At temperatures $\textit{T}$ < 180 K, layered compound 1T-TaS$_2$ is a Mott insulator. Magnetic ground state of this phase was recently proposed to be a quantum spin liquid. Properties of this state are yet unclear and further experiments are required. One possible approach is a destabilization of quantum spin liquid with doping. With investigation of low-temperature excitations we want to observe how a quantum spin liquid progresses into the metalic or the superconducting phase across the 1T-TaS$_{2-x}$Se$_x$ phase diagram.
In this master's thesis, we report a $^{77}$Se nuclear magnetic resonance (NMR) and $^{181}$Ta nuclear quadrupole resonance (NQR) study of magnetic properties of selenium-doped 1T-TaS$_2$. Shapes and positions of spectral lines were measured to follow changes in static local magnetic fields, while temperature dependence of spin-lattice relaxation rate was explored to determine low-energy excitations.
$^{181}$Ta NQR spectra for the lightly doped 1T-TaS$_{1.94}$Se$_{0.06}$ sample exhibit significant line broadening in comparison to pristine samples, indicating Se doping-induced increase of local disorder. Temperature dependence of spin-lattice relaxation rate follows a power law. $1/T_1 \propto T^{2.7}$ over the broad temperature range 10 K < $\textit{T}$ < 160 K and indicates a linear gapless spinon dispersion relation at some point of the Brillouin zone. For the lowest temperatures $\textit{T}$ < 10 K, temperature dependence of spin-lattice relaxation rate changes to $1/T_1 \propto T^\eta$, $\eta \approx 0.6$, which is consistent with a quadratic spinon dispersion band - a hallmark for spinons forming a Fermi surface. Comparison with the data on pristine samples suggests that 1T-TaS$_2$ has similar dispersion relation, but with linear part opening an energy gap on the order of 1 meV for $\textit{T}$ < 50 K. The origin of such gap is still unclear, but it may reflect freezing of charge or structural degrees of freedom.
In the investigated temperature range, the heavily doped 1T-TaSSe is a conductor rather than Mott insulator. The temperature dependence of spin-lattice relaxation rate is roughly linear in temperature, as it is expected for metallic samples. However, the Moriya enhancement factor, which measures the effect of spin correlations in metals, is significantly smaller than 1, implying that antiferromagnetic correlations remain very important even deeply in the metallic regime.
In my master's thesis, I demonstrated how spinon dispersion relation changes with doping in model system 1T-TaS$_2$.
|
