[Paper Review] Investigation of the In-Gap Electronic Structure of LaAlO3 - SrTiO3 Heterointerfaces by Soft X-ray Spectroscopy

A. Koitzsch, J. Ocker+5 more|arXiv (Cornell University)|Mar 29, 2011

Electronic and Structural Properties of Oxides3 citations

TL;DR

This study uses soft X-ray spectroscopy to investigate the in-gap electronic states at LaAlO3-SrTiO3 heterointerfaces grown under low- and high-oxygen pressure conditions. It identifies two distinct Ti³⁺-related charge carriers: one at the Fermi level linked to conduction band filling (present only under low oxygen pressure), and another at ~1 eV binding energy independent of oxygen pressure, likely due to defects. The results show that oxygen pressure critically controls defect density and carrier concentration, with high oxygen pressure suppressing metallic conductivity despite potential charge transfer from the polar catastrophe.

ABSTRACT

We investigated \LAO\ - \STO\ heterointerfaces grown either in oxygen rich or poor atmosphere by soft x-ray spectroscopy. Resonant photoemission across the Ti L$_{2,3}$ absorption edge of the valence band and Ti 2p core level spectroscopy directly monitor the impact of oxygen treatment upon the electronic structure. Two types of Ti$^{3+}$ related charge carriers are identified. One is located at the Fermi energy and related to the filling of the \STO\ conduction band. It appears for low oxygen pressure only. The other one is centered at E$_{B}$ $\approx$ 1 eV and independent of the oxygen pressure during growth. It is probably due to defects. The magnitude of both excitations is comparable. It is shown that low oxygen pressure is detrimental for the Ti - O bonding. Our results shed light on the nature of the charge carriers in the vicinity of the \LAO\ - \STO\ interface.

Motivation & Objective

• To understand how oxygen pressure during growth affects the in-gap electronic structure of LaAlO3-SrTiO3 heterointerfaces.
• To distinguish between charge carriers from the polar catastrophe and those from oxygen or other defects.
• To determine the role of oxygen vacancies and defect states in governing electrical conductivity at the interface.
• To clarify the origin of in-gap states observed in previous photoemission studies, particularly in relation to growth conditions.

Proposed method

• Resonant photoemission spectroscopy across the Ti L2,3 edge to enhance sensitivity to Ti 3d states.
• Ti 2p core-level spectroscopy to probe oxidation states and defect-related features.
• Comparison of samples grown under low oxygen pressure (4.5 × 10⁻⁶ mbar) and high oxygen pressure (5 × 10⁻³ mbar).
• Analysis of valence band spectra and core-level line shapes to extract electronic structure and defect contributions.
• Use of angular-dependent photoemission to assess intermixing and interface sharpness.
• Fitting of Ti 2p spectra to extract relative intensities of Ti³⁺ and Ti⁴⁺ components, estimating defect density.

Experimental results

Research questions

• RQ1 How does oxygen pressure during growth affect the population of the SrTiO3 conduction band at the interface?
• RQ2 What is the origin of the in-gap state observed at ~1 eV binding energy, and is it related to defects or electronic reconstruction?
• RQ3 Why is the Fermi edge signal absent in high-oxygen-grown samples despite theoretical expectations from the polar catastrophe?
• RQ4 To what extent do oxygen vacancies or other defects compete with or dominate over the polar catastrophe mechanism in determining carrier density?
• RQ5 Can defect-related states localize charge carriers and suppress metallic conductivity even when charge transfer is expected?

Key findings

• The conduction band of SrTiO3 is populated at the Fermi level only in samples grown under low oxygen pressure, indicating metallic character is oxygen-pressure-dependent.
• An in-gap state centered at ~1 eV binding energy is observed in both low- and high-oxygen-grown samples, suggesting a defect-related origin independent of oxygen pressure.
• The intensity of the Fermi-level state (G1) and the 1 eV state (G2) are comparable, indicating similar population levels of conduction electrons and defects.
• The Ti 2p spectrum of the high-oxygen sample (OR) shows a Ti³⁺ component with a minimal sheet defect density of ~2 × 10¹⁴ cm⁻² near the interface.
• The defect density in OR is sufficient to localize charge carriers, potentially explaining the absence of metallic conductivity despite theoretical charge transfer from the polar catastrophe.
• The results suggest that oxygen pressure controls defect density, which can dominate over the polar catastrophe in determining the electronic properties of the interface.