This summer I examined carbonate melt and the relationship between carbonate and silicate melt in mantle conditions through high pressure and temperature experiments.  I worked in the University of Minnesotaís Experimental Petrology Laboratory under the advice of Marc Hirschmann.  I used a piston cylinder machine to expose various samples of rock or carbonate to mantle condition pressures and temperatures, then polished and analyzed the different run products with the electron microprobe.

The carbonate with trace element project consisted of making a composition of carbonates that could be tested to address previous research's problems with accurately analyzing the composition of quenched carbonate matte.  These problems occur because carbonates form a matte of crystals when their melting is suddenly stopped, whereas silicate based experiments produce a more easily analyzed glass when quenched.  One of the biggest problems in analysis is deflated sodium values, which are thought to be caused by the temperamental nature of polishing a carbonate matte for analysis.  The carbonate composition I used is based after Wallace and Green (1988), but with the trace elements increased tenfold.  The composition was verified using a mass ICP-MS.  My first carbonate experiment was put at 3GPa at 1175º C for 6.5 hours.  The second run was at 3GPa, 1200º C for 19 hours.  The experiments produced a matte of quenched dolomitic crystals. Pictures of both experiments were taken on the electron microprobe, and an oxide analysis was run on the second experiment.
To figure out the optimum conditions for carbonate analysis, we ran a variety of electron microprobe currents with differing surface areas.  The focus of the microprobe tests was to analyze the major oxides, including Na2O, within the dolomite matte of the experiment.  We felt that the 50 micron spot with a 20 nA current gave us the most consistent results in analyzing the quenched melt.

The second half of my summer was spent looking at the transition between carbonate and silicate melt in carbonated eclogite.  The Salt Lake eclogite used for this series of experiments is from a xenolith in Hawaii.  It had been ground to a powder with five weight percent carbon dioxide added.  Work on the same sample has previously been completed to determine the solidus (Dasgupta et al., 2003).  To continue with melt experiments, I am trying to determine the transition between carbonate melt and silicate melt, more specifically whether the two melts mix together or occur separately.

The pressure of 3GPa and a variety of temperatures ranging from 1225ºC to 1400ºC were used on the eclogite samples to create melt.  The eclogite was quenched at pressure and temperature, and then analyzed.  Photos were taken with the backscatter feature of the electron microprobe.  The images show us that melt within the eclogite begins as carbonate.  As temperatures increase, silica becomes more and more soluble within the carbonate melt.  At 1315º C we see two separate melts forming, one with a carbonate composition and one as a silicate.  At the highest temperature, 1400º C, we see that only one melt formed because of the mixed textures and lack of grain boundaries. Upon being quenched, the mixed texture separated into a silicate glass and carbonate crystal matte.

References:
Dasgupta R., Withers A.C., Hirschmann M.M. (2003).  Generation of carbonatitic melt through partial melting of carbonated eclogite under mantle conditions. Abstract volume, 4th Eurocarb Workshop. 15-17.

Wallace M. E., Green D. H. (1988).  An experimental determination of primary carbonatite magma composition.  Nature.  335, 343-346.
 

Back