Actinides in HD 101065 (Przybylski's Star)

There are many strange things about the abundances in Przybylski's star (HD 101065). The most recent study (MN, 217, 299, 2000) finds that among the elements through copper, the abundances scatter with no apparent pattern. The largest deviations from solar are found for magnesium, which may be deficient by somewhat more than 1 dex, and cobalt, which appears to be in excess by about the same amount.

The heavier elements especially those beyond barium, and continuing through the actinides uranium and thorium, show a clear pattern and a remarkable coherence. We plot solar and stellar abundances on a logarithmic plot vs. atomic number Z, and displace the solar points upward by some 3 dex. While individual points do not overlap, there is a remarkable similarity in the overall trends from Z = 58-80, including the maxima caused in the solar case by the third r-process peak. Points for the displaced solar actinides fall near their stellar counterparts.

The processes that produced the large enhancements of these heavy elements have not caused large fractionations of adjacent elements.

We have new spectra from the ESO UVES spectrograph on UT2. Resolution is 80,000 (blue) and 110,000 (red), with S/N > 300. This material, and new oscillator strengths (AA 381, 1090,2002; AA 382, 368, 2002), allow a more complete analysis of U and Th. We find abundances near +2.5 for both elements (log(H) = 12), but uncertainties at the moment are surely +/- 0.3 dex. We are currently working to improve the accuracy. Current errors preclude the use of these observations for accurate cosmochronometry, even if the chemical differentiation were not relevant.

If a U/Th ratio of unity could be taken at face value, it would imply an r-process event some 10³ years ago. Thanks to B. Pfeiffer help and advice.