The best way is to add to your sample a certain amount of unambiguoully crystallized phase (let me say something like Si or NaCl), precisely weighed. Afterward you make a quantitative Rietveld refinement and if you have add 50% of cristallized Si, and you refine 70%, it means that you have some "x-ray amorphous phase". A simple rule of three will give you the amout of "invisible" phase.

Hope it will help...

The formula, you mentioned in your question, would give you an approximate value. I think you can also find out the fraction of crystalline phase from DSC (Differential Scanning Calorimetry).

BTW, your pattern doesn't seem to contain amorphous. In my opinion, fitting of pattern will not be stable or reliable when you have more background noise in your XRD pattern. You could try increasing the step size and time per step a little and see if you get a better pattern. I guess its a BCC pattern, probably Fe based, if its so try using monochromator at the detector side

Hi Praveen ! thank you for your answer.

My material is Fe base metallic glasses. I want to know the fraction crystalline phase created after heating by electron beam. I am not sure if DSC can use to calculate the volume fraction of crystalline phase in my situation ( the result is reliable if my sample is heating one more time under DSC).

Btw, my sample may become crystalline after heating.I show you the diffraction pattern of Fe based metallic glass before heating ( It shown amorphous).

One more question, how can I calculate the fraction of crystalline after fitting of pattern.

Thank you so much

If you would do it exactly you have to perform Rietveld in Highscore plus. Your scan is bad. If you have an amorpheous phase there should be a broad hill.

I suggest to repeat and to measure at least 24 h between 20-70°. Best results you will get with a parallel beam by a Goebel mirror. Then you can seperate Peak broadening from crystallite size effects.

The best way is to add to your sample a certain amount of unambiguoully crystallized phase (let me say something like Si or NaCl), precisely weighed. Afterward you make a quantitative Rietveld refinement and if you have add 50% of cristallized Si, and you refine 70%, it means that you have some "x-ray amorphous phase". A simple rule of three will give you the amout of "invisible" phase.

Hope it will help...

A very pratical approach but reasonable

Why would you calculate volume fraction?

Generally speaking I think that the method you plane to use will be not reliable enough, moreover the present experimental data are too poor in terms of statistics. Also it is somewhat difficult to separe what one can said amorphous, and nanocrystalline, and from nanocrystalline to µm crystallite size there are considerations to account in terms of scattering.

SEM and TEM can also be usefull for imaging the various size of crystallites and amorphous domains.

Since your material contains iron, you have a chance to consider your present problem in terms of 57Fe Mössbauer spectroscopy, where it could be expected to account for the experimental extend of the hyperfine field, EFG and Isomer shift typical parameters to be correlated and compared to the XRD analysis.

Also by using neutron diffraction, ones can compare the diffraction and "diffusion" contribution and relate to the true absortion of your materials established by cross counting.

I agree with Pascal Roussel, you would to add crystalline material, this is also known as the spiking method.

Additionally you get weight percent from XRD, so you would need to do a recalculation to get the volume percent:

1: first divide the weight percentages by the densities of each phase. You can get these for all the crystalline phases but not for the amorphous phase, so you would have to make a guess for that...

2: The numbers you get in volume does not sum to 100%, so you would then need to re-normalise, such that the sum gets 100%, by dividing each volume fraction by the total of the volume fractions. Note that the accuracy of this method depends on your assumptions on the density of the amorphous phase.

I also recommend that you get better data quality before you start calculation. I you can get chemical analysis of the total mixture and the crystalline and amorphous phase, then you can check if the element sums from the phase mixture matches your bulk chemistry.

have a look at this publication:

http://journals.cambridge.org/action/displayAbstract?fromPage=online&aid=8502525

I am not sure, whether the following suggestion would work out.

1) If you could make complete amorphous of the same composition by some other route, then do XRD on that and measure the FWHM of it.

2) Now using some software (Xpert High score plus), try to deconvolute your most intense peak (pattern from which you need to find out the fraction) in to two peaks.

For one of the peak, you fix the FWHM value, here you need to give the value of FWHM of complete amorphous of your alloy which you get it from the step 1. By doing so, we are assuming this peak to be coming from amorphous fraction of your material.

Then fit the pattern, now the software will fit the other peak accordingly and which will be the peak from crystalline fraction of your material.

3) Now you will have integrated area for two peaks and apply those values in the formula you have mentioned in your questions.

"Amorphous compounds are invisible for the Rietveld method. Only crystalline phases are taken into account and their sum is normalized to 100%. The amount of the crystalline phases is overestimated in case amorphous material is present too. This effect can actually be used to determine the amount of amorphous material. Once you added an internal standard to the sample (a known amount of a pure crystalline phase), the overestimation of this phase (and of all other phases) is obvious and can be corrected."

U can use the HSP autoscale analysis as the tutorial result below:

Many thank for your answers,

Let me explain a little bit about my experiment.

My sample is Iron based metallic glass ( 80% Fe, 13%B, 7% Si), 30 micron in thickness. This material is heated by electron beam, then is characterized by XRD. I would like to know the effect of electron beam on crystallization behavior of Iron based metallic glass. My results shown that Iron metallic glass was crystallized after heating. However, I do not know how many percentage of crystallinity presented in material.

I don't know why my scanning is so poor. What parameters do I need to change so that I can improve the diffraction pattern during XRD scanning.

How can I perform Rietveld method? Do you have any tutorials about that ? Can I use highscore software (not highscore plus) to do that?

Could you tell me if TEM can use to measure the fraction of crystallinity ?

Internal standard addition procedure.

Well-known and accepted method for any kind of analytical spectroscopy instrumentation. Also you can easily validate the methodology and prepare good standards.

You can see chapter 3 in EMU Notes of MIneralogy vol. 9 at the site

http://www.minersoc.org/EMU-notes-9-3.html

Yes, idealy u need to use a standard sampel.

What i told u is in HSP because i see ur data was in HSP plot.

First u need to have a fully crystalline material without size/strain broadening.

Let say Fe powder of microparticle > 1 micron. Refine and note the profiles and background. Mix ur amourphous Fe (let say 50:50) with the amourpohous one.

Use the same profile parameter and background to refine this one, and chek the weight percentage in the list. The same method u can use with the partial crystalline.

I think TEM analysis cant do that.

As u can see if there's a amourphous phase present than it will contribute to broadening in small angle as the bacgroun increase but not so significant at high angle

The reason you may have a noisy signal is that it is well documented that using Cu-k alpha in Fe leads to fluorescence. I had similar problems with FeGa, but you can either use a different k-alpha ie. Cobalt or Molybdenum instead, this should reduce the noise effectively.

The suggestion of mixing your sample 50/50 with that of a pure crystalline Fe sample seems very logical and that is the route I would choose as this method would eliminate any 'guesswork' involved.

I have a lot of intelligent comments to carefully review. I will, and then comment further.

Some initial observations:

1. "Volume fraction of crystalline phase= Areas of crystalline/ (Areas of crystalline + Area of amorphous)". I suggest replacing the "=" with "Proportional to" in the above equation. Also as Ferenc Kristály points out - is it volume fraction or weight fraction?

2. The use of "known standards" (internal if possible) would be extremely helpful in calibrating your method. I'd personally prefer an independent internal standard with diffraction peaks well separated from the target materials to avoid any interference. A single crystal "zero background" sample holder may serve both functions. Here is an example of how we used, in one instance, the polycrystalline Al sample holder as a calibration standard: https://www.researchgate.net/post/Is_this_a_reasonable_way_to_eliminate_effects_of_sample_surface_misalignment_with_the_diffractometer_axis_Theta-2Theta_Diffractograms

3. The sampling volume and its consistency is critical in quantitative analyses. Use of transmission mode XRD should also be considered besides the reflection mode. Pay attention to any potential "preferred orientation" in your sample.

4. Relative intensity must be normalized w.r.t. the total integrated area under the entire diffraction profile. In order to accomplish this successfully, the "back-ground" signal has to be unequivocally determined. It must also be minimized (fluorescence etc. as Dr. Chris Quinn points out). Individual constituent contributions to the diffraction profile must be carefully deconvoluted and then relative integrated intensity ratios determined to compute constituent compositions using "calibration charts" created from "known samples".

5. I'd encourage you to master the following book but meanwhile certainly read the excerpt included regarding the contrast between the ubiquitous 0D point counter and photographic film used in XRD for nearly a century now:

X-ray Diffraction In Crystals, Imperfect Crystals and Amorphous Bodies by Andre Guinier. 1962 - Great book! Worth acquiring for your XRD collection.

(pp. 162-163) "6.3.5. Comparison between the Photographic and Counter Methods

It is obvious that the Geiger Counter and the ionization chamber have the advantage of precision and sensitivity. They alone permit a quantitative study of scattering. However, point by point measurements in reciprocal space are lengthy, while the photographic method gives directly a picture of the scattering in a single experiment over a large surface of reciprocal space. The resolving power of the film is excellent. Thus, sharply defined scattering regions, for example on rows or on planes of reciprocal lattice, show up very clearly on photographic film but are difficult to detect with a counter.

Let us note also that the counter is "blind". In many cases inaccuracies in the setup produce stray scattering which is easy to detect on the photographic film but which can spoil a series of experiments made with a counter.

The two methods are therefore complimentary and it would not be advisable to neglect one in favor of the other. The photographic method is admirably suited to the qualitative exploration of an unknown pattern, since one can then find totally unexpected phenomena. The counter is necessary for quantitative measurements on a pattern which is already known qualitatively."

This was printed in 1962-1963. This has to be considered in light of the present day 2D detectors that combine the benefits of the counter & the film in one. These writings indicate possible causes of the present day inertia for moving from the 0D to the modern 2D detectors. The spatial "blindness" of the 0D scintillation counter/detector is the principal reason even experienced experts using XRD are ambivalent about the 0D XRD data/results.

http://books.google.com/books/about/X_Ray_Diffraction.html?id=vjyZFo5nGUoC

Nguyen Tung!

1. How deep do you estimate the laser heating will effect the 30um thick sample? Do you expect this to be a surface phenomenon or bulk?

2. Can you switch the incident beam wavelength to avoid fluorescence?

3. Do you have the ability to use the transmission XRD?

4. Do you have access to photographic film?

5. Is the packing density of your sample consistent? Is it powder or bulk glass?

Please post some photos of your sample and experimental set-up when convenient.

There are several aproches. I think that the best way is to let the software working and to controle the results. Yes, now it is possible to quantify the amorpous phases and XRD is able to mesure the crystallinity rate.

That's right Habib, as long as the assumptions used in the software are accurate. This usually is highly subjective and operator dependent. Hence, the ambiguity on occasions. The conventional 0D method (which I'm sure is the method presently being discussed) is inherently disadvantaged due to its insensitivity to preferred orientation, instrumental aberrations, sample morphology (packing density), sampling surface orientation/alignment, experimental optics/geometry, beam conditioning apparatus etc.

Sure, you are right. You should take a look in the ILL forum discussion. There are many intresting questions and more convincing answers about what you are asking.

What is ILL forum discussion?

Thank you very much for your answers,

My material is not powder !

My material is a thin film metallic glass ( 30 um).

I have 3 problems here

1) some of your comments suggest me to add a certain amount crystallized, precisely weight and then use the Rietveld refinement or internal standard methods to quantify the fraction of crystalline phase. Until now, I have no ideal with these method. However, I think these method apply for Powder material only, right ? Is it true for thin film material?

2) After XRD scanning, I can get three kinds of intensities " I obs - I cal - I back " from highscore software. I know that we can estimate the fraction of crystalline phase by using relative intensity, but I don't know the formula. One other suggestion is that I copy the data from highscore and paste to "excel", then we can use excel to calculate the fraction of crystalline phase. How can I do that ? Do you have any video or tutorials teach us about that ?

3) Let me talk about my material. After heating by electron beam, the melting deep of my sample is about 5 micron. Thus, I think the crystallized deep is less than 5 micron.

Can you tell me how deep the XRD can go inside the sample ? I believe that my sample is crystallized not only on the surface but also inside thin film.

I show you the SEM of material surface after heating.

@ Habit, could you tell me which software I can use ? are they free?

You should told in the first place, it has different technique to measure thin film such as XRR. Conventional Bragg-Brentano were not very helpful in thin film .

Maybe this article can help u understand:

At low angles the XRD penetration depth changes with the angle of incidence... See here for a succinct overview: http://mrl.illinois.edu/sites/default/files/pdfs/Workshop08_X-ray_Handouts.pdf

Your initial peak looks like a [110] for a bcc iron alloy, which has a lattice parameter of ~2.785 A, slightly less than that of alpha-Fe. This material crystallises around 345C but also has a BCT phase at ~500C. I have requested a paper that a colleague of mine wrote back in 1983 on amorphous FeBSi to crystalline FeBSi using a TEM with heat stage; I think you will find it useful.

I will upload the paper to this discussion when I receive it.

Regards,

Dr. Quinn.

@ Dr. Quinn, thank you for document, It very useful for me

@Nguyen, there are several softwares fou QPA. Some are free like GSAS, FULLPROF, ... and others are not. Actually I am using TOPAS (Bruker) not free.

Concerning the ILL (Institut Laue Langevin) forum, you can subscribe at [email protected] for free.

Your film is not thin at all, so XRR will not work.

It is not a powder either...

SEM shows peculiar periodic structure of your surface - it would be nice if you undestand it.

TEM can be very good if the grains are large enough.

XRD is good for the crystalline phase (not for amorphous). So if you had "in-situ XRD on a hot plate" you would be able to see the progress of crystallization. If at a certain point you crystallize it 100% - this could be your reference point. Still an imaging technique will be necessary to confirm the results of XRD.

Have fun.

Sergey has a good point, but I think TEM with hot stage is the way to accurately see the formation of crystallites, then you could confirm this using XRD and DSC.

Agree with Sergey... Forgot to pay attention to SEM image, is it the cross section of your sample? than it's not a thin, XRR work for about 0.1 - 1000 nm. Perhaps u grind your sample (500 - 1000 nm) and treat as powder diffraction.

OK! Time to summarize. I'll take the lead and be the target. Lol!

1. Nguyen! Your original diffractogram is quite unique in the sense that the number of crystalline peaks appear to be "one". Why so? What is the volume fraction of the crystalline phase? Why have you started your scan only at 20 deg.? Do you have diffractograms for similar material from literature? How reproducible is your initial diffractogram?

Compared to the wonderful XRD profiles illustrated by others your SNR (signal to noise ratio) is exceptionally low (euphemism for "sucks"). Fix this first! Determine the origin (likely culprit - fluorescence from constituents). Here is an actual example of a composite (RDX+Al+CAB binder+"Al foil standard") XRD Image in transmission Laue geometry compared with its conventional diffractogram: http://www.flickr.com/photos/85210325@N04/7988760159/

http://www.flickr.com/photos/85210325@N04/7988731098/in/photostream/

Here, fluorescence was not our challenge but the preponderance of the CAB amorphous binder in the composite was. We were more concerned with the crystalline constituents and their nanostructure. Even for a trained XRD specialist, interpreting the conventional diffractogram in this case would be challenging. No mater how many iterations using "fancy" algorithms and software that you diligently perform. Use film if you can, once you have resolved the "background" issue, in order to get a sense of "what's happening".

If crystallization is limited to 5um of the top, then the reflection mode would be more appropriate. Shallow incidence angle XRD may be an alternative.

2. Dr. Quinn! The link you posted is a fantastic XRD resource. Thanx! I'll share. I sure will take it up with Mario Sardela for keeping it a secret from us in our LinkedIn group "X-ray Diffraction Imaging for Materials Microstructural QC": http://www.linkedin.com/groupItem?view=&gid=2683600&type=member&item=168811359&qid=2f9f38e0-1a07-4168-a925-7de0af1a95dc&goback=%2Egna_2683600

3. Volker! For some of us it may take a few iterations. It took me decades to grasp!

@Sergey, you are right! my film is not thin, Therefore, I can not use TEM to check my sample( the thickness of sample should be less than 100nm for TEM viewing). In situ XRD or DSC is good to know the crystallization process. However, my target is to check the effect of Electron beam on the metallic glass film (for example, there is 30% of crystalline phase is created after electron heating) . If I use in situ XRD or DSC, it mean that the treated sample is heated again and more crystalline. How can I separate the effect of Electron beam heating and the heating process in DSC or in situ XRD on my sample ?

@Ravi Ananth, the original of my material is 100% amorphous in ribbon shape. I use the ion beam to heat the surface of material, then the heated material is characterized by XRD to know if material is still amorphous. In my case, this material may be partially crystallized. Thus, I think this is a reason why the diffraction pattern is shown only one peak. Assume that my material is partially crystallized, It may be 20%, 30% , or 80% crystalline phase. How can I know it from the diffractogram ?

Until now, I have one difficulty in my research. I do not know the process of crystallization during electron heating ( when is material crystallized ?).

Can be quantified crystalline solid mixture and the presence of amorphous phase using the FullProf or GSAS programs, among others.

Volker & J. Amigó are correct. TEM would take you (how many PhD human hours?) significantly longer time to get the exact result of a miniscule of the total sampling area unequivocally. Whereas, XRD would yield near real time Nanoscopic data at a spatial resolution approaching sub-micron range today for up to several inches in diameter. TEM must be used to calibrate the XRD method without any doubt. Without getting in to an advocacy of one or the other, my thought is, they complement each other exceptionally. There is little doubt regarding the efficacy of either method depending upon the ultimate goal. Nanoscopic XRD Imaging!

In my knowledge, based on the materials used for this discussion and the X-ray incident beam energy utilized, the probability of any phase changes including heating of the specimen being irradiated as a result of these two elements of the experiment would be infinitesimal. Almost out of breath with that past sentence! Heating due to XRD is a NO in this case is my opinion!

Nguyen Tung!

1. "the original of my material is 100% amorphous in ribbon shape." Get the diffraction pattern for the initial material before any "ion beam" exposure. Post it!

2. "I use the ion beam to heat the surface of material, then the heated material is characterized by XRD to know if material is still amorphous. In my case, this material may be partially crystallized. Thus, I think this is a reason why the diffraction pattern is shown only one peak. Assume that my material is partially crystallized, It may be 20%, 30% , or 80% crystalline phase."

If you are certain that the peak you see is all you need to establish the presence of crystallinity in the examined surface layer of the specimen (calculate the penetration depth of the X-rays), then use its position, peak height, FWHM, integrated area etc. to monitor the behavior with varying ion beam exposure. This will help verify your postulation.

You'll clearly be able to distinguish "20%, 30% , or 80% crystalline phase" through empirical XRD charts that you may calibrate with TEM.

Just remember, garbage in & garbage out. Meaning the validity and precision of the XRD method is only as good as the incoming data. So, high SNR is critical for precision! Your data lacks this presently. However, you'll still be able to use it empirically. Use photographic film and digitize with a scanner if needed.

3. "How can I know it from the diffractogram ?" By learning and understanding what others have done with film, scintillation counters, LPSD (1D) & 2D Area detectors in XRD for the past century successfully.

4. "Until now, I have one difficulty in my research." Be thankful! For once you have found the solution to this challenge, you'll find many more and your difficulties will be gone! You'll love the challenge! Lol!

5."I do not know the process of crystallization during electron heating ( when is material crystallized ?). " With XRD and the right imaging tools you may be able to watch such events in real time, in situ!

Whatever you find with XRD must be corroborated with TEM! Except the in situ part for now!

I forgot one thing, if look at ur data u must be use Cu source. Fe and Co absorb the Cu wavelength radiation, and then fluoresce that energy as their characteristic X-rays become background noise. Try change the source to Co. It also can increase the depth of penetration to ~30 microns. So at list can cover 2 - 3 layers of your sample.

@Ravi Ananth, thank you very much.

Regarding to your comment, I must combine XRD and TEM to get a full picture about crystallization. The problem is my sample thickness. I should make the irradiated sample become thinner before using TEM. I worry that there is a formation of crystalline phase during thinning process. Therefore, the result of TEM viewing is not correct anymore.

let talk about in situ XRD, I think in situ XRD is better used in case of original sample which is not heated before. The threshold of crystallization of material can be determined. However, if I use in situ XRD to characterize the irradiated sample, the threshold of crystallization will be different.

Btw, Do you have any tutorials about XRD or TEM analysis ?

If you use PIPS (Precision Ion Polishing) at very low angles and low energy, it would take several hours to thin the sample, but this in my opinion would be the best way to prep your TEM sample. I would ask someone experienced in TEM work to assist you here as it takes many years to be good on a TEM setup.

@Christopher, thank you. Hope I can find a good method to thin my sample

Nguyen! "However, if I use in situ XRD to characterize the irradiated sample, the threshold of crystallization will be different." Not sure what you mean by this?

If you mean, that the degree of crystallization may change as a function of depth, then that is a challenge. However, by using various incidence angles, you may be able to adjust the penetration depth of the incident beam. The absorption factor must be accounted for appropriately. Never-the-less, you will get a qualitative idea quickly which may be fine tuned eventually a lot quicker than preparing the correct TEM sample. Use photographic film or 2D detector when possible.

Please join & post your query in the LinkedIn group - "X-ray Diffraction Imaging for Materials Microstructural QC": http://www.linkedin.com/groups?gid=2683600&trk=myg_ugrp_ovr

Here are some helpful links:

3D XRD: http://www.linkedin.com/groupItem?view=&gid=2683600&type=member&item=260221141&qid=31cec000-bd58-4e63-92cf-3259aab5f7de&trk=group_most_recent_rich-0-b-ttl&goback=%2Egmr_2683600

Panalytical XRD Webinar: http://www.linkedin.com/groupItem?view=&gid=2683600&type=member&item=257395175&qid=9550ab2d-9839-49af-9f2a-674db7d93900&trk=group_most_recent_rich-0-b-ttl&goback=%2Egmr_2683600%2Egde_2683600_member_260221141%2Egmr_2683600

I'm trying to get you this paper written a while ago by a colleague of mine:

http://link.aip.org/link/doi/10.1063/1.330795

I don't have off-campus access to Journ. Applied Physics, maybe someone here could get it and upload it, as I wouldn't mind a quick read of it as well - I think it would give all of us a clue on the amorphous to crystalline transition.

Prof. Grundy and Dr. Jones are excellent physicists and magnetism specialists so their papers are usually very good.

@Christopher Quinn , paper attached.

Nguyen Tung , about your mention formula (Volume fraction of crystalline phase= Areas of crystalline/ (Areas of crystalline + Area of amorphous).I found two papers which used this equation. I hope it can help you.

Many Thanks to Ahmed for uploading that paper, it seems there are some good similarities here between the results. Also the electropolishing method used in that paper is a good choice. This should give you some ideas on how to thin your samples. I think VSM or SQUID with heating stage would also be complimentary here, as the magnetisation should increase with crystallinity due to some short-range order being induced into the sample. Also DSC measurements also look to give very good indications of phase change, thus I would also recommend this as it is easy to calculate activation energy of a phase from the volume of the DSC peaks using the Arrhenius equation.

Lots of interesting work looks like it could be performed to give a much clearer picture for this material.

Here is an attempt at understanding the variation of Epi film thickness (volume fraction in the examined VOXEL) for a 0.5um MBE Epi of InAs/InAsSb on GaSb substrate 25nm SL periodicity:

Figure 1: http://www.flickr.com/photos/85210325@N04/10225984176/

Figure 2: http://www.flickr.com/photos/85210325@N04/10255348743/

Here are the links to the related discussion in LinkedIn and RG for more details. I'm hoping some of you experts would be able to review and critique the methodology employed. I'm close to the eventual solution but could use some help to hasten understanding.

LinkedIn: http://www.linkedin.com/groupItem?view=&gid=2683600&type=member&item=238135655&commentID=141338448&goback=%2Egmr_2683600%2Egde_2683600_member_245989945%2Egmr_2683600&report%2Esuccess=8ULbKyXO6NDvmoK7o030UNOYGZKrvdhBhypZ_w8EpQrrQI-BBjkmxwkEOwBjLE28YyDIxcyEO7_TA_giuRN#commentID_141338448

RG: https://www.researchgate.net/post/X-ray_Rocking_Curve_Analysis_of_Super-lattice_SL_Epitaxial_Structures_What_are_some_of_your_practical_experiences_with_this_method

In my opinion, the key to precise XRD data analysis includes:

1. Correct estimation of "back-ground" signal.

2. Correct XRD relative intensity data "normalization".

http://www.flickr.com/photos/85210325@N04/10225966185/in/photostream/lightbox/

You have picked a rather difficult problem!

First, as others have pointed out, using Cu K-alpha results in two problems: limited depth of penetration, and a large fluorescent background. This could be avoided by using a different x-ray source, but I appreciate that you may not have that luxury.

Second, your sample is a solid thick film, so you don't have the ability to easily add an internal standard.

Third, from the pattern of the bonded film (from your initial post) it is extremely likely that you have a significant amount of crystallization (based on the presence of a relatively sharp diffraction peak), with a significant amount of preferred orientation (based on the lack of intensity for other Bragg reflections).

Fourth, from your electron micrograph, the surface is not homogeneous, so your crystalline material varies not only as a function of depth, but in-plane as well. In such as case, what is the value of knowing a crystallinity, when it is distributed so non-randomly.

There are likely other issues, but already you can see that whatever "% crystallinity" number you obtain is going to be difficult to interpret micro-structurally. You may however be able to qualitatively compare samples fabricated under different conditions, and that might very well meet your immediate needs. Otherwise I can only recommend destructive techniques - such as sectioning, polishing, and micro structural/elemental analysis via electron microscopy (including EBSD which should distinguish crystalline from noncrystalline regions) - to obtain quantitative numbers that are meaningful. Perhaps someone else has a better idea?

What is the sample thickness? I couldn't locate it in the discussion yet. Still looking through.

Can you get any XRD signal in the transmission mode?

Do you have a reference sample with ZERO crystallinity to correctly estimate the back-ground signal with the fluorescence?

Figure A1: Composite with crystalline and amorphous content and high back-ground signal similar to yours. In your case you would most likely get crystalline Debye-Scherrer rings in 2D with a high background signal due to fluorescence and amorphous content.

http://www.flickr.com/photos/85210325@N04/7988731098/in/set-72157632728981912

Figure A2: 3D Perspective of the 2D XRD Diffractogram

http://www.flickr.com/photos/85210325@N04/7988760159/in/set-72157632728981912

http://www.flickr.com/photos/85210325@N04/7988760159/in/set-72157632728981912

Once you have correctly identified the intensity components from the fluorescence and the amorphous content for each pixel/VOXEL (they have distinctly different shape factors), then you could use the integrated intensity under the crystalline phase reflection to estimate the relative crystalline content.

1. You may choose to minimize the fluorescence by changing the incident beam wavelength (energy).

2. You may use a known standard to estimate the fluorescence.

3. You must eliminate any potential effects of "preferred orientation" by:

a. Using a modern real time 2D detector and/or collecting 2D diffractograms. Use photographic film. You'll get an idea right away without a fuss. You may even digitize the film with the good old scanner (photodensitometer). RG doesn't even have this word in its dictionary. The danger of knowing such facts is that it dates me way back to the 1950's. I came around way later than that, but saw the method used in India in the 1970's. The quality of information from the photographic film would be better than a single conventional diffractogram with a "fancy shmancy" 0D scintillation counter. Straight out of the Jurasic Age! Or with the 1D detector from the more recent Abrahamic era.

b. If you must only use the conventional ubiquitous 1D diffractogram, then you ought to at least rotate the sample about its surface normal and obtain multiple diffractograms.

4. Sprinkle a known standard powder on the sample (I assume it flat) with minimal fluorescence to create a reference pattern. Try it! You'll be amazed. Double stick tape and graphite powder may do the trick. This would illuminate any instrument based aberrations.

5. Make relative measurements. Avoid absolute, unless you have an excellent calibration standard!

Below is an example of back-ground signal ID and deconvolution:

http://www.flickr.com/photos/85210325@N04/8020997504/in/set-72157632728981912

http://www.flickr.com/photos/85210325@N04/8020997504/in/set-72157632728981912

Nguyen! "the original of my material is 100% amorphous in ribbon shape"

What is the ribbon thickness?

Where is the diffractogram from this as received ribbon?

Why have you not posted it yet?

Is it thin enough for transmission XRD Imaging? Test it with photographic film instead of 2D Detector..

http://www.flickr.com/photos/85210325@N04/10221065324/

The Rietveld method suggested by Pascal Roussel can solve the problem of quantifying crystalline and amorphous phases.

The Background Signal and Signal-to-Noise Ratio (SNR) seem to be the limiting/challenging issues besides potential preferred orientation in the original XRD data posted. How are corrections made in the Rietveld method to account for these effects? Especially, not knowing the 2D XRD pattern/diffractogram.

What is the essence of the Reitveld Method? If someone could explain it better, I'd like to "re-learn it all over again"! Some good links would be awesome to share! Thanks!

Ravi : Please read the full discussion to find answers to many of your questions. The film is 50 microns thick, the amorphous pattern was given about the second discussion entry. If you really want to learn Rietveld, get a recent book (I personally like Pecharsky and Zavalij, but there are some other decent ones) and do as many practice examples as you can. Rietveld is powerful, but very easy to refine your way into a wrong answer, especially if you are a beginner, or use an inappropriate model. Most Rietveld programs can model texture, but accuracy will be limited by accuracy of the selected model.

Thanks Andrew Edward! I like the references. I will (and must) read up as you have suggested (couldn't drive it into my cognition last time I tried). Since Nguyen conceded that he was a beginner, I thought a wiser one among us could possibly share the wealth. That's all!

30-50 Micron sample thickness would be near ideal for 2D transmission XRD! Try it! You could even use an external standard. My favorite is Aluminum foil. But you choose! Use photographic film and scanner! Be A Pioneer of the Future not a Prisoner of the Past! Post the images! You'll see the effect of preferred orientation, if any, instantaneously. You'll also get a pretty good idea about the contribution of the fluorescent component and its effect on the back-ground (BG) signal vividly. Ideally, you need the signal from the 100% amorphous sample as the reference. You may use the 2D signal ("shape factor") from the 100% amorphous sample to ID and deconvolute the BG signal and isolate the XRD signal from the crystalline phase and other "non-amorphous" constituents from the sample. You'd then need to ID the center of diffraction (assuming you use a 0D "point" source in transmission). Once the center of diffraction (transmitted beam direction) has been precisely identified, you may then use "polar integration" along the Debye-Scherrer Arcs to get back the correct conventional linear diffractogram minimizing the effect of "preferred orientation" in the sample. You may then be able to use the relative normalized integrated intensity beneath the "crystalline" peak to estimate the relative crystalline component contribution. You'll need some serious software tools. However, once the images are digitized, there are plenty of "freeware" image processing software tools available. I doubt if the present "cast of characters" among the XRD equipment manufacturers can wrap their minds around this concept. Good Luck!

I'd be curious to examine the entire diffractogram and not just a portion of it. I'd also verify invariance of the diffractogram by rotating the sample about its surface normal, if you continue to use the conventional 0D point/scintillation counter method.

"You have picked a rather difficult problem!" I agree!

But, it would seem that it shouldn't be so, after over a century of learning following the Braggs. It should be straight forward. The theory is clear. Especially, with the awesome plethora of quantitative XRD tools available since at least the 1970's that I know of. In my humble opinion, it is due to the lack of real time XRD imaging capabilities. I discovered it through Dr. Robert E. Green Jr. @ JHU Baltimore (1984-1988). Changed the paradigm for me! This one aspect inhibits materials experts from using XRD for routine materials Nano structural analyses.

To me, that is a GIGANTIC opportunity waiting to be harnessed! Nguyen, you are headed in the right direction. Bragg XRD Microscopy is a powerful tool! Experts like Dr. Payzant and others have used XRD correctly & effectively to unlock many Nano structural mysteries. This NDE tool may be used for in situ Nano structural examinations like none other.

[email protected]

My favorite external standard, Aluminum foil, right out the kitchen:

http://www.flickr.com/photos/85210325@N04/7893352226/

http://www.flickr.com/photos/85210325@N04/7893352226/

Here is some interesting work with "Polar Integration" of 2D XRD Transmission data collected from Beam Line X14A (Dr. Jian Ming Bai) Brookhaven National Lab for LaB6 standard powder:

1. Raw Data in YouTube Video BNL Beam Line X14A:

https://www.youtube.com/watch?v=IU0m4yI7D-k&list=PL7032E2DAF1F3941F

2. "Polar Integration" work done by Andrew Matseevsky, voluntarily:

http://www.flickr.com/photos/85210325@N04/8614015417/

We continue to improve. Lots of room!

[email protected]

http://www.flickr.com/photos/85210325@N04/10515372183/

LaB6 Standard Powder in Beam Line X14A BNL:

http://www.flickr.com/photos/85210325@N04/7890468142/

Thanks to Dr. Edward Andrew Payzant of ORNL for the introduction to BNL Brookhaven through Dr. Jianming Bai that facilitated this work. We are deeply in gratitude!

We are also in gratitude to Dr. Jianming Bai for inviting us and accommodating us at the beam line in short notice.

http://www.flickr.com/photos/85210325@N04/7890468142/

FYI - Jianming Bai is himself on ResearchGate:

Your formula is ok for the beginner. But in the next time if you want to calculate accurately, first you must collect XRD data from the standard sample, and after that collect XRD data from your sample. By comparing both sample you can find the volume fraction of amorphous and crystalline phases.

This question is usually for beginners :). You can find so nice answer which you got at the previous answer from other colleagues. Before you start to calculate anything you have to be sure that your data are without of any systematic errors which you can eliminated by using some standards as usually used standards such as quartz, LaB6 ore some other analytical pure very well crystallized compounds. They can be applied as external or internal standards. Only what you have to care that you haven't diffraction peaks overlapping between standard and your sample if you use internal standard (standard added into sample). X'pert highscore give you opportunity to reduce this systematic errors, and background. From results which you present in this question peak it appear so close to the aluminum sample holder! Results looks as that you haven't sample :). Did you use that sample holder? Did you subtracts the background?

Programs (FullProf, GSAS, RIBOLS,...) based on Rietveld refinement allow to obtain this quantitative information as Pascal Roussel, Ravi Ananth, Edward Andrew Payzant and Drazan Jozic have comment it before.

@Sornadurai Dharmadoss. It's a typical Cu-anode that cause Fe floresence.

I think that you can not do a lot of things with the shown raw diffraction (just one reflection, poor crystallinity, diffuse background...) and I am not seeing the amorphous phase signature that you want to quantify !!!

Dear Sornadurai Dharmadoss, thanks for the correction, sorry for missread ur statement...

It seems the optical (need to provide more info) and the crytallinity of the sample rise up the background.

For flurosence effect: I had a sample for Fe nano particles (~40nm) with (011) signal to noise ratio~7 and buried the other peaks. After change to Co anode, (011) signal to noise ratio~20 and the other peaks apeared.

Every piece of XRD data is good data. It is up to us to correctly extract the reciprocal space parameters including reciprocal space volume correctly. It is not just the weak "Signal-to-background" but the poor SNR (signal-to-shot noise) that could use enhancing for this data. Never-the-less, the one intense peak visible has reasonably good characteristics to compute Integrated Intensity from. The challenge in this case is the use of the conventional linear "equatorial diffractogram" which is spatially blind to several other Nano structural characteristics including "preferred orientation" in the sample.

I saw an article about Aluminum foil using the conventional diffractogram (http://www.icdd.com/resources/axa/vol40/v40_626.pdf ) where they have excellent signal from the (111) reflection but not the (200) due to preferred orientation. Well, if they had rotated the sample surface about the "Phi" axis (normal to sample surface), they'd have found the other reflection with ample SNR. Due to the "spatial blindness" inherent with the conventional approach this vital information is lost/missed even by the experts.

If you review the real time 2D diffractogram below, you'll notice that along the equator with maximum (111) activity the (200) intensity/activity is near "zero". However, the reverse is true if you rotated the equatorial plane by about 40 degrees (clockwise or counter clockwise).

I do see Sornadurai's point that one reflection alone is insufficient to image the entire reciprocal space volume. The diffractogram displayed may have entire crystalline peak but certainly not the entire amorphous diffraction pattern. There appears to be a large "DC component" in the signal.