[Ifeffit] Artemis fitting using FEFF paths plusexperimentalreferences

Matthew Marcus mamarcus at lbl.gov
Tue Sep 25 12:30:05 CDT 2007

The particles are ~8nm diameter, and the distances involved are 0.3-0.34nm, so that's only a 6% effect.  Also, the
fitted Fe-Fe coordinations vary with oxidation temperature, with the biggest deficits at the lowest temps.  The particles
actually hollow more at the higher temperature, so one would expect it to go the other way.  It actually makes some sense that
there be vacancies because these particles are known to become hollow via a Kirkendall mechanism, and the structure of maghemite
shows that some vacancies are possible even in equilibrium.

Naturally, I did do XANES linear-combo fits.  I assume that's what you meant?
----- Original Message ----- 
From: "Wayne Lukens" <WWLukens at lbl.gov>
To: "XAFS Analysis using Ifeffit" <ifeffit at millenia.cars.aps.anl.gov>
Sent: Tuesday, September 25, 2007 9:07 AM
Subject: Re: [Ifeffit] Artemis fitting using FEFF paths plusexperimentalreferences

> Hi Matthew,
> It sounds like you have a problem caused by the finite size of the
> nanoparticles, which results in lower than expected amplitude of
> the shells in comparison to the amplitude of the shells generated
> from the xtal structures of maghemite and magnetite.  If this is
> the case and you have a good idea how large the nanoparticles are,
> you can use Scott Calvin's formula for the amplitude of a scattering
> shell in a nanoparticle:
> APPLIED PHYSICS LETTERS 87 (23): Art. No. 233102 DEC 5 2005.
> This works very well for metallic nanoparticles, but I have never
> tried it on oxide nanoparticles.
> Maybe you could try using Scott's formula for the amplitudes and the
> distances from the crystal structures.  If I recall correctly, Scott
> uses 3 Debye-Waller parameters as variables: one each for the first
> two shells and one for all other shells.
> The other possibility is to fit the XANES and low k portion of the
> EXAFS, but I assume that you have already tried that.
> Sincerely,
> Wayne

More information about the Ifeffit mailing list