1-atm experiments with alkaline melts: problems, solutions and some results

Borisov A.A.

Institute of Geology of Ore Deposits, Petrography, Mineralogy and Geochemistry

Russian Academy of Sciences, Moscow, Russia; e-mail: aborisov@igem.ru

A vertical tube furnace with controlled oxygen fugacities in combination with a loop technique is a standard experimental design for one atm studies in geochemistry, cosmochemistry and petrology. However, the investigation of alkalis-containing system, especially at low fO₂ values, is very limited. The main reason is the loss of Na and K in open systems where the oxygen fugacity is imposed by a gas mixture (CO/CO₂, H₂/CO₂ etc). Therefore, a new experimental device CSLT (crucible supported loop technique) was designed to prevent Na losses of experimental charges primarily for experiments at low oxygen fugacities (Borisov et al., 2006). Later CSLT was modified to allow not only preventing alkalis losses, but also saturating of initially alkalis-free melts with Na, K and Rb.

In brief, a single or a few sample loops place above alkali-rich melts in a quartz crucible. A lid loosely covers the crucible, so that exchange with the surrounding gas is possible. Exchange with the furnace atmosphere is necessary for controlling fO₂ inside the crucible. The alkali-rich atmosphere inside the crucible allows maintenance of alkali concentrations in the loop samples in the weight-percent level for hours or even days. In our practice we have used two types of crucibles. Small crucibles (9.3 mm inner diameter, 20 mm height) were used for a single loop and larger crucibles (15 mm inner diameter, 22 mm height) for a few loops.

CSLT has some limitations, however a few experimental studies in different areas of earth science were conducted:

1.) We have experimentally studied the effect of Na₂O on the solubility of Fe, Co, and Pd (Borisov et al., 2006). All solubilities clearly decrease (correspondent activity coefficients increase) with increasing Na₂O-contents. The increase of FeO activity coefficients with increasing sodium found in this study may provide an explanation for decreasing FeO with decreasing degrees of partial melting in experimentally produced mantle melts (Hirschmann et al. 1998), as melts at low melting degrees are sodium-rich.

2.) We have reported experimentally determined 1 atm olivine/melt D^Na partitioning data for low fO₂, a variety of melt compositions and a temperature range of 1325-1522ºC (Borisov et al., 2008). We demonstrated that high-current electron microprobe analyses (EPMA, I = 500 nA, 600 s on the peak) allow quantitative determination of Na₂O in olivine down to about 10 ppm. The mean olivine/melt D^Na from 12 experimental runs is 0.0031╠0.0007 (1σ). This is the recommended value for low pressures and a wide range of natural compositions.

3.) We have found relative activity coefficients of alkalis in multicomponent silicate melts (Borisov, 2008; Borisov, submitted). Model melts (Di-An eutectic modified with variable amounts of silica and Ca₂Si₃O₈ modified with variable amounts of alumina) were experimentally saturated with Na, K and Rb at temperatures of 1300√1470° C. Either binary Na₂O-SiO₂ and K₂O-SiO₂ or complex Na₂O-K₂O-SiO₂ and Na₂O-K₂O-Rb₂O-SiO₂ melts were used as alkalis sources. It was shown that alumina content in many cases has little influence on the degree of silicate melt saturation with alkalis. In contrast, silica content strongly affects the solubility of alkalis in silicate melts. Model calculations showed that the perfectly mobile behavior of alkalis leads to a potassium specifics in basalts contaminated by crustal materials and, correspondingly, a sodic specifics in silicic melts from crustal xenoliths contaminated by the enclosing basaltic melt.

References:

Borisov A.A. Experimental investigation of K and Na partitioning between miscible liquids. // Petrology. 2008. Vol. 16. P. 552-564.

Borisov A.A. Dependence of alkalis activity coefficients on silica and alumina contents in silicate melts: an experimental investigation. // Petrology. Submitted.

Borisov A., Lahaye Y. and Palme H. The effect of sodium on the solubilities of metals in silicate melts. // American Mineralogist. 2006. Vol. 91. P. 762-771.

Borisov A., Pack A., Kropf A. and Palme H. Partitioning of Na between olivine and melt: An experimental study with application to the formation of meteoritic Na₂O-rich chondrule glass and refractory forsterite grains. // Geochimica Cosmochimica Acta. 2008. Vol. 72. P. 5558-5573.

Hirschmann M.M., Baker M.B. and Stolper E.M. The effect of alkalis on the silica content of mantle-derived melts. // Geochimica et Cosmochimica Acta. 1998. Vol. 62. P. 883-902.