Electric-Field-Induced Spin Flop in ${\mathrm{BiFeO}}_3$ Single Crystals at Room Temperature

Bismuth ferrite, ${\mathrm{BiFeO}}_3$, is the only known room-temperature magnetic ferroelectric material. We demonstrate here, using neutron scattering measurements in high quality single crystals, that the antiferromagnetic and ferroelectric order parameters are intimately coupled. Initially in a single ferroelectric state, our crystals have a canted antiferromagnetic structure describing a unique cycloid. Under electrical poling, polarization reorientation induces a spin flop. We argue here that the coupling between the two orders may be stronger in the bulk than in thin films where the cycloid is absent.

Figure 1

(a) Neutron intensity around the ($\frac{1}{2}$, $\frac{-1}{2}$, $\frac{1}{2}$) Bragg reflection in the virgin, single domain, state. The two diffraction satellites indicate that the cycloid is along the $[10-\text{}\text{}1]$ direction. (b) Schematics of the 64 nm antiferromagnetic circular cycloid.

Figure 2

Mapping of the neutron intensity in reciprocal space. Two sets of splitting appear for the nuclear intensity (yellow or light gray spots) due to the presence of two ferroelastic domains [see (a)]: one because of the presence of two rhombohedral distortions along [111] and $[1\text{}-\text{}\text{}11]$, and the second because of a physical buckling of the crystal induced by the twinning. Magnetic peaks are further split because of the cycloids. Note that because the splitting is small, the scale has been magnified by a factor of 10 around each peak position.

Figure 3

Neutron intensity around the ($\frac{1}{2}$, $\frac{-1}{2}$, $\frac{1}{2}$) Bragg position in the multidomain state. Theoretical positions are indicated by the black and white spots. Diffraction satellites are visible in the 0° (bottom half) and 71° (top half) domains of polarization. The difference in vertical spot shape likely originates from the position of reversed domains at opposite ends of the sample because of a preferred nucleation near the edges. Any warping of the sample splits the new $(1\text{}-\text{}\text{}11)$ peak while recovering an improved resolution for the original domain located near the center.

Figure 4

Schematics of the planes of spin rotations and cycloids ${\overrightarrow{k}}_1$ vector for the two polarization domains separated by a domain wall (in light gray).