Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
DIAMONDS IN METEORITES Edward Anders Enrico Fermi Institute for Nuclear Studies University of Chicago Diamonds have been reported in five meteorites; in one case, Carcote, the identification is uncertain and has never been confirmed. The remaining diamond-bearing meteorites comprise two ureilites (a rare sub-class of achondrites), Novo Urei and Goalpara, and two coarse octa- hedrites, Magura and Canyon Diablo. The orthodox interpretation of the origin of meteoritic diamonds is that they were formed from graphite under high pressures. The pres- sures required are about 1. 6 x 104 atm. at 298o K and about 3. 2 x 104 atm. at 1000o K. Assuming that the source of pressure was the gravita- tional potential of a meteoritic parent body of reasonable density (~3. 3 g cm-3), the minimum radius of such an object must have been ~1000 km. In order for one-half of the interior of the parent body to be at or above the minimum pressure, it must have been of approximately lunar size (radius ~1700 km). These considerations impelled Harold Urey to postulate a generation of lunar-sized parent objects as a principal feature of his proposed history of the meteorites. Parent objects of lunar size, however, are hard to reconcile with evidence discussed in two papers by members of our group (Goles, Fish and Anders, 1960; Fish, Goles and Anders, 1960) and with the model of meteoritic synthesis developed in the latter paper. Accordingly, we felt it was necessary to re-examine critically the orthodox interpretation of meteoritic diamonds to determine whether sustained high pressures were required by the evidence. This investigation, which was carried out mainly by Mr. M. E. Lipschutz, took two forms: a study of the thermo- dynamics of diamond synthesis in the presence of metallic iron, and an empirical study of the pressure and temperature conditions under which Canyon Diablo diamonds were formed. Since excess Fe is always present in diamond-bearing meteorites, and since diamonds in Canyon Diablo at least are not found associated with graphite nodules but with other minor phases, it appears that the Fe-C-Fe3C system is the appropriate one to consider in investigating diamond formation in meteorites. Cohenite (Fe3C) can decompose to 77
either diamond or graphite over a wide range of temperatures, but the reactions are quite sensitive to pressure. When the loci of AF = 0 for these decomposition reactions are plotted in the P-T plane, a very in- formative diagram results (Figure 1; see Lipschutz and Anders, 1960, for discussion of the derivation and significance of this diagram). In the absence of a solvent, this graph may be interpreted as a phase diagram indicating the stable and metastable forms of carbon at a given tempera- ture and pressure. Note that in the presence of excess iron, both graphite and diamond are unstable with respect to cohenite in the area to the 100 80 60 1*1 o s °- 40 20 T Stability Fields For Diamond, Graphite And Cohenite in The Presence of Excess Iron. Cohenite Cohenite 4OO 60O 8OO Temperature (°K) 1000 Figure 1. Stability fields for diamond, graphite and cohenite in the presence of excess iron. Metastable phases are indicated by brackets. In the absence of a solvent, this diagram is equivalent to a phase dia- gram, whereas in the presence of a solvent, it may be used to infer the shift in equilibria upon pressure or temperature change. 78
right of lines CH and HB. From this, it can be shown that it is virtually impossible to construct a plausible history that would permit the preter- restrial formation of the alleged "graphite pseudomorphs after diamond" and the Widmanstatten pattern. In the presence of a solvent, the decomposition reactions of cohenite may be treated as ordinary chemical equilibria at small values of A F. All stable reactants and products will be present in appreciable amounts at any temperature and pressure, their proportions being governed only by the value of the equilibrium constant. One can still use Figure 1 to infer qualitatively the shift in equilibrium for any temperature or pressure change. Of course, diamond may form metastably below the line EH, and graphite, above this line. Our study of diamond-bearing specimens of Canyon Diablo has strongly supported the suggestion by Nininger that the diamonds in this meteorite were synthesized during impact, rather than pre-terrestrially. This conclusion is based on the following line of reasoning: It has long been noted that diamonds in Canyon Diablo: 1) are not associated with graphite; 2) are always associated with troilite; 3) are imbedded in areas which contain cohenite, though not necessarily in contact with the diamonds; 4) are submicrocrystalline, with an ultimate particle size of a few hundred angstroms, as determined by x-ray diffraction; and 5) are irregularly dis- tributed, so that one specimen may be rich in diamonds while other large pieces are completely free of them. In addition, Nininger has pointed out that the diamond-bearing specimens are usually small (<5 kg), are found only on the rim of the crater, and display evidence of reheating. We may infer that there are two distinct types of Canyon Diablo specimens, that the distribution and other properties of the diamond-bearing ones indicate a correlation between conditions during impact and the presence of dia- monds, and that the precursor of the diamonds was, as argued above, cohenite rather than graphite. Since there seems to have been a causal relationship between reheating and diamond synthesis (or, at least, they occurred simultaneously), study of the conditions of reheating should provide information on the conditions of diamond formation. Microscopic examination of diamond-bearing Canyon Diablo speci- mens reveals that, in contrast to the diamond-free type, the kamacite (a-phase) is poly crystalline, the Neumann lines have been annealed out, and there are borders of martensite (metastable iron-carbon alloy, formed only by very rapid cooling of carbon-rich ^-phase) around cohenite lamellae. These features may be reproduced quite accurately by heating meteorite specimens to 850 - 950 oC for a few seconds and quenching in cold water, as determined by experiments on the structurally very similar Odessa octahedrite. Trial-and-error experiments with Odessa, using various minor phases (Fe3P, Fe3C, etc.) and their eutectics as 79
temperature indicators, show that diamond-bearing Canyon Diablo specimens were reheated to temperatures between 925 and 1000 OG. Rapid cooling, on a time scale of <2 min., is indicated by the formation of martensite rather than pearlite around cohenite inclusions. Since no one would seriously propose that the original mass of the Canyon Diablo meteoroid (~2 x 106 tons, according to Opik) could be quenched so rapidly, it is clear that these structures were developed after the fragments attained their present small size, i. e., during impact. Thus, we would hypothesize that the Canyon Diablo diamonds formed from cohenite during impact. The fact that diamond rather than graphite was formed may be due to the high pressures prevailing in the decelerat- ing meteoroid during its penetration into the ground, to localized stresses due to differential thermal expansion, or to metastable formation of diamond due to preferential nucleation. We may estimate a lower limit to the time during which the meteoroid was under compression by equating this with the minimum time for penetration to the floor of the crater. For an initial velocity of 15 km/sec, one obtains >1. 7 x 10-2 sec for this "compressive deformation" time. At growth rates of ~0. 1 mm/min (Bovenkerk et al., 1959), diamonds of ~300 A size could have been formed in this time. For the other processes suggested, similar time scales should apply. What of the other diamond-bearing meteorites? In the case of Magura, the mode of diamond synthesis may be similar to that proposed for Canyon Diablo. Many fragments of Magura have been recovered, and they display clear evidences of different thermal histories. We cannot say whether or not there is a one-to-one correlation between the presence of diamonds and intense reheating in specimens of this meteorite, but it is at least possible that these features were produced during impact. For the ureilites, on the other hand, it may easily be established that they arrived at the Earth's surface with relatively low terminal velocities, so that this explanation would not apply. Nevertheless, at least once in their preterrestrial history the ureilites must have been involved in a violent collison (upon the occasion of the break-up of their parent body). It is tempting to conjecture that the diamonds found in the ureilites were formed in that event. Unfortunately, nothing is known regarding the mode of occurrence of these diamonds, so that it is not possible to decide whether they could have been synthesized in such an impact. Cameron: Can you set a lower limit to the size of the crater produced by, for example, Magura, if this meteorite contains diamonds which were formed on pact? Anders: It would be possible to do this, using the diamond growth-rate mentioned and an estimate of initial velocity, but I would expect that 80
the value obtained for the crater depth could be uncertain by a factor of one hundred and therefore of little use. REFERENCES Bovenkerk, H. P., Bundy, F. P., Hall, H. T., Strong, H. M., and Wentorf. R. H., Jr. (1959) Nature (London) 184, 1094. Fish, R. A., Goles, G. G., and Anders, E. (1960) Astrophys. J. 132, 243. Goles, G. G., Fish, R. A., and Anders, E. (1960) Geochim. et Cosmo- chim. Acta. 19, 177. Lipschutz, M. E., and Anders, E. (1960) submitted to Geochim. et Cosmochim. Acta. 81
