User Tools

Site Tools


Element Specific Modelling

The standard modelling of thin films is done in the layer table and consists of a list of layers of different compound materials. A compound here is defined as a group of elements which have a homogeneous density inside the layer (e. g. crystals). The optical constants can be achieved by a linear combination of element-specific scattering factors (e. g. off-resonant Chantler-tables). The main parameters for each layer are the thickness, interface roughnesses and the optical constants. If the properties of the compound change (e. g. by strain, oxidation, contaminations) you have to introduce additonal layers with assumptions about the properties of the layer.

The element-specific idea is to introduce an advanced way of modelling the layer. The main assumption is that you have seperated layer profiles for each element. In contrast to the compound-modelling the parameters for each element specific layer are the element densities, thicknesses and the element interface roughnesses. Each element has their own scattering factors f1, f2.

Activating Element Mode

At the moment there is only one way to initially change to element mode. After the initialization step the change of the mode can be done in the menu under “mode”.

1. Work in the compound mode. Introduce the initial layer configuration and begin to make a “coarse” fit of the system. (coarse means: as good as possible)

2. Type in the initial material density [g/cm^3] and the chemical formula in the properties dialog for each layer (Right Click→Properties). E. g.

  • “Si” (Silicium)
  • “H2O” (Water)
  • “LaNiO3” or “La1Ni1O3” or “La3Ni3O9” (Lanthan-Nickelate)
  • “SiO2C0” Oxidized Silicium with possible Carbon contamination.

3. In the final step you have to go to the Optical Constants Frame in the Data register and start the converter by clicking the button “Material to Element”. You are in element mode now.

The conversion algorithm uses an internal database to get the molar mass of each element.

Using Element Mode

The most important additional options for the element mode are in the “Element Mode” Frame in the “Data” Register. Here you can change the active element for the layer table. The layer table is different for each element and consists of the entries thickness in Angstrom, density in mol/cm^3, and roughness in Angstrom. The scattering factors f1/f2 are the same for each element.

  1. Additional you can display the element specific profiles and f1/f2 for each element.
  2. Instead of .db files the program needs .ff files for the scattering factors. E. g. Ni.ff, H.ff, …..
  3. Fitting works like in compound mode.

The current version has following limitations

  1. After initialization step no additional element types can be added
  2. In total eight different elements are supported. Otherwise the display of the element density profiles will not work.

What is the strength of this element mode? (Valency, Magnetic profiles, ...)

The strength of of the element mode lies in the fact you can choose completely different profiles for each “element”. The word ““element” is meant in a more general way.

For example if an element has different valency in one system the on-resonant scattering factors might change. To take into account such systems different valencies can be introduced by an additional element. This should work straight forward. You can choose an arbitrary additional element.

For example you can introduce “CoCu0” for an mixed valence system (e. g. Co(2+) and Co(3+)). In this case Cu is used as the second valency with an initial density of zero. The element Cu is chosen because of the similarity of names.

The other possibility is to use a element just for the magnetic profile. This means, that for this element the scattering factor f is set to zero and only the magnetic scattering factor fm is unequal zero. The “density” is then proportional to the magnetization and can be even negative. ReMagX supports such kind of profiles.

documentation/element_specific_model.txt · Last modified: 2015/11/30 10:45 by macke