It depends on the intensity of peaks of the element/ componund which you are measuring. Generally one takes 1000 counts of the least intense peak in order to get a clear/good XRD Pattern. According to my information there is no standard Formula for time per step. 

you may need to try different step sizes and scan speeds in order to optimise the low angle peaks. Also the choice of X-ray tube with different wavelength x-ray may have an effect on peak positions and intensity.   I would also suggest running a standard of some kind to measure low angle peaks.

Even more important the number of counts you are getting is the signal to noise ratio. If you have a very large background (which is often the case at low angles if great care is not taken to reduce the low angle scatter into the detector via shielding or flight tubes) you have to count for a much longer time to get the same signal to noise. Reducing the background is your friend. You may even want to consider using an analyzer crystal. Although it reduces the intensity significantly, it can allow you to bring the background down to practically the electronic noise level of your detector. Suddenly you will be able to get the same signal to noise at low angles with tens of counts per second that you had with thousands of counts per second and a high background.

BTW what is "low angle" in your case? Please post some of your diffractograms to make it interesting and help us understand your paradigm better. Examples of how to attach a variety of files is illustrated below :-)

Also shown is a 2D XRR signal from a NIST 2000 SRM (standard reference materials) and its rocking curve. Is that low enough angle for you? Imaging the incident beam!

https://www.flickr.com/photos/85210325@N04/11343736043/in/album-72157645018820696/

 Artur! In fact, I have no doubt about "background intensity" having signal from the structure of the material based on our real time observations in reciprocal space. I'm not sure how to deconvolute (extract) it yet. I suppose it would have to be done through careful modeling and simulation using software such as Bruker LEPTOS  or other. Any suggestions or references would be welcome :-)

Thank you for answers to my question.

I did low angle-low speed analysis with 5 seconds time per step and the step is 0.01 degree, but I did not observe the peak shift. Should I consider more time per step in order to observe peak shift?

Peak 100 of  XRD analysis for Each material was selected for this analysis and observation of solid solution forming.

Thanks Mr. Braun to send the paper and all recommendations of others.

You will help yourself dramatically when you disclose all the details.

• Are you expecting a peak shit?
• Are you able to resolve the peak clearly?
• Why would you expect the peak shift to become more or less evident with changing step size or dwell time? Is SNR an issue in your case?
• Are there elements in the sample that may fluoresce with the source wavelength?
• What is the flight path (length) to the detector from the sample?
• What size source and receiving slits are you using?

Mr Ananth

Yes, I expect that I will see peak shift.

My project is research on nonstoichiometric in dielectric ceramics.

However, thanks to guide me. I need to more studying on this problem.

This is why you must minimize the background with an empty sample holder. If you want information from scattering between bragg peaks due to disorder it is even more important to reduce scatter from the sample holder or air or other artifacts. I take it as a given that the definition of background is scatter that is not coming from your sample, and any scatter coming from your sample is signal (expect for fluorescence, which can be removed through detector electronics).

Like what Artur? Please share some literature references and data if possible. I'm keenly interested in seeing XRD "diffused scattering data" and analyses. Thanks!

How does one deconvolute sample Nano structure from the reciprocal space data using "diffused scattering" signal, either 2D XRD observations or for that matter from a conventional linear diffractogram?

How far from the Bragg Peak should I consider as "diffused scattering"? See example of a Y-Omega map for a production grade 3" (75mm) diameter SiC 6H (0006) oriented wafer polished "optically flat". Observe the lateral correlation in reciprocal space (Y-Om & X-Om Maps).

X-Omega Map is a series of Bragg profiles along a horizontal cursor line at the center of  the XRD topograph (109)- reflection while Y-Omega Map is the same along a vertical cursor line at the center of the topograph. Each pixel is 27.5um along X & Y (both on the inordinate axis), while along Omega (ordinate axis) each Pixel is about 0.288 Arcsec (1.39626E-06 Radian, about 1.4 micro Radian). Both grey scale and pseudo-color maps are attached. You may download and magnify to measure. 

My interest in this case of SiC 6H is at the tails of the Bragg peak. The "diffused scattering" at the Bragg shoulder/tails contains much information about the defect structure (which causes the "disorder") in real space. I need to understand this aspect better. It is most likely possible to ID the defect type based on this sort of "diffused scattering" signal.

Bottom Line: Before subtracting or removing any part of the signal one must ID, deconvolute and verify. Larger the dell time lower the standard deviation. Smaller the step size better the angular resolution.