Home > Research activities > Oxitronics


Our research is about oxide materials, assembled into thin film heterostructuctures that we study to understand the intimate relationship between structural and physical properties, craft to achieve new, artificial properties, or use to define prototype devices. For more complete information you can visit our group website.

The vast majority of materials that we investigate are perovskites that show a very broad range of properties in the same structural family. This makes it possible to assemble them as lego bricks into multifunctional architectures. Practically, these heterostructures are grown by pulsed laser deposition, a technique by which a high energy laser pulse impinges onto a ceramic pellet of the material to grow (the “target”), creating a plasma plume containing the atomic/ionic species present in the target that then deposit onto a nearby single-crystalline substrate, usually heated at high temperature (typically 500-900°C). When all growth conditions are properly optimized, the material grows coherently onto the substrate, atomic layer by atomic layer (as in molecular beam epitaxy).
(image credit J. Huijben Univ. Twente)

Currently, our research concentrates on Multiferroics (materials that show several orders, such as magnetism and ferroelectricity), Nanoferronics (an emerging area of nanoscience focusing on the interplay between charge and spin currents and ferroic orders) and Oxide interfaces (where novel unexpected electronic phenomena can appear).

Multiferroics and magnetoelectrics
Multiferroics are a relatively rare class of multifunctional materials that simultaneously exhibit several (anti)ferroic orders and can be for instance ferroelectric and antiferromagnetic. We started to work on multiferroics in 2002, first focusing on BiMnO3 thin films for spin-filtering and moving toward BiFeO3 (BFO) in 2004. Among multiferroics, BFO is special: it is both ferroelectric and antiferromagnetic with high ordering temperatures (1100K for the ferroelectric Curie temperature and 640K for the magnetic Néel temperature). In addition, it has a very large ferroelectric polarisation of 100 µC/cm², the highest of all known ferroelectrics. Among our recent research results on BFO thin films, we higlight our discovery of novel structural and magnetic phases through strain engineering. The figure below shows a phase diagram of BFO films in a broad range of strain.

Related publications :

- Evidence for room-temperature multiferroicity in a compound with a giant axial ratio
H. Béa et al, Phys. Rev. Lett. 102, 217603 (2009)

- Bridging multiferroic phase transitions by epitaxial strain in BiFeO3
I. C. Infante et al, Phys. Rev. Lett. 105, 057601 (2010)

- Crafting the magnonic and spintronic response of BiFeO3 films by epitaxial strain
D. Sando et al, Nature Mater. 12, 641 (2013)

- BiFeO3 epitaxial thin films and devices: past, present and future
D. Sando et al, J. Phys. Condens. Matter. 26, 473201 (2014)

The renaissance of multiferroics was initially motivated by the appeal of an electric-field control of magnetism. However, room-temperature multiferroics are rare and magnetoelectric coupling is limited. Alternatively, one can design magnetoelectric hybrids in which a ferroelectric or piezoelectric is combined with a magnetic material. Using this approach in BaTiO3/FeRh we have been able to switch between ferromagnetism and antiferromagnetism by an electric field, just above room temperature (see figure below).

Related publications :

- Electric-field control of magnetic order above room temperature
R. O. Cherifi et al., Nature Mater. 13, 345 (2014)

Nanoferronics exploits the influence of nanoscale ferroic orders on charge and spin current. As many ferroic materials are insulating – notably all ferroelectrics – this influence will often manifest itself through changes in tunnel-type transport transmission, rather than by variations in diffusive transport parameters.

For instance, spin-filters are tunnel junctions in which the tunnel barrier is made of a ferro- or ferrimagnetic insulating material. Because of the exchange splitting of the density of states in this insulator, the energy of the conduction band bottom is different for spin up and for spin down, which produces two different barrier heights for spin up and spin down electrons. Accordingly and because the tunneling current depends exponentially on the barrier height, electrons tunneling across such a barrier are transmitted differently depending on their spin. Practically, the barrier filters out electrons of one spin direction and the tunneling current is highly spin polarized, cf figure below.

We have used different ferromagnetic and insulating oxides as spin filters, including BiMnO3. The figure below shows (a) the field dependence of the resistance of a LSMO/BMO/Au spin-filter junction at 3 K at 10 mV. Insets: bias dependence of the TMR (left); I(V) curve of the junction (right) ; and (b) the field dependence of the resistance at different temperatures for a second junction.

Related publications :

- Spin filtering through ferromagnetic BiMnO3 tunnel barriers
M. Gajek et al, Phys. Rev. B 72, 020406R (2005)

- NiFe2O4 : A versatile spinel material brings new opportunities for spintronics
U. Lüders et al, Adv. Mater. 18, 1733 (2006)

- Tunnel junctions with multiferroic barriers M. Gajek et al, Nature Mater. 6, 296 (2007)

Tunnel junctions with a ferroelectric barrier (‘ferroelectric tunnel junctions’, FTJs) were proposed by Nobel laureate Leo Esaki in 1971, but their experimental realization had to wait over 30 years for the progress in oxide thin film growth. In 2007, we began to investigate tunnel junctions based on BaTiO3 (BTO). We combined scanning probed techniques to write and image ferroelectric domains (piezoresponse force microscopy) in very thin BTO layers, and read the tunnel resistance across these domains (with conductive atomic force microscopy). With this approach, we could evidenced a giant TER effect of up to 75000% at room temperature, see figure.

Subsequently, we defined top electrode pads by electron-beam lithography and lift-off, and measured tunnel electroresistance on solid-state ferroelectric tunnel junctions. We obtained TER effects as large as 1000000%, that is an OFF/ON ratio of 10000. We also found that these devices can switch very fast (10 ns) and have reproducible switching characteristics. Fatigue tests over thousands of cycles did not show a degradation of their performance.We are now using FTJs as memristors, in which the resistance can be quasi-continuously tuned between the ON and OFF states by controlling the ferroelectric domain configuration (through the application of voltage pulses of varying amplitude, duration or repetition number), see figure below. We now work with the Cognitive Information Processing group at CNRS/Thales to develop brain-inspited computational architectures based on these ferroelectric junctions.

Related publications :

- Giant tunnel electroresistance for non-destructive readout of ferroelectric states
V. Garcia et al, Nature 460, 81 (2009)

- Solid-state memories based on ferroelectric tunnel junctions
A. Chanthbouala et al, Nature Nanotech. 7, 101 (2012)

- A ferroelectric memristor
A. Chanthbouala et al, Nature Mater. 11, 860 (2012)

- Giant Electroresistance of Super-tetragonal BiFeO3-Based Ferroelectric Tunnel Junctions
H. Yamada et al, ACS Nano 7, 5385 (2013)

- High-performance ferroelectric memory based on fully patterned tunnel junctions
S. Boyn et al, Appl. Phys. Lett. 104, 052909 (2014)

Oxide interfaces
At the interface bewteen complex oxides, charge, orbital and spin reconstructions can produce new electronic phases that have no equivalent in the bulk of the materials involved. The prototypical oxide interface system is LaAlO3/SrTiO3 (LAO/STO). Although both LAO and STO are wide bandgap semiconductors, a two-dimensional electron gas (2DEG) with a high electron mobility appears at their interface.

A few years ago, we used conductive-tip AFM imaging in cross-section LAO/STO samples to determine that the thickness of the 2DEG at room and low temperatures is on the order of a few nm. This strongly suggests that he electron gas is truly two-dimensional, and thus emerges as a new model system for studying quantum transport phenonema in oxide nanostrucutures (see figure).

Recently, we have also been able to inject spins into the LAO/STO interface, which paves the way towards a new era in oxide spintronics, beyond perpenidcular devices such as tunnel junctions or spin filters.

Related publications :

- High Mobility in LaAlO3/SrTiO3 Heterostructures: Origin, Dimensionality, and Perspectives
G. Herranz et al, Phys. Rev. Lett. 98, 216803 (2007)

- Mapping the spatial distribution of charge carriers in LaAlO3/SrTiO3 heterostructures
M. Basletic et al, Nature Mater. 7, 621 (2008)

- Towards Two-Dimensional Metallic Behavior at LaAlO3/SrTiO3 Interfaces
O. Copie et al , Phys. Rev. Lett. 102, 216804 (2009)

- Gate-Controlled Spin Injection at LaAlO3/SrTiO3 Interfaces
N. Reyren et al, Phys. Rev. Lett. 108, 186802 (2012)