Strain Control of Fermiology and Many-Body Interactions in Two-Dimensional Ruthenates

Here we demonstrate how the Fermi surface topology and quantum many-body interactions can be manipulated via epitaxial strain in the spin-triplet superconductor ${\mathrm{Sr}}_2{\mathrm{RuO}}_4$ and its isoelectronic counterpart ${\mathrm{Ba}}_2{\mathrm{RuO}}_4$ using oxide molecular beam epitaxy, in situ angle-resolved photoemission spectroscopy, and transport measurements. Near the topological transition of the $\gamma$ Fermi surface sheet, we observe clear signatures of critical fluctuations, while the quasiparticle mass enhancement is found to increase rapidly and monotonically with increasing Ru-O bond distance. Our work demonstrates the possibilities for using epitaxial strain as a disorder-free means of manipulating emergent properties, many-body interactions, and potentially the superconductivity in correlated materials.

Figure 1

(a)–(d) Fermi surface maps and (e)–(h) spectral weight along the $(0,k_y)$ direction [thick red line in (a)] for select strain states of ${\mathrm{Sr}}_2{\mathrm{RuO}}_4$ and ${\mathrm{Ba}}_2{\mathrm{RuO}}_4$. The data in (e) were measured at an elevated temperature ($T=100\mathrm{K}$) to thermally populate the states above the Fermi level; the rest of the data in the Letter were taken at 15 K. To show the dispersion near the vHs above $E_F$, the spectral weight was divided by the Fermi function in (e) and (f). The substrate number line shows the room temperature lattice constants and strain values relative to bulk ${\mathrm{Sr}}_2{\mathrm{RuO}}_4$.

Figure 2

(a) Schematic showing the evolution of the $\gamma$ Fermi surface and density of states at $E_F$ as a result of strain and negative chemical pressure by $A$-site substitution. (b) Tight-binding parametrization of ARPES Fermi surfaces and LDA Fermi surfaces. (c) Luttinger volume of experimental Fermi pockets as a function of the in-plane lattice parameter. The total number of electrons adds up to $n=4.00\pm 0.05$ showing negligible overall doping. (d) Lindhard susceptibility calculated for the two-dimensional $\gamma$ and one-dimensional $\alpha /\beta$ pockets from the experimental Fermi surfaces.

Figure 3

Normalized resistivity fitted to $\rho \propto T^n$ [values of $n$ shown in (c) with the open symbol from Ref. 36]. The inset shows $\log (d\rho /dT)\approx (n-1)\log (T)$ with an offset. (b) Dispersion of the $\gamma$ band along $(0,0)-(0,\pi )$, and (d) the deviation of $E_{\mathrm{VHS}}$ from the TB model. (e) $\mathrm{Re}\mathrm{\Sigma }(\omega )$ (offset at $E=E_F$ is included for clarity). $\mathrm{Re}\mathrm{\Sigma }(\omega )\propto \omega$ implies quadratic energy dependence of the quasiparticle scattering rate $\mathrm{\Gamma }(\omega )\propto \mathrm{Im}\mathrm{\Sigma }(\omega )\propto {\omega }^2$ expected for a Fermi liquid. ${\mathrm{Ba}}_2{\mathrm{RuO}}_4/\mathrm{STO}$ acquires additional kinklike features in the real part at the energy scale of $15\pm 10\mathrm{meV}$ near $\mathbf{k}=(0,\pi )$ (red line cut). This flattens the $\gamma$ band and pushes the vHs slightly below the Fermi level.

Figure 4

Dispersion $E(k)$ of the $\alpha$ band along the BZ boundary (a)–(c) and $\beta$ band along (0,0)–(0, $\pi$) (d)–(f). Spectral weight is shown for the single crystal ${\mathrm{Sr}}_2{\mathrm{RuO}}_4$ (a),(d) and ${\mathrm{Ba}}_2{\mathrm{RuO}}_4/\mathrm{GSO}$ (b),(e). The colors of the open symbols in (c) and (f) are consistent with the colors in (g)–(i). Quasiparticle renormalizations for the $\alpha$ band (g) and $\beta$ band (h) have strong monotonic dependence on the strain value. QP renormalization for the $\gamma$ band (i), calculated as a ratio of the LDA bandwidth to ARPES bandwidth from the corresponding tight-binding fits, shows a similar monotonic increase as a function of tensile strain. The open circles in (i) show the band renormalization from the slope of the real part of self-energy $1-\partial \mathrm{Re}\mathrm{\Sigma }(\omega )/\partial \omega$ near $E_F$ calculated at $\mathbf{k}=(\pi ,0)$. The deviation for ${\mathrm{Ba}}_2{\mathrm{RuO}}_4/\mathrm{STO}$ is due to the band flattening shown in Fig. 3.