Rit (Vísindafélag Íslendinga) - 01.06.1984, Blaðsíða 159

MINERALCHEMISTRY AND RELATIONSHIPS
155
carry out. Some experimental investigations have been made of simplií’ed
spinel systems which omit one or more of the major components of the spinel
structure, but these only provide an indication of possible trends in the
naturally occurring spinels. These limited experiments do not explain for
example why one spinel suite shows increased Cr relative to Al, whereas
another shows the reversed trend, and so on.
The most useful experimental studies of spinel relationships are those of
Hill & Roeder (1974). These authors showed that at constant f02 and
decreasing temperature (1200—1150°C) there is an increase in both the Fe-
oxides and in Ti, a slight decrease in Mg and A1 and a strong decrease in Cr
tn the chromite crystals of two basalt samples which they studied. At
constant temperature (1200°C) and decreasing Í02, Al, Ti and FeO increase
slightly, and Cr increases rapidly while Fe2Os and Mg decrease. In terms of
spinel prism ratio changes, this means a decrease in Mg/(Mg+Fe2+) and
Cr/(Cr+Al) and an increase in Fe3+/(Fe3++Cr+Al) with decreasing
temperature at constant f02 (10-7 atm). At constant temperature (1200°C)
and decreasing f02, F'e3+/(Fe3++Cr+Al) decreases and Cr/(Cr+Al)
increases while Mg/(Mg+Fe2+) reaches a minimum between ~10-4—10-6
atm. Hill & Roeders’ study is carried out on two basalt samples. The bulk
composition is constant and the results thus only indicate what may happen
during fractional crystallization. The results are, nevertheless, in general
agreement with the apparent evolution of numerous natural spinel suites.
Jackson (1969) presented an equation, based on thermodynamic data, to
calculate the temperature of equilibrium crystallization ofspinel and olivine
from their compositions. There has not been general agreement on the
validity of this thermometer. Evans & Wright (1972), in a study of the
liquidus chromites of the 1959 and 1965 eruptions of Kilauea, obtained
much too high a temperature using this equation and concluded that it was
not applicable in its present form. Other authors have used the equation
with quite satisfactory results; e.g. Jackson (op.cit.) and Littlejohn &
Greenwood (1974). This thermometer is extremely sensitive to small com-
positional changes, especially in the Mg/Fe2+ ratio ofeither the spinel or the
olivine. Changes in the forsterite content of the olivine from 90 to 91 may
result in a temperature change of over 250°C. Thus, if the composition of the
olivine in equilibrium with the spinel used in the calculations is not known
exactly the temperature result will be quite erroneous. Examples of tcmpera-
ture changes as a result of changed olivine or spinel composition are shown
in Table 22. There is general agreement between these temperature changes
and compositions, which are obtained through thermodynamics, and the
changes found by Hill & Roeder (op.cit.) by their experiments. There seems
therefore to be no reason to reject Jackson’s equation, but it must be used
with extreme caution.
The early spinels of the Jan Mayen ankaramites occur as inclusions in
olivine and clinopyroxene, as freely occurring small phenocrysts, or as small