No abstract

The magnetization process of an S = 1 2 antiferromagnet on the kagomé lattice, [Cu3(titmb)2(OCOCH3)6]·H2O{titmb= 1,3,5-tris(imidazol-1-ylmethyl)-2,4,6 trimethyl-benzene} has been measured at very low temperatures in both pulsed and steady fields. We have found a new dynamical behavior in the magnetization process. A plateau at one third of the saturation magnetization Ms appears in the pulsed field experiments for intermediate sweep rates of the magnetic field and disappears in the steady field experiments. A theoretical analysis using exact diagonalization yields, J1 = −19 ± 2K and J2 = 6 ± 2K, for the nearest neighbor and second nearest neighbor interactions, respectively. This set of exchange parameters explains the very low saturation field and the absence of the plateau in the thermodynamic equilibrium as well as the two-peak feature in the magnetic heat capacity observed by Honda et al. [ Z. Honda et al., J. Phys.: Condens. Matter 14, L625 (2002)]. Supported by numerical results we argue that a dynamical order by disorder phenomenon could explain the transient appearance of the Ms/3 plateau in pulsed field experiments.

The Knight shift K and the nuclear spin–lattice relaxation time T1 of 7Li have been measured under high pressure up to 4.74 GPa. 1/T1T becomes larger on applying higher pressure below 10 K and does not obey the T1T = constant relation down to 1 K. Meanwhile, K is independent of pressure above 2 GPa, indicating that the uniform component of the susceptibility does not change under high pressure. These results indicate that some antiferromagnetic fluctuations with wavevector dominate the relaxation rate in LiV2O4 near the boundary of the pressure induced insulating phase. It is noted that the pressure dependence of T1 for LiV2O4 is opposite to that of typical Ce HF compounds, such as CeCu5.9Au0.1 and CeCu2Si2. At the highest pressure of 4.74 GPa, we found that 1/T1 obeys a power law dependence of T2/3 over the wide T range between 60 mK and 10 K. The spin dynamics under high pressure will be discussed.

No abstract

No abstract