I confuse with cation exchange and surface complexation model

[Jeonghwan Hwang]

Hi, This is Jeonghwan Hwang.
Thank you for reading.

I'm studying about cesium sorption process on illite.
In previous studies, some of them used cation exchange model to evaluate the sorptive properties, another used surface complexation model to solve, and the other used both of them and divided the effects of them by cesium concentration (they think that cation exchange mainly effects at high concentration).

The first of my question is about cation exchange model.
When the model set up, i define the EXCHANGE_MASTER_SPECIES (X, D, Z).
And I input the exchanged_species of them in my code and distribute them into solution with (EXCHANGE 1).

The code looks like below:
==============================================
EXCHANGE_MASTER_SPECIES
   X   X- #planar
   D   D- #TypeII
   Z   Z- #FES
   
EXCHANGE_SPECIES
   D- = D- #TypeII
   log_k 0.0
   Z- = Z- #FES
   log_k 0.0   
   X- = X- #planar
   log_k 0.0
   H+ + X- = HX #planar
   log_k 2.5
   K+ + X- = KX #planar
   log_k 0.5
   Na+ + X- = NaX #planar
   log_k 0.0
   Na+ + D- = NaD #TypeII
   log_k 0.0
   K+ + D- = KD #TypeII
   log_k 0.6
   H+ + D- = HD #TypeII
   log_k 3.6
   Ca+2 + 2D- = CaD2 #TypeII
   log_k 0.8
   Mg+2 + 2D- = MgD2 #TypeII
   log_k 0.6
   Na+ + Z- = NaZ #FES
   log_k 0.0
   K+ + Z- = KZ #FES
   log_k 1.5
   H+ + Z- = HZ #FES
   log_k 1.78 #Poinssot et al 1999, Fuller et al 2014
   Ca+2 + 2Z- = CaZ2 #FES
   log_k 0.8
   Mg+2 + 2Z- = MgZ2 #FES
   log_k 0.6
   
EXCHANGE 1
   X 6.5e-3
   D 1.6e-4
   Z 4.0e-6
-equilibrate solution 1
save exchange 1
end

EXCHANGE_SPECIES
   Cs+ + X- = CsX #planar
   log_k 1.2
   Cs+ + D- = CsD #TypeII
   log_k 3.6
   Cs+ + Z- = CsZ #FES
   log_k 7.0
end
==============================================

In this case, it seems to me that in this setup the exchange sites X, D, and Z are designed to exist in the form of ions in water.
On the other hand, surface complexation model looked as:

==============================================
SURFACE_MASTER_SPECIES #Bradury and Baeyens 1997
   Il_s  Il_sOH
   Il_aw  Il_awOH
   Il_bw  Il_bwOH
   
   SURFACE_SPECIES   
   Il_sOH = Il_sOH
      log_k 0.0
   Il_awOH = Il_awOH
      log_k 0.0
   Il_bwOH = Il_bwOH
      log_k 0.0
   Il_sOH = Il_sO- + H+
      log_k   -7.9
   Il_sOH + H+ = Il_sOH2+
      log_k   4.5
   Il_awOH = Il_awO- + H+
      log_k   -7.9
   Il_awOH + H+ = Il_awOH2+
      log_k   4.5   
   Il_bwOH = Il_bwO- + H+
      log_k   -10.5
   Il_bwOH + H+ = Il_bwOH2+
      log_k   6      
   Il_sOH  + Zn+2 = Il_sOZn+ + H+
      log_k   1.6
   Il_awOH  + Zn+2 = Il_awOZn+ + H+
      log_k   -2.7

SURFACE 1 # automatically calculate the surface charge
     Il_sOH        6.60E-07
     Il_awOH     1.32E-05
     Il_bwOH     1.32E-05

==============================================

What I'm most confused with is that both models have a nearly similar form.
We even found a study using the selectivity coefficient (k) defined in Exchange_species in combination with k of surface complexation.
Looking at the code alone, I find it hard to see what the computational differences are.
Of course, the two are different cases, so don't compare them directly.
Do you have any documentation or data to find out what computational properties these two have directly in PHREEQC?

I want to know whether i have to combine them or select the suitable model.
I still compare and read many papers, and I hope you can make a little better choice with your advice.
Thank you...

Respectfully,

Jeonghwan Hwang

[dlparkhurst]

The implementation of exchange requires exact charge balance such that all negative exchange sites are occupied by cations. Activities are assumed to be equivalent fraction of the sites occupied by a given exchange species, although it is possible to include a activity coefficients set to the same values as the aqueous counterparts.

The implementation of surface complexation does not require a charge-balanced surface, and in fact the surface usually develops a positive or negative charge. Activities are assumed to be are again assumed to be equivalent fractions. In addition, the diffuse double layer model includes a factor related to the surface potential; it is essentially accounts for the work to bring a charged species to the surface.

So, if you consider a change from low to high pH, the exchange model would react simply by a competition of H+ for sites (assuming a H-Exchange species is defined), occupying more sites at low pH and few sites at high pH.  Usually, only cation reactions are defined.

With surface complexation, the surface charge will vary from positive at low pH to negative at high pH, which will not only vary the H-Surface and OH-Surface species, but the potential term will have an effect on all ions such that anions will dominate at low pH and cations will dominate at high pH.

For more, you will need to find the appropriate literature or texts. Dzombak and Morel is the classic on diffuse double layer model, and Appelo and Postma is a text that uses PHREEQC for nearly all examples.

[Jeonghwan Hwang]

Thank you for kind comment.
I know the difference between CE and SCM, right now.

The most important thing, if i sufficiently understand, is that CE have to balance charge, and SCM don't have to balance because it mainly effected by surface charge.

When i used CE model to make cesium sorption, i found that every exchange species (ex FES, PS) were saturated by cesium. But it doesn't agreeable results with my experimental results.

Because i experienced cesium sorption on illite, the maximum sorbed concentration by Langmuir model showed only 20 or 50% of total CEC.
I think that this result means CE model is not suitable at high concentration.

For this reason, i think that i combined two models, CE at low concentration and SCM at high concentration.
I think that this approach sufficiently coupled with my experimental results.

Do you agree with my idea?
Thank you again for kind help.

Respectfully,

Jeonghwan Hwang