A study of Cygnus OB2: Pointing the way toward finding our galaxy's super-star clusters

New optical MK classification spectra have been obtained for 14 OB star candidates identified by Comeron et al. and presumed to be possible members of the Cyg OB2 cluster as recently described by Knodlseder. All 14 candidate OB stars observed are indeed early-type stars, strongly suggesting the remaining 31 candidates by Comeron et al. are also early-type stars. A thorough investigation of the properties of these new candidate members compared with the properties of the Cyg OB2 cluster star has been completed, using traditional as well as newly revised effective temperature scales for O stars. The cooler O star effective temperature scale of Martins et al. gives a very close distance for the cluster (DM=10.4). However, even using traditional effective temperature scales, Cyg OB2 appears to be slightly closer (DM=10.8) than previous studies determined (DM=11.2; Massey & Thompson), when the very young age of the stellar cluster (similar to2x10(6) yr) is taken into account in fitting the late-O and early-B dwarfs to model isochrones. Of the 14 new OB stars observed for this study, as many as half appear to be significantly older than the previously studied optical cluster, making their membership in Cyg OB2 suspect. So, while some of the newly identified OB stars may represent a more extended halo of the Cyg OB2 cluster, the survey of Comeron et al. also picked up a large fraction of nonmembers. Presently, estimates of the very high mass of this cluster (M(cl)approximate to10(4) M-circle dot and over 100 O stars) first made by Knodlseder remain higher than this study can support. Despite this, the recognition of Cyg OB2 as a more massive and extensive star cluster than previously realized using 2MASS images, along with the recently recognized candidate super-star cluster-Westerlund 1 only a few kiloparsecs away (Clark & Negueruela), reminds us that we are woefully underinformed about the massive cluster population in our Galaxy. Extrapolations of the locally derived cluster luminosity function indicate that tens to perhaps a hundred of these very massive open clusters (M(cl)approximate to10(4) M-circle dot, M(V)approximate to-11) should exist within our galaxy. Radio surveys will not detect these massive clusters if they are more than a few million years old. Our best hope for remedying this shortfall is through deep infrared searches and follow-up near-infrared spectroscopic observations, as were used by Comeron et al. to identify candidate members of the Cyg OB2 association.