Quantum Fisher information in a strange metal

Nat. Phys. 22, 1064-1070 (2026) · DOI: 10.1038/s41567-026-03298-0

License: CC BY 4.0.

Abstract

A strange metal is an exotic state of correlated quantum matter, and intensive efforts are ongoing to understand its nature. Here we show that the quantum Fisher information, a concept from quantum metrology, may provide helpful insights. We use inelastic neutron scattering and quantum Monte Carlo simulations to study quantum critical fluctuations of the Kondo destruction type, which are considered to underlie strange metal behaviour in heavy-fermion compounds. We find that the associated quantum Fisher information increases strongly and without a characteristic scale as the strange metal forms with decreasing temperature. This provides evidence for a quantum state with high multipartite entanglement and offers a positive descriptor of strange metallicity that points towards its microscopic basis. Our work opens a direction for future studies across a range of strange metal platforms.

Figures

7 panels with data across 3 figures. Each panel page shows the plot, its columns and its files; each data.csv begins with a header naming the paper, the panel, the source, the license and the provenance route.

Fig. 1

The underlying data is not available. The figure is shown in the paper PDF. Page 2 of the PDF.

Fig. 2

• panel (a): Dynamical spin correlation function $S(\mathbf{q},\omega)$ against energy $\hbar\omega$ at $\mathbf{q} = (0\bar{1}0)$ and $B = 1.73$ T, for isotherms at 0.06, 0.125, 0.5, 1.5, 3.2 and 10 K, with error bars and the thin line drawn through each isotherm in the print. Error bars of the 0.06 K points that the print hides under the symbol are left empty. data.csv
• panel (b): Scaling collapse: $S(\mathbf{q},\omega)$ from panel a multiplied by $(k_{\mathrm{B}}T)^{\alpha}$ with $\alpha = 0.88$, against $\hbar\omega/k_{\mathrm{B}}T$ on log axes, for twelve temperatures from 0.06 to 5 K (circles to 0.5 K, triangles from 0.78 K) at $B = 1.73$ T and $\mathbf{q} = (0\bar{1}0)$, with error bars. The print's legend reads 3.5 K for the series the deposit labels 3.2 K. data.csv
• panel (b-inset): Inset of panel b: quality factor $\chi^2_{\mathrm{reduced}}$ of the scaling collapse against the exponent $\alpha$, with its minimum at $\alpha_{\mathrm{min}} = 0.88$. data.csv

Fig. 3

• panel (a): Quantum Fisher information density $f_{\mathrm{Q}}$ of Ce$_3$Pd$_{20}$Si$_6$ against temperature: $B = 1.73$ T at $\mathbf{q} = (0\bar{1}0)$ (black) and one point at $B = 5.8$ T, $\mathbf{q} = (100)$ (blue), with error bars, and the line drawn through the 1.73 T points in the print. data.csv

Fig. 4

• panel (a): Quantum Monte Carlo QFI density $f_{\mathrm{Q}}$ of the spin degree of freedom against temperature $T$ at $J_k = 2.0$, $L = 18$, for $q = 2\pi/9$, $5\pi/9$, $7\pi/9$ and $\pi$, and for $q = \pi$ at $L = 22$ (dashed). Finite-size effects set in below the dotted line ($T = 0.08$, shaded in the print). The print's inset, a magnified view of the lowest $q = \pi$ points, is not hosted. data.csv
• panel (b): Quantum Monte Carlo QFI density $f^{\Psi}_{\mathrm{Q}}$ of the composite fermion against temperature $T$ at $J_k = 2.0$, $L = 18$, for $q = 2.0$, 0.4, 0.7 and $1.1q_{\mathrm{F}}$ with $q_{\mathrm{F}} = \pi/2$. Below the dotted line ($T = 0.08$, shaded in the print) finite-size effects set in. The print puts the $f^{\Psi}_{\mathrm{Q}}$ axis on the right. data.csv
• panel (b-inset): Inset of panel b: the wavevector-independent sum-rule contribution $f^{\Psi,\mathrm{sum\ rule}}_{\mathrm{Q}}$ to $f^{\Psi}_{\mathrm{Q}}(T)$ against temperature $T$ for the same four wavevectors, which overlap. data.csv

Fig. 5

Illustrative figure, no extractable data. Shown in the paper PDF. Page 5 of the PDF.

Cite

Federico Mazza, Sounak Biswas, Xinlin Yan, Andrey Prokofiev, Paul Steffens, Qimiao Si, Fakher F. Assaad, Silke Paschen. Quantum Fisher information in a strange metal. Nat. Phys. 22, 1064-1070 (2026). https://doi.org/10.1038/s41567-026-03298-0

When you use hosted data, cite the original paper and give the panel's URL so a reader can find the exact values you used.

All papers · About · Statistics · Submit data · API