NanoSIMS isotope studies of rare types of presolar silicon carbide grains from the Murchison meteorite: Implications for supernova models and the role of 14C

Abstract

We have conducted a NanoSIMS ion imaging survey of about 1800 presolar silicon carbide (SiC) grains from the Murchison meteorite. A total of 21 supernova (SN) X grains, two SN C grains, and two putative nova grains were identified. Six particularly interesting grains, two X and C grains each and the two putative nova grains were subsequently studied in greater detail, namely, for C-, N-, Mg-Al-, Si-, S-, and Ca-Ti-isotopic compositions and for the initial presence of radioactive 26Al (half life 716,000yr), 32Si (half life 153yr), and 44Ti (half life 60yr). Their isotope data along with those of three X grains from the literature were compared with model predictions for 15M⊙ and 25M⊙ Type II supernovae (SNe). The best fits were found for 25M⊙ SN models that consider for the He shell the temperature and density of a 15M⊙ SN and ingestion of H into the He shell before the explosion. In these models a C- and Si-rich zone forms at the bottom of the He burning zone (C/Si zone). The region above the C/Si zone is termed the O/nova zone and exhibits the isotopic fingerprints of explosive H burning. Satisfactory fits of measured C-, N-, and Si-isotopic compositions and of 26Al/27Al ratios require small-scale mixing of matter originating from a region extending over 0.2M⊙ for X and C grains and over 0.4M⊙ for one of the putative nova grains, involving matter from a thin Si-rich layer slightly below the C/Si zone, the C/Si zone, and the O/nova zone. Simultaneous fitting of 14N/15N and 26Al/27Al requires a C-N fractionation of a factor of 50 during SiC condensation. This leads to preferential incorporation of radioactive 14C (half life 5700yr) over directly produced 14N and can account for the 14N/15N along with 26Al/27Al ratios as observed in the SiC grains. The good fit for one of the putative nova grains along with its high 26Al/27Al points towards a SN origin and supports previous suggestions that some grains classified as nova grains might be from SNe. Apparent problems with the small-scale mixing scheme considered here are C/O ratios that are mostly <1 if C-, N-, and Si-isotopic compositions and 26Al/27Al ratios are simultaneously matched, underproduction of 32Si, and overproduction of 44Ti. This confirms the limitations of one-dimensional hydrodynamical models for H ingestion and stresses the need to better study the convective-boundary mixing mechanisms at the bottom of the convective He shell in massive star progenitors. This is crucial to define the effective size of the C/Si zone formed by the SN shock. The comparison between the Si isotope data of the SN grains and the models gives a hint that the predicted 30Si is too high at the bottom of the He burning shell.