In this paper
AbstractIntroductionMethodsResultsAmbient-pressure structural properties.Variable-pressure neutron diffraction.DiscussionConclusionAcknowledgementsSupplementary InformationReferencesWorksPhysics and Astronomy
Pressure tuning the Jahn-Teller transition temperature in NaNiO
Abstract
NaNiO is a layered material consisting of alternating layers of NaO and Jahn-Teller-active NiO edge-sharing octahedra. At ambient pressure it undergoes a broad phase transition from a monoclinic to rhombohedral structure between 465 K and 495 K, associated with the loss of long-range orbital ordering. In this work, we present the results of a neutron powder diffraction study on powdered NaNiO as a function of pressure and temperature from ambient pressure to 5 GPa between 290 K and 490 K. The 290 K and 460 K isothermal compressions remained in the monoclinic phase up to the maximum pressures studied, whereas the 490 K isotherm was mixed-phase throughout. The unit-cell volume was fitted to a 2nd-order Birch-Murnaghan equation of state, with GPa. We observe at 490 K that the fraction of Jahn-Teller-distorted phase increases with increasing pressure, from 67.8(6)% at 0.71(2) GPa to 80.2(9)% at 4.20(6) GPa. Using this observation, in conjunction with neutron diffraction measurements at 490 K from 5.46(9) GPa to 0.342(13) GPa, we show that the Jahn-Teller transition temperature increases with pressure. Our results are used to present a structural pressure-temperature phase diagram for NaNiO. To our knowledge, this is the first diffraction study of the effect of pressure on the Jahn-Teller transition temperature in materials with edge-sharing Jahn-Teller-distorted octahedra, and the first variable-pressure study on a JT-active edge-sharing nickelate.
to 80.2(9)% at 4.20(6) GPa. Using this observation, in conjunction with neutron diffraction measurements at 490 K from 5.46(9) GPa to 0.342(13) GPa, we show that the Jahn-Teller transition temperature increases with pressure. Our results are used to present a structural pressure-temperature phase diagram for . To our knowledge, this is the first diffraction study of the effect of pressure on the Jahn-Teller transition temperature in materials with edge-sharing Jahn-Teller-distorted octahedra, and the first variable-pressure study on a JT-active edge-sharing nickelate.
Introduction
Many transition metal oxides exhibit a Jahn-Teller (JT) distortion due to degeneracy in the orbitals, manifesting as an elongation or compression of the ( = transition metal ion) octahedra, generally with associated orbital ordering. Previous studies on the effect of pressure on materials containing JT-active ions have found that pressure can entirely suppress the JT distortion and orbital ordering. [1, 2] It has also been observed that application of pressure reduces the magnitude of distortion in octahedra. [3, 4, 5, 6, 7]
One well-studied material under pressure is (with JT active ions). [2, 3, 8] At ambient pressure it adopts the perovskite structure with corner-sharing octahedra. An ordered JT distortion results in an orthorhombic symmetry. At K the JT distortion is suppressed, and there is an increase in symmetry first to a cubic phase with octahedral tilting, and then at higher temperatures to a rhombohedral phase. [9] The temperature-driven suppression of the JT distortion coincides with a marked increase in electronic conductivity. [10] On application of pressure at room temperature GPa, the JT distortion is decreased through reduction of the long Mn-O bond lengths. [3] At GPa a rhombohedral phase with no JT distortion co-exists with the distorted orthorhombic phase, [2] becoming single-phase at GPa. Similarly, the manganese (III) quadruple perovskite , exhibits a complete suppression of the JT distortion at GPa. [1]
There are several interesting studies on JT-distorted compounds with edge-sharing octahedra. tahedra. Here we describe three different examples classes, all containing JT-active Mn. MnO, a spinel containing both Mn and Mn, has been found to exhibit different pressure-dependence of JT-distorted octahedra depending on morphology; for instance, in single-crystal MnO the JT distortion survives to 60 GPa, [11], whereas there are observed transitions to JT-free phases at much lower pressures in powdered [12] and nanorod [13] MnO.
ZnMnO, also with spinel-type structure but with Zn on the Mn of MnO, has been studied to very high pressure ( GPa), [14] with a transition reported at GPa which has been alternately described as a transition from JT-elongation to a slight JT-compression[14] or a spin-crossover transition resulting in an insulatormetal transition. [15] CuMnO, with delafossite structure, has also had the pressure-dependence of its JT distortion studied. [16] It exhibits a higher compressibility in the long Mn-O bond than the short Mn-O bond similar to LaMnO [3] and other materials [4, 5] up to around around GPa; above this pressure there is an isostructural phase transition associated with a collapse in the interlayer (-axis) and an increase in the volume of the MnO JT-distorted octahedra.
Nickelates containing JT-active Ni are far less studied than the manganates under pressure. This may be partly because many materials containing Ni octahedra do not exhibit a co-operative JT distortion, where the JT distortion is long-range ordered. NdNiO, which has been subject to a variable-pressure structural study, [17] is not considered to contain a JT distortion, [18, 19, 20] as is the case for most Ni-containing perovskites. [21] Similarly, AgNiO is widely accepted to not contain any kind of JT distortion. [22, 23] LiNiO is an interesting case as it does not display long-range magnetic or orbital ordering, likely due to Li/Ni site mixing; some experimental results have been interpreted as evidence for a non-cooperative JT distortion, [24, 25] although this is debated. [26, 27, 28] Similarly various nickel-containing perovskites [29] are subject to discussion regarding whether there exists any kind of JT distortion.
NaNiO is a layered nickelate. The presence of the JT distortion in NaNiO is not subject to debate, [28, 30, 31] even amongst proponents of alternative theories for degeneracy-breaking in LiNiO. [28] NaNiO is therefore an ideal choice for studying the effect of pressure. on the JT distortion in a material which is both a nickelate and has edge-sharing octahedra.
The room-temperature phase of is a semiconductor, based on its black colour and by analogy with [32], but we do not know of any measurement of the conductivity properties of the high-temperature phase. is of interest because of its magnetic ground state, consisting at ambient pressure of intra-layer ferromagnetism and inter-layer anti-ferromagnetism.[33, 34, 35] It has also been studied in recent years because ( alkali metal) is the template compound for Ni-rich alkali metal-transition metal oxides within the field of batteries.[36, 37]
has an ordered JT distortion at room temperature due to degeneracy in orbitals in low-spin . It exhibits a first-order phase transition between 465 K and 495 K to an undistorted phase. The crystal structures are shown in Figure 1. The monoclinic () JT-distorted phase consists of alternating layers of edge-sharing and octahedra. Both the and octahedra exhibit angular and bond-length distortions from geometrically regular octahedra. Ni, Na, and O ions occupy the , , and Wyckoff sites respectively. The rhombohedral () phase consists of the same arrangement of alternating and layers of edge-sharing octahedra, with octahedra bound within layers by tetrahedra. In this phase, Ni, Na, and O ions occupy the , , and Wyckoff sites respectively. The unit cell remains centrosymmetric, with the change in symmetry due solely to the repression of the JT distortion and the resulting non-variable bond lengths.

Figure 1. (a) and (b) show the monoclinic, Jahn-Teller distorted phase along the -axis and -axis respectively. (c) and (d) show the rhombohedral, JT-inactive phase along the and axes respectively. cations are shown in green, anions in red, and cations in cyan. ions and octahedra are hidden in (b) and (d). The solid black quadrilaterals denote the unit cell. The black arrows represent the directions of principal axes of compression projected into the (a,c) -plane and (b,d) -plane.
In this work, we present a structural study of as a function of temperature between 290 K and 500 K, and pressure up to 5.46(9) GPa. We demonstrate using the 490 K isotherm that the JT transition temperature increases between 2 GPa and 4.2 GPa, increasing more rapidly with pressure at higher pressures, while the degree of distortion decreases over this pressure range.
Methods
Sample preparation and characterisation. Samples were prepared by solid state synthesis. (Alfa Aesar; 95%) and (Alfa Aesar; 99.995%) were mixed and pelletised in a 1.05:1 molar ratio of Na:Ni, with excess Na to account for Na-loss during heating. Sample was heated to 973 K for 70 hrs in a tube furnace under constant flow of . To prevent reaction with moisture, the sample was stored and handled in an inert Ar-atmosphere. X-Ray Diffraction (XRD) data were obtained using a Bruker D8 Discover powder (Cu K, Å) diffractometer. A Mira3 TESCAN Scanning Electron Microscopy was used to obtain SEM images of the morphology of , with an accelerating electron voltage of 3 kV (for SEM images, see SI).
Ambient-pressure neutron diffraction. Ambient-pressure neutron diffraction was performed using the NOMAD instrument [38] at the Spallation Neutron Source, Oak Ridge National Laboratory, USA. was sealed in a glass ampoule for the measurements. Heating was performed using a furnace. The sample was measured during heating at 293 K, 450 K, 500 K, and after cooling at 316 K.
Variable-pressure neutron diffraction. Variable-temperature and -pressure neutron diffraction studies were performed at the PEARL instrument [39] (), ISIS Neutron and Muon Source, UK, using a V3 Paris-Edinburgh press. The sample was measured between 0.107(8) GPa and 4.24(5) GPa at 290 K, 0.130(10) GPa and 5.29(8) GPa at 460 K, and 0.254(17) GPa and 4.20(6) GPa at 490 K. NaNiO was packed into a encapsulated null scattering TiZr gasket which was loaded in a zirconia-toughened alumina toroidal profile anvil, with a lead pellet for pressure calibration. [40] Anhydrous deuterated methanol:ethanol (4:1 by volume) was used as a pressure transmitting medium for the ambient temperature isothermal compression experiment. Preliminary measurements indicated that NaNiO reacted with the methanol:ethanol solution at higher temperatures [SI Figure S2], and so a 1:1 ratio by volume of FC77:FC84 fluorinert (purchased from 3M) was used for the 460 K and 490 K isotherms. The data were processed and corrected using Mantid. [41]
Diffraction Analysis. Diffraction data were analysed using the software package TOPAS 5, [42] utilising Pawley fitting[43] and Rietveld refinement. [44] For NaNiO, preliminary analysis of NOMAD data indicated Na occupancy was 1 within error, hence site occupancy of all sites during all further refinement was fixed at 1. Thermal B parameters were allowed to refine but constrained to be positive and not exceed a value of . All atomic positions were refined within symmetry constraints. The background was fitted by a Chebyschev polynomial (order 6 for PEARL data, order 11 for NOMAD data, order 19 for XRD data). For XRD data, a TCHZ peak-shape was used. [45] Peak-shapes used for neutron data are discussed in SI Section II. For PEARL only the detection bank was used, but for NOMAD a combined refinement was performed using banks 2-5 (, , , respectively).
Results
Ambient-pressure structural properties.
Powder X-ray diffraction of the as-synthesised NaNiO indicated the formation of a phase-pure product. SEM on the material indicates the sample is polycrystalline with particulates between 0.2 m and 5 m in diameter [SI Figure S13]. Rietveld refinement using the reported monoclinic space group [SI Figure S1; Table S1] yielded lattice parameters consistent with prior reports. [30, 46]
The reported monoclinicrhombohedral phase transition in NaNiO was investigated using neutron powder diffraction at ambient pressure on the NOMAD instrument. Rietveld refinement [Figure 2] shows the phase transition occurs between 450 K and 500 K, and is reversible on cooling. The lattice parameters [SI Table S2] all exhibit positive thermal expansion, and are consistent with previous measurements. [30, 31]
Figure 2. Rietveld refinements for the ambient-pressure, variable-temperature neutron diffraction measurements of NaNiO on bank 5 of NOMAD () at 293 K (top) and 500 K (bottom). Black dots: measured data; red line: calculated diffraction pattern from Rietveld refinement; grey line: . Green and orange tick marks show expected reflections for the monoclinic and rhombohedral phase respectively.
In the monoclinic structure the NiO octahedra exhibit a cooperative JT distortion with 2 longer Ni-O bonds, whereas in the high-temperature rhombohedral phase all six Ni-O bonds are equivalent [Figure 1]. The degree of bond length distortion within individual NaO and NiO octahedra can be evaluated using a number of distortion metrics, calculated using TOPAS 5. [42] Here we consider the effective coordination number [47] and the bond length distortion indices, [48] which measure distortion in octahedra by quantifying the difference from the average value of the distances between the central cation and the coordinated oxygen anions. The general form of effective coordination, ECoN, and bond length distortion index, , is given in the SI. The equations as applicable to monoclinic NaNiO are:
where is a modified average bond length defined in SI, and;
where is the average bond length, and and are the long and short bonds respectively.
In the rhombohedral structure the effective coordination number and bond length distortion index of both the and octahedra are constrained by symmetry to values of 6 and 0 respectively. In the monoclinic structure, departure from these values indicates bond length disproportionation and is primarily attributable to the JT distortion. These changes are significantly larger for the JT-active octahedra than the octahedra. Throughout the measurement, the bond length distortion index {effective coordination} of octahedra remains very near its high-symmetry value of 0{6}, for example at 293 K the value of bond length distortion index {effective coordination} in octahedra is 0.00581(11){5.99232(19)}, compared with 0.05463(14){5.309(3)} in the octahedra. This is indicative of much greater distortion in bond lengths for octahedra, consistent with the JT distortion. The values of bond length distortion index are on the same order of magnitude as recent studies on JT-distorted -containing compounds.[16, 49]
Inconsistency in bond length is not the only distortion of the octahedra from regular octahedra. A regular octahedron would have bond angles for nearest-neighbour O anions. However, in both the JT-active monoclinic phase and the JT-inactive rhombohedral phase there is variance from this ideal bond angle. Non-nearest-neighbour oxygen anions are constrained to have bond angles via the central cation, and so the 12 bond angles in an octahedron are each paired with another O-M-O bond, with the paired bond angles sharing one oxygen in common and with their non-shared oxygen anions occurring along a straight line through the central cation (for a visual representation, see SI Figure S14). We define these bond angles as , where the two angles in a pair have opposite sign preceding the . can also be thought of as a measure of the extent of angular distortion. In the rhombohedral structure, there is only one value of for each type of octahedron, with half the O--O bond angles being and the other half being . In the monoclinic unit cell where octahedra have two long -O (=Na,Ni) bonds and four short -O bonds, there are four nearest-neighbour bond angles between short and short bonds and eight nearest-neighbour bond angles between short and long bonds. We therefore must define two values of for the bond angles in the monoclinic phase, and , respectively. Table 1 shows these values of at each temperature. It is clear that NaO octahedra exhibit far higher bond angle distortion than NiO octahedra, in contrast to the bond length distortion which is greater for NiO octahedra. This is not unexpected, given that crystal field effects will result in much greater stability for open-shell Ni in an octahedral configuration, minimising bond angle variance, whereas this won’t be a factor for closed-shell Na cations.
| Phase | T (K) | NiO (°) | NaO (°) | ||
| 293 (-) | 6.134(17) | 5.456(19) | 14.494(12) | 9.708(14) | |
| 450 (↑) | 6.163(19) | 5.50(2) | 14.564(14) | 9.844(16) | |
| 500 (↑) | 6.135(15) | 11.777(13) | |||
| 316 (↓) | 6.121(18) | 5.46(2) | 14.505(13) | 9.731(15) |
Table 1. This table shows the value of for bond angles O--O (=Na, Ni) which take the value , as a function of temperature. For definitions, check the main text. The arrow next to the temperature indicates whether the data were collected on warming or cooling of the sample.
Variable-pressure neutron diffraction.
The effect of pressure on the JT distortion in NaNiO was explored at 290 K, 460 K, and 490 K, with an example Rietveld refinement shown in Figure 3 and data shown in Figure 4-Figure 8. Over the entire pressure and temperature range studied NaNiO could be described using the previously-reported ambient pressure crystal structures. Diffraction data also included contributions from alumina and zirconia in the sample environment and the lead used to determined the applied pressure - these are also included in the structural refinements.
Figure 3. Rietveld refinements for the variable-pressure neutron diffraction data of NaNiO. Top: a representative plot with monoclinic NaNiO only; bottom: a representative plot with both monoclinic and rhombohedral NaNiO. Black dots: measured data; red line: calculated diffraction pattern from Rietveld refinement; grey line: . Unfitted peaks are marked with an asterisk and arise from crystalline fluorinert [SI Figure S3].
Figure 4. Fractional contraction of lattice parameters obtained by Rietveld refinement [44] as a function of temperature and pressure for the monoclinic and rhombohedral phases of NaNiO. Error bars are shown for all plots; where they are not visible it is because the error is smaller than the data point. Lines are a guide for the eye.
Figure 8. The fraction of NaNiO which is in the monoclinic phase at 490 K, as a function of pressure.
In addition, at higher temperatures (460 K and 490 K) and pressures additional peaks attributed to crystallisation of the fluorinert pressure media [SI Figure S3] are observed in the measurements.
Rietveld analysis, Figure 3, shows that NaNiO remained in the monoclinic phase at 290 K (up to 4.24(5) GPa) and 460 K (up to 5.29(8) GPa). However, the measurements at 490 K capture NaNiO midway through its transition from JT-distorted monoclinic to JT-inactive rhombohedral, and throughout this isotherm the NaNiO is mixed-phase.
The lattice parameters [Figure 4] show the expected variation with temperature and
pressure. The rhombohedral and monoclinic phases have similar compressibility, and in both is considerably more compressible in the inter-layer direction (-axis) than the intra-layer (-) plane. With reference to Figure 1, we note that compression within the plane results in changes to the highly ionic - interactions and the less ionic, but still repulsive - interactions, whereas compression in the inter-layer direction will compress the - and - bonds which are softer due to the nearest-neighbour interaction lacking a Coulomb repulsive force. This higher compressibility in the inter-layer direction is consistent with that seen in another material with alternating layers of edge-sharing octahedra, the honeycomb iridate .[50, 51]
Within the plane, in monoclinic , the -axis is less compressible than the -axis. A reason for this might be that - and - interactions are parallel to the direction of compression for the -axis, maximising the increase in Coulomb repulsion with decreasing lattice parameter due to compression, whereas there are no -/- interactions with components only along the -axis. Another contribution may be that the - and - ionic distances parallel to the axis are considerably shorter than the distances which can be projected onto the -axis ( Å and Å respectively at 290 K and 0.107(8) GPa).
PASCal[52] was used to obtain the bulk modulus for each isotherm, using a 2nd-order Birch-Murnaghan equation of state.[53] A plot of the unit cell volume obtained by Rietveld refinement, as a function of pressure, with a fit of this equation of state, is shown in Figure 5 and tabulated in Table 2. For the monoclinic phase, which is consistent with a structure with positive thermal expansion. decreases with increasing temperature, meaning that compressibility increases with temperature. At 290 K, is 119.6(5) GPa. This is comparable with a similar JT-distorted material with edge-sharing octahedra, CuMnO, which has bulk modulus 116(2) GPa. [16] It is, however, substantially less than the reported bulk modulus for ZnMnO of 197(5) GPa, [14] and although there are several different reported values for MnO depending on phase and morphology, [11, 12, 13] all are higher than what we report for NaNiO. LaMnO is not entirely comparable owing to the LaO units and corner-sharing octahedra, but for reference it has a reported bulk modulus of 108(2) GPa. [3]
| Structure | Temperature | (Å) | (GPa) |
| 490 K | 119.83(2) | 113(1) | |
| 490 K | 79.900(16) | 110(1) | |
| 460 K | 79.798(9) | 113.5(6) | |
| 290 K | 79.258(7) | 119.6(5) |
Table 2. Parameters determined from the 2nd-order Birch-Murnaghan Equation of State, obtained using PASCal[52] [SI Section V, Table S12].
Figure 5. Variation in unit cell volume per formula unit for monoclinic phase. Solid data points show experimental derived values and the solid line the determined 2nd-order Birch-Murnaghan equation of state. Full lattice parameters are found in SI Tables S3-5. Where error bars are not visible it is because they are smaller than the data points.
The directions of the principal axes of compression are determined using PASCal[52] [SI Figure S13]. These are the axes in which compression occurs linearly with pressure, and do not necessarily align with the crystallographic axes in crystalline materials. The principal axis directions projected onto the -plane do not change between the monoclinic and rhombohedral phases. However, the inter-layer direction is a principal axis for the rhombohedral phase, but not for the monoclinic phase where two principal axes are at an angle to the inter-layer direction [Fig. 1]. Interestingly, the axis of JT elongation does not correspond to any of the principal axes SI Table S11. There is some temperature dependence in the principal axis directions, likely owing to variation in the lattice parameters with temperature SI Tables. S3-S5. The compressibilities of in each of the principal axes is consistent with variation in lattice parameters SI Table S11.
We now consider pressure-dependence of the bond length distortion index and effective coordination SI Figure S8; Tables S8-10. As in the ambient pressure measurements the bond length distortions are significantly larger in the compared to the octahedra, with the most significant variation being the increase in effective coordination (5.387(10) at 0.107(8) GPa to 5.504(13) at 4.24(5) GPa at 290 K) and decrease in bond length distortion index (from 0.0512(5) to 0.0458(7) in the same pressure range at 290 K) of the octahedra in the monoclinic phase on application of pressure. octahedra exhibit far smaller changes in bond length distortion index and effective coordination, with the overall behaviour not seeming to exhibit a consistent change with pressure; effective coordination remains between 5.98 and 5.99 throughout the 290 K isotherm.
The differing behaviour of bond length distortion index between and octahedra is likely attributable to the fact that is JT-active and is not, and suggests that pressure is decreasing the magnitude of JT distortion. We investigate this by considering the direct manifestation of the JT effect in . The Ni-O bond lengths of both the monoclinic and rhombohedral phases as a function of pressure are shown in Figure 6. The short Ni-O bonds are less sensitive to the effect of pressure than the long Ni-O bonds, indicating that the difference between long and short Ni-O bond lengths is decreasing with pressure. We also observe that the average monoclinic bond length is consistently larger than the rhombohedral bond length at 490 K SI Figure S9. In the octahedra, SI Figure S5, there is an approximately linear variation of the Na-O bond lengths with pressure. We conclude that anisotropy of the Ni-O bond compression is a consequence of the JT distortion in octahedra.
Figure 6. Ni-O bond lengths, as a function of pressure, and associated error of monoclinic NaNiO at 290 K, 460 K, and 490 K, with the JT-inactive rhombohedral phase bond lengths shown for 490 K. Lines are a guide for the eye.
The observed decrease in difference between long and short Ni-O bonds with pressure is also reported for other materials containing a JT distortion, such as LaMnO[3], KCuF[4], and CuAsO[5]. This is equivalent to the observed tendency with pressure of NiO octahedra bond length distortion index and effective coordination towards their symmetry-constrained values of 0 and 6, respectively. It indicates that the symmetry of JT-distorted octahedra increases with application of pressure in monoclinic NaNiO, consistent with prior reports. [3, 4, 5, 6, 7]
A previous study on LaMnO attempted to extrapolate a linear fit to the pressure-dependence of JT-distorted bond length, and estimated a critical pressure of GPa.[3] Such an extrapolation could be performed for NaNiO yielding a critical pressure of GPa, converging at a Ni-O bond length of 1.85 Å at 290 K. However, this value is unlikely to be representative of the true critical pressure of the JT distortion in NaNiO. A later study on LaMnO found that the JT distortion was suppressed at a lower pressure of around GPa, suggesting such extrapolation does not yield accurate predictions. Further, studies of other JT-distorted materials such as [(CH)NH][Cu(HCOO)][6] and CuMnO[16] have found that this pressure-dependence of JT-disproportionated bond length exists only up to a certain pressure, beyond which there is a change in behaviour which renders such extrapolation of low-pressure behaviour meaningless.
We earlier defined the bond angles and () for monoclinic NaNiO, and the associated values which reduce the number of parameters needed to describe the behaviour. We plot these values in Figure 7 for the 290 K isotherm. These plots show that throughout the studied pressure range, the degree of angular distortion is far greater for NaO than NiO, as was the case at ambient pressure [Table 1]. We can also see that with application of pressure, is decreasing in value; this indicates increasing symmetry towards the 90° degree bond angle for a perfect octahedron, analogous to the increasing symmetry with pressure we see with bond length distortion index.
Figure 7. Values of for NiO and NaO octahedra as a function of pressure at 290 K, representing the magnitude of angular distortion as nearest-neighbour bond angles take the value 90° ± . The two different values in monoclinic NaNiO are between two short bonds (red) and between a short and long bond (blue), where bonds are short or long due to the JT distortion. Lines are a guide for the eye.
The pressure dependence of the and octahedral volume in [SI Figure S7] show that the changes in volume displays different pressure-dependence for octahedra and octahedra, as compared with the unit cell. octahedra have higher relative compressibility than the entire unit cell, and octahedra are much more resistant to compression. It has been shown that for perovskites with and polyhedra the parameters and can be used to predict the relative compressibility of the polyhedra via , in which is the bond compressibility, is the distance between the central cation and the ith O anion, and is a bond-valence parameter defined in SI.[54] We apply this model to and find that throughout the 290 K isotherm [SI Figure S12]. Accounting for the different values of , this indicates that , which is consistent with our observation that octahedra are more compressible than octahedra. This may be due to differences in electronic configuration for closed-shell and open-shell , or being a much larger ion than .
We now consider a related model proposed by Angel et al., again in the context of perovskites,[55] whereby a transition temperature associated with an octahedral phase transition will exhibit if octahedra are more compressible than the extra-framework cation sites (analogous to the octahedra in ), and if octahedra are less compressible. Our structural analysis shows the enhanced compressibility of octahedra when compared to , and so this model predicts the observed increase in with pressure. It is worth noting that there are more degrees of freedom in the layered structure so the relationships between distortions in and may not be so strongly coupled as in the perovskites. However, the basic hypothesis of the model in Angel et al. appears to be applicable to .
Along the 490 K isotherm, both monoclinic and rhombohedral were observed in coexistence. The fraction of in the low-temperature, JT-distorted monoclinic phase is shown in Figure 8. The fraction remains approximately stagnant to GPa, beyond which it consistently increases with increasing pressure. In the range where it is increasing, the monoclinic fraction at 490 K increases from 67.8(6)% at 0.71(2) GPa to 80.2(9)% at 4.20(6) GPa. This indicates that increases with increasing pressure beyond GPa, consistent with our prediction based on octahedral compressibility.
To explore the P-T dependence of the transition, the sample was heated at 5.29(8) GPa from 460 K to 490 K after measuring the variable-pressure 460 K isotherm. At ambient pressure, this would result in a mixed monoclinic/rhombohedral phase. However, at the resulting high pressure of 5.46(9) GPa, we did not observe emergence of any rhombohedral peaks in the diffraction pattern. Subsequent reduction in pressure to 0.342(13) GPa at the same temperature, 490 K, did yield emergence of rhombohedral peaks [Figure S4], further supporting our interpretation that is increasing with pressure.
Discussion
The results of our P-T study on NaNiO are summarised in a phase diagram, Figure 9.
Figure 9. Tentative phase diagram showing the structure of NaNiO as a function of pressure and temperature. Triangles denote diffraction measurements and point left/right if was decreasing/increasing or up/down if was increasing or decreasing. The precise boundaries of the three regions are estimates based on available data, with the results in Refs. [31, 46] used to estimate the broadness of the transition.
To our knowledge, this is the first study on the effect of pressure on the JT transition temperature in a material containing edge-sharing octahedra, and the first variable-pressure study on a JT-active edge-sharing nickelate. Comparison between the results of this study and previous works must therefore rely on the work done on non-nickelate materials.
Like the perovskite materials LaMnO [3, 8] and KCuF, [4] NaNiO exhibits far greater compressibility in the JT-elongated O-Ni-O axis than the JT-compressed O-Ni-O axes, with the JT distortion in both NaNiO and the previously-discussed perovskites decreasing in magnitude with pressure. The consistent behaviour with other JT-active materials is also clear evidence that the charge disproportionation model proposed for LiNiO [26, 27, 28] and some Ni-containing perovskites [21] is not applicable to NaNiO.
A novel behaviour we observe in NaNiO is that increases with application of pressure NaNiO. For comparison, in LaMnO the JT distortion is suppressed at GPa, indicating that is reduced to room temperature from K by 12 GPa. [2] This mechanism seems unlikely in NaNiO due to the increasing with pressure, although we cannot exclude the possibility that a reversal above our maximum measured pressure may result in a decrease in . Additionally, there is a trend observed in this, and other, [3, 4, 5, 6, 7] works, that the magnitude of distortion due to the JT effect decreases with pressure. This could be interpreted as meaning that there is some pressure where the distortion is entirely suppressed and the octahedra achieve a bond length distortion index of zero, consistent with absence of an ordered JT distortion. However, this is at odds with recent reports [6, 16] which show that at some pressure the long and short bonds in JT-distorted octahedra eventually stabilise at different lengths. It is therefore not clear how exactly the JT distortion is suppressed in with high pressure, and further investigation is needed to elucidate this.
We should once again note that the conductivity behaviour of the high-temperature phase of also remains unexplored. There is significant reduction in resistivity with JT suppression in , [10] and there is a possibility for similar behaviour in high-temperature rhombohedral . Density Functional Theory calculations on rhombohedral, JT-free (which is isostructural with the high-temperature phase of ) have suggested metallic behaviour. [56] On a similar note, a broad first-order transition between two structures with a group-subgroup relationship in [57] featured coexistence of electronic phases. If the high-temperature phase of were indeed metallic, this metallic behaviour could explain why in this material, as application of pressure may result in narrowing of - bands, pushing the metal-to-insulator phase transition to higher and higher temperatures, and electronic phase coexistence could provide an explanation for the very broad nature of the transition.
Conclusion
The key finding of this study is that in , increases slightly with application of pressure while JT-distorted octahedra become more symmetric, as demonstrated by the pressure-dependence of two distortion metrics (effective coordination and bond length distortion index). While the latter is a well-documented property of JT-distorted materials, the former is in contrast to the JT distortion in . [2] is more resistant to pressure than other similar materials, having a higher bulk modulus ( GPa at 290 K) than similar perovskites, [3, 4] Prussian Blue analogues, [7] and layered honeycomb structures, [50] although its bulk modulus is very similar to JT-distorted edge-sharing [16] and is less than [11, 12, 13], [58], and [14]. also displays a much smaller magnitude of than , with shifting from K at ambient pressure to room temperature in 12 GPa$[2] compared with a very small shift from K at ambient pressure in .
Further variable-pressure diffraction measurements, at several temperatures and to higher pressures, are needed to fully understand the process of suppressing the JT distortion in . Variable-pressure Raman spectroscopy measurements on could also be useful and may help identify phase transitions at higher pressures.
Additionally, future investigations are needed to investigate whether other JT-distorted materials exhibit a pressure dependence, for example a study building on previous work on [16] by measuring at multiple isotherms.
Data availability: The variable-pressure neutron diffraction data from the PEARL instrument at ISIS is available at doi:10.5286/ISIS.E.RB2000219.[59] All other data can be found at doi:10.17863/CAM.81605.[60]
Acknowledgements
This work was supported by the Faraday Institution grant number FIRG017. LNC acknowledges a scholarship EP/R513180/1 to pursue doctoral research from the UK Engineering and Physical Sciences Research Council (EPSRC). Experiments at the ISIS Neutron and Muon Source were supported by a beamtime allocation RB2000219 from the Science and Technology Facilities Council of the United Kingdom. A portion of this research used resources at the Spallation Neutron Source, a DOE Office of Science User Facility operated by the Oak Ridge National Laboratory of the United States of America, with data collection performed by Joerg C Neuefeind and Jue Liu (ORNL). Figure 1 was generated using CrystalMaker®: a crystal and molecular structures program from CrystalMaker Software Ltd, Oxford, UK. Heather Greer assisted with the SEM images in Supplementary Information. We also thank others whose ideas and comments were useful: Joshua D Bocarsly, Farheen N Sayed, Andrew G Seel, Siddharth S Saxena, Euan N Bassey, Nicola D Kelly, Venkateswarlu Daramalla, Chloe S Coates, Camilla Tacconis, and Debasis Nayak.
Supplementary Information
The Supporting Information is available free of charge at ————-. Additional neutron diffraction data and refinement details. SEM images. Tabulated diffraction data and distortion parameters. Transformation matrices.
References
References
- [1]Bhadram, V.; Joseph, B.; Delmonte, D.; Gilioli, E.; Baptiste, B.; Godec, Y. L.; Lobo, R.; Gauzzi, A. Reentrant phase transition and suppression of Jahn-Teller distortion in the quadruple perovskite structure under high pressure. arXiv preprint arXiv:2102.09998 2021,DOI
- [2]Zhou, J.-S.; Uwatoko, Y.; Matsubayashi, K.; Goodenough, J. Breakdown of magnetic order in Mott insulators with frustrated superexchange interaction. Phys. Rev. B 2008, 78, 220402.DOI
- [3]Loa, I.; Adler, P.; Grzechnik, A.; Syassen, K.; Schwarz, U.; Hanfland, M.; Rozenberg, G. K.; Gorodetsky, P.; Pasternak, M. Pressure-induced quenching of the Jahn-Teller distortion and insulator-to-metal transition in LaMnO₃. Phys. Rev. Lett. 2001, 87, 125501.DOI
- [4]Zhou, J.-S.; Alonso, J.; Han, J.; Fernández-Díaz, M.; Cheng, J.-G.; Goodenough, J. Jahn-Teller distortion in perovskite KCuF₃ under high pressure. Journal of Fluorine Chemistry 2011, 132, 1117–1121.DOI
- [5]Caslin, K.; Kremer, R.; Razavi, F.; Hanfland, M.; Syassen, K.; Gordon, E.; Whangbo, M.-H. Competing Jahn-Teller distortions and hydrostatic pressure effects in the quasi-one-dimensional quantum ferromagnet CuAs₂O₄. Phys. Rev. B 2016, 93, 022301.DOI
- [6]Collings, I. E.; Bykov, M.; Bykova, E.; Hanfland, M.; van Smaalen, S.; Dubrovinsky, L.; Dubrovinskaia, N. Disorder–order transitions in the perovskite metal-organic frameworks [(CH₃)₂NH₂][M(HCOO)₃] at high pressure. CrystEngComm 2018, 20, 3512–3521.DOI
- [7]Boström, H. L.; Collings, I. E.; Cairns, A. B.; Romao, C. P.; Goodwin, A. L. High-pressure behaviour of Prussian blue analogues: interplay of hydration, Jahn-Teller distortions and vacancies. Dalton Transactions 2019, 48, 1647–1655.DOI
- [8]Pinsard-Gaudart, L.; Rodriguez-Carvajal, J.; Daoud-Aladine, A.; Goncharenko, I.; Medarde, M.; Smith, R.; Revcolevschi, A. Stability of the Jahn-Teller effect and magnetic study of LaMnO₃ under pressure. Phys. Rev. B 2001, 64, 064426.
- [9]Rodriguez-Carvajal, J.; Hennion, M.; Moussa, F.; Mouldden, A.; Pinsard, L.; Revcolevschi, A. Neutron-diffraction study of the Jahn-Teller transition in stoichiometric LaMnO₃. Phys. Rev. B 1998, 57, R3189.
- [10]Zhou, J.-S.; Goodenough, J. Paramagnetic phase in single-crystal LaMnO₃. Phys. Rev. B 1999, 60, R15002.DOI
- [11]Ovsyannikov, S. V.; Aslandukova, A. A.; Aslandukov, A.; Chariton, S.; Tsirlin, A. A.; Korobeynikov, I. V.; Morozova, N. V.; Fedotenko, T.; Khandarkhaeva, S.; Dubrovinsky, L. Structural stability and properties of marokite-type γ-Mn₃O₄. Inorganic Chemistry 2021, 60, 13440–13452.DOI
- [12]Paris, E.; Ross II, C. R.; Olijnyk, H. Mn₃O₄ at high pressure: a diamond-anvil cell study and a structural modelling. European Journal of Mineralogy 1992, 87–94.DOI
- [13]Li, J.; Liu, B.; Dong, J.; Li, C.; Dong, Q.; Lin, T.; Liu, R.; Wang, P.; Shen, P.; Li, Q., et al. Size and morphology effects on the high pressure behaviors of Mn₃O₄ nanorods. Nanoscale Advances 2020, 2, 5841–5847.DOI
- [14]Åsbrink, S.; Waśkowska, A.; Gerward, L.; Olsen, J. S.; Talik, E. High-pressure phase transition and properties of spinel ZnMn₂O₄. Phys. Rev. B 1999, 60, 12651.DOI
- [15]Choi, H.; Shim, J.; Min, B. Electronic structures and magnetic properties of spinel ZnMn₂O₄ under high pressure. Phys. Rev. B 2006, 74, 172103.
- [16]Lawler, K. V.; Smith, D.; Evans, S. R.; Dos Santos, A. M.; Molaison, J. J.; Bos, J.-W. G.; Mutka, H.; Henry, P. F.; Argyriou, D. N.; Salamat, A., et al. Decoupling Lattice and Magnetic Instabilities in Frustrated CuMnO₂. Inorganic Chemistry 2021, 60, 6004–6015.DOI
- [17]Medarde, M.; Mesot, J.; Rosenkranz, S.; Lacorre, P.; Marshall, W.; Klotz, S.; Loveday, J.; Hamel, G.; Hull, S.; Radaelli, P. Pressure-induced orthorhombic-rhombohedral phase transition in NdNiO₃. Physica B: Condensed Matter 1997, 234, 15–17.DOI
- [18]García-Muñoz, J.; Rodríguez-Carvajal, J.; Lacorre, P. Neutron-diffraction study of the magnetic ordering in the insulating regime of the perovskites RNiO₃ (R = Pr and Nd). Phys. Rev. B 1994, 50, 978.pubmed.ncbi.nlm.nih.gov/9975764
- [19]Mizokawa, T.; Khomskii, D.; Sawatzky, G. Spin and charge ordering in self-doped Mott insulators. Phys. Rev. B 2000, 61, 11263.DOI
- [20]García-Muñoz, J.; Aranda, M.; Alonso, J.; Martínez-Lope, M. Structure and charge order in the antiferromagnetic band-insulating phase of NdNiO₃. Phys. Rev. B 2009, 79, 134432.
- [21]Johnston, S.; Mukherjee, A.; Elfimov, I.; Berciu, M.; Sawatzky, G. A. Charge dispro- portionation without charge transfer in the rare-earth-element nickelates as a possible mechanism for the metal-insulator transition. Phys. Rev. Lett. 2014, 112, 106404.
- [22]Wawrzýnska, E.; Coldea, R.; Wheeler, E. M.; Mazin, I. I.; Johannes, M.; Sörgel, T.; Jansen, M.; Ibberson, R. M.; Radaelli, P. G. Orbital degeneracy removed by charge order in triangular antiferromagnet AgNiO₂. Phys. Rev. Lett. 2007, 99, 157204.DOI
- [23]Kang, J.-S.; Lee, S.; Kim, G.; Lee, H.; Song, H.; Shin, Y.; Han, S.; Hwang, C.; Jung, M.; Shin, H., et al. Valence and spin states in delafossite AgNiO₂ and the frustrated Jahn-Teller system ANiO₂ (A= Li, Na). Phys. Rev. B 2007, 76, 195122.DOI
- [24]Rougier, A.; Delmas, C.; Chadwick, A. V. Non-cooperative Jahn-Teller effect in LiNiO₂: an EXAFS study. Solid State Communications 1995, 94, 123–127.DOI
- [25]Chung, J.-H.; Proffen, T.; Shamoto, S.; Ghorayeb, A.; Croguennec, L.; Tian, W.; Sales, B. C.; Jin, R.; Mandrus, D.; Egami, T. Local structure of LiNiO₂ studied by neutron diffraction. Phys. Rev. B 2005, 71, 064410.DOI
- [26]Chen, H.; Freeman, C. L.; Harding, J. H. Charge disproportionation and Jahn-Teller distortion in LiNiO₂ and NaNiO₂: A density functional theory study. Phys. Rev. B 2011, 84, 085108.DOI
- [27]Foyevtsova, K.; Elfimov, I.; Rottler, J.; Sawatzky, G. A. LiNiO₂ as a high-entropy charge-and bond-disproportionated glass. Phys. Rev. B 2019, 100, 165104.DOI
- [28]Green, R.; Wadati, H.; Regier, T.; Achkar, A.; McMahon, C.; Clancy, J.; Dabkowska, H.; Gaulin, B.; Sawatzky, G.; Hawthorn, D. Evidence for bond-disproportionation in LiNiO₂ from x-ray absorption spectroscopy. arXiv preprint arXiv:2011.06441 2020,arxiv.org/abs/2011.06441
- [29]Zhou, J.-S.; Goodenough, J. Chemical bonding and electronic structure of RNiO₃ (R= rare earth). Phys. Rev. B 2004, 69, 153105.DOI
- [30]Dick, S.; Müller, M.; Preissinger, F.; Zeiske, T. The structure of monoclinic NaNiO₂ as determined by powder X-ray and neutron scattering. Powder Diffraction 1997, 12, 239–241.
- [31]Chappel, E.; Nunez-Regueiro, M.; Chouteau, G.; Isnard, O.; Darie, C. Study of the ferrodisorsive orbital ordering in NaNiO₂ by neutron diffraction and submillimeter wave ESR. The Euro. Phys. Journ. B-Cond. Matt. and Complex Systems 2000, 17, 615–622.
- [32]Galakhov, V.; Kurmaev, E.; Neumann, M.; Kellerman, D.; Gorshkov, V., et al. Electronic structure of LiNiO₂, LiFeO₂ and LiCrO₂: X-ray photoelectron and X-ray emission study. Solid State Communications 1995, 95, 347–351.DOI
- [33]Lewis, M.; Gaulin, B.; Filion, L.; Kallin, C.; Berlinsky, A.; Dabkowska, H.; Qiu, Y.; Copley, J. Ordering and spin waves in NaNiO₂: A stacked quantum ferromagnet. Phys. Rev. B 2005, 72, 014408.DOI
- [34]Baker, P.; Lancaster, T.; Blundell, S.; Brooks, M.; Hayes, W.; Prabhakaran, D.; Pratt, F. Thermodynamic and magnetic properties of the layered triangular magnet NaNiO₂. Phys. Rev. B 2005, 72, 104414.
- [35]Darie, C.; Bordet, P.; De Brion, S.; Holzapfel, M.; Isnard, O.; Lecchi, A.; Lorenzo, J.; Suard, E. Magnetic structure of the spin-1/2 layer compound NaNiO₂. The Euro. Phys. Journ. B-Cond. Matt. and Complex Systems 2005, 43, 159–162.
- [36]Vassilaras, P.; Ma, X.; Li, X.; Ceder, G. Electrochemical properties of monoclinic NaNiO₂. Journal of The Electrochem. Soc. 2012, 160, A207.DOI
- [37]Han, M. H.; Gonzalo, E.; Casas-Cabanas, M.; Rojo, T. Structural evolution and electrochemistry of monoclinic NaNiO₂ upon the first cycling process. Journal of Power Sources 2014, 258, 266–271.DOI
- [38]Neuefeind, J.; Feygenson, M.; Carruth, J.; Hoffmann, R.; Chipley, K. K. The nanoscale ordered materials diffractometer NOMAD at the spallation neutron source SNS. Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 2012, 287, 68–75.
- [39]Bull, C. L.; Funnell, N. P.; Tucker, M. G.; Hull, S.; Francis, D. J.; Marshall, W. G. PEARL: the high pressure neutron powder diffractometer at ISIS. High Pressure Research 2016, 36, 493–511.DOI
- [40]Fortes, A. D. RAL Technical Reports, RAL-TR-2019-002. 2019,
- [41]Arnold, O.; Bilb e ux, J.-C.; Borreguero, J.; Buts, A.; Campbell, S. I.; Chapon, L.; Doucet, M.; Draper, N.; Leal, R. F.; Gigg, M., et al. Mantid—Data analysis and visualization package for neutron scattering and μ SR experiments. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment 2014, 764, 156–166.DOI
- [42]Coelho, A. A. TOPAS and TOPAS-Academic: an optimization program integrating computer algebra and crystallographic objects written in C++. Journal of Applied Crystallography 2018, 51, 210–218.DOI
- [43]Pawley, G. Unit-cell refinement from powder diffraction scans. Journal of Applied Crystallography 1981, 14, 357–361.DOI
- [44]Rietveld, H. A profile refinement method for nuclear and magnetic structures. Journal of Applied Crystallography 1969, 2, 65–71.DOI
- [45]Thompson, P.; Cox, D.; Hastings, J. Rietveld refinement of Debye-Scherrer synchrotron X-ray data from Al₂O₃. Journal of Applied Crystallography 1987, 20, 79–83.DOI
- [46]Sofin, M.; Jansen, M. New route of preparation and properties of NaNiO₂. Zeitschrift für Naturforschung B 2005, 60, 701–704.DOI
- [47]Hoppe, R. Effective coordination numbers (ECoN) and mean fictive ionic radii (MEFIR). Zeitschrift für Kristallographie-Crystalline Materials 1979, 150, 23–52.DOI
- [48]Baur, W. The geometry of polyhedral distortions. Predictive relationships for the phosphate group. Acta Crystallographica Section B: Structural Crystallography and Crystal Chemistry 1974, 30, 1195–1215.DOI
- [49]Kimber, S. A. Charge and orbital order in frustrated Pb₃Mn₇O₁₅. Journal of Physics: Condensed Matter 2012, 24, 186002.DOI
- [50]Hermann, V.; Ebad-Allah, J.; Freund, F.; Pietsch, I.; Jesche, A.; Tsirlin, A. A.; Deisenhofer, J.; Hanfland, M.; Gegenwart, P.; Kuntscher, C. A. High-pressure versus isoelectronic doping effect on the honeycomb iridate Na₂IrO₃. Phys. Rev. B 2017, 96, 195137.DOI
- [51]Layek, S.; Mehlawat, K.; Levy, D.; Greenberg, E.; Pasternak, M.; Itié, J.-P.; Singh, Y.; Rozenberg, G. K. Electronic and structural properties of the honeycomb iridates A₂IrO₃ (A = Na, Li) at elevated pressures. Phys. Rev. B 2020, 102, 085156.DOI
- [52]Cliffe, M. J.; Goodwin, A. L. PASCal: a principal axis strain calculator for thermal expansion and compressibility determination. Journal of Applied Crystallography 2012, 45, 1321–1329.DOI
- [53]Birch, F. Finite elastic strain of cubic crystals. Physical Review 1947, 71, 809.DOI
- [54]Zhao, J.; Ross, N. L.; Angel, R. J. New view of the high-pressure behaviour of GdFeO₃-type perovskites. Acta Crystallographica Section B: Structural Science 2004, 60, 263–271.pubmed.ncbi.nlm.nih.gov/15148429
- [55]Angel, R. J.; Zhao, J.; Ross, N. L. General rules for predicting phase transitions in perovskites due to octahedral tilting. Phys. Rev. Lett. 2005, 95, 025503.pubmed.ncbi.nlm.nih.gov/16090697
- [56]Chen, Z.; Zou, H.; Zhu, X.; Zou, J.; Cao, J. First-principle investigation of Jahn-Teller distortion and topological analysis of chemical bonds in LiNiO₂. Journal of Solid State Chemistry 2011, 184, 1784–1790.
- [57]Ortega-San-Martin, L.; Williams, A. J.; Rodgers, J.; Attfield, J. P.; Heymann, G.; Huppertz, H. Microstrain sensitivity of orbital and electronic phase separation in SrCrO₃. Phys. Rev. Lett. 2007, 99, 255701.
- [58]Eto, T.; Endo, S.; Imai, M.; Katayama, Y.; Kikegawa, T. Crystal structure of NiO under high pressure. Phys. Rev. B 2000, 61, 14984.DOI
- [59]L. A. V. Nagle-Cocco, S. E. Dutton, and C. L. Bull. DOI:10.5286/ISIS.E.RB2000219. Behaviour of Jahn-Teller distorted NiO₆ octahedra in NaNiO₂ under pressure. 2021; https://data.isis.stfc.ac.uk/doi/STUDY/113601628/.
- [60]L. A. V. Nagle-Cocco, C. L. Bull, Christopher J. Ridley, S. E. Dutton. DOI:10.17863/CAM.81605. Data associated with ”Pressure tuning the Jahn-Teller transition temperature in NaNiO₂”. 2022; doi.org/10.17863/CAM.81605.







