We are performing MOKE experiments in polar and longitudinal geometries. Our samples range from single magnetron-sputtered or electrodeposited magnetic metals to multilayers. We use a PEM and lock-in amplifier in the standard configuration. Most of the time, we get MOKE signals that are shaped liked a butterfly. See attached. This means, reversing the field does not invert the signal. Sometimes, however, the desired inversion does take place.
Are we not properly understanding the lock-in phase, or is the magnetization not really reversing its alignment, or is there something more subtle behind all this?
Can you do the control experiment without the lock-in amplifier? Just measure the Kerr Rotation without modulation by the PEM. Might need some more averaging to get rid of noise, but it will show if the lock-in is the troublemaker.
sometimes, the existence of a square component in your signal is related to the fact that a part of your sample/beam is touching the magnet itself when your drive the magnetic field....take care of your optical mounting...
Can I ask the specific setup of your whole measurement instruments? I mean your system is based on the paper W S Kim 1993, too? Maybe we can discuss some more about this.
It should be a problem connected with phase. Signal should change direction. In fact this is how one gets the hysteresis loop using MOKE.
Your control sample (Co) shows hysteresis, suggesting your setup is okay. Did you resolve the problem observed in your sample?
Besides, what is the reason the loop of Co is horizontally flipped? Is it due to the lock-in phase? I see you only measured lock-in X. Did you press a "auto phase" at zero field condition? I was wondering you should see a mild value in lock-in Y.
Dear Chung,
Yes indeed, the horizontal flip is due to phase. The phase is adjusted so to minimize the signal in Lock-in Y/or you can auto-phase at zero B field. Once auto-phased, Lock-in X signal is usually much larger than a mild value in Lockin-Y. If not, you can certainly switch to Lockin-R which will also remove the horizontal flip.
Regards
Dear Ali Akbar,
Thank you for your help to answer my questions.
I have a few follow-up questions. I am using a PEM for MOKE/MCD measurement. My understanding is that the phase detected by a lock-in amplifier comes from the interaction of an excitation laser and a target sample. Sure, I can auto-phase at zero B-field and the sample exhibits randomly-orientated magnetic domain. Basically, it sets a reference/zero. But whenever you apply and sweep a B-field to the sample, it is hard to have a good reference point to do auto-phase because the sample is somewhat magnetized (along its hysteresis loop) even at a zero B-field. Do you know any good practices to auto-phase with consistency over a series of measurements on the sample? Or you suggest to manually offset a data? Or we can account only lock-in R instead of lock-in X?
Regards
Dear chung
Once autophased at zero field, one does not need to perform auto-phase during the multiple B sweep for a single sample. If you intended to change your sample, autophase procedure has to be repeated. As far as lock-R is concerned, you can simply collect lockin-x and y data and extract R from it. I hope this might answer your querries.
Regards
Thanks, @Ali Akbar. I have phase-locked my lock-in amplifier and PEM. I can further do autophase at zero magnetic field. But it should be okay as lock-in R or the photo-voltage signal (from a photodetector) is the most important parameter in the measurement, not lock-in X, Y as they depend on the phase difference with the reference frequency.
- Badges
- Science topic
- Similar topics
- Chemistry
- Nanotechnology
- Nanostructures
- Thin Films