My XRD data when plotted have some noise, baseline is not smooth and it is not on the zero of intensity axis. There are some unnecessary small peaks similar to noise I want to avoid those peaks. What are the things I can do to make my XRD plot better? I am attaching the XRD plot of my material. Please tell me step by step procedure to tackle these things. You can also help me by referring any tutorials about origin software. I am using origin Pro- 2018 version.
Unniyarcha K.K,
First of all, the y-axis "intensity" is arbitrary units, i.e. you can remove the numbers and change the scale to create a more aesthetically pleasing plot.
Second, in terms of analysis, you can do a qualitative analysis on the crystalline phases by comparison of the peak positions to a database like the PDF4+ from ICDD, Crystallographic Open Database (COD), etc. Avoiding "peaks" that look like noise is not scientifically accurate, unless you can systematically discount them, i.e. sample holder effects, spiking agents, etc.
The elevated background between 20-45 degree 2theta indicates that you have an amorphous phase that may cover some minor phases within your sample. This requires some high level Rietveld analysis and spiking the sample with standards to properly quantify, among other techniques.
All of this said, you cannot do all of this just in Origin. You need specialized software to do Rietveld analysis to quantify any of your data. Qualitiatively, you can compare your data to some of the patterns in COD or PDF4+.
Good luck!
Dear Unniyarcha K.K ,
I totally agree with all the answers above from Alec Ladonis , Steven Van Petegem and Ahmed Y. Yassin .
You should not modify your XRD pattern.
You have to interpret it with respect to the properties of your actual sample and/or shortcomings of your experiment.
With respect to 'Origin' tutorials I want to add, that you may visit the RG home page of our colleague Prof. Shahid Ali (https://www.researchgate.net/profile/Shahid-Ali-4) and have a look at his questions and answers. There he provided a lot of tutorials with respect to 'Origin' applications...
But as Alec mentioned above, you cannot do all XRD interpretation with 'Origin'.
There is one thing I want to add: a background, such as in your case, may not only be due to amorphous phase contribution but may also be partly due to x-ray fluorescence of your sample. When for example you have got iron (Fe) in the sample*) and you use a Cu based x-ray tube, you will get Fe K-fluoresecence...
This is one example of a 'shortcoming' of an XRD experiment as I indicated above.
*) you may share details of your sample such as expected composition, expected thickness, bulk/powder/thin film etc.
'A priori' knowledge makes interpretation easier...
Thank you Steven Van Petegem. I have seen in youtube videos that xrd plots are modified with smoothening, baseline correction etc. By smoothening what I understood is that to remove noise or noise like peaks. On the basis of those videos I came to a conclusion that my plot should also be smoothened.
Thanks for feedback. But Unniyarcha K.K ,
for what pupose you will smooth your pattern. Just for 'better' display/appearance?
Any smoothing will destroy a bit of information...
Any smoothing by software will pretend better photons statistics, which in reality you do not have. Please do not cheate the readers.
Smoothing by increase of scan time is ok and reflects real experienced photon statistics.
Thank you @Gerhard Martens. I was unaware of whether my XRD plot requires smoothening or not. Some videos have done smoothening that's why I also wanted to do that. I thought smoothening is a necessary thing to be done before analysis. If smoothening destroys data as you said and if it is not a good scientific practice I will not do anything like that.
As mentioned above, don't do things to your data. All weird looks are essential.
Better to share your raw data with detailed instrument configuration and your sample information, i.e., elements, synthesis, etc.
Try not to use Origin software, or it's kind to analyze diffraction data.
Dear Unniyarcha,
It is obvious that background noise must be taken into account when fitting the data.
But these small peaks are not always due to the background noise when fitting the data. The peaks are closer together because the mesh is larger, as in the case of oxides for example, and when the mesh is small, the symmetry is higher. The larger the mesh size, the more small peaks are found, which is due to the fact that a small part of the grain structure is deformed. Attention but not the whole structure!
Best kinds.
- Badges
- Science method
- Similar topics
- Methodology
- Crystallography
- X-ray Diffraction