Search

We have conducted a survey of little-known N ivand O ivmultiplets near 3400 Å in an extensive sample of well-classified, very early O-type spectra. The initial motivation was to search for additional useful classification criteria for these types, but an unexpected result is the high sensitivity of these features to evolutionary CNO processing. We have found a useful discriminant between O2 and later types in the relative strengths of the O ivmultiplets, one of which is subject to selective emission in the hottest spectra; the overall strengths of these lines also decrease between spectral types O4 and O5. More remarkable, however, are the variations in the N/O ratios among both individual stars and clusters. For instance, several O4 If+ spectra have very large ratios, while main-sequence stars in the Carina Nebula generally have smaller values than others of the same spectral types in other regions. These effects correspond to different degrees of mixing of processed material as a function of evolutionary age and initial rotational velocities; the second effect provides significant further evidence that very massive stars mix while still on the main sequence. Thus, further analysis of these features will likely provide valuable diagnostics of important evolutionary parameters.

We have conducted a survey of little-known N iv and O iv multiplets near 3400 A in an extensive sample of well-classified, very early O-type spectra. The initial motivation was to search for additional useful classification criteria for these types, but an unexpected result is the high sensitivity of these features to evolutionary CNO processing. We have found a useful discriminant between O2 and later types in the relative strengths of the O iv multiplets, one of which is subject to selective emission in the hottest spectra; the overall strengths of these lines also decrease between spectral types O4 and O5. More remarkable, however, are the variations in the N/O ratios among both individual stars and clusters. For instance, several O4 If+ spectra have very large ratios, while main-sequence stars in the Carina Nebula generally have smaller values than others of the same spectral types in other regions. These effects correspond to different degrees of mixing of processed material as a function of evolutionary age and initial rotational velocities; the second effect provides significant further evidence that very massive stars mix while still on the main sequence. Thus, further analysis of these features will likely provide valuable diagnostics of important evolutionary parameters.

From a survey of the 3400 A region in the earliest O-type spectra, we have found that two of the four O2 giants observed in the Large Magellanic Cloud have O IV lines there that are stronger than the N IV lines, while the other two have the opposite. A Small Magellanic Cloud counterpart also has N IV stronger than O IV. Inspection of the blue spectra of these stars shows that the former pair have weaker N lines in all ionization states (III, IV, and V) present as well as lines of C IV l4658, while the latter three have stronger N lines and greater He/H. Space ultraviolet observations of two of the N-strong stars show N V wind profiles substantially stronger than those of C IV, while in the N-weak stars the C IV features are equal to or stronger than the N V. The N-strong stars are now reclassified as ON2 III(f*), newly defining that category. These characteristics strongly suggest a larger fraction of processed material in the atmospheres of the ON2 stars, which we confirm by modeling the optical spectra. In the context of current models, it is in turn implied that the ON2 stars are in a more advanced evolutionary state than the others, and/or that they had higher initial rotational velocities. The recent formulation of the effects of rotation on massive stellar evolution introduces an additional fundamental parameter, which the CNO abundances are in principle able to constrain. We present some illustrative comparisons with current Geneva evolutionary models for rotating massive stars. It is possible that these very hot, nitrogen-rich objects are products of homogeneous evolution. Our results will provide motivation for further physical modeling of the atmospheres and evolutionary histories of the most massive hot stars.