nanoparticle size determination

[Simon]

Hi Guys, I copy and pasted this from Mengqiangs earlier thread about
multi-phase refinements. I think it is better if discussion on this
new topic be in this new thread with an appropriate title to it.
Please post responses to the particle size question as responses to
this post, and responses on the topic of multi-phase refinement on the
other.

Thanks,

S

Thanks to Tony and Chris for your reply. I have figured out how to
calculate phase contents.
Another question comes to me recently. I need to figure out the
particle size of some nanoscale minerals. I know that if a particle is
not at nanoscale, the spdiameter parameter will grow very big and not
reliable. But what is the maximum size of particles that can be
reliably determined by PDF spdiameter? Is it true that using PDF to
calculate bigger size of particles, e.g. 100nm, are less approporiate
than smaller size of particles , e.g. 5 nm? Thanks.

Best wishes,

Mengqiang Zhu

[Simon Billinge]

It is definitely true that the PDF will give more accurate sizes for
smaller particles, but where the cutoff occurs depends on the
resolution of your measurement. It also depends on your structural
model being correct.

1) instrument resolution issue:
Peaks in the PDF are attenuated by particle size effects and by
instrument resolution effects. The recommended approach is to have a
bulk material that has the same structure as your nanoparticle (as
near as possible). Measure it at roughly the same time, and with the
diffractometer in the same setup, as when you measure your NPs. User
PDFgui to refine the bulk data. Turn off spdiameter so there is no
particle size effect in there and refine sigma_Q which is a parameter
that is accounting for the instrument resolution. Then fix sigma and
refine your nanoparticle data, letting spdiameter refine. You should
get a good estimate of the NP size. However, if your instrument
resolution is very broad and your NPs are very big, then even this
approach runs into problems of loss of accuracy.

2) accuracy of the model issue:
If the structural model you are using is incorrect, then all bets are
off. The program will use spdiameter to damp out ripples that it
doesn't like at high-r and you will get a hugely incorrect NP size.
You can check for this by looking at a plot of the data-PDF with the
model on top. If at high-r there are clear ripples in the data but
not in the model you need to beware. These could be inter-particle
correlation peaks (and therefore OK) or they could be that your
structural model is not perfect and spdiameter has damped the model.
How do you know if the ripples are particle-particle correlations?
You don't! But if the particles are orientationally or spatially
disordered any inter-particle correlations are broad, much broader
than intra-particle correlations.

Hope it helps,

Simon

--
Prof. Simon Billinge
Applied Physics & Applied Mathematics
Columbia University
500 West 120th Street
Room 200 Mudd, MC 4701
New York, NY 10027
Tel: (212)-854-2918 (o) 851-7428 (lab)

Condensed Matter and Materials Science
Brookhaven National Laboratory
P.O. Box 5000
Upton, NY 11973-5000
(631)-344-5387

email: sb2896 at columbia dot edu
home: http://nirt.pa.msu.edu/

[Mengqiang Zhu]

Thanks, Simon.

 

 I just found that Chris already discussed spdiameter to determine particle size as follows.
 
"This depends on the size of the particles. If the range of the PDF is on the order of the particle size, you can use the spdiameter parameter to get an approximate particle size (which assumes the particle is a sphere). If the PDF range is much shorter than the particle diameter, you will probably get uncertain results, at best. Note that spdiameter is very correlated with qdamp, since they both attenuate the PDF. You'll need a good measurement of qdamp in order to get a meaningful for spdiameter."
 
The Q space in my data is up to 18 k and I trunct  PDF data to 20~25 aus. for fitting. In terms of this PDF data range, what is particle size range that can be certainly determined from spdiameter? Thanks.
 
Best wishes,
 
Mengqiang

On Tue, Oct 27, 2009 at 4:10 PM, Simon <simon.b...@gmail.com> wrote:

[chris farrow]

Hi Mengqiang,

On Tue, Oct 27, 2009 at 4:13 PM, Mengqiang Zhu <zhu...@gmail.com> wrote:

Thanks, Simon.

 

 I just found that Chris already discussed spdiameter to determine particle size as follows.
 
"This depends on the size of the particles. If the range of the PDF is on the order of the particle size, you can use the spdiameter parameter to get an approximate particle size (which assumes the particle is a sphere). If the PDF range is much shorter than the particle diameter, you will probably get uncertain results, at best. Note that spdiameter is very correlated with qdamp, since they both attenuate the PDF. You'll need a good measurement of qdamp in order to get a meaningful for spdiameter."
 
The Q space in my data is up to 18 k and I trunct  PDF data to 20~25 aus. for fitting. In terms of this PDF data range, what is particle size range that can be certainly determined from spdiameter? Thanks.
 

I don't think I can give you a straight answer to this question. There are many factors involved with size determination, as Simon mentioned.

If you need this information for experiment planning, perhaps use PDFgui to create some hypothetical data from a proposed structure model and then fit to it. (Create a calculation, export the calculated signal, and then reimport that as a data set for another fit.) The results panel of your fit will give the uncertainty on the spdiameter determined in this way, which is a lower bound on the uncertainty one might expect from real data. For example. I just created fake nickel nanoparticle data (spdiameter = 200) from the PDFgui nickel example. I fit that data out to 20 A and found that the refined Spdiameter was 200 +/- 34. A similar exercise might help you find the answers you need.

I hope this helps. Good luck, and please let us know if you find anything interesting!

Cheers,

Chris

[Chris Farrow]

Hi again,

I forgot to mention that your fabricated data needs reasonable
uncertainty values to get reasonable bounds on spdiameter. (This is
one of those many factors in determining nanoparticle size.) When I
run the example above using an uncertainty of about 0.1 on the data
points (I had to add an extra column to the data by hand), then I get
spdiameter = 200 +/- 12. Anyway, still a useful exercise if you know
the relevant characteristics of your model and data.

Chris

[Mengqiang Zhu]

Thank Simon and Chris for answering my questions and sorry for the late response.

Recently, I took some HR-TEM images on my material and found some contradiction with PDF results

 

1) HR-TEM indicates my material is a mixture of amorphous area and crystalline domains and the former looks predominant. Because the structures are different of the two part, as I understand, for an ideal fitting, we should use two structural models for PDF fitting, one accounting for amorphous area and the other for crystalline domains. However, my PDF fitting was fairly decent (RW=30%) with only one crystalline phase. How does it come? Does it mean that amorphous area indicated by HR-TEM is not really amorphous?   Is that true, as long as its PDF can be fitted with a crystal structural model, it should be nanoscale and poly-crystalline?

 

 

2) Does particle size determined by PDF fitting with a crystal structural model account for the size of crystalline domain or a whole particle?

 

3) When the particle size determined by HR-TEM does not agree with PDF, which one should I believe?

 

Thanks.

 

Best wishes,

 

Mengqiang Zhu

 

 

 

 

 

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "diffpy-users" group.
To post to this group, send email to diffpy...@googlegroups.com
To unsubscribe from this group, send email to diffpy-users...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/diffpy-users?hl=en
-~----------~----~----~----~------~----~------~--~---

[Beatriz Moreno]

Hi,

Please, for nanoparticles with needle shape and defined orientation along the main axis, is there some way to model/fit, other than using spdiameter?

Kind regards, Beatriz.

[Simon]

It depends a bit how large the NPs are, but if they are small enough you should use SrFit and the DebyeCalculator

S

Sent from my iPhone

--

You received this message because you are subscribed to the Google Groups "diffpy-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to diffpy-users...@googlegroups.com.
To post to this group, send email to diffpy...@googlegroups.com.
Visit this group at http://groups.google.com/group/diffpy-users.
For more options, visit https://groups.google.com/d/optout.

[Beatriz Moreno]

Many thanks!

Beatriz

[Simon Billinge]

THere are some tools called "cookie cutters" which may be used to create the object by chopping it out from a bulk crystalline structure, then building fitRecipe using the Debye Calculator.  Hopefully there are enough examples in the Diffpy-CMI Exchange for you to get going with this, but if not, please let us know!  Also, if you have a good template for something, it would be great if you could share it back into the Exchange for others to use.  We can help with that if and when the time comes.....

S

--

----

Prof. Simon Billinge

Applied Physics & Applied Mathematics
Columbia University
500 West 120th Street
Room 200 Mudd, MC 4701
New York, NY 10027
Tel: (212)-854-2918 (o) 851-7428 (lab)

Condensed Matter Physics and Materials Science Dept.

Brookhaven National Laboratory
P.O. Box 5000
Upton, NY 11973-5000

(631)-344-3594

email: sb2896 at columbia dot edu

home: http://bgsite.apam.columbia.edu

[Beatriz Moreno]

Thanks! I'm going after the examples and the cookie cutters :-)

Beatriz