Astron. Astrophys. 339, 759-772 (1998)

1. Introduction

Ultracompact H II regions (UCH II s) have long been known to be the postmark of the formation of massive stars. They are luminous, compact sources of infrared (IR) and radio continuum radiation caused by embedded stars of (Henning 1990, Churchwell 1991). The strong ultraviolet radiation of such stars produces dense, photo-ionized nebulae which are invisible at optical wavelengths (Wood & Churchwell 1989). Van Buren et al. (1990) suggested the so called bow-shock model for UCH II s of cometary shape, where the appearance is explained by motion through the ambient matter. This model also provides a containment mechanism for the ionized region to account for the so called "lifetime problem" (Wood & Churchwell 1989, hereafter WC89). Hollenbach et al. (1994) published a hypothesis about UCH II s being photo-evaporating circumstellar disks. This model also supplied an explanation for the longevity of the ultracompact phase (See also Yorke & Welz 1996). Even circumstellar disks and/or globules around lower mass stars that are being photo-evaporated by nearby massive stars ("proplyds", see, e.g., Churchwell et al. 1987; O'Dell & Wen 1994; McCullough et al. 1995) may appear as UCH II s. Testi et al. (1997) and Molinari et al. (1998) suggest an evolutionary model for UCH II s, where the radio emission might be suppressed during very early phases of massive star formation due to infall of surrounding dust. Recent observational studies suggest that for most of the models examples can be found which fit their implications (see Howard et al. 1994, Persi et al. 1997, Watson et al. 1998, Stecklum et al. 1998).

While observations in the near-infrared are common and needed because UCH II s are deeply embedded in large amounts of dust, it is also mandatory to use as high a spatial resolution as possible. This not only enables us to detect weak point sources against the extended nebular emission, but also is necessary for a precise identification of such sources with counterparts at different wavelengths. Now that adaptive optics systems have become widely available, it is worthwhile to make a new attempt in this direction. This paper is intended to be the first in a series of publications of high-resolution observations of UCH II s. The sources in this series where selected as bright sources from the WC89 sample with suitable (m_V 13 mag) optical stars close () to them to serve as wavefront sensor for the adaptive optics system.

In addition, we are able to use this example to show that UCH II s do not necessarily form around a single young massive star, but may also accompany the formation of a whole cluster of such stars. Indeed, we are able to compare G45.45+0.06 to the most well known OB cluster in our vicinity, the Orion Trapezium and the KL/BN region (named after Kleinmann & Low 1967 and Becklin & Neugebauer 1967).

In 1971, Wynn-Williams et al. made the first 6 and 11 cm aperture synthesis maps of the G45.5+0.1 complex using the Cambridge One Mile Telescope. They, and later Matthews et al. (1977), found G45.45+0.06 to be part of a cluster of three H II regions. The other two members of the cluster are G45.48+0.13 and G45.47+0.05 ¹. Throughout the paper we will refer to the ultracompact H II region G45.45+0.06 itself as G45, while other regions are denoted separately by their galactic coordinates. The cluster members all share a similar radial velocity of approximately 58 kms^-1 (Matthews et al. 1977). We adopt a distance of 6.6 kpc for G45, which was determined by Churchwell et al. (1990) after the Brand rotation curve (Brand 1986, Brand & Blitz 1993) from NH₃ measurements. Note that distance estimates for this source differ considerably in the literature but we consider direct line measurements and the Brand rotation curve the most reliable method for determining the distance. All authors resolve the near/far ambiguity to the near solution. The projected distances between the three regions range up to approximately 3 minutes of arc or 5.8 pc. Baud (1976) reported the discovery of a CO cloud coincident with the complex in both position and velocity.

G45 itself was classified as a cometary UCH II by WC89, after 6 cm VLA observations. From the measured radio flux, they determined the region to be ionized by a zero-age-main-sequence (ZAMS) star of spectral type O7.5. According to them, the IRAS fluxes suggest either an O4 star as the ionization source or a cluster with an O5.5 star as the brightest member. The cometary appearance was attributed to motion through ambient material. This motion provides at the same time a possible containment mechanism to explain the longevity of the ultracompact phase found for H II regions. However, Wilner et al. (1996) (hereafter W96) showed that the cometary structure of G45 is only a part of a larger shell structure, visible on their 3.6 cm VLA map. The shell structure was also observed at 6 cm by Garay et al. (1993). W96 also have detected clumps in the surrounding cloud, traced by HCO⁺(1-0) emission. They speculate about the expansion of G45 having provided the trigger for the more recent star formation in G45.47+0.05. Since all their clumps contain sufficient mass () to form massive stars, they speculate about the whole region being in an early stage of the formation of an OB star cluster. Mooney et al. (1995) used 1.3 mm continuum observations to determine the total mass of G45 to be .

In this paper we present the result of our infrared imaging campaign of G45. This campaign included high-resolution adaptive optics imaging in the near-infrared (NIR, H- and -band) as well "conventional" NIR imaging (Br) and imaging in the thermal infrared (MIR, L- and N-band). The very high resolution of the AO images reveals 15 compact sources inside G45. The wide wavelength range of our observations shows that all the sources detected are deeply embedded, some of them invisible even in the NIR. In Sect. 2, we briefly describe our observations and the follow-up data reduction techniques used, while the results are presented in Sect. 3. In the discussion (Sect. 4), we argue that the point sources are deeply embedded, young massive stars. This is also verified by using the Br image as an additional estimator for the extinction towards G45. From a multi-wavelength comparison between the morphological structures in the different images starting with our NIR images and ending with VLA data taken at cm wavelengths, we speculate about the history of the object and the possibility of triggered star formation.

© European Southern Observatory (ESO) 1998

Online publication: October 22, 1998
helpdesk.link@springer.de