Table of Contents
Fetching ...

Ambipolar doping-induced surface in-gap state on Mott-insulating Ca$_2$RuO$_4$

M. Horio, T. Wada, V. Granata, R. Fittipaldi, A. Vecchione, J. Chang, I. Matsuda

Abstract

We report an x-ray photoemission spectroscopy study of Ca$_2$RuO$_4$ surface-dosed with Cs alkali atoms and C$_{60}$ molecules. Due to its small ionization energy (large electron affinity), deposited Cs atoms (C$_{60}$ molecules) are expected to provide a solid surface with electrons (holes). Upon dosing the dopants to Mott-insulating Ca$_2$RuO$_4$, we found a new Ru $3d$ photoemission peak emerging on the lower binding-energy side, suggesting the creation of a core-hole screening channel associated with coherent Ru $4d$ states around the Fermi level. For both the Cs and C$_{60}$ dosing, this change occurred without an appreciable chemical potential jump. The coherent state, therefore, develops within the Mott gap through hybridization with the impurity level of the dopants. The present work highlights the flexibility of Mott-insulator surfaces as a playground for metal-insulator transitions.

Ambipolar doping-induced surface in-gap state on Mott-insulating Ca$_2$RuO$_4$

Abstract

We report an x-ray photoemission spectroscopy study of CaRuO surface-dosed with Cs alkali atoms and C molecules. Due to its small ionization energy (large electron affinity), deposited Cs atoms (C molecules) are expected to provide a solid surface with electrons (holes). Upon dosing the dopants to Mott-insulating CaRuO, we found a new Ru photoemission peak emerging on the lower binding-energy side, suggesting the creation of a core-hole screening channel associated with coherent Ru states around the Fermi level. For both the Cs and C dosing, this change occurred without an appreciable chemical potential jump. The coherent state, therefore, develops within the Mott gap through hybridization with the impurity level of the dopants. The present work highlights the flexibility of Mott-insulator surfaces as a playground for metal-insulator transitions.
Paper Structure (6 sections, 2 equations, 4 figures)

This paper contains 6 sections, 2 equations, 4 figures.

Figures (4)

  • Figure 1: Chemical-potential shift by Cs deposition on Ca$_2$RuO$_4$. (a) Ca $2p$ spectra recorded at $h\nu = 500$ eV plotted in the order of Cs deposition steps from top to bottom. The vertical bars mark the half-maximum position on the lower energy side. After five times of Cs deposition, a new sample was cleaved and deposited with Cs to reach higher dosing (steps 6-8). (b),(c) Ca $3p$ and O $2s$ spectra of Cs-dosed Ca$_2$RuO$_4$, respectively, measured at $h\nu = 500$ eV. All the spectra in (a)--(c) have been normalized to the peak height. (d) The shifts of Ca and O core-level peaks plotted against Cs deposition steps. The binominal smoothing filter Smoothing was applied to the spectra before evaluating the half-maximum positions, and the difference from the estimate without smoothing was defined as the error bar.
  • Figure 2: Evolution of the Ru $3d$ peak upon Cs dosing. (a) Ru $3d$ spectra of Ca$_2$RuO$_4$ recorded at $h\nu = 500$ eV and plotted in the order of Cs deposition steps from top to bottom. After five times of Cs deposition, a new sample was cleaved and deposited with Cs to reach higher dosing (steps 6-8). All the spectra have been normalized to total spectral intensity in the displayed region. (b) Ru $3d$ spectra of the pristine (Cs-0) and Cs-dosed (Cs-3 and Cs-8) Ca$_2$RuO$_4$ overlaid with fitting curves. See text for details of the fitting procedure. (c) The proportion of peak B to the overall Ru 3$d$ peak plotted versus Ca $2p$ peak shift. The vertical error bar has been derived from the deviation of the experimental spectrum from the fitting curve. The gray shade is a guide to the eyes. (d) Valence-band spectra of the pristine and Cs-dosed Ca$_2$RuO$_4$ measured at $h\nu = 250$ eV and normalized to total intensity. The amount of deposited Cs is nominally the same as that for Cs-6 in panel (a).
  • Figure 3: Effect of C$_{60}$ dosing on the core-level spectra of Ca$_2$RuO$_4$. (a) Ca $2p$ spectra recorded at $h\nu = 500$ eV plotted in the order of C$_{60}$ deposition steps from top to bottom. The vertical bars mark the half-maximum position on the lower energy side. (b) Comparison of the Ru $3d$ spectra of the pristine and C$_{60}$-dosed Ca$_2$RuO$_4$ measured at $h\nu = 350$ eV. The amount of dosed C$_{60}$ is nominally the same as that for C$_{60}$-1 in (a). The green arrow marks the additional lower-energy component emerging after C$_{60}$ deposition. The steep increase of intensity above 284 eV after C$_{60}$ dosing is due to the C 1$s$ peak of C$_{60}$.
  • Figure 4: Smooth evolution of the coherent in-gap state. (a) Ru $3d$ spectra of the pristine, Cs-dosed, and C$_{60}$-dosed Ca$_2$RuO$_4$. The spectra of the pristine and Cs-dosed ones have been taken from Fig. \ref{['Fig2']}(a), and the C$_{60}$-dosed one from Fig. \ref{['Fig3']}(b). (b) Schematic energy diagram for the pristine Ca$_2$RuO$_4$. UHB and LHB denote the upper and lower Hubbard bands, respectively.(c) The shift of the Ca $2p$ peak, which represents the chemical potential shift, as a function of Cs or C$_{60}$ dosing. No abrupt jump is observed in either direction. (d),(e) Schematic energy diagrams of in-gap state formation after Cs and C$_{60}$ dosing, respectively.