Experimental investigations of the partitioning of elements between coexisting melt and aqueous fluid often employ bulk analytical techniques coupled with chemical and/or mechanical separation to determine the composition of one or both phases. In our experiments the fluid phase coexisting with the melt is encapsulated at elevated temperature and pressure in vesicles in the melt and then quenched as fluid inclusions in the glass. Glass and inclusions are analysed using the Laser Ablation-ICP-MS technique. In this method the laser drills through the glass matrix towards the fluid inclusion. Thus, within one analysis, glass immediately next to the fluid inclusion and the inclusion itself are sampled, and the mechanical or chemical separation of both phases prior to analysis is no longer necessary.
Sets of experiments have been performed at the following conditions in rapid-quench cold-seal vessels:
(1) 850°C and 2 kbar using either pure or doped haplogranitic glass powder (doping with Nb, Cs, Mo and W, each at ~ 1000 ppm) and NaCl-solutions at 2.5 or 5 wt% NaCl as starting materials.
(2) 1050°C and 2 kbar in TZM vessels using a haplogranitic glass powder doped with 20 elements, each at ~ 1000 ppm. The concentration of the solution varied from 2.5 to 5 wt% NaCl.
(3) 850°C und 2 kbar using a glass powder doped with Cs, Nb, Mo, and W at 1000 ppm and aqueous solutions containing a mixture of salts (NaCl, KI, KBr) at different concentration levels up to a total salinity of ~ 20 wt%.
The experimental durations were mostly in the range of 10-15 days and one experiment has been carried out for a duration of 26 days.
Analyses were performed using the ELAN 6000/Perkin Elmer ICP-MS in combination with a 193 nm ArF excimer laser at the Institut für Isotopengeologie und Mineralische Rohstoffe, ETH Zürich. NIST SRM 610 glass and halite were used as external standards for the glass matrix and the fluid inclusions, respectively. Si, determined by electron microprobe, and Na calculated from microthermometric measurements, served as internal standards for the glass and the inclusions. Br and I analyses were calibrated externally using Br- and I-glasses. Cl, Br, and I in the fluid inclusion analyses were calibrated using a standard solution.
To quantify the fluid inclusion signal in the spectra the integrated intensity values have to be corrected for the contribution of the matrix to the signal intensity. This correction requires at least one highly compatible matrix element. We applied a correction based on the matrix element Si. This effect strongly limits the detection of elements in the fluid inclusions. Highly compatible matrix elements as Nb, Sm, Hf, Tb, Ba, Mg, Ca, and Nd could be analysed in the glass matrix but not in the inclusion, so for these elements no partition coefficients could be determined. Calculated fluid-melt partition coefficients for the major elements Na and K and the trace elements Li, B, Rb and Cs agree with literature data. The values determined for Mo and W range from 5.4 to 25.3 and from 1.8 to 8.2, respectively. These values are at least 10 times higher than previous results for the same system, despite the fact that for Mo and W equilibrium partitioning was apparently not obtained in our experiments.
The study of the halogens shows that both Br and I do not affect the element partitioning between fluid phase and melt. The calculated partition coefficient for Cl, DCl = 37 – 91, agrees with literature data. Partition coefficients for Br and I, DBr =39 - 130 and DI = 8 – 48, vary mainly due to apparent Br heterogeneities in the glass matrix and the large scatter of the I concentration values in the matrix.