One of the fundamental assumptions in using fluid inclusion data to study
mineral genesis is that the inclusions have behaved as closed systems
since their formation. Hall et al. [1991] and Mavrogenes and
Bodnar [1994] showed that H
diffusion into and out of fluid
inclusions during metamorphism or laboratory heating could significantly
modify the chemical compositions of the inclusions. Diffusion will be
most rapid at high temperatures and when the hydrogen fugacity
difference between the inclusion and its surroundings is large. Failure
to recognize or expect diffusion problems could result in flawed
reconstruction of the formation conditions of some minerals and ores.
In spite of these diffusion problems, it is still possible to retain primary
gas ratios in fluid inclusions from unmetamorphosed low-temperature
ores. Graney et al. [1991] studied the gas compositions of fluid
inclusions in epithermal jasperoids from various gold deposits. They
noted a correlation of high H
S/CO
and other gas parameters with
mineralized jasperoids, suggesting the utility of the technique for
exploration.
In the case of aqueous fluids trapped as inclusions during boiling,
knowledge of the temperature of boiling and fluid composition can lead
directly to an estimate of the depth of formation if the P-V-T properties
of the fluid are known. Many economic geologists have use the pure
H
O system as a proxy to interpret data from low-salinity fluid
inclusions in epithermal gold deposits. Barton and Chou [1993]
reviewed P-V-T data for the H
O-CO
system and demonstrated
that large errors in hydrostatic paleodepth reconstructions of epithermal
systems may occur if the presence of significant amounts of CO
in
fluid inclusions is not recognized. For example, if one observes the
formation of CO
clathrates upon freezing of an inclusion, then the
inclusion must have formed under relatively high CO
pressure.
This pressure would add at least 1 km to the paleodepth that would
otherwise be estimated if one used P-V-T data for boiling of pure water.
Therefore, if CO
is not detected because clathrate does not form or
is not recognized upon freezing, then large errors may result when
reconstructing the original depth of formation of the host minerals.
Kesler [1991], Bodnar [1992] and McKibben et al. [1994] edited special journal issues containing several other papers on U.S. research on fluid inclusions applied to ore deposits.