Monolithic INtegration of functional Oxides on Silicon for novel micro-system devices – MINOS

The objective of this project is to integrate high quality functional (piezoelectric and ferromagnetic) oxide heterostructures on Si using SrTiO3/Si(001) templates, in order to develop a novel generic demonstrator device which is expected to validate the ability of reversible electrical modulation in functional oxides through the application of a dynamic strain. This demonstrator should allow producing for the first time a multiferroic composite working at room temperature. Perovskite-type metal oxides (ABO3), in particular, are becoming a very important class of materials due to their unique dielectric, piezoelectric, ferroelectric, ferromagnetic, optical, electro-optic, and catalytic properties. It is now highly desirable to integrate these functional metal oxides with mature silicon manufacturing technology. As a consequence, oxide growth has to be performing onto Si(001) substrates. As Molecular Beam Epitaxy (MBE) growth processes for SrTiO3/Si was developed for CMOS gate oxide application (including at INL), the integration of new oxide functionalities on silicon can now be expected since in addition, SrTiO3(001) is the common substrate for most perovskite growth studies. The challenge is to obtain oxides on Si with the same crystalline quality and stability as obtained on SrTiO3 substrates. Ideally the zero default templates should corresponded to the 'pasting' of a SrTiO3 substrate on silicon. The first step of our project will be to optimize the epitaxy of a conductive La-doped SrTiO3 buffer layer on silicon for bottom electrode. On this perovskite electrode, piezoelectric BaTiO3 will be grown by MBE at INL and piezoelectric (Mg,Nb)O3-PbTiO3 (PMN-PT) by Pulsed Laser Deposition (PLD) at IEF. Then an heterostructure will be prepared by growing on the piezoelectric layer a ferromagnetic (La,Ba)MnO3 (LBMO) layer by PLD at IEF. PMN-PT has a high piezoelectric coefficient (2000 pm/V) which leads to a large atomic displacement in the lattice. Well known Si micromachining methods (Si back side anisotropic etching) will be used to fabricate a microsystem device with a membrane free from static strain from the substrate. The device will allow controlling the application of a dynamic strain onto the ferromagnetic layer via the piezoelectric oxide in order to modulate the LBMO physical properties (Curie temperature, carrier mobility, etc.). Indeed, the physical properties of manganites (A1-xA'xMnO3) are fully governed by structural deformations. Any deformation of the perovskite lattice leads to a modulation of the magnetic properties (due to a modification of the orbital overlaps in Mn-O bonding). Dynamic evolution of structural deformation (Mn-O length, Mn-O-Mn angle value) under the piezoelectric strain will not only observed but also quantified using Exafs, Raman and IR characterizations at SOLEIL synchrotron radiation facility. These evolutions will be correlated to the modulation measured on the ferromagnetic LBMO physical properties (transport properties). A shift of around 100K of the ferromagnetic LBMO layer Curie temperature is expected in our micro-system. More generally the project will validate the quality of state of the art SrTiO3/Si templates, the concept of combining MBE grown templates with standard oxide growth techniques, and the possibility to fabricate novel high quality devices (as our prototype MEMS).