I have grown thin films of a ferrite on three different substrates. It is under compressive (2.35 %), weak tensile (1.06 %) and strong tensile (4.83 %) strain. The thickness of films is about 1000nm. Thick films are generally relaxed and the reciprocal space maps of these showed the same. This should mean that the epitaxial strain will not influence the properties such as magnetism of these films. But my magnetization data shows different behavior for all three films. I want to know whether epitaxial strain has any role to play or not? Is it possible that when the film is growing the epitaxial strains are seeding defects or dislocations which get propagated through the volume of thin film? For the sake of understanding, I am also attaching the magnetization versus magnetic field plots of these films compared together.
Your 3 substrates are different. Did you remove the substrate magnetic signal from the M(H) loops ? How accurate is your estimate for thickness ? Do you know if the films are textured (out of plane and in-plane ?)
If the 3 films are relaxed and their thicknesses are identical then either there is a strong contribution from the substrates or you measured along different in-plane anisotropy axes (easy or hard depending on texture relations). 1 micron thick films are likely to look relaxed but close to the interface you may have a gradient of relaxation across the first 10-100 nm. Different substrates may induce different in-plane orientation of crystalographic axes.
Relaxation is typically attained through defects. In elements like iron that can have different oxidation states defects can change the nominal charge, and therefore the magnetism. We observed something like this in a recent article: http://journals.aps.org/prl/abstract/10.1103/PhysRevLett.112.087202
I agree with Dr. Piramanayagam. I wish to add that the magnetic properties would be influenced by stresses and this concerns the anisotropy. This also would be percetible only if the magnetostriction of the material is high because it is this property which induces anisotropy if streses (tensile or compressive) are present.
Ibn your samples apparently there is considerable magnetostriction.I fo you tell us the compostion it can be easily answered.
You might consider that different strains can result in different critical thicknesses and different strain layers at the interfaces between the films and substrates. Although the x-ray data show that the strains were released (your thickness is much larger than the critical thicknesses), the released mechanisms can be different, ie. edge dislocation, antiphase domain boundary, stacking fault, etc. and the strain may not be fully released. Especially, the different strain distributions along the interfaces will change the spin orbital interaction, which result in completely different magnetic properties or electronic band structures. To further understand these issues, you should conduct the cross section TEM to study the interface structures for more details. Also, you can refer Chunrui Ma's paper at Appl. Mater & Interf. 6 (2014) 2540-2545. dx.doi.org/10.1021/am404951v.
Dear all...thank you for your answers. Well the films have different out of plane parameters on different substrates. The film on maximum tensile strain showed a symmetric square shape growth of grains.
Can you please name the substrates that you used. The magnetic data you attached showed that magnetization of the sample grown on substrate with 4.83% tensile strain is still not saturated until 5 Tesla. Can it be the background signal from the substrate? Did you measure the bare substrate to verify it? In the sample on substrate with 1.06% tensile strain, sample can be fully strained but in sample with 4.83% mismatch, it is expected that strain relaxation will occur through creation of dislocation and defects. However, almost one order of magnitude difference in saturation magnetization between samples with 2.35% strained and 1.06% strained sample indicates that anisotropy could be a deciding factor as already mentioned by Dr. Krishnan.
mam I have used MgO(100) to grown my films. The ZFC-FC curves of it show almost reversible behavior with barely any splitting between them. I checked the MH of the substrate I am attaching it for reference.
It is then quite clear that the substrate has quite significant magnetic contribution. Above 3T, it is clearly diamagnetic which will suppress the actual signal from the film. Please try to convert the substrate data to same scale of unit (emu/cc) and subtract the background signal from the film signal. Please also do the same for other substrates as well in case they also have some magnetic signals. Then only you will be able to compare the film properties correctly.
Do you have the XRD rocking curve profiles for these samples?
https://www.youtube.com/watch?v=JESFFQpwumY&list=PL7032E2DAF1F3941F&index=3
Compare for instance our paper
Article Magnetic response of (Cr,Al,Si)N nanocrystallites on the mic...
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Physical Phenomena
- Luminescence