Phase Equilibria in Low Alumina and High Alumina Pelites

Summary

Pressure — temperature pseudosections, or phase diagrams for specific rock compositions, constructed for Waterville formation pelites studied by Ferry (1982) predict the equilibrium mineral assemblages for metamorphism. This study has recently been published in Geological Materials Reseach Our results predict that:

1. the triple point aluminum silicate assemblage: kyanite + andalusite + sillimanite is not stable for the range of compositions that we considered,
2. aluminum silicate minerals are produced by consumption of muscovite+chlorite+quartz or muscovite + chlorite + staurolite + quartz during isobaric heating,
3. garnet is produced by consumption of chlorite, muscovite, & quartz at pressure > ca. 1.5, and
4. at pressures below ca. 5 kbar, garnet is only stable at temperatures above those of the aluminum silicate triple point.

Changes in the alumina concentration cause significant changes in the predicted mineral assemblages. A decrease in potash concentration (and hence a proportional increase in other system components) drives paragonite stability to lower pressures, increases the pressure-temperature range of staurolite stability, and causes growth of aluminum silicate before biotite.

Data

The average chemical composition of eight biotite zone pelitic rocks published by Ferry (1982) was chosen. Biotite zone rocks were chosen in order to determine variability of the paragenesis with respect to change in composition. The concentrations of K₂O and Al₂O₃ were varied independently (Table 1). Figure 1 shows an AFM diagram for the original and two modified compositions. Ferry's biotite zone average corresponds to low Al pelites according to composition fields in Spear (1993). The two modified compositions correspond to high Al pelites.

1 Average pelite — 8 analyses from the biotite zone

2 Composition 1 modified with increased Al

3 Composition 1 modified with decreased K

Figure 1. AFM diagram for average pelite, high Al pelite, and low K pelite

The nine component MnCaNaKFMASH model compositional system was chosen for calculations. Note that a pseudosection for average pelite (Shaw, 1956) in the eight component CaNaKFMASH system can be found on the perplex website:

The internally-consistent thermodynamic data of Holland and Powell (1998) were used for all of the calculations. The following minerals appear as stable phases or were considered in our pseudosections: garnet, plagioclase, biotite, muscovite, chloritoid, chlorite, cordierite, paragonite, clinozoisite, zoisite, sillimanite, andalusite, kyanite, staurolite, and K feldspar. The activity models for minerals were calculated using the formulations and THERMOCALC data files described by Tinkham http://bama.ua.edu/~tinkh001/Pages/Activities.html.

The program THERMOCALC (Holland & Powell, 1988) was used for the following:

1. First, Gibbs minimization calculations were considered over a selected pressure and temperature range in order to determine the potential stability of phases.
2. Second, Mode 1 of THERMOCALC is used to locate mode=0 curves, representing the addition or removal of a phase from the equilibrium assemblage. The following rules are considered in order to determine pseudosection topology:

1. Mineral assemblage variance only changes by 1 across all mode=0 curves, except for univariant curves, where the assemblage is divariant on both sides of the univariant curve,
2. Where two non-univariant mode=0 curves intersect, the variance changes by 2 in two opposing fields, and is equal in the other 2 opposing fields.
3. Last, pressure-temperature locations for all equilibria curves and points are copied into an EXCEL spreadsheet for graphing.

Recent presentations on average pelite psuedosections:

2001 SE Geological Society of America

2001 NE Geological Society of America

Mineral stability fields are colored according to variance; divariant fields are shown in yellow (see high Al section at high T/moderate P), trivariant are orange, quadrivariant are blue, pentavariant are tan, and hexavariant are gray. Pseudosections for the three rock compositions (Table 1) are shown in Figures 2, 3, and 4.

Figure 2. Pseudosection for average pelite

Figure 3. Pseudosection for high Al pelite

Figure 4. Pseudosection for low K pelite

The location of several reaction lines change with increased alumina. This is especially pronounced for the clinozoisite out and biotite-in reactions. The biotite-in reaction shifts to higher temperatures and lower pressures and the clinozoisite out reaction line shifts to lower temperatures and higher pressures with increased alumina (Fig. 3). As a result biotite and clinozoisite do not appear in the same parageneses as in the original pseudosection (Fig. 2).

Dramatic changes occur with decreased potash. Consumption of chlorite, quartz, and white mica components produce the initial growth of sillimanite (isobaric heating). This is accompanied by growth of garnet and plagioclase. The biotite out and clinozoisite out reactions shift in the same way as shown for increased alumina (above). There are significant changes in the locations of other reactions. For example, the cordierite out reaction line shifts to lower temperature, and the stability field for staurolite is wider. In addition, paragonite is stable in the high pressure — low to medium temperature parts of this pseudosection.

Ferry, J. 1982. A comparative geochemical study of pelitic schists and metamorphosed carbonate rocks from south-central Maine, USA. Contributions to Mineralogy and Petrology, 80, p. 59-72.

Holland, T & Powell R. 1998. An internally consistent thermodynamic dataset for phases of petrological interest. Journal of Metamorphic Geology, 16, p. 309.

Powell, R., and Holland, T. HP98 dataset, http://www.esc.cam.ac.uk/astaff/holland/ thermocalc.html. THERMOCALC.

Spear, F. 1993. Metamorphic phase equilibria and pressure-temperature-time paths. Monograph. Mineralogical Society of America, Washington, D.C.