if you are in presence of completely unknown products , the unique way to detect two or more phases by XRPD is to analyze several preparations and see if a class of peaks change the relative intensity with respect to other peaks ( be careful of the powders preparation in order to avoid orientation effects). In the lucky event that a class of peaks are related to a known phase, described in the various crystallographic databank, you can separate them and isolate the unknown peaks.
Yes you need to perform a search match. You can choose between the free option using eg one of the free databases like the COD with a large unknown on the quality of the result or the presence of the phase in the database or you can use one of the pricey databases like the PDF 4+ of ICDD that have thoroughly checked entries. It largely depend on the purpose of your research and the critical nature of the result
It all depends on what the compounds are. If you do not have access to a legally obtained database and you don't want to use a pirated copy then an alternative would be to understand your reactions and purposefully synthesise or obtain known containments, side reaction products etc for the purpose of collecting their PXRD and creating your own database. I have built a personalised database using this method. You can then execute a search match function using your customised database.
I suggest the use of X'Pert Plus or X'Pert Highscore. These software programs are very helpful and easier for phase characterization based on XRD data.
You will have two sets of XRD peaks each set corresponds to one phase. If you are able to identify any one phase, you mix pure sample of that phase with your two phase unknown sample. Then you will find the intensity of the peak of the phase you mixed will increase relative to the unknown.
Depending on the starting elements/ compounds and also the synthesis procedure, you must get some idea from the literature what could be the end products. Then search for the possible phases of these materials from published literature and JCPDS database. The first three most intense peaks are the main guide for identifying the phases, in case of random orientation. For oriented or epitaxial growth and in case of severe overlapping, the task becomes difficult. In that case, you can apply other tools as well like TEM SAD or Raman Spectroscopy to compliment / supplement your findings.
You should make a list of all peaks, identify the known phase , the rest peaks belong to the unknown phase. Yet there is a strong possibility of peak overlap between the known and unknowm phase. Be careful then. And you can follow th ideas given by Jocelyne Bouzaid, Giovanni Giunchi and some others.
If you have reference samples, it is not a difficult question. On the other hand, if you have a quite unknown sample, some theoretical analysis may help. According to Bragg's law, sin(theta)=k*lambda/2/d, so the peaks of one phase can be collected in one series. This task is usually covered by the softwares. Another possibilty is to try other methods, e.g solid-state 13C-NMR, here the doubling of peaks proves that two phases are present.
I want to add "get a single crystal structure and simulate a powder pattern for comparison" but that would be mean. May I ask what type of materials you are playing with? Maybe we can point you to a database that might be useful.
The answer given by Matteo Leoni is for me the correct. I have always used PDF files, initially, now ICDD files. If the reflections of some of these phases are not indicated in this ICDD references, use ITO, TREOR or DICVOL indexing methods.
Sachin, your question is bit vague. Instead of everyone guessing whether your sample is oriented or not, you please specify the conditions or any treatment given to sample. In normal case, as Suchitra Sen mentioned first three most intense reflection's intensity and position give lot of information. Getting two phases present in a sample is very easy actually with ICDD itself. you know what have you added. Look for corresponding compounds and see the match for intensity and position. Most of the cases can be solved by this.
Must be identified by crystallographic methods (commercial -ICDD- or experimental related obtaining these materials) if observed reflections in the diffraction pattern corresponding to one or more crystalline phases.
The ICDD data preassume an ideal sample preparation. This is sometimes challenging if preferred orientation (texture) is formed which may change the expected intensities remarkably. Other disturbing effects are beta reflections for very strong peaks. There the use of monochromators are more beneficial. I also worked with an X-ray tube where the anode material (Cu) had a very small amount of Cr for increasing heat stabilization. Usually disturbing these mixed anodes may have certain advantages, cf.
Brückner, U.; Epishin, A. and Nolze, G.:Determination of the sign of the gamma/gamma'-misfit in nickel-base superalloys by the use of spectral impure {Cu(Cr)} radiation. Scripta mater. 36 (1997),1279--1282
Bottom-line! If you have two phases then you should be able to detect the difftraction pattern from both those phases. All reflections ('whole pattern")! The question of the incident beam wavelength dispersion which Gert raises is often neglected. So whether it contains K alpha 1, K Alpha 2, K Beta or an arbitrary convoluted spectrum of all, it behooves the experimentalist to record the characteristics of the incident beam by running "known" internal or external standards. This is often neglected in practice and contributes to the ambivalence of XRD analyses. As noted, for precise use of relative intensity, Bragg peak position, FWHM & profile shape data, contributing factors other than the sample Nano structure must be evaluated and deconvoluted. E.g., instrumental profile, preferred orientation etc. The conventional method of recording diffractograms using 0D "point counters" introduces another element of "spatial blindness" which is debilitating in the XRD lab.
I have to usually do a lot of "tooth pulling" to even find out if a Ni filter was used or not. Information like that should be a part of the data set just as dimensions are in an engineering drawing.
If the two phases present in a powder pattern (once the unit cell parameters and possible space group identified) can be refined by the Rietveld method (FULLPROF, XLENS, ...) is a good structural confirmation of its presence.
1. Please acknowledge the original author if you intend to use other's words as-is, because all of us are reading this and it just seems fair to avoid the appearance of plagiarism.
2. If you do use other's quotations then please use meaningful clips rather than entire paragraphs and comments. Simply regurgitating the exact same words of others in agreement is not really helpful to any of us in the discussion and certainly not edifying to one's intellect. Additionally, this will insure that the discussion doesn't get boring and folks don't have to negative vote your comment. If you do see the negative votes, don't be shy to delete the comment and eliminate the risk to your RG score. That shouldn't detract the mature audience here.
3.. Let's keep it nice and focused. Make your opinion known. If you like someone's comment do not hesitate to "up vote" with the same temerity as "down vote". I seldom use my "down vote" but will surely disclose my disagreement in content.
4. Most importantly I appreciate the question and I'm reading your answers and learning a lot. To some this may seem obvious but to me it is always new learning experience. Thanks to all contributors.
5. It is OK to learn and change. Stay open-minded all! Methods of articulation that you will learn from other experts in RG will make you a better author, undoubtedly. But learn to use your own words (instead of cut/paste entire discussion and words of others) to help you with publications autonomously.
I'm not RG thought police. Just my humble opinion.
I think we have enough conversation to find out the phases now. I am sure, Sachin would have got the answer now. Sachin you should react here to tell whether you have got the answer.
Hi S. El.Aleem: There are good answers in this debate. However one should be aware that in these debates one forget details. In articles and books related to the debate theme is the scientifically correct answer normally. These online are interesting but can not replace the published scientific results.
Ravi, I completely agree with you. I am really surprised to see how each and every word of your comment has been simply duplicated by another member of this site. There should be some kind of check on the posts.
Research methods always spend more or less time; but in 21 century crystallographic existing methodologies are faster than those used in the early 20th century. If you want is only use powder diagram to identify use ICDD methods.
Although it is easy to do it with search and match patterns but even though first you should have idea that which new may be produced during the reaction. If you don't guess it then its quite difficult and some time very confusing to find the exact phase by search and match method.
Hi Asif: Prior to any crystallographic analysis of phases using X-ray diffraction methods or identification with index cards of ICDD is important to have chemical information of the elements present in the crystalline solid.
Thanks for correction. Actually I also wanted to say the same which you have mentioned Mr. Jose. Everybody should know the chemical and elemental history of the material then he can go for the next step.
In the first use search match (softward EVA) to confirme the presense of one of the two phases, the pics wich are not assigned in the pattern of data base is about the other phase.