Monday 27 July 2015 - Daniela Stornaiuolo - Tunable superconductivity and spin polarization in an oxide two-dimensional system

Daniela's talk challenges us with a thought-provoking idea: two-dimensional electron systems at oxide

interfaces can exhibit spin polarization as well as superconductivity. Are these two phenomena coexisting in the same region of the sample or is phase separation occurring?

Daniela opens with a nice introduction on LaAlO3/SrTiO3 (LAO/STO) interfaces and points out a number of peculiar electronic  properties: 1) Orbital reconstruction: the low energy bands have a 3dxy character at variance with bulk STO. 2) Depending on the band filling, controlled by the electrostatic field effect, a superconductor-insulator transition can be observed as well as an enhancement of spin-orbit coupling.

Daniela discusses nanostructures fabricated using electron beam lithography. Using local gates, superconductivity as well as spin-orbit coupling can be tuned at the nanoscale. In particular she makes the point that the enhancement of spin-orbit coupling is accompanied by the filling of the 3dzy and 3dxz bands.

Magnetic phenomena are also observed at these interfaces. Spectroscopic evidence points at a Ti 3d character of the magnetic moments. Importantly, Daniela remarks that various experiments indicate that in conventional LAO/STO interfaces, magnetism is not affected by the electrostatic field effect. She points out that controlling the interplay of superconductivity, magnetism and spin-orbit coupling by field effect would be appealing for exotic application such as quantum computation.
Motivated by these considerations she considers a new design for oxide interfaces in which a bilayer of the ferromagnetic material EuTiO3 (ETO) is embedded between a TiO2-terminated STO crystal and a ultrathin film of LAO. ETO orders ferromagnetically below 8K.

She discusses a number of interesting features of this new conducting oxide interface:

1) They exhibit a critical thickness for conductivity (4 u.c. of LAO).

2) If STO is SrO terminated, no conducting state is observed.

3) X-ray absorption spectroscopy shows that 3dxy states are lowest in energy.

4) X-ray magnetic circular dichroism shows ferromagnetic ordering of both Eu and Ti occurs below 8K (same temperature as observed in the bulk).

5) The carrier density at the interface can be gated between 1.5 to 2.7 10^13 cm-2.

6) A change in slope of the Hall effect is observed below 8K in the accumulation regime. She argues that this is a signature of the anomalous Hall effect and interprets this as a spin polarization of charge carriers occurring below 8 K. Hysteresis in Hall effect and magneto resistance are observed.

7) Superconductivity is observed below 120 mK. It is gate tunable.

8) The magnetoresistance is analysed on the basis of antilocalisation. The corresponding spin-orbit fields are of the same order as observed for conventional LAO/STO.

She argues that above a critical density ~ 1.9 10^13 cm-2 ferromagnetism with spin polarised carriers, spin-orbit and superconductivity coexist.

She discusses that this is accompanied by multiband occupancy also as evidenced by non-linearities in the Hall effect observed at high fields. The occupied bands have dispersion along the out-of-plane direction. She argues that this may mediating magnetic coupling.

She leaves us with a hard nut to crack: are spin-polarization and superconductivity coexisting in the same region of the sample? This imply that superconductivity has an unconventional character. Or is phase separation occurring?

Blogged by Andrea Caviglia.