ON DENSITY AND COMPRESSIBILITY AT PRESSURES UP TO THIRTY MILLION ATMOSPHERES

Abstract

Summary A critical examination is made of curves presented by W. M. Elsasser on the density and compressibility of materials at zero temperatures and pressures up to 30 million atmospheres, and of Elsasser's inference that the Earth's core consists of iron or nickel-iron. An argument leading Elsasser to suggest that the writer's Earth Models A and B may require adjustments because of discrepancies with Elsasser's results is shown to be invalid. New tables are presented which fit the data used by Elsasser at 30 million atmospheres and are also consistent with geophysical data. The calculations leading to these tables imply that the atomic number to be associated with the material of the outer central core should be at least six units less than the value derived using simple extrapolations from quantummechanical calculations for high pressures. If the reduction in the atomic number is no more than six units, the most probable composition of the material in question would still be nickel-iron, and to this extent Elsasser's main conclusion is supported. But there are other aspects of the new calculations which suggest that the needed reduction may be greater than six units; these aspects raise the probability that the outer central core consists of a modification of ultra-basic rock. A caution is issued against forming too definite conclusions on either of the two opposing theories on the whole evidence so far available. Some support is found for the view that the Earth's present inner core is more likely to consist of an accumulation of iron and denser materials than of ultra-basic rock that has undergone a second pressure transformation. Support is also found for the view that the density gradients of Model B require a progressive change of composition with increase of depth in the inner core. The hypothesis that the inner core is solid is strengthened. Strong support is given to Elsasser's argument against the presence of large amounts of compressed hydrogen in the central core. The new tables suggest that the density of uranium (Z=92) may be a fairly smooth function of pressure from zero pressure right through to 30 million atmospheres, but that, for elements of smaller atomic numbers, there are likely to be significant density jumps at specific intermediate pressures. Tables are presented which give newly computed densities of uranium from 0.1 to 30 million atmospheres, and may serve as a useful basis of discussion in other contexts. As an incidental to discussing Elsasser's arguments, an Earth model (Model X) has been constructed with the property that the density in the central core near its outer boundary is 6.5 g/cm3. The model has a density gradient in the mantle only a little steeper than that in Models A and B, and has somewhat greater densities in the inner core. The outer part of the central core (between depths of 2900 and 4980km), however, differs radically from that of Models A and B through having a far steeper density gradient. The new model has to be regarded as implausible inasmuch as the last property implies far greater changes of composition with depth in the fluid part of the core than anywhere in the mantle.