ABUNDANCES OF ACTINIDES AND SHORT-LIVED NONACTINIDES IN THE INTERSTELLAR MEDIUM: DIVERSE SUPERNOVA SOURCES FOR THE r -PROCESSES G. J. WASSERBURG The Lunatic Asylum, Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125 M. BUSSO Osservatorio Astronomico di Torino, Str. Osservatorio 20, 10025 Pino Torinese, Italy AND R. GALLINO Dipartimento di Fisica Generale, Via P. Giuria 1, 10125 Torino, Italy Received 1996 April 11; accepted 1996 May 20 ABSTRACT Abundances of 244 Pu, 235 U, 238 U, and 232 Th in the early solar system are about those expected for uniform production over most of galactic history. The inferred abundance of 182 Hf is also compatible with this model. We here associate production of 182 Hf with the same r -process SN sources that produce actinides (SNACS). This requires that r -process nucleosynthesis in SNACS took place rather uniformly over the age of the galaxy until 110 7 yr prior to solar system formation. The low abundance of 107 Pd and 129 I in the early solar system indicates that SNACS cannot produce these nuclei at the high yields expected from standard r -process models. We propose that there are distinctive SN sources for different r -process nuclei with a sharp distinction in different SN contributions below and above A 1 140. Abundances in stars with very low metallicities will vary depending on the type of SN contributing to the local region of star formation. A time scale of 110 7 yr is much shorter than the 10 8 yr time usually associated with processes in the galaxy and with the last time of r -process injection accounting for 129 I, but may be compatible with the rate of SN occurrence. The hypothesis of a nearby SN polluting the protosolar nebula is critically discussed. Subject headings: ISM: abundances — nuclear reactions, nucleosynthesis, abundances — solar system: formation — stars: formation — supernovae: general 1. INTRODUCTION The discovery of a deficiency of 182 W/ 184 W in iron meteorites by Lee & Halliday (1995) and Jacobsen & Harper (1996) indicates that 182 Hf was present in the early solar system with 182 Hf/ 180 Hf 2 2.8  10 -4 . The upper bound of 182 Hf/ 180 Hf  5  10 -5 by Ireland (1991) from a mesosiderite may be due to a later time of formation (see Ireland & Wlotzka 1992; Stewart, Papanastassiou, & Wasserburg 1994). A correlation of 182 W with Hf/ W, which is necessary to demonstrate the presence of 182 Hf, has yet to be found. We here assume the deficit in 182 W/ 184 W is due to 182 Hf. This observation was hinted at by Harper et al. (1991) and Harper & Jacobsen (1994). Production of 182 Hf by r - or s -processes was investigated by Cameron (1993) and Cameron et al. (1993). Norman & Schramm (1983) proposed 182 Hf as an r -process chronometer. Wasserburg et al. (1995) investigated AGB sources and found 26 Al, 107 Pd, 41 Ca, and 60 Fe can be explained by the special circumstances of injection from a nearby AGB source. How- ever, the 182 Hf predicted by Wasserburg et al. (1994, 1995) is low by a factor of 100 with respect to the Hf data due to the low neutron density required to reproduce branchings in the main s -component. Gallino et al. (1996) increased this yield times 2. We pursue an analysis of these results assuming 182 Hf is an r -process product and will show that uniform production of actinides and 182 Hf over the history of the galaxy yields self-consistent results which then require distinctive produc- tion sites of many r -process nuclei. We are not interested here in seeking self-consistent long-term nucleosynthetic chronolo- gies but rather in attempting to estimate abundances of short-lived nuclei in the ISM and relating them to the actin- ides. Our approach follows Schramm & Wasserburg (1970). Some calculations presented here were given by Cameron (1993), who considered long-term, uniform production. To explain the low ratio ( 129 I/ 127 I) J = 10 -4 , it is usually assumed that this r -process nuclide was the residue of SN ejecta from a time 110 8 yr before the solar system formed. This is at odds with the inferred abundance of 182 Hf. 107 Pd is not a severe constraint as it may be produced by r - or s -processes. 2. UNIFORM PRODUCTION FOR ACTINIDES We use 232 Th for reference and exhibit N i AC / N 232 Th for various models. Here N i AC is the number of actinide (AC) species “i .” For the limiting case of steady state production of 232 Th, 238 U, 235 U, 244 Pu, and 247 Cm, we have N i AC / N 232 Th = P i AC ¯  i AC where ¯  i AC is the mean lifetime of “i ” and P i AC the production rate. Figure 1 shows N i AC / N 232 Th as a function of ¯  i AC for cases A, steady state reference line with P i AC = P j AC ; B, steady state produc- tion (SSP) with the factors P i AC from standard estimates in Schramm & Wasserburg (1970; see Cowan et al. 1991 for other estimates; these rates are not well established); C, uniform production (UP) over a time T = 10 10 yr using standard production factors; and D, measured (M) or inferred values for the early solar system (cf. summary in Wasserburg 1985). Case C will be called the uniform production (UP) model. The abundances given here are the values immediately after ces- sation of nucleosynthesis. Figure 1 shows that these results cluster about the steady state reference line and give reason- able agreement between UP and M. ( 244 Pu/ 232 Th) J is only a THE ASTROPHYSICAL JOURNAL, 466 : L109–L113, 1996 August 1  1996. The American Astronomical Society. All rights reserved. Printed in U.S. A. L109