betaFeedback

WorksPhysics and Astronomy

Dichotomous electronic system in a bilayer Ni1+^{1+} nickelate

Abstract

"Infinite layer" nickelates R{\cal R}NiO2_2 (R{\cal R}=rare earth), having empty apical O sites, become superconducting upon hole doping, stimulating research into the related sequence Ndn+1_{n+1}Nin+p^{+p}_nO2n+2_{2n+2}, formal charge state pp=1+1n\frac{1}{n}, nn=2,3,4,...., with the nn=5 member being found to be superconducting. The two layer system La3_3Ni2_2O7−δ_{7-δ}, with δδ=0,12\frac{1}{2},1 (pp=2.5,2,1.5) approaches the peak in the nickelate superconducting dome but shows no superconductivity. Newly reported La3_3Ni2_2O5_5F reaches the Ni1+^{1+} goal while, as we show, introducing a partially occupied electron band E∗E^*, based on an interstitial density that extends over the three open "apical" layers and leads to a single cylindrical electron Fermi surface giving self-doping. The commonly inert Ni dxz,dyzd_{xz},d_{yz} orbitals partner with interstitial E∗E^* to provide an incipient non-analytic Dirac point, with the critical point being reachable by pressure or further F insertion. The E∗E^* electron cylinder and the conventional Ni dpσdpσ hole carriers combine to provide a two-fluid dichotomy of hole and electron quasiparticles, affecting normal state properties that should verify the dichotomous aspect of transport.

Background

Electride crystals entered the lexicon of solid state physics and chemistry as an interstitial density, containing one electron and not ascribable to any atomic orbital: an ‘isolated’ electron acting as an anion e−e^{-}. These occurred in molecular crystals, where empty spaces could provide a suitable potential, and the available space was sufficient to obviate a positive energy from confinement. An example that began the excitement, and potential for novel properties, was Dye’s discovery [1, 2], and a subsequent calculation [3], that verified the interstitial density. Two crown ether molecules (15-crown-5) encased Cs, expelling its weakly bound valence electron into a space between the roughly spherical pair of crown ether molecules – the Cs+^{+}(1-crown-5)2_2:e−^{-} electride. This crystal provided a singly occupied (magnetic) electride gap state [3], closely related to color center defects. In the intervening decades several review articles have related how molecular electrides display unexpected properties: electrical conduction, magnetism (namely, the localized electron mentioned above), surface catalytic activity, and even superconductivity (SC) [4, 5, 6]. In the meantime, interest has turned more toward inorganic electrides, with application of the term ‘electride’ having undergone generalization, being adopted whenever there is an unexpected density maximum not ascribable to an atomic or molecular orbital. Alkali metals under pressure have assumed unanticipated structures with ‘electride’ regions [7, 8]. At layered van der Waals interfaces an interstitial density can arise [9, 10]. The advent of rapid computational methods has led to a rise in predicted inorganic electrides, often at high pressure. For example, it is predicted that at high pressures of the Earth inner core, electride formation can stabilize the hcp iron lattice [11].

These uses of ‘electride’ have been applied where the amount of localized charge is only a fraction of one electron, and to materials where the interstitial charge is delocalized. Their properties extend the use of ‘electride’ to a much broader class of crystalline solids, and distinguishing them from materials with related properties can become unclear. So far relatively few proposed electrides have been realized in the laboratory [4, 5, 9].

Nickelates. While at first glance appearing to be a distinct topic from the above, we extend the vigorous activity following the discovery [12] of long-sought nickelate SC in hole-doped ‘infinite layer’ nickelate (ILN) NdNiO2_2, with formal Nip+^{p+} valence p≈1.2p \approx1.2. SC was next found in the n=5n=5 member of the series Ndn+1_{n+1}Nin_nO2n+2_{2n+2} system [13], formal valence p=1+1np=1+\frac{1}{n}, with the same formal value p=1.2p=1.2. This series is now one focus of attention, a particular point being that the p=1p=1 (Ni1+^{1+}) formal charge state has been challenging to stabilize. Recently the n=4n=4–7 systems have been found to be superconducting, providing another SC dome similar to that of hole-doped ILN, peaking at p=1.2p=1.2 [14].

A related system is La3_3Ni2_2O7−δ_{7-\delta}, which for the varying O concentration δ=0,12,1\delta=0,\frac{1}{2},1 gives p=2.5p=2.5, 2, 1.5, respectively. La3_3Ni2_2O6_6 [15] has received the most attention, experimentally [16, 17, 18, 19] (it is insulating, possibly due to charge disproportionation), in the peculiarities of electronic structure [20, 21, 22], and study of possible SC [23, 24, 25, 26]. The properties of these compounds vary greatly and include an insulating regime, with no SC [19]. If the anion concentration could be reduced further, to δ=1.5\delta=1.5, the Ni1+^{1+} charge state would be achieved. This goal has been reached with the announcement by Wernert et al. [27] of La3_3Ni2_2O5_5F=La3_3(NiO2_2)2_2OF, replacement of one O ion in the blocking layer by F, thus the removal of 12\frac{1}{2} additional hole per Ni, p=1p=1.

From density functional theory studies we have found that La3_3Ni2_2O5_5F has an interstitial density that provides a different viewpoint on the behavior of, and independent insight into, interstitial density upon physical behavior. At the electronic carrier level, La3_3Ni2_2O5_5F provides two types of quasiparticles: one is the usual nickelate dpσdp_\sigma Fermi surface (FS), and the other is essentially a plane-wave-like density confined to three channels. The spectrum of low-energy excitations (hence transport and far-IR properties) consists of each of these two distinctly different – dichotomous – quasiparticles, and their coupling via inter-FS scattering.

Results. Unexpectedly, we find that La3_3Ni2_2O5_5F contains a ‘blocking layer’ seemingly typical of related cuprates and nickelates, but that results in extreme blocking, indicated by all 3d3d bands of interest having negligible kzk_z dispersion – uncoupled bilayers. The bottom of the two dpσdp_\sigma bands (indistinguishable in plots) disperse by ∼0.5\sim0.5 meV, with similar (lack of) dispersion of the interstitial band (see below) at −0.5-0.5 eV at MM – truly isolated electronic layers.

Representing the disordered O0.5_{0.5}F0.5_{0.5} by the virtual crystal average X, the layer stacking sequence, shown in the left panel of Figure 1, is La3_3Ni2_2O5_5F = [(La2)X(La2)][(NiO2_2)(La1)(NiO2_2)], i.e., a blocking layer and a NiO2_2 bilayer. The virtual crystal treatment of O/F is justified because (i) O and F are neighbors in the periodic table, and (ii) the closed shell 2p2p states in this layer are strongly bound (∼5\sim5–7 eV) and serve mainly to specify the electron doping of the NiO2_2 layers. The formal Ni1+^{1+} valence is isovalent to Cu2+^{2+} but unusual, and very different in behavior [28, 29, 30, 31]. The methods of calculation, using WIEN2k [32], are described in a separate study of the magnetic behavior of La3_3Ni2_2O5_5F [31].

Tetragonal structure and isosurface plot of La3Ni2O5F

Figure 1. Left: Tetragonal I4/mmmI4/mmm structure of La3Ni2O5F\mathrm{La}_{3}\mathrm{Ni}_{2}\mathrm{O}_{5}\mathrm{F}, consisting of double NiO2_2 infinite layers isolated by the La2 and La1 layers. Right: Isosurface plot of the occupied E∗E^{*} wavefunction ∣Ψ(r)∣|\Psi(r)| at MM (see text). The E∗E^{*} density, arising from a single band, consists of roughly equal parts in each of the three La-apical site layers, with connections (see text).

Some basic points related to the La3_3Ni2_2O5_5F bands and magnetic tendencies have been noted separately [31], we elaborate here. In discussing the electronic structure, it is most instructive to begin above EF_F in the band structure, shown with orbital characters in Figure 2, with more detail shown in the supplemental material (SM) [33].

Nonmagnetic band structure of La3Ni2O5F

Figure 2. Nonmagnetic band structure of La3_3Ni2_2O5_5F. (a) Ni 3dd orbital decompositions are distinguished by color. The blue arrows indicate the large splitting of the dz2d_{z^2} orbitals at the Γ\Gamma point due to coupling with, or through, the interstitial (II) density in the intervening La1 layer. Points are as follows: Except for the usual dpσdp_\sigma band, the other Ni dd bands lie at −1.3±0.5-1.3 \pm0.5 eV, hence are electrically inactive; t2gt_{2g} splitting is small. (b) Fatband representation of the II character, that is, the amplitude of the state not falling within any atomic sphere. The E∗E^* band dips down from 2 eV at Γ\Gamma to −0.5-0.5 eV at MM. See the text for other II density-related features.

  • The usual focus, the pair of dpσdp\sigma bands, forms a twofold degenerate (except for some splitting near XX) pair of large cylindrical hole FSs around MM. The FSs are displayed in Figure S1 of SM [33].

  • A single “E∗E^{*}” band dips from 1.5–2 eV above EFE_{\mathrm{F}} to −0.5 eV at MM, with linear bands (except extremely near MM) along both symmetry directions from MM, up to +1 eV. This E∗E^{*} band derives from the II density, emphasized in the right panel of Figure 1. Being unassociated with any atom, this is essentially a 2D plane-wave-like band, confined by avoidance of the blocking layer and O2−\mathrm{O}^{2-} ions. This band forms an electron cylinder around MM enclosing 0.18 electrons. As a result, the dpσdp\sigma bands are not half filled, with each Ni ion valence becoming +1.09. This self-doping pushes La3_{3}Ni2_{2}O5_{5}F toward the SC dome of Ni1+^{1+} nickelates [13, 34, 35, 36] and the itinerant E∗E^{*} carriers can be expected to hinder magnetic ordering.

  • The dz2d_{z^{2}} orbitals, though lying 1–2 eV below EFE_{\mathrm{F}} with small dispersion, are even-odd split by ∼1\sim1 eV throughout the zone due to strong coupling to and through the II density, see Figure 2. The coupling produces a lower even symmetry hybrid band dz2ed_{z^{2}}^{e} with large II density in the La1 layer apical site, and an upper odd partner dz2od_{z^{2}}^{o} with II density maxima in the La2 layers, both pictured in Figure 3(a,Figure 3b) and Figure 3(d,Figure 3e).

  • The II density avoids the dx2−y2d_{x^{2}-y^{2}} and O pσp_{\sigma} lobes, minimizing coupling between the dpσdp\sigma and E∗E^{*} bands, thereby producing a dichotomous electronic structure.

  • Less evident is a shadow feature involving the dxz,dyzd_{xz},d_{yz} bands lying 1–2 eV below EFE_{\mathrm{F}} with only 1 eV dispersion. Upon approaching MM from both symmetry directions, there is a dxz,dyzd_{xz},d_{yz} band approaching the cluster of t2gt_{2g} bands at MM, with the E∗E^{*} band just above at −0.5 eV. This “D∗D^{*}” band is parallel to the E∗E^{*} band, with the same velocity and linearity.

  • This D∗D^{*} band acquires some II density character upon nearing MM; in complementary fashion, the E∗E^{*} wave function at EFE_{\mathrm{F}} acquires dxz,dyzd_{xz},d_{yz} character creeping in. These features (wave-function plots are provided in Figure S7 of the SM [33]) indicate the source of the D∗D^{*}-E∗E^{*} pairing that gives an incipient Dirac point. Additional hole doping or compressive strain (see below) produces a nonanalytic critical Dirac point with unusually distinct, physically separated orbital character partners above and below the degeneracy. Moreover, this coupling provides the ‘glue’ that cements the three layers of II density into a single entity and band.

Wave-function isosurfaces and projected color-scale plots for the $d_{z^{2}}$ pairs and the $E^{*}$ state

Figure 3. Top row: Isosurface plots of wave functions ∣ψ(r)∣|\psi(\mathbf{r})|: the even [panel (a)] and odd [panel (b)] dz2d_{z^{2}} pairs at the MM point, at −2.4 eV and −1.5 eV respectively, and (c) the E∗E^{*} state at −0.5 eV. Bottom row [panels (d)-(f)]: Corresponding 2D-projected color-scale plots on the [110] plane, passing through La and Ni atoms, making the zz reflection symmetry clear, as well as the involvement of II density. Figure 2 shows that the dz2d_{z^{2}} pair can be followed throughout the zone, with splitting of ∼1\sim1 eV. The D∗D^{*} state approaching MM, part of the network as described in the text, is shown in Figure S7 of the SM [33].

Two-dimensional projected and three-dimensional Fermi surfaces of nonmagnetic La3Ni2O5F

Figure S1. (a) Two-dimensional projected Fermi surfaces (FSs) of nonmagnetic La3Ni2O5F\mathrm{La}_3\mathrm{Ni}_2\mathrm{O}_5\mathrm{F} within GGA, colored by the orbital characters. The small cylinder arises from the interstitial E∗E^* state. (b-c) Corresponding three-dimensional FS plots, colored by Fermi velocities. The two large dpσdp\sigma FSs are degenerate except near the zone edge XX point, but the velocity varies over one of the FS sheets.

Isosurface plots of wave functions for each eigenvalue from the bottom to the top at the M and $\Delta$ points

Figure S7. Isosurface plots of wave functions for each eigenvalue from the bottom to the top at the MM and Δ\Delta points.

The E∗E^{*} band in La3_{3}Ni2_{2}O5_{5}F differs in a few ways from the electride band in La3_{3}Ni2_{2}O5_{5} besides the distinctive difference in the distribution of II density. First, strong kzk_{z} dispersion (2 eV) of the ILN electride band leaves only a pillbox FS centered on corner point AA versus the cylindrical E∗E^{*} FS along the zone corner MM-AA line. Second, previous calculations on ILNs [37, 38, 39, 40, 41, 42] have introduced a “pseudo-atomic sphere” with a zero nuclear charge ‘ss’ orbital in the interstitial region, enabling representation with maximally localized Wannier functions including the ‘ss’ orbital. The interstitial orbital corresponds to a spherical [37, 38, 39, 40] or barrel-shaped [41] shape, with details provided for NdNiO2_{2} [41]. References [37, 38] have provided densities of individual states. Finally, our attempts to Wannierize including the E∗E^{*} band with a minimal basis set were unsuccessful.

It is common for bands several eV above EFE_{\mathrm{F}} to be interstitial plane-wave-like, avoiding atoms; it is very uncommon for such a band to dip down across EFE_{\mathrm{F}} and complicate the electronic structure as it does in La3_{3}Ni2_{2}O5_{5}F, and in a linear band fashion. The connection between D∗D^{*} and E∗E^{*} is seen by taking the La1 site as the center of 4/mmm4/mmm symmetry. On the bottom Ni layer, rotating the Ni dxz,dyzd_{xz},d_{yz} around the plaquette successively by π4+mπ2\frac{\pi}{4}+\frac{m\pi}{2} (m=0,1,2,3m=0,1,2,3), the linear combination is symmetric under 90° rotations. It is odd under zz reflection, so taking the odd combination with the upper Ni layer gives an orbital of 4/mmm4/mmm symmetry, the same as E∗E^{*}. Figure 3(c) shows that the E∗E^{*} wavefunction indeed has maxima in these directions toward La1 (and La2). One can form a 4/mmm4/mmm symmetric state from dz2d_{z^{2}} functions using the even combination of lower and upper layers, essentially the dz2ed_{z^{2}}^{e} band discussed above, but most of its spectral weight is expended in that low dz2d_{z^2} band.

The Dirac point. The D∗D^* and E∗E^* bands are anomalous and coupled as mentioned: (i) in approaching MM linearly, (ii) doing so with parallel dispersions and equal velocities in both high symmetry directions, and (iii) thus they are evident continuations of each other across a small to vanishing gap. The origin is an incipient degeneracy and non-analyticity of the subeigensystem. There is only minor sensitivity of these bands to shifts of positions (internal parameters) of the Ni, La, and O atoms. A related unoccupied band is evident in the study of the Lan+1NinO2n+2\mathrm{La}_{n+1}\mathrm{Ni}_n\mathrm{O}_{2n+2} system with empty apical sites [43, 44].

We have found sensitivity to the composition of the X virtual crystal atom: As the F composition of X=O/F is increased from 0.5, the E∗E^* and D∗D^* bands at MM close up and touch at an F fraction near 0.700. This Dirac point is a tuned degeneracy protected by preservation of the crystal symmetry (see the top row of Figure 4) and it signals a nonanalytic eigensubsystem. As the bands approach, the local band structure that includes other dd orbitals becomes intricate near the critical point. As mentioned, the orientation of the dxz,dyzd_{xz},d_{yz} orbitals enables mixing with the E∗E^* band (see the SM [33]), finally closing to the Dirac point.

Evolutions of the $E^{*}$ and $D^{*}$ bands

Figure 4. Evolutions of the E∗E^{*} and D∗D^{*} bands, meeting and crossing, enlarged from the corresponding region of Figure 1. Top row: Increasing the F concentration through the critical point at F(0.700) in La3_3Ni2_2O4_4(O1−x_{1-x}Fx_x)2_2. Note the switching of eigenvalue order at F(0.705). Bottom row: Applying symmetric strain, with a critical point at Δa=−1.84%\Delta a=-1.84\%. The lattice parameter aa is changed as noted, with fixed volume. Decreasing aa could be achieved by pressure, since the change in cc may have negligible effect. In each case, the upper E∗E^{*} band shifts downward at MM, meeting the D∗D^{*} band.

Increasing the F fraction from 0.5 would be accomplished by extending the O→\toF replacement process during synthesis. This replacement increases electron doping, evident in the lowering of the E∗E^* minimum at MM from −0.5-0.5 eV to −1-1 eV for F(0.7). Moving EFE_F from half-filling will further lessen the tendency toward magnetic order, and move the Ni formal valence closer toward the peak in the SC dome. We have found that compressive strain, see the bottom row of Figure 4, moves the system through the critical degeneracy with a somewhat different local band structure and band inversion.

This D∗D^*-E∗E^* Dirac point has characteristics in common with one in three-dimensional (3D) skutterudite CoSb3_3 [45, 46]. In that compound, linear bands meet at Γ\Gamma and lie (with much import) at EFE_F. The upper partner is a triplet at Γ\Gamma; only one linear combination of the triplet mixes with the lower linear band. The other point of interest is that the “interstitial” band does not have a density maximum in the large void in the skutterudite structure, instead it arises from a linear combination of Sb “4pz4p_z” orbitals perpendicular to the skutterudite ‘squares’ surrounding the void. This void is large enough to hold an atom, hence the intense interest in ‘filled skutterudites’ with unusual properties [47].

The interstitial density band in ILNs, discovered in 2004 [28] and for 15 years assigned to “La dd” character until recently interpreted as “electride-like”, has been little discussed in terms of its contribution to physical properties. The dichotomous electronic structure of La3_3Ni2_2O5_5F involves 2D carriers of opposite sign, separated in both kk and real spaces, and with distinct characters – a bilayer of correlated local dpodpo orbitals within layers of interstitial electrons confined to three channels.

Transport. Our separate study [31] indicates that the dpodpo FS will experience the magnetic fluctuations that are the likely dominant scattering, and pairing, mechanism, but that the E∗E^* FS is almost unaffected by Ni moments, confirmed by Madani et al. [48]. Our calculations of La and NiO2_2 optical phonon coupling to the E∗E^* FS (see the SM [33]) give deformation potentials of ∼0.1\sim0.1 eV/Å, an order of magnitude smaller than those in conventional phonon-coupled SCs. Moreover, Hausoel et al. [49] have shown that dynamical electron-electron interactions in ILNs leave the corresponding band in Ndn+1_{n+1}Nin_nO2n+2_{2n+2} compounds (n=4n=4-7) unaffected except for band filling due to renormalization of the dpodpo band. The E∗E^* band in La3_3Ni2_2O5_5F thus seems almost immune to disturbances.

Based on this evidence, we propose a two-band picture with very different values of scattering time τ\tau for the two carriers. The factor of 5 larger value of ΩP,dpo2\Omega_{P,dpo}^2 for the dpodpo bands will therefore be compensated (perhaps over-compensated) by the smaller value of τdpo\tau_{dpo}, so the E∗E^* FS may contribute strongly to the conductivity, somewhat short-circuiting the resistivity to values lower than those in other nickelates.

Within Bloch-Boltzmann theory, the conductivity is the sum of contributions from each FS (neglecting the very small difference in velocities)

σxx(T)=24π[ΩP,dpo2τd(T)+ΩP,E∗2τE(T)]→e2vh[τp(2πn)1/2+τn(2πp)1/2],(1)\sigma_{xx}(T)=\frac{2}{4\pi}\left[\Omega_{P,dpo}^{2}\tau_d(T)+\Omega_{P,E^{*}}^{2}\tau_E(T)\right] \rightarrow\frac{e^{2}v}{h}\left[\tau_p(2\pi n)^{1/2}+\tau_n(2\pi p)^{1/2}\right], \tag*{(1)}

where the factor of 2 is for the square lattice case, and σyy=σxx\sigma_{yy}=\sigma_{xx} by symmetry. The simple band structure (circular, isotropic) allows the FS quantities to be expressed in terms of the hole and electron densities pp, nn and τp,τn\tau_p,\tau_n (dpodpo, E∗E^*, respectively). Their separate FS density of states, Fermi velocities, and Drude plasma frequencies ΩP\Omega_P are given, respectively, (dpodpo, E∗E^*) by N(O)=(1.40, 0.33)/(eV-f.u.), vFv_F=(5.1, 4.6)×107\times10^7 cm/s, and ℏΩP2=4πe22N(O)vF2\hbar\Omega_P^2=\frac{4\pi e^2}{2}N(O)v_F^2=(3.8, 1.6) eV, quantities required for analysis of data when they become available.

In the isotropic relaxation time approximation, the Hall conductivity σxy\sigma_{xy} involves the integral of the curvature [50] over each FS. For a single circular FS (the dpodpo FS is also nearly circular; see the SM [33]), with constant velocity vv and curvature ∝1/kF\propto1/k_F on each surface, each FS contribution is analytic: σxy=e3B2ℏ2υ2τ2\sigma_{xy}=\frac{e^3B}{2\hbar^2}\upsilon^2\tau^2, depending on the square of τ\tau. The Hall coefficient becomes

RH=σxyσxxσyy=Be(τp2−τn2)[τpp1/2+τnn1/2]2.(2)R_H=\frac{\sigma_{xy}}{\sigma_{xx}\sigma_{yy}}=\frac{B}{e}\frac{(\tau_p^2-\tau_n^2)}{[\tau_p p^{1/2}+\tau_n n^{1/2}]^2}. \tag*{(2)}

The expression can be generalized to accommodate the slightly different carrier velocities if desired. In the limit of large τn/τp\tau_n/\tau_p this becomes RH/B=−1/neR_H/B=-1/ne, i.e., the Hall coefficient becomes dominated by the small E∗E^* (nn). contribution. Hall measurements on La3_3Ni2_2O5_5F will be instructive, potentially verifying the E∗E^{*} contribution to transport.

Possible two-gap SC. The evidence above of very weak coupling of E∗E^{*} to the two most prominent interactions (and mechanisms) promotes a two-band, perhaps extreme two-gap, picture of possible La3_3Ni2_2O5_5F SC. In a simple two-gap system, a dominant order parameter (giving TcT_{\mathrm c}) is coupled to a subservient order parameter (its Tc=0T_{\mathrm c}=0). The BCS treatment of such a two-gap system is simple [51]: A small E∗E^{*} gap would be induced by the stronger coupling partner, with the ratio of gaps depending on the interband coupling parameter and independent of temperature. A more extensive exposition of Ginzburg-Landau theory [51] should become relevant. This state of well-separated gaps would make specific heat a crucial experiment (as well as others, such as NMR), as well as impacting other SC state parameters.

Summary. The properties discussed here pinpoint La3_3Ni2_2O5_5F as a unique Ni1+^{1+} system. Supplemented by a single E∗E^{*} band of II density spanning all three La layers, it is found that the II density effects span a network of dd bands: a dz2d_{z^2}, dz2od_{z^2}^{\mathrm o} strongly bound pair yet split by 1 eV: the unanticipated D∗D^{*} pair dxz,dyzd_{xz},d_{yz} approaching degeneracy from below, and the E∗E^{*} band itself; the last two form an incipient degeneracy (Dirac point) that leads to the linearity of the E∗E^{*} band crossing EFE_{\mathrm F}. Whereas the dpσd_{p\sigma} band will show Ni spin fluctuations requiring correlation effects and presenting a short mean free path, the E∗E^{*} band will likely experience weak renormalization due to only indirect effects from moments and lattice motion, with a longer relaxation time. Effects on electronic transport properties have been outlined.

While the E∗E^{*} band provides hole doping to valence Ni1.09+^{1.09+}, further doping will promote La3_3Ni2_2O5_5F into the SC regime, where the E∗E^{*} band should affect properties in novel ways. Higher doping (higher fraction of anionic F or cationic La→\toSr substitution) and in-plane compressive strain will lead toward the Dirac nonanalyticity of the D∗D^{*} and E∗E^{*} bands at the zone corner. The out-of-plane coherence length will likely be on the order of the blocking layer separation, with the next question being whether Josephson coupling between bilayers can promote a coherent 3D SC state. It remains clear, however, that Ni1+^{1+} is fundamentally different from Cu2+^{2+}.

*Acknowledgments—*KWL was supported by National Research Foundation of Korea (NRF) Grant (RS-2024-00392493). YJS was supported by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – CRC 1487, “Iron, upgraded!” – project number 443703006.

*Data availability—*The data that support the findings of this article are provided in the main text and supplemental information. Additional data would be available from the authors upon reasonable request.

References

tailed variations of E∗E^{*} and D∗D^{*} bands by changing the F concentration and lattice parameter.

End of Paper

Supplemental Material

This Supplemental Material provides additional information to provide support to the text. The information is described in the figure captions.

Fermiology

Plots of supercell band structure and electron localization function

Comparison of band structures obtained using VCA and a supercell approach

Figure S2. Comparison of the band structures of La3_3Ni2_2O5_5F obtained using (left) the virtual crystal approximation (VCA) and (right) a supercell approach having twice as many bands (not so obvious). In the ordered O/F supercell, we placed two O ions and two F ions in the square. As expected, the two band structures are extremely similar, since the 2p2p orbitals of these atoms are far below EFE_\mathrm{F} and both ions have filled 2p2p shells.

atomsmultiplicityF(0.5)F(0.700)
Ni20.5560.501
O4−1.244-1.244−1.247-1.247
F/O2−1.118-1.118−1.028-1.028
La111.9381.924
La222.0872.062
sum–0.0100.001

Table S1. Bader charges (ZZ minus electronic charge) of La3_3Ni2_2O4_4(O1−x_{1-x}Fx_x)2_2 at F(0.5) and F(0.700) for comparison. Note that Bader charges do not reproduce the formal charges very closely, as is common. Certain charges change with F content; Ni net charge decreases 0.05e−0.05e^- with F content. It is worthy of note that the F/O layer X atom does not lose 0.5e−0.5e^- charge, but only (Bader) 0.109e−0.109e^- per X ion. Ni layer O and all La charges hardly change.

Figure S3. Electron localization function plots on the [110] plane at (a) F(0.5) and (b) F(0.7). The white arrow indicates a density maximum between the NiO2_2 and La2 layers, and double-crescent-shaped weaker local maxima in the La1 layer. These maxima become somewhat more prominent with increasing F content. These results are roughly consistent with the plots of the E∗E^* wavefunction.

Symmetry Analysis and Isosurface Contours

GGA band structure of La₃Ni₂O₅F showing irreducible representations at high-symmetry points

Figure S4. GGA band structure of La3_3Ni2_2O5_5F (the X 2p bands lie lower), providing the irreducible representations of each band at several high symmetry points in the I4/mmmI4/mmm space group. The E∗E^* band dips well below the Fermi level (the zero of energy) at MM.

Isosurface plots of wave functions for each eigenstate at the Γ and Σ points

Figure S5. Isosurface plots of wave functions ∣Ψ(r)∣|\Psi(\mathbf{r})| for each eigenstate from the bottom to the top at the Γ\Gamma and Σ\Sigma points. Extended lime colored regions indicate contributions from the interstitial density. This is the case for several following plots. The upper labels indicate the symmetry, while the lower labels give a description of the character of the state.

Isosurface plots of wave function plots at the X and XM points

Figure S6. Isosurface plots of wave function plots for each point from the bottom to the top at the XX and XMXM points, with labeling as above.

Variations of D∗D^{*} and E∗E^{*} bands by F concentration, Ba-doping, phonons, lattice parameter, and pressure

Evolution of the $E^{*}$ band and related bands versus F occupancy

Figure S8. Evolution of the E∗E^{*} band and related bands versus F occupancy from 0.650 to 0.725 in La3Ni2O4(O1−xFx)2\mathrm{La}_{3}\mathrm{Ni}_{2}\mathrm{O}_{4}(\mathrm{O}_{1-x}\mathrm{F}_{x})_{2}. At the critical doping level of 0.705, corresponding to La3Ni2O4.59F1.41\mathrm{La}_{3}\mathrm{Ni}_{2}\mathrm{O}_{4.59}\mathrm{F}_{1.41}, the E∗E^{*} band [M1+(Ag)M_{1}^{+}(A_{g})] forms a Dirac point at around −0.94-0.94 eV with the odd-parity dxz,dyzd_{xz},d_{yz} D∗D^{*} band [M2−(B2u)M_{2}^{-}(B_{2u})]. Upon further doping, the crossing lifts and the bands separate.

Band structures and wave-function isosurface plots for La3Ni2O4.59F1.41

Figure S9. (a) Band structure of La3_3Ni2_2O4.59_{4.59}F1.41_{1.41}, i.e. at the F fraction of 0.705, together with the irreducible labels at the MM point where the singular point occurs. (b) The corresponding band structure is zoomed in (twice) near the linear crossing point. At the bottom are a series of wave-function isosurface plots of the wavefunctions at the MM point, from the bottom energy up, together with their 2D-projected isocontour maps on the [110] plane. The first two are those designated in the text dz2d_{z^2}, dz2od_{z^2}^{\mathrm{o}} (even, odd), which illustrates the participation of the interstitial density; they are very much the same for F(0.5). The third (E∗E^*) eigenvalue (blue M1+M_1^+), with the interstitial density, has crossed the D∗D^* (M4−M_4^-) state by a tiny amount. The involvement of the M4−M_4^- density with interstitial density, though relatively minor, identifies it as the D∗D^* partner of E∗E^*.

Band structures and Fermi surfaces at different F concentrations

Figure S10. Variations in the size of the E∗E^* electron FS with changing F concentration in La3_3Ni2_2O4_4(O1−x_{1-x}Fx_x)2_2. (a), (b) show the band structures at F(0.5) and F(0.7), respectively, with (c) the corresponding FSs. Increasing the F ratio from 0.5 to 0.7 leads to a some enlargement of the E∗E^* electron pocket and electron density. The dpσdp\sigma band thereby acquires more additional holes than the F difference (0.2) provides. Specifically, the E∗E^* electron pocket has a 45% larger radius kFk_F at F(0.7) than at F(0.5), so the Ni valence is +1.19, just below the dome peak.

Band structures and Fermi surfaces of (La,Ba)3Ni2O5F at different Ba doping levels

Figure S11. Band structures of (La,Ba)3_3Ni2_2O5_5F, at different Ba doping levels from 0.1 to 0.3, together with (rightmost panels) the variations of the FSs. Note that the VCA approach is applied to either (upper row) La1 or (lower row) La2. The main change is in the self-doping, which is very much as would be expected from a rigid band approximation. At 30% Ba-doping level, the electron carriers contained in the E∗E^* FSs reduce to 0.072e−0.072e^- and 0.052e−0.052e^- for the La1 and La2 cases, respectively. The result is that doping of holes by the La to Ba substitution negates the intrinsic electron doping of La3_3Ni2_2O5_5F, as would be expected from usual carrier counting.

Frozen phonon calculations for oxygen breathing and La ion vertical modes in La3Ni2O5F

Figure S12. Frozen phonon calculations for the oxygen breathing mode in the NiO2_2 layers and vertical vibrating modes of La ions, in La3_3Ni2_2O5_5F. Top: Sketches of ion vibrations for (a) the oxygen breathing mode, (b) La2 A1gA_{1g}, and (c) La A2uA_{2u} modes. Bottom: Corresponding variations of bands at EFE_F for each mode. These yield tiny deformation potentials of 0.1–1 eV/Å for the E∗E^* band, indicating minor phonon effects.

Evolution of the $E^{*}$ band with variation of the $a$ lattice constant

Figure S13. Evolution of the E∗E^{*} band with variation of the aa lattice constant, fine-tuned from −1.78% to −1.89%, with fixed volume in La3Ni2O5F\mathrm{La}_{3}\mathrm{Ni}_{2}\mathrm{O}_{5}\mathrm{F}. The multiband complex complicates the behavior, which is quite sensitive to compression, but the Dirac point degeneracy occurs very close to the −1.84% compression. It is interesting that the “initial” and “final” band structures look virtually the same, in spite of the changing of the order of states. A larger scale variation is provided in the next figure.

Evolution of the $E^{*}$ band position with variation of the $a$ lattice constant

Figure S14. Evolution of the E∗E^{*} band position with variation of the aa lattice constant, broad-tuned from −4% to +4%, with fixed volume in La3Ni2O5F\mathrm{La}_{3}\mathrm{Ni}_{2}\mathrm{O}_{5}\mathrm{F}. In addition to the larger scale view of variation of self-doping, the figure provides a close-up picture of the band complex very near (−2%) the critical Dirac point.

Band structures of La$_3$Ni$_2$O$_5$F at hydrostatic pressures of 0, 10, and 25 GPa

Figure S15. Variations of band structure of La3_3Ni2_2O5_5F at hydrostatic pressure PP of (a) 0 GPa, (b) 10 GPa, and (c) 25 GPa. By applying external hydrostatic pressure, the E∗E^{*} and D∗D^{*} FSs increase in area (density). Fine-tuning to somewhat higher PP values will produce the Dirac crossing as does the simple reduction of the aa lattice parameter.

References

  1. [1]J. L. Dye, Anionic electrons in electrides, Nature 365, 10 (1993).DOI
  2. [2]M. J. Wagner, R. H. Huang, J. L. Eglin, and J. L. Dye, An electride with a large six-electron ring, Nature 368, 726 (1994).
  3. [3]D. J. Singh, H. Krakauer, C. Haas, and W. E. Pickett, Theoretical determination that electrons act as anions in the electrides Cs⁺(15-crown-5)₂e⁻, Nature 365, 39 (1993).
  4. [4]C. Liu, S. A. Nikolaev, W. Ren, and L. A. Burton, Electrides: a review, J. Mater. Chem. C 8, 10551 (2020).DOI
  5. [5]H. Hosono and M. Kitano, Advances in materials and applications of inorganic electrides, Chem. Rev. 121, 3121 (2021).
  6. [6]F. Sun, Z. Guo, Y. Lu, J. Li, T.-N. Ye, H. Hosono, and J. Wu, Electrides: Emerging electronic materials for catalysis, Fund. Res. 6, 400 (2026).
  7. [7]W. Cheng, K. Li, T. Tada, H. Hosono, and J. Wang, Electride Formation in Ba–P System and the Unexpected Structure Transition of Electrides under Pressure, J. Phys. Chem. C 126, 16815 (2022).DOI
  8. [8]M.-S. Miao and R. Hoffmann, High Pressure Electrides: A Predictive Chemical and Physical Theory, Acc. Chem. Res. 47, 1311 (2014).DOI
  9. [9]H. Y. Song, R. I. Yoo, J.-H. Choi, S.-H. Kang, J. Bang, W. Li, C. N. Nandadasa, D. Thapa, D. Yoon, M. J. Han, K. H. Lee, S. G. Kim, K. Lee, and S. W. Kim, Van der Waals electride: Toward intrinsic two-dimensional ferromagnetism of spin-polarized anionic electrons, Mater. Today Phys. 20, 100473 (2021).DOI
  10. [10]J. Zhou, J.-Y. You, Y.-M. Zhao, Y. P. Feng, and L. Shen, Van der Waals Electrides, Acc. Chem. Res. 57, 2572 (2024).DOI
  11. [11]I. Park, Y. He, H.-k. Mao, J. H. Shim, and D. Y. Kim, Electride formation of hcp-iron at high pressure: Unraveling the origin of the superionic state of iron-rich compounds in rocky planets, Adv. Sci. 11, 2308177 (2024).DOI
  12. [12]D. Li, K. Lee, B. Y. Wang, M. Osada, S. Crossley, H. R. Lee, Y. Cui, Y. Hikita, and H. Y. Hwang, Superconductivity in an infinite-layer nickelate, Nature 572, 624 (2019).DOI
  13. [13]G. A. Pan, D. F. Segedin, H. LaBollita, Q. Song, E. M. Nica, B. H. Goodge, A. T. Pierce, S. Doyle, S. Novakov, D. Córdova Carrizales, A. T. N’Diaye, P. Shafer, H. Paik, J. T. Heron, J. A. Mason, A. Yacoby, L. F. Kourkoutis, O. Erten, C. M. Brooks, A. S. Botana, and J. A. Mundy, Superconductivity in a quintuple-layer square-planar nickelate, Nat. Mater. 21, 160 (2022).DOI
  14. [14]G. A. Pan, D. F. Segedin, S. F. R. TenHuisen, L. Bhatt, H. LaBolitta, A. Y. Jiang, Q. Song, A. B. Turkiewicz, D. R. Baykusheva, A. Nag, S. Agrestini, K.-J. Zhou, J. Pelliciari, V. Bisogni, H. Zhou, M. P. M. Dean, H. Paik, D. A. Muller, L. F. Kourkoutis, C. M. Brooks, M. Mitrano, A. S. Botana, B. H. Goodge, and J. A. Mundy, Superconducting phase diagram of multilayer square-planar nickelates, Science 392, 1390 (2026).
  15. [15]V. V. Poltavets, K. A. Lokshin, S. Dikmen, M. Croft, T. Egami, and M. Greenblatt, La₃Ni₂O₆: A New Double T’-type Nickelate with Infinite Ni1+/2+O₂ Layers, J. Am. Chem. Soc. 128, 9050 (2006).
  16. [16]N. apRoberts Warren, J. Crocker, A. P. Dioguardi, K. R. Shirer, V. V. Poltavets, M. Greenblatt, P. K. Kłavins, and N. J. Curro, NMR evidence for spin fluctuations in the bilayer nickelate La₃Ni₂O₆, Phys. Rev. B 88, 075124 (2013).DOI
  17. [17]K. A. Lokshin, D. Mitchell, M. V. Lobanov, V. Struzhkin, and T. Egami, Synthesis and Characterization of Pure Infinite Layer Ni⁺ Nickelates: LnNiO₂ (Ln = La, Nd, Pr) and La₂Ni₂O₆, ECS J. Solid State Sci. Technol. 11, 044008 (2022).DOI
  18. [18]Z. Liu, H. Sun, M. Huo, X. Ma, Y. Ji, E. Yi, L. Li, H. Liu, J. Yu, Z. Zhang, Z. Chen, F. Liang, H. Dong, H. Guo, D. Zhong, B. Shen, S. Li, and M. Wang, Evidence for charge and spin density waves in single crystals of La₃Ni₂O₇ and La₃Ni₂O₆, Sci. China Phys. Mech. Astron. 66, 217411 (2023).DOI
  19. [19]R. Gao, L. Jin, S. Huyan, D. Ni, H. Wang, X. Xu, S. L. Bud’ko, P. Canfield, W. Xie, and R. J. Cava, Is La₃Ni₂O₆₅ a Bulk Superconducting Nickelate?, ACS Appl. Mater. Interfaces 16, 66857 (2024).
  20. [20]V. Pardo and W. E. Pickett, Metal-insulator transition in layered nickelates La₃Ni₂O₇−δ (δ = 0.0, 0.5, 1), Phys. Rev. B 83, 245128 (2011).
  21. [21]A. S. Botana, V. Pardo, W. E. Pickett, and M. R. Norman, Changing ordering in Ni1+/Ni2+ nickelates: La₄Ni₃O₈ and La₃Ni₂O₆, Phys. Rev. B 94, 081105 (2016).DOI
  22. [22]A. S. Botana, K.-W. Lee, M. R. Norman, V. Pardo, and W. E. Pickett, Low valence nickelates: Launching the nickel age of superconductivity, Front. Phys. 9, 813532 (2022).DOI
  23. [23]Y. Zhang, L.-F. Lin, A. Moreo, T. A. Maier, and E. Dagotto, Electronic structure, magnetic correlations, and superconducting pairing in the reduced Ruddlesden-Popper bilayer La₃Ni₂O₆ under pressure: Different role of d₃z²−r² orbital compared with La₃Ni₂O₇, Phys. Rev. B 109, 045151 (2024).
  24. [24]F. Lechermann, S. Bötzel, and I. M. Eremin, Interplay of orbital-selective Mott criticality and flat-band physics in La₃Ni₂O₆, Phys. Rev. B 112, 245125 (2025).DOI
  25. [25]Y. Wang, Z. Chen, Y. Zhang, K. Jiang, and J. Hu, The Mott­ness and the Anderson localization in bilayer nickelate La₃Ni₂O₇−δ, arXiv preprint arXiv:2501.08536 10.48550/arXiv.2501.08536 (2025).
  26. [26]S. Kamiyama, R. Kohno, Y. Hoshi, K. Ushio, D. Nakaoka, H. Sakakibara, and K. Kuroki, Theoretical proposal of superconductivity in hole-doped reduced bilayer nickelate La₃Ni₂O₆: a manifestation of orbital-space bilayer model with incipient bands, arXiv preprint arXiv:2603.11771 10.48550/arXiv.2603.11771 (2026).
  27. [27]R. Wernert, R. D. Smyth, and M. A. Hayward, Synthesis of the Double Infinite-Layer Ni(I) Phase La₃Ni₂O₅F via Sequential Topochemical Reactions, J. Am. Chem. Soc. 148, 6109 (2026).DOI
  28. [28]K.-W. Lee and W. E. Pickett, Infinite-layer LaNiO₂: Ni1+ is not Cu2+, Phys. Rev. B 70, 165109 (2004).DOI
  29. [29]M.-Y. Choi, K.-W. Lee, and W. E. Pickett, Role of 4f states in infinite-layer NdNiO₂, Phys. Rev. B 101, 020503 (2020).
  30. [30]M.-Y. Choi, W. E. Pickett, and K.-W. Lee, Fluctuation-frustrated flat band instabilities in NdNiO₂, Phys. Rev. Res. 2, 033445 (2020).DOI
  31. [31]Y.-J. Song, W. E. Pickett, and K.-W. Lee, Anomalous Behavior of the Ni1+ moment and interstitial band in bi-infinite-layered La₃Ni₂O₅F, arXiv preprint arXiv:2606.28735 10.48550/arXiv.2606.28735 (2026).
  32. [32]P. Blaha, K. Schwarz, F. Tran, R. Laskowski, G. K. H. Madsen, and L. D. Marks, WIEN2k: An APW+lo program for calculating the properties of solids, J. Chem. Phys. Soc. 152, 074101 (2020).DOI
  33. [33]See Supplemental Information at [URL will be inserted by publisher] for Fermi surfaces, symmetry analysis of bands, various isosurface plots of wave functions, and de tailed variations of E* and D* bands by changing the F concentration and lattice parameter.
  34. [34]D. Li, B. Y. Wang, K. Lee, S. P. Harvey, M. Osada, B. H. Goodge, L. F. Kourkoutis, and H. Y. Hwang, Superconducting Dome in Nd₁₋ₓSrₓNiO₂ Infinite Layer Films, Phys. Rev. Lett. 125, 027001 (2020).
  35. [35]S. Zeng, C. S. Tang, X. Yin, C. Li, M. Li, Z. Huang, J. Hu, W. Liu, G. J. Omar, H. Jani, Z. Slim, K. Han, D. Wan, P. Yang, S. J. Pennycook, A. T. S. Wee, and A. Ariando, Phase Diagram and Superconducting Dome of Infinite-Layer Nd₁₋ₓSrₓNiO₂ Thin Films, Phys. Rev. Lett. 125, 147003 (2020).
  36. [36]M. Osada, B. Y. Wang, K. Lee, D. Li, and H. Y. Hwang, Phase diagram of infinite layer praseodymium nickelate Pr₁₋ₓSrₓNiO₂ thin films, Phys. Rev. Mater. 4, 121801 (2020).
  37. [37]Y. Nomura, M. Hirayama, T. Tadano, Y. Yoshimoto, K. Nakamura, and R. Arita, Formation of a two-dimensional single-component correlated electron system and band engineering in the nickelate superconductor NdNiO₂, Phys. Rev. B 100, 205138 (2019).
  38. [38]Y. Nomura and R. Arita, Superconductivity in infinite-layer nickelates, Rep. Prog. Phys. 85, 052501 (2022).DOI
  39. [39]P. Adhikary, S. Bandyopadhyay, T. Das, I. Dasgupta, and T. Saha-Dasgupta, Orbital-selective superconductivity in a two-band model of infinite-layer nickelates, Phys. Rev. B 102, 100501 (2020).
  40. [40]Y. Gu, S. Zhu, X. Wang, J. Hu, and H. Chen, A substantial hybridization between correlated Ni-d orbital and itinerant electrons in infinite-layer nickelates, Commun. Phys. 3, 84 (2020).
  41. [41]K. Foyevtsova, I. Elfimov, and G. A. Sawatzky, Distinct electridelike nature of infinite-layer nickelates and the resulting theoretical challenges to calculate their electronic structure, Phys. Rev. B 108, 205124 (2023).DOI
  42. [42]J.-Y. You, Interstitial s states and chemical pressure as key drivers of enhanced electron-phonon coupling in infinite-layer nickelates, Phys. Rev. B 113, 064510 (2026).DOI
  43. [43]H. LaBollita and A. S. Botana, Electronic structure and magnetic properties of higher-order layered nickelates: Laₙ₊₁NiₙO₂ₙ₊₂ (n = 4 − 6), Phys. Rev. B 104, 035148 (2021).DOI
  44. [44]A. K. Gavrilov, L. C. Santander, and A. S. Botana, Electronic structure of higher-order layered palladates: Laₙ₊₁PdₙO₂ₙ₊₂ (n=4-7), arXiv preprint arXiv:2603.27456 10.48550/arXXiv.2603.27456 (2026).arxiv.org/abs/2603.27456
  45. [45]J. C. Smith, S. Banerjee, V. Pardo, and W. E. Pickett, Dirac point degenerate with massive bands at a topological quantum critical point, Phys. Rev. Lett. 106, 056401 (2011).
  46. [46]V. Pardo, J. C. Smith, and W. E. Pickett, Linear bands, zero-momentum weyl semimetal, and topological transition in skutterudite-structure pnictides, Phys. Rev. B 85, 214531 (2012).DOI
  47. [47]B. C. Sales, Filled skutterudites, in Handb. Phys. Chem. Rare Earths, Vol. 33, edited by K. Gschneidner, J.-C. Bü̈nzli, and V. Pecharsky (Elsevier, 2003) pp. 1–34.
  48. [48]K. Madani, Q. N. Meier, and A. Cano, Structural stability, electronic structure, and magnetism of the d⁹ double infinite-layer La₃Ni₂O₅F under chemical pressure and epitaxial strain, arXiv preprint arXiv:2607.22029 10.48550/arXiv.2607.22029 (2026).
  49. [49]A. Hausoel, S. D. Cataldo, M. Kitatani, O. Janson, and K. Held, Superconducting phase diagram of finite-layer nickelates Ndₙ₊₁NiₙO₂ₙ₊₂, npj Quantum Matter. 10, 69 (2025).
  50. [50]N. P. Ong, Geometric interpretation of the weak-field Hall conductivity in two-dimensional metals with arbitrary Fermi surface, Phys. Rev. B 43, 193 (1991).pubmed.ncbi.nlm.nih.gov/9996203
  51. [51]M. E. Zhitomirsky and V.-H. Dao, Ginzburg-landau theory of vortices in a multigap superconductor, Phys. Rev. B 69, 054508 (2004).DOI

Paper details

vlnc:W_vjxs4bdd
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 1

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents