Practical Aspects of Mineral Thermobarometry
This area of the site will be used for illustrating some applications of THERMOCALC in phase diagram mode to particular chemical systems. It could expand to several pages in due course, but for now, here is one offering:
Modelling with Na and Ca phases
Two issues are worth dealing with here.
Aluminous phases with Na and Ca: paragonite and margarite
Na and Ca-bearing sheet silicates are Al-rich, and lock up Al which might otherwise appear as aluminous porphyroblast phases. In this example, the rock R76 (Hoosac schist from Nick Gardiner) has the molar composition
| Al2O3 | FeO | MgO | CaO | Na2O | K2O |
| 20.11 | 5.43 | 2.41 | 0.76 | 2.30 | 3.92 |
The AKF diagram below shows the bulk composition projected from plagioclase (Al-Na, red square) and from paragonite (Al-3Na, blue circle). See the dramatic difference that this makes to the plotted position on the A-FM join when the compositions are then projected from muscovite.
In this particular case, the bulk composition was derived from XRF analysis of a bulk sample. The part of the rock to be modelled, i.e. that seen in thin section, contains paragonite together with garnet and some staurolite. Its true local bulk composition, therefore, should have plotted above the Ms-Chl tie line, so that it projected into the observed AFM assemblage. When you use the XRF-based composition for modelling you fail to find the observed mineral assemblages.
Moral: A small error in bulk composition estimation for Na and Ca can have a big effect when that composition is projected into AFM. Always check that the bulk composition is consistent with what you see.
General phase relations of plagioclase with other Na and Ca phases
Here are some results of an exploration of relationships among plagioclase, paragonite, margarite, clinozoisite and kyanite, displayed as composition diagrams projected onto Ky-Ab-An. To use them, you'll have to imagine your rock composition projected from all the K-Fe-Mg phases. In most pelites there will be significant Na in muscovite and Ca in garnet. The calculations for 6 kbar and a(H2O) = 1 are shown. Similar phase relationships are seen at other pressures.
 In these diagrams the plagioclase solid solution runs along the base of the triangle. The limiting Pl composition at 450° is controlled by Pg + Czo, and is rather Na-rich. |
 The univariant reaction Pg + Czo + Qtz = Mrg + Pl has been crossed, and the limiting Pl composition is now in Mrg + Czo + Pl. The univariant reaction lies at 473°, with plagioclase of an31 composition. |
 At higher T, but with the same phase topology, we see that the limiting Pl composition has become more Ca-rich, but the Pl in equilibrium with Pg and Mrg is little changed. Before 600° the full range of Pl compositions becomes stable, and Czo will be confined to Al-undersaturated rocks. |
 The next univariant reaction Pg + Mrg + Qtz = Pl(an34) + Ky has just been crossed, and over a fairly short T interval both margarite and paragonite will become unstable in the presence of quartz, leaving the full range of plagioclase compositions stable with Al-silicate. Compositions below the ab-an join may still contain Czo. |
This page last modified 12 October 2004