How to Respresent Solid Solution Minerals in PHREEQC?

[swhit]

Hi,

I'm wondering if there is a consensus on the best way to represent solid solution minerals in a forward model.

I need to include plagioclase that is roughly 25% Anorthite 75% Albite, and also have several other solid solution minerals such as Fe-Mg Hornblendes, and Calcite-Dolomite, and Fe-Mg micas.

I see two options (but am open to any other approaches):

1.  Include representative portions of the end-members.

2.  Use solid solutions.

The issue I see for option 2 is that I am not sure where to find thermodynamic data for the various mixtures.

I see a related issue for option 1 in that if one end-member has a larger equillibrium constant, it would dominate the dissolution relative to what occur in reality with the solid solution.

What are the pros and cons of each approach?  Anything else I should consider?

Thanks,

Thanks,

[dlparkhurst]

I don't think you want to use SOLID_SOLUTIONs for phases that are only dissolving. You can think of solid solutions completely dissolving and re-precipitating in equilibrium with the solution, which is not reasonable for your minerals.

The minerals could be considered as fixed-composition solids that dissolve stoichiometrically, in which case you would need to come up with a log K if they are to approach equilibrium, either assuming an ideal solid solution log K related to the proportions of end members, or if available, based on non-ideal solid solution.

Given that equilibrium will probably not be approached for many of these minerals, you may simply be looking at kinetic dissolution. You could use the end-member proportions. On the other hand, anorthite probably preferentially dissolves relative to albite, in which case, you could dissolve each end member at a different rate.

So, no, I don't have a consensus even in my own thinking about how I would approach the problem. If you have a natural water, I would probably try inverse modeling like Garrels and McKenzie in example 16 to see if you can gain any insights.

[swhit]

I think that calculating the ideal (or non-ideal, if available) Kss for the minerals and allowing them to dissolve only will be my initial approach, except for Calcite-Otavite, which I plan to use SOLID_SOLUTIONS for.

However, I have been doing some conceptual examples to make sure I understand what is going on, and I have a question about something I am seeing with  the following code, where I dissolve a fixed composition solid solution of otavite and calcite, and allow calcite to precipitate if oversaturated (I recognize that SOLID_SOLUTIONS is probably the more realistic way to represent this mineral pair), but am struggling to understand what I'm seeing in the model output.

SOLUTION 2
pH 7
pe 4

EQUILIBRIUM_PHASES 2

Otacal 0 1
Calcite 0 0

PHASES

Otacal
   Cd0.01Ca0.99CO3 = 0.01Cd+2 + 0.99Ca+2 + CO3-2
   -log_k   -8.50

END


My confusion is that this simulation precipitates calcite, when KOtacal is < KCalcite.  In my mind, calcite saturation shouldn't be reached (although I admit the K values are very close).  I thought maybe this was a numerical issue, so I adjusted -convergence_tolerance in KNOBS to 1e-12 and still saw calcite precipitation.

Here is the relevant output:

Reaction step 1.

Using solution 2.   
Using pure phase assemblage 2.   

-------------------------------Phase assemblage--------------------------------

                                                      Moles in assemblage
Phase               SI  log IAP  log K(T, P)   Initial       Final       Delta

Calcite           0.00    -8.48     -8.48    0.000e+00   9.020e-05   9.020e-05
Otacal            0.00    -8.50     -8.50    1.000e+00   9.998e-01  -2.145e-04

-----------------------------Solution composition------------------------------

   Elements           Molality       Moles

   C                 1.243e-04   1.243e-04
   Ca                1.221e-04   1.221e-04
   Cd                2.145e-06   2.145e-06

----------------------------Description of solution----------------------------

[dlparkhurst]

It is subtle, but one way to look at it is to break down the IAP for the two phases as follows:

Code: [Select]

IAP(Cal)     = LA(CO3-2) + 0.99LA(Ca+2) + 0.01LA("Ca+2")
IAP(Otacal) = LA(CO3-2) + 0.99LA(Ca+2) + 0.01LA("Cd+2")

Using phreeqc.dat, 0.01LA(Ca+2) = -0.040 and 0.01LA(Cd+2) = -0.062.

So, IAP(Cal) is greater than IAP(Otacal), and it is enough that calcite should precipitate.

[swhit]

That makes sense, thanks!