PACROFI VI - Electronic Program

Modeling the P-V-T-X Properties of Haplogranite Melt Inclusions During Heating and Cooling

James J. Student and Robert J. Bodnar

Department of Geological Sciences, Fluids Research Laboratory, Virginia Polytechnic Institute and State University, Blacksburg, VA, 24061, U.S.A

A model is presented which predicts the P-V-T-X evolution of crystalline melt inclusions during cooling and heating. In the model, hydrous haplogranite melts (Ab-Or-Qtz-H2O) are trapped as inclusions in quartz, then allowed cool and crystallize. Mass and volume relationships for melt-solid-volatile phases in the inclusions during cooling and heating are calculated based on available phase equilibria for the haplogranite system. As a result of second boiling during crystallization and cooling, the trapped melt releases exsolved H2O into the inclusion cavity. A positive volume change is associated with the release of water from the melt, and during cooling to solidus temperatures, the internal pressure within the inclusion can increase significantly. This overpressure can cause inclusion decrepitation during cooling. The calculated P-T path and the increase in internal pressure are highly dependent upon the value for the partial molar volume of H2Ototal in the melt used in the model, and several values have been selected (0.0, 14.5, 17.0, and 22.0 cm3/mole) which span the range of reported values in the literature.

Six different trapping scenarios are described. Both minimum and non-minimum haplogranite melt compositions are considered, and for each of these, H2O-saturated, H2O-undersaturated, and mixed trapping (H2O + H2O-saturated melt) are considered. Each of these six scenarios is further described for trapping pressures 500, 2000, 5000 bars. For minimum melt compositions, the pressure at the time of trapping fixes both the melt composition and temperature. Phase relationships during the homogenization process can distinguish each trapping scenario. For the H2O-saturated examples, total homogenization occurs with the simultaneous consumption of quartz, feldspar and vapor by the melt. The entrapment temperature (Tt) is determined by the temperature of complete melting of feldspar and quartz (Tmsil) and the vapor-melt homogenization temperature (Th) which are both equal to Tt. In the mixed scenario, total homogenization occurs by vapor-melt homogenization (Th) at an unreasonably high temperature and does not correspond to Tt. In the mixed trapping scenario, the temperature of complete melting for feldspar (Tmfeld) corresponds to Tt, at a temperature lower than vapor-melt homogenization (Th). Initial melt compositions for the H2O-undersaturated scenarios are determined for an activity of water in the melt equal to 0.5. The minimum entrapment temperature for these H2O-undersaturated melts can be determined by the temperature of complete melting of feldspar and quartz (Tmsil). In this case, Th occurs at a lower temperature than Tmsil, and Th defines the water saturated liquidus temperature.

As an example of the results predicted by the model, we have calculated the homogenization P-T path and phase volume relationships for an H2O-saturated minimum melt inclusion trapped at 500 bars and 782oC, using a partial molar volume of H2Ototal in the melt of 22.0 cc/mole. During heating, the P-T path follows the H2O liquid-vapor curve until the water phases in the bubble homogenize to the vapor phase at 365oC (Fig. A, point I). With continued heating the path follows the H2O isochore to the haplogranite minimum melt curve were first melting begins (Fig. A, point II). During heating from room temperature to the temperature of first melting, no changes in the phase volumes of the solids in the inclusion are recognized. Note, however, that the inclusion volume at room temperature is less than the original inclusion volume, owing to precipitation of quartz on the inclusion walls (Fig. B, I & II). The amount of melt shown at the first melting temperature (Fig. B, II) is greater than the actual amount of melt generated in order to facilitate its graphical representation. With continued heating, the P-T path follows along the H2O-saturated solidus and the internal pressure decreases as H2O dissolves in the melt (Fig. A, path II, III, IV, V). Similarly, the quartz and feldspar phases are reduced in size as they melt (Fig. B, III & IV). Total homogenization occurs at the trapping conditions with the simultaneous consumption of H2O (Th) and feldspar-quartz (Tmsil) by the melt. During cooling from trapping to ambient conditions, the P-T path followed by the inclusion is the reverse of the path followed during homogenization.