Sulfur CMAS glass standard

[Jeremy Wykes]

We have been doing a lot of sulfur analyses at ANU on synthetic glasses to calibrate sulfur solubility models. Like most places we use VG2 as a secondary standard for sulfur. I have felt that VG2 is not bad, but it is also not ideal, because it has ~2 wt. % Na₂O, so we know it will not be stable under the beam at the beam currents (50-100 nA) and the timescales (5-10 mins) we analyse for sulfur. Also, VG2 probably represents a mixture of S^2- and S⁶⁺.

So I have attempted to make a new glass to be used as a secondary standard for sulfur analyses. It is CMAS E composition (21.1 wt% CaO, 18.7 wt.% MgO, 9.9 wt.% Al₂O₃, 50.1 wt.% SiO₂) and should contain ~4000 ppm S when all S is dissolved as sulfide, and ~1000 ppm S when all S is dissolved as sulfate. I made several grams of the CMAS E starting mix, and we get sulfur into to by equilibrating ~5 mm diameter glass beads in a CO-CO₂-SO₂ gas stream at 1400 °C. We make 5-7 of these beads per 24 hour run. I currently have about 30 beads synthesised in 7 batches under reducing conditions, and Hugh O'Neill is producing a similar number of oxidised beads for me, so we will have material with the same glass major element composition but end member S^2- and S⁶⁺ speciation. I would imagine that 1/8th of a bead would be more than sufficient material to use as a standard, so potentially 240 aliquots of each valence state are available.

I am now at the stage where I would like to distribute some of this material to different EPMA labs. I have a number of questions that I need to answer:

(1) What is the S content and do different labs agree
(2) Is a single beads homogeneous in S content (previous work has demonstrated this is the case, otherwise the method would not be useful for calibrating sulfur solubility models
(3) Are all beads from a single batch identical in S content
(4) Are different batches identical in S content

Unfortunately, I did not make enough of the initial CMAS E mix, so we had to make another batch so not all batches use exactly the same starting material, but our weighing errors are likley to be insignificant, so major element composition should be identical, but this needs to be tested.

We also intend to take one batch an analyse that for S isotopic composition such that it can be useful as a SIMS standard.

I am looking for volunteers from labs with an interest in sulfur in glasses who would like to be part of the characterisation of this material.