Jökull - 01.01.2010, Blaðsíða 168

A. Stefánsson
bilities were selected. The key aqueous species in the
data set included H2O, OH−, H+, Na+, K+, Mg2+,
Ca2+, Al(OH)4−, Fe2+, Fe(OH)4−, H4SiO4(aq),
Ti(OH)4(aq), H2S(aq), SO42−, HCO3−, Cl−, F−
and H2(aq). Their Gibbs energy values were selected
from Supcrt92 slop07.dat (Johnson et al., 1992), Gun-
narsson and Arnórsson (2000), Benezeth et al. (2001),
Ziemniak et al. (1993), Hill (1990), Diakonov et al.
(1999) and Marshall and Franck (1981). Values for
the apparent Gibbs energy formation of secondary
minerals were selected from the data sets given by
Holland and Powell (1998), Robie and Hemingway
(1995), Neuhoff (2000), Arnórsson and Stefánsson
(1999), Fridriksson et al. (2001) and Stefánsson and
Gíslason (2001).
The saturation of the basaltic glass was estimated
according to its dissolution mechanism proposed by
Oelkers and Gíslason (2001) and Gíslason and Oelk-
ers (2003). The mechanism assumes the solubility of
the basaltic glass to be controlled by a leached layer
enriched in Al and Si that may be approximated as a
mixture of amorphous silica and aluminium hydrox-
ide.
Reaction path modelling calculations were carried
out to simulate the interactions of basaltic glass with
water. The initial water composition was either that
of pure rainwater from Langjökull, Iceland, or dilute
non-thermal groundwater from Thingvellir, Iceland,
that had gained some alkalinity. The water compo-
sition is given in Table 1.
The initial water was allowed to react with basaltic
glass in accordance with the dissolution rates reported
by Gíslason and Oelkers (2003) and saturated sec-
ondary minerals in each step allowed to precipitate.
Commonly, 0.1 to 1 mol of basaltic glass (∼10 to
∼100 g) was dissolved in 1 kg of water in >1000 steps.
The composition of the basaltic glass used for the cal-
culations is given in Table 2 and the secondary miner-
als incorporated in the calculations are given in Table
3. The effect of acid supply was conducted by addi-
tion of CO2 (weak acid) and H2SO4 (strong acid) to
the starting solutions with concentrations of 0.1, 1 and
10 mmol/kg.
The thermodynamic properties for phyllosilicates
and carbonate solid solutions were calculated using
the approach of Tardy and Fritz (1981), assuming
ideal solid solution behaviour and, in the case of phyl-
Table 1. Chemical composition of the starting water
composition used in the model calculations. The con-
centrations are in µmol/kg. – Upphafleg efnasetnings
vatns sem notast var við í líkanreikningum. Efnastyrk-
urinn er gefin í míkrómól/kg.
Vellankatla ground water Langjökull precipit.
Thingvellir, Iceland a Iceland b
pH/◦C 7.54/22 6.05/20
Si 256 3.3
Na 269 61.1
K 11.9 1.01
Ca 71 0.7
Mg 38 6.7
Al 1.09 0.046
Fe 0.16 0
Cl 120 60.2
CO2 354 55.1
SO4 15 0
a Gysi and Stefánsson (2010).
b Pogge von Strandmann et al. (2008).
Table 2. The chemical composition of Stapafell bas-
altic glass (after Oelkers and Gíslason, 2001). –
Efnasamsetning basaltglers frá Stapafelli (samkvæmt
Oelkers and Gíslason, 2001).
wt. %
SiO2 48.12
TiO2 1.564
Al2O3 14.62
Fe2O3 1.11
FeO 9.82
MnO 0.191
MgO 9.08
CaO 11.84
Na2O 1.97
K2O 0.29
P2O5 0.195
168 JÖKULL No. 60