class i methanol masers: signposts of star formation? - IOPscience

The Astronomical Journal, 135:1718–1730, 2008 May c 2008. The American Astronomical Society. All rights reserved. Printed in the U.S.A.

doi:10.1088/0004-6256/135/5/1718

CLASS I METHANOL MASERS: SIGNPOSTS OF STAR FORMATION? Preethi Pratap1 , P. A. Shute1 , Thomas C. Keane2 , Cara Battersby3,4 , and Sarah Sterling2 1

MIT Haystack Observatory, Off Route 40, Westford, MA 01886, USA; [email protected] 2 Russell Sage College, 45 Ferry Street, Troy, NY 12180, USA 3 Astronomy Department, Boston University, Boston, MA 02215, USA Received 2007 September 21; accepted 2008 February 15; published 2008 April 2

ABSTRACT Class I methanol masers appear to probe very early stages of star formation. An observational survey of the 44 and 36 GHz methanol lines toward several star-forming regions was conducted using the Haystack Observatory 37 m telescope. Examining the intensities of the 36 GHz Class I maser line as compared to the 44 GHz maser line, it is seen that the 36 GHz line is enhanced toward sources where there is no apparent sign of star formation. Sources where the 36 GHz emission is absent, but the 44 GHz emission is strong, appear to be those where ultracompact H ii regions and millimeter continuum sources are present. Existing models for the excitation of Class I methanol masers show strong temperature and density dependences for the presence or lack of certain methanol transitions. The 36 GHz masers appear in regimes where the temperatures are low—below 100 K. The 44 GHz masers are excited in a wider range of gas temperatures (80–200 K), supporting the hypothesis that these transitions are still masing even when the 36 GHz masers are quenched. Key words: ISM: molecules – masers – stars: formation

issue, the Haystack Observatory methanol maser search has been conducted in a manner designed to try to increase the number of known Class I masers, and the environments in which they exist. The search process involved taking known molecular clouds and searching the entire cloud for maser emission at 44 GHz. The 44 GHz line has traditionally been the one to use in searches for Class I methanol masers. It appears to be the most ubiquitous maser line and theoretical studies appear to indicate that this transition has the broadest range of excitation conditions (Voronkov et al. 2005). The molecular clouds in this survey were chosen to have known high-mass star formation. In the search process, once maser emission was detected at 44 GHz, observations of the 25 GHz, 36 GHz and 95 GHz lines were also made in order to put constraints on the excitation conditions. The maser search was conducted by undergraduate students from colleges around the United States as part of the MIT Haystack Observatory undergraduate research program, funded by the National Science Foundation.

1. INTRODUCTION Methanol masers have been classified in two categories— Class I and Class II, based on the transitions that show maser action. Class II masers, which show strong maser action in the 6 and 12 GHz lines of methanol, are often associated with H ii regions or near-infrared (near-IR) sources and are radiatively pumped (Sobolev et al. 2007 and references therein). Class I methanol masers, on the other hand, are also associated with sites of star formation but have often been found offset from regions where other masers, such as H2 O and OH, are found (Kurtz et al. 2004). These masers, which show maser action in the 25 (the J2 → J1 E series), 36 (4−1 → 30 E), 44 (70 → 61 A+ ) and 95 GHz (80 → 71 A+ ) lines, are thought to be collisionally pumped (Cragg et al. 1992). Interferometric studies of Class I masers have shown that these masers are often associated with outflows (Sandell et al. 2003, 2005; Plambeck & Menten 1990). The conclusion from the first of these studies was that the outflows impacting the molecular material enhances the methanol abundances at the interfaces and creates the path lengths necessary for maser action. However, there are several cases where these masers are seen in regions with no obvious outflow activity (Kurtz et al. 2004; this paper). In such cases, a possible hypothesis is that the masers arise from a region of very young star formation—one that has not yet been detected by any other means. In this scenario, an outflow may be involved in the maser pumping scheme, but the outflow is spatially small and associated with the protostar. Such a result would need to be confirmed by not only mapping the maser emission but also the possible outflow emission. Lack of any outflow detection would require a modification of the maser excitation theories. In order to test this hypothesis we need to have a large, unbiased sample of maser sources. In the past, all the searches for the Class I masers were made toward known star-formation sites. Masers that were outside the primary beam of the observing telescope were found serendipitously. To solve this selection 4

2. OBSERVATIONS 2.1. Haystack Observations Observations of methanol masers were made during the period between 2000 November and 2005 January with the Haystack 37 m telescope. The sources observed and the observing parameters are shown in Table 1. Observations of each source were first made in the 44069.488 GHz 70 → 61 A+ transition of methanol. The beamsize of the telescope at this frequency was ∼46 . The searches were done using a sampling interval of 40 in a right ascension declination grid around the nominal center of the known star-forming region. In most cases, the grid was extended to include the entire molecular cloud (mapped in CO). The grid sizes are shown in Table 1. When emission was detected, a half beam sampled (20 spacing), smaller grid was made in most sources. The source positions given in Table 1 correspond to the maser peak positions (with a positional accuracy of ∼5 ), obtained from Nyquist sampled 44 GHz maser maps, that were then used for the 36 GHz observations.

REU student at MIT Haystack Observatory, summer 2004.

1718

No. 5, 2008

CLASS I METHANOL MASERS

1719

Table 1 Observing Parameters Source OMC-2 S235 S255 M8E W75N W75N-Off DR21-C DR21-W DR21(OH) DR21N S140 NGC 7538S NGC 7538-IRS 9 NGC 7538-M NGC 7538-W

R.A. (J2000)

Decl. (J2000)

Observed frequencies (GHz)

Grid size (40 spacing)

Grid size (20 spacing)

05:35:27.5 05:40:53.4 06:12:56.4 18:04:53.3 20:38:37.2 20:38:32.0 20:39:00.0 20:38:55.0 20:39:00.8 20:39:02.0 22:19:18.1 23:13:45.0 23:14:01.8 23:13:45.0 23:13:40.0

−05:09:36.5 35:41:48.8 17:59:53.8 −24:26:42.0 42:37:58.3 42:36:32.5 42:19:38.0 42:19:22.5 42:22:47.8 42:25:43.0 63:18:49.0 61:26:55.0 61:27:20.5 61:27:35.0 61:27:25.0

44, 36, 95 44, 36, 95 44, 36 44, 36 44, 36, 95 44, 36, 95 44, 36 44, 36 44, 36 44, 36 44, 36, 95 44, 36 44, 36 44, 36 44, 36

120 × 520 160 × 160 ... 120 × 120 320 × 160 160 × 160 320 × 560b 320 × 560b 320 × 560b 320 × 560b 160 × 160 320 × 240 320 × 240 320 × 240 320 × 240

80 × 140 120 × 120 110 × 110 ... 120 × 120a 80 × 80a 180 × 140c 180 × 140c 120 × 120 120 × 120 ... . . .d . . .d . . .d . . .d

Notes. a The original 40 spaced grid included the entire W75N cloud. The 20 spaced grid focused around the detected masers. b The original 40 spaced grid included the entire DR21 cloud. c The 20 spaced grid focused around the detected masers, covering both the DR21-C and DR21-W masers. d Only a 40 grid was done to cover the entire NGC 7538 cloud.

The individual source maps and the corresponding spectra are shown in Figures 1–10. In sources where methanol emission was detected, observations of the 36169.240 GHz 4−1 → 30 E and the 24959.079 GHz 52 → 51 E transitions were conducted. The status of these observations is also indicated Table 1. No methanol emission was seen at 25 GHz. The observations at 36 GHz and 44 GHz were made in a beamswitching mode, with a beam throw of 7 , using a 17.8 MHz bandwidth, and a resulting spectral resolution of 0.002 MHz. The mapping was done by integrating for 5 min at each position, resulting in an rms noise per channel that varied between 2.5 and 4 Jy. Maser detections were considered for emission that was at least 3σ of the rms noise. The central position was used as a check of the pointing and gain which resulted in deeper integrations at these positions. The 44 GHz line was observed in two linearly polarized channels. The aperture efficiency at these frequencies was measured using observations of Venus and Jupiter. An average aperture efficiency of 25% is assumed in all the analysis. The average value was calculated from measurements of the aperture efficiency made during each observing run. 2.2. FCRAO Observations Observations of the 95169.516 GHz 80 → 71 A+ transition of methanol were made toward four sources with the FCRAO 14 m antenna in 2005 January with a beamsize of ∼56 . The observations were made of an 8 × 8 region (Nyquist sampled in each source, using the on-the-fly (OTF; Mangum et al. 2000) mapping technique. The spectral resolution was 0.024 MHz and the rms noise was ∼3 Jy channel−1 . 3. RESULTS 3.1. NGC 7538 NGC 7538 is a star-forming region situated in the Perseus arm of the Galaxy. The region consists of a visible H ii region with several stars and an associated molecular cloud containing several centers of star formation. The molecular cloud is situated south of the H ii region with a density gradient and massive

star formation at the interface between the two (Campbell & Thompson 1984). Within the molecular cloud there are several centers of star formation which have apparently formed sequentially (Ojha et al. 2004). The Haystack survey found several centers of methanol maser emission in the extended molecular cloud. Figure 1(a) shows the entire mapped grid with the detected positions marked. The most evolved center of star formation, the IRS 1–3 cluster, did not contain any Class I masers. The southern source, NGC 7538S, which is deeply embedded and a source of H2 O masers, has methanol masers with multiple velocity components (Figure 1(b)). Toward this source, the 36 GHz line is as strong as the 44 GHz line and covers the same velocity extent; the 44/36 flux ratio ranges between ∼0.5 and ∼10 (Table 2). Strong 44 GHz maser emission was found in two spots situated in the molecular cloud between the IRS 1–3 cluster and the NGC 7538S star-forming site (NGC 7538M and NGC 7538W in Figure 1(a)). There is also strong 36 GHz emission at these two spots that cover the same velocity extent as the 44 GHz maser line (44/36 flux ratio ∼0.6 to ∼2). Maser emission was also found toward IRS 9 (Figure 1(a)), which contains several protostellar cores with surrounding reflection nebulosity and outflow activity. The emission toward IRS 9 emission is weak in both lines. 3.2. W75N W75N is a bright star-forming region with a central infrared cluster containing mid- to early-B-type stars. It contains an extended and massive CO outflow and a cluster of ultracompact H ii regions that appears to have multiple outflows with a total flow mass greater than 250 M (Shepherd et al. 2003, hereafter STS03). A 5 × 5 region of the source was mapped in the 44 GHz methanol line (Figure 2(a)). Two sites of 44 GHz methanol maser emission were detected. The maser emission detected toward W75N(A) was previously known (Haschick et al. 1990). A new 44 GHz maser was detected ∼1 southwest of the H ii regions. This maser appears to be situated in the extended CO outflow (Davis et al. 1998).

1720

PRATAP ET AL.

Vol. 135

Table 2 Characteristics of Methanol Maser Emission Source

Velocity (km s−1 )

S(44/36)a

S(44/95)a

OMC-2

11.2 11.4 −16.0 10.6 11.5 11.7 11.2 11.5 8.9 9.8

14.9 15.4 7.3 13.2 31.1 15 2.6 10.3 17.8 1.08

3.2 2.3 6.25 ... ... ... ... ... 21.6 ...

10.5 −3.5 −3.2 −2.1 −1.2

5.7 4.3 13.4 2.5 5.8

... 1.5 ... 2.2 ...

0.4 0.9 −4.8 −3.8 −2.8 −3.5 0 0.8 −7.8 −56.5

11.3 23.3 4.3 1.8 4.7 0.13 0.06 0.1 4.7 0.5

1.6 ... ... ... ... ... ... ... 6.3 ...

−55.2 −54.1 −53. −57.3 −56.9 −53.0 −57.1 −56.9 −53.0

1.8 4.3 9.9 2.1 1.5 1.2 0.6 2.0 2.0

... ... ... ... 1.9 1.7 ... ... ...

S235 S255

M8E W75N W75N-Off

DR21-C DR21-W

DR21(OH) DR21N

S140 NGC 7538S

NGC 7538-IRS 9 NGC 7538-M NGC 7538-W

Comments

No detected IR or radio emission No 36 GHz emission, UC H ii region No 36 GHz emission No 36 GHz emission

No 36 GHz emission, compact mm sources No 95 GHz detection, no observations of compact sources No 36 GHz detection No 36 GHz detection 36 GHz at different velocity: −2.8 km s−1 Masers in extended outflow, no detected star formation Multiple mm continuum sources in region 36 GHz emission at different velocities

Velocity corresponds to 36 GHz emission Velocity corresponds to 36 GHz emission Velocity corresponds to 36 GHz emission 36 GHz emission maybe thermal; millimeter continuum sources and H2 O masers

Outflows detected No detected continuum sources No detected continuum sources

Note. a Flux ratios taken at the same velocities. Cases where peak emission velocities are different are indicated in the comments column.

Toward the H ii region, W75N(A), the 44 GHz maser transition shows evidence of multiple maser features with one strong spike (Figure 2(b)). Interferometric observations of this maser (Kurtz et al. 2004) indicate that this maser is situated on the edge of a millimeter continuum peak (MM5 from STS03). The other masers in this source are distributed on the edges of a more extended millimeter source which contains several ultracompact H ii regions. STS03 also detected several other deeply embedded objects that do not have near-IR counterparts that are distributed to the north of the compact UCHII region cluster. While several high-velocity CO outflows have been detected emanating from this cluster, it is not clear whether these outflows are in any way responsible for the maser excitation. The velocity range of the masers is within the central core of the CO lines. It is possible, that as in the case of NGC 7538, the outflow structure may need to be mapped in a different molecular tracer. However, these results show that there is certainly sufficient star-formation and outflow activity to result in the methanol maser activity. In examining the maser action from the 36 and 95 GHz transitions of methanol (Figure 2(b)), we observe that while there is low level maser action over the same velocity range as the 44 GHz line, the strongest maser, which is at a velocity of

8.9 km s−1 , is about seven times weaker in the 36 GHz spectrum and about 2.5 times weaker in the 95 GHz spectrum (Table 2). The new masers discovered in this source are situated about 1. 7 southwest of W75N(A). These masers would not have been detectable in the interferometer maps (Kurtz et al. 2004) since they are outside the primary beam of the Very Large Array (VLA). The new masers are well within the envelope of H2 knots identified by STS03 and on the southern edge of the H2 bow-shock. The offset position masers also appear to have multiple velocity components although the range of velocities is narrower than that around W75N(A) (Figure 2(b)). The peak velocity of these masers is also offset by about 1 km s−1 from the velocity of the strongest maser in W75N(A). The 36 GHz line toward the offset position is as strong as the 44 GHz maser and has the same velocity extent. However, there was no detection of the 95 GHz maser line toward this position. 3.3. W75S/DR21 W75S (also known as DR21(OH)) and DR21-C are situated in the optically obscured Cygnus X region. The molecular emission in the source is elongated in a north–south direction, within

No. 5, 2008


1721

(a)

Dec. offset (arcsec)

100

0 M

IRS9

W

S

-100

100

200

(b)

0 R.A. offset (arcsec)

NGC 7538W - 44 GHz

-100

27.2

NGC 7538M - 44 GHz

27.4

13.6

Flux (Jy)

13.6

0

0

27.2 36 GHz

36 GHz

13.6

13.6

0

-70

0

-60

-50 Velocity (km/s)

NGC 7538S - 44 GHz

20.4 13.6

-60

-70

-40

-50 Velocity (km/s)

-40

NGC 7538 - IRS9 - 44 GHz

13.6 6.8

6.8 Flux (Jy)

0

0 -6.8 36 GHz

36 GHz 6.8

13.6 6.8

0

0

-70

-6.8

-60

-50 Velocity (km/s)

-40

-70

-65

-55 -60 Velocity (km/s)

-50

-45

Figure 1. (a) Mapped grid of the 44 GHz line toward the NGC 7538 molecular cloud made with a 40 spacing. The peak positions considered in the analysis are indicated. (b) The 44 GHz and 36 GHz spectra toward four positions in NGC 7538. The source names are indicated in the panels and correspond to the pointing positions given in Table 1 and those indicated in (a).

1722

PRATAP ET AL.

Vol. 135

(a)


100

W75N(A)

0

W75N-Off

-100

100

-100


(b) 40.8 W75N(A) 44 GHz

40.8

W75N-Off 44 GHz

27.2

27.2 13.6

13.6

0

0 36 GHz

36 GHz

6.8

13.6

0

0

95 GHz

2.7 0 -2.7 -10

95 GHz Flux (Jy)

Flux (Jy)

5.4

2.7 0 -2.7

0

10 20 Velocity (km/s)

30

-10

0

10 20 Velocity (km/s)

30

Figure 2. (a) Map grid of the 44 GHz line toward the W75N molecular cloud made with a 40 spacing. The positions of the two centers of maser emission given in Table 1 are marked. (b) The 44, 36 and 95 GHz spectra toward W75N and the offset position to the southwest of W75N. The pointing position for this source is given in Table 1 as W75N-Off.

which several dense cores have been detected in ammonia and C18 O (Wilson & Mauersbeger 1990); these multi-wavelength observations led the authors to suggest the possibility of an evolutionary trend in the cloud from south to north. Figure 3(a) shows the peak positions of the masers (Table 2) overlaid on a Spitzer 8µ image (obtained from the Spitzer archive: PID-623, IRAC CampaignW(SV-3), P.I.

G. Fazio and PID-1021, DR21 and its Molecular Outflow, P.I. A. Marston; Marston et al. 2004). Several centers of 44 GHz methanol maser emission have been detected toward this extended complex (Kurtz et al. 2004; Kogan & Slysh 1998). These previously detected masers were re-observed in this survey and are marked in Figure 3(a) as DR21-C, DR21-W and DR21(OH). DR21N is a newly detected maser source.

No. 5, 2008


1723

(a) DR21N

DR21(OH)

DR21-C DR21-W 1'

(b)


50

0 DR21-C

DR21-W

-50

-50 0 R.A. offset (arcsec)

50

(c)

-100

54 44 GHz DR21-C

44 GHz DR21-W

136

27 68

0 0 41 Flux (Jy)

Flux (Jy)

36 GHz 27

27 13.6 0

0

-10

36 GHz

-5

0 Velocity (km/s)

5

10

-13.6 -10

-5

0 Velocity (km/s)

5

10

Figure 3. (a) Spitzer IRAC 8µ image of the entire DR21 molecular cloud region. The peak 44 GHz maser positions, from the survey in this paper, are marked. (b) The 20 grid of the 44 GHz spectra covering the DR21-C and DR21W maser regions. (c) The left panel shows the 44 GHz and 36 GHz spectra toward DR21-C. The right panel shows the methanol spectra toward DR21-W.

PRATAP ET AL.

Vol. 135

50


The southernmost source is DR21-C, which has a large, east–west outflow seen in vibrationally excited H2 (Garden et al. 1986). The methanol masers toward DR21-C are offset several arcseconds south of a 3.6 cm radio continuum source. However, they appear to be coincident with an intense 8µ source (Figure 3(a)). The 44 GHz maser line was detected in this survey but no 36 GHz emission was detected within the flux limits (Figure 3(c)). DR21-W is situated about 1 west of DR21-C (Figure 3(b)). The 44 GHz maser emission is stronger than that in DR21-C. The 36 GHz emission is also present but offset in velocity from the 44 GHz line, possibly indicating that the masers are being generated in different parts of the cloud (Figure 3(c)). Interferometric observations of the 44 GHz maser (Kogan & Slysh 1998) indicate that the masers toward DR21-W consist of two strong components separated by about 3 . No continuum source has been detected at this position, but the masers are situated within an extended outflow detected in 2.2µ H2 emission (Garden et al. 1986). The 95 GHz methanol line has previously been detected in both DR21-C and DR21-W (Plambeck & Menten 1990). DR21(OH), which is situated about 3 north of DR21-C, is a source of OH and H2 O masers and also contains a hot ammonia core (Mauersberger et al. 1986). Toward DR21(OH), one of the strongest Class I masers known, maser lines at 84 GHz (Batrla & Menten 1988), 95 GHz (Plambeck & Menten 1990) and 44 GHz (Kogan & Slysh 1998) have been imaged with interferometers. All these observations indicate that the masers are distributed in an east–west direction and appear in two groups separated by about 30 —this is also apparent in the 44 GHz map made in this survey (Figure 4). The velocities of the two groups are different—the western group has a range between −1.8 and 0.8 km s−1 while the eastern group velocities range between −6.1 and −3.3 km s−1 . The physical distribution and the velocity separation are strong indications for the masers being situated in a bipolar outflow. Although ammonia observations (Mangum et al. 1992) show the presence of multiple dense cores in the region, implying that the masers could arise in individual dense cores. DR21(OH) also contains two compact millimeter continuum sources (Woody et al. 1989). The masers appear to be situated east and west of the compact MM2 source, in a region with emission from OH and H2 O masers and molecular outflows detected in lines of CS, HCO+ and H2 (Richardson et al. 1986; Mangum et al. 1992; Chandler et al. 1993)—all providing strong evidence for the presence of very young stellar objects. Since detailed interferometric measurements have been made of several Class I methanol maser lines in DR21(OH), we concentrate here on comparing the intensities and velocities of the different maser transitions. The 44 GHz and 36 GHz spectra exhibit similar velocity structure (Figure 4). The strongest component is about 14 times stronger in the 44 GHz spectrum as compared to the 36 GHz (Table 2). However, in the weaker velocity components the 44 GHz line is only about twice as strong as the 36 GHz line. It is also apparent that the strongest 44 GHz line does not have a counterpart in the 36 GHz spectrum again indicating that the two maser lines may be coming from different gas in the cloud. A new maser source was discovered about 3 north of the W75S/DR21(OH) source (Figures 3(a) and 5). The maser appears to be associated with a molecular condensation seen in C18 O and C34 S (Wilson & Mauersbeger 1990). The 44 GHz spectrum (Figure 5) shows several velocity features spread between velocities of −6 and −2 km s−1 , which is con-

0

-50


50

-50

44 GHz 272

136

Flux (Jy)

1724

0 27

36 GHz

13.6

0

-10

-5

0 Velocity (km/s)

5

10

Figure 4. The top panel shows the 40 grid map of the 44 GHz line toward DR21(OH). The bottom panel shows the 44 GHz and 36 GHz spectra toward the peak position in the grid.

sistent with the velocities seen in the C18 O maps. The 36 GHz spectrum toward this source shows strong maser emission— however the strongest lines are at velocities different from the 44 GHz lines. In fact, the 44 GHz lines appear at velocities that are within the velocity range spanned by the H2 O masers (Figure 5). 3.4. OMC-2 OMC-2 is an extended, filamentary region containing multiple dense cores and is situated north of the Orion A molecular cloud. A 2 × 8 region of this cloud was mapped in the 44 GHz methanol line (Figure 6). The only source of maser emission was detected toward one of the cores in the source that also contains millimeter and submillimeter continuum sources (Chini et al. 1997; Lis et al. 1998). Observations of CO and HCO+

No. 5, 2008


The 36 GHz line is much weaker than the 44 GHz line (Figure 6)—the 44/36 line ratio is 14.9 (Table 2). The stronger 44 GHz line does not appear to have a 36 GHz counterpart (within the flux limits of these observations). The 95 GHz line is about the same strength as the 44 GHz line but also appears to coincide with the velocities of the weaker component.

50


1725

3.5. S140 S140 consists of an H ii region separated from a dark cloud by a nearly edge-on ionization front (Hayashi et al. 1985). The associated molecular cloud contains a compact, embedded, infrared cluster of nearly 50 protostars. The 44 GHz methanol line is weak in this source (Figure 7) and appears to be situated toward the dense core located about 1 northeast of a bright optical rim. The core contains a deeply embedded cluster of early B stars (Evans et al. 1989) and an energetic outflow emanating from IRS 1 (Hayashi et al. 1987). IRS 1 has an H ii region detected at 6 cm (Beichman et al. 1979). These observations imply that the source is evolved beyond the young protostellar stage. No 36GHz emission was detected and while the 95 GHz spectrum shows the hint of an emission feature, the detection is less than 2σ (Figure 7).

0

-50

50

0

-50

R.A. offset (arcsec) 21 44 GHz 14 7 0

36 GHz

28

14

0

H2O

Flux (Jy)

12

6

0 -20

-10

0

10

20

velocity (km/s)

Figure 5. The top panel shows the 20 grid map of the 44 GHz line toward DR21N. The bottom panel shows the 44 GHz, the 36 GHz, and the H2 O line spectra toward the peak position in the grid.

toward the cores detected an outflow that was oriented in an northeast–southwest direction (Aso et al. 2000). There is also an H2 jet emanating from the source and which is coincident with the CO outflow (Yu et al. 1997). Interferometric observations of the 44 GHz masers showed that the maser spots were also distributed in a line oriented along the outflow (Kogan & Slysh 1998), with the strongest maser spots situated farthest from the continuum source.

3.6. S235 S235A is a small optical nebulosity with an associated compact H ii region. 40 south of S235A is S235B, which is a smaller optical and near-IR nebula without a radio counterpart (Felli et al. 2006). The lack of any radio emission toward S235B appears to make is less evolved than S235A. A water maser was detected in the region between the two sources (Tofani et al. 1995) and does not appear to have any associated radio continuum at the longer wavelengths. Recent high-resolution observations of several molecular tracers of dense gas have found a compact core at the maser position with associated millimeter and submillimeter continuum emission (Felli et al. 2004). Interferometric observations of the 44 GHz maser line show several maser features situated around the water maser peak. This survey detected methanol emission at 44, 36, and 95 GHz (Figure 8). The 44/36 GHz line intensity ratio is about 7 (Table 2). The 36 GHz line has the same line shape as the 44 GHz line—including the appearance of a blue line wing and the line shapes of the maser lines are similar in structure to the HCO+ line (Felli et al. 2004). The strongest methanol maser has a velocity of −16.6 km s−1 and is situated to the south of the H2 O maser—the HCO+ gas in this region has velocities between −20 and −18 km s−1 . The HCO+ observations show the presence of two outflows situated perpendicular to each other. The 44 GHz masers are situated on the southern edge of the 3 mm and 1.2 mm continuum sources, which appears to be the origin of the two outflows, and distributed around the H2 O maser. A Spitzer IRAC image of the region shows that the masers are clustered around a very faint source (L. Chavarria 2005, private communication). All the observations point to the presence of a very young stellar object at this position. 3.7. S255 S255-IR is situated in a region of high mass star formation with molecular outflows, H ii regions, and multiple IR sources. The complex region contains an elongated molecular cloud that is situated at the interface between two large optical H ii regions, S255 and S257 (Miralles et al. 1997). The methanol maser in

1726

PRATAP ET AL.

Vol. 135 44 GHz

136

68

0 13.6


0

36 GHz

6.8 0 -6.8 95 GHz

190 Flux (Jy)

-200

95

0 8 50

0

-50

10

12 Velocity (km/s)

14

R.A. offset (arcsec)

Figure 6. The left panel shows the 44 GHz maser line grid (40 ) mapped toward OMC-2. The right panel shows the 44, 36 and 95 GHz spectra toward the peak position.

S255 has multiple components (Figure 9) and peaked about 50 to the west of the nominal survey pointing position. VLA observations (Kurtz et al. 2004) indicate that the components are spread out over a 10 region. The two strong components are separated by 1 km s−1 in velocity (Figure 9). The methanol masers are coincident with a radio continuum source G192.580.04 (Snell & Bally 1986; Kurtz et al. 2004). The radio continuum source also has two peaks, seen in 2.2µ H2 , associated with it (Miralles et al. 1997). No 36 GHz emission was detected toward this region. 3.8. M8E M8E is an infrared source situated to the east of M8. Simon et al. (1984) detected a weak 6 cm radio source toward this region which is situated about 10 to the west of the IR source. The methanol maser in this source is one of the brightest Class I masers detected so far (Figure 10). The 36 GHz line in this source is about 2.6 times weaker than the 44 GHz line (Table 2) and spans the same velocity range and has similar velocity structure (Figure 10). VLA observations of the Class I masers show the presence of two maser features situated about 15 to the east of the radio source and about 10 east of the IR source (Kogan & Slysh 1998). An OH maser was detected in this region—situated between the IR and radio sources. The methanol maser is situated close to the OH maser position (Va’ltts 1999 and references therein). The two lobes of the CO outflow toward this source are separated along a north-south line (Mitchell et al. 1992). The interpretation provided in the analysis by Val’tts is that the lack of an ultracompact H ii region, the presence of high density tracers such as ammonia and CS, and the age of the CO outflow imply that M8E is a very young object. Recent high resolution mid-IR results have shown that M8E-IR is a cluster of six separate sources clustered around a point source (Billmeier et al. 2003). These results emphasize the conclusions that the source is very young.

4. DISCUSSION 4.1. Evolutionary Stage of Star Formation Class I methanol masers are thought to be associated with the earliest stages of high-mass star formation (Ellingsen 2006). However, a search for correlation between the Class I methanol masers and infrared emission did not detect any statistically significant correlation. The Class I masers do not often appear to be associated with other maser species (H2 O and OH) and are not coincident with Class II masers. We can hypothesize from the results presented in Section 3 that the Class I masers are excited by deeply embedded sources that have yet to display significant near-IR emission. In examining the results presented here we look at the relative strengths of the 44 GHz and 36 GHz masers and the star-forming environment in which the masers are produced. Table 2 presents the line ratios of the 44, 36, and, where available, the 95 GHz masers. A detailed examination of the 44/36 line ratios and the evidence for ongoing star formation at the observed maser sites leads to an intriguing possibility (Table 2). It appears that in sources where the star formation is at a sufficiently evolved state the 44/36 GHz is greater than about 5, i.e. the 44 GHz line intensity is much more enhanced compared to that of the 36 GHz line. In sources with evidence of very young star formation or with no obvious evidence of protostellar activity, the 36 GHz line emission is enhanced, leading to 44/36 GHz line ratios that are less than 5. The evidence for this hypothesis can be examined in terms of specific sources. First, in the case of NGC 7538, which contains several centers of star formation, the most evolved region in the molecular cloud, viz. IRS 1–3, does not show any evidence of Class I methanol maser emission. The maser emission that was detected by Bachiller et al. (1990) is situated about 30 south

No. 5, 2008


1727

50



50

0

0

-50

-50 100

50

0

-50

-100

R.A. offset (arcsec) 50

-50

0


44 GHz 6.8

44 GHz

14

0

7

36 GHz

0

6.8

36 GHz

0

7 -6.8

95 GHz

0

4.8 0 -4.8 -9.5 -15

95 GHz

21

-10

-5

0

Flux (Jy)

Flux (Jy)

9.5

14 7

Velocity (km/s)

Figure 7. The top panel shows the 44 GHz maser line grid toward S140. The 44, 36 and 95 GHz spectra toward the peak position are shown in the bottom panel.

of the IRS 1–3 complex. These masers are in the vicinity of H2 O masers and are in an extended ridge of ammonia emission (Zheng et al. 2001). Submillimeter and millimeter maps of this region show the presence of an extension to the southwest of the IRS 1–3 region (Sandell & Sievers 2004). While the 44 GHz methanol masers are not associated with a distinct peak they are definitely situated in the extended region of continuum emission. Comparing the interferometric positions of the 44 GHz masers (Kurtz et al. 2004) with the highest resolution 450µ image indicates that at least one of the maser sites is situated close to a submillimeter peak, but the other two are not.

0 -25

-20

-15 Velocity (km/s)

-10

-5

Figure 8. Grid map of the 44 GHz line toward S235, with 20 spacing. The 44, 36 and 95 GHz spectra toward the peak position are shown in the bottom panel.

The second maser site in this source is situated toward the NGC 7538S star-forming region. This region is contained within a dense molecular core and has H2 O masers as well as a millimeter/submillimeter continuum source. The methanol masers are situated in an outflow detected in HCO+ (Sandell et al. 2003). The 36 GHz methanol emission toward this region is strong and has either broad, thermal emission with maser components superimposed on it or multiple maser features

1728

PRATAP ET AL.

Vol. 135



50

50

0

0

-50

-100

0

-50 R.A. offset (arcsec)

136

50

44 GHz

0

-50


102 544

44 GHz

68 408

34

272

0 36 GHz

136 0 204

36 GHz

0

-6.8 0

5

10 Velocity (km/s)

15

20

Figure 9. 20 spacing grid toward S255. The bottom panel shows 44 and 36 GHz spectra toward the peak position in S255.

Flux (Jy)

Flux (Jy)

6.8

136

68

0

5

spread out in velocity, and similar to the 44 GHz line structure. A compact outflow has been detected in this source. The third maser source in the NGC 7538 molecular cloud has been detected toward the IRS 9 reflection nebula. Davis et al. (1998) found several knots of H2 emission and a collimated jet that do not appear to be associated with the IRS 9 source, suggesting the presence of other young stellar objects in the region. This source has been resolved into multiple young sources with molecular outflows (Sandell et al. 2005) and the masers are associated with the outflows. Molecular line emission observations toward NGC 7538 (Sandell et al. 2003, 2005) have shown the presence of several outflows in the NGC 7538 molecular cloud. The outflows toward NGC 7538S and IRS 9 have methanol masers associated with them and the masers are situated in the lobes of the outflows. The NGC 7538M masers are within 10 of H2 O masers found by Kameya et al. (1990) and situated within an ammonia clump (Zheng et al. 2001). Examining the 44 GHz spectra of the masers (Figure 1(b)) we see that there is strong indication

10 Velocity (km/s)

15

Figure 10. 40 spacing map grid toward M8E. The bottom panel shows 44 and 36 GHz spectra toward the peak position in M8E.

for outflow activity, given that the two stronger components are separated by about 4 km s−1 . However, the velocities of the methanol maser (−52 to −58 km s−1 ) are redshifted relative to the H2 O masers (−63 km s−1 ). This behavior is also seen toward IRS 9 (Sandell et al. 2005). In examining the 44/36 GHz line ratios, we can hypothesize that the maser source with no corresponding signs of star formation is the youngest site in the cloud. Toward this source (NGC 7538M), the 36 GHz emission is as strong as the 44 GHz maser emission, and while both lines span the same range of velocities, the 36 GHz emission does appear to have a thermal component. A similar situation is seen toward NGC 7538S, where the 36 GHz emission appears to be mostly thermal. However, offset from the (0, 0) position, there seems to be evidence of narrow maser emission in the 36 GHz line—the

No. 5, 2008


velocity range of the line corresponds closely with that of the 44 GHz line. There is currently no way of ascertaining the maser nature of the 36 GHz emission, since there are no interferometers that operate at the 36 GHz frequency. In the case of the DR21 complex, there appears to be a clear correlation between the ratios of the two maser lines and the evolutionary stage of the star formation. The northernmost source, DR21N, has strong 36 GHz maser emission, however the strongest maser line is at a different velocity from the 44 GHz line. The situation is similar to NGC 7538S where the two peaks are at different velocities. In this source no traditional signs of star formation, such as optical/near-IR or radio continuum emission, have been detected although the region has extensive submillimeter continuum emission (Davis et al. 2007). Also, the velocity structure of the methanol lines strongly indicates the presence of an outflow. In fact, the strongest 36 GHz line velocities are similar to that of the H2 O masers. Farther south, DR21(OH), has strong 44 GHz maser emission but weak 36 GHz emission. This region contains a cluster of young stellar objects that have been detected at other wavelengths. South of DR21(OH), the source DR21-C, has no detected 36 GHz emission. The south to north evolutionary structure has been proposed by other authors (Wilson & Mauersbeger 1990). In W75N, the masers toward the main cluster of IR sources show no evidence for 36 GHz line emission. However, to the southwest of the main cluster there is strong emission from both 36 and 44 GHz emission. The line has multiple velocity components that are redshifted with respect to the velocities at the main maser site. These masers are situated within an extended CO outflow. The excitation source for these masers will need to be identified by far-infrared or submillimeter observations which trace deeply embedded young stellar objects (YSOs). The other sources in this study also provide evidence for our main hypothesis. S235 appears to be a very young source, situated between two H ii regions, S235A and S235B. Recent observations have shown that the methanol masers are situated in a dense core with molecular outflows and an H2 O maser (Felli et al. 2004). The other two sources, S255 A and OMC-2, are somewhat more evolved, with the presence of ultracompact H ii regions. In both these cases, there is no (S255) or weak (OMC-2) 36 GHz emission. In the case of M8E, one of the strongest 44 GHz maser sources, the presence of strong 36 GHz emission and Class I maser emission from the higher excitation lines provide additional evidence for the correlation between the young age of the exciting object and the presence of Class I maser emission. 4.2. Methanol Excitation Class I methanol maser transitions arise from collisional excitation. The abundance enhancement that is necessary for maser emission could either be caused by the evaporation of methanol from the ice grains or by chemical reactions occurring in the interaction region between outflows and the ambient molecular gas (Plambeck & Menten 1990; Blake et al. 1987). Collisions then excite the freed molecules to higher rotational states. The resulting cascade of spontaneous decay overpopulates the upper levels of the masing transitions. Several models have been proposed to explain the observed Class I maser activity (Cragg et al. 1992; Kalenskii 1995; Slysh et al. 2002; Voronkov et al. 2005). Most of these models have attempted to explain the observed intensities of the Class I and Class II masers. The models from Voronkov et al. and Slysh et al.

1729

attempted to explain the masing action in the E-type methanol transitions at 25 and 36 GHz as well as those in the millimeter and submillimeter regime, the model from Kalenskii looked at the 44 and 95 GHz maser excitation, while the Cragg et al. model examined the whole range of methanol transitions. Leurini (2007, private communication) has performed statistical equilibrium calculations for both A- and E-type methanol for the ground torsional states using the LVG method with spherical geometry. These calculations have used new rate coefficients for the collisions of methanol with helium, for both symmetry states and for levels up to (J,K) = 9. For the Class I methanol maser transitions, a strong dependence of the line optical depths on the molecular hydrogen density emerges. In particular, for the transitions considered in this paper, maser action in the 36 GHz line is maximized at densities in the range of 104 – 105 cm−3 , while the 44 GHz line shows maximum maser action at densities of 105 –106 cm−3 . The dependence of the line optical depth as a function of kinetic temperature is not as dramatic but qualitatively it is shown that the 36 GHz transition mases more readily at lower temperatures (30–100 K) while the 44 GHz transition has a peak depth at temperatures between 80 and 200 K. With all the uncertainties in the modeling resulting from simplified collision models and simplified radiative transfer models, one can make a qualitative conclusion that the presence of 36 GHz maser emission points to gas at lower temperatures and densities. One can anticipate such conditions in an environment where the effects of the star formation have not fully influenced the surrounding gas—an early evolutionary stage. In contrast, the 44 GHz maser emission is predicted for higher temperatures and densities, indicative of a possibly more evolved star-formation site. 5. CONCLUSIONS Class I methanol masers appear to probe very early stages of star formation. Sources where the 36 GHz emission is enhanced, relative to the more commonly observed 44 GHz line, have no apparent near-IR or centimeter radio continuum emission. Sources in which the 36 GHz maser line is not present but the 44 GHz line is appear to be more evolved—at least to the point of showing the presence of ultracompact H ii regions and millimeter continuum sources. Existing models for the excitation of Class I methanol masers show strong temperature and density dependences for the presence or lack of certain methanol transitions. The 36 GHz masers appear in regimes where the temperatures are low— below 100 K. The 44 GHz masers are excited in a wider range of gas temperatures (80–200 K), supporting the hypothesis that these transitions are still masing even when the 36 GHz masers are quenched. While these models appear to explain the basic presence of the masers, there is still much that needs modeling. For example, the enhancement of methanol in the gas has been thought to come from shocks knocking methanol molecules off the grain surfaces. However, shocks with the necessary energy for this mechanism are associated with higher temperatures. One possibility is that the maser emission occurs after the shocks have passed through and the gas has cooled. While this is a plausible explanation, one would need to reconcile the age of the YSOs that potentially create the outflows that may lead to the shocks with the lack of near-IR or radio continuum signs of star formation. Weak shocks that release the methanol from grain mantles and then dissipate, leaving an enhanced methanol

1730

PRATAP ET AL.

abundance is one plausible explanation (Voronkov et al. 2006). Another valuable piece of the excitation puzzle is to understand the spatial relationship between the 36 GHz (E-type) and the 44 GHz (A+ ) types of Class I masers. While the spectral structure indicates that they may arise from different regions, high-resolution mapping is needed to confirm this conclusion. Undergraduate research at MIT Haystack Observatory is supported by grants from the Division of Astronomy and the Division of Undergraduate Education at the National Science Foundation. Undergraduate students who contributed to this project and their host institutions (at the time of the participation) are as follows: Adrienne Schwartz (University of Minnesota, Morris), Dan Brubeck (College of Wooster), Christian Clerc (University of Dallas), Rudy Montez (University of Texas, Austin), Carrie Perkowski (Mt. Union College), Emily Bowsher, Morgan Carberry, Bena Chang, Tara Donner, Julia Levine (Wellesley College), Kristina Barkume, Andrew Essin (Reed College), Maggie McKeon (Amherst College), Lisa Tackett (Roane State Community College), Karen Bland (James Madison University), and Reimi Hicks (Brown University). We would also like to acknowledge the help of Professor Vladimir Strelnitskii (Maria Mitchell Observatory) for useful comments on the manuscript. REFERENCES Aso, Y., Tatematsu, K., Sekimoto, Y., Nakano, T., Umemoto, T., Koyama, K., & Yamamoto, S. 2000, ApJS, 131, 465 Bachiller, R., Gomez-Gonzalez, J., Barcia, A., & Menten, K. M. 1990, A&A, 240, 116 Batrla, W., & Menten, K. M. 1988, ApJ, 329, 117 Beichman, C. A., Becklin, E. E., & Wynn-Williams, C. G. 1979, ApJ, 232, 47 Billmeier, R. R., Jayawardhana, R., Marengo, M., Mardones, D., & Alves, J. 2003, BAAS, 35, 1362 Blake, G. A., Sutton, E. C., Masson, C. R., & Phillips, T. G. 1987, ApJ, 315, 621 Campbell, B., & Thompson, R. I. 1984, ApJ, 279, 650 Chandler, C. J., Moore, T. J. T., Mountain, C. M., & Yamashita, T. 1993, MNRAS, 261, 694 Chini, R., Reipurth, B., Ward-Thompson, D., Bally, J., Nyman, L.-A., Sievers, A., & Billawala, Y. 1997, ApJ, 474, 135 Cragg, D. M., Johns, K. P., Godfrey, P. D., & Brown, R. D. 1992, MNRAS, 259, 203 Davis, C. J., Kumar, M. S. N., Sandell, G., Froebrich, D., Smith, M. D., & Currie, M. J. 2007, MNRAS, 374, 29 Davis, C. J., Smith, M. D., & Moriarty-Schieven, G. H. 1998, MNRAS, 299, 825 Ellingsen, S. P. 2006, ApJ, 638, 241

Vol. 135

Felli, M., Massi, F., Navarrini, A., Neri, R., Cesaroni, R., & Jenness, T. 2004, A&A, 420, 553 Felli, M., Massi, F., Robberto, M., & Cesaroni, R. 2006, A&A, 453, 911 Evans, N. J., II, Mundy, L. G., Kutner, M. L., & Depoy, D. L. 1989, ApJ, 346, 212 Garden, R., Geballe, T. R., Gatley, I., & Nadeau, D. 1986, MNRAS, 220, 203 Haschick, A., Menten, K. M., & Baan, W. A. 1990, ApJ, 354, 556 Hayashi, M., Hasegawa, T., Omodaka, T., Hayashi, S. S., & Miyawaki, R. 1987, ApJ, 312, 327 Hayashi, M., Suzuki, S., Omodaka, T., & Hasegawa, T. 1985, ApJ, 288, 170 Israel, F. P., & Felli, M. 1978, A&A, 63, 325 Kalenskii, S. V. 1995, Astron. Rep., 39, 465 Kalenskii, S. V., Promyslov, V. G., Slysh, V. I., Bergman, P., & Winnberg, A. 2005, Astron. Rep., 50, 289 Kameya, O., Morita, K., Kawabe, R., & Ishiguro, M. 1990, ApJ, 355, 562 Kogan, L., & Slysh, V. 1998, ApJ, 497, 800 Kurtz, S., Hofner, P., & Alvarez, C. V. 2004, ApJS, 155, 149 Lis, D. C., Serabyn, E., Keene, J., Dowell, C. D., Benford, D. J., Phillips, T. G., Hunter, T. R., & Wang, N. 1998, ApJ, 509, 299 Mangum, J., Emerson, D., & Greisen, E. 2000, ASP Conf. Proc. 217, Imaging at Radio through Submillimeter Wavelengths, ed. J. G. Mangum, & S. J. E. Radford, 179 Mangum, J. G. , Wootten, A., & Mundy, L. G. 1992, ApJ, 388, 467 Marston, A. P., et al. 2004, ApJS, 154, 333 Mauersberger, R., Henkel, C., Wilson, T. L., & Walmsley, C. M. 1986, A&A, 162, 199 Miralles, M. P., Salas, L, Cruz-Gonzalez, I., & Kurtz, S. 1997, ApJ, 488, 749 Mitchell, G. F., Hasegawa, T. I., & Schella, J. 1992, ApJ, 386 Ojha, D. K., et al. 2004, ApJ, 616, 1042 Plambeck, R. L., & Menten, K. M. 1990, ApJ, 364, 555 Richardson, K. J., White, G. J., Phillips, J. P., & Avery, L. W. 1986, MNRAS, 219, 223 Sandell, G., Goss, W. M., & Wright, M. 2005, ApJ, 621, 839 Sandell, G., & Sievers, A 2004, ApJ, 600, 269 Sandell, G., Wright, M., & Forster, J. R. 2003, ApJ, 590, 45 Shepherd, D. S., Testi, L., & Stark, D. P. 2003, ApJ, 584, 882 Simon, M., Cassar, L., Felli, M., Fischer, J., Massi, M., & Sanders, D. 1984, ApJ, 278, 170 Slysh, V. I., Kalenski, S. V., & Val’Tts, I. E. 2002, Astron. Rep., 46, 49 Snell, R. L., & Bally, J. 1986, ApJ, 303, 683 Sobolev, A. M., et al. 2007, Astrophysical Masers and their Environments, Proc. IAU, IAU Symposium, 242, 81 Tofani, G., Felli, M., Taylor, G. B., & Hunter, T. R. 1995, A&AS, 112, 299 Val’tts, I. E. 1999, Astron. Rep., 43, 149 Voronkov, M. A., Brooks, K. J., Sobolev, A. M., Ellingsen, S. P., Ostrovskii, A. B., & Caswell, J. L. 2006, MNRAS, 373, 411 Voronkov, M., Sobolev, A., Ellingsen, S., Ostrovskii, A., & Alakoz, A. 2005, Ap&SS, 295, 217 Wilson, T. L., & Mauersbeger, R. 1990, A&A, 239, 305 Woody, D. P., Scott, S. L., Scoville, N. Z., Mundy, L. G., Sargent, A. I., Padin, S., Tinney, C. G., & Wilson, C. D. 1989, ApJ, 337, 41 Yu, Ka, Chun Bally, J., & Devine, D. 1997, ApJ, 485, 45 Zheng, X.-W., Zhang, Q., Ho, P. T. P., & Pratap, P. 2001, ApJ, 550, 301