SIT - Specific Ion interaction Theory
This model was adopted by the OECD-NEA thermodynamic database (TDB) project,
and it is described in its published reviews and books, see
list of references.
In the SIT model the activity coefficient
of an ion i with electric charge Zi is given by:
| |
log γi = |
−Aγ Zi2
(√I /
(1 + åBγ √I
))
+ ∑j ε(i,j) mj
|
where I is the ionic strength;
Aγ and Bγ are the
Debye-Hückel
parameters, and the ion-size parameter å is selected to give
åBγ = 1.5
(mol/kg)−½ at 25°C.
ε(i,j) are the specific interaction parameters
between an ion i and another ion j of opposite charge sign;
that is, the summation is taken over all ions j such that Zi×Zj > 0.
Often the ε-values are assumed to be independent of
the ionic strength. A fundamental principle of the SIT model is that
ε(i,j) = ε(j,i),
that is, the specific ion interactions are symmetrical. When making a diagram,
the values of ε(i,j) are read from a
separate file (see below).
The temperature dependence of
åBγ (= 1.5
(mol/kg)−½ at 25°C) and of the
ε(i,j) values is discussed by
Lemire et al. (2020) (Section B.1.2),
as well as by Bretti et al. (2004),
Xiong (2006) and
Plyasunov and Popova (2013).
See Table
of Debye-Hückel parameters at different temperatures.
See also temperature in DATABASE/SPANA.
The activity of water (H2O) is given by:
| |
log aH2O = |
− Φ ∑mk /
(ln(10) 55.508)
|
where k are all solute species in the aqeuous solution,
and Φ is the osmotic coefficient, calculated using
the SIT as follows:
| |
Φ = |
1 −
(2 AΦ I3/2
(1/∑mk)
σ(åBγ √I))
+ (ln(10)/∑mk)
∑i ∑j
ε(i,j) mi mj
|
where AΦ = (1/3) ln(10) Aγ and
i are cations or neutral species
and j are anions,
and
| |
σ(x) = |
(3/x3) ((1+x) − 1/(1+x) − 2 ln(1+x))
|
Note: in the calculations the values of log γi
are cut-off to be in the range ±(20 Zi2), and
log aH2O is cut-off to be in the
range ±4. Also the concentrations mi
are cut-off at 100 molal.
The ionic strength is nevertheless calculated
without restrictions in the values of mi.
The ε-values:
When making a diagram the ε-values are first searched in a file named:
SIT-coefficients.dta
This is a text file that is distributed and stored in the installation folder.
The format of the data in this file is described in the file itself.
If you wish to add or to change ε-coefficients you can
create a copy of the file (with the same name) and store it either in your user directory
(system dependent, shown with the Help / About menu), or in the
working directory (where your input data file is located). Any
ε-values in the working directory will
supersede those in the user directory, and the data in the user directory will
supersede those in the installation directory.
If the first non-comment line in any of the three ε-files
described above is equal to
NoDefaults (not case-sensitive) then all undefined
ε-values will be set to zero. Otherwise all undefined
ε-values will be set to the default
values as follows:
Default ε-values between any ion and either
Na+, Cl− or
ClO4− are assigned with the
estimation method proposed by Hummel (2009):
ε(Na+,XY)
= 0.05 Y
ε(MZ,Cl−)
= −0.05 + (0.1 Z)
ε(MZ,ClO4−)
= 0.2 Z
where Z and Y are the electric charges for the cation
(Z>0) and anion (Y<0), respectively. For example,
ε(Sr2+,Cl−) = 0.15, and
ε(Na+,MoO42−) = −0.1.
Any ε-value not yet defined is set to:
ε(MZ,XY)
= 0.15 + 0.15 (Z + Y)
where Z and Y are the electric
charges for the cation (Z>0) and anion (Y<0), respectively.
For example:
ε(K+,MoO42−) = 0, and
ε(H+,OH−) = 0.15.
For neutral species ε-values not yet defined
are set to:
ε(N0,MZ) =
ε(N0,XY) = 0.1
where N is a neutral species, Z and Y are the electric
charges for the cation (Z>0) and anion (Y<0), respectively.
For example:
ε(CO2,Na+) =
ε(HAc,OH−) = 0.1
Convergence problems
The nature of the SIT equation is such that in concentrated solutions
some of the values of
log γi become linearly dependent on the concentrations being calculated.
This may create convergence problems in the calculations, especially when making diagrams where the pH
is very high or very low, or where the redox potential,
Eh, is outside the stability range of H2O.
If the redox potential is very high or very low, then unrealistically large
amounts of O2(aq) or of of H2(aq)
are calculated, respectively. As a consequence, the calculated
∑mk in the equations above
also becomes unrealistically large.
References
-
Bretti C, Foti C, Sammartano S (2004) Chem. Speciat. Bioavailab., 16, 105–110. doi: 10.3184/095422904782775036
-
Giffaut E, Vitorge P, Capdevila H (1994) J. Alloys Compd., 213–214, 278–285. doi:10.1016/0925-8388(94)90916-4
-
Grenthe I, Fuger J, Konings R J M, Lemire R J, Muller A B, Nguyen-Trung C, Wanner H
(1992) Chemical Thermodynamics of Uranium. Amsterdam, The Netherlands: North-Holland, Elsevier Sci. Publ.
-
Grenthe I, Plyasunov A (1997) Pure Appl. Chem., 69, 951–958. doi: 10.1351/pac199769050951
-
Grenthe I, Plyasunov A V, Spahiu K (1997)
Estimations of medium effects on thermodynamic data. In: Modelling In Aquatic Chemistry
(I. Grenthe & I. Puigdomenech, eds.), pp.325-426. Paris, France:
OECD Nuclear Energy Agency (NEA).
-
Hummel W (2009) Ionic strength corrections and estimation of SIT ion interaction
coefficients, Report PSI-TM-44-09-01. Paul Scherrer Institut, Switzerland.
-
Lemire R J, Palmer D A, Taylor P, Schlenz H (2020) Chemical Thermodynamics of Iron. Part 2.
Paris, France: OECD Nuclear Energy Agency (NEA).
-
Plyasunov A V, Popova E S (2013) J. Solution Chem., 42, 1320–1335. doi: 10.1007/s10953-013-0029-3
-
Xiong Y (2006) Geochem. Trans., 7, 4. doi: 10.1186/1467-4866-7-4
|
|
Models for activity coefficients
Comparison between experimental data and models