For each particle: I used microprobe(EMPA) to analyze the rock, the main minerals in the rock are: muscovite, lepidolite, quartz, topaz, albite, heredite and other minerals contain lithium: amblygonite, montebrasite. i can calculate H2O and Li2O in micas, but in phosphates i cannot.
For general of rock: i used XRF, AAS to analyze major and trace elements.
My question is: how can i show the mineral composition of my rock by diagram?
Have you used normative mineral classification based on your analysis? e.g. http://en.wikipedia.org/wiki/Normative_mineralogy references at the end of the article explain that. or are you saying that you are having difficulty in deterining H2O and Li2O in the phosphate rocks?
i did try with CIPW, but it didn't succeed, you can see the attached file. I can calculated H2O and Li2O in trioctahedral micas, but i cannot do the same thing with amblygonite, montebrasite.
Hien in doing normative minerals we often found difficulties with our totals when fluorine was present in appreciable amounts. This seemed to be due to compounds which were much more complex than the simpler silicates. In http://webmineral.com/data/Amblygonite.shtml the formula for amblygonite is provided and perhaps you can use that to calculate the lithium and fluorine contributions for the minerals you have had problems with and subtracting those percentage from the CIPW calculations and normalizing based on the remaining concentrations. Then ratio the amblygonite and montebrasite back into the totals and adjust the final distribution. Would that work?