Suppression of electronic susceptibility in metal–Mott-insulator alternating material $(\mathrm{Me}\text{−}3,5\text{−}\mathrm{DIP})\left[\mathrm{Ni}\right(\mathrm{dmit}{)}_{\mathbf{2}}{]}_{\mathbf{2}}$ observed using ${}^{13}\mathrm{C}$ NMR

Frequency shifts and nuclear relaxations of ${}^{13}\mathrm{C}$ NMR of the metal-insulator alternating layered material $(\mathrm{Me}\text{−}3,5\text{−}\mathrm{DIP})\left[\mathrm{Ni}\right(\mathrm{dmit}{)}_2{]}_2$, are presented. The NMR absorption lines originating from metallic and insulating layers are well resolved, which evidences the coexistence of localized spins $\left({\pi }_{\mathrm{loc}}\right)$ and conduction $\pi$ electrons. The NMR line originating from the insulating layers is found to be wiped out at about $2.5\mathrm{K}$, suggesting antiferromagnetic ordering (AFO) or strong fluctuations of ${\pi }_{\mathrm{loc}}$ toward AFO. In the metallic layer, we found significant suppressions of static and dynamical susceptibilities of conduction electrons below $35\mathrm{K}$, where antiferromagnetic correlation in the insulating layer evolves. We propose a dynamical effect through strong $\pi \text{−}{\pi }_{\mathrm{loc}}$ coupling between the metallic and insulating layers as an origin of the reduction of the density of states.

Figure 1

(Color online) (a) Crystal structure of $(\mathrm{Me}\text{−}3,5\text{−}\mathrm{DIP})\left[\mathrm{Ni}\right(\mathrm{dmit}{)}_2{]}_2$. The metallic and insulating $\mathrm{Ni}(\mathrm{dmit}{)}_2$ layers are hatched by different colors. Side projections of the metallic (b) and insulating (c) layers.

Figure 2

Ratio of interplane-to-inplane resistivities. The raw data of ${\rho }_b$ and ${\rho }_c$ are plotted in the inset.

Figure 3

(a) ${}^{13}\mathrm{C}$ NMR spectra of $(\mathrm{Me}\text{−}3,5\text{−}\mathrm{DIP})\left[\mathrm{Ni}\right(\mathrm{dmit}{)}_2{]}_2$. Distinct lines are labeled as $A$, $B$, and $C$. (b) Spectra obtained by reducing the time interval between saturating and $\pi ∕2$ pulses to 1/15 of $T_1$ for lines $A$ and $B$. At $2\mathrm{K}$, line $C$ is missing and only the partially recovered components of lines $A$ and $B$ are observed.

Figure 4

(Color online) Frequency shifts of lines $A$, $B$, and $C$. The peak positions $(○)$ below $20\mathrm{K}$ are defined by deconvolution of the spectra. Center of gravities of lines $A$ and $B$ $(K_A,◼)$ and line $C$ $(K_C,▴)$ are plotted. The dotted line is the bulk susceptibility measured by SQUID scaled by $A_{\parallel }\chi \left(T\right)∕N_A{\mu }_B$.

Figure 5

(Color online) Second moments of lines $A$, $B$, and $C$.

Figure 6

(a) $1∕T_{1}T$ $(●,○)$, $1∕T_1$ ($◻$, right axis) and (b) $1∕T_{2G}$ of lines $A$ and $B$. The dashed line shows the level of $⟨\mathrm{\Delta }{\omega }_{\mathrm{direct}}^2⟩$ estimated by Eq. (2).