Is anyone here using the "NIST SANS data analysis program" written in Igor?
Igor_user
Thu, 02/09/2012 - 05:49 pm
I'm currently using the above mentioned program for analysing some data I acquired from small-angle scattering experiments.
It's a free program downloadable from NIST website and fully compatible with Igor.
So, I'd like to ask if here is any small-angle scattering experts who is familiar with this program.
Best Wishes,
February 9, 2012 at 06:05 pm - Permalink
February 9, 2012 at 06:37 pm - Permalink
Because I'm new to small-angle scattering, I fitted my experiment data using only the form factor model (a list of which is included in the NIST suite), which describes the properties of the scattering objects.
I'm still figuring out how I can extract further information like their inter-particle spacing by combining what I already got with a structure factor.
This is the manual I'm reading:
ftp://ftp.ncnr.nist.gov/pub/sans/kline/Download/SANS_Model_Docs_v4.10.p…
February 9, 2012 at 07:22 pm - Permalink
February 9, 2012 at 10:33 pm - Permalink
I'm at melbourne university. I've been to ANSTO several times but I got my data from Korea and Taiwan.
Do you also work in the field of SAS?
The whole datasets are loaded into the NIST program as an Igorfile. Please let me know if you want to take a look of that.
February 9, 2012 at 10:58 pm - Permalink
February 10, 2012 at 12:36 am - Permalink
Do you know anything of how to combine the form factor, which fitted your data, with a structure factor that could provide inter-particle spacing information when there are interactions between the particles, and how to obtain information about the particle size distribution? I've heard of some functions like global unified method or the inverse fourier method, but i don't know how to use it, and they are not described in the manual either.
February 14, 2012 at 08:12 pm - Permalink
February 14, 2012 at 11:37 pm - Permalink
ok, so i assume i have to find a combined one which fits my data in order to get the information that are provided from the structure factor?
February 15, 2012 at 08:40 pm - Permalink
Hi Andy
I have two reduced datafiles of two Ti-based materials of similar structure. We know their structures:
the 1st one has a random distribution of polydisperse ellipsoid like Ti-Mo particles
the 2nd one has also ellipsoid like, but more uniformly distributed, Ti-Mo particles.
I am using the model list that includes both form factor and structure factor in this NIST program to carry out the analysis.
Our goal is to obtain some information like the size, shape & inter-particle spacing of these particles. But there is no suitable model that fits our data for these two materials.
Can please advise if there is a way to edit or create our own model for fitting our experimental curves?
I have attached the two files here for your reference.
Regards,
February 28, 2012 at 06:36 pm - Permalink
Dear Igor_user, there is always some way to analyse your data - the NIST stuff is quite powerful. What you wish to do is quite within reason, but you will have to experiment. First, try model independent methods: are there any power law decays (look on a log log plot, this can tell you about the geometry of the system), do a guinier and porod plot, etc. You could use GNOM to examine particle shape.
Then you go and try analytic models. I know for a fact that you can use the NIST code to describe ellipsoids.
apart from this I can't help you much more, the drain on my time would be too big.
regards,
Andrew.
February 29, 2012 at 02:52 pm - Permalink
April 19, 2012 at 10:05 pm - Permalink
The table is embedded in a control panel so you can't export it as a picture. That would not be a great way to export to Word anyway.
I recommend that you just select the cells and choose Edit->Copy. Then create a Word table in Word. Then you have all the formatting flexibility that Word provides. You will have to add the column headers.
Alternatively you could create a standalone table and then use File->Save Table Copy. To create the standalone table, create a recreation macro from the panel (Windows->Control->Window Control), edit the recreation macro to remove all the panel stuff leaving just the table stuff, and run it to create the standalone table. Then choose File->Save Table Copy, import into Word and convert to a Word table.
April 19, 2012 at 10:42 pm - Permalink
I used the GuinierPorod model which fits well to my SAXS data. Initially, I got a value of Rg=70A which reflects the object size. But it's quite small compared to electron microscope (Rg=104A).
But I continued to reduce the 'chisq' value for the fitting of 'SAXS_txt' by clicking the "Do fit" button, the Rg was eventually refined from 70A to114A, which is a huge difference before & after refinement.
Actually, to what level of refinement can we say that the results from fitting are reasonably reliable?
April 25, 2012 at 01:34 am - Permalink
THe best way of getting Rg is to plot ln(I) vs Q^2. _Iff_ the plot at low Q is linear, and you are in the regime QRg < 1, then the radius of gyration may be determined. The slope of the plot for QRg<1 should be linear and have gradient -Rg/3. This is how you should determine Rg. Don't forget to transform the error bars when you calculate ln(I). A linear fit is easy, but make sure you only fit the QRg<1 region. For your data this region is probably 0 < Q < 0.01, if the Rg is ~100.
April 26, 2012 at 02:26 am - Permalink
Finding Rg is ok.
But do you know how to also add the structure factor that could give the inter-particle distance?
I selected the model which fits my data with the '_Sq' after it. But there are lots of them (attached), and it's not clear what each of these stands for.
May 6, 2012 at 10:08 pm - Permalink