Jökull - 01.01.2010, Page 180

A. Stefánsson
reasonable extensive reactions or by both these fac-
tors. However, this effect of the acid supply or extent
of reaction has little effect on the Si concentrations,
whereas all the factors of pH, extent of reaction and
temperature seem to influence the Ca concentration.
DISCUSSION
The findings of the geochemical model calculations
and the comparison with natural low-temperature
geothermal alteration and water chemistry raise some
important questions on the role of temperature, initial
fluid composition (acidity) and extent of reaction on
the alteration process of basalts. The models strongly
suggest that the alteration process is largely controlled
by very fine variations in the pH of the water and that
temperature is even less important. The pH in turn
is determined by the steady state conditions between
the acid supply, extent of reaction and ionisation con-
stants of the weak acids and bases in the water.
One of the primary implications of the results has
to do with the application of low-temperature alter-
ation mineralogy and water chemistry for estimating
geothermometry temperatures. At 50 to 150◦C the
system temperature does not seem to be the primary
variable whereas pH and extent of reaction seem more
important. However, the geochemical models assum-
ing stoichiometric dissolution of basaltic glass fail to
distinguish the fine temperature dependence reported
for zeolite formation. On the other hand, this does
not mean that detailed alteration mineralogy is not
very important for understanding the geologic history
of a given rock formation and cannot be applied as
geothermometry. Combined with the understanding
of various other parameters affecting the alteration
mineralogy, water chemistry and time, much more in-
formation may be obtained than generally assumed.
The common alteration mineralogy observed in pores
in the tertiary basalts of east and west Iceland is con-
sistent with the last low-temperature alteration cate-
gory, i.e. a basalt buffered system. The occurrence of
phyllosilicates, chalcedony (or quartz) and zeolites in
the above order suggests either low-acid supply sys-
tems and/or considerably reacted systems. In the ab-
sence of any traces of the acid in the secondary min-
eralogy, like carbonates, the former is probably true,
whereas the abundant carbonates suggest the latter.
In addition, estimation of the mass of the alteration
product to the unaltered rocks may be further used to
constrain the extent of reactions and the time during
which the systems have reacted. One of the problems,
however, is that dissolution rates of basaltic glass and
basalts increases with increasing temperature, and so
do the mass fluxes. Therefore, extensively reacted
rocks may have resulted from reactions at higher tem-
peratures than being old and altered. Other minerals
may also be indications of acid supply, like CO2 rich
fluids. The appearance of carbonates like siderite and
magnesite, abundant SiO2 minerals, as well as kaoli-
nite may suggest lower pH values and CO2 enriched
alteration. This is in very good agreement with obser-
vations on the low-temperature alteration of basalts on
Disko Island, Greenland, with Mg-Fe carbonates and
SiO2 predominating under alteration characterized by
an elevated CO2 supply (and possibly other organic
acids) and the regional and probably low CO2 alter-
ation of basalts with phyllosilicates, SiO2 and zeolites
predominating (Neuhoff et al., 2006; Rogers et al.,
2006). Similar observations in secondary mineralogy
resulted during experiments with CO2-enriched and
depleted waters with basalt at low temperatures (Gysi
and Stefánsson, 2010). The detailed composition of
the phyllosilicates and zeolites may further reflect
changes in the reaction history, both temperature and
time. In the present work, changes from celadonite to
Ca-Mg-Fe smectites to chlorite with increased alter-
ation were observed, comparable to observations form
Teigarhorn and Hvalfjördur, Iceland (Neuhoff et al.,
1999; Weisenberger and Selbekk, 2009). In addition,
very fine relative changes in the phyllosilicate compo-
sition may further reflect changes in temperature and
time, yet such fine details are impossible to model at
present with the very simplified models for phyllosil-
icate solid solutions used in this study.
In the case of solution geothermometers, tempera-
ture seems to be the dominant factor that influences Si
concentration. For many other elements like Na and
Ca, pH and extent of reaction or reaction time seem
to be equally, if not more important. This could ex-
plain the success of the silica geothermometer, i.e. it
is independent of most factors other than temperature.
180 JÖKULL No. 60