@S.N. Piramanayagam: thank you sir for your reply. But rotating the sample possibly gives the information of anisotropy in the system and again it will also depends on the kind of sample we are dealing with viz; polycrystalline sample, oriented or epitaxial thin film. If you have any literature relevent to this please make it available to me.

@Oleg Petracic: Thank you for your nice reply. But my curiosity is that; Even though the available magnetic field is not enough to saturate, the observed behavior in which hysteresis loop is centred about a non-zero magnetic field may be the real feature of system. So how one can confirm the EB effect in minor loop?

@Oleg Petracic: thank you very much sir for your valuable information/suggestions. We have a nice equipment from Quantum Design "squid-vsm" temperature range(1.8K-1000 K) and field upto 7 Tesla. In my case in zero field cooled M-H at 2K, there is no shift means that the loop in centred about zero magnetic field. Whereas, in case of 7T field cooling the two Hc-values are different. Well, what you suggested may be interesting and I will check this. But at higher temperature (above 150K) system is paramagnetic. What you suggested is a kind of M-T at max field. Since I am working in thin film so at that much high field substrate (which are generally diamagnetic) moment will dominate.

Sir,

Exchange bias phenonmenon from minor to major will be only explanable for the case of B and H at their lower points.

@oleg Petracic: you are right sir.One can easily subtract the diamagnetic contribution.But as you suggested to heat up the sample until one reaches the satutarion at that field,in that sense I have mentioned substrate contribution and in such cases it will become difficult to heat up the sample in order to guess for a saturation since diamagnetic contribution from substrate will not saturate.

There are two more features of exchange biased magnetic materials:

1) the loop gradually shifts horizontally when the field is cycled ("training effect")

2) upper and lower branches of a loop sometimes have different shapes.

Observing such a behavior, even in minor loops, is a good reason to call the phenomenon the exchange bias, IHMO. With no precise quantitative characterization, unfortunately.

Besides, I think the term "unidirectional anisotropy" should be avoided as it doesn't fit to the classical description of magnetic anisotropy, neither magnetocrystalline nor shape anisotropy.

@ Marek Gutowski: Thank you sir. 1> Is training effect exist only in the exchanged biased system? What if system show measuring field dependence means that if vertical or horizontal shift depend on the measuring field value of hysteresis loop?

2> Is it possible for system possessing genuine feature of exchange bias to show measuring field dependece?

Dear Pankaj,

Vertical shift of the hysteresis loop during field cycling bears the name of "reptation" (Neel, around 1950, at least 10 years before "exchange bias" has been observed for the first time). It may be called training effect, but ordinary dipolar interactions together with domain structure seem sufficient to explain its origin. It is possible to simulate it by Jiles-Atherton model of magnetic hysteresis.

I'm not sure what do you have in mind in your second question. I've encountered the case of thin magnetic layer (MnFe, if I remember well) deposited in presence of 100 Oe field and still exhibiting clear exchange bias effect in ferromagnetic resonance experiment, when the applied field was as high as 5000 Oe.

Pankaj,

I've seen the article you mentioned and I'm in trouble here. See, you cannot speak seriously about exchange bias effect when your hysteresis loop is always positive - there is no clearly defined coercive field there, so how you determine H_E?

I think we are dealing with two separate branches in magnetic investigation. The figures (and measurements) like the one in Fig. 5 are characteristic for permanent magnet industry. Permanent magnets are with us since more than thousand years, long before the exchange effects were identified. Exchange bias was discovered in 1957 by Meiklejohn and Bean and at first it was a kind of scientific curiosity, called by them simply "new magnetic anisotropy". Since then no widely agreed experimental protocol has been established, even after the training effect has been discovered (unintentionally). Therefore you seem to be on your own.

Is exchange bias effect present in permanent magnets? Is it the rule or exception? Such questions seem unanswered today. In fact, the behavior of permanent magnets was routinely explained by the high anisotropy of their constituents and the dipolar interactions between them. But .. it is no secret that permanent magnets usually have granular structure, so exchange interactions between grains cannot be excluded. It still remains to be explained whether or not the antiferromagnetic interactions are essential for exchange bias type effects. Why strong ferromagnets should exhibit antiferromagnetism? I think it is only possible via RKKY-type exchange interactions, especially at or between grain boundaries.

The reptation is still something else (usually reported for not-so-hard magnetic materials, but also in NdFeB nagnets), see the following papers:

P J Thompson and R Street

Viscosity, reptation and tilting effects in permanent magnets

J. Phys. D: Appl. Phys. 30 (1997) 1273-1284

O. Hovorka, G. Friedman

Onset of reptations and critical hysteretic behavior in disordered systems

Journal of Magnetism and Magnetic Materials 290-291 (2005) 449-455

J. Takacs

Analytical way to model magnetic transients and accommodation

Physica B 387 (2007) 217-221

And finally: don't call me "sir", please. I'm Marek.