Hi everyone,
I tried to use Rhyolite-MELTS 1.2.0 to simulate the LLD evolution in my study area. It works quite well in most of the major oxides like MgO, FeO, Al2O3, CaO, Alkalis, and SiO2. Our phase-equilibria experiments also show a similar LLD pathway under similar conditions.
However, the main issue is that the concentration of MnO and P2O5 keep increasing throughout the magma evolution from my parental composition to rhyolite composition. This problem was also pointed out by some reviewers for my submitted paper. The phase assemblage does not contain any manganese- or phosphorus-rich minerals (the starting composition is a high-Mg andesite with plag + cpx + opx + ol + spinel + ilmenite). I am not sure if this is the reason.
Thank you!
The main problem is that there aren't many solid phases in the MELTS models that take MnO or P2O5.
The only Mn-bearing solid phases are olivine and rhombohedral oxide (ilmenite). If ilmenite is not joining the assemblage early enough, you could try playing around with the fO2.
The only P-bearing solid phases are (hydroxy)apatite and whitlockite. MELTS underestimates the stability of apatite, mainly because of the lack of fluorine and chlorine in the system. There is a F, Cl-bearing version of MELTS under development. At the moment the most recent citation is this:
[CITATION] MELTS+ DEW: Modeling major element+ Cl+ F+ S phase equilibria, redox reactions and elemental partitioning in magmatic-hydrothermal systems (https://conf.goldschmidt.info/goldschmidt/2023/meetingapp.cgi/Paper/13890)
MS Ghiorso (https://scholar.google.com/citations?user=_DOXBSwAAAAJ&hl=en&oi=sra), S Matthews (https://scholar.google.com/citations?user=JUDrd6sAAAAJ&hl=en&oi=sra)... - Goldschmidt 2023 ..., 2023 - conf.goldschmidt.info
Look out for more developments in the next few months.
Paula