Intrinsic Spin Susceptibility and Pseudogaplike Behavior in Infinite-Layer ${\mathrm{LaNiO}}_2$

The recent discovery of superconductivity in doped infinite-layer nickelates has stimulated intensive interest, especially for similarities and differences compared to that in cuprate superconductors. In contrast to cuprates, although earlier magnetization measurement reveals a Curie-Weiss-like behavior in undoped infinite-layer nickelates, there is no magnetic ordering observed by elastic neutron scattering down to liquid helium temperature. Until now, the nature of the magnetic ground state in undoped infinite-layer nickelates was still elusive. Here, we perform a nuclear magnetic resonance (NMR) experiment through ${}^{139}\mathrm{La}$ nuclei to study the intrinsic spin susceptibility of infinite-layer ${\mathrm{LaNiO}}_2$. First, the signature for magnetic ordering or freezing is absent in the ${}^{139}\mathrm{La}$ NMR spectrum down to 0.24 K, which unambiguously confirms a paramagnetic ground state in ${\mathrm{LaNiO}}_2$. Second, a pseudogaplike behavior instead of Curie-Weiss-like behavior is observed in both the temperature-dependent Knight shift and nuclear spin-lattice relaxation rate ($1/T_1$), which is widely observed in both underdoped cuprates and iron-based superconductors. Furthermore, the scaling behavior between the Knight shift and $1/T_{1}T$ has also been discussed. Finally, the present results imply a considerable exchange interaction in infinite-layer nickelates, which sets a strong constraint for the proposed theoretical models.

Figure 1

Crystal structure, XRD result, and ${}^{139}\mathrm{La}$ NMR spectrum in ${\mathrm{LaNiO}}_2$. (a) Sketch of structural change from ${\mathrm{LaNiO}}_3$ to ${\mathrm{LaNiO}}_2$ by topotactic reduction with metal hydrides ${\mathrm{CaH}}_2$. (b) X-ray diffraction pattern of the ${\mathrm{LaNiO}}_2$ powder sample. The main structural phase is proved to be ${\mathrm{LaNiO}}_2$, and the asterisk indicates a small amount of impurity after topotactic reduction. (c) The full NMR spectra of ${}^{139}\mathrm{La}$ nuclei with spin number $I=7/2$ and related fitting result (for details, see Sec. S4 in Supplemental Material [24]).

Figure 2

Temperature-dependent ${}^{139}\mathrm{La}$ NMR spectrum in ${\mathrm{LaNiO}}_2$. (a) Temperature-dependent central transition peak with the temperature range from 463 to 2.2 K. The value of external magnetic field is 12 T. (b) The comparison of full NMR spectrum between 2.2 and 0.24 K. No significant broadening effect has been observed. (c) Temperature-dependent linewidth of the central transition peak in (a). The linewidth is determined by a Gaussian fitting, and the corresponding error bar is the fitting error bar.

Figure 3

Temperature-dependent Knight shift and nuclear spin-lattice relaxation in ${\mathrm{LaNiO}}_2$. (a) The temperature-dependent $1/T_1$ of the ${\mathrm{LaNiO}}_3$ precursor and ${\mathrm{LaNiO}}_2$. (b) Upper panel: the temperature-dependent Knight shift extracted from Fig. 2; lower panel: the temperature-dependent $1/T_{1}T$ in ${\mathrm{LaNiO}}_2$. A similar pseudogaplike behavior has been observed in both the Knight shift and $1/T_{1}T$. For a detailed discussion, see the main text. The error bars for both the Knight shift and $1/T_1$ are determined by the fitting error bar.

Figure 4

Comparison of different scalings between the Knight shift and $1/T_{1}T$. Based on the data above 150 K in Fig. 3, different scalings between the Knight shift and $1/T_{1}T$ have been checked. (a) Modified Korringa relation between the Knight shift and $1/T_{1}T$; (b) conventional Korringa relation between the Knight shift and $1/T_{1}T$. The gray bold line indicates a linear relationship between the Knight shift and $1/T_{1}T$ in a different scaling method. For detailed discussion, see the main text.