CO2 in fractional melting models

[Ri Cao]

Dear Paula,
This is Ri. Happy New Year and hope you have a good holiday break. Thank you very much for your advice in using Rhyolite-MELTS last year.

I just have a further question there:
Following your suggestions, I have successfully run the fractional melting modelling utilizing Rhyolite-MELTS. Basically, I run the batch crystallization first to find the solidus temperature of the system, and then get the temperature just a little bit above the solidus temperature and switch to fractional melting (FM) mode ('Fractionate liquids'), then run the FM with temperature step size < 3 oC.

However, I just found that if I add 0.2 wt.% CO2 into the system, I could only get liquid at the first step of the FM, which is equivalent to the amount of liquid of the last step of Batch crystallization. Then in the further step of FM, the liquids just disappear from the system at all.

I was wondering whether I could interpret this small amount of liquids at the first step of FM as ''carbonatite melts'' or this is an artefact? Any thoughts on this will be very helpful. Thank you very much.

Sincerely,
Ri Cao

[Paula]

Hi Ri,

Probably a bit of both. The model is not intended for small degree melts (see the answer to "What is rhyolite-MELTS?") and I think I'm right in saying that no experiments with carbonatites were used in development of versions 1.2.0 and 1.1.0. Rhyolite-MELTS can reproduce a carbonatite - silicate liquid miscibility gap if multiple liquids is switched on, but the location of this miscibility gap it not calibrated in PTX space.

If the composition of the small degree melt looks like carbonatite then you might interpret it as such, but I would not treat the temperature at which this melt is produced / exhausted seriously. Presumably if you continue the calculation then melting starts again at the CO₂-free solidus?

Paula

[Ri Cao]

Hi Paula,
Thank you very much for your response and clarification regarding the carbonatite melts problem.

I believe you are right that since the Rhyolite-MELTS modelling I am using is version 1.0.2, so probably it is possible to have carbonatite melts in the system for FM. I have another three naïve following up questions there:

1: '' silicate liquid miscibility gap if multiple liquids is switched on''

I was wondering if that's possible in Rhyolite-MELTS graphical interface software that I could control the mode between ''multiple liquids'' or ''single liquid''? I mean is that possible to switch off the ''multiple liquids'' button? If so, how shall I do that?


2:   ''Presumably if you continue the calculation then melting starts again at the CO2-free solidus?''

Yes, if I continue the FM calculation after finding the solidus temperature (batch melting procedure) but without CO2 situation, then the melts/liquids could start as normal.

3: Recently, I did another Batch melting modelling with a little bit different initial Venusian composition. After checking the output files (i.e., melts.out), it sounds like I got several unusual minerals  with smaller weight/mass (i.e., garnet: 0.0045 (gm); spene: 0.002 (gm), corundum: 0.001 (gm) (this one only exist at a higher pressure, 1 GPa) given that the total mass of liquid +solid in the system is 100 (gm)).

I was wondering shall I treat these unusual minerals as an artefact as well? Perhaps, they are here based on the chemical equilibrium of the system (i.e., similar to essenite we discussed before)? Any thoughts on this will be very helpful! Thank you very much and I look forward to your response.


Sincerely,
Ri Cao

[Paula]

If you have CO₂ in the system you should be using Rhyolite-MELTS 1.2.0. Any CO₂-bearing liquid with Ryolite-MELTS 1.0.2 will definitely be an artefact. For more details see:

https://melts.ofm-research.org/MELTS-decision-tree.html

A1. By default, the model runs in single liquid mode. There is a menu option under "Options" that allows you to turn "multiple liquids" on or off. Or you can add a line to your .melts file to turn it on:

Mode: multiple liquids

A3. I think so. If the model has run correctly then the final result will be in chemical equilibrium (although occasionally things misbehave and then it takes some trickery to get the right result...).

Best wishes,
Paula

[Ri Cao]

Dear Paula,
Thank you very much for your prompt response and clarification and that all make sense now.

Following your suggestions, I have run the Rhyolite-MELTS 1.2.0 in order to add CO2 in. I have a question there for the calculation procedure.

I am trying to run the batch melting for an Venusian initial composition. However, after adding 0.2 wt.% CO2, and when I trying to find the liquidus temperature of the system, the system keep searching the liquidus and do not stop, which reach to a surprisingly high value (i.e., 1956200 oC). Originally, if that's anhydrous condition in rhyolite-melts 1.0.2, then the solidus temperature is just approximately 1300 oC.

One possible way I was wondering is whether that make senses to skip this ''find liquidus temperature'' step and just nominate a start temperature and stop temperature to let the system run directly under version 1.2.0, which works.

In this case, after checking the output files and there is a file called fluid.tbl, which documented the mass of fluid we get at each step. After checking the chemical composition from the melts.out file and that said it is made of co2duan. Are there any practical meaning for this liquid? or I could just ignore this if I am doing the mineralogical calculation? Any thoughts on this will be very helpful! Thank you very much and I look forward to your response.

Sincerely,
Ri Cao

[Paula]

In the MELTS GUI, anything that is not silicate liquid is treated as solid. So Find Liquidus does not work if the system is fluid saturated... as you have found. MELTS for Excel has a Find Wet Liquidus function that gets around this issue. alphaMELTS 1.9 also has a workaround that I still need to transfer to alphaMELTS 2.

co2duan just means it is a pure CO₂ fluid. The "duan" part refers to the equation of state; for CO₂ fluid this is always Duan & Zhang (2006), but for H₂O there are different options, which is why there are multiple Rhyolite-MELTS versions). See Ghiorso & Gualda (2015) for more details. Both references are listed in this thread:

How is the Density of Fluid Phase Computed?

If you want the exsolved fluid to be extracted with the silicate liquid then turn on Fractionate fluids, either with the menu option or in the .melts file.

Paula

[Ri Cao]

Hi Paula,
Thank you very much for this detailed clarification.

I believe the issue for running CO2 on the 1.2.0 version has been resolved now.

Another question I have spotted just now is that, if I add H20 to the system (i.e., 0.2 wt.% H2O), the solidus temperature of the system would diminish compared with an anhydrous condition. However, when utilizing the 1.2.0 version to add 0.2 wt.% CO2 instead, the solidus temperature of the system does not change at all compared with the anhydrous case. I was wondering did I miss some fundamental understanding of this CO2 calculation? Thank you very much.

Sincerely,
Ri Cao 

[Ri Cao]

Dear Paula,
Sorry for my interruption again but I have another question there for choosing the ''multiple liquids'' section in utilizing Rhyolite-MELTS version 1.2.0.

Following your suggestions, I turn on the ''allow multiple fluids'' button in the Options section. When I tried to run the calculation, the Rhyolite-MELTS software just shut down immediately. All the other settings are the same as I did before with ''single liquid'' option (i.e., same FMQ, same start, and stop temperature, and same initial chemical composition).

I was wondering are there some problems with my running procedure here when adding ''multiple liquids''? I also tried to add a line manually in the .melts file but the problem is still there. Any thoughts on this will be very helpful, too! Thank you very much and I look forward to your response.

Sincerely,
Ri Cao

[Paula]

Hmm, that's probably a starting solution problem for solving the melt speciation reaction. I looked at this a bit a few years ago. If you send me your input .melts file (psmith AT gps.caltech.edu) that would be useful for debugging. Can't promise much but I'll give it a go.

Thanks,
Paula

[Ri Cao]

Dear Paula,
Thank you very much for your response.

I have sent you the input file. Please have a look through it.

Best regards,
Ri Cao

[Paula]

Got it, thanks. I'll try to get back to you early next week.

Nice talk at VMSG the other day!

Paula

[Ri Cao]

Dear Paula,
Sorry for my interruption again and hope you are doing well.

Just saw the MELTS workshop for Goldsmith this year (https://magmasource.caltech.edu/workshops/goldschmidt/). Just want to double-check with you whether this workshop will be a hybrid version or just an in-person workshop? Does that mean I should double-register for both Goldsmith and this workshop as well? Thank you very much and I look forward to your response.

Sincerely,
Ri Cao 

[Paula]

Hi Ri,

The workshop is in-person only, unless the conference ends up being completely virtual. More details:

https://magmasource.caltech.edu/workshops/goldschmidt/#latest
https://2022.goldschmidt.info/goldschmidt/2022/meetingapp.cgi/Session/3303

You can register for the workshop at the same time as you register for the conference, or you should be able to add it later. The in-person workshop is $65 for two days for early career researchers. Should the conference/workshop end up being virtual then the fees will be adjusted accordingly.

I'm not sure when registration closes for the workshops, but I sent a message to find out. It usually coincides with the close of standard online registration for the conference. (Note that Early Bird registration closes May 31st, 2022.)

Best wishes,
Paula

[Ri Cao]

Dear Paula,
Thank you very much for this clarification and will keep an eye on this registration.

Best regards,
Ri Cao