Varun,

Sorry if my message will appear quite harsh, but I think the answer is already in your question. If you know what Rietveld refinement is, how to use it, what are the requirements and limits, then you would have not placed this question. Otherwise it is definitely time to learn it first (as too many times I see it used as a plug and play black box) or to ask for help or collaboration on it.

Topas can be very dangerous as it is fast, relatively friendly and always give you a solution (even a non physical one). This is the reason why school on this are often organized (I organized one this year both on EXPO and Topas).

The suggestion of using EXPO is even more dangerous, especially if you don't know what you're doing (and again you better know what direct methods are before taking a software, providing a pattern and obtaining some structural data without knowing how to check if it has some sense or not). My suggestion is ALWAYS use a tool once you what is useful for. Otherwise you risk to reinvent hot water. EXPO is definitely not a first choice when analyzing a diffraction pattern.

Try first to use the knowledge on the material/specimen, your brain and a good search match tool to see what's in the pattern if you don't know anything about it. Unless you are doing frontier synthesis or you are sure you're working with something new, It is quite unlikely that you don't get any hit by doing a search match. And by adding some chemical constraints and relaxing some of the instrumental ones you can usually get some idea of what's inside.

In this case you talk about "Fe-Co-Cr/Mn system". What do you mean? a steel? In this case it is unlikely that you find a hit as anyone working on those materials can already give you the answer by heart. In any database you can easily find alpha-Fe, whereas austenite (gamma-Fe) is harder to find if you don't know where to look for (you can find it looking for a FeNiCr stainless steel). Both alpha and gamma structures are quite elementary (im-3m and fm-3m respectively, atom in the origin). The doping elements are usually substitutional for low concentration. Co can be tricky as it is hcp and would probably lead to a second phase. Plus if you do some heat treatment you can find some martensite and the situation can be more complex....

If you have an idea of the structure in similar materials (e.g. your material is the same structure-type as another material), search the ICSD (subscription) or mineral database (http://rruff.geo.arizona.edu/AMS/amcsd.php, free) and use those atomic positions and cell constants as your starting parameters. It should work reasonably well for an initial guess, which you can refine against your data.

If you do not have any idea about space group and unit cell parameters, use EXPO2009.

Dear Michael and Mona

Thanks for your reply

I am actually relatively new for rietveld refinements. I can do it in very easy way for well known materials/spinals by finding the ICSD data with the help of "FIND IT" and insert this data to "TOPAS". But , There is no ICSD data for some materials (like Fe-Co-Cr/Mn system). Is any other software in which we can refine this type of system..?

How EXPO2009 work.. Sorry to bother you by asking many question.

For EXPO2009, you only need the XRD pattern. The software tells you about possible space groups and cell parameters (usually more than one). Choose the best suggested parameters and use them as initial input for your TOPAS refinement.

Varun,

Sorry if my message will appear quite harsh, but I think the answer is already in your question. If you know what Rietveld refinement is, how to use it, what are the requirements and limits, then you would have not placed this question. Otherwise it is definitely time to learn it first (as too many times I see it used as a plug and play black box) or to ask for help or collaboration on it.

Topas can be very dangerous as it is fast, relatively friendly and always give you a solution (even a non physical one). This is the reason why school on this are often organized (I organized one this year both on EXPO and Topas).

The suggestion of using EXPO is even more dangerous, especially if you don't know what you're doing (and again you better know what direct methods are before taking a software, providing a pattern and obtaining some structural data without knowing how to check if it has some sense or not). My suggestion is ALWAYS use a tool once you what is useful for. Otherwise you risk to reinvent hot water. EXPO is definitely not a first choice when analyzing a diffraction pattern.

Try first to use the knowledge on the material/specimen, your brain and a good search match tool to see what's in the pattern if you don't know anything about it. Unless you are doing frontier synthesis or you are sure you're working with something new, It is quite unlikely that you don't get any hit by doing a search match. And by adding some chemical constraints and relaxing some of the instrumental ones you can usually get some idea of what's inside.

In this case you talk about "Fe-Co-Cr/Mn system". What do you mean? a steel? In this case it is unlikely that you find a hit as anyone working on those materials can already give you the answer by heart. In any database you can easily find alpha-Fe, whereas austenite (gamma-Fe) is harder to find if you don't know where to look for (you can find it looking for a FeNiCr stainless steel). Both alpha and gamma structures are quite elementary (im-3m and fm-3m respectively, atom in the origin). The doping elements are usually substitutional for low concentration. Co can be tricky as it is hcp and would probably lead to a second phase. Plus if you do some heat treatment you can find some martensite and the situation can be more complex....

Rietveld refinement needs some a priori information (because is a refinement of a model): the space group of your compound; approximate values of the cell parameters (a,b,c, alpha, beta, gamma); and positions of the atoms. Other parameters like instrumental peak width, zero of the diffractometer, Debye-waller factor,... will help to get a better solution.

Now we suppose that the main information is not available; you can still do something: to get the space group, you can index the reflections; to obtain the cell parameters, you can use Le Bail's method; and to figure out the atoms sites, you can use simulated annealing. FullProf Suite can handle with those methods

http://www.ill.eu/sites/fullprof/php/downloads.html

You can find more programs in

http://www.ccp14.ac.uk/solution/indexing/

Finally, you will find an example of solving a structure from a powder diffraction pattern in

www.epcryst.com/archives/Tutorial_La2CuO4.pdf

(epcryst is a program that generates trial structure models; I have never use it, but it seems interesting)

hi Pablo,

can you tell me the steps in Fullprof suite for rietveld refiment ? I have all the crystal structure oriented information of my material but i dont how Fullprof works.

hi shiddaling

if you know the space group, then see the International table for crystallography for atomic positions

guys, sorry if I keep insisting... but would you take a scalpel and try open-heart surgery on him? it seems many people on this site would do...

At my times people were used to attend school on this stuff and learn the hard way (litle to no internet available)and to ask only questions of a reasonably high level (nobody wants to look dumb, right?)...

Anyways if you really want to check how fullprof works, what about typing "fullprof tutorial" in google?

http://www.ill.eu/sites/fullprof/php/tutorials.html

i know getting through them takes longer than waiting for someone to do it for you... but what happens if some joking guy posts here something reasonable but wrong? how would you guess?

"a bisturi" What's that Leoni? Does it hurt? Do I just drop the word in Google? Its a lot more fun hearing from you. Sinks in better! Your students must luv ur style.

"At my times" What time was it Leoni? Wonder if it was the same as mine? Because way then "no internet" as well. I was dragged in kicking & screaming even though I taught FORTRAN back then. Have you ever seen the IBM punch cards? BTW this stuff never sunk in to me way back then. Internet really helped!

Only dumb question I know is the one I never asked. I still have a bunch left. The more I ask, the more I need to ask. "Ask and yee shall receive" is most appropriate for knowledge. Some of the younger have no patience to read. Info these days is available "on demand". The key is knowing which tree has the right fruit (which market has the right fruit, for this case). Thanks for sharing the fruit!

" if some joking guy posts here something reasonable but wrong?" Won't last long. Will probably get torn up by "pit bulls" like me and you. That, I can envision easily based on RG participation so far.

lol.. stupid automatic correction set on Italian lol... it did not correct it for scalpel! Thanks Ravi.. fixed now!

I have seen (but fortunately never used) the punch cards but I still play also with fortran.

Talking about fruits, if you keep buying apples in the supermarket you will grow convinced that apples are produced there.. no apple trees, no people caring for the trees, no pollination no insects. So welcome the "ask and ye shall receive", but not with my apples, please.. let's contribute to reduce the number of people answering "in supermarkets" to the question "where do apples grow?"

Leoni! Bisturi=scalpel! Multo Graci!

I learnt a long while ago (since punch cards) that there would be no point in getting "stressed out" about things I have no control over. My kids taught me that when they were teenagers only a few years ago. My wife is still trying to teach me that. She isn't a teenager but still looks like one to me! This being a public forum where all your prior blogs are not being thoroughly read (except by me for sure) by participants, will leave you no option but be tolerant. At least here the participants have the excuse of not being well versed yet. However, as you have mentioned many time in the past, many seasoned practioners of XRD are still "groping in the dark" and making egregious (possibly erroneous) assumptions and publishing those results.(e.g., the Scherrer equation). So in this case I'd suggest being a little more tempered and chalk it to reiteration. Besides, not all of us may have had as good an instructor as you are.

"but not with my apples, please.. " You still seem to keep "giving" despite your protestation. Forget it! You can't help giving. That is my diagnosis.

I certainly empathize with you as I feel similarly about folks that are supposed to "know it all" in XRD but are still hanging on to the 0D point/scintillation detector after nearly a century, like a "baby hanging on to its pacifier".

These days, to enjoy apple cider or other apple products there doesn't seem to be any need to know the shape, size or color of the apple tree. For all intents and purposes it may have been "grown in the market"! In fact, today 90% (maybe more) of those buying apples from supermarkets have never ever seen an apple tree. We are lucky in NJ, we get to see the apple trees in orchards. No such luck in southern India.

Back to topic! Varun has asked a fundamental question that I do not know the complete answer to yet, but have gained a lot of insight from all your comments. I'm glad he had the courage to ask the question and I appreciate everyone's perspective. The reason for my interest is that such analyses have been a critical part of XRD for nearly a century. My intent is to thoroughly learn, use and apply these methods for real time 2D XRD data.

Dumb question 1: What is "CIF data file for a particular sample"?

Answer: "A file with the ending .cif is a Crystallographic Information File"

BTW come join us and learn: Friday August 30th, 2013 10AM EST! TAKE ADVANTAGE! "The 3rd Dimension & Beyond in XRD 2013" by Weiguang Guan of McMaster University! http://www.chemistry.mcmaster.ca/facilities/xray/221-max3d

http://www.linkedin.com/groupItem?view=&gid=2683600&item=268765048&type=member&commentID=159428290&trk=hb_ntf_COMMENTED_ON_GROUP_DISCUSSION_YOU_CREATED#commentID_159428290

Matteo and all of you! We need you there! 9:50am EST Friday 8/30/2013. Google Hangout!

We had a phenomenal LIVE preview at 4PM EST today with Dr. Weiguang Guan himself. Here are some still shots from the preview on Google Hangout. If you would like to attend, please include [email protected] in your Google Circle as Friend. This will allow me to invite you. Please send me a R.S.V.P [email protected]

A file with the ending .cif is a Crystallographic Information File, containing the lattice parameters and atomic positions. It can be used by various software packages to create a simulated diffraction pattern, which can then form the basis of Rietveld refinement.

.cif files may be created using one of a number of crystallographic packages, including VESTA and CrystalMaker. Alternatively, they may be downloaded from online databases.

Again, I want to reiterate the point that no software package and no input data will give you good results if you don't understand the process and know how to use it properly!

Thanks Philip!

I'll join my voice to those above who suggest that if you want to learn how to use a Rietveld refinement program, you will not find that answer here. You can get suggestions of which software may be best for you, and where to find it and find tutorials and literature on the method, but if you want to be able to use the method and get meaningful results you must either take one of the many courses available around the world and as workshops at many conferences, have someone who is an expert who is nearby and willing to spend the time to help you learn, or read up on the method, read the instructions for the software and go through multiple tutorials that are available which give you a sample data set and take you step by step through a fitting process.

You must be aware, though, that just because you have a good fit and low residual does not mean you have the correct solution. It is very easy to converge to a false minimum with such a large parameter space. You must always verify that you have found a global minimum by redoing the fit with starting parameters that vary in a number of different directions in parameter space. If they all converge to the same solution you are probably not at a local minimum.

Stats @ 08:50hrs EST 6-14-14, 10 contributors, 33 followers,

2 / 0 · 19 Answers · 1454 Views

Suggestions for Varun and others:

1. Add 13 more topics with large followings "up top".

2. Share the question on RG & LI.

3. Share with network & followers.

4. Add real data to enhance discussion.

5. Add critical comments and move the conversation to the next level.

Thanks matteo leoni

Thank you very much to all for a nice discussion 

Regards

Varun 

 please explain me what are the differences between "popa rules" and "anisotropy no rules" in Maud

Varun,

Look up the lattice parameters and crystallographic information for your sample in the literature; that should basically contain the information in the .cif files.

To all the other people, this is a community of researchers. If you don't have anything nice or useful to say, maybe don't say anything at all. Have some tact and/or manners.