干旱区生态环境知识资源中心

Interpretation of the mineralizing environment of magnetite-apatite deposits remains controversial with theories that include a hydrothermal or magmatic origin or a combination of those two processes. To address this controversy, we have analyzed the trace element content of magnetite from precisely known geographic locations and geologic environments from the Precambrian magnetite-apatite ore and host rocks in Kiruna, Sweden, and the Pliocene-Holocene El Laco volcano in the Atacama desert of Chile. Magnetite samples from Kiruna have low trace element concentrations with little chemical variation between the ore, host, and related intrusive rocks. Magnetite from andesite at El Laco, and dacite from the nearby Lascar volcano, has high trace element concentrations typical of magmatic magnetite. El Laco ore magnetite have low trace element concentrations and displays growth zoning in incompatible elements (Si, Ca, and Ce), compatible elements (Mg, Al, and Mn), large-ion lithophile element (Sr), and high field strength element (Y, Nb, and Th). The El Laco ore magnetite are similar in composition to magnetite that has been previously interpreted to have crystallized from hydrothermal fluids; however, there is a significant difference in the internal zoning patterns. At El Laco, each zoned element is either enriched or depleted in the same layers, suggesting the magnetite crystallized from a volatile-rich, iron-oxide melt. In general, the compositions of magnetite from these two deposits plot in very wide fields that are not restricted to the proposed fields in published discriminant diagrams. This suggests that the use of these diagrams and genetic models based on them should be used with caution.