Letters

https://doi.org/10.1038/s41567-020-01128-5

Isotropic Pauli-limited superconductivity in the

infinite-layer nickelate Nd0.775Sr0.225NiO2
Bai Yang Wang 1,2 ✉, Danfeng Li 2,3, Berit H. Goodge 4, Kyuho Lee 1,2, Motoki Osada 2,5,
Shannon P. Harvey 2,3, Lena F. Kourkoutis 4,6, Malcolm R. Beasley3 and Harold Y. Hwang 2,3 ✉

The recent observation of superconductivity in thin-film In good agreement with prior observations, c-axis oriented crys-
infinite-layer nickelates1–3 offers a different angle from which talline coherence is observed throughout the film, together with
to investigate superconductivity in layered oxides4. A wide extended defects that are primarily based on vertical Ruddlesden–
range of candidate models have been proposed5–10, which Popper-type faults. The temperature-dependent resistivity ρ(T) in
emphasize single- or multi-orbital electronic structure, the low-temperature region, with varying magnetic field applied
Kondo or Hund’s coupling and analogies to cuprates. Further along the c axis (H⊥) and the a–b plane (H||, transverse geometry),
experimental characterization of the superconducting state is shown in Fig. 1b,c. Although minimal magnetoresistance is
is needed to develop a full understanding of the nickelates. observed in the normal state, the superconducting transition is sup-
Here we use magnetotransport measurements to probe the pressed with increasing magnetic field strength, which allows us to
superconducting anisotropy in Nd0.775Sr0.225NiO2. We find that extract the temperature dependence of Hc2.
the upper critical field is surprisingly isotropic at low tem- The main result of this study, the magnetic field–temperature
peratures despite the layered crystal structure. In a magnetic phase diagram of Hc2,⊥ and Hc2,|| down to a temperature of 0.5 K, is
field, the superconductivity is strongly Pauli-limited, such that shown in Fig. 2. Here, Tc and Hc2 are identified as the temperature
the paramagnetic effect dominates over orbital de-pairing. and field strength at which the resistive transition reaches 50% of
Underlying this isotropic response is a substantial anisotropy the normal state, as discussed below. Overall, the most striking
in the superconducting coherence length, which is at least feature is that Hc2 is largely isotropic at low temperatures (Fig. 2a,
four times longer in-plane than out-of-plane. A prominent inset), despite the layered crystal structure, the quasi-2D nature of
low-temperature upturn in the upper critical field indicates the dx2 y2 hole band5,6 and the thin-film geometry. Does this reflect
the presence of an unconventional ground state. a rather
I isotropic electronic structure, and superconducting coher-
A key characteristic of layered superconductors is an upper ence length, as unusually found in (K,Ba)FeAs2 (ref. 14), or does it
critical magnetic field Hc2 that often exhibits substantial anisotropy have a different fundamental origin? We aim to address this central
γ ≡ Ηc2,||/Ηc2,⊥, where Ηc2,|| and Ηc2,⊥ indicate fields in the a–b plane question. As we argue below, Hc2 is dominated by the isotropic spin
and along the crystal c direction, respectively. In layered oxides such response, rather than orbital effects, over much of the phase dia-
as cuprates11, ruthenates12 and cobaltates13, the anisotropy reflects gram. Such strongly Pauli-limited superconductivity in nickelates
the two-dimensional (2D) nature of both the crystal and electronic appears to be unique among the layered oxides.
structure. The case for the recently discovered infinite-layer nick- We start by examining the magnetoresistive transition observed
elates is a priori interesting: although it shares the same crystal in Fig. 1b,c. The visible broadening can have contributions from
structure as infinite-layer cuprates, calculations indicate that both a static compositional inhomogeneity, superconducting fluctuations
quasi-2D hole band and three-dimensional electron bands are pres- and vortex motion in this type-II superconductor1,9. In terms of
ent in the nickelates for a broad range of electron interactions5,6. The inhomogeneity, the Ruddlesden–Popper-type faults probably give
relative importance of these bands is a subject of much interest with rise to a spread in Tc. Regarding superconducting fluctuations, we
respect to the superconducting state. More generally, nickelates pro- fit the onset of superconductivity following the 2D Aslamazov–
vide a new system within which to explore longstanding debates on Larkin (AL) formulism15,16, with representative fits shown in the
the role of dimensionality for superconductivity. insets of Fig. 1b,c. From this we obtain a corresponding TcAL, above
With these motivations, we have studied the superconduct- which superconducting fluctuations substantially impact the resis-
ing anisotropy through resistive measurements of Hc2 in thin-film tive transition. We note that this regime also includes effects arising
Nd0.775Sr0.225NiO2. This composition has the highest crystallinity in from static inhomogeneities.
our recent investigations, as well as a superconducting transition As for vortex motion, we investigate its contribution in scaled
temperature Tc among the peak values observed across the super- Arrhenius plots, as shown in Fig. 3a,b. Between the normal state
conducting dome2. As previously reported, our optimal samples and the measurement noise floor, exponentially decreasing resis-
are capped with SrTiO3 and are bounded in thickness to less than tivity is seen for both orientations across several orders of magni-
~10 nm; above this we observe mixed-phase formation or decom- tude. This is well described by thermally activated vortex motion17:
position. A high-angle annular dark-field scanning transmission ρ(T,H) = ρ0(H) exp(−U0/kBT), where ρ0 is the resistivity prefactor,
electron microscopy (HAADF-STEM) image is shown in Fig. 1a. U0 is the activation energy and kB is Boltzmann’s constant. Figure 3c

1
Department of Physics, Stanford University, Stanford, CA, USA. 2Stanford Institute for Materials and Energy Sciences, SLAC National Accelerator
Laboratory, Menlo Park, CA, USA. 3Department of Applied Physics, Stanford University, Stanford, CA, USA. 4School of Applied and Engineering Physics,
Cornell University, Ithaca, NY, USA. 5Department of Materials Science and Engineering, Stanford University, Stanford, CA, USA. 6Kavli Institute at Cornell
for Nanoscale Science, Cornell University, Ithaca, NY, USA. ✉e-mail: [email protected]; [email protected]

Nature Physics | VOL 17 | April 2021 | 473–477 | www.nature.com/naturephysics 473

Letters NATUre PHySIcS

SrTiO3

Nd0.775Sr0.225NiO2

SrTiO3
2 nm

b c
250

12 T H⊥ 12 T H∥
200
0T 0T

150
ρ (µΩ cm)

200 200

150 150
ρ (µΩ cm)

ρ (µΩ cm)
1T 1T
4T 4T
100 6T 6T
100 100
8T 8T
10 T 10 T
50 50
50
0 0
0 4 8 12 16 0 4 8 12 16
T (K) T (K)
0
0 10 20 30 40 0 10 20 30 40
T (K) T (K)

Fig. 1 | Structure and magnetotransport properties of thin-film Nd0.775Sr0.225NiO2. a, Cross-sectional HAADF-STEM image of an Nd0.775Sr0.225NiO2 thin
film, with the size scale indicated in the lower left. b,c, Resistivity ρ(T) below 40 K at H = 0, 0.2, 0.4, 0.6, 0.8, 1, 2, 3, 4, 6, 8, 10 and 12 T (ranging in colour
from dark red to dark blue) for magnetic fields along the c axis (b) and transverse in the a–b plane (c). The insets show representative resistive transitions
with 2D Aslamazov–Larkin fits shown as dashed lines for field along the c axis (b) and in the a–b plane (c). Grey arrows indicate the direction of increasing
field strength.

shows U0 extracted using this form. Similar to cuprates17,18, U0 for H⊥ the energy cost of the field-induced supercurrent), with the char-
and H|| exhibits regions of power-law dependence on the magnetic acteristic (Tc − T)1/2 dependence of Hc2,|| that arises from 2D con-
field strength. Note that the success of this fitting implies that vor- finement. This is commonly observed for superconductors that are
tex motion is relevant for H||, and the large difference in U0 values thinner than the Ginzburg–Landau coherence length20,21, and it is a
between H⊥ and H|| indicates substantial vortex-motion anisotropy. reasonable possibility here, given the thickness limits for stabilizing
We define TcVM as the temperature above which resistivity deviates Nd0.775Sr0.225NiO2. In such geometries, Hc2 is well described by the
from thermally activated behaviour, and attribute the resistive tran- linearized Ginzburg–Landau form22 (generalized to an anisotropic,
sition below TcVM predominantly to vortex motion. c-axis thickness-limited superconductor) as
Summarizing the onset of fluctuations and vortex motion, we  
construct the phase diagrams in Fig. 4a,b. In addition to HcVM(T) and ϕ0 T
Hc2;? ðT Þ ¼ 1� ; ð1Þ
HcAL(T), we plot Hc90%(T), Hc50%(T) and Hc1%(T), which correspond 2πξ2ab ð0Þ Tc
to the field strength at which the resistivity reaches 90%, 50% and
1% of the normal state, respectively, and which are representative pffiffiffiffiffi  1
candidate proxies for Hc2. Between HcVM(T) and HcAL(T) (extrapo- 12ϕ0 T 2
Hc2;k ðT Þ ¼ 1� ; ð2Þ
lating to 27% and 80% of the resistive transition for H = 0 T), there 2πξab ð0Þd Tc
is a regime in which the magnetoresistance can serve as a reliable
indicator of Hc2, free from vortex creep or superconducting fluc- where ϕ0 is the flux quantum, ξab(0) is the in-plane zero-temperature
tuations, which tend to under- or overestimate Hc2 as well as distort coherence length and d is the superconducting thickness.
its functional form19. Hence, we use Hc50% to measure Hc2, and our Using these forms, we extract ξab(0) = 42.6 ± 0.1 Å and
conclusions are robust to variations within this regime aside from d = 225 ± 7 Å. In previous studies on other material systems, d
specific numerical values. derived from such an analysis was in good quantitative agreement
Next, we focus on magnetotransport near Tc and take advantage with the physical thickness of the film20,23, except for cases with
of the simplification in the critical regime that arises from diverg- dominant spin–orbit scattering24,25, for which the thickness is under-
ing length scales. Here, the central observation is a T-linear depen- estimated. Here, although the obtained ξab(0) is reasonable and
dence of Hc2,⊥ and a (Tc − T)1/2 dependence of Hc2,|| (Fig. 4c). First, we comparable to that of the cuprates (Supplementary Information),
consider purely orbital de-pairing (superconductivity suppressed by we deduce a thickness that is 2.3–3 times larger than the observed

474 Nature Physics | VOL 17 | April 2021 | 473–477 | www.nature.com/naturephysics

NATUre PHySIcS Letters
a b
16 HPauli, µ = 1 µB HPauli, µ = 1 µB
H⊥ 5 H∥
12
H∥ 4
14

µ0Hc2 (T)
Normal state

µ0Hc2 (T)
8 3
2
12 Normal state 4 Sample 1 (74 Å)
1 Sample 2 (90 Å)
Hc2,⊥ GL
10 0 0
0 2 4 6 8 0.80 0.85 0.90 0.95 1.00
µ0Hc2 (T)

T (K) T/Tc
8
H⊥ H∥
HPauli, µ = 2.4 µB HPauli, µ = 2.4 µB
Hc 50%
Hc50%
6

Pauli limited Pauli limited

4

2 Superconductivity Superconductivity
Orbital
ξab(0) = 42.6 Å a,b = 3.905 Å ξc(0) ≲ 10 Å c = 3.37 Å
limited
0
0 2 4 6 8 0 2 4 6 8
T (K) T (K)

Fig. 2 | Hc2 versus T phase diagram for Nd0.775Sr0.225NiO2. a,b, Hc2–T phase diagrams for magnetic fields along the c axis (a) and in the a–b plane (b), where
μ0 is the vacuum permeability. Hc50% is the field strength at which the resistive transition reaches 50% of the normal state, which will be identified as Hc2
for the rest of the paper. The regions above Hc2 (open and filled circles) are shaded orange, representing the normal state, while the regions under Hc2
are shaded as a transition from red regions near Tc to green regions near 0 K, representing a shift from superconductivity limited by orbital de-pairing,
to Pauli-limited superconductivity. The solid green lines are WHH fits of the Hc2 data above the low-temperature upturn, constrained to second-order
superconducting transitions at all temperatures. The solid red line (a) is a three-dimensional Ginzburg–Landau (GL) fit of the Hc2,⊥ data near Tc. The black
arrows indicate the free-electron Pauli-limited field HPauli ; μ ¼ 1 μB ¼ 1:86 ´ Tc ðH ¼ 0 TÞ, where μ is the magnetic moment in units of the Bohr magneton,
μB. The red arrows indicate the effective Pauli-limitedI field HPauli ; μ ¼ 2:4 μB, which corresponds to the enhanced magnetic susceptibility deduced near
Tc. The obtained values of the zero-temperature coherence length I along the a–b plane, ξab(0), and c axis, ξc(0), as well as the lattice parameters a, b and c
along the corresponding directions are listed at the bottom of each panel for comparison. The inset of a overlays the Hc2 data and the corresponding WHH
fits for both field orientations. The inset of b overlays the Hc2,|| data of two samples with different thicknesses, with solid lines as guides to the eye.

a b c

2 U0 ∝ H −0.22
100
U0 ∝ H −0.27
0.2 T 0.2 T
1
U0/kB (K)

12 T 10
logρ

0 H⊥ U0 ∝ H −5.34
H∥
1 U0 ∝ H −5.86
12 T

–1 H⊥ H∥

0.1
0 4 8 12 16 20 24 0 10 20 30 40 50 60 70 2 3 4 5 67 2 3 4 5 67
1 10
U0T –1 U0T –1 µ0H (T)

Fig. 3 | Transport signatures of vortex motion, and the magnetic field dependence of the activation energy. a,b, Resistivity ρ(T) in scaled Arrhenius plots
at field strengths H = 0.2, 0.4, 0.6, 0.8, 1, 2, 3, 4, 6, 8, 10 and 12 T (ranging in colour from dark red to dark blue) for magnetic fields along the c axis (a) and
in the a–b plane (b). Grey arrows indicate the direction of increasing field strength. The x axes are scaled by the activation energies U0 that are obtained
through fits of the resistive transition in the thermally activated dissipation region according to ρ(T,H) = ρ0(H) exp(−U0/kBT). The dashed black lines are
representative linear fits. c, The extracted U0 as a function of magnetic field in double logarithmic scale for both field orientations. The error bars are within
the symbol size and correspond to one standard deviation. The solid lines are power-law fits of U0 below 2 K for fields along the c axis and in the a–b plane,
and the dashed lines are power-law fits of U0 above 5 K.

Nature Physics | VOL 17 | April 2021 | 473–477 | www.nature.com/naturephysics 475

Letters NATUre PHySIcS

a b c
7 40
12 Normal state H⊥ Normal state H∥
H⊥
6
Hc1% Hc1% H∥
10 30
HcVM HcVM
Hc50% Hc50% 5
Su
Su HcAL pe HcAL
8 pe rc
rc Hc90% on Hc90% 20

(µ0Hc2)2 (T2)
on du 4

µ0Hc2 (T)
µ0Hc (T)

du cti
ct ng
in flu
6 g ctu
flu 3
ct at 10
ua ion
tio s
ns Vo
4 rte 2
x
m
Vo ot 0
io
rte n
2 x 1
m
ot
io SC
n
SC
0 –10
0
5.0 6.0 7.0 8.0 9.0 10.0 5.0 6.0 7.0 8.0 9.0 10.0 0.60 0.70 0.80 0.90 1.00
T (K) T (K) T/Tc(H = 0 T)

Fig. 4 | Hc versus T phase diagram and temperature dependence of Hc2 near Tc. a,b, Hc–T phase diagrams for magnetic fields along the c axis (a) and in the
a–b plane (b). The shaded regions separated by Hc1%, HcVM, HcAL and Hc90% indicate superconductivity (SC), vortex motion, superconducting fluctuations and
the normal state. c, Temperature dependence of Hc2 (for which we use Hc50%) and (Hc2)2 near Tc for fields along the c axis and in the a–b plane, respectively.
The blue and red arrows indicate the axes for the Hc2 data points of the corresponding colours. The dashed lines are linear fits of Hc2 and (Hc2)2 above 85%
of T/Tc(H = 0 T). These fits are also shown in a and b as blue lines.

thickness of the film (Supplementary Table 1), which we take to be effective mass meff. A number of consistency checks (Supplementary
an unphysical discrepancy. Furthermore, this analysis presumes Information) confirm that the spin–orbit scattering time is much
confinement, which would prevent vortex entry for H|| and therefore longer than the elastic-transport scattering time (assumed in
is inconsistent with the evident vortex motion in this orientation WHH), a reasonable electronic diffusion coefficient is deduced,
(Fig. 3c). Finally, 2D confinement of an orbital-limited supercon- ξab(0) is sensibly related to the clean-limit coherence length and the
ductor usually leads to a dramatic enhancement of Hc2,|| over Hc2,⊥ transport mean free path, and α is consistent with estimates from
that is inversely proportional to thickness, unlike the response the critical regime near Tc.
observed here (Fig. 2a,b, Supplementary Fig. 9 and Supplementary Such overall consistency and the high quality of the fits allow us
Information). Given these discrepancies, we can rule out the 2D to draw some qualitative conclusions robust to the fitting details.
geometric confinement scenario. The most prominent feature is the high values of α for both ori-
Next, we investigate a different origin for the behaviour of Hc2,||. entations, which indicates the strong presence of the paramagnetic
In fact, if the superconductor is Pauli-limited, the paramagnetic effect. In particular, the exceptionally large α|| supports the purely
effect can also give a (Tc – T)1/2 dependence near Tc (refs. 26–29), Pauli-limited scenario for Hc2,||. Furthermore, isotropy of the effective
which arises from the Zeeman energy gain of the normal state electronic magnetic moment can be deduced from the observation
(~μH) competing against the superconducting condensate energy of a nearly isotropic Hc2 approaching 4 K, as the paramagnetic effect
(~H2/8π). In this case, fitting the (Tc – T)1/2 dependence allows us to plays an increasingly important role (Supplementary Information).
extract the in-plane electronic effective magnetic moment28 μ near This transition is visible through the substantial suppression of Hc2,⊥
Tc to be 2.37 ± 0.01 Bohr magnetons, which indicates an enhanced below the purely orbital-limited response with decreasing tem-
magnetic susceptibility of the normal state electrons, as seen in perature, as shown by the pair-breaking picture (Supplementary
other unconventional superconductors30. Additionally, as the purely Information). This manifestation of a dominant paramagnetic effect
orbital-limited Hc2,|| necessarily exceeds the measured (Pauli-limited) shows that the nickelate superconductor has singlet pairing and,
Hc2,||, the observation of vortex motion below TcVM provides only an therefore, even parity. The large α values also indicate that super-
upper-bound estimate of the zero-temperature coherence length conductivity here is in the dirty limit, in agreement with transport
along the c axis ξc(0) ≤ 10 Å, or 3.3 NiO2 layers (Supplementary estimates2. Finally, we note that the enhanced susceptibility deduced
Information), which corresponds to ξc(0)/ξab(0) ≤ 0.23. independently from Fig. 4c (Hc2,||) is consistent with the scale for
To quantitatively describe both orientations and lower tem- the Pauli limit that is relevant for the WHH fit (with enhancements
peratures beyond the critical regime, clearly both orbital and associated with λso).
paramagnetic effects need to be considered, as captured in Below 4 K, the anomalous upturns of Hc2,⊥ and Hc2,|| require
Werthamer–Helfand–Hohenberg (WHH) theory31. Recognizing considerations beyond conventional WHH theory19,28,32,33. Similar
the unusual nature of the prominent upturn at lower tempera- upturns, of much debated origin, have been observed for a number
tures, first we focus on T > 4 K and fit the data (independently for of layered superconductors19,34. We are not in a position to suggest
each orientation) to the single-band WHH form, constrained to a most-likely scenario, but briefly discuss a few possibilities here.
a second-order superconducting transition for all temperatures. The behaviour is tantalizingly similar to expectations for the Fulde–
We use single-band WHH, given that no signatures of multiband Ferrell–Larkin–Ovchinnikov state33; however, the fact that we are
superconductivity were observed near Tc (ref. 32, Supplementary in the dirty limit is inconsistent with anticipated requirements.
Information). The three fit parameters that are allowed to vary The upturn could indicate the onset of two-band superconductiv-
(Supplementary Table 1) are the Maki parameter α (which char- ity32, but the second-band contribution also would be expected to
acterizes the ratio between the orbital-limited and Pauli-limited saturate to an isotropic, Pauli-limited value (see the Supplementary
upper critical fields), the spin–orbit scattering parameter λso and the Information for a multiband analysis of Hc2). Finally, we discuss the

476 Nature Physics | VOL 17 | April 2021 | 473–477 | www.nature.com/naturephysics

NATUre PHySIcS Letters
possible onset of a first-order superconducting transition28,31. This 11. Ihara, H. et al. How to make superconducting-anisotropy least in high-Tc
tendency is expected for a strongly Pauli-limited superconduc- cuprate superconductors. Physica C 282, 1973–1974 (1997).
12. Yoshida, K., Maeno, Y., Nishizaki, S. & Fujita, T. Anisotropic
tor at low temperatures, although with different curvature35. If we superconductivity of Sr2RuO4. Physica C 263, 519–522 (1996).
remove the second-order restriction to our WHH fit, a comparable 13. Chou, F. C. et al. Thermodynamic and transport measurements of
description of the data can be obtained in the high-temperature superconducting Na0.3CoO2·1.3H2O single crystals prepared by
region, although with substantially smaller λso. Therefore, λso should electrochemical deintercalation. Phys. Rev. Lett. 92, 157004 (2004).
14. Yuan, H. Q. et al. Nearly isotropic superconductivity in (Ba,K)Fe2As2. Nature
be considered under-constrained by our data, whereas the other
457, 565–568 (2009).
conclusions we draw from WHH are robust to the possibility of 15. Aslamazov, L. G. & Larkin, A. I. Effect of fluctuations on the properties of a
a first-order transition and quite reproducible (Supplementary superconductor above the critical temperature. Sov. Phys. Solid State 10,
Information). 875–880 (1968).
In summary, our studies of Hc2 in Nd0.775Sr0.225NiO2 indicate that 16. Fulde, P. & Maki, K. Fluctuations in high field superconductors. Z. Phys. 238,
233–248 (1970).
superconductivity is strongly and nearly isotropically Pauli-limited 17. Palstra, T. T. M., Batlogg, B., van Dover, R. B., Schneemeyer, L. F. &
with singlet pairing and even parity, although ξc(0)/ξab(0) ≤ 0.23. Waszczak, J. V. Dissipative flux motion in high-temperature superconductors.
Therefore, it is fortuitous that we can extract quasi-intrinsic prop- Phys. Rev. B 41, 6621–6632 (1990).
erties from measurements of thickness-limited films. The micro- 18. Kim, J. -J. et al. Flux-creep dissipation in epitaxial YBa2Cu3O7−δ film:
scopic origins of the enhanced magnetic susceptibility remain to magnetic-field and electrical-current dependence. Phys. Rev. B 43,
2962–2967 (1991).
be understood and may reflect electronic correlations and proxi- 19. Brandow, B. Characteristic features of the exotic superconductors. Phys. Rep.
mate magnetic instabilities. It is intriguing that the low-temperature 296, 1–63 (1998).
behaviour of Hc2,|| extrapolates to the bare Pauli limit, which sug- 20. Maloney, M. D., De La Cruz, F. & Cardona, M. Superconducting parameters
gests that the anomalous low-temperature upturn and enhanced and size effects of aluminum films and foils. Phys. Rev. B 5, 3558–3572 (1972).
susceptibility may have a common origin. 21. Kozuka, Y. et al. Two-dimensional normal-state quantum oscillations in a
superconducting heterostructure. Nature 462, 487–490 (2009).
22. Harper, F. E. & Tinkham, M. The mixed state in superconducting thin films.
Online content Phys. Rev. 172, 441–450 (1968).
Any methods, additional references, Nature Research report- 23. Kim, M., Kozuka, Y., Bell, C., Hikita, Y. & Hwang, H. Y. Intrinsic spin–orbit
ing summaries, source data, extended data, supplementary infor- coupling in superconducting δ-doped SrTiO3 heterostructures. Phys. Rev. B
mation, acknowledgements, peer review information; details of 86, 085121 (2012).
24. Wu, X. S., Adams, P. W., Yang, Y. & McCarley, R. L. Spin proximity effect in
author contributions and competing interests; and statements of ultrathin superconducting Be–Au bilayers. Phys. Rev. Lett. 96, 127002 (2006).
data and code availability are available at https://doi.org/10.1038/ 25. Shalom, M. B., Sachs, M., Rakhmilevitch, D., Palevski, A. & Dagan, Y. Tuning
s41567-020-01128-5. spin–orbit coupling and superconductivity at the SrTiO3/LaAlO3 interface: a
magnetotransport study. Phys. Rev. Lett. 104, 126802 (2010).
26. Chandrasekhar, B. S. A note on the maximum critical field of high-field
Received: 27 June 2020; Accepted: 17 November 2020; superconductors. Appl. Phys. Lett. 1, 7–8 (1962).
Published online: 4 January 2021 27. Clogston, A. M. Upper limit for the critical field in hard superconductors.
Phys. Rev. Lett. 9, 266–267 (1962).
References 28. Maki, K. & Tsuneto, T. Pauli paramagnetism and superconducting state. Prog.
1. Li, D. et al. Superconductivity in an infinite-layer nickelate. Nature 572, Theor. Phys. 31, 945–956 (1964).
624–627 (2019). 29. Tedrow, P. M., Meservey, R. & Schwartz, B. B. Experimental evidence for a
2. Li, D. et al. Superconducting dome in Nd1–xSrxNiO2 infinite layer films. first-order magnetic transition in thin superconducting aluminum films. Phys.
Phys. Rev. Lett. 125, 027001 (2020). Rev. Lett. 24, 1004–1007 (1970).
3. Zeng, S. et al. Phase diagram and superconducting dome of infinite-layer 30. Zocco, D. A., Grube, K., Eilers, F., Wolf, T. & Löhneysen, H. V. Pauli-limited
Nd1–xSrxNiO2 thin films. Phys. Rev. Lett. 125, 147003 (2020). multiband superconductivity in KFe2A2. Phys. Rev. Lett. 111, 057007 (2013).
4. Norman, M. R. Entering the nickel age of superconductivity. Physics 13, 31. Werthamer, N. R., Helfand, E. & Hohenberg, P. C. Temperature and purity
85 (2020). dependence of the superconducting critical field, Hc2. III. Electron spin and
5. Lee, K. -W. & Pickett, W. E. Infinite-layer LaNiO2: Ni1+ is not Cu2+. Phys. Rev. spin–orbit effects. Phys. Rev. 147, 295–302 (1966).
B 70, 165109 (2004). 32. Gurevich, A. Enhancement of the upper critical field by nonmagnetic
6. Botana, A. S. & Norman, M. R. Similarities and differences between impurities in dirty two-gap superconductors. Phys. Rev. B 67, 184515 (2003).
LaNiO2 and CaCuO2 and implications for superconductivity. Phys. Rev. X 10, 33. Matsuda, Y. & Shimahara, H. Fulde–Ferrell–Larkin–Ovchinnikov state in
011024 (2020). heavy fermion superconductors. J. Phys. Soc. Jpn 76, 051005 (2007).
7. Wu, X. et al. Robust dx2 y2 -wave superconductivity of infinite-layer nickelates. 34. Lortz, R. et al. Calorimetric evidence for a Fulde–Ferrell–Larkin–
Phys. Rev. B 101, 060504
I (2020). Ovchinnikov superconducting state in the layered organic superconductor
8. Jiang, M., Berciu, M. & Sawatzky, G. A. Critical nature of the Ni spin state in κ-(BEDT-TTF)2Cu(NCS)2. Phys. Rev. Lett. 99, 187002 (2007).
doped NdNiO2. Phys. Rev. Lett. 124, 207004 (2020). 35. Tedrow, P. M. & Meservey, R. Measurement of the supercooling curve of a
9. Bernardini, F., Olevano, V. & Cano, A. Magnetic penetration depth and Tc in paramagnetically limited superconductor. Phys. Lett. 63A, 398–400 (1977).
superconducting nickelates. Phys. Rev. Res. 2, 013219 (2020).
10. Zhang, G. -M., Yang, Y. -F. & Zhang, F. -C. Self-doped Mott insulator Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
for parent compounds of nickelate superconductors. Phys. Rev. B 101, published maps and institutional affiliations.
20501 (2020). © The Author(s), under exclusive licence to Springer Nature Limited 2021

Nature Physics | VOL 17 | April 2021 | 473–477 | www.nature.com/naturephysics 477

Letters NATUre PHySIcS

Methods Acknowledgements
Thin-film synthesis. The samples were grown on SrTiO3 substrates using We thank R. L. Greene, A. Kapitulnik, S. A. Kivelson, P. B. Littlewood, B. Maiorov, G. A.
pulsed-laser deposition under growth and reduction conditions previously Sawatzky, H. Takagi, R. Thomale, A. Viswanathan and Y.-H. Zhang for discussions. The
reported2,36. work at SLAC and Stanford was supported by the US Department of Energy, Office
of Basic Energy Sciences, Division of Materials Sciences and Engineering (contract
HAADF-STEM characterization. Electron transparent STEM samples were no. DE-AC02-76SF00515) and the Gordon and Betty Moore Foundation’s Emergent
prepared on a Thermo Fischer Scientific Helios G4 X focused ion beam Phenomena in Quantum Systems Initiative (grant no. GBMF9072, synthesis equipment).
using the standard lift-out method, as was previously reported2,37. Samples B.H.G. and L.F.K. acknowledge support by the US Department of Defense Air Force
were thinned to < 30 nm with 2 kV Ga ions, followed by a final polish at 1 kV Office of Scientific Research (no. FA 9550-16-1-0305). This work made use of the
to reduce effects of surface damage. Then, HAADF-STEM characterization of Cornell Center for Materials Research Shared Facilities, which are supported through the
the sample was obtained on an aberration-correction FEI Titan Themis at an US National Science Foundation (NSF) MRSEC Program (no. DMR-1719875). The FEI
accelerating voltage of 300 kV with a convergence angle of 30 mrad and inner Titan Themis 300 was acquired through NSF grant no. MRI-1429155, with additional
collection angle of 68 mrad. support from Cornell University, the Weill Institute and the Kavli Institute at Cornell.
The Thermo Fisher Helios G4 X focused ion beam was acquired with support from the
NSF Platform for Accelerated Realization, Analysis and Discovery of Interface Materials
Transport measurements. The samples were contacted using wire-bonded (PARADIM) under Cooperative Agreement no. DMR-1539918.
aluminium wires in a 6-point Hall bar geometry. The 50% resistive criterion is
applied to both temperature-sweep and field-sweep data to determine Tc/Hc2. In
our measurement, the sample is aligned parallel or perpendicular to the applied Author contributions
magnetic field by performing angle sweeps of the sample resistance at 8 K in 3 T on D.L., K.L. and M.O. grew the nickelate films and conducted the reduction experiments
a rotator. Specifically, the angular dependence of the resistance is symmetric about and structural characterization. B.H.G. and L.F.K. conducted electron microscopy. B.Y.W.
the principal axes that correspond to the conditions in which the magnetic field is performed the transport measurements and analysis with M.R.B. B.Y.W., S.P.H. and
aligned along the c axis or in the a–b plane of the sample. H.Y.H. wrote the manuscript with input from all authors.

Competing interests
Data availability The authors declare no competing interests.
Source data are provided with this paper. All other data that support the
plots within this paper and other findings of this study are available from the
corresponding author upon reasonable request. Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/
s41567-020-01128-5.
References Correspondence and requests for materials should be addressed to B.Y.W. or H.Y.H.
36. Lee, K. et al. Aspects of the synthesis of thin film superconducting
infinite-layer nickelates. APL Mater. 8, 041107 (2020). Peer review information Nature Physics thanks Neil Harrison and the other, anonymous,
37. Goodge, B. H. et al. Doping evolution of the Mott–Hubbard landscape in reviewer(s) for their contribution to the peer review of this work.
infinite-layer nickelates. Preprint at https://arxiv.org/abs/2005.02847 (2020). Reprints and permissions information is available at www.nature.com/reprints.

Nature Physics | www.nature.com/naturephysics