2009ApJ...693..634O

We present the results of deep and high-resolution (FWHM ∼ 0".35) JHK near-infrared (NIR) observations with the Subaru telescope, to search for very low mass young stellar objects in the W3 Main star-forming region. The NIR survey covers an area of ∼ 2.6 arcmin² with 10σ limiting magnitude exceeding 20 mag in the JHK bands. The survey is sensitive enough to provide unprecedented details in W3 IRS 5 and IRS 3a regions and reveals a census of the stellar population down to objects below the hydrogen-burning limit. We construct JHK color-color and J - H/J and H - K/K color-magnitude diagrams to identify very low luminosity young stellar objects and to estimate their masses. Based on these color-color and color-magnitude diagrams, we identified a rich population of embedded YSO candidates with infrared excesses (Class I and Class II), associated with the W3 Main region. A large number of red sources (H - K > 2) have also been detected around W3 Main, which are arranged from the northwest toward the southeast regions. Most of these are concentrated around W3 IRS 5. We argue that these red stars are most probably pre-main-sequence (PMS) stars with intrinsic color excesses. We find that the slope of the K-band luminosity function of W3 Main is lower than the typical values reported for young embedded clusters. Based on the comparison between theoretical evolutionary models of very low mass PMS objects with the observed color-magnitude diagram, we find there exists a substantial substellar population in the observed region. The mass function does not show the presence of cutoff and sharp turnover around the substellar limit, at least at the hydrogen-burning limit. Furthermore, the mass function slope indicates that the number ratio of young brown dwarfs and hydrogen-burning stars in the W3 Main is probably higher than those in Trapezium and IC 348. The presence of mass segregation, in the sense that relatively massive YSOs lie near the cluster center, is seen. The estimated dynamical evolution time indicates that the observed mass segregation in the W3 Main may be the imprint of the star formation process.