Figure 1

(a) ${\chi }^{\prime }$ as a function of temperature at various magnetic fields up to 49 mT, measured during upward temperature sweeps at $0.6\mathrm{mK}/\min$. The frequency and amplitude of the ac field were 0.21 Hz and 0.94 mT. The magnetic field increases from top to bottom, above the peaks. (b) ${\chi }^{\prime }$ as a function of magnetic field up to 0.25 T at various temperatures. The temperature decreases from top to bottom. At and above 85 mK, the frequency and amplitude of the ac field and the field-sweep rate were 87.6 Hz, 0.94 mT, and $4.2\mathrm{mT}/\min$. For lower temperatures, they were reduced to 9.6 Hz, 0.20 mT, and $0.625\mathrm{mT}/\min$. (With the ac-field amplitude of 0.94 mT, the 100 and 85 mK data were taken in a slightly nonlinear regime, at least at $H=0$. See the Supplemental Material [33].) (c) The data at and below 70 mK over a wider field range. Lines are field-sweep data. Circles are equilibrium data at 40 mK. The temperature decreases from top to bottom, near the lowest fields. (d) The 16 mK data up to 1.5 T, arrows indicating anomalies discussed in the text. See the Supplemental Material [33] for ${\chi }^{\prime \prime }$.Reuse & Permissions

Figure 2

$H\mathrm{\text{−}}T$ phase diagram determined from the ${\chi }^{\prime }$ peak positions. Triangles are from the temperature sweeps, with peak positions in downward and upward sweeps averaged. Other symbols are from field sweeps. Red circles and magenta diamonds are from the data shown in Figs. 1b, 1c; blue square from 0.21 Hz data, for which the ac amplitude and the field-sweep rate were 0.94 mT and $0.51\mathrm{mT}/\min$. Open symbols are the same data plotted against $H_{\mathrm{int}}$. Lines are guides to the eye. Inset: Contour plot of ${\chi }^{\prime }$ in the $H\mathrm{\text{−}}T$ plane in and near phase II, constructed from field and temperature sweeps at 9.6 Hz.Reuse & Permissions

Figure 3

Equilibrium ${\chi }^{\prime }$ (a) and ${\chi }^{\prime \prime }$ (b) near the ${\chi }^{\prime }$ peak in zero field. (c) ${\chi }^{\prime }$ at 40 mK as a function of magnetic field at various frequencies. In (c), the frequency decreases from top to bottom. The field was swept upward at $4.2\mathrm{mT}/\min$. For all panels, the ac-field amplitude $H_{\mathrm{ac}}$ was 0.94 mT. Thus, in (a) and (b), the 87.6 Hz data near 100 mK were taken in a slightly nonlinear regime. Similarly, the 87.6 Hz data in (c) were from a slightly nonlinear regime, at least at $H=0$ (see the Supplemental Material [33]). (d) ${\chi }^{\prime }$ at $H=0$ and 0.36 T as a function of frequency at various temperatures. $H_{\mathrm{ac}}$ was 0.20 mT. For the 40 mK data at $H=0$, arrows indicate the directions of frequency steps, showing a weak hysteresis. Hysteresis is absent from all the other data. See the Supplemental Material [33] for the imaginary part of the data shown in (c).Reuse & Permissions

Figure 4

(a) $|\chi |$ as a function of time at 40 mK in zero field. Before $t=0$, the ac frequency was held at 2.6 Hz for 12 h until equilibrium was reached. Thereafter, the frequency was changed stepwise, each time after new equilibrium was reached. For clarity, relaxation curves at 189.6, 269.6, and 389.6 Hz—between the two 87.6 Hz curves—are not shown. From left to right, the lines represent 9.6, 19.6, 49.6, and 87.6 Hz, followed by the same frequencies in descending order. (b) Relaxation time $\tau$ as a function of frequency, extracted from the data. Arrows indicate the frequency-step sequence. (c) $\tau$ versus temperature. At each temperature, the frequency was first held at 389.6 Hz until equilibrium was reached, then stepped down to 0.21 Hz, the frequency at which the measurements were made. (d) $1/\tau$ versus $T_p/T$, where $T_p=140\mathrm{mK}$. For all the panels, $H_{\mathrm{ac}}=0.20\mathrm{mT}$.Reuse & Permissions