Both Sn and SnBi are Tetragonal. S.G. I41/amd. A rapidly solidified Bi-Sn alloy was produced and some SEM and EDX and DSC indicate the presence of SnBi but in XRD the peaks, due to Sn and SnBi, overlap.
This can occur when both the materials have the same crystal structure, which is indeed, and same lattice constant or at least lattice constants very very close to each other (d-spacing vs. 2-theta). Alternatively this takes place when the volume fraction of one of the phases is so low that it cannot be detected with the normal XRD Cu-source. One has to use high-energy Snychrotron radiation source to resolve such tiny tiny phases within the matrix. One can try with extensive Rietveld refinement to differentiate them from each other.
I agree with Steve and Jayram. Relative intensity should be different for Sn and SnBi. Cell parameters should also differ. My suggestion is to try to do a Rietveld analysis on a good pattern collected over a large Q range and see if it fits with Sn or with SnBi or with a mixture of two phases with slightly different cell parameters.
If you are in trouble cause both fit well or if the peaks are quite broad then you need to work with hard X-rays. Probably checking the very high angle or getting a PDF from that pattern can help you clearly separate the two.
Thank you very much for your contribution. My sample is in the form of metallic ribbon about 5 micron thick and 7 mm width rapidly solidified using melt-spinning technique. The structure of the alloy is phase mixture, the matrix is Bi in which a small particles of Sn and SnBi are dispersed. Of course the peaks due to SnBi are very small, and overlaped with that of Sn.You know one of the advantages of rapid solidification is the production of metastable crystaline phases and I want to prove that. I think the suggestion of Jayaram to use Snychrotron radiation will be useful. Attached the XRD of wheel side and the other side.
This can occur when both the materials have the same crystal structure, which is indeed, and same lattice constant or at least lattice constants very very close to each other (d-spacing vs. 2-theta). Alternatively this takes place when the volume fraction of one of the phases is so low that it cannot be detected with the normal XRD Cu-source. One has to use high-energy Snychrotron radiation source to resolve such tiny tiny phases within the matrix. One can try with extensive Rietveld refinement to differentiate them from each other.
Hi Tarek,
I also agree w/ Mateo, Steve and Jaran.
When I was doing my master thesis, I also had the same problem. I found many overlapped peaks in my x-ray data. However, we still can distinguish them, you might want to try Rietvield analysis as Mateo suggested, but it takes time, and not easy to do.
I did the simple analysis by using Renishaw Wire 2.0 software. Actually, this software is able to simulate the overlapped peaks for Raman data, nevertheless I also use this software to distinguish overlapped peaks on my x-ray data.
I know it may not as accurate as Rietvield, since x-ray data is provided in Lorentzian approach and Gaussian for Raman.
If your main goal is to distinguish the peaks, I encourage you to use this software, but if you intend to do more deep analysis such as crystalline size calculation, etc, this software might not be the best choice, but you still can get the tendency anyway.
A screenshot of my work can be seen here
http://imageshack.us/photo/my-images/20/peakssimulationviarenis.jpg/
Another idea is to take advantage of the difference in Z numbers for Sn (50) and Bi (83). Using a SEM to capture backscattered electrons might be useful in this case. Sn, Bi, and SnBi alloy ought to each have significant differences in contrast. Atoms with the highest Z are brighter. So I imagine Bi can be contrast adjusted to be light grey, SnBi, to be a medium grey, and Sn a dark grey. A picture is worth a thousand words.
X-ray core photoelctron spectroscopy (XPS) would be adequate.
You can also try to measure neutron diffraction intensities and refine the data with a program like Fullprof or GSAS using three phases Bi, Sn and SnBi. You may have have preferred orientation (ribbons) but these programs can trat preferred orientations as well. You should be able to get the relative amount of these phases ideally.
How about separating them chemically? Not being a chemist, I don't know if there are any acids or bases which could preferentially dissolve one or two of the phases, but this would be a very easy way to remove one phase to allow you to study the other in isolation.
Otherwise, due to your solidifcation approach, we can expect a strong crystallographic texture. In this case, by tilting the sample if various directions you should be able to isolate different peaks. The various phases can have different textures if the preferred growth direction on solidifcation is different. If they have the same crystal structure, however, then the growth directions may be the same and this won't help.
Lastly, the suggestion by Mr. Morales is almost certainly the most direct solution and would indeed give you beautiful pictures which offer the clearest proof.
best of luck,
Gijs
The best is to correlate with XPS on the one hand and, on the other hand, You can use the negative images from SEM (Please check the answer of Dr. Morales).
Surely, tin and bismuth do not form an intermetallic. At the eutectic composition for example (~62%Bi) you have a two phase alloy consisting of tin in bismuth solid solution and the bismuth in tin solid solution. If you look at the phase diagram the solubilities are not that high in either case.
The lattice parameter of the phases are different even that have the same crystal structures. So, the peaks will be separated at high angle of 2 Theta. Using different X-rays (wave length coming from different targets) will help you to differentiate between the two phases at high angles
There are several methods which may work; it all depends on the nature of the sample and the equipment that you have available. You can determine the information from the XRD pattern by using programs to deconvolute the peaks. But I do not think this is conclusive proof. It just backs up what you see in the SEM/EDX and DSC data. TEM can be a good way to go, but sample prep can be difficult and time-consuming. Using a different wavelength (or neutron scattering) could provide the answer. BUT, you should check the databases...do the peaks still overlap at these different wavelengths? Personally, I believe that XPS and/or ToF-SIMS would provide the most conclusive answers for this. These methods would provide pretty conclusive proof on the existence of the intermetallic, with minimal sample preparation.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science
- Material Characterization