Finding a point of phase diagram with defined concentrations of some components

[PrestoConFuoco]

Hello.
I have some experimental points of isothermal phase diagrams of systems salt(1)-salt(2)-H2O.  And for comparing calculated data with experimental one, for example, i need to get the point with concentration of one salt identical to experimental, and concentration of second salt i will just compare.
So, the problem is - how to calculate solubility of salt(1) while constant concentration of salt(2)?

At the moment i managed to do it only for cases with no double salts and no crystallohydrates, for example for system NaCl - KCl - H2O at 25 C, where equilibrium phases are halite (anhydrous NaCl) and sylvite (anh. KCl), as amount of water in solution doesn't change during precipitation of solids, and so the concentration of unsaturated component.
And if crystallohydrates are precipitated, water is going to solid phase and the concentration of unsaturated component is rising! :(
Thus,  i need to calculate more complicated phase diagrams of salt-salt-water systems, and also more complicated 4- and 5-component systems.

Big thanks! And sorry a lot for my poor english.
Input file and appropriate database for mentioned Na-K-Cl-H2O system are attached.

[dlparkhurst]

I don't quite understand what you need to calculate. Your examples define a solution made with two salts, and then equilibrium phases finds equilibrium. That seems a reasonable physical experiment.

But you want the elements of one salt to have fixed concentrations? Because you measured them I assume, but you also measured all the constituents, so why not just calculate the SI of the second salt?

Here is an example that fixes Na and Cl and allows K to vary to obtain equilibrium. Seems sort of what you are asking.

Code: [Select]

SOLUTION
-units mol/kgw
Na 0.5
Cl 8.5
K  8.0 Sylvite 0
END


[PrestoConFuoco]

Well, i didn't have to deal with Saturation indices in my practice, i do not realize adequately what it is.
Where may i get information about this and its application in PHREEQC?

And about your solution definition. How can we vary only K+ not varying Cl- at the same time?

Okay, i tried your offered code and obtained very big percent error (~45%). This was quite reasonable.
Then i tried next version of it:

SOLUTION
-units mol/kgw
Na 0.5
Cl 8.5 Sylvite 0
K  8.0 Sylvite 0
END

This, surprisingly, turned into 11% percent error. What am i doing wrong?

Update: the next two enterings gave the same result which looks like it is true.

SOLUTION
   units mol/kgw
   Na 0.5
   Cl 8.5 Sylvite 0
   K  8.0 charge
   END

SOLUTION
   units mol/kgw
   Na 0.5
   Cl 8.5 charge
   K  8.0 Sylvite 0
   END

[dlparkhurst]

That is my confusion. If you fix Na and Cl, K will be known by charge balance, yet I thought you said you wanted to fix Na and Cl, but let the saturation index of sylvite vary.

Using sylvite in SOLUTION for two elements is not a well defined problem. Higher K and lower Cl (or the opposite) will also satisfy equilibrium with sylvite.

If you look at the result of the initial solution calculations (before  the RUN_CELLS calculations) for your original example, you will find saturation indices for halite and sylvite for each solution.

I think you said you were satisfied with the results of your original example, that defined charge balanced solutions and let halite and sylvite precipitate as needed. So, can you give an example of another system that you do not know how to solve? Please state which concentrations are fixed and which are variable (presumably because of precipitation of salts).

[PrestoConFuoco]

So, can you give an example of another system that you do not know how to solve? Please state which concentrations are fixed and which are variable (presumably because of precipitation of salts).

For example, system Na-Ca-Cl-H2O. Let's apply some entries to this problem:

Code: [Select]

SOLUTION 1
units mol/kgw
Na 0.5
Cl 16.5
Ca 8
EQUILIBRIUM_PHASES 1
Antarcticite 0 0
END
SOLUTION 1
units mol/kgw
        -water 1
Na 0.5
Cl 16.5
Ca 8
EQUILIBRIUM_PHASES 1
Antarcticite 0 0
ENDThis results with:

Elements           Molality       Moles

   Ca                7.229e+00   4.474e+00
   Cl                1.527e+01   9.448e+00
   Na                8.079e-01   5.000e-01

after implicit batch-reaction calculation. As you can see, Na molality increased from 0.5 to ~0.8 mol/kgw.

Then i tried next entry:

Code: [Select]

SOLUTION 1
units mol/kgw
Na 0.5
Cl 16.5 charge
Ca 8 Antarcticite 0
END

Code: [Select]

SOLUTION 1
    -water    1                 # kg
    -units mol/kgw
    K 2
    S 1
    pH 7 charge

EQUILIBRIUM_PHASES 1
    Arcanite 0 100 # 100 moles

REACTION_TEMPERATURE 1
    -5.0 105.0 in 23 steps

SELECTED_OUTPUT 1
    -reset   false
    -file                 K2SO4.sel
    -pH   true
    -temperature   true
    -molalities   K+ SO4-2 HSO4-

This worked well for system Na-K-Cl-H2O, but in this case it wrote some warnings:

WARNING: Maximum iterations exceeded, 200

WARNING: Maximum iterations exceeded, 200

WARNING: The program has failed to converge to a numerical solution.

The following equations were not satisfied:
ERROR:                   Ca solution phase boundary has not converged.    Residual: 1.076634e-01

And results:

Elements           Molality       Moles

   Ca                7.101e+00   7.101e+00  Equilibrium with Antarcticite
   Cl                1.470e+01   1.470e+01  Charge balance
   Na                5.000e-01   5.000e-01

Do not know if i can trust the result after such warnings.

Code: [Select]


SOLUTION 1
    -water    1          # kg
    -units mol/kgw
    Mg 1
    S 1
    pH 7 charge


[dlparkhurst]

You can normally trust results if there are only "Warning" messages, although you should look at what the message is.

"Error" messages mean that the calculation failed.

This seems like the correct calculation.

Code: [Select]

SOLUTION 1
units mol/kgw
Na 0.5
Cl 16.5
Ca 8
EQUILIBRIUM_PHASES 1
Antarcticite 0 0
END

I do not see the problem with the concentration of Na increasing because of loss of water. It is a real effect. What is it that you want to happen? Add water until Na is again 0.5 mol/kgw? Tell me what you think should happen. I assume you expect antarcticite to precipitate; do you not expect water to be lost from solution?

[PrestoConFuoco]

I do not see the problem with the concentration of Na increasing because of loss of water. It is a real effect. What is it that you want to happen? Add water until Na is again 0.5 mol/kgw? Tell me what you think should happen. I assume you expect antarcticite to precipitate; do you not expect water to be lost from solution?

Yes, i understand fully that effect. My goal is to evaluate solution's saturated composition with fixed concentration(s) of some component(s) in it.
This is useful for calculating, for example, isothermal phase diagram of quaternary system with fixed concentration of one component. For example, for calculating Li-Na-K-Cl-H2O system with KCl or NaCl constant in solution.

[dlparkhurst]

I don't see how it is physically possible. With the sylvite, halite example, if you fix Na and Cl, you also fix the K by charge balance. So you need to let Cl adjust, include another ion, or allow charge imbalance to be able to allow K to be adjusted. Are you willing to allow one of these conditions?

[PrestoConFuoco]

I don't see how it is physically possible. With the sylvite, halite example, if you fix Na and Cl, you also fix the K by charge balance. So you need to let Cl adjust, include another ion, or allow charge imbalance to be able to allow K to be adjusted. Are you willing to allow one of these conditions?

Only Na is fixed, K is varying and Cl amount is defined by charge balance.
In fact, i want to get a solution with 0.5 molality NaCl, which is saturated by KCl. It became possible by your offer:

Code: [Select]

SOLUTION
   units mol/kgw
   Na 0.5
   Cl 8.5 Sylvite 0
   K  8.0 charge
   END
Okay, after calculation here will be 0.5 molality of NaCl and saturated concentration of KCl.
This is almost perfect. Next step is to get these conditions not entering that only Sylvite can be in equilibrium
with solution. I mean, i need to calculate system using general equilibrium condition.

Calculation with general equilibrium condition is possible by following code:

Code: [Select]

SOLUTION 1
units mol/kgw
Na 0.5
Cl 16.5
Ca 8
EQUILIBRIUM_PHASES 1
Antarcticite 0 0
END
Here  other phases too can be specified, not only antarcticite. This way of calculating has a priority for me.
But this way, i said, concentration of Na is not constant.
All what i need is to calculate batch reaction in the second mentioned way with fixed NaCl concentration.

[dlparkhurst]

Seems like you want:

SOLUTION 1
   units mol/kgw
   Na 0.5
   Cl 16.5 charge
   Ca 8     Antarcticite
        K  1     Sylvite
END

[PrestoConFuoco]

Code: [Select]

SOLUTION 1
units mol/kgw
Na 0.5
Cl 16.5
Ca 8
EQUILIBRIUM_PHASES 1
Antarcticite 0 0
END
THIS is what i want. I would prefer using This type of input.
But i want some concentrations to be constant during calculations, that use this input.

[dlparkhurst]

If that is what you want, then so be it.

Here are my last suggestions:

The first allows you to fix the activity of Na+. This will not give you the concentration you want, but may get you close if you can estimate the activity coefficient for the solution.

Code: [Select]

PHASES
Fix_Na
Na+ = Na+
log_k 0
SOLUTION 1
units mol/kgw
Na 0.5 
Cl 16.5
Ca 8     
END
USE solution 1
EQUILIBRIUM_PHASES
Antarcticite 0 0
Fix_Na  2.0 NaCl 10
END

This calculation adjusts the Na concentration incrementally. You can pick out the point where Na concentration crosses the desired concentration.

Code: [Select]

SOLUTION 1
units mol/kgw
Na 0.5 
Cl 16.5
Ca 8     
END
USE solution 1
REACTION
NaCl -1
0.4 mol in 10 steps
EQUILIBRIUM_PHASES
Antarcticite 0 0
USER_GRAPH 1
    -axis_titles            "NaCl removed, mol" "Na, mol/kgw" ""
    -initial_solutions      false
    -connect_simulations    true
    -plot_concentration_vs  x
  -start
10 graph_x rxn
20 graph_y TOT("Na")
  -end
    -active                 true
END