Daniel C. Kiminki

A radial velocity survey of the Cyg OB2 association

We conducted a radial velocity survey of the Cyg OB2 association over a 6 yr (1999-2005) time interval to search for massive close binaries. During this time we obtained 1139 spectra on 146 OB stars to measure mean systemic radial velocities and radial velocity variations. We spectroscopically identify 73 new OB stars for the first time, the majority of which are likely to be association members. Spectroscopic evidence is also presented for a B3 Iae classification and temperature class variation (B3-B8) on the order of 1 yr for Cyg OB2 No. 12. Calculations of the initial mass function with the current spectroscopic sample yield Γ = -2.2 ± 0.1. Of the 120 stars with the most reliable data, 36 are probable and 9 are possible single-lined spectroscopic binaries. We also identify three new and eight candidate double-lined spectroscopic binaries. These data imply a lower limit on the massive binary fraction of 30%-42%. The calculated velocity dispersion for Cyg OB2 is 2.44 ± 0.07 km s-1, which is typical of open clusters. No runaway OB stars were found.

First Orbital Solution for the Non-thermal Emitter Cyg OB2 No. 9

After the first detection of its binary nature, the spectroscopic monitoring of the non-thermal radio emitter Cyg OB2 No. 9 (P = 2.4 yr) has continued, doubling the number of available spectra of the star. Since the discovery paper of 2008, a second periastron passage has occurred in 2009 February. Using a variety of techniques, the radial velocities could be estimated and a first, preliminary orbital solution was derived from the He I λ5876 line. The mass ratio appears close to unity and the eccentricity is large, i.e., 0.7-0.75. X-ray data from 2004 and 2007 are also analyzed in quest of peculiarities linked to binarity. The observations reveal neither large overluminosity nor strong hardness, but it must be noted that the high-energy data were taken after the periastron passage, at a time where colliding wind emission may be low. Some unusual X-ray variability is however detected, with a 10% flux decrease between 2004 and 2007. To clarify their origin and find a more obvious signature of the wind-wind collision, additional data, taken at periastron and close to it, are needed.

Erratum: “A Radial Velocity Survey of the Cygnus OB2 Association” (ApJ, 664, 1102 [2007])

The velocity dispersion of Cygnus OB2 was incorrectly quoted in the original manuscript as V 1⁄4 2:44 0:07 km s 1 for Vavg (the weighted average heliocentric velocities), and V 1⁄4 3:41 0:11 km s 1 for Vmid (the averages of the largest and smallest heliocentric velocities). The FWHM values for the Gaussian fits in Figure 23 are also incorrectly quoted as 5:70 0:17 km s 1 and 8:01 0:26 km s 1 for Vavg and Vmid, respectively. The correct one-dimensional velocity dispersions and corresponding FWHM values, as indicated by the Gaussian fits to the data in Figure 23, are V 1⁄4 5:74 0:15 km s 1 (Vavg), FWHM 1⁄4 13:50 0:36 km s 1 (Vavg), V 1⁄4 8:03 0:26 km s 1 (Vmid), and FWHM 1⁄4 18:87 0:60 km s 1 (Vmid). These values are larger than typical OB associations and open clusters (1Y2.5 km s 1; Bate & Bonnell, MNRAS, 356, 1201 [2005]). The larger dispersion estimates may be attributed to the uncertain systemic velocities of binaries (both identified and unidentified), owing to the limited orbital phase information from the current data set. The Astrophysical Journal, 681:735, 2008 July 1