two new 19-component CCCs series, 19CCC-P|F (, ) and 19CCC-P|F (, ), were designed and synthesized to investigate the influence of stoichiometry (maintaining or deviating from an ideal 2:2 A/B cation ratio) on phase stability and ODTs in ultrahigh-entropy CCCs. Figure 3 schematically illustrates the compositions investigated in both series. Table 2 and Table 3 summarize the designed compositions, phase, the averaged cation-radius ratio , size-disorder parameters (, , and ), lattice parameter, measured densities, relative densities, and ideal configurational entropies of mixing per cation () for all compositions in the two new CCCs series.
Table 2. Phases, averaged cation-radius ratios (), two size-disorder parameters ( and ), lattice parameters, measured and relative densities, and calculated ideal configurational entropies of mixing () for nine specimens in the second series of 19-component CCCs exhibiting ultrahigh-entropy fluorite or pyrochlore phases (denoted as “19CCC-P|F”), , where represents the fraction of the disordered defect-fluorite endmember, . This chemical formula can also be expressed in simplified form as . For the ordered pyrochlore phase, represents the fraction of the (smallest) ions occupying the B site in the structure, and a lower Ti fraction (i.e., larger ) decreases the ratio, thereby promoting disorder.
| Phase | (%) | (%) | (%) | or (Å) | Measured Density (g/cm) | Relative Density (%) | Fluorite | Pyrochlore A-site | Pyrochlore B-site | Pyrochlore Mean | ||
| 0 | P | 1.711 | 17.47 | 6.56 | 4.64 | 5.0757 ± 0.0002 | 5.96 | 93.9 | 2.71 | 0 | 1.35 | |
| 0.375 | P | 1.555 | 15.50 | 15.05 | 10.64 | 5.1516 ± 0.0008 | 6.26 | 91.1 | 2.71 | 1.07 | 1.89 | |
| 0.5625 | P | 1.487 | 14.29 | 15.43 | 10.91 | 5.2062 ± 0.0025 | 7.04 | 99.9 | 2.71 | 1.30 | 2.01 | |
| 0.75 | P | 1.425 | 12.89 | 14.56 | 10.30 | 5.2196 ± 0.0008 | 7.18 | 98.2 | 2.71 | 1.39 | 2.05 | |
| 0.8125 | P | 1.405 | 12.37 | 14.02 | 9.92 | 5.2507 ± 0.0010 | 7.31 | 97.0 | 2.71 | 1.38 | 2.04 | |
| 0.8281 | P | 1.401 | 12.24 | 13.87 | 9.81 | 5.2227 ± 0.0077 | 7.20 | 95.6 | 2.71 | 1.37 | 2.04 | |
| 0.8438 | F | 1.396 | 12.10 | 13.71 | 9.69 | 5.2548 ± 0.0004 | 7.29 | 95.8 | 2.73 | |||
| 0.875 | F | 1.386 | 11.83 | 13.36 | 9.44 | 5.2615 ± 0.0004 | 7.20 | 94.1 | 2.72 | |||
| 1 | F | 1.350 | 10.62 | 11.58 | 8.19 | 5.2425 ± 0.0004 | 7.42 | 95.3 | 2.60 |
Table 2.
| Phase | (%) | (%) | (%) | or (Å) | Measured Density (g/cm³) | Relative Density (%) | Fluorite | Pyrochlore A-site | Pyrochlore B-Site | Pyrochlore Mean | ||
| 0 | P | 1.425 | 12.89 | 14.56 | 10.30 | 5.2196 ± 0.0008 | 7.18 | 98.2 | 2.71 | 1.39 | 2.05 | |
| 0.125 | P | 1.422 | 12.62 | 13.85 | 9.79 | 5.2374 ± 0.0003 | 7.38 | 100.0 | 2.77 | 1.38 | 2.08 | |
| 0.25 | P | 1.419 | 12.33 | 13.07 | 9.25 | 5.2412 ± 0.0011 | 7.28 | 98.4 | 2.75 | 1.35 | 2.05 | |
| 0.3125 | P | 1.418 | 12.17 | 12.66 | 8.95 | 5.2399 ± 0.0005 | 7.40 | 99.7 | 2.72 | 1.34 | 2.03 | |
| 0.375 | P | 1.417 | 12.02 | 12.23 | 8.65 | 5.2412 ± 0.0009 | 6.96 | 93.5 | 2.68 | 1.31 | 2.00 | |
| 0.4062 | P | 1.416 | 11.94 | 12.01 | 8.49 | 5.2404 ± 0.0013 | 7.29 | 97.6 | 2.66 | 1.30 | 1.98 | |
| 0.4375 | F | 1.415 | 11.86 | 11.78 | 8.33 | 5.2377 ± 0.0003 | 7.27 | 94.7 | 2.54 | |||
| 0.4688 | F | 1.414 | 11.77 | 11.54 | 8.16 | 5.2395 ± 0.0002 | 7.41 | 96.5 | 2.52 | |||
| 0.5 | F | 1.414 | 11.69 | 11.30 | 7.99 | 5.2392 ± 0.0023 | 6.51 | 84.7 | 2.50 | |||
| 0.75 | F | 1.408 | 10.97 | 9.05 | 6.40 | 5.2434 ± 0.0002 | 6.55 | 84.2 | 2.32 | |||
| 1 | F | 1.402 | 10.15 | 5.82 | 4.12 | 5.2425 ± 0.0004 | 7.32 | 92.2 | 2.06 |
Table 3. Table 3. Phases, averaged cation-radius ratios (), two size-disorder parameters ( and ), lattice parameters, measured and relative densities, and calculated ideal configurational entropies of mixing () for 11 specimens in the third series of 19-component CCCs exhibiting ultrahigh-entropy fluorite or pyrochlore phases (denoted as “19CCC-P|F”), , where () represents the fraction of the disordered fluorite endmember, . This chemical formula can also be expressed in a simplified form as . For the ordered pyrochlore phase, also represents the fraction of the Ce (presumably ) ions occupying the A site in the structure, and a larger value of promotes disorder.

Figure 3. Schematic illustration of the pyrochlore-fluorite order-disorder transitions (ODTs) in the stoichiometric 19CCC-PF series, (), and the non-stoichiometric 19CCC-PF series, (), where or represents the fraction of the disordered fluorite endmember in each case. All 20 fabricated and characterized compositions exhibit single ultrahigh-entropy phases, with abrupt pyrochlore–fluorite ODTs occurring near in the 19CCC-PF series and in the 19CCC-PF series. Due to space constraints and for clarity of presentation, the 19CCC-PF series is not plotted to scale; the range from to has been compressed. Note that the composition in the 19CCC-PF series corresponds to in the 19CCC-PF series.
Full XRD patterns confirm that all compositions in both series form exclusively single-phase fluorite or pyrochlore structures (Supplementary Figures S5-S6), showing a strong capability for single-phase formation similar to what we observed in other ultrahigh-entropy systems [17, 19]. To examine the pyrochlore-fluorite ODTs in greater detail, slow XRD scans were performed over , encompassing the fluorite (200), the pyrochlore (400), and the characteristic pyrochlore (331) superstructure peaks (Figure 4(a) and Figure 5(a)). Enlarged views of the pyrochlore (331) superstructure peaks in Figure 4(b) and Figure 5(b) clearly shows that the (331) reflection progressively weakens and eventually disappears with increasing compositional variables and (i.e. increasing - and - fraction of fluorite endmembers). The pyrochlore-to-fluorite ODTs are identified at for 19CCC-P|F series and for 19CCC-P|F series.

Figure 4. (a) Evolution of XRD patterns showing the fluorite–pyrochlore transition in the 19CCC-P|F series, , where () represents the fraction of the disordered defect-fluorite endmember, which can also be expressed in a simplified form as . (b) Intensity plots showing the disappearance of the characteristic pyrochlore superstructure peak occurring between and . (c) Normalized intensity of the characteristic pyrochlore superstructure peak , relative to the peak, as a function of the compositional variable , revealing the fluorite–pyrochlore ODT consistent with a linear projection of the superstructure peak intensity.

Figure 5. (a) Evolution of XRD patterns showing the fluorite–pyrochlore transition in the 19CCC-P|F series, , where () represents the fraction of the disordered fluorite endmember, which can also be expressed in a simplified form as . (b) Intensity plots showing the disappearance of the characteristic pyrochlore superstructure peak between and . (c) Normalized intensity of the characteristic pyrochlore superstructure peak , relative to the peak, as a function of the compositional variable . A linear projection of the peak intensity suggests a pyrochlore-to-fluorite ODT at , while the actual ODT occurs more abruptly between and .
Figure 6 presents the full width at half maximum (FWHM) analysis for the fluorite (200), pyrochlore (440), and characteristic pyrochlore (331) peaks for the stoichiometric 19CCC-P|F (Figure 6(a, Figure 6c)) and non-stoichiometric 19CCC-P|F series (Figure 6(b, Figure 6d)). The results show that, as the compositions approach the ODT, the pyrochlore (331) peaks undergoes a sudden broadening, while the pyrochlore (400) peaks initially narrow and then gradually broaden as structural disorder increase with the compositional variable and , respectively. These observations suggest that, during the ODT, (331) planes experience increased lattice strain, while the (400) planes undergo a reduction in lattice strain. Overall, FWHM evolution provides complementary information on the phase evolution during the ODT process beyond that observed from XRD.

Figure 6. Evolution of the full width at half maximum (FWHM) of the pyrochlore and peaks and the fluorite peak as a function of the compositional variable for (a, c) the stoichiometric 19CCC-P|F series, (), and as a function of the compositional variable for (b, d) the non-stoichiometric 19CCC-P|F series, (). The gray vertical lines indicate the fluorite–pyrochlore ODTs.
Linear projections were performed using the pyrochlore superstructure peak intensity (331)/(440) as an order parameter to quantitatively estimate critical points of these ODTs (Figure 4(c) and Figure 5(c)). Six data points from each series were used: , , , , , and for the stoichiometric 19CCC-P|F series, and , , , , , and for the non-stoichiometric 19CCC-P|F series. The linear regression yielded the following relationships:
Stoichiometric 19CCC-P|F series: ()
Non-stoichiometric 19CCC-P|F series: ()
These results suggest a projected pyrochlore-fluorite ODT at for the stoichiometric 19CCC-P|F series, in good agreement with the experimentally observed ODT at . These demonstrate an excellent agreement between the predicted and measured transition compositions in the stoichiometric series.
For the non-stoichiometric 19CCC-PF series, the linear projection predicts an ODT at . However, the experimentally observed ODT occurs at , indicating a significantly more abrupt transition than projection, as evidenced by the rapid disappearance of the characteristic pyrochlore (331) superstructure peaks near ODT critical points (Figure 5(b); and ). A similarly abrupt ODT behavior has been reported in another non-stoichiometric 20CCFBO series (, ), where the linear regression of pyrochlore superstructure peak intensity vs. compositional variable yielded: , , suggesting a projected pyrochlore-fluorite ODT at . This projection is also considerably lower than the experimentally observed ODT at . Together, these results suggest an abrupt nature of the pyrochlore–fluorite ODTs in non-stoichiometric CCCs series.
Raman spectroscopy is widely recognized as an effective probe for distinguishing ordered, characterized by well-resolved vibrational bands, from disordered structures, which exhibit broadened spectra [20]. It therefore provides structural information complementary to XRD for investigating the ODTs in the 19CCC-PF and 19CCC-PF series. Figure 7 presents selected Raman spectra from both series, including the fluorite and pyrochlore endmembers ( and ), as well as compositions immediately before and after the pyrochlore-to-fluorite ODTs ( and ).

Figure 7. Raman spectra of selected compositions from two 19-component CCCs series exhibiting pyrochlore-fluorite ODTs: (a) the stoichiometric 19CCC-F|Py series, (), and (b) the non-stoichiometric 19CCC-F|Pz series, (). The selected compositions include the fluorite and pyrochlore endmembers (; ) and compositions immediately before and after the pyrochlore-fluorite ODTs (, ; , ).
As shown in Figure 7(a) for the 19CCC-PF series, the ultrahigh-entropy pyrochlore titanate endmember () exhibits prominent F, E, and A modes, confirming that this composition adopts a well-ordered pyrochlore structure. As the composition approaches the pyrochlore-to-fluorite ODT critical point, the Raman bands for progressively broaden but still retain features characteristic of the pyrochlore structure. Once the ODT is completed, composition exhibits a loss of spectral resolution, indicating the breakdown of long-range order. The fluorite endmember () likewise exhibits a broadened spectral profile.
In contrast, for the 19CCC-PF series (Figure 7(b)), the ultrahigh-entropy pyrochlore endmember () exhibits only weak features reminiscent of the pyrochlore structure, rather than well-resolved characteristic modes. This behavior is expected given the non-stoichiometric nature of the composition, which suppresses the formation of a well-crystallized pyrochlore structure. A similar trend of spectral broadening is observed with increasing , indicating progressive structural disorder.
Overall, the evolution from well-resolved spectra to broadened features across the pyrochlore-to-fluorite ODTs is consistent with the disordering behavior commonly reported for ternary pyrochlores in the literature. However, the Raman spectra does not provide additional quantitative information on the ODTs beyond that obtained from XRD.
Phase stability and descriptors
The phase stability was further evaluated through a combined analysis of multiple variables: the compositional variables and , the averaged cation-radius ratio , the size-disorder parameters and , and the non-stoichiometric parameter (Figure 8). The non-stoichiometric parameter describes compositions of the form , quantifying the deviation from the ideal A/B = 2:2 cation ratio. For the stoichiometric 19CCC-P|F series, , , while for the non-stoichiometric series 19CCC-P|F series, , .

Figure 8. Averaged cation-radius ratio vs. (a) compositional variable for the stoichiometric 19CCC-P|F series, (), (b) compositional variable for the non-stoichiometric 19CCC-F|P series, (). Panels (c) and (d) show vs. size disorder and the modified size disorder , respectively, for both 19CCC-P|F and 19CCC-F|P series. (e) vs. for both series. (f) vs. , where for the 19CCC-P|F series and for the 19CCC-F|P series.
As shown in Figure 8(a, Figure 8b), the pyrochlore-fluorite ODTs occur at averaged cation-radius ratios of for the stoichiometric 19CCC-P|F and for the non-stoichiometric 19CCC-P|F series (see black dashed lines; calculated by forcing an A/B = 2:2 ratio). Both values are notably lower than the established threshold for ternary zirconates and hafnates (, black solid line). Furthermore, the linear-projection-predicted critical composition for the pyrochlore-to-fluorite transition in 19CCC-P|F series occurs at , corresponding to , which is also significantly below the conventional threshold of . These indicate that ultrahigh-entropy CCCs exhibit a stronger tendency toward ordering compared to ternary ceramics. Comparison of the pyrochlore-fluorite ODT thresholds in terms of for the stoichiometric 19CCC-P|F series and non-stoichiometric 19CCC-P|F series indicates that the stoichiometric series is more prone to ordering. This is consistent with expectations, as an stoichiometric ratio promotes the preferential occupation of 3+ and 4+ cations on the A and B sites, respectively, in the typical pyrochlore structure.
The averaged cation-radius ratio vs. size disorder and are shown in Figure 8(c) and (d), respectively. Both sets of descriptors clearly define that the fluorite phases are stable near the lower-left corner, corresponding to small (favor cations mixing in a single lattice) coincide with small , and (corresponding to lower lattice strain and enhanced lattice mixing), while the pyrochlore phases are stable near the upper-right corner with larger values of these parameters. The dashed lines, which connect the ODT critical points in both series, delineate the boundary between the regions where the fluorite phase is stable and those where the pyrochlore phase is stable.
Figure 8(e) shows the modified size disorder vs. size disorder diagram. It indicates that the fluorite phases are stable in the region with smaller , whereas the pyrochlore phase is table in the region with larger , with no pronounced trend associated with increasing . This observation suggests that is the more effective descriptor for distinguishing the two phases. The dashed line delineates the phase stability boundary between these stability regions.
Figure 8(f) presents a diagram of the averaged cation-radius ratio vs. non-stoichiometry parameter . The parameter quantifies the degree of deviation from the ideal 2:2 A/B ratio and is defined for both series by the general composition . For the stoichiometric 19CCC-P|F series, , whereas for the non-stoichiometric 19CCC-P|F series, . As shown in the figure, it is evident that plays a more dominant role than in determining phase stability.
In the non-stoichiometric 19CCC-P|F series, the pyrochlore phases remains stable up to , corresponding to an ODT critical A/B ratio of 0.66 (A-site deficient). Beyond this threshold, progressive disordering occurs. The influence of non-stoichiometry on the ODT has been discussed in different CCCs systems. In ternary rare earth zirconates and hafnates , the ODT occurs when , corresponding to a critical A/B ratio of 0.82 - 1.22[21]. In a 11-component CCCs series , the ODT was reported at a critical A/B ratio between 1.16 and 1.20 [16]. In a recent study of a 20-component 20CCFBO series , the pyrochlore structure remained stable until [17]. These observations suggest that ultrahigh-entropy CCCs are better able to tolerate non-stoichiometry while maintaining an ordered structure compared with ternary CCCs.
We computed and plotted ideal configurational entropy of mixing vs. compositional variable and for stoichiometric 19CCC-P|F and non-stoichiometric 19CCC-P|F series, respectively, as shown in Figure 9. For the pyrochlore phase with two sublattices, we calculated and plotted the mixing entropy for each sublattice separately ( and ). We then plotted weighted average entropies for the pyrochlore phase using: . The resulting weighted average ideal configurational entropy of mixing is generally greater than per cation for both the stoichiometric 19CCC-P|F and non-stoichiometric 19CCC-P|F series, indicating that these compositions fall within the ultrahigh-entropy CCCs, even though short-range ordering (if present) may reduce the actual configurational entropies for fluorite composition. The weighted average configurational entropy reaches its maximum near ODT critical point in the stoichiometric 19CCC-P|F series, suggesting that the ODT is possible entropy-driven (primarily driven from the B-site configurational entropy ). In contrast, shows no clear correlation with phase formation or the ODT in the non-stoichiometric 19CCC-P|F series, consistent with our previous observations in another non-stoichiometric 20CCFBO series [17]. It is reasonable, as oxygen vacancies and anti-site defect in the non-stoichiometric 19CCC-P|F series complicate accurate estimation of the configurational entropy. In addition, the difference in configurational entropies of mixing during the ODT, , was calculated and is indicated by black arrows in the figure. For the stoichiometric 19CCC-P|F, per cation, whereas for the non-stoichiometric 19CCC-P|F series, per cation, indicating that the non-stoichiometric CCCs are more prone to disordering. The configurational entropy difference at the ODT in ternary CCCs is 0.693 pre cation. This value is close to what we observe in stoichiometric ultrahigh-entropy CCCs, suggesting an inherent tendency toward ordering in stoichiometric systems.

Figure 9. Computed ideal configurational entropy of mixing (in units of per cation) as a function of the compositional variables (a) for the 19CCC-P|F series, , and (b) for the 19CCC-P|F series, . For the pyrochlore phase with two sublattices, the solid data points and solid lines represent the average configurational entropy, whereas the hollow data points and dashed lines represent the configurational entropies of the individual sublattices.
The lattice parameters, normalized to that of the primitive cubic fluorite cell () for stoichiometric 19CCC-P|F and non-stoichiometric 19CCC-P|F series are shown in Figure 10. In both series, the normalized lattice parameters generally increase gradually with the compositional variables, with a more pronounced trend observed in the stoichiometric 19CCC-P|F series reflecting the larger lattice parameter difference between its endmembers.

Figure 10. Measured lattice parameters ( for the fluorite phase and for the pyrochlore phase) for (a) the 19CCC-P|F series () and (b) the 19CCC-P|F series (). The lattice parameters were obtained from unit-cell refinements of the XRD patterns and normalized to the primitive fluorite cell parameter () for direct comparison.
Conclusion
In this work, we systematically synthesized and investigated three groups of 32 ultrahigh-entropy compositionally complex ceramics (CCCs) containing 16–19 cations. In the first group of 15 CCCs, , we identified 12 single-phase ultrahigh-entropy fluorite or pyrochlore ceramics, along with two compositions exhibiting fluorite–fluorite dual phases—reported here for the first time. Compared with ternary (which undergo the fluorite– pyrochlore transition at ), these ultrahigh-entropy CCCs show a stronger tendency toward ordering (pyrochlore formation), even at higher size disorder levels where disordering would typically be favored. Among the structural descriptors examined, is found to be more effective for predicting phase formation in ultrahigh-entropy CCCs than either or .
Moreover, no fluorite–pyrochlore dual-phase CCCs were detected within the conventional stability window (, ) previously established for high-entropy CCCs, indicating a pronounced suppression of dual-phase formation in the ultrahigh-entropy regime. The pyrochlore-fluorite ODT was further investigated in two newly designed 16–19-component CCC series: the stoichiometric 19CCC-P|F and the non-stoichiometric 19CCC-P|F series, both of which form exclusively single-phase pyrochlore or fluorite structures across their entire compositional ranges.
Phase evolution and transition behaviors were systematically characterized through detailed analysis of XRD peak features. In the stoichiometric 19CCC-P|F series, the ODT occurs at a predictable composition consistent with the linear projection of the pyrochlore (331) superstructure peak intensity used as an order parameter. In contrast, the non-stoichiometric 19CCC-P|F series exhibits a significantly more abrupt transition. The effect of non-stoichiometry on phase stability and the ODT was evaluated through a combined analysis of multiple structural descriptors, including , , and .
Both the stoichiometric 19CCC-P|F () and the non-stoichiometric 19CCC-P|F series () exhibit a stronger propensity for ordering than ternary CCCs, which undergo the fluorite–pyrochlore transition at . This tendency is particularly pronounced in the stoichiometric 19CCC-P|F series, which is consistent with the fact that half of the 3+ cations naturally occupy the A site and half of the 4+ cations occupy the B site in the ordered pyrochlore structure.
Acknowledgment: The work is supported by the U.S. National Science Foundation (NSF), primarily through NSF Grant No. DMR-2026193 from July 2020 to March 2025, and we also thank partial support from NSF Future Manufacturing Research Grant No. CBET-2328044 after March 2025 to complete this study related to the general topic of ceramic processing and manufacturing.
References
- [1]A.J. Wright, Q. Wang, C. Huang, A. Nieto, R. Chen, J. Luo, From high-entropy ceramics to compositionally-complex ceramics: A case study of fluorite oxides, J. Eur. Ceram. Soc. 40 (2020) 2120–2129. https://doi.org/10.1016/j.jeurceramsoc.2020.01.015.
- [2]C.M. Rost, E. Sachet, T. Borman, A. Moballegh, E.C. Dickey, D. Hou, J.L. Jones, S. Curtarolo, J.P. Maria, Entropy-stabilized oxides, Nat. Commun. 6 (2015). https://doi.org/10.1038/ncomms9485.
- [3]A.J. Wright, J. Luo, A step forward from high-entropy ceramics to compositionally complex ceramics: a new perspective, J. Mater. Sci. 55 (2020) 9812–9827. https://doi.org/10.1007/s10853-020-04583-w.
- [4]M. Qin, J. Gild, C. Hu, H. Wang, M.S. Bin Hoque, J.L. Braun, T.J. Harrington, P.E. Hopkins, K.S. Vecchio, J. Luo, Dual-phase high-entropy ultra-high temperature ceramics, J. Eur. Ceram. Soc. 40 (2020) 5037–5050. https://doi.org/10.1016/j.jeurceramsoc.2020.05.040.
- [5]K. Song, J. Luo, Fluorite-pyrochlore dual-phase compositionally complex ceramics, J. Eur. Ceram. Soc. 46 (2026). https://doi.org/10.1016/j.jeurceramsoc.2025.117972.
- [6]J. Yang, W. Pan, Y. Han, M. Zhao, M. Huang, C. Wan, Mechanical properties, oxygen barrier property, and chemical stability of RE3NbO7 for thermal barrier coating, Journal of the American Ceramic Society 103 (2020) 2302–2308. https://doi.org/10.1111/jace.16952.
- [7]Q. Xu, W. Pan, J. Wang, C. Wan, L. Qi, H. Miao, K. Mori, T. Torigoe, Rare-earth zirconate ceramics with fluorite structure for thermal barrier coatings, in: Journal of the American Ceramic Society, 2006: pp. 340–342. https://doi.org/10.1111/j.1551-2916.2005.00667.x.
- [8]X.Q. Cao, R. Vassen, D. Stӧver, Ceramic materials for thermal barrier coatings, J. Eur. Ceram. Soc. 24 (2004) 1–10. https://doi.org/10.1016/S0955-2219(03)00129-8.
- [9]E.C. O’Quinn, D.L. Drey, M.K. Lang, Defining the Structural Stability Field of Disordered Fluorite Oxides, Front. Chem. 9 (2021). https://doi.org/10.3389/fchem.2021.733718.
- [10]C.R. Stanek, C. Jiang, B.P. Uberuaga, K.E. Sickafus, A.R. Cleave, R.W. Grimes, Predicted structure and stability of A4 B3 O 12 δ -phase compositions, Phys. Rev. B Condens. Matter Mater. Phys. 80 (2009). https://doi.org/10.1103/PhysRevB.80.174101.
- [11]Z. Teng, Y. Tan, S. Zeng, Y. Meng, C. Chen, X. Han, H. Zhang, Preparation and phase evolution of high-entropy oxides A2B2O7 with multiple elements at A and B sites, J. Eur. Ceram. Soc. 41 (2021) 3614–3620. https://doi.org/10.1016/j.jeurceramsoc.2021.01.013.
- [12]W. Fan, Y. Bai, Y. Liu, T. Li, B. Li, L. Zhang, C. Gao, S. Shan, H. Han, Principal element design of pyrochlore-fluorite dual-phase medium- and high-entropy ceramics, J. Mater. Sci. Technol. 107 (2022) 149–154. https://doi.org/10.1016/j.jmst.2021.08.018.
- [13]A.J. Wright, Q. Wang, Y.T. Yeh, D. Zhang, M. Everett, J. Neufeind, R. Chen, J. Luo, Short-range order and origin of the low thermal conductivity in compositionally complex rare-earth niobates and tantalates, Acta Mater. 235 (2022). https://doi.org/10.1016/j.actamat.2022.118056.
- [14]Y. Wang, Y.J. Jin, T. Wei, Z.G. Wang, G. Cao, Z.Y. Ding, Z.G. Liu, J.H. Ouyang, Y.J. Wang, Y.M. Wang, Size disorder: A descriptor for predicting the single- or dual-phase formation in multi-component rare earth zirconates, J. Alloys Compd. 918 (2022). https://doi.org/10.1016/j.jallcom.2022.165636.
- [15]D. Zhang, Y. Chen, H. Vega, T. Feng, D. Yu, M. Everett, J. Neufeind, K. An, R. Chen, J. Luo, Long- and short-range orders in 10-component compositionally complex ceramics, Advanced Powder Materials 2 (2023). https://doi.org/10.1016/j.apmate.2022.100098.
- [16]A.J. Wright, Q. Wang, C. Hu, Y.T. Yeh, R. Chen, J. Luo, Single-phase duodenary high-entropy fluorite/pyrochlore oxides with an order-disorder transition, Acta Mater. 211 (2021). https://doi.org/10.1016/j.actamat.2021.116858.
- [17]K. Song, D. Zhang, K.M. Chung, R. Chen, J. Luo, Fluorite-pyrochlore-weberite phase transitions in a series of 20-component ultrahigh-entropy compositionally complex ceramics, J. Eur. Ceram. Soc. 44 (2024) 7825–7836. https://doi.org/10.1016/j.jeurceramsoc.2024.05.058.
- [18]A.J. Wright, Q. Wang, S.T. Ko, K.M. Chung, R. Chen, J. Luo, Size disorder as a descriptor for predicting reduced thermal conductivity in medium- and high-entropy pyrochlore oxides, Scr. Mater. 181 (2020) 76–81. https://doi.org/10.1016/j.scriptamat.2020.02.011.
- [19]M. Qin, H. Vega, D. Zhang, S. Adapa, A.J. Wright, R. Chen, J. Luo, 21-Component compositionally complex ceramics: Discovery of ultrahigh-entropy weberite and fergusonite phases and a pyrochlore-weberite transition, Journal of Advanced Ceramics 11 (2022) 641–655. https://doi.org/10.1007/s40145-022-0575-5.
- [20]M.L. Sanjuán, Raman spectroscopy study of disorder phenomena and size effects in pyrochlores, in: Pyrochlore Ceramics: Properties, Processing, and Applications, Elsevier, 2022: pp. 95–159. https://doi.org/10.1016/B978-0-323-90483-4.00003-9.
- [21]E.R. Andrievskaya, Phase equilibria in the refractory oxide systems of zirconia, hafnia and yttria with rare-earth oxides, J. Eur. Ceram. Soc. 28 (2008) 2363–2388. https://doi.org/10.1016/j.jeurceramsoc.2008.01.009.