New model calculations for the production rates of cosmogenic nuclides in iron meteorites

Abstract

Here we present the first purely physical model for cosmogenic production rates in iron meteorites with radii from 5 cm to 120 cm and for the outermost 1.3 m of an object having a radius of 10 m. The calculations are based on our current best knowledge of the particle spectra and the cross sections for the relevant nuclear reactions. The model usually describes the production rates for cosmogenic radionuclides within their uncertainties; exceptions are 53Mn and 60Fe, possibly due to normalization problems. When an average S content of about 1 +/- 0.5% is assumed for Grant and Carbo samples, which is consistent with our earlier study, the model predictions for 3He, 21Ne, and 38Ar are in agreement. For 4He the model has to be adjusted by 24%, possibly a result of our rather crude approximation for the primary galactic particles. For reasons not yet understood the modeled 36Ar/38Ar ratio is about 30-40% higher than the ratio typically measured in iron meteorites. Currently, the only reasonable explanation for this discrepancy is the lack of experimentally determined neutron induced cross sections and therefore the uncertainties of the model itself. However, the new model predictions, though not yet perfect, enable determining the radius of the meteoroid, the exposure age, the sulphur content of the studied sample as well as the terrestrial residence time. The determination of exposure ages is of special interest because of the still open question whether the GCR was constant over long time scales. Therefore we will discuss in detail the differences between exposure ages determined with different cosmogenic nuclides. With the new model we can calculate exposure ages that are based on the production rates (cm3 STP/(gMa)) of noble gases only. These exposure ages, referred to as noble gas exposure ages or simply 3,4He , 21Ne , or 36,38Ar ages, are calculated assuming the current GCR flux. Besides calculating noble gas ages we were also able to improve the 41K-40Kand the 36Cl-36Ar dating methods with the new model. Note that we distinguish between 36Ar ages (calculated via 36Ar production rates only) and 36Cl-36Ar ages. Exposure ages for Grant and Carbo, calculated with the revised 41K-40K method, are 628 +/- 30 Ma and 841 +/- 19 Ma, respectively. For Grant this is equal to the ages obtained using 3He, 21Ne, and 38Ar but higher than the 36Ar- and 36Cl- 36Ar ages by ~30%. For Carbo the 41K-40K age is ~40% lower than the ages obtained using 3He, 21Ne, and 38Ar but equal to the 36Ar age. These differences can either be explained by our still insufficient knowledge of the neutron-induced cross sections or by a long-term variation of the GCR.