Keep it simple: Feldspar forms a solid solution of 3 end-members:

K-feldspar: KAlSi3O8

Albite: NaAlSi3O8

Anorthite: CaAl2Si2O8

Using the end-member composition you can easily solve your questions for min and max values.

I recommend to check any mineralogical textbook (e.g. Deer, Howie, Zussman) to get more inside into compositional variations of feldspar and plagioclase

I agree with Tobias. If one takes the molecular ratios of the given oxides the maximum is 1:1 and the minimum is 0:1 for both anorthite and albite. This happens only in the end-member compositions.

I agree with both of you. But, this (weird, is it?) idea crossed my mind when I went though the attached diagram (from a journal article). Nothing much is mentioned about the K/Al ratio lines. Could you please explain where I have mistaken and why 1:1 in albite or anorthite? If needed I can attach the article.

Based on my experience. Agree with Tobias but great feldespat is commonly mixed has K, Na and Ca Al2O3 2SiO2 2H2O and Caolinite. Some materials a I,ve know its relation in Na/K is from 1 to 5, from 1 to 5 % and a total alcalis from 9 to 13%. As mentioned by Tobias you can solve how much of each one feldspar you have according to chemical analysis. The Al2O3 goes from 14 to 44 %. Hope this help but the quality varies from one to other mine including in the same mine.

Tharanga: Without knowing the paper, I assume that the ratio refers to the Kfs composition in particular study.

Tharanga,

I looked at the Alkali feldspar analyses given in the Mineralogy reference (student edition) given above by Tobias. In the diagram you are making reference K2O % goes only up to 10 suggesting that they are dealing with Na2O rich alkali feldspars. In very potassic feldspar K2O% goes up to about 16. Also from Deer et al it is clear that in some natural feldspars Fe+3 gets into Al+3 positions thus the theoretical K2O/Al2O3 ratio might go beyond 1 for such K2O rich alkali feldspar, However, irrespective of whether K or Na rich the Al2O3 % in alkali feldspar according to Deer et al is around 17 -21, but the diagram you are having problems indicate Al2O3 percentages as low as 10, which is a bit strange. Perhaps that is why you have posted your question.

The literature must to mention how kind of feldspat, this is because some feldespats are processed and washing for separate or purify the caolinite from feldspart, this coul be one of the reasons that exposed by Perera. But  a crude Feldspar in mine is according to i've mentioned above. So that paper from Tharanga some Potasic Felspar includig  large K2O as 10% and Al2O3 18-20%. ie when Al2O3 its on 44% then that is a mixure of caolinite and the potasium decrease too low. 

Thank you very much for your inputs. Now I am going through the original article (Cox et al., 1995), the base of the above diagram. It's mostly clear now. One more question, has any one published this sort of diagram for Plagioclases according to your knowledge? (that would ease some of my extra burden) 

Not need for further explanation, your question has been answered through the discussion thanks to the input of all the participants. By the way, that figure is from one of my papers (Fig 3f): Campos Alvarez and Roser 2007. Geochemistry of black shales from the Lower Cretaceous Paja Formation, Eastern Cordillera, Colombia: Source weathering, provenance, and tectonic setting, Journal of South American Earth Sciences 23, 271–289, and simply reflects the abundance of clay minerals is indicated by low K2O/Al2O3 ratios (

I agree with Perera.

Hi Taranga

Have you tried an AK–C–N ternary plot (c.f., Fedo et al., 1997).You can use it as a proxy for the degree of albitization of plagiocalse?

Are you working with Whole rock data? do you have data for trace elements?

if that is the case, you cal plot for example a binary Be Vs Na/Al plot, and relate any systematic distribution or trends to the remobilization of Be (an other trace elements) and Na, to the effect of albitization, be it due to weathering or hydrothermal alteration, etc. ...you can look for a study about that was carried by Giuliani, G., et al 1999 but unfortunately I do not remember the complete reference...sorry about that