Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.arcetri.astro.it/~lt/preprints/w75n_sio/w75n_sio.ps.gz Äàòà èçìåíåíèÿ: Tue Sep 11 16:57:04 2007 Äàòà èíäåêñèðîâàíèÿ: Sat Dec 22 08:00:57 2007 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: n 49

ApJ, in press (2004, v 601, Feb 1 issue)
The Nature of the Massive Young Stars in W75 N
D. S. Shepherd 1 , S. E. Kurtz 2 , L. Testi 3
ABSTRACT
We have observed the W75 N massive star forming region in SiO(J=2{1 &
J=1{0) at 3 00 5 00 resolution and in 6 cm, 2 cm, and 7 mm continuum emission
at 1:4 00 0:2 00 resolution. The abundance ratio of [SiO]/[H 2
]  5 7  10 11
which is typical for what is expected in the ambient component of molecular
clouds with active star formation. The SiO morphology is diuse and centered
on the positions of the ultracompact HII regions - no collimated, neutral jet was
discovered. The ionized gas surrounding the protostars have emission measures
ranging from 1 15  10 6 pc cm 6 , densities from 0:4 5  10 4 cm 3 , and
derived spectral types of the central ionizing stars ranging from B0.5 to B2.
Most of the detected sources have spectral indicies which suggest optically thin
to moderately optically thick HII regions produced by a central ionizing star.
The spread in ages between the oldest and youngest early-B protostars in the
W75 N cluster is 0:1 5  10 6 years. This evolutionary timescale for W75 N
is consistent with that found for early-B stars born in clusters forming more
massive stars (M ? > 25M ).
Subject headings: circumstellar matter { jets and out ows { stars: formation {
stars: mass loss { HII regions
1 National Radio Astronomy Observatory, P.O. Box 0, Socorro, NM 87801
2 Centro de Radioastronoma y Astrofsica, Universidad Nacional Autonoma de Mexico, Apdo. Postal
3-72, C.P. 58089, Morelia, Mich. Mexico
3 INAF { Osservatorio Astrosco di Arcetri, Largo Enrico Fermi 5, I-50125 Firenze

{ 2 {
1. INTRODUCTION
Molecular out ows from young, early-B protostars have many characteristics in common
with those from lower mass young stellar objects (YSOs) while the HII regions produced by
the central stars often look similar to those produced by O stars. For example, the out ow
momentum and the mass of circumstellar material both scale with the bolometric luminosity
of the driving source (e.g. Levreault 1988; Cabrit & Bertout 1992; Rodrguez et al. 1996;
Shepherd & Churchwell 1996; Chandler & Richer 2000). Some mid- to early-B YSOs have
ionized or molecular jets that are well-collimated close to the YSO (e.g. IRAS 20126+4014:
Hofner et al. 1999; Ceph A HW2: Torrelles et al. 1993; Rodrguez et al. 1994; Garay et
al. 1996); at least one source, HH 80{81, has a well-collimated, parsec-scale ionized jet that
appears to be a scaled version of a Herbig-Haro jet from a low-mass YSO (Mart, Rodrguez,
& Reipurth 1993; Heathcote, Reipurth, & Raga 1998). Yet despite these similarities, recent
observations have shown that the characteristics of early-B star out ows and disks may
also be diverging from their low-mass counterparts. In particular, molecular out ows from
early-B stars tend to be less collimated than those from low-mass YSO even when there is
a well-collimated, ionized jet (e.g. HH 80{81: Yamashita et al. 1989; IRAS 20126+4014:
Shepherd et al. 2000), and some out ows show no evidence for a collimated jet (G192.16{
3.82: Shepherd, Claussen, & Kurtz 2001), instead sporting a classic ultracompact (UC) HII
region at the protostellar position.
One early-B star cluster with out ows that may exhibit some dierences from low-
mass ows is W75 N: a massive star forming region with an integrated IRAS luminosity
of 1:4  10 5 L forming mid- to early-B stars (Haschick et al. 1981; Hunter et al. 1994;
Torrelles et al. 1997).
At the heart of the W75 N out ows is a cluster of four UC HII regions embedded in
a millimeter core W75 N:MM 1 4 . Haschick et al. (1981) identied three regions of ionized
gas in W75 N at a resolution of  1:5 00 : W75 N (A), W75 N (B), and W75 N (C). Hunter
et al. (1994) later resolved W75 N (B) with  0:5 00 resolution into three regions: Ba, Bb,
and Bc. Torrelles et al. (1997) then imaged W75 N (B) at  0:1 00 resolution, and detected
Ba and Bb (which they called VLA 1 & VLA 3), along with another weaker, and more
compact HII region, VLA 2. Within a 10 00 radius of MM 1 are three, compact millimeter
cores (MM 2-4). None of these sources have discernible near-infrared counterparts although
there is substantial near-infrared re ection nebulosity in the region (see, e.g., Figs 1, 5, & 10
from Shepherd, Testi, & Stark 2003, hereafter STS03). Mid-infrared emission at 12.5m has
been detected in the vicinity of the UC HII regions however it is unclear which source(s) are
4 Names of millimeter cores are shortened to MM 1-5 for the remainder of this paper

{ 3 {
producing the emission (Persi et al. 2003). An extended millimeter core (MM 5) is located
roughly 30 00 north of MM 1 and has an associated re ection nebula (W75 N A) and central
star visible in the infrared.
Multiple out ows have been identied originating from the cluster of UC HII regions and
millimeter cores with a total ow mass greater than 250 M (Fischer et al. 1985; Hunter et
al. 1994; Davis et al. 1998a, 1998b, Ridge & Moore 2001; Shepherd 2001; STS03, Torrelles et
al. 2003). Davis et al. (1998a,b) suggest the out ow is driven by a powerful, well-collimated
jet while STS03 nd no evidence for a jet. But is there an underlying, undetected neutral
jet driving the ow? And what are the properties of the HII regions and are they consistent
with what is expected for ionized gas around early-B zero-age-main-sequence (ZAMS) stars?
To answer these questions we have observed W75 N in SiO(J=2{1 & J=1{0) line emission to
search for evidence of a neutral jet and in centimeter & 7 mm continuum emission to obtain
a better understanding of the nature of the powering sources.
2. OBSERVATIONS
2.1. Owens Valley Radio Observatory
Observations in SiO(v=0, J=2{1) and SiO(v=1, J=2{1) were made with the Owens
Valley Radio Observatory (OVRO) millimeter-wave array of six 10.4 m telescopes between
1999 May 26 and 1999 November 12. The 64 channel spectral bandpass was centered on the
local standard of rest velocity (v LSR ) of 10 km s 1 with a spectral resolution of 1.726 km s 1
for SiO(v=0) and 1.738 km s 1 for SiO(v=1). Gain calibration used the quasar BL Lac
while observations of Neptune and/or Uranus provided the ux density calibration scale
with an estimated uncertainty of  20%. Calibration was carried out using the Caltech
MMA data reduction package (Scoville et al. 1993). Images were produced and analyzed
using the MIRIAD software package (Sault, Teuben, & Wright 1995). SiO(v=1, J=2{1)
was not detected. SiO(v=0, J=2{1) was detected and images were made with and without a
Gaussian taper of 3 00 FWHM to optimize sensitivity to extended structure and more compact
features, respectively. A summary of the observational parameters is presented in Table 1.
The largest angular scale that can be accurately imaged,  LAS , is  20 00 .

{ 4 {
Table 1: Observational Summary
Rest Beam Beam Peak y Total y
Image Freq FWHM P.A. RMS Flux Density Flux Density
(GHz) (arcsec) (deg) (mJy/beam) (mJy/beam) (mJy)
6 cm continuum 4.88 1:36  1:12 29.2 0.11 4.5 128.5
2 cm continuum 14.96 0:46  0:38 34.5 0.23 4.4 112.7
7 mm continuum 43.34 0:27  0:20 89.4 0.31 5.4 8.3
SiO(v=0, J=1{0) 43.42 0:40  0:36 65.8 4.0 14.0 14.0
SiO(v=0, J=2{1) 86.85 5:35  4:15 {60.3 45.0 640.0 42,300.
SiO(v=0, J=2{1) yy 86.85 3:17  2:56 {61.5 39.0 330.0 17,700.
SiO(v=1, J=2{1) 86.24 4:87  3:83 {68.3 45.0






y Flux densities measured in the primary beam corrected image. Total ux density given is the
combined value for all sources in the eld.
yy Higher resolution, the more extended emission, 58% of the total ux density, has been resolved
out.
2.2. Very Large Array
Observations of 43.3399 GHz (7 mm) continuum emission were made with the National
Radio Astronomy Observatory's 5 Very Large Array (VLA) in the \C" conguration on 2000
April 24 and in the \B" conguration on 2001 March 22. Baselines between 35 m and
11.4 km could detect a largest angular emission scale,  LAS  18 00 . The total bandwidth
was 200 MHz. The quasar 2012+464 was used as a phase calibrator and 3C286 was the
ux calibrator. The estimated uncertainty of the ux calibration is  10%. Calibration and
imaging was performed using the AIPS++ data reduction package. The data were imaged
using natural uv weighting and the image was deconvolved with a CLEAN-based algorithm.
Observations of 4.8851 GHz (6 cm) and 14.9649 GHz (2 cm) continuum emission were
made with the VLA in the \B" conguration on 2001 March 22. Baselines between 0.21 km
and 11.4 km detected a largest angular scale  LAS  26 00 at 6 cm and 9 00 at 2 cm. The quasar
2012+464 was used as a phase calibrator, the quasars 3C286 & 0410+769 were used as ux
calibrators. The estimated uncertainty of the ux calibration is  1% for both 2 cm and
6 cm. Calibration and imaging was performed using the AIPS++ data reduction package.
The data were imaged using robust uv weighting. The 2 cm image was deconvolved with
5 The National Radio Astronomy Observatory is a facility of the National Science Foundation operated
under cooperative agreement by Associated Universities, Inc.

{ 5 {
a CLEAN-based algorithm while the 6 cm image was deconvolved using both a standard
CLEAN-based algorithm and a multi-scale CLEAN algorithm. The multi-scale image was
CLEANed with six component scale sizes with diameters of 0 00 , :6 00 , 1:2 00 , 2 00 , 4 00 , & 6 00 .
Both 6 cm images gave essentially the same ux however the image generated with multiple
scales provided the best sensitivity to extended emission in the HII region W75 N A while
preserving the compact structure in the UC HII regions in W75 N B. In contrast, the 2 cm
(and 7 mm) observations resolved out most of the ux in W75 N A to the point where it
could not be imaged properly and the standard CLEAN algorithm was the most eective
deconvolution method.
SiO(v=0, J=1{0) line emission was observed with VLA in the \C" conguration on
2000 April 24. The observations were made using 32 channels centered on v LSR with
a spectral resolution of 2.7 km s 1 and a total bandwidth of 86.4 km s 1 . The quasar
2012+464 was used as a phase calibrator and 3C286 was the ux calibrator. The estimated
uncertainty of the ux calibration is  10%. Baselines between 35 m and 3.4 km provided a
synthesized beam of approximately 0:4 00 over the 1 0 eld of view. The largest angular scale
that the VLA is sensitive to at this frequency,  LAS , is  18 00 . Calibration and imaging
was performed using the AIPS data reduction package. The data were imaged using robust
uv weighting and deconvolved with a CLEAN-based algorithm. At 0:4 00 resolution, a single
14 mJy beam 1 peak (3) in one channel was recovered near the UC HII region positions.
Convolving the uv data with a Gaussian taper to yield a 3 00 beam resulted in a peak emission
of  30 mJy beam 1 .
3. RESULTS
3.1. Continuum Emission
The locations of the HII regions discussed in Section 1 are shown in images of the ionized
gas in 6 cm, 2 cm, and 7 mm continuum emission (Fig. 1). Peak and total ux densities are
given in Table 2 and the spectral energy distributions (SEDs) for each UC HII region are
shown in Fig. 2.
Continuum emission at 7 mm wavelength can be due to a mixture of warm dust and
ionized gas emission. Figure 2 (and Table 2) shows that the ux density of the UC HII regions
at 7 mm is consistent with or lower than what is expected for ionized gas (either optically
thin emission, S  /  0:1 , or moderately optically thick). How much 7 mm continuum
emission is expected to be due to warm dust? Using the 3 mm ux densities of Shepherd
(2001) and assuming the thermal dust emission between 3 & 7 mm has a spectral index of

{ 6 {
Table 2: Measured Flux Density (S  ) of Continuum Sources
Total Peak Total Peak Total Peak
Position S 6cm S 6cm S 2cm S 2cm S 7mm S 7mm Spectral
Source (h m s) ( ô 0 00 ) (mJy) ( mJy
beam ) (mJy) ( mJy
beam ) (mJy) ( mJy
beam ) Index y
VLA 1 (Ba) 20 38 36.455 +42 37 34.80 5.3 3.5 4.0 1.5





0:2  0:3
VLA 2 20 38 36.491 +42 37 34.30





1.5 1.2 2.6 2.2 0:4  0:1
VLA 3 (Bb) 20 38 36.491 +42 37 33.50 2.7 2.7 5.8 4.5 5.7 5.4 0:5  0:3
Bc 20 38 36.527 +42 37 31.50 4.4 2.8 1.7 1.2





-0:3  0:6
W75 N (A) 20 38 37.780 +42 37 59.00 116.1 4.5 > 99:7 0.8









y Spectral index derived from this data and data from Torrelles et al. (1997) and Hunter et al.
1994.
2, we expect to detect about 36 mJy of combined ux at 7 mm from thermal dust emission
near the UC HII regions. Only 8.3 mJy is recovered suggesting that at least 80% of the
thermal dust emission is being resolved out. In the millimeter cores MM 2, MM 3, & MM 4,
6{9 mJy is expected while none is detected, again consistent with the thermal dust emission
being resolved out by the 0:2 00 resolution.
Source Bc is detected at 6 and 2 cm with a spectral index of  = 0:3  0:6. In a
1.3 cm continuum image, Torrelles et al. (1997) found that Bc was marginally recovered at
the 3 level when a Gaussian taper was applied to the data suggesting that the source was
being resolved out by their higher 0:1 00 resolution. Thus, the negative spectral index derived
for Bc may be due to missing ux at short wavelengths (higher resolution) rather than an
intrinsic property of the source. Bc is not detected in 7 mm continuum emission with  0:2 00
resolution. Based on the peak ux density at 2 cm with 0:4 00 resolution, the expected peak
emission at 7 mm is  0:3 mJy beam 1 which is just under 1 in the 7 mm image. Thus,
the surface brightness of the emission in Bc is below our sensitivity limit.
VLA 3(Bb) is detected at all observed wavelengths. The source is marginally resolved
at 2 cm & 7 mm and the spectral index between 6 cm and 7 mm,  = 0:5 0:3, is consistent
with a moderately optically thick HII region.
VLA 2 is detected between 2 cm and 7 mm with a spectral index  = 0:40:1, consistent
with a moderately optically thick HII region. At 6 cm wavelength, the 1:2 00 resolution was
not adequate to isolate VLA 2 from VLA 1(Ba) or VLA 3(Bb).
VLA 1(Ba) is detected at 6 and 2 cm. The elongation of the 2 cm emission along the
H 2
O maser axis is consistent with that found by Torrelles et al. (1997). The spectral index
between 6 and 1.3 cm is  = 0:2  0:3. Assuming constant  = +0:2 between 1.3 cm

{ 7 {
and 7 mm, the expected total ux density at 7 mm due to ionized gas is 8.9 mJy while
the expected peak ux density should be roughly one fourth the peak found at 2 cm, i.e.,
 0:4 mJy beam 1 (assuming the peak emission scales with the area of the synthesized
beam). Figure 1 shows that there are a few peaks at the 3 level near the position of
VLA 1(Ba) however the surface brightness sensitivity is not adequate to recover the source
structure.
Continuum emission associated with the extended HII region W75 N A is detected at a
wavelength of 6 cm. At 2 cm, most of the ux has been resolved out, leaving only low-level
emission and a few 3 peaks. Previous images of the ionized gas in 6 cm continuum emission
were not able to recover the complex structure of this extended source due to limited uv
coverage (Haschick et al. 1981).
3.2. SiO emission
SiO(v=0, J=2{1) channel maps with  5 00 and 3 00 resolution are shown in Fig. 3. The
higher resolution image resolves out 58% of the ux density suggesting that a signicant
fraction of the emission is diuse. SiO(v=0, J=2{1) red-shifted (10 to 17.8 km s 1 ) and
blue-shifted (2.23 to 10 km s 1 ) emission are shown in Fig. 4. A zeroth moment image
covering the full velocity range is shown in Fig. 5 along with representative SiO spectra at
dierent locations in the cloud.
Despite the strong detection of SiO(v=0, J=2{1) with the OVRO interferometer, the
SiO(v=0, J=1{0) line could not be imaged well at a resolution of  0:4 00 . For low J-transition
lines, assuming local thermodynamic equilibrium, optically thin emission, and high excita-
tion temperatures (T ex  50 K), the brightness temperature of SiO(v=0, J=2{1) should
be roughly four times that of SiO(v=0, J=1{0) (Goldsmith 1972). Assuming the emission
arises in the same region, then:
S  (1 0)
S  (2 1)
= TB (1 0)
TB (2 1)
 2 (2 1)
 2 (1 0)
= 0:0625 (1)
Thus, SiO(J=1{0) peak ux densities of the order of 40 mJy beam 1 should be observed,
assuming similar uv coverage between the VLA and OVRO. Unfortunately, the uv coverage
of the interferometers diered signicantly. To compare the two images, a 10 k hole was cut
out of the OVRO data to match the hole in the VLA uv coverage. This provided a peak ux
density in the SiO(v=0, J=2{1) line of about 240 mJy beam 1 at a resolution of 3 00  2:5 00 .
Roughly one third of the total ux was recovered in this image. The VLA continuum-
subtracted data cube then had a Gaussian taper applied until the resolution matched that

{ 8 {
of OVRO. Although the taper down-weighted a signicant fraction of the VLA data, the
resulting peak emission in the SiO(v=0, J=1{0) line was  30 mJy beam 1 which is close to
what is predicted based on equation 1 above. This exercises demonstrates that the SiO(v=0,
J=1{0) emission in W75 N is relatively diuse.
SiO line widths range from about 3 { 10 km s 1 within the W75 N cloud (see Fig.
5). Assuming a rotational temperature of 50K, the total SiO gas mass traced by the (v=0,
J=2{1) line is 2  10 6
M while the average column density is 3:5  10 14 cm 2 . The
estimated total mass of the cloud core based on submillimeter continuum and CS(J=7-6)
observations is  1000 2000 M (Moore, Mountain, & Yamashita 1991; Hunter et al. 1994)
which implies an SiO/[H 2
] ratio of  5 7  10 11 . The measured, average SiO fractional
abundance, [SiO]/[H 2
], in typical low-mass dark clouds is between 10 11 and 10 12 (Ziurys,
Friberg, & Irvine 1989) while the average ambient abundances in clouds with active star
formation range from 10 10 to 10 11 (Codella, Bachiller, & Reipurth 1999). In comparison,
SiO abundance in the quiescent ridge of Orion is less than 310 10 (Blake et al. 1987). The
SiO abundance calculated for the W75 N cloud is consistent with the typical abundance for
ambient gas in both high and low-mass star forming regions.
Note, SiO abundance can be as high as 10 5 10 6 in high-velocity gas (> 20 km s 1
from the systemic velocity of the cloud) which is directly associated with high and low-
mass molecular out ows (see, e.g., Martin-Pintado, Bachiller, & Fuente 1992; Shepherd,
Churchwell, & Wilner 1997). However, the average abundance enhancement is generally
several orders of magnitude less as discussed above.
4. DISCUSSION
4.1. SiO emission and the molecular out ows
Based on a comparison between CO(J=1{0), H 2
, and [FeII] emission, STS03 suggested
that only slow, non-dissociative J-type shocks exist throughout the parsec-scale out ows
produced by the central stars in the W75 N (B) UC HII regions. Fast, dissociative shocks,
common in jet-driven low-mass out ows, appear to be absent in W75 N. Thus, the energetics
suggest that the out ows from the mid- to early-B protostars in W75 N are not simply scaled-
up versions of low-mass out ows. Further, there was no evidence for well-collimated, parsec-
scale jets such as those seen in ows from lower mass protostars. However, the observations
of STS03 could not rule out the presence of an underlying neutral jet that could drive the
CO out ows.
SiO emission in molecular ows is excited in shocks where silicon is rst removed from

{ 9 {
dust grains and then reacts with OH radicals to form SiO in the post-shock cooling zone. The
gas phase abundance of SiO can increase up to a factor of 10 6 over that found in quiescent
molecular clouds and can delineate the axis of highly collimated jet-driven out ows and/or
the bow-shock where the head of a jet interacts with dense molecular material (e.g. Haschick
& Ho 1990). SiO 'jets' have been detected in well-collimated out ows from low-mass YSOs
(e.g. L1448: Guilloteau et al. 1992; HH 211: Chandler & Richer 2001) and in at least one
massive out ow from an early-B star (IRAS 20126+4014: Cesaroni et al. 1997, 1999). Thus,
SiO has the potential to uncover collimated, molecular jets that may not be obvious in other
tracers.
Our SiO observations cover the central 60 00 eld of the molecular out ows mapped by
STS03 (the full 5 0  1:5 0 mosaic was not covered). There is physically diuse SiO(J=2{
1) and SiO(J=1{0) emission centered near the positions of the UC HII regions. The SiO
abundance is roughly a factor of 10 higher than abundances toward dark, quiescent clouds
and is consistent with what is expected for ambient gas in in both low- and high-mass star
forming regions. Figure 4 shows the relation between the SiO emission, the locations of
the UC HII regions and millimeter cores 2{4, and the proposed out ows from VLA 1 (Ba),
VLA 3 (Bb), and MM 2 suggested by STS03. There is no clear relationship between the
SiO distribution and the proposed out ows. The higher resolution (3 00 ) SiO images (Fig. 3
and right panel of Fig. 4) were made in an attempt to resolve out the extended emission
associated with the ambient gas and search for compact, high-velocity, red- and blue-shifted
emission that may delineate collimated out ows. There is no clear indication of a jet-like
structure from any specic source despite the presence of VLA 1 (Ba), an HII region that
appears to be excited by a thermal jet.
On a larger scale than was observed in SiO in this work, STS03 found clear evidence for
shocked gas associated with the out ows as seen from the H 2
line morphology. The shocked
gas is diuse and appears to be caused by interactions between the ambient medium, wide-
angle out owing gas, and ionized gas. Our SiO observations of the center 60 00 near the
out ow driving sources show FWHM line widths up to 10 km s 1 (Fig. 5) which suggests
that the SiO is produced in shocks. However, the resolution and sensitivity are not suôcient
to determine if the SiO abundance enhancement is associated with the CO ows or if it arises
from a shocked boundary between the ambient medium and the ionized wind from the UC
HII regions.

{ 10 {
4.2. Physical properties of the HII regions
Three of four UC HII regions in W75 N (B) (VLA 1 (Ba), VLA 2, & VLA 3 (Bb))
display evidence for on-going out ow/accretion based on high-velocity molecular gas traced
to the source and/or the presence of H 2
O or OH masers (Baart et al. 1986, Hunter et
al. 1994, Torrelles et al. 1997, STS03, Torrelles et al. 2003). Thus, it is likely that the
ionized gas produced by the central star can escape along the out ow axis. The remaining
two HII regions (Bc in W75 N (B) & W75 N (A)) are more extended structures that
are recovered in  1:2 00 resolution images at 6 cm but are resolved out at 2 cm with
higher resolution ( 0:4 00 ). As discussed in Wood & Churchwell (1989, hereafter WC89),
complicated geometries present diôculties for interpretation because physical parameters
such as density and surface brightness along a line of sight depend on source structure.
Following the method outlined by WC89, we use the integrated ux densities when knowledge
of the source structure is not required and, when geometry is important, we estimate
peak values using the peak ux densities per beam. Table 3 presents the derived physical
parameters of the HII regions in W75 N. For each source, the values listed are: , the
frequency at which the derivations were made;
s, line-of-sight depth at the peak position
(taken to be the projected diameter of a sphere on the sky); T b , the synthesized beam
brightness temperature;   , the peak optical depth assuming the beam is uniformly lled
with T e = 10 4 K ionized gas; EM, the emission measure in units of 10 7 pc cm 6 ; n e , the
RMS electron density in units of 10 4 cm 3 ; U, the excitation parameter of the ionized gas;
N L , the number of Lyman continuum photons required to produce the observed emission
assuming an ionization-bounded, spherically symmetric, homogeneous HII region; and nally,
the spectral type of the central star assuming a single ZAMS star is producing the observed
Lyman continuum ux (Panagia 1973, WC89).
Table 3: Derived Parameters for HII regions
Peak Values from Observed Integrated Values from
Flux Density per Synthesized Beam Integrated Flux Density

s T b   EM/10 7 n e =10 4 U LogNL Spectral
Source (GHz) (pc) (K) (pc cm 6 ) (cm 3 ) (pc cm 2 ) (s 1 ) Type
VLA 1 (Ba) y 14.96 0.006 52 0.005 1.5 5.0 3.2 45.01 B1
VLA 2 14.96 0.004 42 0.004 0.38 3.2 2.3 44.59 B2
VLA 3 (Bb) 14.96 0.007 159 0.016 1.1 4.1 3.6 45.17 B1
Bc 4.88 0.024 108 0.011 0.09 0.6 3.2 45.00 B1.5
W75 N (A) 4.88 0.121 172 0.017 0.15 0.4 9.9 46.42 B0.5
y Should be considered an upper limit due to likely strong contamination by ionizing ux produced
by shock waves in the jet (see text for discussion).

{ 11 {
There are a number of errors associated with the values in Table 3 that are diôcult to
estimate due to our limited knowledge of the detailed source structure. First, peak properties
in Table 1 assume the beam is uniformly lled with 10 4 K gas. However, the UC HII regions
are unresolved, thus peak properties should be considered lower limits due to beam dilution
eects. Second, derivations based on the integrated ux density should be considered lower
limits for sources with known out ows since the ionized gas can escape along the out ow
axis. Third, there is no correction for dust absorption within the ionized gas, which would
tend to underestimate N L , and hence the spectral type of the star. Fourth, shock waves
within the out ow are expected to contribute to the total ionizing ux, which would result
in an overestimate of N L . Finally, accretion rates above 10 5 to 10 4
M yr 1 , typical
for early-B protostars, may inhibit the formation of a UC HII region near the equatorial
plane where accretion is highest (see, e.g., Churchwell 1999 and references therein). Despite
these uncertainties, the derivations are probably accurate to within a spectral type, except,
perhaps, for VLA 1 (Ba). Due to the elongation of the ionized gas along the molecular
out ow axis and the presence of H 2
O masers along the axis, Torrelles et al. (1997) & STS03
argue that VLA 1 (Ba) is a thermal jet source. Thus, a signicant fraction of the observed
centimeter continuum emission is likely due to the ionized jet rather than emission from an
ionization-bounded UC HII region produced by a central, massive star. If this is true, then
the estimated spectral type of VLA 1 (Ba) should be considered an upper limit.
For the sources VLA 2 and VLA 3 (Bb), our estimates of the physical properties of
the UC HII regions are lower than those of Torrelles et al. (1997) for two reasons: 1) our
estimates based on peak emission likely suer from beam dilution; and 2) the spectral indicies
we estimate from ts of the SEDs suggest optically thin or slightly optically thick emission
while Torrelles et al. estimated that the emission was signicantly optically thick (based on
only two data points). Spectral types from both estimates still only dier by a spectral type.
Comparison of the values in Table 3 with those in WC89 (their Table 17), shows that
the physical parameters of the ionized gas in the W75 N sources are consistent with ZAMS
stars with spectral types later than B0. Peak values of the emission measure and n e in
the more extended sources Bc & W75 N (A) are roughly an order of magnitude less than
what is found in the more compact HII regions. In the absence of connement, the radius
of an HII region is expected to increase with time as the ionization front expands to form
an increasingly larger Stromgren sphere with subsequently smaller electron densities. Thus,
the lower peak values and increased size of the HII region are consistent with these sources
being more evolved than the more compact UC HII regions in W75 N (B). Bc and W75 N
(A) also show no evidence for driving an out ow; again consistent with the sources being
more evolved.

{ 12 {
STS03 assumed all millimeter continuum emission from W75 N (A) was due to thermal
dust and they calculated a mass of gas and dust to be 68 M . With a good centimeter
continuum image, it is now possible to estimate the likely contribution due to ionized gas
and obtain a better estimate for the molecular cloud mass traced by warm dust emission
surrounding the Stromgren sphere. Assuming the 6 cm emission from W75 N (A) is optically
thin (S  /  0:1 ), the expected ux density at 2.7 mm due to ionized gas is 85 mJy. STS03
measured a total ux density at 2.7 mm of 129.5 mJy, thus the expected mass in the dust
shell surrounding W75 N (A) is  23 M (see STS03 for a discussion of the assumptions
and errors associated with this estimate).
4.3. Timescale for the formation of early-B stars in W75 N
The timescale for the formation of O star clusters for which M ? > 25 M appears
to be less than 3 Myrs (Massy, Johnson, & DeGioia-Eastwood 1995). In all the clusters
studied by Massy et al., there was evidence for the continued formation of mid to early-B
stars (5-10 M ) at least 1 Myrs after the formation of the O stars. Do clusters where the
most massive members are early-B stars have a similar age spread as that found for clusters
forming stars more massive than 25 M ?
Within a radius of 30 00 ( 0:3 pc) from the W75 N (B) UC HII regions are three young
early-B stars: the central star in W75 N (A), IRS1 and IRS2 (see, e.g., Fig. 10 of STS03).
The early-B stars are embedded in a  1000 2000 M molecular cloud from which multiple
ows are emerging with a combined out ow mass of at least 250 M (Moore et al. 1991,
Hunter et al. 1994, STS03). The out ows are driven by at least two of the stars embedded
in UC HII regions and one millimeter core (MM 2). No high velocity molecular gas can be
traced to the early B stars which are seen in the infrared (W75 N (A), IRS1, and IRS2)
which suggests that the infrared stars are older than the central stars of the UC HII regions.
From the size and velocity of the CO out ows, STS03 estimate that the central stars of the
UC HII regions are  10 5 years old.
W75 N (A), a classic example of an expanding Stromgren sphere (Fig. 6), appears to be
the oldest B-type star of the cluster. The central B0.5 star is detected in the near-infrared
with an irregular re ection nebula surrounding the star. A 12 00 diameter sphere of ionized
gas surrounds the star (0.12 pc at a distance of 2 kpc) which, in turn, is enclosed in an
18 00 (0.17 pc) diameter shell of molecular gas. The near-infrared colors are consistent with
foreground extinction from the molecular shell (A V  20, STS03). IRS1 and IRS2, on the
other hand, have excess emission at 2m suggesting that they have not had time to disperse
their circumstellar material via photoevaporation and stellar winds. Thus, IRS1 & IRS2

{ 13 {
appear to be of an intermediate age between W75 N (A) and the embedded stars in the UC
HII regions.
Using the 6 cm observations of W75 N (A), we derive the age of W75 N (A) and hence
an estimate of the age spread of the early-B stars in W75 N. The age of W75 N (A) can
be derived in two ways: 1) from an estimate of the expansion time required for the HII
region to reach its current radius; and 2) assuming the central star formed via accretion,
from an estimate of the time it would take for the remnant accretion disk to be photo-
evaporated. For the rst case, the expansion of an ionization-front at the boundary of an
expanding Stromgren sphere increases with time from some initial radius which depends on
the ionizing ux from the central star and the density of the ambient medium (Dyson &
Williams 1980). Assuming a strong shock approximation at the boundary of the ionization
front, the initial radius, r i , is given by:
r i =

3N L
4 n 2
o  2
! 1
3
(2)
and the time required for the HII region to expand to a radius r(t) in a uniform density
environment is:
 =
0
@
"
r(t)
r i
# 7
4
1
1
A
 4r i
7c i

(3)
where
N L = the ux of ionizing photons
n o = density of the ambient medium
 2 = recombination coeôcient to all levels except the ground state at temperature, T e
c i = sound speed in ionized gas ( 10 km s 1 )
Assuming n o = 2  10 7 cm 3 (DePree, Rogrguez, & Goss 1995), T e = 10 4 K,  2
=
2:6  10 10 T 3=4
e cm 3 s 1 , and N L = 2:6  10 46 s 1 for this B0.5 star, then the initial
radius of the Stromgren sphere is r i = 4  10 9 km and the expansion timescale for the
W75 N (A) HII region is  = 1:2  10 6 years.
As discussed in DePree, Rogrguez, & Goss (1995), this calculation assumes that the
molecular gas has a constant density and innite extent. If this were the case, then one
would expect W75 N (A) to reach pressure equilibrium at a radius of only 0.01 pc, which
does not match the observed radius of > 0:1 pc. Instead, we expect the molecular gas
density to decrease with radius which suggests that, for timescales >
 10 5 years, UC HII
regions are probably not in pressure equilibrium and should be expanding. At the same
time, the molecular material surrounding the Stromgren sphere is expected to expand at a
slower rate ( 1 km s 1 ) due to the lower temperature and molecular composition. Thus,

{ 14 {
the exact expansion timescale of the HII region and the stability of the molecular envelope
surrounding the ionized gas are quite uncertain.
For the second method, we assume the central star formed via accretion. Although
competing theories exist for how the most massive O stars formed (e.g. coalescence or
accretion), observational evidence for disks around early-B stars and the similarity between
out ows produced by low-mass stars and those from mid- to early-B stars, suggest that stars
up to spectral type B0 or O9 most likely form via accretion (see, e.g., the review by Shepherd
2003 and references therein). If the central star of W75 N (A) once had a massive accretion
disk, then the lifetime of the UC HII region could be lengthened due to the photoevaporation
of the circumstellar disk by the stellar wind (Hollenbach et al. 1994). The disk material
would have provided high-density, ionized gas thus, the UC HII region would persist as long
as the disk survived the mass loss (assuming the disk is no longer being fed by material from
the surrounding molecular core). Since the near-infrared colors suggest there is no current
disk, we assume the disk has been completely photo-evaporated and the timescale for this
to occur would represent a lower limit to the age of W75 N (A). For the \weak wind" case
of Hollenbach et al. appropriate for early-B stars, the lifetime of the disk is given by:
 disk = 7  10 4  1=2
49
M 1=2
1
M d [yrs] (4)
where
 49 = ionizing Lyman continuum ux in units of 10 49 s 1
M 1
= the mass of the central star in units of 10 M
M d = disk mass in units of M
For W75 N (A),  49 = 2:6  10 3 and M 1  1:5. Shu et al. (1990) showed that an accretion
disk becomes gravitationally unstable when it reaches a mass of M d  0:3M ? where M ? is the
mass of the central protostar. During the initial collapse of the cloud core, the disk mass may
be maintained close to the value of 0:3M ? . When infall ceases and the disk mass falls below
the critical value, disk accretion onto the star may rapidly decline and photoevaporation
may be the dominant mechanism which disperses the remaining gas and dust (Hollenbach
et a. 1994). Based on this scenario, we assume an initial disk at the edge of stability, that
is M d  0:3M ? = 4:5M  . Errors in the estimate for the photoevaporative timescale would
scale directly as M d . We nd that  disk = 5  10 6 years.
Both estimates for the lifetime of W75 N (A) have numerous assumptions about, e.g.
characteristic cloud densities, disk mass, & temperature of the ionized gas. None-the-less,
they are probably reasonable to within an order of magnitude. These derivations suggest
that the B0.5 star in W75 N (A) is roughly 1 5  10 6 years old while the youngest B stars
forming are  10 5 years old. Thus, the spread in ages between young B-stars in this cluster
is
 = 0:1 5  10 6 years.

{ 15 {
Efremov & Elmegreen (1998) and Elmegreen et al. (2000) suggest that the duration of
star formation tends to vary with the size, S, of the cluster as something like the crossing time
for turbulent motions, e.g.
 / S 0:5 . Given a cluster diameter of W75 N of  1 0 (0.6 pc),
the expected age would be roughly 0.8 Myrs, which is somewhat lower than our estimates
although still within the errors. Comparing with other observations, clusters forming stars
with M ? > 25 M have a typical spread in ages,
 , of about 2 Myrs for the O stars while
mid- to early-B stars continue to form for at least another million years (Massy, Johnson, &
DeGioia-Eastwood 1995). Thus, the early-B stars in W75 N are formed over a period that
is consistent with the timescale for early-B stars formed in clusters with more massive stars.
5. SUMMARY
We have observed the W75 N massive star forming region in SiO(J=2{1) & (J=1{0)
to search for well-collimated neutral jets from the early-B protostars and in centimeter and
7 mm continuum emission to examine the nature of the driving sources. The SiO emission
is diuse with no clear indication of a neutral, collimated jet from the region. This does
not, however, completely rule out the presence of a jet since enhanced SiO emission does not
always trace known jet structure.
The ionized gas surrounding the protostars have emission measures, densities, and
derived spectral types which are consistent with early-B stars. Most of the detected sources
have spectral indicies which suggest optically thin to moderately optically thick HII regions
produced by a central ionizing star.
By comparing the oldest and youngest B stars in the cluster, an estimate for the duration
of early-B star formation in W75 N is obtained. The oldest star in the cluster is roughly
1 5  10 6 years old while the youngest, B protostars are  10 5 years old. Thus, the spread
in ages is 0:1 5  10 6 years. The age spread for W75 N is consistent with that found for
early-B stars born in clusters forming more massive stars (M ? > 25M ).
Acknowledgments: Research at the Owens Valley Radio Observatory is supported by
the National Science Foundation through NSF grant number AST 96-13717. Star formation
research at Owens Valley is also support by NASA's Origins of Solar Systems program, grant
NAGW-4030 and by the Norris Planetary Origins Project. S. Kurtz acknowledges support
from Project IN118401 and CONACyT Project E-36568. D. Shepherd would like to thank
Sally Oey for useful discussions on the evolutionary timescales of young clusters.

{ 16 {
REFERENCES
Baart, E. E., Cohen, R. J., Davies, R. D., Norris, R. P., & Rowland, P. R. 1986, MNRAS,
219, 145
Blake, G. A., Sutton, E. C., Masson, C. R., & Phillips, T. G. 1987, ApJ, 315, 621
Cabrit, S. & Bertout, C. 1992, A&A, 261, 274
Cesaroni, R., Felli, M., Testi, L., Walmsley, C. M., & Olmi, L. 1997, A&A, 325, 725
Cesaroni, R., Felli, M., Jenness, T., Neri, R., Olmi, L., Robberto, M., Testi, L. & Walmsley,
C. M. 1999, A&A, 345, 949
Chandler, C.J. & Richer, J.S. 2000, ApJ, 530, 851
Chandler, C. J. & Richer, J. S. 2001, ApJ, 555, 139
Churchwell, E. 1999, in \The Origin of Stars and Planetary Systems", ed. Charles J. Lada
and Nikolaos D. Kylas (Kluwer Academic Publishers), 515
Codella, C., Bachiller, R., & Reipurth, B. 1999, A&A, 343, 585
Davis, C. J., Moriarty-Schieven, G. H., Eisloel, J., Hoare, M.G., & Ray, T.P. 1998a, AJ,
115, 1118
Davis, C. J., Smith, M. D., & Moriarty-Schieven, G. H. 1998b, MNRAS, 299, 825
DePree, C. G., Rogrguez, L. F., Goss, W. M. 1995, Rev. Mex. Astron. Astros., 31, 39
Dyson, J. E., Williams, D. A., 1980, Physics of the Interstellar Medium (New York: Halsted
Press), p 204.
Efremov, Y. N. & Elmegreen, B. G. 1998, MNRAS, 299, 588
Elmegreen, B. G., Efremov, Y., Pudritz, R. E., & Zinnecker, H. 2000, in \Protostars and
Planets IV", ed. V. Mannings, A. Boss & S. Russell (Tucson: Univ. of Arizona
Press), 179
Fischer, J., Sanders, D. B., Simon, M., & Solomon, P. M. 1985, ApJ, 293, 508
Garay, G., Ramirez, S., Rodriguez, L. F., Curiel, S., & Torrelles, J. M. 1996, ApJ, 459, 193
Goldsmith, P. F., 1972, ApJ, 176, 597
Guilloteau, S., Bachiller, R., Fuente, A., Lucas, R. 1992, A&A, 265, 49
Haschick, A. D. & Ho, P. T. P. 1990, ApJ, 352, 630
Haschick, A. D., Reid, M. J., Burke, B. Fl, Moran, J. M., & Miller, G. 1981, ApJ, 244, 76

{ 17 {
Heathcote, S., Reipurth, B., & Raga, A.C. 1998, AJ, 116, 1940
Hofner, P., Cesaroni, R., Rodrguez, L. F., & Mart, J. 1999, A&A, 345, L43
Hollenbach, D., Johnstone, D., Lizano, S., & Shu, F. 1994, ApJ, 428, 654
Hunter, T. R., Taylor, G. B., Felli, M., & Tofani, G. 1994, A&A, 284
Levreault, R.M. 1988, ApJ, 330, 897
Massy, P., Johnson, K. E., & DeGioia-Eastwood, K. 1995, ApJ, 454, 151
Mart, J., Rodrguez, L.F., Reipurth, B. 1993, ApJ, 416, 208
Martin-Pintado, J., Bachiller, R., & Fuente, A. 1992, A&A, 254, 315
Moore, T. J. T. Mountain, C. M., Yamashita, T. 1991, MNRAS, 248, 79
Panagia, N. 1973, AJ, 78, 929 properties of HII regions.
Persi, P., Marenzi, A. R., Tapia, M. & Bohigas, J. 2003, Mem S.A.It., 74, 146
Ridge, N. A. & Moore, T. J. T. 2001, A&A, 378, 495
Rodrguez, L.F., Garay, G., Curiel, S., Ramirez, S., Torrelles, J.M., Gomez, Y., Velazquez,
A. 1994, ApJ, 430, L65
Rodrguez, L.F., Garay, G., Curiel, S., Ramirez, S., Saraceno, P., Andre, P., Ceccarelli, C.,
Griôn, M., & Molinari, S. 1996, A&A, 309, 827
Sault, R. J., Teuben, P. J., & Wright, M. C. H. 1995, in \Astronomical Data Analysis
Software and Systems IV", ed. R. A. Shaw, H. E. Payne, & J. J. E. Hayes, PASP
Conf Series 77, 433
Scoville, N. Z., Carlstrom, J. E., Chandler, C. J., Phillips, J. A., Scott, S. L., Tilanus, R. P.
J., & Wang, Z. 1993, PASP, 105, 1482
Shepherd, D. S., & Churchwell, E. 1996, ApJ, 472, 225
Shepherd, D. S., Churchwell, E., & Wilner, D. J. 1997, ApJ, 482, 355
Shepherd, D. S., Yu, K. C., Bally, J., & Testi, L. 2000, ApJ, 535, 833
Shepherd, D. S. 2001, ApJ, 546, 345
Shepherd, D. S., Claussen, M. J., & Kurtz, S. E. 2001, Science, 292, 1513
Shepherd, D. S., Testi, L., Stark, D. P. 2003, ApJ, 584, 882 (STS03)
Shepherd, D. S. 2003, in \Galactic Star Formation Across the Stellar Mass Spectrum," ASP
conference series, in press

{ 18 {
Shu, F., Tremaine, S., Adams, F., & Ruden, S. 1990, ApJ, 358, 495
Torrelles, J.M., Verdes-Montenegro, L., Ho, P.T.P., Rodriguez, L.F., Canto, J. 1993, ApJ,
410, 202
Torrelles, J. M. Gomez, J. F., Rodrguez, L. F., Ho, P. T. P., Curiel, S., & Vazquez, R.
1997, ApJ, 489, 744
Torrelles, J. M., Patel, N. A., Anglada, G., Gomez, J. F., Ho, P. T. P., Lara, L., Alberdi,
A., Canto, J., Curiel, S., Garay, G., & Rodrguez, L. F., 2003, ApJL, submitted
Wood, D. O. S. & Churchwell, E. 1989, ApJS, 69, 831 (WC89)
Yamashita, T., Suzuki, H., Kaifu, N., Tamura, M., Mountain, C. M., Moore, T. J. T. 1989,
ApJ, 347, 894
Ziurys, L. M., Friberg, P., & Irvine, W. M. 1989, ApJ, 343, 201
This preprint was prepared with the AAS L A T E X macros v4.0.

{ 19 {
Figure Captions
Figure 1. Continuum emission from W75 N (A) and UC HII regions within W75 N (B) at
6 cm (top), 2 cm (lower left), and 7 mm (lower right) wavelength is plotted in both contours
and greyscale. Top: The 6 cm image RMS is 0.11 mJy beam 1 ; contours are plotted at
{3, 3, 5  and continue with spacings of 5 ; greyscale is displayed on a linear scale from
0.33 mJy beam 1 to 4.5 mJy beam 1 . The synthesized beam, plotted in the lower left corner,
is 1:36 00  1:12 00 at position angle 29.2 ô . The extended emission from W75 N (A) coincides
with the millimeter core MM 5 and a near-infrared re ection nebula and star (Shepherd,
Testi, & Stark 2003). Locations of UC HII regions in W75 N (B) that are detected in 6 cm
continuum emission are identied by lled triangles. The dashed box around the W75 N B
sources shows the area displayed in the lower two images at 2 cm and 7 mm. Lower left: The
2 cm image RMS is 0.23 mJy beam 1 ; contours are plotted at {3, 3, 4, 5, 6, 8,  and continue
with spacings of 2 ; greyscale is displayed from 0.46 mJy beam 1 to 4.42 mJy beam 1 .
The synthesized beam, plotted in the lower right corner, is 0:44 00  0:36 00 at P.A. 31.4 ô .
Lower right: The 7 mm image RMS is 0.31 mJy beam 1 ; contours are plotted at {3, 3,
5, 7  and continue with spacings of 2 ; greyscale is displayed from 0.62 mJy beam 1 to
5.4 mJy beam 1 . The synthesized beam, plotted in the lower right corner, is 0:27 00  0:20 00
at P.A. 89.4 ô . UC HII region Bc is not detected. The locations of the UC HII regions in the
eld are indicated with lled triangles. Water masers are shown as crosses (Torrelles et al.
1997); OH maser positions are represented as lled circles (Baart et al. 1986).
Figure 2. The spectral energy distributions of the UC HII regions in W75 N (B). Asterisks
represent data from this work, squares represent data from Hunter et al. (1994), circles are
from Torrelles et al. (1997), and the triangle represents the 0:9 00 resolution 1 mm data
from Shepherd (2001). Upper limits (3) are shown as symbols with arrows. In all cases,
estimated errors are smaller than the symbols. Solid lines show linear least squares ts to
the SEDs for data between 6 cm and 7 mm. The slope of the t (spectral index) is shown
in the lower right corner of each plot.
Figure 3. Top: SiO(v=0, J=2{1) channel maps at 1.726 km s 1 spectral resolution
between 0.5 and 19:5 km s 1 . The velocity is indicated in the upper right of each panel. The
RMS is 45 mJy beam 1 . Contours are plotted from 3; 4; 5; 6  and continue with a spacing
of 2 . The last panel (velocity 19.5 km s 1 ) shows the synthesized beam in the lower right
corner (5:35 00  4:15 00 at P.A. 60:3 ô ) and a scale size of 0.1 pc is represented by a bar in
the lower left corner. The locations of four millimeter continuum peaks in W75 N (B) are
shown as plus signs (MM 1 { MM 4, Shepherd 2001) while positions for the UC HII regions
VLA 1(Ba), VLA2, VLA 3(Bb), and Bc are indicated by small triangles (Hunter et al. 1994;
Torrelles et al. 1997). Bottom: SiO(v=0, J=2{1) channel maps made with a higher spatial

{ 20 {
resolution of 3:17 00  2:56 00 at P.A. 61:5 ô (beam shown in lower right panel). The RMS is
39 mJy beam 1 . Contours are plotted from 3; 4; 5; 6  and continue with a spacing of 2 .
Figure 4. Blue-shifted (2.23 to 10 km s 1 ; thin lines) and red-shifted (10 to 17.8 km s 1 ;
thick lines) SiO(v=0, J=2{1) emission contours. The left panel shows the lower resolution
data while the right panel shows the higher resolution data. The synthesized beam is shown
in the lower right corner of each image. Millimeter cores MM 2, MM 3, and MM 4 are shown
as lled circles, UC HII regions within MM 1 are indicated by lled triangles. Position angles
of out ows proposed by Shepherd, Testi, & Stark (2003) are illustrated by arrows.
Figure 5. W75 N SiO(v=0, J=2{1) zeroth moment map and SiO spectra. The SiO(v=0)
image (lower left) has an RMS of 0.3 mJy beam 1 km s 1 ; contours begin at 3; 4; 6 and
continue with spacings of 2. The synthesized beam of 5:35 00  4:15 00 at P.A. 60:3 ô is
shown in the lower right corner while a scale size of 0.1 pc is shown in the lower left corner.
Millimeter cores MM 2, MM 3, and MM 4 are shown as lled circles, UC HII regions within
MM 1 are indicated by lled triangles. SiO(v=0, J=2{1) spectra are shown at dierent
locations in the cloud. The dashed vertical line in each plot represents v LSR = 10 km s 1 .
Figure 6. W75 N (A): White contours represent the 6 cm continuum emission tracing
ionized gas in the Stromgren sphere surrounding the central B0.5 star (contours are the same
as in Fig. 1). Greyscale shows the near-infrared K-band re ection nebula and central star
while the black contours represent the 3 mm continuum emission from the warm dust shell
surrounding the ionized gas (from Fig. 5 of STS03).

{ 21 {
Figure 1. Continuum emission at 6 cm (top), 2 cm (lower left), and 7 mm (lower right).

{ 22 {
Figure 2. SEDs of UC HII regions

{ 23 {
Figure 3. SiO(v=0, J=2{1) channel maps,  5 00 (top) and 3 00 (Bottom) resolution.

{ 24 {
Figure 4. SiO(v=0, J=2{1) red and blue-shifted emission. Low-resolution image (left)
and higher resolution image (right).

{ 25 {
Figure 5. SiO zeroth moment map and spectra.

{ 26 {
Figure 6. W75 N A: greyscale = K band infrared - re ection nebulosity; white contours =
6cm continuum - ionized gas; black contours = 3mm continuum - warm dust.