There are many ways of performing quantitative XRD and it will depend a lot on the software you have available. However the first thing you will need to do is to identify all components in the system as the methods are relative rather than absolute, meaning if you miss out a phase, the other phases will be overestimated. If you have access to PDF-2 database or similar and suitable software a simple way is to use RIR method (reference intensity ratios). If the phases identified have an I/Ic value then you can scale the stick patterns to the 100% peaks and obtain a semi-quant analysis. This method suffers if there are any preferred orientation effects as the relative peak intensities will be incorrect. Also not every phase in the database has I/Ic value. A more robust method is to use full pattern fitting and Rietveld refinement. Again you will need to have identified all components and have the appropriate software. Have a look on these pages to get you started.

http://www.ccp14.ac.uk/ccp/web-mirrors/hugorietveld/stxnews/riet/faq/progs/riet-pc.htm

There are many ways of performing quantitative XRD and it will depend a lot on the software you have available. However the first thing you will need to do is to identify all components in the system as the methods are relative rather than absolute, meaning if you miss out a phase, the other phases will be overestimated. If you have access to PDF-2 database or similar and suitable software a simple way is to use RIR method (reference intensity ratios). If the phases identified have an I/Ic value then you can scale the stick patterns to the 100% peaks and obtain a semi-quant analysis. This method suffers if there are any preferred orientation effects as the relative peak intensities will be incorrect. Also not every phase in the database has I/Ic value. A more robust method is to use full pattern fitting and Rietveld refinement. Again you will need to have identified all components and have the appropriate software. Have a look on these pages to get you started.

http://www.ccp14.ac.uk/ccp/web-mirrors/hugorietveld/stxnews/riet/faq/progs/riet-pc.htm

First of all, does your pattern shows other reflections in addition to those of hydroxyapatite?

If it doesn't, then the only crystalline phase is hydroxyapatite, and possible impurities must be amorphous. In this case, you could get the amount of amorphous matter by Rietveld refinement, which is a quite complex procedure.

If it does, then you need to identify to which phase (or phases)  the additional reflections belong. This could be quite easy, if the spurious phase(s) is(are) known.

Once you have identified the various phases present in your pattern, you have to perform quantitative Rietveld refinement, in order to understand the relative amount of each phase in your powder.

In any case, I don't think you can have a reliable estimation of the percentage of each phase just by comparing the intensities of the reflections belonging to different phases.

Dear Shruti Mohan,

I suggest you first do a good data collection of X-ray diffraction After this step, you should make a structural refinement using the Rietveld method, which will allow you a quantification of phases much more accurate than that obtained by simple intensity of the diffraction peaks. There are several free software suites that allow this refinement, such as the GSAS and FullProf are free and available for download. In both software after each refinement cycle is provided the percentage of the phases present in the material. If you have this amorphous material in a sample, I suggest you to make use of an internal standard to determine the amorphous fraction in a material. In this case the internal standard must be refined as another crystalline phase. As the internal standard is introduced in a known quantity in the calculations are performed in order to provide the same amount after each refinement cycle. That is, after each cycle the ratio is determined and multiplied by a scaling factor to provide the same amount added. All other phases are corrected by the same scale factor.

Best regards.

Taking into account the comments above, i would strongly recommend using full pattern fitting. The use of the I/Icorr values is not particularly reliable. We did some work on comparing full pattern fitting with other methods and full pattern fitting is the best. It requires some expertise as indicated by the other commenters but if you can find someone who is good at Rietveld this shouldn't be an arduous task, assuming the phases are not unusual.

Thank you very much for your replies. They were really helpful and now I am working on my sample with these.

I also agree with all.

I also agree with above comments