Hello Philip,

I measured Co-Pt multilayers by MOKE several years ago without seeing this effect. The possible explanation which first comes to my mind is an undesired tilting of the magnetic field with respect to the surface normal. Alternatively, the intensity "zero" may be shifted ... but this should rather lead to an asymmetry in the loop. All other explanations (which I experienced myself) work only with normal MOKE, not with the MOKE microscope. And the multilayer stack is surely too thick to "see" the substrate, isn't it?

By the way - which resolution does your MOKE microscope have? We have some nanostructured Py samples which we measured with MOKE (longitudinal / transverse) and simulated (Magpar / OOMMF), but it might be interesting to have a deeper look into the spatially-resolved magnetization reversal processes (esp. in combination with simulations).

Have a nice day,

  Andrea

Hello Philip,

next try, after RG deleted my answer ...

Measuring Co-Pt multilayers with normal MOKE myself, I didn't see this effect; thus I assume it's not related to the physics of the sample. The first explanations coming to my mind would be an undesired tilting of the magnetic field with respect to the sample normal or a shifted "zero" intensity in the evaluation (which would probably rather lead to an asymmetry in the loop). Or is there a chance to "see" through the complete multilayer and to have an influence of the substrate? All other experimental problems are only relevant for normal MOKE, not for your technique. Or is it possible that you see some longitudinal or polar magnetization additional to the desired polar effect due to the light impinging on the sample not exactly perpendicularly?

By the way, which resolution does your Kerr microscope have? We have some nanostructured Py samples which we measured with normal MOKE (longitudinal / transverse) and simulated (Magpar / OOMMF), but it might be interesting to have a deeper look into the spatially resolved magnetization reversal processes.

Have a nice day,

 Andrea

can this be due to a diamagnetic substrate? If your film is thin compared to the skin depth, the substrate could play a role...

The film thickness would really be interesting.

However, here http://beach.mit.edu/wp-content/uploads/2012/07/Emori_Beach_CoPt_Buffer_JAP_2011.pdf you can also see a slight additional diamagnetic signal for the out-of-plane measurements, independent of the buffer layer - and "seeing" through 120 nm Pt, which is one of the buffer layers, doesn't seem probable to me. Especially since Pt is, as far as I know, paramagnetic.

So - maybe it's indeed real physics. What does your normal MOKE or a SQUID state about this problem?

Have you tried measuring a nonmagnetic sample (e.g. Ag mirror)? Is your measurement free from such tilt? Since you measure in a microscope, (parts of) your lens system might be exposed to magnetic field and produce Faraday rotation?

Farren, you are stating the obvious.

The question remain, what is causing the para of diamagnetism (in the original problem stated above) in these measurements? 

This may due to the Faraday effect of the lenses. It can be removed by liner subtraction.

Thank you all for your informative answers. I appreciate every suggestion. Unfortunately, I have not yet found the time to check on the things you suggested, but I will definitely do so, as soon as possible. Best Regards!

PS.: The substrate is SiO2, so diamagetic. And as far as I know, the film of CoPt is only a few ML thick.

You will surely be able to see through a few ML, I thought more about a thick multilayer system. So maybe it's simply the substrate which you see in addition to the ferromagnetic loop.

Are you guys sure that a diamagnetic response, say, of Silicon would produce a sizeable MOKE signal?? (I never pondered that but I am surprised)

In my old MOKE measurements on GaAs samples I assumed to see a (very small, of course) signal, and GaAs and SiO2 should have similar magnetic properties, as far as I know. But of course it's always hard to be sure whether this is really a signal or simply an effect of the cryostat windows or something similar. Nevertheless I wouldn't definitely exclude an effect from the substrate as long as this hasn't been checked with other samples.

I would hope, the question could be den Sweed on a more fundamental level: does the state of diamagnetic pobenzawou of an insulator or semicouductor give rise to Kerr rotation? Is there an off- diagonal element in the optical conductivity tensor?

In diamagnetic or paramagnetic materials, the Zeeman splitting of the energy niveaus relevant for optical transitions should create magneto-optical effects. Of course the possible measurements depend on the wavelength of the laser used for the experiment, but that's also not unimportant for measuring ferromagnets.

However, as you have stated before - measuring a reference sample would be easier than these theoretical discussions ;-)

My experimental setup has no microscope thus no lens but still I see this effect on Fe/MgO sample. I haven't tried measuring Ag mirror but soon i will just to check, did any one find a solution ?

Well I, too, were always puzzled to see those "tilted" loops obtained from Kerr-type measurements.  Especially, when those result were presented together with more traditional (VSM or SQUID) measurements, not exhibiting this feature.  On the other hand, we can occasionally see hysteresis loops taken at direction perpendicular to the exciting field.  They are similarly "tilted" with saturation equal to zero at high fields (no surprise, right?).  Perhaps your observation has its roots in"non-purity" of Kerr apparatus arrangement, thus recording both components of magnetization?  Just a wild guess ...

We recently observed the effects of this exact type and also some even more unexpected hysteresis loop shapes in exchange-coupled magnetic garnet trilayers (soon to be reported in literature), in Faraday geometry with out-of-plane magnetisation. The loop shape tilting of this exact type (somewhat unexpected in its direction/sign) was also noted and is still under investigation.

We measured these rather unexpected hysteresis loops of Faraday rotation in garnet multilayer structures on GGG substrates. The paramagnetic effect of 0.5-mm thick substrate should have induced a tilt in the opposite direction, but not in some of our samples.