부산대학교 도서관

No state of matter can be defined categorically by what it is not; yet spin liquids are often conjectured to exist based on the nonexistence of magnetic order as $T \to 0$. An emerging concept designed to circumvent this ambiguity is to categorically identify each spin liquid type by using its spectrum of spontaneous spin noise. Here we introduce such a spectroscopy to spin liquid studies by considering Ca$_{10}$Cr$_7$O$_{28}$. This is a spin liquid, but whether classical or quantum and in which specific state, are unknown. By enhancing the flux-noise spectrometry techniques introduced for magnetic monopole noise studies, here we measure the time and temperature dependence of spontaneous flux $\varPhi(t,T)$ and thus magnetization $M(t,T)$ of Ca$_{10}$Cr$_7$O$_{28}$ samples. The resulting power spectral density of magnetization noise $S_M(\omega,T)$ along with its correlation function $C_M(t,T)$, reveal intense spin fluctuations spanning frequencies $0.1\ \mathrm{Hz} \leq \omega/2\pi \leq 50\ \mathrm{kHz}$, and that $S_M(\omega,T)\propto \omega^{-\alpha(T)}$ with $0.84 < \alpha(T) < 1.04$. Predictions for quantum spin liquids yield a frequency-independent spin-noise spectrum, clearly inconsistent with this phenomenology However, when compared to Monte Carlo simulations for a 2D spiral spin liquid state that are accurately parameterized to describe Ca$_{10}$Cr$_7$O$_{28}$, comprehensive quantitative correspondence with the data including $S_M(\omega,T)$, $C_M(t,T)$ and magnetization variance $\sigma_M^2(T)$ fingerprint the state of Ca$_{10}$Cr$_7$O$_{28}$ as a spiral spin liquid.
Comment: 42 pages, 4 main figures, 8 supplementary figures, 2 supplementary movies