[Paper Review] Metal-insulator transition on SrTiO$_{3}$ surface induced by ionic-bombardment

Heiko Groß, Namrata Bansal+2 more|arXiv (Cornell University)|Apr 28, 2011

Electronic and Structural Properties of Oxides30 references18 citations

TL;DR

This study demonstrates that argon ion-milling induces a metal-insulator transition on SrTiO3 (STO) surfaces by creating oxygen vacancies, leading to a highly conductive layer even at cryogenic temperatures where thermal diffusion is suppressed. The conductance arises from oxygen vacancy doping, but vacancy clustering reduces carrier mobility, explaining discrepancies between experimental conductance and simple models.

ABSTRACT

SrTiO$_{3}$ is one of the most popular insulating single-crystal substrates for various complex-oxide thin film growths, because of its good lattice match with many complex oxide films. Here, we show that a common thin film processing technique, argon ion-milling, creates highly conducting layer on the surface of STO, not only at room temperatures but also at cryogenic temperatures at which thermal diffusion is completely suppressed. Systematic \emph{in situ} four-point conductance measurements were taken on single-crystal STO substrates inside vacuum environment. The evolution of metallicity out of insulating STO follows simple models based on oxygen vacancy doping effect. At cryogenic temperatures, ion milling created a thin - but much thicker than the argon-penetration depth - steady-state oxygen-vacant layer, leading to a highly-concentric metallic state. Near room temperatures, however, significant thermal diffusion occurred and the metallic state continuously diffused into the bulk, leaving only low concentraion of electron carriers on the surface. Analysis of the discrepancy between the experiments and the models also provided evidence for vacany clustering, which seems to occur during any vacancy formation process and affects the observed conductance. These observations suggest that the transport properties of films processed on STO substrates using energetic methods such as ion milling need to be taken with caution. On the other hand, if properly controlled, ionic bombardment could be used as a way to create selective conducting layers on the surface of STO for device applications.

Motivation & Objective

• Investigate the origin of metallic surface states on SrTiO3 induced by argon ion-milling.
• Determine whether thermal diffusion or vacancy clustering dominates conductance evolution during ion bombardment.
• Clarify the role of oxygen vacancies and their clustering in altering electronic transport in STO.
• Assess the reliability of transport measurements on STO substrates processed with energetic techniques.
• Explore the potential of ionic bombardment as a controlled method for creating conductive oxide heterostructures.

Proposed method

• Performed in situ four-point conductance measurements on single-crystal STO substrates under ultra-high vacuum (base pressure <10−7 Torr).
• Used a gridless ion source (50–500 eV) to deliver argon ion beams at temperatures from −160 °C to 700 °C.
• Monitored sample temperature with a thermocouple and oxygen partial pressure with a residual gas analyzer (RGA).
• Measured conductance using a Keithley 2636A source meter after each ion-milling step.
• Modeled conductance evolution assuming each oxygen vacancy donates two mobile electrons, comparing to experimental data.
• Conducted long-term oxygen exposure experiments to distinguish vacancy filling from clustering effects on conductance decay.

Experimental results

Research questions

• RQ1 Can ion-milling induce a metallic surface layer on insulating SrTiO3 at cryogenic temperatures where thermal diffusion is negligible?
• RQ2 How does thermal diffusion influence the depth and evolution of the metallic layer formed during ion-milling near room temperature?
• RQ3 To what extent do oxygen vacancy clusters reduce carrier mobility and explain the discrepancy between measured conductance and simple doping models?
• RQ4 What mechanisms underlie the observed conductance decay over time, and can they be separated into vacancy filling and clustering?
• RQ5 Can ionic bombardment be used as a controlled method to engineer conductive layers on STO for oxide electronics?

Key findings

• At cryogenic temperatures (100–200 K), a steady-state, highly conductive layer formed on STO after ion-milling, with conductance saturating to a metallic value despite negligible thermal diffusion.
• Near room temperature, conductance continuously increased without saturation due to thermal diffusion of oxygen vacancies into the bulk, leading to deeper metallic states.
• The observed conductance drop by a factor of four after long-term exposure to oxygen cannot be explained by vacancy filling alone, indicating significant vacancy clustering.
• Vacancy clustering, particularly with more than two vacancies per cluster, reduces the number of mobile carriers and explains the discrepancy between measured conductance and simple two-electron-per-vacancy models.
• Even at oxygen partial pressures below 10−9 Torr, a slow, background conductance decay persisted, indicating that vacancy clustering occurs spontaneously and actively at room temperature.
• The results imply that transport measurements on STO substrates processed with energetic techniques like ion-milling or pulsed laser deposition must account for unintended metallic surface layers and vacancy clustering effects.