New Research Places Mars Bulk Composition In Question

Connie Bertka and Yingwei Fei of Carnegie's Geophysical Laboratory and Center for High Pressure Research report in this week's Science magazine that the bulk elemental composition of Mars does not match the composition of a type of primitive meteorite called a C1 carbonaceous chondrite. The abundance ratios of non-volatile elements in C1 chondrites, especially the iron/silica (Fe/Si) ratio, has long been believed to be a standard for the terrestrial planets. C1 chondrites evidence refractory element abundance ratios similar not only to those of the sun's atmosphere, but to lunar and terrestrial samples as well. Because of this, scientists for over forty years have assumed that C1 chondrites represent the original parent material from which the inner solar system accreted, and that the terrestrial planets (with the exception of Mercury) evidence the same basic non-volatile element composition. The differences in mean densities were thought to arise from differences in the amount of reduction that the originally oxidized C1 material experienced. (Some elements in their reduced form favor the formation of denser mineral phases than in their oxidized form. For example, metallic iron, Fe, is much denser than an Fe+2- or Fe+3-bearing silicate mineral phase.)

Previous studies had suggested that the C1 model might not work for Mars, but those studies were based on questionable assumptions. Bertka and Fei entered the fray last year, after the Mars Pathfinder mission brought home a definitive value for Mars's moment of inertia, designated C. C describes the mass distribution within a planet's interior; essentially it tells how the elements may be partitioned into a silicate mantle and a denser metallic core. C is one of the factors necessary to determine a planet's bulk composition. Before the Mars data were derived from Pathfinder results, C was known only for the Earth and Moon. That value for Earth, combined with knowledge of the Earth's mean density and an understanding of high-pressure mineral phase transitions in its interior, can indeed lead to a calculated non-volatile element bulk composition equivalent to that of a C1 chondrite.

Bertka and Fei did their best to come up with similar results for Mars. However, they could not make Mars fit a C1 composition and still conform to known geophysical and geochemical constraints (including the new value for C and a bulk composition derived from a set of martian meteorites). The problem arises in the martian core. In order to conform to C1 and other constraints, the core cannot be made only of iron, sulfur, and nickel, as many previous models had assumed. That combination is much too dense. Therefore, Bertka and Fei mixed in the lighter elements carbon and hydrogen. They calculated core densities resulting from a variety of element combinations as functions of pressure and temperature all with the final elemental end product of C1. However, the core remained too dense. The C1 model had failed. The elemental composition of Mars was clearly different from that of C1--and of Earth.

If the C1 model doesn't work with Mars, says Bertka, then it can't be assumed as a standard for the other terrestrial planets, and the variations in mean density of the inner planets must be explained some other way not by the oxidation and reduction of a common bulk elemental composition. "In our heart of hearts, we suspected that the C1 model was an oversimplification," Bertka says. "But it was the best we had."

The Bertka-Fei results suggest that a variation in bulk Fe/Si ratios among the terrestrial planets is possible. At first appearances, this would mean that Mercury, Venus, Earth, and Mars all accreted from different materials that they had their own local "feeding zones." However, Carnegie's George Wetherill, who has developed a widely accepted accretion model based on the assumption that the planets accreted from material contributed from a common area, has suggested a scenario that would explain the discrepancy, at least for Venus, Earth, and Mars. (The high density of Mercury is owed to something else.) He sees a correlation between the final distance of a planet from the sun and the location of the average area, or "provenance," from which the material that accreted to form the planet originated. Thus, if the original planetesimal swarm orbiting the sun was not entirely homogeneous, that is, if it evidenced fluctuations in its elemental composition, then it might be possible that the resulting planets would reflect those fluctuations and evidence the differences in bulk composition and density we see today.

The work was partially supported by a grant from NASA.

###

The Geophysical Laboratory is one of five science research departments of the Carnegie Institution of Washington, a nonprofit organization devoted to advanced research and education in the physical and biological sciences. It's new director, Wesley T. Huntress, Jr., assumes his responsibilities at the end of the month. The Carnegie Institution is led by its president, the biologist Maxine F. Singer.