Candidate for a fully frustrated square lattice in a verdazyl-based salt

We present an experimental realization of an $S=1/2$ fully frustrated square lattice (FFSL) composed of a verdazyl-based salt $\left(p\text{-MePy-V}\right)\left(\mathrm{TCNQ}\right)·{\left({\mathrm{CH}}_3\right)}_2\mathrm{CO}$. Ab initio molecular orbital calculations indicate that there are four types of competing ferro- and antiferromagnetic nearest-neighbor interactions present in the system, which combine to form an $S=1/2$ FFSL. Below room temperature, the magnetic susceptibility of the material can be considered to arise from the $S=1/2$ FFSL formed by the $p$-MePy-V and indicates that the system forms a quantum valence-bond solid state whose excitation energy is gapped. Furthermore, we also observe semiconducting behavior arising from the one-dimensional chain structure of the TCNQ molecules.

Figure 1

(a) Molecular structure of ($p$-MePy-V)(TCNQ). (b) Single crystals of $\left(p\text{-MePy-V}\right)\left(\mathrm{TCNQ}\right)·{\left({\mathrm{CH}}_3\right)}_2\mathrm{CO}$. (c) 1D chain structure composed of TCNQ radical anions. (d) 2D structure composed of $p$-MePy-V radical cations. (e) Crystal structure viewed parallel to the $a$ axis. Hydrogen atoms are omitted for clarity. The dashed lines indicate N-C and C-C short contacts.

Figure 2

$S=1/2$ FFSL in the $ab$ plane of $\left(p\text{-MePy-V}\right)\left(\mathrm{TCNQ}\right)·{\left({\mathrm{CH}}_3\right)}_2\mathrm{CO}$ crystal formed by AF $J_1, J_2, J_3$, and ferromagnetic $J_4$.

Figure 3

Temperature dependence of magnetic susceptibility ($\chi =M/H$) of $\left(p\text{-MePy-V}\right)\left(\mathrm{TCNQ}\right)·{\left({\mathrm{CH}}_3\right)}_2\mathrm{CO}$ at 0.5 T. The open circles denote raw data, and the solid ones are corrected for the paramagnetic term due to the impurity. The solid line represents the calculated results for the $S=1/2$ AF dimer coupled through $J_1$. The inset shows calculated $\chi$ in terms of the $S=1/2$ 1D AF chain with $J_0=1796$ K.

Figure 4

Temperature dependence of (a) specific heat $C_{\mathrm{p}}$ and (b) resistivity $\rho$ along the $a$ axis of $\left(p\text{-MePy-V}\right)\left(\mathrm{TCNQ}\right)·{\left({\mathrm{CH}}_3\right)}_2\mathrm{CO}$.