We report magnetization, nuclear magnetic resonance (NMR), nuclear quadrupole resonance (NQR), and transmission electron microscopy (TEM) studies on the quasi-two-dimensional spin-gap system (CuCl)LaNb 2 O 7 , a possible candidate for the J 1 -J 2 model on a square lattice. A sharp single NQR line is observed at the Cu and Cl sites, indicating that both Cu and Cl atoms occupy a unique site. However, the electric field gradient tensors at the Cu, Cl, and La sites do not have axial symmetry. This is incompatible with the reported crystal structure. Thus the J 1 -J 2 model has to be modified. We propose alternative twodimensional dimer models based on the NMR, NQR, and TEM results. The value of the hyperfine coupling constant at the Cu sites indicates that the spin density is mainly on the dð3z 2 À r 2 Þ orbital (z k c). At 1.5 K, Cu-and Nb-NMR signals disappear above the critical field B c1 ' 10:3 T determined from the onset of the magnetization, indicating a field-induced magnetic phase transition at B c1 .

We report nuclear magnetic resonance (NMR) studies in the antiferromagnetic state of the quasi-two-dimensional (CuBr)LaNb2O7. The NMR spectra at zero magnetic field and 4.2 K indicate a unique Cu and Br sites with an internal field of 5.7 T (at Cu) and 16.4 T (at Br), confirming a magnetic order. For the large internal field at the Br sites to be compatible with the collinear antiferromagnetic order observed by neutron diffraction experiments (N. Oba et al., J. Phys. Soc. Jpn. 75, (2006) 113601), the Br atoms must move significantly off the center of the square of the Cu sublattice so that the Br nuclei couple predominantly to two parallel Cu moments. While invalidating the frustrated J1-J2 model defined on a C4-symmetric square lattice, our results are compatible with the structural model proposed for (CuCl)LaNb2O7 by Yoshida et al. (J. Phys. Soc. Jpn. 76, (2007) 104703).