Effect of Pressure and Oxygen-Isotope Substitution on Density-Wave Transitions in La$_4$Ni$_3$O$_{10}$

Abstract

Understanding the interplay between magnetism and superconductivity in nickelate systems is a key objective in condensed matter physics. Here, we present a systematic muon-spin rotation/relaxation ($μ$SR) and resistivity study of the trilayer Ruddlesden-Popper nickelate La$_4$Ni$_3$O$_{10}$ under ambient and applied pressure, combined with oxygen-isotope substitution. At ambient pressure, two incommensurate spin-density-wave (SDW) transitions are identified at $T_{SDW}\simeq132$ K and $T^\ast\simeq80-90$ K. Comparison of the internal magnetic fields with dipole-field calculations reveals a magnetic structure consistent with antiferromagnetically coupled SDW order on the outer two Ni layers, with smaller moments on the inner layer. Above $T^\ast$, the moments lie mainly in the $ab$ plane, whereas below this temperature they develop a $c$-axis component. The internal fields at the muon stopping sites appear abruptly at $T_{SDW}$, suggesting a first-order-like SDW transition closely linked to the charge-density-wave (CDW) order occurring at the same temperature ($T_{SDW}=T_{CDW}$). Under pressure, all transition temperatures -- $T_{SDW}$, $T^\ast$, and $T_{CDW}$ -- are suppressed at a nearly uniform rate of $\simeq-13$ K/GPa. This contrasts with bilayer La$_3$Ni$_2$O$_7$, where pressure enhances the separation between the SDW and CDW transitions. Oxygen-isotope substitution ($^{16}$O $\rightarrow$ $^{18}$O) shifts $T_{CDW}$ to higher values. The isotope effect on $T_{SDW}$ and $T^\ast$ differs markedly: when CDW and SDW are intertwined, a notable isotope effect is observed on $T_{SDW}$, yielding nearly identical isotope shifts for $T_{CDW}$ and $T_{SDW}$, whereas no isotope effect is detected at $T^\ast$, where the SDW transition occurs independently of the CDW.

Figures (10)

• Figure 1: (a)--(b) X-ray diffraction patterns of pristine [panel (a)] and oxygen-isotope ($^{16}$O/$^{18}$O) substituted [panel (b)] La$_4$Ni$_3$O$_{10}$ samples taken at room temperature. The x-ray data were refined using the $P2_1/a$ [panel (a)] and $Bmab$ [panel (b)] structures. (c)--(d) Thermogravimetric curves of the pristine [panel (c)] and oxygen-isotope substituted [panel (d)] La$_4$Ni$_3$O$_{10}$ samples. (e)--(f) Room-temperature Raman spectra collected for pristine [panel (e)] and oxygen-isotope-substituted [panel (f)] La$_4$Ni$_3$O$_{10}$ samples. The dashed lines indicate several Raman modes of the samples containing the lighter oxygen isotope. (g) Photograph of polycrystalline La$_4$Ni$_3$O$_{10}$, La$_4$Ni$_3\,^{16}$O$_{10}$, and La$_4$Ni$_3\,^{18}$O$_{10}$ samples mounted on the measurement chip and prepared for resistivity measurements. (h) Temperature dependence of resistivity normalized to its 300 K value $R(T)/R(300)$ for pristine and isotope-substituted La$_4$Ni$_3$O$_{10}$ samples.
• Figure 2: (a)--(b) Zero-field $\mu$SR time spectra of the pristine La$_4$Ni$_3$O$_{10}$ sample collected at $T = 5$ K [panel (a)] and $T = 110$ K [panel (b)]. (c)--(d) Fourier transforms of the data shown in panels (a) and (b). Solid lines represent the fit (red) and individual fit components. Dashed lines indicate the positions of internal field peaks expected from the 6-component fits. (e) Temperature evolution of the Fourier transform spectra, showing the transition of the magnetic field distribution from a 5-peak to a 2-peak structure. (f) Temperature dependence of the ZF-$\mu$SR signal fractions. The solid line corresponds to a fit of Eq. \ref{['eq:magnetic-fraction']} to $f_{\rm m}(T)$. (g) Temperature dependencies of the internal fields. The gray and pink lines in panels (f) and (g) indicate the positions of the magnetic ordering temperatures $T^\ast$ and $T_{\rm SDW}$, respectively.
• Figure 3: (a) The three muon stopping sites with their relation to the Ni trilayer structure. Symmetrically equivalent positions are all shown, although in the experiment only one position is occupied at any one time. (b) The magnetic structure used in simulations of the ZF $\mu$SR spectra. There is a large moment on the outer Ni layers, with a small one on the inner Ni layer. All moments point in the $\bm{b}$ direction, and form a SDW. There is a phase shift of 90$^\circ$ between each layer. (c) and (d) Simulations of the dipole field at the muon stopping sites with the moments either pointing in the $a$b-plane ($\theta~=~0^\circ$), or rotated towards the $c$-axis ($\theta~=~20^\circ$). Colours under the curves correspond to the colours of the different fit components in Fig. \ref{['fig:Ambient-pressure_muons']} (c) and (d), highlighting the qualitative similarity.
• Figure 4: (a)--(b) Temperature dependence of the magnetic volume fraction $f_{\rm m}$ measured in WTF-$\mu$SR experiments at $p = 0.0$ GPa [panel (a)] and $p = 1.95$ GPa [panel (b)]. The solid lines represent fits of Eq. \ref{['eq:magnetic-fraction']} to the data. (c)--(d) Temperature evolution of the internal fields measured in ZF-$\mu$SR experiments at $p = 0.0$ GPa [panel (c)] and $p = 1.95$ GPa [panel (d)]. The gray and pink lines in panels (a)--(d) indicate the positions of the magnetic ordering temperatures $T^\ast$ and $T_{\rm SDW}$, respectively. (e) Pressure dependence of the magnetic ordering temperatures $T_{\rm m,onset}$, $T_{\rm SDW}$, and $T^\ast$, as well as the charge-density-wave ordering temperature $T_{\rm CDW}$. The solid lines are linear fits. (f) Pressure dependence of the internal field components measured in ZF-$\mu$SR experiments at $T = 20$ K. The solid lines are linear fits.
• Figure 5: (a) Temperature dependencies of resistivity $R$ measured at pressures $p = 0.22$, 0.64, 1.50, and 2.10 GPa. (b) First derivatives of the resistivity curves. Red dots indicate the CDW transition temperature $T_{\rm CDW}$.