# Magnetic units in journals

1. Apr 10, 2005

### nbo10

What is with magnetic units used in published papers? You can only compare ~1% of the published results. Some of the units I've seen emu, emu/g, Arb, a few don't even give units. How do these papers make it through the review process?

2. Apr 10, 2005

### dextercioby

It is surprising and for me annoying,indeed.These experimentalists should be using SI units.Good thing i'm not one of them or one interested in reading their works...

Daniel.

3. Apr 10, 2005

### ZapperZ

Staff Emeritus
The most common units that I have encountered in my line of work is "Gauss". So most of the people in my field publish using those units.

Keep in mind that it is VERY "culture" and field-of-study dependent. In some cases, the relative magnitude is the only thing that is meaningful, so sometime you see in it arbitrary units (arb). Giving exact value may either be not meaningful, or providing something the experiment cannot provide. So you really need to look at this case by case.

It would have helped if you cite specific examples so we have something more definite to go by, rather than just trying to explain this wholesale.

Zz.

4. Apr 10, 2005

### Gokul43201

Staff Emeritus
The cgs unit for magnetization is emu/cc which is 1000 A/m (in SI). If magnetization is specified in either emu/gm (or just emu), there will also be specification (or an implicit assumption of knowledge) of the sample density (or volume).

5. Apr 10, 2005

### Meir Achuz

One problem is that SI units are very awkward for describing magnetic phenomena.
They were adopted i a close vote by an international congress, but not by the people who were really caluclating or measuring things. People who do those two things still prefer to use unts more appropriate to the physics.

6. Apr 11, 2005

### nbo10

I don't have specific reference handy. I'm going back and looking at magnetization data of YBCO and how the YBCO was processed. Of course you can see the qualitative behavior, ie diamagnetic signal at 90K. But I want to compare the size of the signal for different processing. Only a few of the papers give all the information needed to make a quantitative statements.

7. Apr 11, 2005

### ZapperZ

Staff Emeritus
OK, so now we're getting to the specifics here.

To be fair, the size of the signal is often irrelevant. Unless you think the response is non-linear (which I don't think it is), the amplitude of the signal is not important. What is definitely important is the width of the transition. This is what characterize the degree of purity of the sample, for example. So in your fabrication process, this is what you want to pay attention to more than anything else, I would think....

I see the same thing when people do NMR measurement. The resonance curve often are simply quoted in terms of arbitrary amplitude, because it is the width of the resonance curve that usually is the important quantity. Only in the case where you have a non-linear response would there be a definite need to know the quantitative value.

Zz.

8. Apr 11, 2005

### Gokul43201

Staff Emeritus
This is also especially true in susceptibility measurements. You will often see $\chi$ plotted (vs. T) in arbitrary units. What's important is whether or not there is weirdness in the curve (a sharply peaked susceptibility, for instance), and where (at what T) it happens.

9. Apr 15, 2005

### shyboy

well, the size of the signal can be relevant. The question is what kind of method do you use to measure $$\chi$$. The simplest thing is to look for AC magnetization, but this is usually a very crude test and the interesting value here is indeed a transition width. Even if people give you the absolute values, they are usually of little use, because you may get the similar (but broader) curve from a good metal.

However the situation is different if you are looking on SQUID data. Now
that is the more close to the material property. Ideal superconductor should expell all the field, which means that its susceptibility is $$\chi=-1/4\pi$$ in the CGS units ($$B=H+4\pi M$$). You can also see the susceptibility per unit of mass, volume, or amount of the matter (per mole).
In SI the susceptibility of the ideal superconductor will be -1.

So if you are looking on the SQUID from ceramic, you may say how much superconducting material do you have. There are other method wich can give you the "real" DC $$\chi$$

First you need to figure out what