Hi Rafaeldi,

Sample prep is extremely important for good quality XRD. With respect to the 4 cases you mention:

1) smooth, flat and level surface (wrt sample holder): great

2) rough surface: can lead to "spots" on the diffractogram, meaning that some reflections will be more intense than they should. If not flush with the sample that will trigger systematic displacement (peaks are shifted systematically)

3) large particle size: same as 2) for the peak intensities but much more pronounced

4) low surface (wrt sample holder): could lead to "flat specimen error" and skew the 2theta values.

I suggest you read this great document from Jim Connolly (U. New Mexico)

http://epswww.unm.edu/media/pdf/07-Errors-Sample-Prep-PPT.pdf

The orientation of the crystal planes to the x-ray beam affects the intensity. That's why interpretation of peak heights is a minefield. Try grinding the sample further to randomize the orientations.

Hi Rafaeldi,

Sample prep is extremely important for good quality XRD. With respect to the 4 cases you mention:

1) smooth, flat and level surface (wrt sample holder): great

2) rough surface: can lead to "spots" on the diffractogram, meaning that some reflections will be more intense than they should. If not flush with the sample that will trigger systematic displacement (peaks are shifted systematically)

3) large particle size: same as 2) for the peak intensities but much more pronounced

4) low surface (wrt sample holder): could lead to "flat specimen error" and skew the 2theta values.

I suggest you read this great document from Jim Connolly (U. New Mexico)

http://epswww.unm.edu/media/pdf/07-Errors-Sample-Prep-PPT.pdf

In addition to both Alan and Nico comments, you can calibrate your XRD using some standard, like quartz. Run XRD for long time for better accuracy (2-4 h), it is depends on noise level you are getting there.

I completely agree with the answer of Prof. Dr. Alan F Rawle

The orientation of the crystal planes exposed to the XRD beam affects greatly the intensity. That's why interpretation of peak heights is a minefield.

The introduction of Jim Connolly is great.

Dear Rafaeldi,

I agree with all the answers given above.

In addition to all these answers please have a look at a very similar question and subsequent discussion on the XRD sample preparation topic from some days ago here on RG.

https://www.researchgate.net/post/What_is_the_effect_of_the_sample_preparetion_on_the_XRD_diffractogram

Rafaeldi, it is nice to meet you, even if only electronically.

The answers you've gotten so far are spot on. I would also especially caution you on making judgments solely on peak height, it has been stated here that it is a minefield, and I couldn't agree more.

Using the reflectance XRD technique as you are, sample prep is of the utmost importance. You're looking to grind to as fine a particle size as possible. Even more important, is the consistency of sample preparation. Remember that XRD is a comparative technique, you are looking to make judgment calls and compare sample to sample, and if your preparation technique is not consistent, then you will never be able to accurately judge one specimen against the other.

So I think to start with, most of your variations are coming from inconsistent sample preparation, as soon as you get this part of it under control, I think everything else will fall into place.

I agree with all that has been said, but wish to emphasise the difficulties posed by working with coarse-grained powders (or large ordered domain sizes). How large is acceptable or, more likely, how small is acceptable? This depends in part on the linear absorption coefficient of the sample as well as the grain/crystallite size - see pages 365 - 367 of the classic text by Klug and Alexander (second edition, 1974). Note, especially, Figure 5-43 which helps us estimate intensity reproducibility (+-%) according to chemical composition, x-ray wavelength and grain size. We see why about 5 microns is generally safe in terms of intensity reproducibilty, whereas the intensity reproducibility at 50 microns is dreadful.

I always consult the figure when planning XRD data measurements for new sample types.

Kind regards.

Brian O'Connor