Robust but disordered collapsed-volume phase in a cerium alloy under the application of pulsed magnetic fields

We report synchrotron x-ray powder diffraction measurements of Ce${}_{0.8}$La${}_{0.1}$Th${}_{0.1}$ subject to pulsed magnetic fields as high as 28 T. This alloy is known to exhibit a volume collapse upon cooling at ambient pressure, which is a modification of the $\gamma$-$\alpha$ transition in elemental cerium. Recently, it has been suggested on the basis of field-cooled resistivity and pulsed-field magnetization measurements that the volume collapse in this alloy can be suppressed by the application of magnetic fields. Conversely, our direct diffraction measurements show a robust collapsed phase, which persists in magnetic fields as high as 28 T. We also observe nanoscale disorder in the collapsed phase, which increasingly contaminates the high-temperature phase upon thermal cycling.

Figure 1

Top: The incident x-ray pulse (the narrow pulse) is timed to arrive during the magnetic field peak (broader pulse). In this case, a 5.75-ms magnetic field pulse to 15.4 T weighted by an x-ray exposure with a 1-ms full width gives an average magnetic field value of $\sim$14.9 T, integrated over $\pm$0.5 T for the exposure. Middle: Raw data collected on the image plate for a single 1 ms exposure. Bottom: The same data, radially integrated and plotted as a function of momentum transfer.

Figure 2

Zero-field temperature dependence of the $\langle$111$\rangle$ Bragg reflection. The peak shifts to higher $\left|\mathbf{q}\right|$ and broadens upon cooling. The sample is a single phase at all temperatures. Resolution-limited peaks at high temperature are well described by Gaussians, while a Voigt shape at low temperature is consistent with an Ornstein-Zernike scattering function with a short correlation length (solid lines).

Figure 3

Zero-field volume collapse in Ce${}_{0.8}$La${}_{0.1}$Th${}_{0.1}$. Sample was warmed from 8 K to 250 K ($\blacksquare$) and then cooled back to 10 K ($•$). This thermal cycling led to a reduction in the magnitude of the volume collapse, shown by the dashed lines. The inset shows the Ornstein-Zernike correlation length extracted from the cooling measurements (see text). The correlated regions of the collapsed phase have an average size of only $\sim$20 nm. Conversely, the Bragg peaks above $T\sim 70$ K in the $\gamma$ phase are resolution limited, with $\xi$ greater than 1 $\mu$m.

Figure 4

Observed and expected effects of 28 T pulsed magnetic fields on the $\langle$111$\rangle$ Bragg peak at $T=35$ K. There is no discernible change in the position or width of the peak. The expected curve is calculated based on the position and width change that would arise from an 8-K shift in the transition temperature, consistent with the results of Ref. 5. The measured curves averaged results of five (ten) shots taken in zero field (28 T).

Figure 5

Volume collapse ($\Delta V/V$) and Ornstein-Zernike correlation length ($\xi$), comparing zero-field-cooled measurements and measurements where the system was repeatedly pulsed to 28 T upon cooling. No field effect is seen.

Figure 6

Gradual contamination by nanoscale disorder, induced by thermal cycling through the $\gamma$-$\alpha$ transition. The effects are clear in both the magnitude of the volume collapse ($\Delta V/V$) and the Ornstein-Zernike correlation length ($\xi$). All data were collected upon cooling in zero magnetic field. The three data sets represent the second ($•$), seventh ($\blacksquare$), and eleventh ($▴$) thermal cycle.