Multilayers

In a wider sense the term multilayer (ML) describes a stack of thin (homogeneous) films of several materials. The thickness of the individual layers reaches from a few Å up to the μm-range; The layer materials can be insulators, metals, polymers, peptides or the like. This already tells that this is a wide field, extending from biology (cell membranes) over chemistry (sensors) to physics (optical and electronic devices, storage media). Here we refer to inorganic systems, used as reflective optical devices or as models for the investigation of magnetic or diffusion properties.

A characteristic feature of multilayers is the strong anisotropy of their structure. In the direction perpendicular to the interfaces the chemical composition changes and the characteristic length scale is of 1 to several hundred unit cells. The deviations from a mean in-plane structure are due to formation of domains, non-smooth interfaces and strain. Within a layer the mean electron density or the mean scattering length density are constant. This means that the interfaces are of great importance for the properties of a ML.
  • One application field for periodic MLs is the use as monochromators for neutrons or x-rays. Unlike with crystals the lattice spacing can be adjusted to the needs and higher harmonics can be suppressed by adequately choosing the layer thicknesses or by modifying the interfaces. A third advantage is, that the lattice constant can be laterally graded, allowing for the production of focusing monochromators or the reduction of divergence.
Fig. 1: Intensity map of a Ni/Ti multilayer with Cr interlayers to prevent diffusion. The intensity of a reflected neutron beam is plotted as a function of the incomming angle 2θ-ω and twice the exit angle θ. The vertical line at 2θ-ω=0 tells about the vertical density profile and thus about layer thicknesses and interface sharpness. The off-specular reflectivity (the rest) is caused by lateral inhomogenieties in the films and a vertical correlation of this roughness'.
Fig. 1: Intensity map of a Ni/Ti multilayer with Cr interlayers to prevent diffusion. The intensity of a reflected neutron beam is plotted as a function of the incomming angle 2θ-ω and twice the exit angle θ. The vertical line at 2θ-ω=0 tells about the vertical density profile and thus about layer thicknesses and interface sharpness. The off-specular reflectivity (the rest) is caused by lateral inhomogenieties in the films and a vertical correlation of this roughness'.
  • If several multi-bilayers with different periodicity are deposited on top of each other several Bragg peaks occur. Thus the range of total external reflection can be enlarged by an appropriate tuning of the layer thicknesses. Coatings of this kind are called supermirror. They are used as coatings for neutron beam guides, or in optical devices where a wide wavelength band or a wide divergence is to be treated. Typical supermirrors consist of several hundred layers.
  • Real interfaces in a ML are neither sharp nor smooth. It is possible to deduce graded compositions and roughnesses from reflectivity measurements. Interdiffusion between the adjacent layers causes a reduction of the Bragg peak height. This effect in turn can be used to determine interdiffusion with an effect to external treatment e.g. heating, applying pressure, etc. In particular, using the fact that scattering length densities vary significantly from isotope to isotope, it is possible to extract self-diffusen.
  • Information-processing technology has so far relied on purely charge-based devices. Only recently a new technology has emerged called spintronics (spin-based electronics) where not only the electron charge but also its spin carries information. The quantum nature of the spin state offers enormous opportunities for a new generation of devices with spin-dependent effects that have tremendous advantages in terms of increased data processing speed and integration densities.
Very promising candidates for new devices include materials with strongly correlated charge carriers, whose mobility is linked to their spin degree of freedom and its exchange coupling to magnetic ions. Superconducting materials are also very attractive since their charge carriers are condensed in a macroscopic quantum state, the phase of which can be easily manipulated, for example, by an externally applied magnetic field. Based upon these considerations, heterostructures made from perovskite-oxides, such as cuprate high Tc superconductors and ferromagnetic manganite compounds that exhibit the so-called colossal magneto resistance effect seem to be ideal candidates for novel spintronic devices.
Fig. 2: Neutron reflectometry data of a multilayer consisting of the ferromagnet LCMO and the high-Tc superconductor YBCO. The structurally forbidden 2nd Bragg peak appears at about 140 K, indicating that the magnetic induction profile deviates from the layer profile. Modulation of these measurements and other techniques tell that the magnetic field penetrates into YBCO with an antiparallel orientation compared to LCMO.
Fig. 2: Neutron reflectometry data of a multilayer consisting of the ferromagnet LCMO and the high-Tc superconductor YBCO. The structurally forbidden 2nd Bragg peak appears at about 140 K, indicating that the magnetic induction profile deviates from the layer profile. Modulation of these measurements and other techniques tell that the magnetic field penetrates into YBCO with an antiparallel orientation compared to LCMO.
The primary goal of our research program is to understand the fundamental interactions and related quantum-phenomena that are expected to emerge when the two materials are brought in contact with each other. Polarised neutron reflectometry is able to probe the relevant structural and magnetic properties on a nanometer scale and is therefore one of the techniques of choice for the analysis of the interface properties as well as the study of proximity-induced interactions between the superconducting and ferromagnetic order parameter.

Funding: ETH internal grant, SNF, EU, Others 

Partners: 
Max-Planck-Institut für Festkörperforschung, Stuttgart, Germany, http://www.fkf.mpg.de/start.html 
Technische Universität München, Germany, http://www.tu-muenchen.de/ 

Contact: jochen.stahn@psi.ch, Jochen Stahn