Yes, this can be done using Siroquant or Highscore + XRD spectra. You may refer this to some papers by Dr. Colin Ward. 

Hi Cheng Yang,

the step I would do are the following:

1) Try to find all the research publication of the outcrop in advance for knowing all the possible mineralogy that could be present there.

2) With XRD diffraction, you can do it by powder analysis or in bulk mode. So, If you have the sample powdered, you should do a very detailed pattern matching of XRD pattern (Match! is a good option, Xpowder12 has a good database and friendly enviroment for quantification). On the other hand, you can do an analysis in bulk at several points (in the fresh part and the altered part). In this regard, you will be able to  compare with the powder results about  which minerals you have in matrix and which are from alteration. 

3) If you have the possibility to use another spectroscopic technique like Raman or IR, you will be able to distinguish the specific minerals that you are looking for and confirm your results

4) You should read about alteration process from parent to secondary minerals, check if your results could match with the different processes.

Mr. Towe, that is true, it will give only semi-quantification. However, I answered taking in consideration the techniques that he can use and are available. Also, it is necessary to prepare oriented measurement and to complement with other techniques.

In theory, the relative abundance of minerals can be estimated using peaks in XRD patterns, but in practice this method can be challenging. One problem is that it is very difficult to obtain consistent results from packed powder samples. You can get a good practical test of the method by analyzing several replicate samples to see how the patterns compare. If several packed powders prepared for the sample give similar results, the XRD method may work for you. Another issue is that XRD patterns from whole rock samples generally are hard to work with, because you mostly get good peaks only from a few of the most abundant minerals, peaks for other constituents tend to be lost in the background.

Well, Thank all of you for the valuable advices, escepcially those detailed advices from Mr lalla and Mr Mustoe. It is indeed a very tough job like Mr Towe said, but i still want to have a try, since there isn't a better way.

 Nothing to thanks, this pages is for that. You can take a look in this book chapter, maybe it can be useful for your objetives 

http://www.sciencedirect.com/science/article/pii/B9780444825681500197t

It can be quantify if u had:

1. Good sample preparation (1-10 um)

2. Good data collection strategy (angular range, step scan and scan time)

3. Good monochromatization of the radiation

    - Kbeta (not good for Fe rich sample / flouresence)

    - Monochromator (huge intensity lost)

    - Energy discriminative detector (no Fluoresence and no intensity lost)

Cement is one of the example of how XRD can quantify more than 15 different phases.

to my sense it is difficult, because you have no idea what element exchange exactly took place during alteration. If it is highly altered you cannot say anything, but if the alteration is partial you can predict something. If you kindly provide me little more information in my personal e-mail ([email protected]) I can help you.

Due to the complexity caused by high alteration degree for most of the samples, i may have to choose another way.

Thank you anyway.