Post entrapment crystallisation of melt inclusions

[mc12g08]

Hello there,

I want to correct for post entrapment crystallisation of my melt inclusions, (e.g. Kress and Ghiorso, 2004). I am trying to follow this approach, in order to calculate the original melt composition from my partially crystallized melt inclusions (hosted in feldspar), but I am bit confused about how exactly this is done using the MELTS program. I basically just need to do reverse fractional crystallization along a liquid line of descent.
I am using RhyoliteMELTS (version 1.0.2) interface through a virtual box/Lubuntu system on my windows laptop and I have also downloaded the new Excel version. Is it possible to do reverse-fractional crystallization with either of these (I am guessing not the excel version since this does not have isochoric capabilities)? I understand that I have to enter both my measured inclusions and host compositions, but in the versions I can see only one place to input my compositions, so I am wondering if I am missing an extra algorithm/plugin.

Any advice on how to do this would be really appreciated.

Many thanks,

Mike

[Paula]

Hi Mike,

Sorry for the delay in replying. At the moment, MELTS_Excel doesn't have any open system behaviour enabled. The only interface that has automated, albeit non-unique, ways to do reverse fractionation is the alphaMELTS one. It's included in the VirtualBox version but the Rhyolite-MELTS thermodynamic model is not available in alphaMELTS (except in alphaMELTS 2, which doesn't have the reverse fractionation / Amoeba routines yet). I've never tried the fractionation correction in isochoric mode, but in principle it should work. I expect it would be better for olivine hosted inclusions, or any mafic phase... your mileage may vary for feldspar.

I have also been looking into the Kress and Ghiorso (2004) algorithm. The Brimstone website that hosts the software was down last week, though it appears to be up again now. The textEntrapment.linux executable from there will run in the 32-bit version of the VM alphaMELTS. You will need to install libstdc++5 - let me know if you want to go that route. There are instructions for using the software here.

In the meantime, Mark Ghiorso kindly found his copy of the Kress and Ghiorso (2004) software. I am trying to build versions that will run on the current 32-bit and 64-bit VM alphaMELTS (without having to install any extra libraries). There is quite a bit to update as some compiler features that were available 10 years ago are no longer... Let me know if you are using the 32-bit or 64-bit VM and I'll send you a copy when I'm done. I'd like to test whether the new executables behave in the same way as the downloaded one, and also whether alphaMELTS, which uses a slightly different strategy, gives similar results. Perhaps you could post or e-mail me a couple of your example compositions so that I can try them out, please?

More soon,
Paula