WorksPhysics and Astronomy
Ambient-pressure superconductivity in bulk single crystals of LaNiO
Abstract
Superconductivity in the Ruddlesden-Popper (RP) nickelates has so far been confined either to multilayer compounds under high pressure or to square-planar infinite-layer-type films stabilized by epitaxial strain. Here we report the high-pressure high-temperature (HPHT) growth of bulk single crystals of the quadruple-layer RP nickelate LaNiO () and reproduce the quintuple-layer LaNiO () synthesis, two members of the series that had previously been accessible only as thin films. Topotactic reduction of LaNiO single crystals with CaH yields the square-planar phase LaNiO, which we show to be a bulk, ambient-pressure superconductor with , a shielding fraction above 60%, and a specific-heat jump that confirms a bulk pairing transition. All crystals show besides superconductivity a coexistence with a density-wave or metal-insulator transition. The upper critical field shows a small anisotropy, , which remains finite as , indicating three-dimensional superconductivity with substantial coupling between the four-layer blocks. These results identify the square-planar nickelates NiO as a bulk platform in which the layer number alone tunes the nickel valence into the superconducting regime, without chemical substitution, pressure, or epitaxial constraint.
The RP nickelates LaNiO provide a model system for studying how dimensionality and band filling jointly control the ground state of a correlated transition-metal oxide. As the number of perovskite layers, , increases, electrons gain the ability to move more freely along the axis, while the nominal Ni valence rises from 2+ at toward 3+ as , following . This structural progression generates a rich phase diagram: LaNiO () is an antiferromagnetic insulator, the and compounds are metals that host density-wave transitions[1, 2] and become superconducting under pressure[3, 4, 5, 6, 7], and LaNiO () is a paramagnetic metal at ambient conditions. The series therefore spans, within a single chemical family, the full range of behaviour from Mott-like insulator to correlated metal to conventional metal, making it an unusually direct testbed for theories that link interlayer coupling, orbital occupation, and superconductivity. Even hybrid forms of RP nickelates exist as single crystals with a mixed stacking of with alternating monolayer and bilayer, i.e. 12/12[7, 8] and 13/13[9, 10, 11], extending Ni charge mixtures and two density wave transitions[12, 13], as well as superconductivity[7, 10].
In bulk, however, this ladder of compounds stopped at . Members with are not thermodynamically stable under conventional high temperature ambient-pressure synthesis and instead appear, if at all, as stacking faults within intergrowths of lower RP phases and the perovskite end member, for example, LaNiO + LaNiO = LaNiO. Essentially everything known about the higher- members comes from metastable epitaxial films[14], and for LaNiO in thin-film form even basic structural parameters, the stacking sequence of the four layers, the number of crystallographically distinct Ni sites, and the oxygen content, must be inferred rather than directly measured, with substrate-imposed strain always present as a confounding variable. A bulk single crystal would resolve these ambiguities. The member was only very recently obtained in bulk form under high pressure[15]; the member, the subject of this work, has not previously been reported in bulk.
There is a second, independent motivation for growing these crystals. Topotactic removal of the apical oxygen converts the RP phases into square-planar compounds for : NiO, in which the nominal Ni valence, , is set purely by the layer number, falling from +1.5 at toward +1 as . In thin films, superconductivity has been reported at [16], with signatures extending from to [17]: precisely the range over which the Ni valence falls within the putative superconducting dome. Reproducing this regime in bulk has proven difficult. Bulk polycrystalline NdSrNiO obtained by topotactic reduction remains insulating and shows no superconductivity up to pressures of about 50 GPa[18]. Topotactic reduction itself is difficult to control uniformly in single crystals, and chemical doping of infinite-layer single crystals by divalent alkaline-earth substitution is especially challenging to achieve[19] as a clear solubility limit is reached[20]. The layer number offers an alternative, purely structural handle: it tunes the Ni valence by self-doping, without any need for aliovalent substitution. Reduction itself, however, tends to introduce block-like domains and microcracks in the case of infinite-layer nickelate crystals[21], degrading their quality. Partly for these reasons, no ambient-pressure nickelate superconductor has, until now, been reliably available in bulk single-crystal form, and the bulk and reduced square-planar compounds have not been found to superconduct[22, 23, 24]. Where, along this series, bulk ambient-pressure superconductivity first sets in has therefore remained an open and important question. Overall, a complete new family of unconventional superconductors is realized in nickelates[25, 26, 27, 28], spanning over a broad range of RP phases with [3, 6], [10], [10] and [7] requiring pressure, [17] and [29] in their reduced forms.
In this study, we report the single-crystal growth of , LaNiO by high-pressure synthesis in a Walker-type multi-anvil module, characterize its structure, perform its topotactic reduction, and demonstrate superconductivity in the resulting square-planar compound LaNiO. We systematically explore the growth conditions required to stabilize these high-order RP nickelates under pressure and to topotactically reduce the superconducting quadruple-layer phase. The resulting quadruple-layer nickelate constitutes only the second bulk, ambient-pressure nickelate superconductor reported to date, with a superconducting volume fraction reaching 60%.
Synthesis of Bulk High-Order RP Nickelates
The high-order Ruddlesden–Popper nickelate with , LaNiO, was synthesized in bulk form using the high-pressure high-temperature (HPHT) technique. By subsequently adjusting the La:Ni stoichiometry of the precursor, we also obtained LaNiO () and find results consistent with the recent report of this phase[15] (see Methods for details).
Single crystals were extracted from the recovered HPHT pellet and characterized by single-crystal X-ray diffraction (XRD). Reciprocal-space maps of the and planes (Figure 1a,Figure 1b) display sharp, well-separated reflections with no evidence of intergrowth, and structural refinement confirms that the quadruple-layer phase crystallizes in the monoclinic space group , with two crystallographically inequivalent Ni sites corresponding to the outer and inner NiO layers of the four-layer block (Figure 1f; refinement details in Table 1).

Figure 1. Synthesis of bulk high order RP nickelates. a,b Single-crystal XRD maps of the (h0l) and (hk0) planes of LaNiO. and c,d, of LaNiO. e, Comparative PXRD patterns of crushed single crystals of LaNiO and LaNiO with underlying Rietveld refinements; the right panel shows an enlarged view of a selected region showing the difference via the (114) reflex distance to the (020) reflex. f,g, Resulting structure of LaNiO and LaNiO. h,i, Temperature dependent resistance and magnetic susceptibility with underlying derivative of LaNiO and j,k, of LaNiO.
| Site | ||||||||||||
| (Å) | (Å) | (Å) | (°) | (Å) | (Å) | (Å) | (°) | (Å) | (Å) | (Å) | (°) | |
| 18.0299(10) | 5.4494(3) | 5.4234(4) | 98.635(6) | 3.9608(2) | 3.9608(2) | 32.752(3) | – | 21.806(2) | 5.4546(6) | 5.4020(7) | 96.773(12) | |
| La01 | 0.5 | 1 | 0.5 | 0.0025(4) | 1 | 1 | 0.39586(6) | 0.0047(7) | 0.73596(6) | 0.2492(3) | 0.6098(2) | 0.0056(4) |
| La02 | 0.28651(5) | 0.5007(2) | 0.89237(17) | 0.0038(4) | 1 | 0 | 0.5 | 0.0042(8) | 0.91192(5) | 0.2505(3) | 0.7042(2) | 0.0041(4) |
| La03 | 0.08011(6) | 0.0006(2) | 0.2893(2) | 0.0075(4) | 1 | 0 | 0.28942(7) | 0.0086(7) | 0.43405(6) | 0.7488(3) | 0.4702(2) | 0.0071(4) |
| Ni04 | 0.39225(11) | 0.5007(4) | 0.4456(3) | 0.0002(6) | 0.5 | 0.5 | 0.44949(14) | 0.0039(11) | 0.82258(12) | 0.7460(5) | 0.6579(5) | 0.0030(7) |
| Ni05 | 0.17416(15) | 0.4972(5) | 0.3364(4) | 0.0036(6) | 0.5 | 0.5 | 0.34954(17) | 0.0076(11) | 0.64348(15) | 0.7518(6) | 0.5674(5) | 0.0047(6) |
| O1 | 0.5 | 0.5 | 0.5 | 0.008(4) | 1 | 0.5 | 0.4494(5) | 0.007(4) | 1 | 0.7620(8) | 0.75 | 0.0014(8) |
| O3 | 0.2836(9) | 0.497(3) | 0.392(3) | 0.016(3) | 1 | 0.5 | 0.3486(6) | 0.008(4) | 0.6452(7) | 0.495(3) | 0.823(3) | 0.001 |
| O4 | 0.1754(10) | 0.245(3) | 0.066(3) | 0.021(4) | 1 | -0.5 | 0.25 | 0.005(5) | 0.7323(9) | 0.758(4) | 0.605(4) | 0.025(5) |
| O2 | 0.1726(8) | 0.742(2) | 0.100(2) | 0.008(3) | – | – | – | – | 0.9088(9) | 0.711(4) | 0.701(3) | 0.019(5) |
| O6 | 0.3914(12) | 0.735(3) | 0.190(4) | 0.033(5) | – | – | – | – | 1 | 0.5 | 0.5 | 0.005(4) |
| O5 | 0.3932(10) | 0.236(3) | 0.201(3) | 0.024(4) | – | – | – | – | 0.5429(8) | 0.718(3) | 0.513(3) | 0.013(4) |
| O7 | 0.0570(16) | 0.521(3) | 0.278(4) | 0.039(6) | – | – | – | – | 0.6345(12) | 1.010(4) | 0.312(4) | 0.025(5) |
| – | – | – | – | – | – | – | – | – | 1 | 1 | 0.5 | 0.084(18) |
| – | – | – | – | – | – | – | – | – | 0.8232(8) | 0.992(3) | 0.913(3) | 0.008(4) |
| – | – | – | – | – | – | – | – | – | 0.816(2) | 0.517(8) | 0.393(8) | 0.096(15) |
Table 1. Lattice parameters and refined atomic coordinates for , , and from single crystal XRD results with the corresponding maps shown in Figure 1 and 2.
To assess the phase homogeneity of the bulk product, crushed single crystals were further analysed by powder XRD (PXRD). A Rietveld refinement using a single phase reproduces the observed pattern well (Figure 1e), confirming that the HPHT synthesis yields a macroscopically homogeneous phase rather than an intergrowth of neighbouring RP members.
We find that pressure is a necessary condition for stabilizing these metastable higher- phases, but is not itself a fine control parameter for selecting a specific (see Supplementary Materials). This observation parallels literature reports on Sr-doped bilayer crystals obtained at both 3 and 20 GPa, where, notably, the lower-doped composition was synthesized at the higher pressure[30], suggesting that pressure primarily provides thermodynamic access to the metastable manifold rather than selecting a unique member within it. The decisive parameter for phase selection is instead the precursor: homogeneous cation mixing, achieved here via sol–gel preparation, is essential for obtaining a single-phase product (see Supplementary Information). By tuning the La:Ni ratio of the sol–gel precursor, we obtained phase-pure batches of either LaNiO or the quintuple-layer LaNiO, which crystallizes in the same space group (Figure 1c,Figure 1d and Table 1). The two phases are readily distinguished by PXRD (Figure 1e): the monoclinic axis, which runs along the stacking direction, increases from 18.0299(10) Å to 21.806(2) Å on going from to . The corresponding stacking period, , grows by 3.8 Å, in excellent agreement with the insertion of one additional NiO octahedral layer per perovskite block.
The temperature-dependent resistance of LaNiO () and LaNiO () is shown in Figure 1h and Figure 1j, respectively. Both compounds are metallic at high temperature, with resistance increasing approximately linearly with for currents majorly running in plane. On cooling, each sample passes through a resistance minimum near 50 K, consistent with defect scattering, which is mirrored by a weak upturn in the susceptibility near 30 K (Figure 1i,Figure 1k) attributable to a Curie-like tail from localized moments associated with oxygen vacancies or trace paramagnetic impurities. However, for LaNiO we see reproducibly an additional anomaly at 150 K, which shows an upturn followed by a downturn reminiscent of density-wave (DW) transitions found in bilayer and trilayer. Notably, we studied 3 different samples S1 is a larger piece, while S3 were single crystals measured in (S1) and out of plane (S3). This anomaly shows anisotropy and we see an anisotropy of metallicity, where S3 with current running majorly out of plane showing insulating behavior. We observe the expected trend of increasing metallicity with layer number , consistent with the end-member LaNiO () being a Stoner-enhanced Pauli-paramagnetic metal. This evolution is captured in transport, where equal contact layouts on equal sized samples where performed. We can see a reduction in resistance by a factor of 10. We can estimate the resistivity around 1 mΩ cm for LaNiO and 0.1 mΩ cm for LaNiO (note here the extreme anisotropy observed in LaNiO³¹). This is also clearly captured in the susceptibility: LaNiO displays clean Pauli-paramagnetic behaviour over a wide temperature range, whereas LaNiO also reveals that but in addition retains signatures of stronger electronic correlation, with an anomaly in susceptibility around 150 K indicating magnetic order. This trend falls well in line with our studies on = 1-3, who³² and the progression suggests that applying pressure to LaNiO is a promising route toward tuning or suppressing its possibly DW-like order, in analogy with the lower- members, whereas LaNiO already appears too metallic, and correspondingly too far from any density-wave instability, for this strategy to be as effective.
Reduction of Quadruple-Layer Nickelates
Topotactic reduction was carried out successfully while preserving single crystallinity throughout, as confirmed by the single-crystal XRD maps in Figure 2a,Figure 2b and by backscattered electron imaging (BSE) shown in Figure 2d,Figure 2f. This outcome contrasts markedly with the reduction of perovskite LaNiO, where extensive domain formation severely limits the retained crystallinity¹⁹,³³. We note that the incongruent melting behaviour characteristic of higher-order RP phases typically produces finite NiO inclusions in the as-grown crystals²⁰, which upon reduction give rise to ferromagnetic Ni metal impurities³⁴; identifying and accounting for this contribution is therefore essential when interpreting the magnetic data below.

Figure 2. Reduction of quadruple-layer nickelates. a,b, single crystal XRD maps of (h0l), and (hk0) of LaNiO. c, relative −axis change by for ²², 3²³, 4 and 5 revealing a clear linear trend and full reduction for our higher order nickelates. d,f, BSE picture of a LaNiO and a LaNiO single crystal. e,g, Visualization of the crystal structure of LaNiO and reduced LaNiO. h, PXRD of crushed single crystals with underlying Rietveld refinement of as grown LaNiO and topotactic reduced LaNiO each refined with a single phase. i, EDX spectra on a LaNiO in comparison to LaNiO. j TGA of as grown LaNiO in comparison to reduced single crystals of LaNiO.
Comparing the relative -axis contraction of our reduced LaNiO crystals with literature values for other (Figure 2c), we find a strikingly linear reduction trend across the series, where the lower- members were previously characterized by neutron diffraction [22, 23], directly demonstrating the migration of rock-salt layer oxygen toward fluorite-type coordination: the structural rearrangement we illustrate schematically in Figure 2e,Figure 2g. The reduced phase, nominally La₅Ni₄O₁₀, is well fitted in the tetragonal space group by PXRD (Figure 2h; structural details in Table 1), consistent with the single-crystal results and confirming a homogeneous reduced phase free of residual higher-oxygen impurities. Energy-dispersive X-ray spectroscopy (EDX) confirms an unchanged La:Ni ratio of 5:4 before and after reduction (Figure 2i); the relative change in oxygen content can also be qualitatively tracked from the reduced intensity of the O K line at 0.52 keV. Quantitative EDX analysis of approximately 20 points yields a reduced-phase stoichiometry of La₅.₀(₃)Ni₃.₉(₆)O₉.₅(₁₇), consistent within error with full reduction to the square-planar composition.
Thermogravimetric analysis (TGA, Figure 2j) independently corroborates this conclusion. Heating drives both the as-grown and reduced samples toward complete decomposition into 2.5La₂O₃ + 4Ni (molar mass 1049.29 g mol⁻¹). The as-grown sample loses 7.6 wt%, corresponding to an oxygen content of , in good agreement with the nominal La₅Ni₄O₁₃ stoichiometry. The reduced sample loses only 3.11 wt%, corresponding to
. Taken together with the EDX results, these data establish that the reduction proceeds essentially to completion, yielding a fully reduced La₅Ni₄O₁₀ phase that is already approaching the onset of thermal decomposition.
Magnetic Screening of La₅Ni₄O₁₀ Reduced Under Varying Conditions
Superconductivity was observed in reduced La₅Ni₄O₁₀ crystals obtained both by direct and by indirect reduction inside sealed ampoules (see Methods). Remarkably, irrespective of the precise reduction conditions, reduction time, or sample mass, every superconducting crystal displays the same transition temperature, K, indicating that is an intrinsic property of the fully reduced phase rather than a parameter sensitive to processing details. Field-dependent magnetization at 2 K shows a clear hysteresis loop (Figure 3a), from which we estimate a lower critical field of Oe, close to the value reported for La₆Ni₅O₁₂ [15]. In the optimally reduced samples, the superconducting volume fraction exceeds 60%, as estimated from the zero-field-cooled (ZFC) shielding signal (Figure 3b) at 10 Oe.

Figure 3. Magnetic screening of La₅Ni₄O₁₀ reduced under varying conditions. a, Field dependent magnetization at 2 K for ZFC and then cycling through -5000 to 5000 Oe. b, Temperature dependent magnetization of optimal reduced La₅Ni₄O₁₀ measured at varied field in ZFC and FC manner at 10 (black), 50 (red), 100 (blue), 500 (green) and 1000 Oe (purple). c, Temperature dependent magnetization of over-reduced crystals revealing the background signal for various samples of similar masses as optimal reduced crystals. The inset shows the mounted Tetragonal platelets of approximately 200 m of La₅Ni₄O₁₀ on a quartz squid holder, which leads to a marginally small demagnetization factor. d,e,f Reproducibility of superconductivity for various crystals with their mass noted visualized in the Temperature dependent magnetization measured at an external field of 50-100 Oe of ampoule 1, 2 and 3 respectively.
To evaluate the reproducibility of the reduction process, we measured multiple single crystals of comparable size, typically 200 × 200 × 100–200 m (Figure 2d), corresponding, given the density of 7.025 g cm⁻³, to a mass of roughly 0.02 mg per crystal. To minimize spurious background contributions, crystals were mounted on quartz holders using a minimal amount of grease (inset of Figure 3c). We isolated and subtracted the ferromagnetic Ni and diamagnetic background contributions by separately measuring several over-reduced crystals under identical temperature and field conditions (Figure 3c); because the crystals were measured with the field in the plane of the platelets, the associated demagnetization factor is minimal. Most samples were additionally confirmed by single-crystal XRD to retain the quadruple-layer structure prior to magnetic screening.
For direct-contact reduction over 100 h (Figure 3d), all crystals of approximately 0.01 mg show a consistent ZFC shielding signal at 2 K of about ; larger crystals show a reduced volume fraction, consistent with the reduction depth being limited by diffusion and therefore sample-volume dependent (see Supplementary Materials). Extending the direct-contact reduction time to 130 h (Figure 3e) suppresses the shielding response to about for samples of comparable mass, indicating that prolonged exposure to the reducing agent degrades, rather than improves, the superconducting volume fraction, likely through over-reduction or the onset of decomposition suggested by the TGA data above.
Indirect-contact reduction (Figure 3f) produces larger sample-to-sample variation: a subset of crystals reaches shielding fractions comparable to the best direct-reduced samples, but the majority show substantially weaker signals, reflecting a less controlled and more spatially inhomogeneous nature of vapor-phase CaH2 reduction, possibly showing an even stronger scaling by volume.
Transport and Thermodynamic Properties of Bulk La5Ni4O10 Superconductor
The superconducting transition is clearly resolved in temperature-dependent resistance measurements (Figure 4a-Figure 4c). All crystals from the same growth and reduction batch exhibit a superconducting transition, although the absolute normal-state resistance varies between individual samples, as expected for crystals of different geometry and residual defect density. Notably, the contacting for reduced samples also has an effect on the final result and in Figure 4b, we show a sample that despite imperfect contacts as non-cured shows zero resistance. When cured epoxy contacts are used we observe finite resistance values (Figure 4a-Figure 4c). On cooling, all samples show a resistive anomaly near 98 K, below which the resistance increases, suggestive of a transition into either a partially gapped density-wave state or metal-insulator like transition. This transition shows hysteretic behaviour upon heating and cooling, but reveals no field dependence (Figure 4e). This upturn is weak in samples S2 and S3 but much more pronounced in S4, where increases to roughly four times ; the microscopic origin of this anomaly, and of its pronounced sample-to-sample variability, remains to be clarified, but plausibly reflects differences in local oxygen stoichiometry or strain introduced during reduction. For sample S2, the superconducting onset (90% of ) occurs at , and the resistance reaches 50% of at 16.5 K. Below the transition, the resistance drops by about 98%, leaving a small residual of roughly 2% of down to 2 K, which likely stems from surface decomposition of the sample in air.

Figure 4. Transport and thermodynamic properties of bulk La5Ni4O10 superconductor a, Temperature dependence of the normalized resistance for three other samples S1, S2, S3 and S4 at zero field from 300 K to 2 K, with cured epoxy silver paint. Inset: optical micrograph of S2 with four contacts. b, Temperature dependence of a sample showing zero resistance in various applied fields untreated silver paint. c, Shown is the Low-temperature resistance under magnetic fields of 0–14 T applied out of plane (top) and in plane (bottom). c, Upper critical field determined from the 90% criterion for both field orientations. Solid and dashed lines are fits to the WHH and GL models, respectively; the dotted line marks the weak-coupling Pauli limit . e, Temperature dependence of the resistance through the 90 K transition for heating and cooling in 0 and 14 T. f, Specific heat of La5Ni4O13 and La5Ni4O10 from 2 to 250 K; the dashed line is the Dulong–Petit value . Inset: versus at low temperature with a fit to . g, below 30 K. The solid line is the normal-state contribution , obtained from a Debye-plus-two-Einstein phonon model constrained by entropy conservation; the dashed line is the phonon part alone. h, Electronic specific heat with a fit to a two-component -model including a residual term (solid line). The shaded areas show the contributions of the two components above , each with a Gaussian distribution of . The superconducting volume fraction is estimated as .
Magnetic fields suppress superconductivity more strongly when applied perpendicular () than parallel () to the NiO planes (Figure 4b,Figure 4c,Figure 4d), but the anisotropy is small. The ratio remains finite near , and is close to linear: a fit to gives –, rather than expected for a two-dimensional superconductor. In nickelate thin films, by contrast, the finite thickness leads to and a diverging anisotropy. The Ginzburg–Landau coherence lengths obtained from the initial slopes, and – depending on the criterion (Supplementary Information), exceed both the spacing of the NiO planes within a four-layer block (0.33 nm) and the distance between neighbouring blocks across the La–O spacer (0.65 nm). LaNiO is thus an anisotropic, three-dimensional superconductor in which the four-layer blocks are appreciably coupled along . For in-plane fields, the orbital-limited extrapolation gives , well above the weak-coupling Pauli limit of about 32 T[35, 36], which makes LaNiO a good candidate to examine the presence or absence of Pauli-limiting behaviour in high magnetic fields.
Specific heat provides thermodynamic evidence for bulk superconductivity (Figure 4f–Figure 4h). shows a clear jump at (Figure 4g), and the electronic specific heat, , rises sharply below about 18.5 K, close to the resistive onset (Figure 4h). The jump, , is a large fraction of the normal-state Sommerfeld coefficient, . At low temperatures, with (Figure 4f, inset), indicating that part of the electronic entropy remains normal. The superconducting part thus contributes , corresponding to a superconducting volume fraction of , in agreement with the magnetic shielding. For this fraction, the weak-coupling BCS ratio predicts a jump of about , close to the measured value. The size of the jump is therefore consistent with bulk superconductivity in about half of the sample.
The normal-state contribution, , was determined with a Debye-plus-two-Einstein phonon model constrained by conservation of the electronic entropy (Supplementary Information). As an independent check of the phonon background, the unreduced LaNiO shows no anomaly in this temperature range, and its agrees with the fitted within 3% between 10 and 18 K (Figure 4g). Varying the fitting ranges changes by only a few percent (53–57%; Supplementary Information).
The transition is broad and shows a main maximum near 16 K and a shoulder near 11 K (Figure 4h). Such a broad transition likely reflects an inhomogeneous degree of oxygen reduction within the crystals, consistent with the dependence of the superconducting fraction on the reduction conditions discussed above. A detailed analysis of the shape of the anomaly, including a comparison of two-component and two-gap models, is given in the Supplementary Information.
Discussion and Conclusion
In this work, we overcome the ambient-pressure thermodynamic limit on the perovskite layer number in RP nickelates through HPHT synthesis. Combined with careful control of precursor homogeneity, this approach yields phase-pure bulk crystals and allows itself to be used as a synthetic tuning parameter. Topotactic reduction of the quadruple-layer () compound LaNiO yields the square-planar phase LaNiO, which we show to be superconducting with .
The superconductivity is unambiguously a bulk property of these crystals: transport, magnetization, and specific heat all register the transition at essentially the same temperature, with superconducting volume fractions of about 60% from zero-field-cooled shielding and 56% from , two independent estimates that agree remarkably well. The reduction step and air sensitivity appear to be the principal factor currently limiting the superconducting fraction: superconductivity is optimized only within a narrow window of reduction time, depends on sample mass through diffusion-limited kinetics, and is most spatially homogeneous when crystals are in direct physical contact with the reducing agent. The measurement of a two-component structure of the specific-heat anomaly, with and , can be explained by regions of differing local oxygen content (which must then cluster and not randomly distribute possibly due to surface and bulk oxygen contents) within the crystal rather than by an intrinsic multi-gap pairing state.
The availability of bulk single crystals also allows the superconducting anisotropy to be measured without the finite-thickness confinement and substrate effects that affect thin films. The anisotropy is small: remains finite as , is close to linear, and the out-of-plane coherence length exceeds the separation of neighbouring four-layer blocks. Together with the four NiO layers per block, this indicates a substantial interlayer coupling across the La–O spacer, making LaNiO an anisotropic but three-dimensional superconductor.
More broadly, these results establish the reduced RP series as a viable bulk platform for nickelate superconductivity at ambient pressure. In this family, the layer number simultaneously controls the electronic dimensionality and the nominal Ni valence, , entirely without chemical doping, a self-doping mechanism with no direct analogue in the cuprates. Now that HPHT synthesis has lifted the long-standing restriction on in bulk RP nickelates, the series can be explored systematically in both directions: toward larger , to map out the evolution of and dimensionality as the system approaches the infinite-layer limit, and through rare-earth substitution at fixed , to introduce chemical pressure or additional doping as a further means of enhancing . The availability of bulk single crystals, rather than films or diamond-anvil-cell samples, also opens the door to probes that are difficult or impossible to apply otherwise: inelastic neutron scattering (for this crystal sizes need to be improved) to map the spin and lattice dynamics, muon spin rotation to probe the superfluid density and pairing symmetry, and angle-resolved photoemission on cleaved surfaces. Such measurements should help to clarify the microscopic pairing mechanism in the nickelates and to position this emerging family of superconductors within the broader landscape of unconventional superconductivity alongside the cuprates and iron-based materials. The coexistence of the superconductivity with a anisotropic MIT around 90 K in all our samples shows that the compound resides in the lower doping regime of the dome and reveals clear analogy to cuprates.
Methods
Precursor synthesis
Polycrystalline precursors were prepared by a sol–gel route. and , in the La:Ni molar ratio 6:5, were dissolved in a 1:1 (by volume) mixture of concentrated and deionized water. Citric acid was added at twice the mass of the targeted precursor (for example, 5 g of citric acid for 2.5 g of precursor). The solution was heated at 120 °C until the water had evaporated and a gel had formed. The temperature was then raised slowly to 230 °C — a rate chosen to moderate the exothermic auto-combustion of the gel — and held overnight. The resulting ash-like powder was ground for 1 h and heated in flowing oxygen to 800 °C at 50 °C h, with a 12 h dwell, to fully remove residual organic components. The product was ground a second time for 1 h prior to use.
High-pressure crystal growth
Crystals of were grown in a Walker-type multi-anvil module (Voggenreiter) at 7 GPa, using an 18/11 octahedral pressure assembly (18 mm octahedron edge length, 11 mm anvil truncation edge length) and a heater, which allowed temperatures sufficient to melt the La–Ni–O system to be reached. The precursor, prepared with a slightly Ni-rich La:Ni ratio of 6:5 relative to the target phase to promote crystal growth from the melt, was sealed in a Pt capsule. After the target pressure was reached, the sample was heated to 1600 °C at approximately 100 °C min, then cooled uniformly to 1480 °C over 8 h, at which point it was quenched to room temperature; the pressure was subsequently released slowly to ambient conditions. Single crystals were extracted from the recovered pellet under an optical microscope. We additionally tested the sensitivity of the growth to pressure and precursor composition: lowering the synthesis pressure to 5 GPa, and separately adjusting the precursor stoichiometry to a La:Ni ratio of 5:4 by the addition of , each produced only minor changes in crystallinity while the product phase remained . In contrast, preparing a new precursor with a La:Ni ratio of 7:6 yielded phase-pure five-layer crystals, demonstrating that precursor stoichiometry, rather than pressure, is the dominant parameter controlling which RP member is obtained.
Topotactic reduction
Single crystals selected from the high-pressure product were reduced using in evacuated, sealed quartz ampoules[19]. Each ampoule contained 200 mg of , and three reduction conditions were compared. In ampoules 1 and 2, crystals were scattered directly onto the powder, placing them in direct physical contact with the reducing agent. In ampoule 3, crystals were spatially separated from the , allowing reduction to proceed only via the gas phase (indirect reduction). All ampoules were evacuated to mbar, sealed, and annealed at 300 °C for 100 h (ampoules 1 and 3) or 130 h (ampoule 2).
Diffraction
Powder XRD data were collected on a Rigaku Miniflex diffractometer in Bragg–Brentano geometry using Cu K radiation with a Ni filter. Rietveld refinements were performed using TOPAS V7 software[37]. Single-crystal XRD was performed on a Rigaku XtaLAB mini II diffractometer using Mo K radiation. Data were processed with CrysAlis(Pro), and final structure refinements were carried out using Olex2 with SHELX.
Transport measurements
Transport measurements were carried out in a Quantum Design Physical Property Measurement System (PPMS, 14 T) using a four-probe van der Pauw configuration. Electrical contacts were made with silver epoxy (EPO-TEK H20E), cured at 350 °C under argon for 10 minutes. Resistance was measured on cooling under fixed magnetic fields of 0–14 T, applied either perpendicular or parallel to the planes.
Magnetization measurements
Magnetic susceptibility measurements were performed on single crystals using a Quantum Design MPMS3 SQUID magnetometer. Specimens were mounted on a quartz holder using a minimal amount of grease (inset of Figure 3c). Temperature-dependent magnetization, , was measured in direct-current (DC) mode between 1.8 and 300 K, and between 300 and 800 K using the oven option. Field-dependent magnetization, , was measured in vibrating sample magnetometer (VSM) mode.
Specific heat measurements
Specific heat was measured using the heat-capacity option of a PPMS, which employs the thermal-relaxation technique. For the superconductivity measurements, several single crystals were placed together on the sapphire sample platform using Apiezon N grease. The addenda heat capacity (platform and grease) was measured separately over the same temperature range and subtracted from the raw data. Each relaxation curve was analysed using the two- model [38]. Data were collected in zero applied field between 1.8 and 300 K, with every temperature point measured three times and the results averaged.
References
References
- [1]Khasanov, R. et al. Pressure-enhanced splitting of density wave transitions in La₃Ni₂O₇−δ. Nature Physics 21, 430–436, DOI: 10.1038/s41567-024-02754-2 (2025).DOI
- [2]Zhang, J. et al. Large orbital polarization in a metallic square-planar nickelate. Nature Physics 13, 864–869, DOI: 10.1038/nphys4149 (2017).DOI
- [3]Sun, H. et al. Signatures of superconductivity near 80 K in a nickelate under high pressure. Nature 621, 493–498, DOI: 10.1038/s41586-023-06408-7 (2023).DOI
- [4]Puphal, P. et al. Unconventional Crystal Structure of the High-Pressure Superconductor La₃Ni₂O₇. Phys. Rev. Lett. 133, 146002, DOI: 10.1103/physrevlett.133.146002 (2024).
- [5]Wang, N. et al. Bulk high-temperature superconductivity in pressurized tetragonal La₂PrNi₂O₇. Nature 634, 579–584, DOI: 10.1038/s41586-024-07996-8 (2024).DOI
- [6]Zhu, Y. et al. Superconductivity in pressurized trilayer La₄Ni₃O₁₀−δ single crystals. Nature 631, 531–536, DOI: 10.1038/s41586-024-07553-3 (2024).DOI
- [7]Shi, M. et al. Pressure induced superconductivity in hybrid ruddlesden–popper La₅Ni₃O₁₁ single crystals. Nature Physics 21, 1780–1786, DOI: 10.1038/s41567-025-03023-3 (2025).DOI
- [8]Li, F. et al. Design and synthesis of three-dimensional hybrid ruddlesden-popper nickelate single crystals. Physical Review Materials 8, DOI: 10.1103/physrevmaterials.8.053401 (2024).DOI
- [9]Chen, X. et al. Polymorphism in the Ruddlesden–Popper Nickelate La₃Ni₂O₇: Discovery of a Hidden Phase with Distinctive Layer Stacking. Journal American Chemical Society 146, 3640–3645, DOI: 10.1021/jacs.3c14052 (2024).DOI
- [10]Huang, C. et al. Superconductivity in Monolayer-Trilayer Phase of La₃Ni₂O₇ under High Pressure. Physical Review X 16, DOI: 10.1103/jc35-gj2l (2026).
- [11]Wang, H., Chen, L., Rutherford, A., Zhou, H. & Xie, W. Long-Range Structural Order in a Hidden Phase of Ruddlesden–Popper Bilayer Nickelate La₃Ni₂O₇. Inorganic Chemistry 63, 5020–5026, DOI: 10.1021/acs.inorgchem.3c04474 (2024).DOI
- [12]Khasanov, R. et al. Magnetism of the alternating monolayer-trilayer phase of La₃Ni₂O₇, DOI: 10.48550/ARXIV.2512.22371 (2025).
- [13]Zhang, M. et al. Spin-density-wave transition in monolayer-trilayer La₃Ni₂O₇ single crystals. Materials Today Physics 67, 102175, DOI: https://doi.org/10.1016/j.mtphys.2026.102175 (2026).
- [14]Li, Z. et al. Epitaxial growth and electronic structure of ruddlesden–popper nickelates (Laₙ₊₁NiₙO₃ₙ₊₁, n = 1–5). APL Materials 8, 091112, DOI: 10.1063/5.0018934 (2020).DOI
- [15]Shi, M. et al. Bulk superconductivity in quintuple-layer square-planar nickelate La₆Ni₅O₁₂, DOI: 10.48550/arXiv.2610.20249 (2026). 2610.20249.DOI
- [16]Pan, G. A. et al. Superconductivity in a quintuple-layer square-planar nickelate. Nature Materials 21, 160–164, DOI: 10.1038/s41563-021-01142-9 (2022).DOI
- [17]Pan, G. A. et al. Superconducting phase diagram of multilayer square-planar nickelates. Science 392, 1390–1395, DOI: 10.1126/science.adp4440 (2026).DOI
- [18]Li, Q. et al. Absence of superconductivity in bulk Nd₁₋ₓSrₓNiO₂. Communications Materials 1, 16, DOI: 10.1038/s43246-020-0018-1 (2020).DOI
- [19]Puphal, P. et al. Topotactic transformation of single crystals: From perovskite to infinite-layer nickelates. Science Advances 7, eabl8091, DOI: 10.1126/sciadv.abl8091 (2021).DOI
- [20]Puphal, P. et al. Phase formation in hole- and electron-doped rare-earth nickelate single crystals. APL Materials 11, 081107, DOI: 10.1063/5.0160912 (2023).DOI
- [21]Hu, K. et al. Atomic-scale disorder and reconstruction in bulk infinite-layer nickelates lacking superconductivity. Nature Communications 15, 5104, DOI: 10.1038/s41467-024-49533-1 (2024).DOI
- [22]Poltavets, V. V. et al. La₃Ni₂O₆: A new double T′-type nickelate with infinite Ni¹⁺₁₊ₓO₂ layers. Journal American Chemical Society 128, 9050–9051, DOI: 10.1021/ja063031o (2006).DOI
- [23]Poltavets, V. V. et al. Crystal structures of Ln₄Ni₃O₈ (Ln = La, Nd) triple layer T′-type nickelates. Inorganic Chemistry 46, 10887–10891, DOI: 10.1021/ic701480v (2007).DOI
- [24]Zhang, J. et al. Spin stripe order in a square planar trilayer nickelate. Physical Review Letters 122, 247201, DOI: 10.1103/PhysRevLett.122.247201 (2019).DOI
- [25]Puphal, P., Schäfer, T., Keimer, B. & Hepting, M. Superconductivity in infinite-layer and ruddlesden–popper nickelates. Nature Reviews Physics 8, 70–85, DOI: 10.1038/s42254-025-00898-2 (2025).DOI
- [26]Wang, B. Y., Lee, K. & Goodge, B. H. Experimental progress in superconducting nickelates. Annu. Rev. Condens. Matter Phys. 15, 305–324, DOI: 10.1146/annurev-conmatphys-032922-093307 (2024).DOI
- [27]Wang, Y. et al. Recent progress in nickelate superconductors. National Science Review 12, DOI: 10.1093/nsr/nwaf373 (2025).DOI
- [28]Zhang, Y., Lin, F.-L., Maier, T. A. & Dagotto, E. Superconductivity in ruddlesden–popper nickelates: a review of recent progress, focusing on thin films. Reports on Progress in Physics 89, 096501, DOI: 10.1088/1361-6633/ae9aae (2026).DOI
- [29]Li, D. et al. Superconductivity in an infinite-layer nickelate. Nature 572, 624–627, DOI: 10.1038/s41586-019-1496-5 (2019).DOI
- [30]Xu, M. et al. Pressure-dependent “insulator–metal–insulator” behavior in Sr-doped La₃Ni₂O₇. Advanced Electronic Materials 10, 2400078, DOI: 10.1002/aelm.202400078 (2024).DOI
- [35]Clogston, A. M. Upper limit for the critical field in hard superconductors. Physical Review Letters 9, 266–267, DOI: 10.1103/PhysRevLett.9.266 (1962).DOI
- [36]Chandrasekhar, B. S. A note on the maximum critical field of high-field superconductors. Applied Physics Letters 1, 7–8, DOI: 10.1063/1.1777362 (1962).DOI
- [37]Coelho, A. A. TOPAS and TOPAS-Academic: an optimization program integrating computer algebra and crystallographic objects written in C++. Journal Applied Crystallography 51, 210–218, DOI: 10.1107/S1600576718000183 (2018).DOI
- [38]Hwang, J. S., Lin, K. J. & Tien, C. Measurement of heat capacity by fitting the whole temperature response of a heat-pulse calorimeter. Review Scientific Instruments 68, 94–101, DOI: 10.1063/1.1147722 (1997).DOI