Dear David,

while you might not think so, your question is still very broad and it is not so easy to come up with an answer which should suit your needs (maybe but for the quest for literature) and not be utterly lengthy.

Maybe you will find more literature when you include SQUID magnetometry in your search. While the sensitivity is not the same, many considerations should take the same direction.

Most important is to realize what exactly it is that you wish to find out. There are so many kinds of measurement protocols you could want to apply...

With thin films, there are some typical issues, though. Usually the mass or volume of the magnetic (-ally interesting) material is many orders of magnitude less than the substrate carrying the film. Even a small amount of dia- or paramagnetic susceptibility from the carrier may easily amount to the same magnitude (or even much more) of a magnetic moment than the moment you're interested in. In such cases, I would say it is mandatory to make accurate reference measurements on empty substrates of same dimensions as the specimen(s) you're interested in. Might effectively double the time required to accomplish the desired task. The magnetometer cannot tell apart where a magnetic moment is coming from. Only you can.

Very important: take massive care that there is strictly NO contamination from any undesired material. Using the wrong (iron based) tweezers just once may spoil your sample for good. 

Also frequent with thin films of magnetic materials: surfaces are easily corroded (i.g. oxidized, or have a different oxide stoichiometry at or near the surface). Thinking about some suitable cap layer might be worthwhile but could also change the characteristics of the film. The better your films are characterized by additional means the higher the chances that you will learn something useful from magnetic measurements.

Finally, do search on RG for previous questions involving VSM & SQUID magnetometry, there are quite a few. These should be instructive in getting ideas about what could possibly go wrong (because this is usually when people seek help here - unlike you). Some typical issues are sample size (stick to the specified volume) and (in-) correct sample mounting.

What is specific about measuring ferromagnets? Quite generally, you have access to mS(H,T), where mS is the total specimen magnetic moment. Unless you measure major magnetization loops, there will be a memory with respect to the "magnetic history" of the sample. Depending on what you'd like to do, it might make sense to think about how to obtain a well defined magnetic state (->, saturation, thermal demagnetization, AC-demagnetization...). You might want to look for the ordering temperature, shape anisotropy, hysteresis, measurement of (partial) magnetic reversal etc..... which all reveal something about the ferromagnet. As I said before, among the most important things is to actually know what type of information you are looking for. You will most probably not be able to do everything that could be done, so you need to judge what is important. You might want to ponder that thoroughly with knowledgeable colleagues around you or at least the folks whose magnetometer you will use.

Thin films contain very low magnetic moments so it is very difficult to accurately measure the magnetization using VSM due to which you have to go for SQUID magnetometry. While with VSM you have to take care about the substrates dia or para magnetic behaviour, either you can subtract the substrate behaviour by doing empty substrates VSM or you can also use the slopes deduction method which can also give you the approximate result.

gud luck.

The final suggestion by Vinay Sharma ("slopes deduction method") is indeed often done. In doing so you implicitly assume that you know well enough that the magnetic film is magnetically saturated. Many times, though, this is not the case and there is a substantial risk of serious error. I strongly discourgage doing this, at least without complementary evidence.

Hi David,

I have carried out vibrating sample magnetometry (VSM) measurements of the magnetic moment of 5nm films of permalloy (Ni80Fe20) on a Si wafer. However, the set-up I used at the University of Leeds, is a SQUID-VSM so it was extremely sensitive to magnetic moment measurements. To make the VSM measurements I had to cut my samples to a size of approximately 3mm2, I imagine this will vary between different machines so it would be a good  idea to ask your collaborator about size restrictions. 

Dear David,

As previously stated, think about what kind of data you want to acquire and the information you are looking for.

2nd make sure that the machine you will use for measurements gives you the needed sensitivity. There are VSM's nowadays with high sensitivity for films but not necessarily all are suitable. (SQUID VSM as mentioned would work of course).

Another point you have to keep in mind is the sample geometry. Especially for films when you measure in plane or out of plane you have to correct for that. The real magnetic moment can vary a lot in comparison to the measured one. It influences the demagnetizing field as well.

There is of course a lot more to consider as mentioned by Mr. Fauth but most of it only makes sense to discuss when knowing what kind of information about the sample you are looking for.

The main criterion for the suitability of a VSM for ferromagnetic thin film measurements is the effective sensitivity of the specific instruments which is determined by the level of noise and background signals. A well-designed VSM with conventional pick-up coils (not SQUID-VSM) will allow useful room temperature measurements down to 1 nm Permalloy for a sample size of 5 mm x 5 mm (e.g. with our lab-built instrument). To measure very low moments it is very important to avoid any kind of contamination of sample and sample holder as pointed out by Kai Fauth. One major advantages of a suitable VSM (compared to SQUID) generally is the precise variation of the sample orientation relative to the applied field. It is important to know what kind of information you need (absolute saturation or remanent magnetic moment, coercivity, magnetic anisotropy etc.) as pointed out in previous comments (e.g. out-of-plane measurements are more difficult to interpret than in-plane ones) in order to give a more specific advice.     

I would like to thank everyone who posted for there feedback, it means a great deal to me. sadly I haven't been able to comprehend 100% of what has been said as I still have a lot to learn.

to provide you all with more information, My material, Bi4Ti3O12 when doped with Iron and cobalt, has been known to show a remnant magnetization of 7.8memu/g in the bulk form. They used VSM in this case, however I do not know if standard VSM is sensitive enough when for meaurign a thin film with a RM of of similar value. If anyone could provide me with insight on this matter. I would be most appreciative.

Best Wishes

Jamie

If the work you mention is published then it would be good to give the references.

Otherwise, combine the bulk magnetization above with the bulk density (g/cm3) such as to get the magnetization in units of moment/volume. Then multiply with the dimensions of your prospective films [this assumes you know what thickness(es) you want to grow and that you explore pretty much the maximum permitted sample size]. That provides you with an estimate of the saturation moment of our sample.

Now set yourself a limiting moment as a fraction of the saturation moment that you still want to be able to detect reliably (be it 10%, 5%, 1% of the saturation moment, or even less. You decide!) That will represent the required sensitivity of the magnetometer. Once you come up with such a number, your colleagues will hopefully be able to tell whether they can cope with it.

Going through that procedure you may notice that in/after your measurements you will have to go the other way round: you will measure some (total) moment and derive the film magnetization (in whatever units - I personally like Bohr magnetons per relevant atom much more than emu/g...). You cannot obtain magnetization values that are more accurate than your knowledge of volume, density and composition (stoichiometry) of your films. To get those quantities may take quite a bit of extra characterization.

With respect to what you didn't comprehend: challenge us on explaining it better or more adequately with respect to your current level of knowledge.... 

 Dear Dr.Fauth,

I am incredibly grateful for your continued support. In inexperience in the matters largely comes from my lack of background in magnetic materials, as I have only know stepped into he field, branching from an electronics background into multiferroics. After reading thoroughly through everything that has been said, the key thing I do not understand is your last paragraph of your first reply, where you talk about measuring ferromagentisation specifically. I understand by MS(H,T) you are referring to the total magnetisation of the sample, regardless of mass, but I unable to comprehend what you are referring to interest of your text.

I have attached two journals discussing what my work is based on below. From what I have gathered from peoples input, I need to calculate what I expect the total magnetisation of my thin films to be. Working from the value of 7.8memu/g I can calculate the expected total magnetisation of the thin films based on their physical dimensions and density. However, this is working upon the assumptions that both the density and magnetisation scale perfectly from bulk to thin films level. So I expect some error.

As with Ferroelectrcity, I expect ferromagnetism to greatly depend on the orientation of the film, however curiously enough neither of the journals presented discuss orientation. I assume they present data regarding the strongest magnetisation they could find.

Any further feedback you can provide would be greatly appreciated.

Best Wishes

Jamie

Article Multiferroic Properties of Layer-Structured Bi5Fe0.5Co0.5Ti3...

Article Effects of Co-substitutes on multiferroic properties of Bi5F...

 Concerning your first question: When you measure a M(H) curve and you don't reach saturation (the magnetisation of your sample won't further increase with higher external fields) you are measuring a minor loop (when you can saturate the sample you measure a major loop). In case of minor loops the sample will develop a memory of the magnetic history as stated. Usually you want to avoid this. In order to delete this memory the safest way is to heat the sample above Tc to destroy "all magnetism". Another possibility to achieve this is by applying an AC field (the field must be high enough to overcome the coercivity of your sample) and steadily lower the amplitude of the AC field.

You are right assuming errors when trying to calculate the magnetization of the film starting from bulk values (density will change due to approx. 2D system and mainly composition in most cases is different from bulk besides some other factors as well). But as a rough guess it will be fine just to get an idea in which range the film magnetization will be. Then you have an idea what kind of magnetometer would make sense to use concerning sensitivity to be able to measure the low moment of the films.

I just had a look at the first paper. In this case they obtain powders which means you have a lot of randomly oriented crystallites so it doesn't make sense to discuss magnetization with respect to crystallographic directions. This is different in case of films. Here you usually grow the film along a specific direction which gives you a strong dependence on the direction of the applied magnetic field (assuming your films don't have very bad quality)

David,

just to make it very clear: the magnetometer measures a moment, not a magnetization. Determining the magnetization from the measured moment will be your task, and as I have tried to indicate, it need not be a simple one.

What the magnetometer does, is independent of whether your specimen exhibits (ferro-) magnetic order or not. Whatever is there will be measured. When the moment is too small, the signal will be buried in noise, when it's "too large", the electronics will saturate.

I think it is the latter aspect which led someone to quote an upper limit for the film thickness - but that evidently depends on the magnetization of the film material.

So you're left with two major issues, I think.

(1) The magnetometer will detect whatever is in the sampling volume to which it is sensitive. Telling apart the film contribution may represent a challenge. Besides the estimate for the saturation moment of the film, it might be wise to also estimate the substrate moment at some typical magnetic fields. Not only will the magnetometer have to be sensitive enough for your film moment, but also, it will have to provide enough accuracy to tell the two apart. This is likely to be a problem only in case the film moment were just a fraction of the substrate moment.

(2) In the OP you wrote "I am unfamiliar with how it specifically measures Ferromagnetism"

From what I wrote above, one could say that nothing is specific to measuring ferrromagnetism with the VSM. What is specific is the behavior of M(H,T) which follows from its being ferromagnet. And there is, halas, a huge playground for both, the ferromagnet (to behave) and the experimenter (to find out about that behavior).

In my answer, I tried to address a few points that might be of interest.

"magnetic history": you may know that in many cases a piece of magnetic material will not be found uniformly magnetized. The magnetization breaks up in domains instead. The reason is a gain in magnetostatic energy. How exactly that proceeds depends on specimen shape and the other energy terms besides magnetostatic energy. Clearly, above some external field, the magnet will essentially align completely. That means that the domain walls have disappeared. They can be trapped at defects, costing extra energy to remove them, there are different mechanisms of annihilation of domain walls etc. Likewise, when you reduce the large field, at some point, domain walls will be generated (they have to nucleate, which takes some mechanism to happen). All these aspects contribute to what your M(H) cycles look like.

Imagine measuring an M(H) loop only partially: you start from satuarion in one direction but decide to reverse the sign of field change before saturation was reached in the opposite direction. Stakes are high that the two branches will not be looking the same and that they will not exhibit symmetry with respect to the point with (H=0,M=0). That means, any M(H) circle needs to start from a well defined state to be itself well defined. Hence my allusion to demagnetization procedures or the restrictin to measuring the so-called major loops.

On the other hand, after starting from a well defined state of the sample, measurement sequences as the partial loops mentioned may actually be of great help in identifying the processes going on in the ferromagnet (if that interests you...)

There exists quite some literature about different kinds of protocols that have been invented for such purposes.

But then, you might be interested in other characteristics of the ferromagnet:

What is MS(T)?

What is the Curie temperature?

How does HC depend on temperature?

How do MR and MR/MS depend on temperature?

Are there changes with respect to the direction in which the field is applied (anisotropies, including shape anisotropy)?

Is there an effect of applying some electrical potentials (since you're looking into a multiferroic) --- but that is probably nontrivial to set up.

So, my answer would be (as before): How you measure a ferromagnet depends on what you want or need to find out.

The VSM detection principle is always the same. But the measurement protocols for finding out the different properties are not.

Hope this may halp a little.