Content area
Full Text
NGC 6791 is a well studied open cluster1 that it is so close to us that can be imaged down to very faint luminosities2. The mainsequence turn-off age (~8 Gyr) and the age derived from the termination of the white dwarf cooling sequence (~6 Gyr) are very different. One possible explanation is that as white dwarfs cool, one of the ashes of helium burning, ^sup 22^Ne, sinks in the deep interior of these stars3-5. At lower temperatures, white dwarfs are expected to crystallize and phase separation of the main constituents of the core of a typical white dwarf (^sup 12^C and ^sup 16^O) is expected to occur6,7. This sequence of events is expected to introduce long delays in the cooling times8,9, but has not hitherto been proven. Here we report that, as theoretically anticipated5,6, physical separation processes occur in the cores of white dwarfs, resolving the age discrepancy for NGC 6791.
White dwarf stars are the most common end-point of stellar evolution. Because they are very old objects, they convey important information about the properties of all Galactic populations, including globular and open clusters. This is particularly true for NGC 6791, a metal-rich ([Fe/H][asymptotically =]+0.4), well populated (~3,000 stars) and very old (~8 Gyr) open cluster that has been imaged down to luminosities below those of the faintest white dwarfs1,2, thus providing us with a reliable white dwarf luminosity function2. The white dwarf luminosity function ofNGC6791 presents two prominent peaks (a rather peculiar feature). The first of these peaks has been interpreted as either the result of a population of unresolved binaries10, or a population of single helium white dwarfs11. The second peak and the subsequent drop-off in the white dwarf luminosity function are a consequence of the finite age of the cluster. The age obtained using the white dwarf luminosity function of NGC 6791 is in conflict with the age of the cluster derived from its main-sequence stars. This discrepancy cannot be attributed to a poor determination of the main-sequence turn-off age, because for this cluster we have a reliable determination of its age, 8.0±0.4 Gyr. The age uncertainty mainly arises from the uncertainty in the metallicity determination and is probably an overestimate2. Several explanations for solving this discrepancy have been proposed1,2. Amongst...