Re: [Ifeffit] particle size and coordination numberin Runanoparticles on carbon

Yes, Marian, that makes sense. Anatoly Frenkel also has a number of good papers based on this kind of approach. While the approaches differ in detail, they are all based on using the ratio of CNs, and thus avoiding the issue of having to extract S02 from a standard. --Scott Calvin Sarah Lawrence College On Oct 11, 2011, at 10:25 AM, Dreher Marian wrote:

Hi Scott,

thanks for the suggestion.

I've looked through your recent publications and also read your post from Fri, 04 Feb 2011, concerning the R dependence of s02. Comparing the CNs of the 1st and 2nd shell (that are the only ones that are properly resolved in my spectra) makes a lot of sense. I also read the following publication:

Jentys, A. Physical Chemistry Chemical Physics 1999, 1, 4059-4063.

Basically, I could use their formula to calculate the theoretical ratio of 1st and 2nd shell CN at a given cluster size and compare that to my data. Their calculations are based on a fcc structure but it should be a good enough approximation for my hcp Ru clusters. Does that sound reasonable?

Regards,

Marian.

Marian Dreher, Dipl. Chem. PhD Student, Catalytic Process Engineering Group

Paul Scherrer Institut General Energy Research (ENE) Laboratory for Bioenergy and Catalysis Office: OVGA/115 CH-5232 Villigen PSI Switzerland

Phone: +41 (0)56 310 2937 email: marian.dreher@psi.ch

-----Original Message----- From: ifeffit-bounces@millenia.cars.aps.anl.gov [mailto:ifeffit-bounces@millenia.cars.aps.anl.gov] On Behalf Of Scott Calvin Sent: Dienstag, 11. Oktober 2011 15:04 To: XAFS Analysis using Ifeffit Subject: Re: [Ifeffit] particle size and coordination numberin Runanoparticles on carbon

Hi Marian,

Have you tried what I suggested in the sections you quote below? With a ruthenium metal catalyst, you can probably see some second and perhaps third shell neighbors. The ratio between the coordination numbers for the shells is a measure of particle size that is independent of S02, and thus might (depending on data quality) provide a more reliable estimate of particle size.

--Scott Calvin Sarah Lawrence College

On Oct 11, 2011, at 8:54 AM, Dreher Marian wrote:

Dear Ifeffit Community,

I have a question regarding the signal damping I mentioned in my last post (I searched the archive but didn't find an answer. If this topic has been discussed before, please just send me the link).

As I mentioned before, I'm trying to determine the particle size of a Ru catalyst. I have 3 different reference spectra from Ru that I can use to determine the s0^2 for bulk Ru. From that I can calculate the Ru-Ru CN in my catalyst.

Now, two of my references give a very similar EXAFS signal in terms of amplitude. When fitting the first Ru-Ru shell I get a s0^2 of 0.6 to 0.8, depending on k-weight. One of the reference spectra was obtained from a pellet containing Ru powder ("Ru pellet Sigma Aldrich"), the other from a thick Ru filter ("Ru filter foil").

I also have a reference spectrum from an unknown Ru sample. The amplitude of the EXAFS is much higher compared to the other two spectra. However, the edge step is considerably smaller. Here, a Ru-Ru first shell fit gives me a s0^2 of 0.9 to 1.1, depending on k-weight. Even before fitting, the experimental data is very close to the theoretical spectrum.

Obviously, the CN I calculate for my Ru catalyst depends on these reference values for s0^2.

I can see how poor sample preparation leads to signal damping and thus to a low amplitude in the EXAFS, but I can't imagine how it could cause an amplitude which is too high. Should I therefore take the unknown reference since it has a higher amplitude?

I attached an Athena project file with all 3 spectra.

Many thanks,

Marian.

________________________________ From: ifeffit-bounces@millenia.cars.aps.anl.gov [mailto:ifeffit-bounces@millenia.cars.aps.anl.gov] On Behalf Of Dreher Marian Sent: Mittwoch, 21. September 2011 21:07 To: XAFS Analysis using Ifeffit Subject: Re: [Ifeffit] particle size and coordination number in Runanoparticles on carbon

Thanks everyone for your help, now everything is working fine. My Ru reference was simply way too thick which caused a lot of dampening in the signal.

From: ifeffit-bounces@millenia.cars.aps.anl.gov [mailto:ifeffit-bounces@millenia.cars.aps.anl.gov] On Behalf Of Wayne W Lukens Jr Sent: Dienstag, 20. September 2011 20:26 To: XAFS Analysis using Ifeffit Subject: Re: [Ifeffit] particle size and coordination number in Ru nanoparticles on carbon

Hi Marian,

There is a very nice paper by Scott Calvin that describes this in some detail:

J. Appl. Phys. 94:778-83 (2003)

Sincerely,

Wayne Lukens

On Tue, Sep 20, 2011 at 11:04 AM, Scott Calvin mailto:scalvin@sarahlawrence.edu> wrote: Hi Marian,

My big tip is to look at the second nearest-neighbor CN as well, and even further out if you can get it. It's sometimes hard to pin down the near-neighbor coordination number because of issues like the one you describe. But the ratio of higher CNs to near-neighbor CN is quite diagnostic of small nanoparticles. (Actually, for reasons too lengthy to get in to at the moment, this method does tend to produce results a bit biased to the small side, so it's best to try the identical fitting model on a bulk standard for comparison.)

--Scott Calvin Sarah Lawrence College

On Sep 20, 2011, at 1:50 PM, Dreher Marian wrote:

Dear Ifeffit Community,

I am currently trying to extract the particle size of Ru supported on carbon from EXAFS data. I also recorded spectra of a Ru foil as a bulk reference sample. So, my approach is to fit the first coordination shell of my bulk reference, setting CN to 12. From that I get a value for s02. I'm then using this value in the first shell fit for the Ru nanoparticles in order to get the coordination number.

However, the fit suggests that CN is still pretty much 12. What is changing drastically, though, is sigma^2 which usually doubles in value compared to the bulk sample. That makes sense, I guess, since there might be higher disorder in the nanoparticles.

Obviously, there is a chance that my nanoparticles are rather big, in which case CN would be close to 12. However, STEM and HRTEM pictures suggest that the Ru particles are between 1 and 2 nm in diameter.

I'm just starting to get into EXAFS analysis and pretty clueless right now about how to approach this.

Happy about any input,

many thanks,

Marian.

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit

_______________________________________________ Ifeffit mailing list Ifeffit@millenia.cars.aps.anl.gov http://millenia.cars.aps.anl.gov/mailman/listinfo/ifeffit