Site-selective observation of spin dynamics of a Tomonaga-Luttinger liquid in frustrated Heisenberg chains Diep Minh Nguyen, Azimjon A. Temurjonov, Daigorou Hirai, Zenji Hiroi, Oleg Janson, Hiroshi Yasuoka, Taku Matsushita, Yoshiaki Kobayashi, Yasuhiro Shimizu Abstract Low-energy spin dynamics is investigated by 35 Cl NMR measurements in a frustrated antiferromagnet Ca 3 ReO 5 Cl 2 . The local spin susceptibility measured with the Knight shift behaves as a one-dimensional Heisenberg antiferromagnet and remains constant down to low temperatures, as expected in a gapless Tomonaga-Luttinger liquid. The nuclear spin-lattice relaxation rate T 1 − 1 T_{1}^{-1} demonstrates a slowing down of atomic motions and a power-law evolution of spin correlation. The Luttinger parameter is enhanced in a site-selective manner depending on the form factor of dynamical spin susceptibility. The strong anisotropy of T 1 − 1 T_{1}^{-1} reflects the strong spin-orbit coupling through Dzyaloshinskii-Moriya interaction. The ground state exhibits an incommensurate antiferromagnetic ordering with low-lying magnon excitations. I Introduction A Tomonaga-Luttinger liquid (TLL) is characterized by fractional gapless excitations and bosonic collective modes in one-dimensional (1D) localized spin chains [ 1 , 2 , 3 ] . The spin correlation function of TLL follows a universal power law. One of the observables is the nuclear spin-lattice relaxation rate T_{1}^{-1} , which measures the dynamical spin susceptibility summed over the wave vector space in a low-energy limit. It approximately obeys the power-law temperature T dependence [ 4 , 5 , 6 , 7 , 8 , 9 ] : T_{1}^{-1}\propto T^{1/2K_{\sigma}-1}, (1) where K_{\sigma} is one of the TLL parameters. A spin-1/2 Heisenberg chain yields K_{\sigma}=1/2 , in which T_{1}^{-1} becomes T -invariant [ 10 , 11 , 12 , 13 ] . K_{\sigma} depends on the Ising anisotropy and the magnetic field strength respectively acting as the interaction and chemical potential of spinless fermions [ 2 ] . As three-dimensional (3D) coupling between the chains sets in, the system heads toward long-range magnetic ordering with the effectively enhanced K_{\sigma} above the transition temperature T_{\rm N} [ 14 , 15 ] . On the spin-1/2 ladder and spin-1 chain exhibiting a nonmagnetic ground state, the spin correlation measured with K_{\sigma} sensitively depends on the magnetic field [ 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 ] . The anisotropic triangular antiferromagnet connects 1D TLL to two-dimensional (2D) quantum spin liquid as a function of the ratio of the interchain interaction J^{\prime} and the intrachain interaction J [ 2 , 24 , 25 , 26 , 27 ] . Geometrical frustration reduces the interchain correlation and hence the system dimensionality [ 28 , 29 ] . The TLL phase is stabilized up to the exchange anisotropy J^{\prime}/J\approx 0.3 for the interchain exchange coupling J^{\prime} and the intrachain coupling J [ 28 , 25 , 30 , 29 , 26 , 27 ] . A spiral magnetic order occurs for J^{\prime}/J>0.3 - 0.7 on the triangular lattice. The quantum spin liquid, if any, near the isotropic point J^{\prime}/J\approx 1 may involve a tiny spin gap and topological order [ 10 , 31 ] , which is distinct from the renormalized 1D liquid phase. An example of the spin-1/2 anisotropic triangular antiferromagnets, Cs 2 CuCl 4 ( J^{\prime}/J=0.3 ), exhibits magnetic orders and TLL behavior with bound spinons, depending on the magnetic field strength and direction, as observed by NMR and inelastic neutron-scattering measurements [ 32 , 33 , 34 , 30 , 35 , 36 , 37 ] . A 5d transition-metal oxychloride Ca 3 ReO 5 Cl 2 (CROC) is served as another example of the spin-1/2 quasi-1D antiferromagnet with the anisotropic triangular lattice, as shown in Fig. 1 [ 38 , 39 , 40 ] . CROC forms an orthorhombic Pnma lattice, where the ReO 5 square-pyramid with the 5d_{xy} orbital forms a chain along the b axis. Despite the strong spin-orbit coupling of Re ions, the orbital moment is quenched under the asymmetric ligand field and perturbatively contributes to the ground state through the Dzyaloshinskii-Moriya (DM) interaction {\bf D}\cdot{\bf S}_{i}\times{\bf S}_{i+1} , where {\bf D} is parallel to the c axis in the ac mirror plane. The interchain coupling within the bc plane constructs an anisotropic triangular lattice, as shown in Fig. 1 . The anisotropy of the triangular lattice has been evaluated as J^{\prime} / J = 0.25 from the high-field magnetization [ 41 ] , consistent with that of the calculation of density functional theory (DFT) ( J^{\prime} / J = 0.295) [ 39 , 42 ] . At low temperatures, \chi behaves as a spin-1/2 1D antiferromagnetic Heisenberg model with J/k_{\rm B} = 41 K [ 39 , 43 ] , which implies a reduction of the dimensionality. The ground state eventually exhibits a long-range magnetic order below T_{\rm N} = 1.13 K. Thus, one can investigate the spin correlation of the TLL state featured by K_{\sigma} in an extensive temperature range of T_{\rm N}28 K. The Knight shift K obtained from the central resonance line scales to the local spin susceptibility \chi_{ii} along the crystal i axis; K_{ii}=\frac{A_{ii}}{N\mu_{\rm B}}\chi_{ii} ( i=a,b,c ), where A_{ii} is the diagonal hyperfine coupling, N the Avogadro number, and \mu_{\rm B} the Bohr magneton. Figure 4 shows the T dependence of K for Cl(1) and Cl(2), denoted as K(1) and K(2) , respectively. -K(1) increases upon cooling, showing a broad peak around 28 K. It becomes nearly constant at low temperatures. The temperature dependence of -K(1) for each crystal axis fits to the Bonner-Fisher model with J/k_{\rm B} = 38 K ( 70.5 , the transverse component is enhanced at low temperatures, whereas the parallel component vanishes and thus becomes negligible. Furthermore, as seen from Eq. (3), \chi_{\parallel} contributes to T_{1}^{-1} only through the off-diagonal hyperfine coupling. Therefore, T_{1}^{-1} would be dominated by \chi_{\perp}(\mathbf{q}) , as shown in Eq.(1). Below 10 K, T_{1}^{-1} increases with a power law \sim T^{n} , as shown in Fig. 5 . The fitting yields n=-0.36 for Cl(1) and -0.59 for Cl(2) in a magnetic field of 9.13 T along the a axis. Applying the relation n=1/(2K_{\sigma})-1 in Eq. (1), we obtained K_{\sigma} = 0.79 and 1.21 for Cl(1) and Cl(2), respectively. Since the dynamical spin susceptibility is carried by Re spins, K_{\sigma} should be independent of the probe nuclear spins. Therefore, the observed Cl site dependence comes from the form factor dependence that filters antiferromagnetic fluctuations at the specific wave vector, as typically observed in tetragonal cuprate and pnictide superconductors [ 11 , 58 ] . Referring to the crystal structure of CROC in Figs. 1 and 9 , Cl(1) is located between neighboring Re atoms, while Cl(2) is on the Re atom along the a axis. Here we consider antiferromagnetic spin fluctuations at the wave vector \mathbf{Q_{0}} = (0, \pi , 0) along the chain. The staggered hyperfine fields from the second-neighbor Re sites are canceled out at Cl(1), as shown by the site-symmetry analysis in Appendix B . Then, the antiferromagnetic correlation at {\bf q} = {\bf Q_{0}} is filtered through the form factor at Cl(1), while it remains at Cl(2). Therefore, T_{1}^{-1} measured at Cl(1) is strongly suppressed in contrast to that of Cl(2), despite the uniform hyperfine coupling of Cl(1) greater than that of Cl(2). T_{1}^{-1} is indeed reversed between Cl(1) and Cl(2) at high temperatures above 50 K, where the uniform {\bf q}=0 mode becomes dominant. In the TLL regime, the external magnetic field works as chemical potential and enhances magnetization or K_{\sigma} , as observed in the spin-1/2 ladder system with the spin gap [ 15 ] . We have investigated the field dependence of K_{\sigma} for the Cl sites at H_{0} = 4.95 T, as shown in Fig. 6 . There is no significant field effect on K_{\sigma} , consistent with the negligible effect of the Zeeman energy in the gapless TLL, as listed in Table 1 . Instead, the energy scale of the DM interaction is comparable to the magnetic field and thus impacts on the anisotropy. Figure 7: Temperature dependence of T_{1}^{-1} for H_{0} = 9.13 T along (a) b and (b) c axes. Solid lines show the power law T^{1/2K-1} fit with a parameter K_{\sigma} . Table 1: TLL parameter K_{\sigma} for Cl(1) and Cl(2) under the different magnetic field orientation and strength. The values in the parentheses are obtained after the RPA analysis [ 14 , 15 ] . Cl(1) Cl(2) H\parallel a , 9 T 0.79 (0.7) 1.21 (1.0) H\parallel b , 9 T 0.87 (0.7) 1.13 (1.0) H\parallel c , 9 T 0.61 (0.5) 0.70 (0.6) H\parallel a , 5 T 0.79 1.20 III.4 Anisotropy of T_{1}^{-1} Finally, we investigate the anisotropy of T_{1}^{-1} by applying the magnetic field H_{0} = 9.13 T along the b and c axes, as shown in Fig. 7 . The result along the b axis is similar to that for the a axis in Figs. 5 (a) and 6 , where T_{1}^{-1} is dominated by atomic fluctuations at high temperatures and then by electric spin fluctuations below 20 K. At low temperatures, we obtain K_{\sigma}=0.87 and 1.13 for Cl(1) and Cl(2), respectively, as listed in Table 1 . In contrast, the anisotropy of T_{1}^{-1} for the c axis is not reversed at low temperatures, suggesting that the {\bf q}={\bf Q_{0}} mode is less dominant in T_{1}^{-1} for Cl(1) and Cl(2). Interestingly, T_{1}^{-1} remains anisotropic in the intermediate temperature range where T_{1}^{-1} exhibits the broad peak. It may indicate significant spin-lattice coupling through the DM interaction. In the TLL regime below 20 K, we obtained K_{\sigma} , 0.61 for Cl(1) and 0.70 for Cl(2), which is much lower than those in the other directions. The reason can be attributed to the anisotropy of dynamical spin susceptibility, as discussed below. IV Discussion The enhancement of K_{\sigma} might be interpreted as the repulsive interaction of spinless fermions based on the XXZ model. In the present system showing isotropic spin susceptibility, the effective spin Hamiltonian would be close to the Heisenberg model with \Delta\sim 1 . Therefore, the large K_{\sigma} for Cl(2) will originate in the antiferromagnetic correlation toward T_{\rm N} , which is not taken into account within the simple TLL theory. In the critical regime, the transverse component of T_{1}^{-1} is expected to follow the power law against reduced temperature, \sim(T-T_{\rm N})^{-\nu(z_{t}-D-\eta_{t}+2)} , within the dynamical scaling hypothesis [ 59 , 14 ] , where \nu is the correlation length exponent, z_{t} the dynamical exponent, D the dimensionality, and \eta_{t} the anomalous exponent. The exponents depend on the universality: e.g. a mean field and a 2D XY model yield n=-0.5 and -0.67 , respectively. Approaching T_{\rm N} , the system exhibits a crossover from 1D to 3D regime, leading to the nominal enhancement of K_{\sigma} . The effect of critical spin fluctuations on T_{1}^{-1} can be calculated from the random phase approximation (RPA) below 10 K [ 14 , 15 ] . A fitting into our experimental result reduces K_{\sigma} , as listed in Table 1 . The value is still close to unity for Cl(2). The result implies that critical spin fluctuations can be reproduced beyond the mean-field approximation near T_{\rm N} . The anisotropy of T_{1}^{-1} can be explained in terms of spinon dynamics under the DM interaction {\bf D}\cdot{\bf S_{i}}\times{\bf S_{i+1}} . In CROC, the DM vector {\bf D} is directed along the c axis within the mirror plane. Spinon dispersion splits by \pi D/2 = 310 GHz at q=0 or \pi along the chain, as observed by the EPR measurement [ 41 ] . The excitation branches split with increasing the magnetic field strength along the c axis, whereas they monotonically increase with magnetic field along the a and b axes. The DM interaction induces the magnetic order with the moments directed within the ab plane. As a result, antiferromagnetic fluctuations are enhanced when the magnetic field is applied to the ab plane, while they are suppressed under the field along the c axis where the effect of DM interaction is minimized. Therefore, the critical spin fluctuations are largely suppressed for the c axis, which extracts the intrinsic K_{\sigma} of the TLL regime. The field dependence of K_{\sigma} differs from that of the spin-ladder system with the gapped ground state [ 15 , 19 , 18 ] . The magnetic field induces the paramagnetic or long-range order phase where the effect of the DM interaction may not be negligible. However, detailed angular dependence measurements of spin fluctuations are absent. Our determination of the TLL parameter through the T_{1}^{-1} anisotropy measurement will promote further theoretical studies on the anisotropic hydrodynamics of quasi-1D quantum liquids. V Conclusion We have investigated the magnetic ground state and the anisotropic low-energy excitation through the 35 Cl NMR measurement on the quasi-one-dimensional antiferromagnet Ca 3 ReO 5 Cl 2 with the DM interaction. We determined the nuclear quadrupole splitting and Knight shift tensors by the angular dependence of the 35 Cl NMR spectrum in agreement with the LDA calculation. The nuclear spin-lattice and spin-spin relaxation rates show slow atomic dynamics in the intermediate temperature range. We observed a power-law devolution of spin correlation, characteristic of one-dimensional quantum liquid. The site and angular dependence of the Luttinger parameter comes from the form-factor filtering of the antiferromagnetic correlation induced by the DM interaction. The incommensurate magnetic order occurs below 1 K with the low-lying magnon excitation. Appendix A Nuclear quadrupole splitting In this section, we show a complete set of the angular dependence of nuclear quadrupole splitting \delta\nu , which is compared with the DFT calculation. The nuclear spin Hamiltonian is expressed as a sum of the magnetic Zeeman interaction and the electric quadrupole interaction \mathcal{H}_{Q} : \mathcal{H}=\mathcal{H}_{Z}+\mathcal{H}_{Q}, (4) where the Zeeman term is given by \mathcal{H}_{Z}=\gamma\hbar\mathbf{I}\cdot(\mathbf{H_{0}}+\bf{H_{\rm hf}}) (5) The hyperfine field is given by \bf{H_{hf}}=\mathbf{K}\cdot\mathbf{H_{0}} with the Knight shift tensor \mathbf{K} . The nuclear quadrupole interaction is given by \mathcal{H}_{Q}=\sum_{\alpha,\beta}\frac{1}{9}\delta\nu_{\alpha\beta}[\frac{3}{2}(I_{\alpha}I_{\beta}+I_{\beta}I_{\alpha})-\delta_{\alpha\beta}I^{2}], (6) where the nuclear quadrupole splitting \delta\nu_{\alpha\beta} is defined by \delta\nu_{\alpha\beta}=\frac{3eQV_{\alpha\beta}}{2I(2I-1)h} (7) with the electric field gradient (EFG) V_{\alpha\beta} and the nuclear quadrupole moment Q . Under an intense magnetic field, the 35 Cl ( I=3/2 ) NMR spectrum splits into three by the interval frequency \delta\nu_{\alpha\beta} . The second-order electric quadrupole interaction gives a frequency shift of the central 1/2\leftrightarrow-1/2 transition in the order of \nu_{Q}^{2}/\gamma_{n}H_{0} . The satellite lines from the 3/2\leftrightarrow 1/2 and -1/2\leftrightarrow-3/2 transitions shift equally by the second-order effect. We obtain \delta\nu_{\alpha\beta} by subtracting the lowest satellite frequency from the highest one. Figure 8: (a-c)Angular dependence of the nuclear quadrupole splitting frequency \delta\nu for two Cl sites at 200 K. The field was rotated along the crystal axis. Solid lines show fitting to Eq. ( 8 ). (d-f) Calculated \delta\nu using LDA. (g-i) Angular dependence of the central frequency for two Cl sites at 200 K. The field was rotated about the crystal axis. Solid curves show fitting to the exact diagonalization calculation of the nuclear spin Hamiltonian including the electric quadrupole interaction and the Zeeman interaction, yielding the Knight shift tensor. (j-l) The second-order quadruple contribution to the central frequency, obtained from the EFG calculation using the LDA functional. We measured the angular dependence of \delta\nu_{\alpha\beta} about three crystal axes at 20 K and 9.13 T, as shown in Fig. 8. Without detailed analysis below, the profiles of the experimental result in Figs. 8(a,b,c) agree well with those of the LDA calculation in Figs. 8(d,e,f). Thus, we can successfully assign the 35 Cl NMR spectra to two Cl sites with different local environment. Since both Cl sites are located on the mirror plane normal to the b axis, the b axis must be one of the principal axes of the EFG tensor [ 60 , 61 ] . Thus, there is only one set of the spectrum from each Cl site, when we rotate the sample in the crystal plane including the b axis. Then the off-diagonal terms vanish in the \delta\nu_{\alpha\beta} tensor: \delta\nu_{ab}=\delta\nu_{ba} = 0, \delta\nu_{bc} = \delta\nu_{cb}=0 . The number of NMR lines doubles in the ca -plane rotation about the b axis without the spatial inversion symmetry. The rotation about the b axis is analyzed with the Volkoff formula [ 62 , 63 ] . On the ac -plane rotation with the angle \theta_{b} measured from the a axis, \delta\nu_{b} is fitted by \delta\nu_{b}=\delta\nu_{1,b}+\delta\nu_{2,b}\cos(2\theta_{b})+\delta\nu_{3,b}\sin(2\theta_{b}),\\ (8) where the fitting coefficients are given by \begin{split}\delta\nu_{1,b}=(\delta\nu_{aa}+\delta\nu_{cc})/2\\ \delta\nu_{2,b}=(\delta\nu_{aa}-\delta\nu_{cc})/2\\ \delta\nu_{3,b}=-\delta\nu_{ac}.\end{split} (9) Thus, we obtained \delta\nu_{\alpha\alpha} ( \alpha=a,b,c) for each crystal axis and the off-diagonal element \delta\nu_{ac} , as listed in Table 2 . The diagonalization of the tensor yields the principal components \delta\nu_{\alpha\alpha} ( \alpha = X, Y, Z) and the asymmetric factor \eta=|\delta\nu_{XX}-\delta\nu_{YY}|/\delta\nu_{ZZ} , where X , Y , and Z are defined by satisfying |\delta\nu_{XX}|<\delta\nu_{YY}<|\delta\nu_{ZZ}| . The results are compared with the LDA calculation based on the electronic structure of CROC in Figs. 8 (d-f) and Table. 3 . Then We can obtain the quadrupole frequency \nu_{Q}=\delta\nu_{ZZ} : \nu_{Q}=\frac{3eQV_{ZZ}}{2I(2I-1)h}, (10) where V_{ZZ} the principal component of the diagonalized electric field gradient tensor [ 51 ] . Table 2: Nuclear quadrupole splitting frequency \delta\nu_{\alpha\alpha} determined from the angular dependence of the 35 Cl NMR spectrum at 200 K. The tensor for the crystal coordinate ( a,b,c ) is diagonalized with the principal axes ( X,Y,Z ). The asymmetry factor \eta is defined in the text. The Y axis for Cl(1) (or Cl(2)) is directed at \pm 37^{\circ} ( \pm 30^{\circ} ) from the a axis in the ac plane. \delta\nu_{aa} \delta\nu_{bb} \delta\nu_{cc} \delta\nu_{ac} \delta\nu_{XX} \delta\nu_{YY} \delta\nu_{ZZ} \eta Cl(1) 0.29 -0.05 -0.24 \pm 0.98 0.05 0.99 -1.04 0.90 Cl(2) 1.09 -1.45 0.36 \pm 0.61 0.02 1.43 -1.45 0.98 Table 3: Nuclear quadrupole splitting \delta\nu_{\alpha\alpha} obtained from the LDA calculation for CROC. The Y (Z) axis for Cl(1) (or Cl(2)) is directed at \pm 40^{\circ} ( \pm 33^{\circ} ) from the a axis in the ac plane. \delta\nu_{aa}^{cal} \delta\nu_{bb}^{cal} \delta\nu_{cc}^{cal} \delta\nu_{ac}^{cal} \delta\nu_{XX}^{cal} \delta\nu_{YY}^{cal} \delta\nu_{ZZ}^{cal} \eta^{cal} Cl(1) 0.47 0.05 -0.52 \pm 1.27 0.05 1.34 -1.39 0.92 Cl(2) 1.29 -1.71 0.36 \pm 1.00 -0.22 -1.71 1.93 0.77 The angular dependence of the central resonance ( -1/2\leftrightarrow+1/2 ) frequency is shown in Figs. 8(g-i) and compared to the second-order nuclear quadrupole effect obtained from the LDA calculation in Figs. 8(k-l). The good agreement between the experimental result and the calculation allows the site assignment for the two Cl sites. Then we obtained the Knight shift tensor K after subtracting the second-order quadrupole contribution by the exact diagonalization of the nuclear spin Hamiltonian of Eq.( 5 ). We obtained the Knight shift tensor \mathbf{K} arising from the hyperfine interaction with Re electron spins at 200 K: \displaystyle\mathbf{K}=\begin{pmatrix}K_{aa}&0&K_{ac}\\ 0&K_{bb}&0\\ K_{ca}&0&K_{cc}\end{pmatrix} \displaystyle= \displaystyle\begin{pmatrix}-0.056&0&-0.002\\ 0&-0.041&0\\ -0.002&0&-0.039\end{pmatrix} , \displaystyle\begin{pmatrix}-0.032&0&-0.001\\ 0&-0.033&0\\ -0.001&0&-0.033\end{pmatrix} in % for Cl(1) and Cl(2), respectively. The diagonalization of the tensor yields (K_{XX},K_{YY},K_{ZZ})=(-0.039,-0.041,-0.056)\% for Cl(1) and (-0.032,-0.033,-0.033) for Cl(2), where Z is directed at \pm 7^{\circ} and \pm 71^{\circ} from the a axis in the ac plane, respectively. We also calculated the magnetic dipole field at Cl sites from Re ions. However, the result disagrees with the experimental result, which indicates that the Knight shift is governed by the transferred hyperfine coupling through the oxygen and calcium ions. Therefore, the hyperfine coupling does not necessarily scale to the atomic distances but depends on the paths through the overlap between the wavefunctions. The following discussion for the form factor is independent of the numerical value of the hyperfine coupling. Appendix B Form factors of dynamical spin susceptibility To explain the site dependence of T_{1}^{-1} , we evaluated the form factor of the dynamical spin susceptibility using the symmetry operation in the crystal structure of CROC [ 38 ] . As shown in Fig. 1, Cl(1) and Cl(2) sites are located between Re ions and on the Re ion, respectively. Here we focus on the original Cl sites with the fractional coordinate of the Pnma 4c position, (x_{1},y_{1},z_{1}) , y_{1}=1/4 , and the surrounding Re sites ( 4c ). Then we semi-quantitatively evaluate the hyperfine coupling based on the local symmetry. For Cl(1), there are one first neighbor Re site (3.716Å), Re {}^{(1)}_{1} , two second ones (4.094Å), Re {}^{(2)}_{1} and Re {}^{(2^{\prime})}_{1} , two third ones (4.892Å), Re {}^{(3)}_{1} and Re {}^{(3^{\prime})}_{1} , and one fourth one (5.868Å) Re {}^{(4)}_{1} . The fractional coordinates of Re are given by Re {}^{(1)}_{1} = (-x+1,-y+1,0) , Re {}^{(2)}_{1}=(x,y,z) , Re {}^{(2^{\prime})}_{1}=(x,y-1,z) , Re {}^{(3)}_{1}=(x+1/2,y,-z+1/2) , Re {}^{(3^{\prime})}_{1}=(x+1/2,y-1,-z+1/2) , and Re {}^{(4)}_{1}=(-x+3/2,-y+1,z+1/2) , where y=3/4 . The relative vectors of the Cl(1)-Re {}^{(i)}_{1} bonds are thus written as {\bf r_{1}^{(1)}}=(1-x-x_{1},0,-z-z_{1}) , {\bf r_{1}^{(2)}}=(x-x_{1},1/2,z-z_{1}) , {\bf r_{1}^{(2^{\prime})}}=(x-x_{1},-1/2,z-z_{1}) , {\bf r_{1}^{(3)}}=(x-1/2-x_{1},1/2,-z+1/2-z_{1}) , {\bf r_{1}^{(3^{\prime})}}=x-1/2-x_{1},-1/2,-z+1/2-z_{1}) , and {\bf r_{1}^{(4)}}=(-x-3/2-x_{1},0,z+1/2-z_{1}) . Here Re {}^{(2)}_{1} and Re {}^{(2^{\prime})}_{1} are connected via mirror reflection normal to the ac plane on Cl(1). Figure 9: Geometry of hyperfine interaction between Cl nuclear spins and Re sites in Ca 3 ReO 5 Cl 2 . Solid and dashed lines indicate nearest and next nearest Re-Cl bonds. Thus the hyperfine coupling tensor can be expressed as \displaystyle{\bf A^{(1)}_{1}} \displaystyle= \displaystyle\begin{pmatrix}A_{1aa}^{(1)}&0&A_{1ac}^{(1)}\\ 0&A_{1bb}^{(1)}&0\\ A_{1ca}^{(1)}&0&A_{1cc}^{(1)}\end{pmatrix} (11) \displaystyle{\bf A^{(2)}_{1}} \displaystyle= \displaystyle\begin{pmatrix}A_{1aa}^{(2)}&A_{1ab}^{(2)}&A_{1ac}^{(2)}\\ A_{1ba}^{(2)}&A_{1bb}^{(2)}&A_{1bc}^{(2)}\\ A_{1ca}^{(2)}&A_{1cb}^{(2)}&A_{1cc}^{(2)}\end{pmatrix} \displaystyle{\bf A^{(2^{\prime})}_{1}} \displaystyle= \displaystyle\begin{pmatrix}A_{1aa}^{(2)}&-A_{1ab}^{(2)}&A_{1ac}^{(2)}\\ -A_{1ba}^{(2)}&A_{1bb}^{(2)}&-A_{1bc}^{(2)}\\ A_{1ca}^{(2)}&-A_{1cb}^{(2)}&A_{1cc}^{(2)}\end{pmatrix}. (12) {\bf A^{(3)}} , {\bf A^{(3^{\prime})}} , and {\bf A^{(4)}} are written similar to {\bf A^{(2)}} , {\bf A^{(2^{\prime})}} , and {\bf A^{(1)}} , respectively. The net hyperfine coupling tensor is expressed as a sum of the hyperfine paths, {\bf A_{1}}=\sum_{i}{\bf A_{1}^{(i)}} . Taking Fourier transformation {\bf A_{1}(q)}=\sum_{i}{\bf A_{1}^{(i)}}({\bf r_{j}}){\rm e}^{i{\bf q\cdot r_{j}}}, (13) we obtain the uniform mode of the hyperfine coupling tensor for Cl(1) \displaystyle{\bf A_{1}(0)}\simeq\begin{pmatrix}A_{1aa}^{(1)}+2A_{1aa}^{(2)}&0&A_{1ac}^{(1)}+2A_{1ac}^{(2)}\\ 0&A_{1bb}^{(1)}+2A_{1bb}^{(2)}&0\\ A_{1ca}^{(1)}+2A_{1ca}^{(2)}&0&A_{1cc}^{(1)}+2A_{1cc}^{(2)}\end{pmatrix} where we omit the higher order terms for simplicity. The staggered mode along the b axis with the specific wave vector {\bf Q_{0}}=(0,\pi,0) is expected to dominate the low-energy spin correlation in the 1D chain. The hyperfine tensor at {\bf Q_{0}} is calculated as \displaystyle{\bf A_{1}(Q_{0})}\simeq\begin{pmatrix}A_{1aa}^{(1)}&2iA_{1ab}^{(2)}&A_{1ac}^{(1)}\\ 2iA_{1ba}^{(2)}&A_{1bb}^{(1)}&2iA_{1bc}^{(2)}\\ A_{1ca}^{(1)}&2iA_{1bc}^{(2)}&A_{1cc}^{(1)}\end{pmatrix}. The second neighbor interactions remain for {\bf A_{1}(0)} but vanished for {\bf A_{1}(Q_{0})} in the diagonal components. Namely, antiferromagnetic spin fluctuations are partly filtered through the hyperfine form factor at Cl(1). As for Cl(2) = (x_{2},y_{2},z_{2}) , y_{2}=3/4 , there are one nearest neighbor (3.555Å) and one second neighbor (4.350Å) Re site. We omit the higher order terms with the distance exceeding 5 Å. Similar to Cl(1), the Cl(2)-Re vectors are expressed as {\bf r_{2}^{(1)}}=(x-1/2-x_{2},0,-z+1/2-z_{2}) and {\bf r_{2}^{(2)}}=(x-x_{2},0,z-z_{2}) . Thus both the uniform and staggered mode of the hyperfine coupling tensor {\bf A(q)} ( {\bf q}={\bf 0},{\bf Q_{0}} ) is expressed as \displaystyle{\bf A_{2}(q)}\simeq\begin{pmatrix}A_{2aa}^{(1)}+A_{2aa}^{(2)}&0&A_{2ac}^{(1)}+A_{2ac}^{(2)}\\ 0&A_{2bb}^{(1)}+A_{2bb}^{(2)}&0\\ A_{2ca}^{(1)}+A_{2ca}^{(2)}&0&A_{2cc}^{(1)}+A_{2cc}^{(2)}\end{pmatrix}. The hyperfine form factor along the field direction, F_{\parallel} is calculated as F_{\parallel}=\sum_{\alpha}{A_{\alpha\beta}(q)A_{\alpha\beta}(-q)} . For r_{1}^{(1)}>r_{2}^{(1)} , A_{2\alpha\beta}^{(1)} is greater than A_{1\alpha\beta}^{(1)} . Since the staggered spin fluctuations at {\bf Q_{0}} are dominant at low temperatures, the reversed T_{1}^{-1} values of Cl(1) and Cl(2) can be attributed to the partial cancellation of the form factor at Cl(1) due to the antisymmetric spin fluctuations at the second neighbor Re spins.