Using microbeam analytical techniques, Lowenstern et al. [1991]
and Lowenstern [1993] discovered Cu sulfides in CO
- and
Cl-bearing vapor bubbles in melt inclusions within
phenocrysts in
pantellerites and rhyolites, thus demonstrating that melt Cu could be
strongly partitioned into an early magmatic vapor phase in
[4]
phenocryst-poor magmas. The possibility of strong partitioning of Cu
into an early vapor phase, prior to extensive crystallization of phases that
would otherwise remove Cu from the melt, means that
crystallization-induced volatile saturation (second boiling) is not
necessary for the creation of metal-rich fluids in shallow H
O- or
CO
-rich silicic magma chambers. They also argued that volcanic
contributions of Cu to the atmosphere may be more significant than
previously thought. Meeker et al. [1991] identified crystalline
elemental gold and gold chloride particles being emitted from Mount
Erebus in Antarctica. This plus consistent Au/Cl ratios of aerosols from
the volcano suggested that the gold is transported as a chloride gas
species. Transport of trace metals in volcanic gases from Mount St.
Helens was modeled by Symonds and Reed [1993], who likewise
concluded that most were volatilized from shallow magma as simple
chlorides and deposited as sublimates upon cooling as oxides, sulfides,
halides, tungstates and native elements.
Rye [1993] summarized the evolution of magmatic-hydrothermal ore-forming fluids based on many years of stable isotopic research on such ore deposits. He reviewed evidence for high-level interactions of deep magmatic components with shallow wall-rock and meteoric waters, and emphasized the episodic, successive input of deep magmatic volatiles and evolved brine into shallower crustal levels to generate acid alteration and ore deposition. His paper and the review papers by Giggenbach [1993] and Hedenquist and Lowenstern [1994] are perceptive, complementary evaluations of the processes that form magma-hydrothermal ore deposits.