I'm not sure I fully understand your question. You want to do high quality XRD analysis, but you don't want to use software to do the analysis??? Why not? In the days before computers and software I used to use a chart recorder, a ruler and pencil, a copy of the Hanawalt Search Method and a databook from JCPDS (when it was still called JCPDS), this worked pretty well for most qualitative applications, but a bit slow. And for quantitation (of clays) I used a pair of scissors and a 4-figure balance. Probably not an option for the multiphase work involved in natural rock assemblages or cements, where I would much rather use the power of Rietveld quantitation.

Origin is software for plotting and will give you a qualitative rather than quantitative view. Note they are not spectra, it is not a spectrographic technique, they are called diffractograms, patterns or traces.

Can you be more specific with your question? What XRD analysis are you trying to achieve, and on what type of samples? - there are many ways in which these data can be interrogated. And why exactly do you not want to use software? - slightly baffled...

Dear Ian,

Many thanks for the message. Actually, I am working on the Chemistry of  molten salts mixture, KF-ZrF4 eutectic. I have recorded the XRD for the pure salts without baking and after baking. I want to know is there any hydrate or oxide in the salts or not. Also, we recorded the XRD of the eutectic too and would like to know is there any hydrate or oxide in the salt eutectic or not. Also, Is there a way to do a quantitative comparison of two diffractograms (recorded and database)? Which software do you use to improve the quality of your diffractograms? I would like to use the software to analysis the patterns but the quality of the figures will be decreased when I used the software. 

Kind regards,

Zahed,

Zahed,

to take you points in order. If you want to know if you have hydrate or oxides it is best if you use software, yes it can be done manually, but it is far quicker to use software - you will need access to a 'search-match' routine that links to a database. The usual databases are ICDD PDF-files (commercial) or COD (free). For my own purposes I use EVA as it is a Bruker diffractometer, but whatever software came with your system should be adequate for basic qualitative work such as identification of hydrates or oxides. If you want to do it manually, without software, go to the databases find the d-spacings for your target phases, measure your peak positions, convert to d-spacings with Bragg equation, and make a comparison. 

Your question about 'quantitative comparison of two diffractograms (recorded and database)' does not make much sense, There will not be a database diffractogram equivalent to your phase mixture. They exist as patterns for pure phases. It sounds as though you are describing qualitative rather than quantitative analysis. This could be done with any plotting software.

Software to improve the quality of the diffractograms! IN MY OPINION, software cannot improve the quality of a diffractogram, and it should never be attempted. Any routines such as smoothing is only throwing data away. The only way to improve the quality of a diffractogram is to re-collect the data with more suitable instrumental parameters. I really don't understand what your software is doing such that it degrades the patterns. 

Maybe it would be best for you to post up some data to show us what you mean, and what issues you having.

You are working in a prestigious University which has ample XRD facilities with both Scintag and Rigaku diffractometers. You have access to JADE for qualitative and quantitative analysis and access to the PDF-4 database. These facilities are more than adequate to meet your needs. If you have trouble using them then it may be sensible to talk to your professors or to the laboratory technical staff who should be able to help you.

Dear Ian,

Many thanks for the comprehensive response. Actually, we want to be sure is there any moisture (hydrate) or oxide species in the salts and the formed eutectic or not. We have used the PDF-files in order to evaluate the possibility of presence of any oxide or hydrate. Please see the attached files. My professor believe that showing overlaid graphs like the ones attached isn't very effective.

Kind regards,

Zahed.

Hello Zahed,

if the Y-axis scale is really the number of counts you have, then the data are very poor indeed. Peak shapes are poor and minor phases will be near impossible to determine. With so few counts you cannot see if hydroxide or other products are present or not.

You must gather better data. You should either count for much longer at each step or open up the slit size - or both. It seems you have a maximum of 3 counts in the 100% peak (on the Eutectic) but only 1.9 counts on the other two. To have good enough data, the counts value should be in the thousands or tens of thousands of counts. Only then can you analyse this properly.

It looks as though the background has been stripped right out. While this may be necessary for some search-match routines it is better to display it with background included for presentation purposes.

There is a problem of peak overlap, near isostructural solution, for Zr O F5 and Zr F7 phases, this can only be resolved with better data.

Ian

Dear Ian,

Many thanks for the useful comments. So, we have to gather better data. Do you have any suggestion on how to gather better data?

Kind regards,

Zahed.

As I suggested, you need to count for longer and/or open up the slits. Both of these approaches will increase the counts at the detector. However, opening the slit will cost in terms of resolution, increasing the count time will cost in terms of data collection time. This will improve your counting statistics, and increase the likelihood of detecting small quantities of other phases. 

I have attached a basic presentation from Bruker on good diffraction practice. There is a lot of useful information there. It should help you decide on a data collection strategy.

There is also the possibility that the poor counts are due to incorrect instrument alignment. This is beyond your control, but you should ask the lab supervisor to show you data that proves the instrument is passing its validation checks which should be carried out annually. There should be a standard sample, corundum, or silica or some such, and a prescribed set of parameters, generator voltage, 2-th range, step size, count time, slit sizes, detector configuration. And to go with that a certificate that shows the numbers of counts achieved on a certain peak on the standard, and that the peak is in the correct 2-th position. If this is all in order then the problem will lie in your data collection.

Dear Ian,

Many thanks for all your kindly help and giving me your precious time.

Kind regards,

Zahed.

Dear Ian,

We tried to increase the counts on the XRD test. The attached is a sample of the result. Do you think the quality of the data is enough good?

Kind Regards,

Zahed.

Yes this is much better! You should be able to identify the minor phases with more confidence.

Dear Ian,

Many thanks for your consideration.

Best,

Zahed.