Mi. Bird et al., OXYGEN-ISOTOPE FRACTIONATION IN GIBBSITE - SYNTHESIS EXPERIMENTS VERSUS NATURAL SAMPLES, Geochimica et cosmochimica acta, 58(23), 1994, pp. 5267-5277
The following oxygen-isotope fractionation factors for the gibbsite-wa
ter system (alpha(g-w)) have been obtained from a series of twenty-two
synthesis experiments: 1.0119 +/- 0.0005 (51 +/- 2-degrees-C); 1.0144
+/- 0.0003 (25 +/- 0.1-degree-C); 1.0139 +/- 0.0003 (22-degrees-C +/-
2-degrees-C); and 1.0162 +/- 0.0012 (8 +/- 2-degrees-C). Isotopic equ
ilibrium is considered to have been reached in the majority of cases b
ecause of the coincidence of results for syntheses using several diffe
rent techniques. All three polymorphs of Al(OH)3, namely, gibbsite, ba
yerite, and nordstrandite, were identified in the synthesis products.
Partial dissolution techniques have been applied to samples from twelv
e bauxite deposits to obtain the deltaO-18 value of both gibbsite and
kaolinite. These data, plus previously published results, have been us
ed to calculate alpha(g-w) values, using two independent methods: (1)
by comparison with the deltaO-18 value of associated kaolinite, and (2
) by comparison with the deltaO-18 value of local meteoric waters. The
results show a spread of values from those obtained by synthesis expe
riments to values up to approximately 6parts per thousand higher. It i
s concluded that the spread of values is real and the deltaO-18 value
of gibbsite is dependent upon its formation mechanism. It is suggested
that the gibbsite samples which exhibit high alpha(g-w) values have e
ither obtained their isotopic signature by inheritance during the desi
lication of kaolinite, or from exchange with oxygen-bearing anions suc
h as sulphate, during the transport of aluminum as Al-OH complexes. Th
e possibility that the high values are the result of evaporative modif
ication of the fluid prior to gibbsite precipitation cannot be ruled o
ut, but is unlikely, given that humid conditions are required for baux
ite formation. Such conclusions suggest that the oxygen isotope compos
ition of gibbsite is unsuitable as a palaeoclimatic indicator but may
prove useful in providing genetic information on bauxite formation.