The magnetic properties of geometrically frustrated systems are among the hot topics in condensed matter physics. In such systems, a variety of unconventional phases appears due to frustrated spin interactions that couple with lattice, orbital and charge degrees of freedom. We can tune the balance between them using strong magnetic field through the Zeeman interaction of spins, and various novel magnetic phases have been found in magnetic fields exceeding 20 T. Neutron scattering is a powerful and valuable technique to directly determine the space- and time-correlation of magnetic moments, and to have a deeper insight into the origin of the novel phases in high fields. So far, however, sufficiently high magnetic fields have not been accessible for neutron scattering studies.
Here, we present the first successful application of pulsed high magnetic fields for neutron diffraction experiments. This new technique was used to investigate the magnetic structure of the geometrically frustrated system TbB4, in magnetic fields up to 30 T. The experiment was carried out on the high-flux, low background triple-axis spectrometer IN22 (Institut Laue Langevin , Grenoble), with a neutron wavelength of 1.53 Å, using a transportable pulsed magnet system built up by Prof. Nojiri’s group (IMR, Tohoku University, Sendai) and the mobile capacitor bank developed at the LNCMI-T.
At room temperature, TbB4 crystallizes in the tetragonal structure P4/mbm. The network of magnetic Tb ions consists of squares and triangles whose configuration within the c-plane is characterized by orthogonal dimers topologically equivalent to the two dimensional Shastry-Sutherland lattice (SSL) . In zero field, TbB4 exhibits two successive antiferromagnetic transitions at TN1 = 44 K and TN2 = 24 K . The magnetic structure at B = 0 T is a XY-type noncollinear structure . At low temperature and high fields (between 16 and 30 T), multi-step magnetization plateaus appear when the magnetic field is perpendicular to the magnetic easy plane (figure below, left) . This behaviour is unusual since successive metamagnetic transitions in rare earth intermetallics are generally expected when the field is parallel to the Ising easy axis . Therefore the determination of magnetic structure of TbB4 in high magnetic field is essential to understand this phenomenon.
Within the constraints of the pulsed magnet system developed for neutron scattering, we have succeeded in measuring the field dependences of four Bragg reflections ((100), (200), (110) and (220)) up to 30 T. The neutron counts and magnetic fields were measured simultaneously using a time analysing system so that the field dependence of Bragg peaks at fixed diffraction angle can be monitored and recorded as a function of time. The accumulation of 100-200 magnetic field pulses for each reflection was required to acquire statistically relevant data. As an example, the field dependence of the intensity of the (100) peak is reported on figure below (right). The relative intensity against the zero field intensity I/I0 and the relative magnetization M/MS are reported in the figure. Stepwise changes which can be related to the magnetization process are clearly observed in the intensity of the magnetic (100) reflection.
From these experimental results, different magnetic models have been calculated. They indicate a much weaker AF correlation in TbB4 than the expected one in a conventional antiferromagnet and also suggest the presence of significant anisotropic interactions. A model introducing a biquadratic interaction term which stabilizes an orthogonal configuration of magnetic moments was proposed to explain the multiple magnetization plateaus. Further investigations of the magnetic scattering diagram in high fields would be needed to confirm this model. Nevertheless, these results clearly demonstrate the large potential of pulsed magnetic fields for neutron diffraction experiments.
 B.S. Shastry and B. Sutherland, Physica B+C 108, 1069 (1981).
 Z. Fisk, M.B. Maple, D.C. Johnston, and L.D. Woolf, Solid State Commun. 39, 1189 (1981).
 T. Matsumura, D. Okuyama, and Y. Murakami, J. Phys. Soc. Jpn. 76, 015001 (2007).
 S. Yoshii, T. Yamamoto, M. Hagiwara, S. Michimura, A. Shigekawa, F. Iga, T. Takabatake, and K. Kindo, Phys. Rev. Lett. 101, 087202 (2008).
 J. Rossat-Mignot, P. Burlet, S. Quezel, J.M. Effantin, D. Delacôte, H. Bartholin, O. Vogt, and D. Ravot, J. Magn. Magn. Mater. 31-34, 398 (1983).
Members of the laboratory implied in this activity
H. Nojiri, K. Ohoyama, S. Yoshii (IMR, Université de Tohoku, Japon)
M. Matsuda (JAEA, Tokai, Japon)
L.-P. Regnault (CEA, Grenoble)