s19RemnantMagnetization
Remanence Variability in Response to Demagnetization Techniques
James Kieke and Austin Schleusner
University of Minnesota, Twin Cities
Methods of Experimental Physics - Spring 2019
Introduction
For understanding the behavior of ferromagnetic materials, one value commonly examined is remanence. Remanence refers to the remaining magnetization of a material after magnetic field is applied and removed. Researches often used remanence measurements in fields such as geophysics, bio-imaging, and magnetic storage. However, researches do not always specify the preparation of the materials tested before drawing conclusions. However, in particulate (i.e. made of particles) materials, it was shown that remanence values were dependent on the preparation of the ferromagnetic tested [1]. Specifically, how the samples were demagnetized before measuring remanece values affected the remanence values measured. Thus, researchers should specify the demagnetization technique used when properly forming conclusions from remanence values.
For our experiment, we looked at bulk ferromagnetic samples of nickel and cobalt to see if a similar dependence arose. We compared measured remanence values to the Stoner-Wohlfarth model for three different demagnetization techniques.
Theory
Two types of remanences (shown by the red circles in the figure above) were measured: Demagnetizing (ID) and Remagnetizing (IR) remanence. Demagnetizing remanence measurements begin with a sample saturated in the forward direction. The terms 'forward' and 'backward' refer to the directions of the applied magnetic field. Once saturated, fields of increasing size (in steps of H) are applied and removed until the sample reaches saturation in the backward direction. Upon removal of a field, remanence values are measured. Remagnetizing remanences measurements begin with the sample in a demagnetized state. Once demagnetized, the same fields as for demagnetizing measurement are applied to get remanence values.
To demagnetize the samples, three methods were used, which are labeled as AC, DC Backwards, and DC Forwards. For AC demagnetization, a sample is first saturated forwards, then a AC field of decreasing magnitude is applied. For DC backwards, a sample is saturated backwards, then a forward field is applied until there is a net zero magnetized state. For DC forwards, a sample is saturated forwards, then a backward field is applied until there is a net zero magnetized state.
The Stoner-Wolfarth Model (SWM) [2] of magnetic remanence assumes independent magnetic domains within the lattice of a ferromagnetic material. From this assumption, predictions are made for the relation between ID and Remagnetizing IR. The figure above shows the predicted relations for each demagnetizing technique, which are normalized by the remanence after saturation. Deviations from these predictions are often used to indicate domain interactions.
Apparatus
Remanence measurements were made using a Vibrating Sample Magnetometer (VSM), show on the left in generality and in detail on the right in the figure above. Samples of nickel and cobalt were placed within the VSM on a mounting rod between two faces of an electromagnet. The samples are oscillated vertically by a VSM controller, perpendicular to the applied fields. This oscillation induces an EMF proportional to the magnetization of the sample within pairs of coils on the magnet faces. This induced signal is then run to a lock-in amplifier, which sends the signal to a labVIEW VI via a DAQ board. The labVIEW also has a control to the magnet power supply, which adjusts the current sent to the electromagnet. This current is proportional to the magnetic field applied to the sample, which is measured by a Hall probe. This field value is also sent to labVIEW, which thus allows the measurement of magnetizations for known applied fields.
Results
The plots of cobalt and nickel (below) show the parametric relation between ID and IR normalized by the remanence after saturation. Observed data points for the three methods of demagnetization are plotted over the lines of predicted values from the SWM (as color coded).
The significant difference between the predictions and the observations for DC forwards demagnetization as well as the noise in the observed data could be the result of several factors in this experiment. First, orientation of the sample within the VSM was very sensitive in the induced EMF values. To address this, minute adjustments were made to the angle of the longest axis of the sample with respect to the applied field until the maximum EMF was induced. Vertical positioning along the driving rod also had a significant impact on induced EMF data. Lastly, and most importantly, the remanence of the field-applying magnet itself necessitated inputting a current that would ideally produce a field that "overshot" the zero applied field point when collecting remanences of our samples. This made returning to a zero magnetization state as observed by the hall probe more difficult than anticipated.
Conclusion
The accuracy of the SWM predictions was not ideal to our observations, but the conclusion can still be drawn that different methods of demagnetization do produced different remanences behaviors. While the differences between different methods of demagnetization may not be completely modeled by the SWM, the difference is still important. As is visible in the different AC series behaviors of the nickel and cobalt samples, some variations between differing ferromagnetic materials will result in different remanences. However, between the observations of this study and the findings of Hamann and Vogt [3], the variations between materials does not approach the extent necessary for demagnetization to become a null factor. For studies involving magnetic remanence, it is necessary to address the technique of demagnetization to draw meaningful conclusions.
Acknowledgements
This project could not have been possible without access to the lab of Dan Dahlberg, coding assistance of Kevin Booth, and the prior works of Aaron Hamann and Ryan Vogt (see spring of 2017 projects).
References
[1] Sharrock, M. P.Time Dependence of Switching Fields in Magnetic Recording Media (Invited).Journal of Applied Physics, vol. 76, no. 10, 1994
[2] Stoner, E. C.; Wohlfarth, E. P.(1948). ”A mechanism of magnetic hysteresis in heterogeneous alloys”. Philosophical Transactions of the Royal Society A: Mathematical,Physical and Engineering Sciences.
[3] Similar MXP Project:
A. Hamann, R. Vogt (2017) Analysis of the Dependence of Remanent Magnetization on De-magnetization Technique, University of Minnesota;
The results as seen in Table I show the areas between the lines predicted by the SWM and our observed data. This was calculated through the trapezoidal method of summation. The DC Forward data sets showed the greatest deviations from the predictions for both ferromagnetic samples. The observations of this model can be best applied to the AC demagnetization of the two samples. In the nickel AC demagnetization observations, the area difference of 0.26 is significantly greater than the area difference of 0.17 for cobalt as both of these values have an error of 0.01. This result lends itself to the conclusion that the nickel sample's lattice structure more closely resembles the non-interacting magnetic domains assumption of the SWM than that of cobalt. This is also seen in the plotted data how the cobalt observations for AC are linear like the model predicts compared to the slope of the nickel sample's AC plot approaching zero as ID approaches 1.
