[Ifeffit] EXAFS reference materials

[Scott Calvin]

Hi Ruoshuang,

Following up on Anatoly's excellent post, I would say that I 
personally distinguish between a "reference" and a "standard."

When I talk about a reference, I usually mean something I'm measuring 
simultaneously with the sample. The primary purpose is energy 
alignment, although it also gives a check on beam quality issues 
(e.g. harmonics). For this purpose, I generally use whatever is 
convenient and well known, such as a metal foil. Sometimes I'll 
measure the reference by itself, particularly if it's a beamline or 
edge that is new to me.

On the other hand, I think of a standard as a known compound that is 
similar to my sample. I often work with nanoparticles, so my standard 
might be the corresponding bulk material. Or if I'm working with a 
doped compound, the standard might be the undoped version. The 
standard is measured under the same conditions as the sample...in 
fact, that means that I measure the standard with a reference in 
place, if that makes sense. :)

So I use a reference for Anatoly's purpose (1) and a standard for 
Anatoly's purposes (2) and (3).

I hope that's helpful.

--Scott Calvin
Sarah Lawrence College

At 03:28 PM 3/22/2007, you wrote:
>Hi Ruoshuang,
>
>I am cc'ing this email to IFEFFIT mailing list. There are several 
>reasons (I am listing just a few) you should measure a reference 
>sample simultaneously with unknown materials.
>
>1) to align the unknown sample scans prior to merging. If those data 
>are noisy, reference sample can be used for the alignment. In 
>addition, if you want to compare samples before and after some 
>changes that may affect their energy positions (e.g., oxidation, 
>reduction, doping, metal-insulator transition, etc.) you want to 
>align all of their references to reliably analyze these changes.
>
>For this purpose, it does not matter whether you use a metal or a 
>metal oxide or any other standard with the same resonant element.
>
>2) to find S02 value. It is usually the same in most samples which 
>have the same resonant element.
>
>For this purpose, what matters the most is whether you can reliably 
>find S02 from your standard (and then impose it on your unknown 
>sample). Thus, some exotic reference which may still be good for 1) 
>may not be as good for 2) since it may be difficult to refine its 
>structure and get precise value for S02. The metals would be the 
>best, but if you have other materials with "open lattices", e.g., 
>fcc, rock salt, etc., with well defined first shell that you can fit 
>with the miniumum parameters, they will all work fine.
>
>3) to use as a model for your analysis.
>
>If you analyze an alloy, as in your case, you want to compare EXAFS 
>results in your alloys against those in relevant pure compounds. For 
>binary alloys, the best reference will be a pure metal, for ternary 
>- an intermetallic compound, etc. An oxide would not be so useful 
>for modeling. For that reason, people often measure more than one 
>reference, as per 1),2) and 3).
>
>