Evolution of the Fermi Surface and Quasiparticle Renormalization through a van Hove Singularity in ${\mathrm{Sr}}_{2-y}{\mathrm{La}}_y{\mathrm{RuO}}_4$

We employ a combination of chemical substitution and angle resolved photoemission spectroscopy to prove that the Fermi level in the $\gamma$ band of ${\mathrm{Sr}}_{2-y}{\mathrm{La}}_y{\mathrm{RuO}}_4$ can be made to traverse a van Hove singularity. Remarkably, the large mass renormalization has little dependence on either $\mathbf{k}$ or doping. By combining the results from photoemission with thermodynamic measurements on the same batches of crystals, we deduce a parametrization of the full many-body quasiparticle dispersion in ${\mathrm{Sr}}_2{\mathrm{RuO}}_4$ which extends from the Fermi level to approximately 20 meV above it.

Figure 1

(a)–(c) Raw ARPES spectral intensity $k$-space maps for ${\mathrm{Sr}}_{2-y}{\mathrm{La}}_y{\mathrm{RuO}}_4$, using an integration window of $E_F\pm 5\mathrm{meV}$, and an energy resolution of 16 meV, each generated from approximately ${10}^4$ energy distribution curves (EDCs). Data have not been symmetrized. (d)–(f) Precise Fermi surface crossings in the 2D projected Brillouin zone determined by fitting the intensity maxima of cuts running roughly perpendicular to the Fermi surfaces, set against the predictions of the tight-binding rigid-band shift model described in the text (black solid lines). Fermi surface sheets are labeled in (d).

Figure 2

Scans along the $\Gamma \mathrm{\text{−}}M\mathrm{\text{−}}\Gamma$ line in the Brillouin zone for samples with $y=0$, $y=0.18$ and $y=0.27$ at 10 K and $\Delta E<8\mathrm{meV}$ energy resolution, (a), (c), (d) and for $y=0$ at 100 K divided by the Fermi-Dirac distribution (b). In all cases the true experimentally measured $E_F$ (solid white line) was referenced to that of an Au target in direct electrical contact with the sample. The dotted white lines in (c) and (d) show the Fermi level for $y=0$, to highlight the magnitude of the doping-induced shift.

Figure 3

dHvA (closed circles) and ARPES (open squares) data for the carrier number of each of the three Fermi surface sheets as a function of $y$. The data from the two techniques is in excellent agreement where there is overlap (at $y=0$ and, for the $\alpha$ pocket, at $y=0.1$). (b) The dHvA effective masses (closed circles) and electronic specific heat coefficient ${\gamma }_N$ (defined as $C_p/T$ at 500 mK, with $C_p$ denoting the specific heat at constant pressure; open triangles) plotted as a function of $y$. In a quasi-two-dimensional material, the sum of the dHvA masses should be proportional to the electronic specific heat, and as seen from the graph, the agreement is excellent both at $y=0$ and $y=0.06$. In all three panels the solid lines are calculated using the model discussed in the text.