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.
COSMIC RAY PRODUCTION OF RADIOACTIVE NUCLIDES IN IRON METEORITES J. R. Arnold School of Science and Engineering University of California at La Jolla Much new information is becoming available on the concentration of various long-lived radioactive nuclides (such as Be10, Al2^ Cl36, K40, and Mn53) produced by cosmic rays in iron meteorites. A number of short-lived species are also currently being measured in the recently fallen iron meteorite Aroos. With the aid of this information and of data on stable isotopes Dr. Honda, Dr. Lai, and I have tried to shed new light on the old problem of the time dependence of the intensity of the cosmic radiation. I am going to discuss some of this information and the prob- lems which have arisen in its interpretation. The following table presents some data (Honda, Shedlovsky, and Arnold 1960) on the activity ratios of Be, Al, K, and Mn in four iron meteorites. Nuclide Be 10 A126 K40 Mn53 Half-life 2. 7xl06 y 7.4xl05 y 1. 3xl09 y 2xl06 y Grant Absolute dis- integration rate (dpm/kg) 4. 03±0. 27 4.29±0.27 4.6±0. 5 299±H Williamstown Grant Odessa Grant Canyon Diablo Grant 0. 87±0. 11 0. 80±0. 13 1. 1±0. 3 0. 95±0. 15 0. 51±0. 04 0. 32±0. 04 0. 4±0. 15 0. 66±0. 03 0. 21±0. 04 0. 20±0. 06 0. 31±0.04 45
The samples are as follows: Grant--saw cuttings from the cut through the meteorite which was made preliminary to the distribution studies on rare gases (Fireman 1959, Hoffman and Nier 1958). These distribution studies tell us that the 500 kg specimen we possess was a single individual in space. Williams town--a slice from the single 30 kg specimen recovered. Odessa--an individual of this large meteorite fall. Canyon Diablo--again a single small individual from a large fall. We expect that the production rates in these two meteorites will depend very much on the specimen chosen. Our chemical procedures have been distinguished by the use of un- usually large samples in order to permit comparatively accurate measure- ments, and by a serious and apparently successful effort to recycle to constant specific activity in each case. We have also been able because of the level of activity observed to apply certain auxiliary tests such as the measurement of beta absorption curves. The accuracy is of course still low compared to that obtainable in rare gas work, but some significant features are evident in the data: 1. The disintegration rates in Williamstown and Grant are practi- cally alike for all activities. For the species Be10, Al26, Mn53 this similarity may be taken to mean that both meteorites have built steady state abundances. Such an interpretation is manifestly not valid for K40, which has a far longer half-life. Presumably the nearly equal potassium decay rates mean that Williamstown and Grant had approximately the same cosmic ray ages. 2. The ratios of the short-lived isotopes of Odessa/Grant and Canyon Diablo/Grant are all less than one, as one would expect for inter- nal samples from large meteorites when they are compared to samples from a smaller meteorite (K40 activities once again need not conform to the other ratios). 3. The absolute disintegration rate of Mn53 is approximately two orders of magnitude higher than those of A126 and Be10 in aii cases. 46
4. The ratio of Mn53 activity to that of Be10 or A126 is substantially higher in both Odessa and Canyon Diablo than in Williamstown or Grant. * Of the models which have been developed to calculate production rates of isotopes in meteorites the most widely used is that of Martin (1953) and Ebert and Wanke (1957). Here the buildup of secondaries with depth was taken into account using the geometric interaction cross section of the primaries, and assuming a constant number of secondaries (about 3) per primary interaction. The cross section for the production of a par- ticular species is determined at the average energy of the primary radia- tion (usually 3 Bev). This cross section is multiplied by the primary flux at a given depth. Cross sections at some average secondary energy (usually 300 Mev) are used to predict the production rate by secondaries. This method is satisfactory in making comparisons of production rates of isotopes which have no important production below 300 Mev. However, no plausible set of cross sections can be found in this model to account for the production of so much Mn53 or other species close to the target nucleus as is actually observed. The physical basis of this model is ad- mittedly primitive; its great advantage is convenience and simplicity. Fireman and coworkers (1958) developed an ingenious method in connection with their studies of tritium and argon-37. A long iron bar is bombarded at various energies in high energy machines to deduce the production rate as a function of depth in iron and of energy, in a colli- mated beam. One then integrates the observed depth effect over angle, which yields a production rate as a function of depth in a meteorite. For accurate results this method would require integration over the primary energy spectrum. The most serious difficulty would appear to be the loss of secondaries which are emitted at large angles to the beam in the machine experiments. This is particularly important for low energy products. Nuclear emulsion studies of the cosmic ray primary and secondary particles within the atmosphere are a potential source of pertinent infor- mation. Especially important are the experiments of Shapiro and co- workers (1951) who flew large blocks of lead of various shapes at balloon altitudes. They placed emulsions at different positions in the blocks and studied the star size distribution. The block which most closely approxi- mated a sphere was about 15 cm in radius. Despite the fact that the * It should be noted that the production of Al26 by spallation of sulfur and phosphorus in an iron meteorite may be significant compared to the production from iron, because of the much higher cross section and lower energy threshold in these elements. The variation in the sulfur and phosphorus content may be a source of variation in the Al2" data. 47
dimensions were much in excess of the mean interaction length neither the total number of stars nor the star size distribution changed very much with depth. Flatter shapes showed depth effects, but the change with depth was not great. This is particularly striking since the primary flux comes only from above in these cases. These results would not be pre- dicted either by the Martin model or by the use of the iron bar target data. Admittedly they should be checked with iron blocks of various dimensions, but they do seem consistent with the fact that one keeps finding small meteorites which all show about the same radioactivity level. Many studies have been made of star size distributions in emulsions in the free atmosphere at various depths. Measurements of this type have also been made using mono-energetic beams of protons between about 100 Mev and 6 Bev. From a combination of these two sets of data it is in principle possible to deduce the energy spectrum of primary and secon- dary particles at any atmospheric depth. In practice this method is not capable of high precision, but the differential spectrum deduced is prob- ably not in error by as much as a factor of two at any energy, and should be more accurate than this when broad energy ranges are considered. Other information such as the primary energy spectrum and the secondary production spectrum of Camerini et al (1950) are helpful in inferring the functional form of the distribution. Shapiro's studies show that the total star production in emulsion in- side a lead block is enhanced over that in emulsion in the free atmosphere, but that the shape of the energy distribution is quite similar. The same should be true for iron. The energy distribution thus determined may be used together with the excitation functions for various nuclides to predict relative and absolute production rates. By measurement of short-lived isotopes from freshly fallen meteorites it is possible to check the validity of this tech- nique under conditions where the cosmic ray intensity is known to have been constant within the accuracy of the predictions (at present no better than 30 percent). The abundance ratios of stable isotopes or of pairs of long lived isotopes of similar half-life may also be used to check the technique. Our current program includes the measurement for this purpose of numerous short-lived isotopes in the meteorite Aroos. The higher ratio of Mn^ to A126 observed in Odessa and Canyon Diablo is easy to understand on the basis of such a model. Since Mn^3 is a "low energy nuclide" and Be10 and A126 are "high energy nuclides" the ratio should be elevated as the energy spectrum is degraded. The lower 48
production rates in Odessa and Canyon Diablo would lead one to expect that they represent greater depth; thus the results seem entirely reason- able. This new model seems to represent a useful advance over earlier ones. Presumably it is capable of being made exact as energy distribu- tions and excitation functions are improved. Preliminary calculations for the production rates of radioactive and stable isotopes are in gratify- ing agreement with experiment. The agreement is gratifying, that is, if the ablation of a meteorite like Grant was of the order of a few centimeters or less. But if one assumes an ablation of 12 cm in Grant (see Signer's paper in this connection) then the theoretical production rates are low by factors of the order two. We cannot yet prove that this discrepancy is significant. Within the accuracy of these preliminary calculations the cosmic ray flux has not altered significantly over the lifetime of species such as Be10, A126, Cl36, and Mn53. In the case of K40 a steady state abundance would not be expected since the apparent cosmic ray ages of the meteor- ites so far studied are less than 109 years. From the point of view of the constancy of the cosmic ray flux, K40 is by far the most interesting iso- tope. Its usefulness would be enhanced if meteorites of cosmic ray age more than one billion years could be found. The longer the time over which a meteorite has actually been bombarded, the greater the difference between the behavior of K40 and that of a stable species. Erosion in space, continued over the lifetime of the meteorite, can also be studied using radioactive isotopes. It seems obvious that not enough erosion can have occurred in the iron meteorites studied over the lifetimes of the million year species to seriously alter their activity ratios. In the case of K40 one might expect to see signs of an apparently increasing flux with time. It would be difficult but not impossible to dis- tinguish this from a true variation in cosmic ray intensity with time. REFERENCES Camerini, U., Fowler, P. H., Lock, W. O., Muirhead, H. (1950) Phil. Mag. 41, 413. Ebert, K. H., and Wanke, H. (1957) Z. Naturforsch 12a. 766. Fireman, E. L. (1958), in "Cosmological and Geological Implications of Isotope Ratio Variations, " Publication 572 of the National Academy of SciencesâNational Research Council (Washington, D. C.). 49
Fireman, E. L. (1959) Planet. Space Sci. 1, 66. Hoffman, J. H. and Nier, A. O. (1958) Phys. Rev. 112, 2112. Honda, M., Shedlovsky, J. P., and Arnold, J. R. (1960) Geochim. et Cosmochim. Acta (in press). Martin, G. R. (1953) Geochim. et Cosmochim. Acta 3, 288. Shapiro, M. M., Stiller, B.. Birnbaum, M., and O'Dell. F. W. (1951) Phys. Rev. 83, 455. (See also Phys. Rev. 86, 86 and 84, 160). Shapiro, M. M. (1951) Phys. Rev. 83, 456. 50
