Spectral imaging of the Sagittarius B2 region in multiple 3-mm molecular lines with the Mopra telescope
P. A. Jones, M. G. Burton, M. R. Cunningham, K. M. Menten, P. Schilke, A. Belloche, S. Leurini, J. Ott, A. J. Walsh
aa r X i v : . [ a s t r o - ph ] J a n Mon. Not. R. Astron. Soc. , 1–22 (2007) Printed 1 November 2018 (MN L A TEX style file v2.2)
Spectral imaging of the Sagittarius B2 region in multiple3-mm molecular lines with the Mopra telescope
P. A. Jones ⋆ M. G. Burton , M. R. Cunningham , K. M. Menten , P. Schilke ,A. Belloche , S. Leurini , J. Ott , A. J. Walsh School of Physics, University of New South Wales, NSW 2052, Australia Max-Planck-Institut f¨ur Radioastronomie, Auf dem H¨ugel 69, 53121 Bonn, Germany European Southern Observatory, Karl-Schwarzschild-Str. 2, 85748 Garching, Germany National Radio Astronomy Observatory, 520 Edgemont Road, Charlottesville, VA 22903, USA School of Maths, Physics and IT, James Cook University, Qld 4814, Australia
Accepted . Received ; in original form 2007 XXX XX
ABSTRACT
Using the Mopra telescope, we have undertaken a 3-mm spectral-line imagingsurvey of a 5 arcmin square area around Sgr B2. We covered almost the completespectral the range from 81.7 to 113.5 GHz, with 2.2 MHz wide spectral channels or ∼ − and have observed 24 lines, with 0.033 MHz wide, or ∼ . − channels.We discuss the distribution of around 50 lines, and present velocity-integrated emissionimages for 38 of the lines. In addition, we have detected around 120 more lines, mostlyconcentrated at the particularly spectral line-rich Sgr B2(N) source.There are significant differences in molecular emission, pointing to both abundanceand excitation differences throughout the region. Seven distinct spatial locations areidentified for the emitting species, including peaks near the prominent star formingcores of Sgr B2(N), (M) and (S) that are seen in IR-to-radio continuum images. Theother features are a ’North Ridge’ and a ’North Cloud’ to the north of the Sgr B2N-M-S cores, a ’South-East Peak’ and a ’West Ridge’.The column density, as evident through C O, peaks at the Sgr B2(N) and (M)cores, where strong absorption is also evident in otherwise generally-bright lines suchas HCO + , HCN and HNC. Most molecules trace a ridge line to the west of the Sgr B2N-M-S cores, wrapping around the cores and extending NE to the North Cloud. This ismost clearly evident in the species HC N, CH CN, CH OH and OCS. They are foundto be closer in distribution to the cooler dust traced by the sub-mm continuum thaneither the warmer dust seen in the mid-IR or to the radio continuum. The moleculeCN, in contrast, is reasonably uniform over the entire region mapped, aside fromstrong absorption at the positions of the Sgr B2(N) and (M) cores.
Key words:
ISM:individual (Sagittarius B2) – ISM:molecules – radio lines:ISM –ISM:kinematics and dynamics.
Sagittarius B2 (Sgr B2) (G0.7-0.0) is a very massiveand well-studied molecular cloud complex near the centreof the Galaxy. It contains multiple centres of (in manycases) spectacular star formation activity. The name de-rives from low resolution radio observations where Sagit-tarius A is the strong source at the Galactic Centre proper(Piddington & Minnett 1951) and B1, B2, C, D and E refer ⋆ E-mail:[email protected] (PAJ) to other radio and mid-IR features nearby (Lequeux 1962;Hoffmann, Frederick & Emery 1971), albeit with some con-fusion in the literature (Palmer & Goss 1996).Sgr B2 is about 100 pc in projected distance from theGalactic Centre and we assume its distance from the Sunto be identical to the latter’s, R o . An R o of 7 . ± . . ± . R o = 7 . ± . c (cid:13) P. A. Jones et al. recently been determined from orbital solutions of a starmoving around the super-massive central black hole, Sgr A ∗ (Eisenhauer et al. 2003). In the following we assume R o = 8kpc.Sgr B2 presents itself as the strongest feature in im-ages of emission in CO, CO (Oka et al. 1998) and CS(Tsuboi, Handa & Ukita 1999) that define the bar-shaped(Sawada et al. 2004) Central Molecular Zone (CMZ), whichstretches over the central few hundred pc of the Galaxy. Thetotal mass of Sgr B2 is > × M ⊙ and its peak H columndensity > cm − (Lis & Goldsmith 1990).Recent star formation is indicated by a giantH II region (Mehringer et al. 1993), with many com-pact and ultra-compact H II regions (Gaume et al.1995). There are multiple centres of maser emis-sion from the water (McGrath, Goss & De Pree 2004),hydroxyl (Gaume & Claussen 1990) and formaldehyde(Mehringer, Goss & Palmer 1994) molecules, as well asclass I and class II methanol masers (Caswell 1996;Mehringer & Menten 1997). The region’s huge far-IR lumi-nosity requires several young O-type stars as power sources,which are deeply embedded in the molecular cores.The star-forming centres are located in a north-southline about 2 arcmin ( ∼ II free-free, millimetre andsub-millimetre (Gordon et al. 1993; Pierce-Price et al. 2000)and infrared (Goldsmith et al. 1992) emission. These coreshave been extensively studied with millimetre spectral-linesurveys (Cummins, Linke & Thaddeus 1986; Turner 1989;Nummelin et al. 1998, 2000; Belloche et al. 2005, 2007).Sgr B2(N) is particularly rich in complex molecules:it has been called the ‘Large Molecule Heimat’(Snyder, Kuan & Miao 1994; Miao et al. 1995) or LMH.Sgr B2(N) is considered to be in a more recent stage of starformation than Sgr B2(M) (Miao et al. 1995), due to thepresence of the complex molecules, stronger H O masers,and the relatively large amount of dust.The surrounding molecular cloud has complex kine-matics. The densest core emits around 60–65 km s − , butthere is a ‘hole’ in the CO and CS emission around 40–50km s − in this area (Sato et al. 2000). This has been at-tributed (Hasegawa et al. 1994) to a collision between the40–50 km s − cloud and a cloud at 70–80 km s − , triggeringthe star formation activity. There is also a cloud 2 arcminnorth of Sgr B2(M), and 1 arcmin north of Sgr B2(N), withchemical enhancement in HNCO and HOCO + (Minh et al.1998), which may be associated with the shock from thiscollision.We present here a multi-line spectral study in the 3-mm band, of the central 12 pc of the Sgr B2 complex, toprobe the chemistry and kinematics with a wide range ofmolecular tracers. The data were obtained with a new 8 GHzwide spectrometer on the Mopra millimetre wave telescopein Australia. These are the initial results of a project to mapthe CMZ in a variety of molecular species emitting in the3-mm band. The observations were made with the 22-m Mopra radiotelescope, in on-the-fly mapping mode (Ladd et al. 2005).During 2005 a new wide-bandwidth digital filterbank,MOPS, was installed. This takes advantage of the widebandwidth of the MMIC receiver, also installed in 2005,which covers the range from 77 to 117 GHz and has a widefront-end bandwidth. The MOPS can cover 8 GHz of band-width simultaneously, in either a broad band mode coveringthe whole band in four 2.2 GHz wide spectra, or a zoommode where several narrower spectral bands of 137 MHzcan be selected within the overall 8 GHz. In both the broadband and zoom modes, two polarisations are detected.The observations made in 2006 June had 1024 channelsin each 2.2 GHz in the broad band mode giving channelwidth 2.15 MHz or around 6.4 km s − (at 100 GHz). Thisis coarser velocity sampling than desirable, but does allowthe whole 8 GHz spectrum to be covered for a single tuning,and most of the 3-mm band in 4 tunings. The narrow bandmode allowed a maximum of 8 zoom spectra of 137 MHzwith 4096 channels (0.033 MHz or 0.10 km s − at 100 GHz)to be observed (with a maximum of 4 zooms in each 2.2 GHzsection) or 8 lines to be selected at high spectral resolutionwithin the 8 GHz band covered by a single tuning.This period was while the MOPS was still being up-graded, and the performance has since improved to allow upto 8192 channels of 0.27 MHz for each 2.2 GHz window in thebroad band mode and up to 16 zoom spectra simultaneously.Further observations in this Mopra CMZ mapping projectin 2007 and onwards use this increased performance. The on-the-fly (OTF) observations covered an area 5 × centred on ( α, δ ) J2000 = 17 h m s , − ◦ ′ ′′ ,i.e., close to Sagittarius B2(N). We observed this area inboth the broad band and zoom modes, in several tunings,as summarised in Table 1. The broad band ranges are cal-culated assuming an overall range of 8 GHz: the data covera bit more spectral range with 2.2 GHz sub-band spectraseparated by 2.0 GHz, but are poor at the sub-band edges.The OTF observations were made in a similar modeas for the Mopra G333-0.5/RCW106 survey (Bains et al.2006). We used position switching for bandpass calibra-tion with an off-source reference position (( α, δ ) J2000 =17 h m s , − ◦ ′ ′′ , or l = 1 .
093 deg., b = − . livedata and gridzilla packages . livedata is theprocessing software originally designed for the Parkes HI multibeam survey and is used to apply system tempera- In 2007 we have observed the frequency range 85.3 to 93.3 GHzin broad band mode, over the area longitude -0.2 to 0.9 deg., andlatitude -0.20 to 0.12 deg., and will discuss these observations ina later paper. (cid:13) , 1–22 -mm spectral imaging of the Sagittarius B2 region Table 1.
Log of Mopra observations. The rms noise of the zoom bands is given for the 9-point Hanning smoothed data with 0.13 MHzchannels, whereas for the broad band data it is 2.15 MHz per channel.Date Time Mode Central Broad band range and sub-band centres rms T MB ture calibration, bandpass calibration, heliocentric correc-tion, spectral smoothing, and to write out the data in sdfits(Garwood 2000) format. gridzilla is a re-gridding softwarepackage that is used to form three dimensional (RA-Dec-velocity) data cubes from bandpass-calibrated sdfits files(usually from livedata ). The raw data files in rpfits for-mat, were corrected with livedata for bandpass by the off-source spectra, a robust second order polynomial fit to thebaseline subtracted and output as sdfits spectra. These werethen regridded into data cubes using gridzilla , with a gaus-sian smoothing function for the interpolation.The resolution of the Mopra beam varies between 36arcsec at 86 GHz and 33 arcsec at 115 GHz (Ladd et al.2005), so the resolution in the final data varies between 39and 36 arcsec after convolution with the 15 arcsec FWHMgaussian in the gridzilla interpolation. The main beam ef-ficiency of Mopra varies between 0.49 at 86 GHz, 0.44 at100 GHz and 0.42 at 115 GHz (Ladd et al. 2005). Theseparameters were measured, however by Ladd et al. (2005)with a previous receiver and correlator. Since we are moreconcerned in this paper with the spatial and velocity struc-ture, we have left the intensities throughout in this paperin the T ∗ A scale, without correction for the beam efficiencyonto the T MB scale (except for the rms noise in Table 1).The zoom mode data, with high resolution in velocity,were output as cubes over the velocity range −
30 to 170km s − , to reduce the file size, using the appropriate restfrequency of the line targeted. The broad band mode cubeswere made with frequency as the third axis, over the whole1024 channels of each sub-band. The pixel size was 12 arcsec.The FITS cubes were then read into the MIRIAD packagefor further analysis.The integrated spectra over the region were plotted forthe broad band mode data cubes, to identify the lines de-tected. Because of ripples in the spectra, particularly at thebandpass edges (mostly correlator ringing or 30 MHz rip-ples due to standing waves between the main dish and thesecondary), we did not always reach the expected thermalnoise sensitivity of around T MB = 0 . nels ( ∼
650 km s − ) from the frequency cubes, and rela-belled the scale as velocity, by putting the appropriate restfrequency into the file headers. These made low velocity res-olution data cubes for the broad band mode data.Since the lines are broad compared to the 0.033 MHzfrequency channels of the zoom mode data, we also madesmoothed versions of the zoom mode cubes, with a 9-pointHanning function, to improve the signal-to-noise of the spec-tra, and using every fourth channel to reduce the file size,making 0.13 MHz channels.For both the zoom mode and broad band mode datacubes, we then made integrated emission images, by sum-ming the data over velocity, using velocity range over whichthe emission was well above the noise level. This velocityrange differed depending on whether the particular line hadstrong line wings. These images are plotted and discussed inSection 3.In addition, we searched the broad band data cubes,visually plane-by-plane, to identify line emission which wasweak or not very extended, and so was not obvious in thespectrum integrated over the whole area. There are around120 of these additional lines, which are listed in Table 3.Most of these are known lines, in the NIST on-line database(Lovas 2002). The line around 107.63 GHz is attributed tomultiple blended transitions of CH CH CN v = 1 (JohnPearson, private communication).The additional lines (Table 3) are discussed below, inSection 3, but as the line emission is weak and noisy, andmostly confined to a small area, the images are not plottedhere.The mapped area of 5 arcmin corresponds to 12 pc, andthe resolution is 1.4 to 1.5 pc (using the Galactic Centredistance R o = 8 . We present here the integrated emission images, analysisof these images and the data cubes. An area of 5 . × . is plotted for each image, generally using the samerifgt acsension and declination scale for the axes, to alloweasy comparison. Images from the broad-band 109-GHz tun-ing cover a region with slight offset in right ascension to theother images, as these data had a small ( ∼
24 arcsec) sys-tematic shift in position, which has been corrected. (Theorigin of this offset is not clear, but is probably due to a c (cid:13) , 1–22 P. A. Jones et al.
Table 2.
Summary of strong lines detected from the broad bandmode observations. The flag Z in the last column indicates linesfor which there is zoom mode data with higher velocity resolution.For most of these lines we show integrated images in Figs. 1, 3 to7 and 9.Rough line ID ExactFreq. molecule transition Rest Freq.GHz GHz81.88 HC N 9 – 8 81.88146284.52 CH OH 5(-1,5) – 4(0,4) E 84.52120685.14 OCS 7 – 6 85.139104 Z85.27 CH CH OH 6(0,6) – 5(1,5) 85.26550785.34 c-C H CCH 5(3) – 4(3) 85.4426005(2) – 4(2) 85.4507655(1) – 4(1) 85.4556655(0) – 4(0) 85.45729985.53 HOCO + CN 1 – 0 F=1-1 86.3387351 – 0 F=2-1 86.3401671 – 0 F=0-1 86.34225686.75 H CO + C 1 – 0 F=0-1 87.0907351 – 0 F=2-1 87.0908591 – 0 F=1-1 87.09094287.32 C H 1 – 0 3/2-1/2 F=2-1 87.316925 Z1 – 0 3/2-1/2 F=1-0 87.32862487.40 C H 1 – 0 1/2-1/2 F=1-1 87.402004 Z1 – 0 1/2-1/2 F=0-1 87.40716587.93 HNCO 4(0,4) – 3(0,3) 87.925238 Z88.63 HCN 1 – 0 F=1-1 88.6304157 Z1 – 0 F=2-1 88.63184731 – 0 F=0-1 88.633936089.19 HCO + N 10 – 9 90.978989 Z91.99 CH CN 5(3) – 4(3) F=6-5 91.971310 Z5(3) – 4(3) F=4-3 91.9714655(2) – 4(2) F=6-5 91.9800895(1) – 4(1) 91.9853165(0) – 4(0) 91.98708992.49 CS 2 – 1 92.49430393.17 N H + =1-1 F=0-1 93.171621 Z1 – 0 F =1-1 F=2-2 93.1719171 – 0 F =1-1 F=1-0 93.1720531 – 0 F =2-1 F=2-1 93.1734801 – 0 F =2-1 F=3-2 93.1737771 – 0 F =2-1 F=1-1 93.1739671 – 0 F =0-1 F=1-2 93.176265 poor pointing correction made just before these data werecollected). Integrated emission images use the mean of broadband and zoom mode, if the zoom mode data were available(or we use the better image if one of the broad or zoom datahad problems). The OTF scanning direction was in right as-cension, and some of the images show stripe artifacts in thisdirection.We plot positions of radio sources with crosses, to makethe alignment of different features more obvious. The ra-dio positions are taken from the 9.1 GHz continuum peaks Table 2 continued.Rough line ID ExactFreq. molecule transition Rest Freq.GHz GHz94.41 CH OH 2(-1,2) – 1(-1,1) E 94.405223 Z2(0,2) – 1(0,1) A+ 94.4071292(0,2) – 1(0,1) E 94.4108952(1,1) – 1(1,0) E 94.42043995.17 CH OH 8(0,8) – 7(1,7) A+ 95.16951695.91 CH OH 2(1,2) – 1(1,1) A+ 95.91431096.41 C S 2 – 1 96.41296196.74 CH OH 2(-1,2) – 1(-1,1) E 96.739393 Z2(0,2) – 1(0,1) A+ 96.7413772(0,2) – 1(0,1) E 96.7445492(1,1) – 1(1,0) E 96.75550797.30 OCS 8 – 7 97.301209 Z97.58 CH OH 2(1,1) – 1(1,0) A- 97.58280897.98 CS 2 – 1 97.980953 Z99.30 SO 3(2) – 2(1) 99.299905 Z100.08 HC N 11 – 10 100.076385 Z100.63 NH CN 5(1,4) – 4(1,3) 100.62950 Z101.48 H CS 3(1,3) – 2(1,2) 101.477764102.07 NH CHO 5(1,5) – 4(1,4) 102.064263 ZH COH + CCH 6(3) – 5(3) 102.530346 Z6(2) – 5(2) 102.5401436(1) – 5(1) 102.5460236(0) – 5(0) 102.547983103.04 H CS 3(0,3) – 2(0,2) 103.040416 Z104.03 SO CS 3(1,2) – 2(1,1) 104.616988 Z105.79 CH NH 4(0,4) – 3(1,3) 105.794057106.91 HOCO + OH 0(0,0) – 1(-1,1) E 108.893929109.17 HC N 12 – 11 109.173638109.25 SO 2(3) – 1(2) 109.252212109.46 OCS 9 – 8 109.463063109.78 C O 1 – 0 109.782173109.91 HNCO 5(0,5) – 4(0,4) 109.905753110.20 CO 1 – 0 110.201353110.38 CH CN 6(3) – 5(3) F=7-6 110.3644696(3) – 5(3) F=5-4 110.3645246(2) – 5(2) F=7-6 110.3750526(1) – 5(1) F=7-6 110.3814046(0) – 5(0) F=7-6 110.383522112.36 C O 1 – 0 112.358988113.17 CN 1–0 1/2-1/2 F=1/2-3/2 113.1441921–0 1/2-1/2 F=3/2-1/2 113.1705281–0 1/2-1/2 F=3/2-3/2 113.191317113.49 CN 1–0 3/2-1/2 F=3/2-1/2 113.4881401–0 3/2-1/2 F=5/2-3/2 113.4909821–0 3/2-1/2 F=1/2-1/2 113.4996391–0 3/2-1/2 F=3/2-3/2 113.508944 of Hunt et al. (1999) obtained with the Australia TelescopeCompact Array (ATCA), supplemented by a few positions ofpeaks from 20-cm Very Large Array (VLA) data for sourcesoutside the area of Hunt et al. (1999). Note in particularthat the peak near the centre is Sgr B2(N) at J2000 17 4720.4, -28 22 12, with Sgr B2(M) at 17 47 20.5, -28 23 05and Sgr B2(S) at 17 47 20.5, -28 23 44 in a line almost ex-actly to the south (labelled in Fig. 1). We also plot withopen squares, some mid-infrared sources with positions fit-ted from the 21 µ m (band E) Midcourse Space Experiment(MSX) data (Price et al. 2001). Note that the four mid-IR c (cid:13) , 1–22 -mm spectral imaging of the Sagittarius B2 region Table 3.
Summary of weaker lines detected from the broad bandmode observations. We do not show the integrated images forthese lines here. The flags in the last column indicate the spatialdistribution of the line: N = peak at Sgr B2(N); M = peak atSgr B2(M); B = peaks at both Sgr B2(N) and Sgr B2(M); E =extended. Lines marked as ‘unidentified’ in column 2, have beennoted in previous surveys and included in the NIST database(Lovas 2002) with the rest frequency in column 4. Lines markedas ‘U’ in column 2 are not in the NIST database. These rest fre-quencies are quoted to the nearest MHz assuming radial velocityaround 61 km s − appropriate for Sgr B2(N) and Sgr B2(M),Section 4.Rough line ID ExactFreq. molecule transition Rest Freq.GHz GHz82.46 CH OCH CH CN 9(1,8)–8(1,7) 82.458611CH OCH OCH CHO 4(2,2)–3(2,1) 85.093268 N85.69 U 85.686 B87.85 NH CHO 4(1,3)–3(1,2) 87.848871 E88.17 H CCCN 10–9 88.166808 N88.24 HNCO 4(1,3)–3(1,2) 88.239027 N89.32 CH OCHO 8(1,8)–7(1,7) E 89.314589 NCH OCHO 8(1,8)–7(1,7) A 89.31666889.57 CH CH CN 10(6)–9(6) 89.562318 NCH CH CN 10(7)–9(7) 89.565034CH CH CN 10(5)–9(5) 89.568100CH CH CN 10(8)–9(8) 89.57305789.59 CH CH CN 10(4,7)–9(4,6) 89.590033 NCH CH CN 10(4,6)–9(4,5) 89.59101790.45 CH CH CN 10(2,8)–9(2,7) 90.453354 N90.60 HC CCN 10–9 90.593059 EHCC CN 10–9 90.60179191.20 HC N 10–9 v =1 l =1 f 91.199796 NHC N 10–9 v =l l =1 e 91.20260791.33 HC N 10–9 v =1 l =1 f 91.333308 N91.55 CH CH CN 10(1,9)–9(1,8) 91.549117 NSO CN 5(0)–4(0) v =1 l =1 92.261440 NCH CN 5(2)–4(2) v =1 l =1 92.26399292.43 CH CHCN 10(1,10)–9(1,9) 92.426260 N93.60 CH CHO 5(-1,5)–4(-1,4) E 93.595238 E93.87 CCS 8(7)–7(6) 93.870098 ENH CHO 3(2,2)–4(1,3) 93.87170094.28 CH CHCN 10(0,10)–9(0,9) 94.276640 N94.54 CH OH 8(3,5)–9(2,7) E 94.541806 N94.76 U 94.759 N94.91 CH CHCN 10(4,7)–9(4,6) 94.913139 NCH CHCN 10(4,6)–9(4,5) 94.91325094.92 U 94.924 N94.94 U 94.940 N95.15 unidentified 95.145 E95.33 CH CHCN 10(2,8)–9(2,7) 95.325490 N95.44 CH CH CN 11(1,11)–10(1,10) 95.442479 Nt-CH CH OH 16(2,14)–16(1,13) 95.44406795.95 CH CHO 5(0,5)–4(0,4) E 95.947439 E95.96 CH CHO 5(0,5)–4(0,4) A++ 95.963465 E
Table 3 continued.Rough line ID ExactFreq. molecule transition Rest Freq.GHz GHz96.49 CH OH 2(1,2)–1(1,1) E 96.492164 Nv t =1CH OH 2(0,2)–1(0,1) E 96.493553v t =196.98 O CS 8–7 96.988123 E97.70 SO SO 3(2)–2(1) 97.715401 M98.18 CH CH CN 11(2,10)–10(2,9) 98.177578 NCH OCHO 8(7,1)–7(7,0) E 98.18219998.90 CH CHO 5(1,4)–4(1,3) A– 98.900951 E99.02 U 99.021 M99.65 HC CCN 11–10 99.651863 NHCC CN 11–10 99.66147199.68 CH CH CN 11(2,9)–10(2,8) 99.681511 N100.03 SO 4(5)–4(4) 100.029565 B100.32 HC N 11–10 v =1 l =1 e 100.322349 N100.41 U 100.406 M100.46 CH OCH OCH OCH CH CN 11(1,10)–10(1,9) 100.614291 N100.71 HC N 11–10 v =2 l =0 100.708837 NHC N 11–10 v =2 l =2 e 100.710972HC N 11–10 v =2 l =2 f 100.714306100.88 SO CO 5(2,4)–4(2,3) 101.024438 NCH SH 4(-1)–3(-1) E 101.029750101.14 CH SH 4(0)–3(0) A 101.139160 ECH SH 4(0)–3(0) E 101.139650101.33 H CO 6(1,5)–6(1,6) 101.332987 N101.98 CH CO 5(1,4)–4(1,3) 101.981426 E103.57 CH CHCN 11(0,11)–10(0,10) 103.575401 N104.05 CH CH CN 12(1,12)–11(1,11) 104.051278 N104.21 CH CHCN 11(2,10)–10(2,9) 104.212655 N104.24 SO OH 11(-1,11)–10(-2,9) 104.300396 NE104.35 CH OH 10(4,7)–11(3,8) 104.354861 NA-104.41 CH CHCN 11(5,*)–10(5,*) 104.408903 NCH OH 10(4,6)–11(3,9) 104.410489A+CH CHCN 11(4,8)–10(4,7) 104.411262CH CHCN 11(4,7)–10(4,6) 104.411485104.49 t-CH CH OH 7(0,7)–6(1,6) 104.487254 E104.80 t-CH CH OH 5(1,5)–4(0,4) 104.808618 E104.96 CH CHCN 11(2,9)–10(2,8) 104.960550 N105.06 CH OH 13(1,13)–12(2,10) 105.063761 NA+105.30 U 105.299 M105.46 NH CHO 5(0,5)–4(0,4) 105.464216 ECH CH CN 12(0,12)–11(0,11) 105.469300105.54 U 105.537 N105.57 CH OH 14(-2,13)–14(1,13) 105.576385 NE105.77 CH OCH OCH OCH (cid:13) , 1–22 P. A. Jones et al.
Table 3 continued.Rough line ID ExactFreq. molecule transition Rest Freq.GHz GHz105.97 NH CHO 5(2,4)–4(2,3) 105.972593 N106.11 U 106.107 N106.13 NH CHO 5(3,3)–4(3,2) 106.134418 B106.35 CCS 9(8)–8(7) 106.347740 E106.54 NH CHO 5(2,3)–4(2,2) 106.541674 N106.64 CH CHCN 11(1,10)–10(1,9) 106.641394 N106.74 SO 2(3)–1(2) 106.743374 M107.01 CH OH 3(1,3)–4(0,4) A+ 107.013770 B107.04 U 107.042 N107.06 SO OH 15(-2,14)–15(1,14) 107.159915 NE107.19 CH CN 6(1)–5(1) 107.194547 N CH CN 6(0)–5(0) 107.196564107.48 CH CH CN 17(2,16)–17(1,17) 107.481465 NCH CH CN 12(7,*)–11(7,*) 107.485181CH CH CN 12(6,*)–11(6,*) 107.486962CH CH CN 12(8,*)–11(8,*) 107.491579107.50 CH CH CN 12(5,8)–11(5,7) 107.502426 NCH CH CN 12(5,7)–11(5,6) 107.502473107.54 CH CH CN 12(11,*)–11(11,*) 107.539857 NCH OCHO 9(2,8)–8(2,7) A 107.543746CH CH CN 12(4,9)–11(4,8) 107.543924CH CH CN 12(4,8)–11(4,7) 107.547599107.59 CH CH CN 12(3,10)–11(3,9) 107.594046 N107.63 CH CH CN v = 1, multiple 107.636 N107.73 CH CH CN 12(3,9)–11(3,8) 107.734738 N107.84 SO CN 1/2–1/2 F=2-1, 108.651297 EF =0, F =1-0 CN 1/2–1/2 F=2-2, 108.657646F =0, F =1-1 CN 1/2–1/2 F=1-2, 108.658948F =1, F =1-1108.71 HC CCN 12–11 108.710523 NHCC CN 12–11 108.721008108.78 CN 3/2–1/2 F=3-2, 108.780201 EF =1,F =2-1 CN 3/2–1/2 F=2-1 108.782374F =1,F =2-1 CN 3/2–1/2 F=1-0 108.786982F =1,F =2-1 peaks all correspond to radio sources, including Sgr B2(M)and Sgr B2(S) but that Sgr B2(N) does not have strongemission at 21 µ m. See Section 4 for plots of the radio andmid IR continuum, and discussion of the alignment of thedifferent molecular lines with the radio and mid-IR contin-uum features.In the figure captions we give peak integrated brightnessand contour level steps, in K km s − , on the T ∗ A scale, that isnot corrected for beam efficiency. The contours are in equallinear steps. In most cases the lowest contour level is thesame as the step size, but this is not the case for some ofthe strongest lines (such as CO) where the whole 5 arcminsquare area is filled with emission well above the zero level.In this section we present maps for many of thelines measured. We summarise these line maps in Table 2,whereas in Table 3 we list all the other (weaker) lines de-
Table 3 continued.Rough line ID ExactFreq. molecule transition Rest Freq.GHz GHz108.94 CH CH CN 12(2,10)–11(2,9) 108.940596 N109.14 CH OH 26(0,26)–26(-1,26) 109.137570 NE109.15 CH OH 16(-2,15)–16(1,15) 109.153210 NE109.44 HC N 12–11 v =1 l =1 f 109.438572 NHC N 12–11 v =1 l =1 e 109.441944109.49 HNCO 5(1,5)–4(1,4) 109.496007 E109.60 HC N 12–11 v =1 l =1 f 109.598751 B109.65 CH CH CN 12(1,11)–11(1,10) 109.650301 N109.75 NH CHO 5(1,4)–4(1,3) 109.753499 ESO N 12–11 v =2 l =2 f 109.870188 BHNCO 5(1,5)–4(1,4) v =1 109.870278HNCO 5(2,4)–4(2,3) 109.872366HNCO 5(2,3)–4(2,2) 109.872773110.29 HNCO 5(1,4)–4(1,3) 110.298098 E110.33 CH CN 6(2)–5(2) 110.320438 NCH
CN 6(1)–5(1) 110.326795CH
CN 6(0)–5(0) 110.328914CH CN 6(5)–5(5) F=7-6 110.330627CH CN 6(5)–5(5) F=5-4 110.330872110.35 CH CN 6(4)–5(4) F=7-6 110.349659 ECH CN 6(4)–5(4) F=5-4 110.349797110.69 CH CN 6(2)–5(2) v =1 110.695506 N l =-1CH CN 6(4)–5(4) v =1 110.698701 l =1110.71 CH CN 6(1)–5(1) v =1 110.706251 N l =-1CH CN 6(3)–5(3) v =1 110.709313 l =+1CH CN 6(0)–5(0) v =1 110.712166 l =1CH CN 6(2)–5(2) v =1 110.716212 l =1111.29 CH OH 7(2,5)–8(1,8) A+ 111.289601 N112.64 CH CH CN 13(1,13)–12(1,12) 112.646233 N112.84 U 112.839 N113.12 CN 1–0 J=1/2-1/2 113.123337 EF=1/2-1/2 tected, for which we do not present maps. We also discussthe velocities and line widths at the emission peaks for thevarious maps presented. These are summarised in Table 4.We use the rough frequency in GHz, rounded to two decimalplaces, in the figures, Tables 2 and 3 and text below, as aconvenient shorthand to refer to the lines. CO, C O and C O The isotopic carbon monoxide CO 1 – 0 (110.20 GHz) andC O 1 – 0 (109.78 GHz) integrated emission is shown inFig. 1. The CO emission is optically thick in the densestregions, with the ratio of the peak integrated emission of CO/C O of around 5, rather than ∼ O, whichshows two peaks associated with Sgr B2(M) and Sgr B2(N),with fitted positions (J2000) 17 47 20.3, -28 23 06 and 17 c (cid:13) , 1–22 -mm spectral imaging of the Sagittarius B2 region
47 19.5, -28 22 15, LSR velocities 63 and 68 km s − and fullwidth at half maximum 21 and 22 km s − respectively.The CO data cube, with intensity as a function ofvelocity (Fig. 2), agrees well with the results of Sato et al.(2000) and Hasegawa et al. (2007), showing the low veloc-ity ‘hole’ at 40–50 km s − and the high velocity ‘clump’ at70–80 km s − . However, the broad band data here are withpoorer velocity and spatial resolution than that of Sato et al.(2000) or Hasegawa et al. (2007), so we do not resolve de-tails in the spatial and velocity structure that they attributeto their cloud-cloud collision model (Hasegawa et al. 1994).The integrated CO and C O images (Fig. 1) also showthe northern emission ridge or ‘Edge’ (Hasegawa et al. 1994)with peak at 17 47 24.2, -28 20 49 (in CO) with cen-tral velocity 65 km s − (width 42 km s − ) from CO and63 km s − (width 36 km s − ) from C O. There is alsothe higher-velocity ridge to the west in CO (Fig. 2) withpeak at 17 47 14.0 -28 22 14, velocity 109 km s − (width 32km s − ).We have also imaged the weaker C O 1 – 0 (112.36GHz) data, which shows the densest CO peak nearSgr B2(M) at around 64 km s − . However, the C O dataare affected by the bandpass ripples, so we do not show theintegrated image here, or consider further quantitative anal-ysis (such as line ratios). CS and C S The carbon monosulphide CS 2 – 1 (97.98 GHz) inte-grated emission is shown in Fig. 3. The CS data cube (notshown here) shows that the main peak near Sgr B2(M)has a minimum around velocity 62 km s − , due to self-absorption at the position and velocity where the brightestCO is found. The CS also traces the low velocity ‘hole’ at ∼
35 km s − similar to the results of Sato et al. (2000) andTsuboi, Handa & Ukita (1999) using the CS 1 – 0 line at48.99 GHz.The CS 2 – 1 emission near Sgr B2(M) shows a velocitygradient, with the emission wings on either side of the 62km s − self-absorption offset: the peak around 85 km s − isat 17 47 19.8, -28 22 56 and the peak around 50 km s − is at17 47 19.5, -28 23 06. This is shown at higher resolution inBIMA observations of Mehringer (1995) who attribute thisto an outflow. The blue-shifted wing is stronger, so that theintegrated CS emission peaks at around 17 47 19.2, -28 2303 to the south-west of Sgr B2(M). There is very little CSemission from Sgr B2(N) indicating that it is underabundantin CS, relative to Sgr B2(M).The CS data cube also shows: the ‘south-east CS peak’noted by Yusef-Zadeh et al. (1996), centred at 17 47 27.1,-28 23 13, at 41 km s − , width 20 km s − ; the north ridgewith peak at 17 47 22.3, -28 20 49, at 61 km s − , width 57km s − ; and the west ridge with peak at 17 47 14.9, -28 2237, at 119 km s − , width 14 km s − (Sato et al. 2000).We also have data (not plotted here) from CS 2 – 1(92.49 GHz) and C S 2 – 1 (96.41 GHz) transitions, whichare much weaker, but are optically thin and do not suffer asmuch from the self-absorption. These confirm the lower CSemission from Sgr B2(N) than from Sgr B2(M), and showthat the peak near Sgr B2(M) is at 17 47 18.7, -28 23 11with velocity around 54 km s − , width 15 km s − . + , HCN, HNC, H CO + , H CN andHN C The integrated emission distributions of formylium (HCO + )1 – 0 (89.19 GHz), hydrogen cyanide HCN 1 – 0 (88.63 GHz)and hydrogen isocyanide HNC 1 – 0 (90.66 GHz) are shownin Fig. 3. The distributions are qualitatively similar, butrequire careful interpretation as they are strongly affectedby self-absorption. In particular, the low level of integratedemission in the centre, near Sgr B2(N) and Sgr B2(M), isdue to absorption, as is shown, for example, in the spectraand integrated images of Jacq et al. (1999).The whole area is filled with emission over a wide ve-locity range. Fitting spectra at the east edge of the imagedarea, away from the strong absorption in the centre, we finda peak velocity of 70 km s − , width 71 km s − for HCO + ,velocity 69 km s − , width 77 km s − for HCN and veloc-ity 55 km s − , width 67 km s − for HNC. This componentis enhanced in the area of the north ridge in HNC, withpeak position 17 47 22.0, -28 20 55, and single componentfit velocity 59 km s − , width 54 km s − (but there is someself-absorption at this position, making the single compo-nent gaussian not a very good fit). The west ridge adds tothis wide component, in the integrated images (Fig. 3), butis not well separated in velocity. Multi-component fits to thespectra show that it peaks at 17 47 14.8 -28 22 36 in HCO + (we cannot get a good fit to this component in velocity),peak 17 47 14.7 -28 22 34, velocity 119 km s − , width 24km s − in HCN and peak 17 47 14.9 -28 22 34, velocity 112km s − , width 21 km s − in HNC.The HCO + , HCN and HNC data cubes show a peaknear Sgr B2(M), much like that in CS, with self-absorptionaround 65 km s − . The spectra from this area show twocomponents around 46 km s − and 90 km s − , which areinterpreted as a single component with an absorption dip.There is also absorption of the Sgr B2 continuum emission,by gas along the line of sight, giving a broad negative fea-ture to the spectra between velocities -120 and 20 km s − .Quantitative analysis of the peak near Sgr B2(M) is affectedby the absorption.The fitted peak positions (at around 90 km s − ) are 1747 20.1, -28 22 34 in HCO + , 17 47 20.1, -28 22 32 in HCNand 17 47 19.8, -28 22 56 in HNC. This is near the CS peakand the C O peak, but as for CS, there may be a gradient ofposition with velocity. We also have data (not plotted here)of the corresponding weaker isotopologue lines H CO + CN 1 – 0 (86.34 GHz) and HN C 1 –0 (87.09 GHz) which also show some self-absorption, butare less affected, and hence better for the velocity fits. Thevelocities are peak 50 km s − , width 8 km s − for H CO + ,peak 47 km s − , width 12 km s − for H CN and peak 52km s − , width 16 km s − for HN C. The HN C fit is ingood agreement with the fit to the CS peak but the othertwo are a bit lower in velocity and narrower, presumably dueto the effect of the absorption.Note that the HCN, HNC, H CN and HN C lines aretriplets with hyperfine splitting, but that the spread of fre-quency for HNC and HN C is only 0.21 MHz, so this willhave negligible effect on the fitted velocity widths. For HCNand H CN however, the frequency range is 3.5 MHz, corre-sponding to velocity range 12 km s − , so that the blending c (cid:13) , 1–22 P. A. Jones et al.
Figure 1.
Integrated emission for CO and C O. In this, and subsequent images, the crosses indicate positions of radio peaks, asdescribed in Section 3, including in particular the positions of Sgr B2(N), (M) and (S). The squares show mid-IR sources. The opticallythin C O peaks near Sgr B2(M), while the CO shows the widespread diffuse emission. The peak brightness and contour steps are 240K km s − and 20 K km s − for CO, and 48 K km s − and 5 K km s − for C O. The beam size is shown in the bottom left cornerof the CO image. of the hyperfine components would contribute to increasingthe fitted velocity width.
The integrated silicon monoxide SiO 2 – 1 (86.85 GHz) emis-sion is also shown in Fig. 3.The SiO data cube and integrated image shows similarfeatures to the CS 2 – 1 data, but the SiO line is weaker and,thus, has lower lower signal to noise ratio. The integratedemission peak near Sgr B2(M) shows absorption at around65 km s − . The emission peak is at 17 47 18.9, -28 22 49, ve-locity 50 km s − , width 11 km s − , but this is affected by theabsorption with a second velocity component to the fit at 87km s − , width 31 km s − , on the redshifted side of the ab-sorption. There are also the north ridge, peak 17 47 22.5, -2821 06, velocity 58 km s − , width 41 km s − , and the south-east ‘CS peak’ at 17 47 27.1, -28 23 12, velocity 45 km s − ,width 29 km s − . These three peaks in integrated SiO, andthe absorption near Sgr B2(M) and Sgr B2(N) are also seenin the integrated SiO image of Martin-Pintado et al. (1997).Higher resolution BIMA data of the peak near Sgr B2(M)are interpreted by Liu et al. (1998) as an outflow (like theCS data.) Emission from the cyanide radical is observed in two groupsof blended hyperfine components at CN 1 – 0 J=1/2–1/2(113.17 GHz) and 1 – 0 J=3/2–1/2 (113.49 GHz), each ofwhich consists of several components. The distribution of theintegrated emission from the two sets of lines is very simi-lar, so the sum of the two sets is plotted here, in Fig. 3. Themost striking feature of the data is the strong absorptionassociated with Sgr B2(M) and Sgr B2(N) giving a deficitin the integrated emission in Fig. 3. This absorption is dueto spiral clouds along the line of sight (Greaves & Williams1994) against the strong continuum of the Sgr B2(M) andSgr B2(N) cores, rather than absorption in the Sgr B2 com-plex itself.Because of the multiple components, the data cubes are rather complicated with overlapping velocity and frequencystructure. The J=3/2–1/2 (113.49 GHz) data cube showsthe peak near Sgr B2(M) with position 17 47 20.1, -28 22 50,velocity 94 km s − (from the strongest component), width19 km s − . There is also absorption over a wide velocityrange down to -100 km s − , at the continuum peaks. Wetherefore interpret the velocity of the peak fit as being bi-ased high due to the absorption. This is confirmed by theoptically thin CN lines having velocity around 52 km s − (Gerin et al. 1984). There is widespread CN emission overthe whole area, with velocity 52 km s − , fitted at the northridge position, with broad lines (but the fitted velocity widthof 113 km s − includes the confusion of the multiple com-ponents). The CN emission is widespread compared to thedistribution of other molecules studied here. The J=1/2–1/2(113.17 GHz) data cube shows deep absorption features atSgr B2(M) and Sgr B2(N) but is too complicated to do muchmore interpretation, with the multiple components blended.We also detect (Table 3) the weak lines of the CN iso-topologue J=1/2–1/2 (108.65 GHz) and J=3/2–1/2 (108.78GHz) in extended emission and absorption at Sgr B2(N) andSgr B2(M). N The integrated emission from cyanoacetylene HC N 9 – 8(81.88 GHz), 10 – 9 (90.98 GHz), 11 – 10 (100.08 GHz) and12 – 11 (109.17 GHz) is shown in Fig. 4. All four transitionsshow similar structure, which is a ridge of emission to thewest of radio continuum peaks Sgr B2(N), Sgr B2(M) andSgr B2(S), looping to the east, north of Sgr B2(N).This is similar to the single-dish results ofLis & Goldsmith (1991) for the 12 – 11 transition,Chung, Ohishi & Morimoto (1994) for the 10 – 9 and 12 –11 transitions, and de Vicente, Martin-Pintado & Wilson(1997) for the 11 – 10 transition. Higher resolution interfer-ometer images of the HC CCN 9 – 8 transition at 81.53GHz are given by Kuan & Snyder (1996) and multipletransitions are given by de Vicente et al. (2000). The highresolution interferometer observations of the HC N 1 – 0transition at 9.10 GHz by Hunt et al. (1999) show weak c (cid:13) , 1–22 -mm spectral imaging of the Sagittarius B2 region Figure 2.
Velocity channel images of CO, separated by the channel spacing of 6 km s − . Note the west ridge peaking at 117 km s − ,the north ridge at 64 km s − , and how the hole at 40 – 50 km s − matches the clump at 70 – 80 km s − . The contours are steps of 0.5K, and the peak is 5.45 K. The crosses and squares are the same as for Figure 1.c (cid:13) , 1–22 P. A. Jones et al.
Figure 3.
Integrated emission for CS (contour step 5 K km s − , peak 74 K km s − ), HCO + (step 5 K km s − , peak 65 K km s − ),HCN (step 5 K km s − , peak 94 K km s − ), HNC (step 5 K km s − , peak 57 K km s − ), SiO (step 2 K km s − , peak 18.4 K km s − )and CN (step 10 K km s − , peak 93 K km s − ). Note that the grey-scale is darker for stronger emission, so the lighter shades nearSgr B2(N) (and Sgr B2(M) for SiO, HNC and CN) indicate lower integrated emission due to absorption. maser emission, and so preferential emission at the radiocontinuum peaks. While interesting in its own right, thisdoes not trace the molecular distribution well.The data cubes show that the emission has several peakswith different velocities, which are merged together in theintegrated emission images. We fit four peaks, from northto south, with a systematic velocity gradient: (a) the northcloud at 17 47 21.4, -28 21 29, north of Sgr B2(N), velocity68 km s − , width 23 km s − ; (b) peak at 17 47 18.7, -28 2212, near Sgr B2(N), velocity 67 km s − , width 23 km s − ;(c) peak at 17 47 18.6, -28 23 04, near Sgr B2(M), velocity60 km s − , width 22 km s − ; and (d) peak at 17 47 19.9, -2823 55, near Sgr B2(S), velocity 58 km s − , width 20 km s − .In addition to these four peaks we fit the north ridge at peak17 47 21.0, -28 20 54, velocity 62 km s − , width 27 km s − ,and the south-east peak at 17 47 26.3, -28 23 04, velocity 55km s − , width 23 km s − .We can calculate column densities of molecules in theupper level N u from the intensities of the transitions, usingthe simple assumption that lines are optically thin and inlocal thermodynamic equilibrium (LTE) by N u = (8 πν k/hc A ul ) Z T B dv (1)where A ul is the Einstein coefficient, and R T dv is the inte-gral over velocity of the brightness temperature T B of theemission line. Using the multiple HC N transitions we can,in principle, plot an excitation diagram of column densityin that level (expressed as ln( N u /g u )) versus the energy of the level (expressed as E u /k ), to determine the total col-umn density N and excitation temperature T ex , using theequation( N u /g u ) = ( N/Q T ) exp( − E u /kT ex ) (2)where Q T is the partition function at excitation temper-ature T ex , and g u is the statistical weight of the upperlevel. In practice, for the lines here in the 3-mm band,we do not have enough range in the energy levels for thisto be very reliable ( E u /k = 20 to 34 K for these lines).However, we can determine that there are spatial varia-tions in the excitation temperature, between the peaks, withthe cloud north of Sgr B2(N) giving T ex = 28 K (20 –46 K in the 1 σ range) and the others hotter with limits >
43 K, >
76 K and >
41 K (at the 1 σ level) for thepeaks near Sgr B2(N), Sgr B2(M) and Sgr B2(S) respec-tively. This is confirmed by considering the spatial vari-ation in ratios of the different transitions, and is consis-tent with the results of Chung, Ohishi & Morimoto (1994)and the higher kinetic temperature in these hot dense cores(de Vicente, Martin-Pintado & Wilson 1997). This analysisis complicated towards Sgr B2(N), as the IRAM 30-m sur-vey of Belloche et al. (2005, 2007) shows that the HC N issomewhat optically thick there.We also detect (Table 3) seven vibrationally excitedlines of HC N at 91.20, 91.33, 100.32, 100.71, 109.44, 109.60and 109.87 GHz, concentrated at Sgr B2(N), as the higherupper energy transitions are excited in this hot region. Wedetect weak lines of the isotopologues H CCCN, HC CCN c (cid:13) , 1–22 -mm spectral imaging of the Sagittarius B2 region and HCC CN at 88.17, 90.60, 99.65 and 108.71 GHz, manyof which appear to peak at Sgr B2(N), but as the lines areweak the spatial distribution is not clear. CN The integrated emission from methyl cyanide CH CN 5 – 4(91.99 GHz) and 6 – 5 (110.38 GHz) is shown in Fig. 4. Thereare multiple components for each of these transitions. Theintegrated emission for the two sets of lines is similar, andsimilar to that of the four HC N lines. It is also similar to thesingle-dish results of de Vicente, Martin-Pintado & Wilson(1997) for the 5 – 4 transition. We fit five components similarto that in HC N above: (a) the north cloud at 17 47 21.3,-28 21 28, velocity 68 km s − , width 35 km s − ; (b) peaknear Sgr B2(N) at 17 47 19.1, -28 22 12, velocity 66 km s − ,width 30 km s − ; (c) peak near Sgr B2(M) at 17 47 18.8,-28 23 11, velocity 61 km s − , width 33 km s − ; (d) peaknear Sgr B2(S) at 17 47 19.9, -28 23 54, velocity 59 km s − ,width 32 km s − ; and (e) the north ridge at 17 47 23.5, -28 21 01, velocity 64 km s − , width 41 km s − The ratioof integrated emission of the two lines indicates that thepeaks near Sgr B2(N) and Sgr B2(M) have higher excitationtemperature than the surrounding area, but the overlappingcomponents and low signal to noise make more quantitativeanalysis difficult.We also detect weak emission (Table 3) from severalmore transitions of CH CN at 92.26, 110.33, 110.35, 110.69and 110.71 GHz, and the isotopologues CH CN at 107.19GHz and possibly CH
CN at 110.33 GHz (as a blend).These are concentrated at the position of Sgr B2(N). OH and CH OH In Fig. 5 we show the integrated emission of five transitionsof methanol CH OH: 5(-1,5) – 4(0,4) E (84.52 GHz), 8(0,8)– 7(1,7) A+ (95.17 GHz), 2(1,2) – 1(1,1) A+ (95.91 GHz),2(0,2) – 1(0,1) A+ blend (96.74 GHz) and 2(1,1) – 1(1,0) A-(97.58 GHz). In addition, we show the integrated emission ofthe isotopologue CH OH 2(0,2) – 1(0,1) A+ blend (94.41GHz), and we have data, not plotted here for the CH OH0(0,0) – 1(-1,1) E (108.89 GHz) transition. The distribu-tion of integrated emission is quite different for the differenttransitions.Methanol is a very useful tracer of physical condi-tions, described as ‘the Swiss army knife of star formation’(Leurini et al. 2005), particularly when using simultaneousfits to multiple lines (Leurini et al. 2004). However, the exci-tation conditions of methanol can be very complicated, withcollisional and radiative excitation. For example, both the84.52 GHz and 95.17 GHz transitions here can be masers(Cragg et al. 1992). Also the A- and E-types can be consid-ered separate species, which have different abundances. Wedo not attempt to model the different CH OH lines here,but restrict ourselves to describing their overall features.The different lines mostly trace the same spatial andvelocity structure, despite the different relative intensitiesof the features. These are: (a) the north cloud at 17 47 21.4,-28 21 20, velocity 68 km s − , width 25 km s − ; (b) the peaknear Sgr B2(N) at 17 47 18.8, -28 22 14, velocity 67 km s − ,width 19 km s − ; (c) the peak near Sgr B2(M) at 17 47 18.2, -28 23 11, velocity 61 km s − , width 22 km s − ; and(d) the peak near Sgr B2(S) at 17 47 19.9, -28 23 57, velocity59 km s − , width 20 km s − . The 96.74 GHz CH OH lineand the 94.41 GHz CH OH line are blends of multipletransitions, so the velocity structure is confused. The 96.74GHz line also shows absorption at Sgr B2(N) and Sgr B2(M).Because it is the strongest line, however, it shows featuresnot seen in the other weaker lines: the south-east peak (seenin CS) at 17 47 26.7, -28 23 07, velocity 56 km s − , width 34km s − ; the western ridge at 17 47 15.0, -28 22 44, velocity120 km s − , width 21 km s − ; and a peak to the north-west of the main ridge-line at 17 47 14.5, -28 21 41, velocity70 km s − (and width unclear due to blending with otherfeatures).There are thirteen more weak CH OH lines detectedhere (Table 3) concentrated at the position of Sgr B2(N),that are higher upper energy lines excited in the hot core. CH OH We have also detected and imaged the ethanol CH CH OH6(0,6) – 5(1,5) (85.27 GHz) transition, but as the line isweak, and the data are affected by scanning stripes, the in-tegrated emission is not shown here. The emission is centredon the north cloud, and the line fit gives velocity 68 km s − ,width 21 km s − . We expect from Requena-Torres et al.(2006) that the ethanol CH CH OH column density followsthat of methanol CH OH, but the distributions of line emis-sion here differ due to excitation differences.The CH CH OH 7(0,7) – 6(1,6) and 5(1,5) – 4(0,4)(104.49 and 104.80 GHz) transitions show weak extendedemission (Table 3).
The integrated emission from isocyanic acid HNCO 4(0,4) –3(0,3) (87.93 GHz) and 5(0,5) – 4(0,4) (109.91 GHz) is shownin Fig. 6. The cloud 2 arcmin north of Sgr B2(M) is partic-ularly prominent in HNCO, as pointed out by Wilson et al.(1996) from observations of the 21.98 GHz 1 – 0 line, andas discussed in Minh et al. (1998) including observations, ashere, of the 4(0,4) – 3(0,3) and 5(0,5) – 4(0,4) lines. Wefind similar integrated emission in the 4(0,4) – 3(0,3) toMinh et al. (1998), and the velocity gradient in the datacubes, which they attribute to collapse. The 5(0,5) – 4(0,4)line at 109.91 GHz here also shows the ridge west of theSgr B2(N), Sgr B2(M) and Sgr B2(S) radio and infraredcontinuum peaks, but the ridge is less clearly broken intoclumps than in other molecules, such as HC N. We fit thenorth cloud at peak 17 47 21.6, -28 21 20, velocity 65 km s − ,width 25 km s − and the peak near Sgr B2(M) at 17 47 18.2,-28 23 01, velocity 66 km s − , width 29 km s − . From theratio of the two lines, the peak near Sgr B2(M) has a higherexcitation temperature, but the difference in energy of theupper levels is too small ( E u /k = 10 . c (cid:13) , 1–22 P. A. Jones et al.
Figure 4.
Integrated emission for HC N (contour step 10 K km s − ; 81.88 GHz, peak 55 K km s − ; 90.98 GHz, peak 72 K km s − ;100.08 GHz, peak 68 K km s − ; 109.17 GHz, peak 83 K km s − ) and CH CN (contour step 5 K km s − ; 91.99 GHz, peak 28 K km s − ;110.38 GHz, peak 43 K km s − ). These two molecules trace an arc from the north cloud, then west of the radio and mid-IR continuumpeaks. + The integrated emission from protonated CO HOCO + − , width 23 km s − , and the peak nearSgr B2(M) at 17 47 18.4, -28 23 21, velocity 63 km s − ,width 22 km s − . We also find the peak near Sgr B2(M)has a higher excitation temperature, from the ratio of thepeaks in the two lines. The difference in energy of the upperlevels is however too small ( E u /k = 10 . The integrated emission from carbonyl sulphide OCS 7 – 6(85.14 GHz), 8 – 7 (97.30 GHz) and 9 – 8 (109.46 GHz) isshown in Fig. 6. The emission traces the north cloud, andridge line, with the peaks near Sgr B2(N), Sgr B2(M) andSgr B2(S) quite compact and hence distinct in the integrated emission. This is unlike the more continuous ridge line seenin HC N (Fig. 4), as shown by the higher resolution datafrom Goldsmith et al. (1987) for the OCS 9 – 8 and HC N12 – 11 transitions. We fit (a) the north cloud at peak 17 4721.3, -28 21 18, velocity 65 km s − , width 23 km s − ; (b)the peak near Sgr B2(N) at 17 47 19.8, -28 22 12, velocity 66km s − , width 21 km s − ; (c) the peak near Sgr B2(M) at17 47 18.6, -28 23 08, velocity 62 km s − , width 21 km s − ;and (d) the peak near Sgr B2(S) at 17 47 19.5, -28 23 53,velocity 58 km s − , width 19 km s − . Despite having threetransitions, we cannot get reliable excitation temperaturesdue to the small range of upper energy levels ( E u /k = 16 . CS 9 – 8 line at96.98 GHz, which has a similar extended distribution, withthe strongest peak near Sgr B2(N), although the IRAM30-m survey of Belloche et al. (2005, 2007) indicates thisline is blended with several other lines at Sgr B2(N) andSgr B2(M).
The integrated emission from sulphur monoxide SO 2(2) –1(1) (86.09 GHz), 3(2) – 2(1) (99.30 GHz) and 2(3) – 1(2)(109.25 GHz) is shown in Fig. 7. The distribution of the86.09 GHz and 109.25 GHz transitions is similar, with com-pact peaks near Sgr B2(N) and Sgr B2(M), as shown by c (cid:13) , 1–22 -mm spectral imaging of the Sagittarius B2 region Figure 5.
Integrated emission for CH OH (84.52 GHz, contour step 10 K km s − , peak 93 K km s − ; 95.17 GHz, step 4 K km s − ,peak 23 K km s − ; 95.91 GHz, step 5 K km s − , peak 30 K km s − ; 96.74 GHz, step 10 K km s − , peak 131 K km s − ; 97.58 GHz,step 5 K km s − , peak 30 K km s − ) and CH OH (94.41 GHz, step 2 K km s − , peak 16.4 K km s − ). These lines show the arc fromthe north cloud, west of the radio and mid-IR continuum peaks, with differences in the relative intensities of the peaks related to thecomplicated excitation of the different levels. Goldsmith et al. (1987) at higher resolution for the 109.25GHz transition. The 99.30 GHz transition, however, showsquite a different distribution tracing the north cloud andridge-line to the west, and with absorption in the data cubeat Sgr B2(N) and Sgr B2(M). This is presumably becausethe 86.09 GHz and 109.25 GHz transitions trace the moreexcited gas ( E u /k = 19 . E u /k = 9 . − , width25 km s − ; (b) peak near Sgr B2(N) (86.09 and 109.25 GHz)at 17 47 19.3, -28 22 08, velocity 66 km s − , width 27 km s − ;and (c) peak near Sgr B2(M) at 17 47 19.8, -28 22 56, ve-locity 61 km s − , width 20 km s − .We also detect (Table 3) the SO 4(5) – 4(4) line at100.03 GHz, concentrated at Sgr B2(M) and Sgr B2(N), andthe isotopologue SO 3(2) – 2(1) and 2(3) – 1(2) lines at97.72 and 106.74 GHz, at Sgr B2(M). The integrated emission from sulphur dioxide SO − , width 26 km s − , whichis lower velocity here than as seen in other lines. The northcloud has velocity 68 km s − , width 21 km s − , and near SgrB2(N) velocity 61 km s − , width 29 km s − . The low level east-west extension is an artifact of the east-west scanningand baseline variations.We detect eight more lines of SO at 83.69, 91.55, 97.70,100.88, 104.24, 107.06, 107.84 and 109.75 GHz (Table 3)concentrated at Sgr B2(N) and Sgr B2(M). H + The integrated emission from diazenylium N H + − , and double-peaked spectra over mostof the area (Fig. 8), which we attribute to widespread ab-sorption at a similar velocity. There are multiple componentsto the 1 – 0 line, which contributes to broadening the fittedline width, but the frequency range is too small to explainthe double profiles. The major features are fitted as: (a) thenorth cloud at 17 47 21.4, -28 21 23, velocity 51 and 81km s − ; (b) peak to the west of Sgr B2(M) at 17 47 17.2,-28 23 06, velocity 47 and 79 km s − ; (c) peak south of SgrB2(S) at 17 47 20.1, -28 24 09, velocity 46 and 71 km s − ;(d) west ridge at 17 47 15.1, -28 22 38, velocity 120 km s − ,width 22 km s − ; and (e) south-east peak at 17 47 27.2, -2823 22, velocity 43 km s − , width 29 km s − . c (cid:13) , 1–22 P. A. Jones et al.
Figure 6.
Integrated emission for HNCO (contour step 10 K km s − ; 87.93 GHz, peak 84 K km s − ; 109.91 GHz, peak 113 K km s − ),HOCO + (85.53 GHz, step 2 K km s − , peak 13.4 K km/s) and OCS (85.14 GHz, step 2 K km s − , peak 15.6 K km s − ; 97.30 GHz,step 2 K km s − , peak 24 K km s − ; 109.46 GHz, step 4 K km s − , peak 24 K km s − ). CCH
The integrated emission from propyne or methyl acetyleneCH CCH 6 – 5 (102.55 GHz) is shown in Fig. 7. We alsohave data for the CH CCH 5 – 4 (85.46 GHz) transition,not shown here as it is qualitatively similar, but weaker andnoisier. The distribution shows the north cloud, and ridge-line west of the radio continuum peaks. The fitted featuresare: (a) the north cloud at 17 47 21.5, -28 21 23, velocity 73km s − , width 23 km s − ; (b) peak near Sgr B2(N) and SgrB2(M) at 17 47 18.9, -28 22 33, velocity 70 km s − , width24 km s − ; (c) peak near Sgr B2(M) and Sgr B2(S) at 1747 19.5, -28 23 22, velocity 65 km s − , width 25 km s − ;and (d) peak south of Sgr B2(S) at 17 47 20.4, -28 24 04,velocity 61 km s − , width 23 km s − . There are multipleblended line components, so the spectra are a fit to the lineblend with the velocity calculated using the rest frequency ofone of the components. The velocity therefore is offset, butthe gradient is shown, similar to that found in CH CCH byChurchwell & Hollis (1983) with lower resolution, but overa larger area. CHO and H COH + The integrated emission from the line at 102.07 GHz isshown in Fig. 9. We identify this as a blend of formamideNH CHO 5(1,5) – 4(1,4) and protonated formaldehydeH COH + CCH 6 – 5(102.55 GHz) above, but because the line is weak, we do not get good positional fits for all of them. The features are: (a)the north cloud with velocity 65 km s − , width 24 km s − ;(b) peak near Sgr B2(N) and Sgr B2(M) at 17 47 20.1, -28 2227, velocity 64 km s − , width 13 km s − ; (c) peak near SgrB2(M) and Sgr B2(S) at 17 47 18.7, -28 23 31, velocity 58km s − , width 12 km s − ; and (d) peak south of Sgr B2(S)with velocity 53 km s − , width 12 km s − . The velocitiesare calculated using the rest frequency of NH CHO 5(1,5) –4(1,4), so again will be shifted due to the blending.We detect eight more weak lines of NH CHO at 85.09,87.85, 93.87, 105.46, 105.97, 106.13, 106.54 and 109.75 GHz(Table 3). The spatial distribution shows excitation differ-ences, with some of these concentrated at Sgr B2(N), andothers extended in the north-south ridge line a bit to thewest. This is consistent with higher upper energy lines be-ing excited in the hot core, although complicated by severalof the lines being blended with other species. CN The integrated emission from cyanamide NH CN 5(1,4) –4(1,3) (100.63 GHz) is shown in Fig. 9. The line is ratherweak, and the data affected by scanning ripples, so we donot fit the positions, but we do see the four peaks in the datacube and fit velocities: (a) the north cloud with velocity 59km s − , width 27 km s − ; (b) peak near Sgr B2(N) and SgrB2(M) with velocity 60 km s − , width 35 km s − ; (c) peaknear Sgr B2(M) and Sgr B2(S) with velocity 55 km s − , c (cid:13) , 1–22 -mm spectral imaging of the Sagittarius B2 region Figure 7.
Integrated emission for SO (86.09 GHz, contour step 2 K km s − , peak 11.1 K km s − ; 99.30 GHz, step 4 K km s − , peak25 K km s − ; 109.25 GHz, step 5 K km s − , peak 36 K km s − ), SO (step 5 K km s − , peak 35 K km s − ), N H + (step 5 K km s − ,peak 52 K km s − ) and CH CCH (step 5 K km s − , peak 35 K km s − ). Figure 8.
Spectra of N H + and HC N near Sgr B2(M) illustrating how some spectral lines show absorption at around 60 km s − , fromgas seen in emission at this velocity in other lines. There are multiple components to the N H + line, which contributes to broadeningthe line, but the frequency range is too small to explain the double profile. width 26 km s − ; and (d) peak south of Sgr B2(S) withvelocity 55 km s − , width 17 km s − . NH The integrated emission from methylenimine CH NH 4(0,4)– 3(1,3) (105.79 GHz) is shown in Fig. 9. The peak closeto Sgr B2(N) is strong, and we detect the north cloud andthe ridge line, with fits: (a) the north cloud with velocity66 km s − , width 19 km s − ; (b) peak near Sgr B2(N) at17 47 20.0, -28 22 21, velocity 61 km s − , width 27 km s − ; and (c) peak near Sgr B2(M) and Sgr B2(S) with velocity59 km s − , width 18 km s − . This line is probably blendedwith HC CCN 12 – 11 at rest frequency 105.799093 GHz. CS The integrated emission from thioformaldehyde H CS 3(0,3)– 2(0,2) (103.04 GHz) is shown in Fig. 9. We also havedata (not plotted here as the images are similar, but noisy)for H CS 3(1,3) – 2(1,2) (101.48 GHz) and 3(1,2) – 2(1,1)(104.62 GHz). We fit four peaks as: (a) the north cloud at c (cid:13) , 1–22 P. A. Jones et al.
Figure 9.
Integrated emission for NH CHO blended with H COH + (contour step 2 K km s − , peak 6.6 K km s − ), NH CN (step 2K km s − , peak 8.4 K km s − ), CH NH (step 2 K km s − , peak 12.3 K km s − ), H CS (103.04 GHz, step 2 K km s − , peak 10.0 Kkm s − ), C H (step 5 K km s − , peak 40 K km s − ) and c-C H (step 2 K km s − , peak 11.1 K km s − ).
17 47 21.4, -28 21 25, velocity 68 km s − , width 21 km s − ;(b) peak near Sgr B2(N) at 17 47 19.1, -28 22 23, velocity 67km s − , width 20 km s − ; (c) peak near Sgr B2(M) and SgrB2(S) at 17 47 19.2, -28 23 21, velocity 59 km s − , width 19km s − ; and (d) peak south of Sgr B2(S) at 17 47 20.2, -2824 05, velocity 57 km s − , width 17 km s − . H The integrated emission from ethynyl C H 1 – 0 J=1/2–1/2(87.32 GHz) and 1 – 0 J=3/2–1/2 (87.40 GHz), is shown inFig. 9. As for CN, above, each set consists of componentsand the integrated emission of two sets of lines are verysimilar, but weak, so the sum of the two sets is plotted here.The integrated emission image in Fig. 9 shows widespreademission peaked at the north ridge and west of Sgr B2(M),and a deficit of emission at the continuum peaks Sgr B2(N)and Sgr B2(M). The fitting of velocity components in thedata cubes is complicated by the blended components, andthe weakness of the emission, but the emission is peakedaround 60 – 65 km s − . The deficit of integrated emissionnear Sgr B2(N) and Sgr B2(M) could be explained by areal deficit of the molecule in this area, but is more likelysimply be due to absorption, as Greaves & Nyman (1996)show absorption features due to intervening clouds along theline of sight to Sgr B2. The offset between the absorption andthe radio continuum peaks is not considered significant, butrather due to the baseline stripes in the east-west scanning direction causing north-south shifts in centres of the weakfeatures. H The integrated emission from the cyclic molecule cyclo-propenylidene c-C H H, that iswidespread emission with a deficit at continuum peaksSgr B2(N) and Sgr B2(M). The emission is weak, however,so the integrated emission does show some spurious strip-ing due to the RA scanning. Vrtilek, Gottlieb & Thaddeus(1987) find rotation temperature T rot = 11 ± H in Sgr B2, so absorption against the continuum is quite plau-sible. As for C H, above, the position offset between the ab-sorption and continuum peaks is not considered significant.
We list in Table 3 nine more molecules, and dozens morelines, than we have plotted and discussed above, as well asweaker transitions of the molecules already discussed. Mostof these lines are confined to Sgr B2(N) or Sgr B2(M). Someof the weaker transitions are higher energy states, some vi-brationally excited, of molecules already discussed, whichtrace these hot cores.Since the main aim of this paper is the wider scale spa-tial distribution, we do not concentrate here on quantitative c (cid:13)000
We list in Table 3 nine more molecules, and dozens morelines, than we have plotted and discussed above, as well asweaker transitions of the molecules already discussed. Mostof these lines are confined to Sgr B2(N) or Sgr B2(M). Someof the weaker transitions are higher energy states, some vi-brationally excited, of molecules already discussed, whichtrace these hot cores.Since the main aim of this paper is the wider scale spa-tial distribution, we do not concentrate here on quantitative c (cid:13)000 , 1–22 -mm spectral imaging of the Sagittarius B2 region analysis of the weaker lines. Our Mopra OTF mapping sacri-fices sensitivity on a single position to get the spatial cover-age. Therefore our data on the spectra at the Sgr B2(N) andSgr B2(M) positions are not particularly sensitive comparedto previous (Turner 1989) and current (Belloche et al. 2005;Hieret et al. 2005; Belloche et al. 2007) dedicated spectralline surveys of these well studied sources. However, by map-ping it is useful to determine whether a particular line isconfined to Sgr B2(N), Sgr B2(M), or both, or whether itis distributed more widely. Of the weaker lines (Table 3)a substantial fraction are identified with blends of differentspecies, complicating the analysis.Four molecules in Table 3 have extended spatial dis-tribution: acetaldehyde CH CHO (93.60, 95.95, 95.96 and98.90 GHz), dicarbon monosulphide CCS (93.87 and 106.35GHz), methanethiol CH SH (101.03 and 101.14 GHz) andketene CH CHO (101.03 GHz). These distributions appearsimilar to that of some other molecules, such as HC N, withthe north cloud and ridge line to the west of the radio con-tinuum peaks, but with much lower signal to noise.The other five molecules are confined to Sgr B2(N),as this region is known to be particularly rich inlarge molecules (Snyder, Kuan & Miao 1994; Miao et al.1995). These molecules are: ethyl cyanide or propioni-trile CH CH CN (22 lines), acrylonitrile CH CHCN (92.43,94.28, 94.91, 95.33, 103.57, 104.21 and 106.64 GHz),methyl formate CH OCHO (89.32, 98.18 and 107.54 GHz),dimethyl ether CH OCH (82.46, 100.46 and 105.77 GHz)and formaldehyde H CO (101.33 GHz).More sensitive observations of Sgr B2(N) andSgr B2(M) with the IRAM 30-m (Belloche et al. 2005,2007) have been modelled with the XCLASS software(Comito et al. 2005), which simultaneously fits multiplelines with the LTE approximation and handles line blendswell. We note here that for Sgr B2(N), in particular, thisallows us to identify some extra lines that may confuse theMopra images. These lines are: CH CH CN 10(1,10) – 9(1,9)at 86.819848 GHz for SiO; CH CHCN multiplet around85.5329236 GHz for HOCO + ; HC N 12 – 11 ν = 1 l = 1 f at 109.244339 GHz for SO; CH CHCN 9(1,8) – 8(1,7) at87.312827 GHz for C H. However, the effect of this line con-fusion does not appear to be significant.
We now consider the comparison of spatial and velocitystructure in the Sgr B2 complex, as traced by the different3-mm lines. Figure 10 shows the positions of the molecularpeaks listed in Section 3, and Table 4 lists these fitted peaks.The strongest lines, such as CO, C O, CS, HCN,HCO + , HNC, SiO, N H + and CH OH (96.74 GHz) show upthree features which we have called here the north ridge, thewest ridge and the south-east peak (Tables 4 and 5). Thesefeatures are detected in the strongest transitions, which arealso the lines which are optically thick in the densest regionsof the complex (near Sgr B2(N) and (M)), so the relativeprominence of these three features (Figure 2) is partly due tothis optical depth effect. However, they do trace the weakersurrounding structure of the complex. We have not imageda large enough area to show the ‘hole’ around 40 km s − (Sato et al. 2000; Hasegawa et al. 2007) well, so we do notconsider the wider surrounding structure.We note that the south-east peak is much more ob-vious in the CS, than in say CO or C O, as noted byYusef-Zadeh et al. (1996), indicating that there is a chemi-cal difference from the main sources.The west ridge and south-east peak are offset both spa-tially and in velocity (at around 117 and 48 km s − re-spectively) from the main north-south axis of the Sgr B2complex. The other features (Table 5), that we have calledthe north ridge, the north cloud, and the three groups ofpeaks near Sgr B2(N), Sgr B2(M) and Sgr B2(S) are in anorth-south line, with a velocity gradient, as shown on theright of Figure 10 and in Table 5.The north ridge is (as noted) seen only in the strongestlines, while the other four features are best traced by weaker,optically thin lines. We find a spatial and velocity differencebetween the north ridge, and nearby chemically enriched(Minh et al. 1998) north cloud. The north ridge is elongatedeast-west, so there is a scatter of the peak positions alongthis axis, but the north cloud has a surprisingly tight distri-bution of peaks fitted from the different lines.The feature near Sgr B2(S) also has a fairly tight dis-tribution of fitted peak positions, given the 36 to 39 arcsecbeamsize of the observations. However, there is a signifi-cant difference in the peak positions, for the groups of fittedpeaks near Sgr B2(N) and Sgr B2(M). This is attributed toa real difference in the positions of the peaks in differentlines, where some more excited lines are associated with thecompact hot cores Sgr B2(N) and Sgr B2(M), or particularlyfor Sgr B2(N) some molecules are concentrated there. Otherlower excitation lines peak in the ridge further to the west ofSgr B2(N) and Sgr B2(M) and avoid the hot core positionsas the molecules are destroyed there. The excitation effectcan be seen clearly in the SO lines (Figure 7) where the86.09 and 109.25 GHz lines are concentrated at Sgr B2(M),while the 99.30 GHz line traces the ridge-line more to thewest.From some of the stronger lines in Table 3 which areconcentrated at Sgr B2(N) and Sgr B2(M) we fit the hotcore positions and velocities as: Sgr B2(N) 17 47 19.9, -2822 11, velocity 63 km s − , width 24 km s − ; and Sgr B2(M)17 47 20.3, -28 22 58, velocity 59 km s − , width 22 km s − .From these lines (mostly CH OH and CH CH CN for SgrB2(N) and SO and SO for Sgr B2(M)) we find that the hotcores are unresolved relative to the 36 to 39 arcsec Moprabeam.The distribution of optically thin C O, which shouldbe a good tracer of CO column density, and hence H col-umn density, peaks at the Sgr B2(N) and Sgr B2(M) cores,whereas there are several molecules, such as HC N, CH CN,CH OH and OCS, which peak in the ridge-line to the west ofthe cores. This is shown in Figure 11, and in the integratedemission images, by the alignment of the distributions rela-tive to the reference crosses (radio peaks) and squares (mid-IR peaks).We also show in Figure 12 the 20-cm radio, from the c (cid:13) , 1–22 P. A. Jones et al.
Figure 10.
The position (left) of the peaks fitted for the 3-mm molecular lines, and (right) the velocity as a function of declination.Note that in the velocity-declination plot, the points for the SE peak have been shifted 1 arcmin north for clarity, to avoid overlappingthe points near Sgr B2(M). In the left diagram, the points are the 3-mm molecular peaks, the crosses radio sources and the open boxesmid-IR sources.
Table 5.
Summary of molecular features in the Sgr B2 complex, from the Mopra 3-mm peaks. We give the mean and standard deviationof position, velocity and velocity width, from the fits to different lines, and include positions in galactic coordinates for reference.Feature R.A. (J2000) Dec. (J2000) σ (R.A.) σ (Dec.) lat. long. Velocity σ (Vel.) Width σ (Width)arcsec arcsec degree degree km s − km s − km s − km s − north ridge 17 47 22.6 -28 20 56 15 7 0.702 -0.024 62 3 43 10north cloud 17 47 21.4 -28 21 24 2 4 0.693 -0.024 66 3 24 4near Sgr B2(N) 17 47 19.3 -28 22 18 7 8 0.676 -0.026 65 3 25 5near Sgr B2(M) 17 47 19.2 -28 23 04 11 15 0.665 -0.032 59 4 22 5near Sgr B2(S) 17 47 20.0 -28 24 00 4 6 0.653 -0.043 58 2 21 6west ridge 17 47 14.8 -28 22 34 5 9 0.664 -0.014 117 5 22 6south-east peak 17 47 26.9 -28 23 12 5 7 0.678 -0.057 48 7 27 6 VLA , the 850- µ m sub-mm from SCUBA , and 21- µ mmid-IR, from MSX .We point out that, for the many molecules here thatpeak in the ridge-line to the west of the Sgr B2(N) andSgr B2(M) cores, this distribution of molecular emission‘wraps around’ the north and west side, where there is lit-tle radio and mid-IR emission tracing recent star forma-tion, and avoids the south-east area where there is recentstar formation. The north cloud, in particular, is quite iso-lated from the recent star formation activity. In contrast,the SCUBA sub-mm emission, tracing cooler dust than themid-IR, shows extended emission around the Sgr B2(N) andSgr B2(M) cores, to the north and west which matches wellthe north cloud and the molecular ridge-line.Each line that we have imaged here has its own partic-ular distribution, corresponding to the spatial distributionof the molecule, and the effect of the excitation of the differ-ent levels. There is also a complicated velocity structure inthe region. However, we can make some generalisations andcomments here to bring some order to the overall results. http://imagelib.ncsa.uiuc.edu/imagelib.html http://irsa.ipac.caltech.edu/applications/MSX/ Figure 11.
The C O integrated emission as grey-scale, with the90.98 GHz HC N integrated emission as contours, showing howmolecules such as HC N peak in the ridge-line to the west of thehot cores.
The CO ( CO, C O and C O) shows that the dens-est region is around Sgr B2(M) at velocity 63 km s − . The CO is optically thick at this core, so the density there isbetter traced by the C O. The column density, would befurther concentrated at Sgr B2(N) and Sgr B2(M), than the c (cid:13) , 1–22 -mm spectral imaging of the Sagittarius B2 region Figure 12.
The continuum emission in radio (left) from the VLA at 20-cm, sub-mm (middle) from SCUBA at 850 µ m and mid-IR(right) from MSX at 21 µ m. The overlaid crosses are for radio sources and open squares for mid-IR sources. Note how the sub-mm tracesdiffuse cool dust to the west and north of the Sgr B2(N), (M) and (S) cores, as well as compact emission from the cores. The radio andmid-IR trace star formation in the cores and to the south-east. C O 1 – 0 integrated line emission shown here. The highertemperature at the cores ( ∼
200 K, compared to ∼
20 K forthe surrounding gas) leads to an extra factor there, whenconverting, with the standard LTE analysis, from integratedline emission to total CO column density, and hence totalH column density.The HCO + , HCN and HNC are strong and widespreadwith absorption at the Sgr B2(M) and Sgr B2(N) cores. Thecolumn density is likely to be peaked at these cores, but theintegrated emission is strongly affected by this absorption,leading to local minima in the emission intensity at the cores.There are differences in the detailed distribution of thesethree lines, as expected: HCN should be a good tracer of highgas density ( > cm − ), the isomer HNC should trace coolquiescent gas, and the ion HCO + should trace ionisation dueto cosmic rays. We have further Mopra data of these linesover a larger area, from broad-band observations over the85.3 to 93.3 GHz range, which show the differences moreclearly, so we postpone further discussion for a later paper.The CS and SiO distributions are also affected by ab-sorption at the Sgr B2(M) and Sgr B2(N) cores, so the col-umn density distribution is hard to determine from the in-tegrated emission images. CS is expected to be, like HCN, agood tracer of high density gas, and SiO is expected to traceshocks, but is quite widespread here.The CN emission is quite uniform over the 5 × H and c-C H also have absorption at theSgr B2(M) and Sgr B2(N) cores and widespread emission,but some excess emission on the ridge-line west of the cores.Most of the lines imaged here trace the ridge-linewest of the Sgr B2(M), Sgr B2(N) and Sgr B2(S) cores,and north-east to what we are calling the north cloud.These lines include HC N, CH CN, CH OH, HNCO, OCS,N H + , CH CCH, NH CHO/H COH + , NH CN, CH NHand H CS. These more complex molecules, as noted above,trace the cooler dust seen at sub-mm wavelengths, and avoidthe areas with the warmer dust (mid-IR) and radio contin-uum associated with the active star formation. N H + and the 96.74 GHz transition of CH OH are strong, and alsoshow some absorption at the Sgr B2(M) and Sgr B2(N)cores.The relative prominence of the peaks in the ridge-linewest of the Sgr B2(M), Sgr B2(N) and Sgr B2(S) cores inthese different molecules, and between different transitionsof the same molecule (e.g.CH OH), indicate differences inchemistry and excitation conditions.The lines of HNCO and HOCO + highlight the northcloud, and are tracers of shock chemistry.The lines of SO and SO are also tracers of shocks, andare concentrated at Sgr B2(M), although the lower excita-tion 99.30 GHz SO line also traces the more extended gasin the north cloud and ridge to the west. We have undertaken a 3-mm spectral-line imaging surveyof the Sgr B2 area, of 5 arcmin square, with the Mopratelescope, at resolution ∼
36 arcsec. We covered almost thecomplete spectral the range 81.7 to 113.5 GHz, with 2.2MHz or ∼ − spectral channels, and have observed 24lines, with 0.033 MHz, or ∼ . − channels. We havediscussed the distribution of around 50 lines, and presentedintegrated emission images for 38 of the lines. In addition, wehave detected around 120 more lines, mostly concentratedat Sgr B2(N).By fitting the peak position and velocity of the emis-sion in the various lines, we find that there are seven distinctmolecular features in the region, which show distinct differ-ences in both molecular abundances and excitation condi-tions. ACKNOWLEDGMENTS
The Mopra telescope is funded by the Commonwealth ofAustralia as a National Facility managed by CSIRO as partof the Australia Telescope. The UNSW MOPS digital fil-terbank was provided with funding from the Australian Re-search Council, University of New South Wales, Sydney Uni-versity, Macquarie University and the CSIRO ATNF. PAJ c (cid:13) , 1–22 P. A. Jones et al.
Table 4.
Compilation of fitted peaks of the molecular features inthe Sgr B2 complex. We are mostly considering here the spatialand velocity structure, but include in this table, for complete-ness, the intensity of the fitted peaks in the T ∗ A scale. For somemolecules with multiple transitions, where we have used the meanspatial position and velocity for higher signal to noise, we list theintensities for the different transitions in consecutive lines, in or-der of frequency, as given in Table 2.Feature / R.A. Dec. Vel. Width T ∗ A Molecule (J2000) (J2000) km s − km s − K N ridge CO 17 47 24.2 -28 20 49 65 42 3.03C O 63 36 0.56CS 17 47 22.3 -28 20 49 61 57 1.01HNC 17 47 22.0 -28 20 55 59 54 0.96SiO 17 47 22.5 -28 21 06 58 41 0.32HC N 17 47 21.0 -28 20 54 62 27 1.151.481.301.33CH CN 17 47 23.5 -28 21 01 64 41 0.370.54
N cloud HC N 17 47 21.4 -28 21 29 68 23 1.481.981.791.88CH CN 17 47 21.3 -28 21 28 68 35 0.510.70CH OH 17 47 21.4 -28 21 20 68 25 3.530.400.692.170.810.53 CH OH 0.52HNCO 17 47 21.6 -28 21 20 65 25 2.944.03HOCO +
17 47 21.1 -28 21 29 67 23 0.540.60OCS 17 47 21.3 -28 21 18 65 23 0.540.680.74SO 17 47 21.3 -28 21 20 66 25 0.100.560.33SO
68 21 0.30N H +
17 47 21.4 -28 21 23 1.39CH CCH 17 47 21.5 -28 21 23 (73) (23) 0.300.46NH CHO (65) (24) 0.16/H COH + NH CN 59 27 0.30CH NH 66 19 0.22H CS 17 47 21.4 -28 21 25 68 21 0.660.290.47 nearSgr B2(N) C O 17 47 19.5 -28 22 15 68 22 1.60HC N 17 47 18.7 -28 22 12 67 23 1.662.292.162.69
Table 4 continued.Feature / R.A. Dec. Vel. Width T ∗ A Molecule (J2000) (J2000) km s − km s − KCH CN 17 47 19.1 -28 22 12 66 30 0.721.09CH OH 17 47 18.8 -28 22 14 67 19 2.691.051.181.771.190.32 CH OH 0.43OCS 17 47 19.8 -28 22 12 66 21 0.621.111.28SO 17 47 19.3 -28 22 08 66 27 0.350.360.87SO
61 29 0.23CH CCH 17 47 18.9 -28 22 33 (70) (24) 0.360.61NH CHO 17 47 20.1 -28 22 27 (64) (13) 0.42/H COH + NH CN 60 35 0.23CH NH 17 47 20.0 -28 22 21 61 27 0.46H CS 17 47 19.1 -28 22 23 67 20 0.670.350.39 nearSgr B2(M) C O 17 47 20.3 -28 23 06 63 21 1.87CS 17 47 19.2 -28 23 03 2.70 CS 17 47 18.7 -28 23 11 54 15 0.35C S 0.64HCO +
17 47 20.1 -28 22 34 1.44HCN 17 47 20.1 -28 22 32 1.66HNC 17 47 19.8 -28 22 56 1.37H CO + (50) (8) 0.48H CN (47) (12) 0.41HN C 52 16 0.32SiO 17 47 18.9 -28 22 49 0.54CN 17 47 20.1 -28 22 50 0.74HC N 17 47 18.6 -28 23 04 60 22 2.022.952.833.72CH CN 17 47 18.8 -28 23 11 61 33 0.741.08CH OH 17 47 18.2 -28 23 11 61 22 3.010.991.181.911.190.65 CH OH 0.41HNCO 17 47 18.2 -28 23 01 66 29 1.832.97HOCO +
17 47 18.4 -28 23 21 63 22 0.230.35OCS 17 47 18.6 -28 23 08 62 21 0.540.810.95SO 17 47 19.8 -28 22 56 61 20 0.590.961.66SO
17 47 20.4 -28 23 04 52 26 1.00c (cid:13) , 1–22 -mm spectral imaging of the Sagittarius B2 region Table 4 continued.Feature / R.A. Dec. Vel. Width T ∗ A Molecule (J2000) (J2000) km s − km s − KN H +
17 47 17.4 -28 23 06 0.92CH CCH 17 47 19.5 -28 23 22 (65) (25) 0.390.79NH CHO 17 47 18.7 -28 23 31 (58) (12) 0.40/H COH + NH CN 55 26 0.21CH NH 59 18 0.35H CS 17 47 19.2 -28 23 21 59 19 0.900.510.67AlF 17 47 19.7 -28 22 56 0.29 nearSgr B2(S) HC N 17 47 19.9 -28 23 55 58 20 1.532.632.452.52CH CN 17 47 19.9 -28 23 54 59 32 0.730.94CH OH 17 47 19.9 -28 23 57 59 20 2.000.751.102.041.070.66 CH OH 0.42OCS 17 47 19.5 -28 23 53 58 19 0.420.580.67N H +
17 47 20.1 -28 24 09 1.41CH CCH 17 47 20.4 -28 24 04 (61) (23) 0.200.46NH CHO (53) (12) 0.32/H COH + NH CN 55 17 0.25H CS 17 47 20.2 -28 24 05 57 17 0.660.430.48
W ridge CO 17 47 14.0 -28 22 14 109 32 1.68CS 17 47 14.9 -28 22 37 119 14 0.42HCO +
17 47 14.8 -28 22 36HCN 17 47 14.7 -28 22 34 119 24 0.63HNC 17 47 14.9 -28 22 34 112 21 0.27CH OH 17 47 15.0 -28 22 44 120 21 0.60N H +
17 47 15.1 -28 22 38 120 22 0.35
SE peak
CS 17 47 27.1 -28 23 13 41 20 1.53SiO 17 47 27.1 -28 23 12 45 29 0.36HC N 17 47 26.3 -28 23 04 55 23 0.700.980.941.06CH OH 17 47 26.7 -28 23 07 56 34 1.72N H +
17 47 27.2 -28 23 22 43 29 0.89 thanks the Max-Planck-Institut f¨ur Radioastronomie, Bonn,for a Visiting Fellowship in 2006, and the anonymous refereefor comments that improved the presentation of the paper.
REFERENCES
Bains I., et al., 2006, MNRAS, 367, 1609Belloche A., Comito C., Hieret C., Menten K. M., SchilkeP., M¨uller H. S. P., 2005, IAUS, 235, 332PBelloche A., Comito C., Hieret C., Menten K. M., M¨ullerH. S. P., Schilke P., 2007, in Molecules in Space & Labo-ratory, eds J.L. Lemaire, F. Combes, in pressCaswell J. L., 1996, MNRAS, 283, 606Chung H. S., Ohishi M., Morimoto M., 1994, JKAS, 27, 1Churchwell E., Hollis J. M., 1983, ApJ, 272, 591Comito C., Schilke P., Phillips T. G., Lis D. C., Motte F.,Mehringer D., 2005, ApJS, 156, 127Cragg D. M., Johns K. P., Godfrey P. D., Brown R. D.,1992, MNRAS, 259, 203Cummins S. E., Linke R. A., Thaddeus P., 1986, ApJS, 60,819de Vicente P., Martin-Pintado J., Wilson T. L., 1997, A&A,320, 957de Vicente P., Mart´ın-Pintado J., Neri R., Colom P., 2000,A&A, 361, 1058Eisenhauer F., Sch¨odel R., Genzel R., Ott T., Tecza M.,Abuter R., Eckart A., Alexander T., 2003, ApJ, 597, L121Garwood R. W., 2000, ASPC, 216, 243Gaume R. A., Claussen M. J., 1990, ApJ, 351, 538Gaume R. A., Claussen M. J., de Pree C. G., Goss W. M.,Mehringer D. M., 1995, ApJ, 449, 663Gerin M., Combes F., Encrenaz P., Linke R., DestombesJ. L., Demuynck C., 1984, A&A, 136, L17Goldsmith P. F., Snell R. L., Hasegawa T., Ukita N., 1987,ApJ, 314, 525Goldsmith P. F., Lis D. C., Lester D. F., Harvey P. M.,1992, ApJ, 389, 338Gordon M. A., Berkermann U., Mezger P. G., Zylka R.,Haslam C. G. T., Kreysa E., Sievers A., Lemke R., 1993,A&A, 280, 208Greaves J. S., Nyman L.-A., 1996, A&A, 305, 950Greaves J. S., Williams P. G., 1994, A&A, 290, 259Hasegawa T., Sato F., Whiteoak J. B., Miyawaki R., 1994,ApJ, 429, L77Hasegawa T., Arai T., Yamaguchi N., Sato F., 2007,Ap&SS, 234Hieret C., Belloche A., Comito C., Menten K. M., SchilkeP., 2005, IAUS, 235, 67PHoffmann W. F., Frederick C. L., Emery R. J., 1971, ApJ,164, L23Hunt M. R., Whiteoak J. B., Cragg D. M., White G. L.,Jones P. A., 1999, MNRAS, 302, 1Jacq T., Baudry A., Walmsley C. M., Caselli P., 1999,A&A, 347, 957Kuan Y.-J., Snyder L. E., 1996, ApJ, 470, 981Ladd N., Purcell C., Wong T., Robertson S., 2005, PASA,22, 62Lequeux J., 1962, AnAp, 25, 221Leurini S., Schilke P., Menten K. M., Flower D. R., PottageJ. T., Xu L.-H., 2004, A&A, 422, 573 c (cid:13) , 1–22 P. A. Jones et al.
Leurini S., Schilke P., Wyrowski F., Menten K., 2005,IAUS, 235, 99PLis D. C., Goldsmith P. F., 1989, ApJ, 337, 704Lis D. C., Goldsmith P. F., 1990, ApJ, 356, 195Lis D. C., Goldsmith P. F., 1991, ApJ, 369, 157Liu S.-Y., Mehringer D. M., Miao Y., Snyder L. E., 1998,ApJ, 501, 680Lovas F.J., 2002, J. Phys. Chem. Ref. Data 21, 181http://physics.nist.gov/PhysRefData/Micro/Html/McGrath E. J., Goss W. M., De Pree C. G., 2004, ApJS,155, 577Martin-Pintado J., de Vicente P., Fuente A., Planesas P.,1997, ApJ, 482, L45Mehringer D. M., 1995, ApJ, 454, 782Mehringer D. M., Menten K. M., 1997, ApJ, 474, 346Mehringer D. M., Goss W. M., Palmer P., 1994, ApJ, 434,237Mehringer D. M., Palmer P., Goss W. M., Yusef-Zadeh F.,1993, ApJ, 412, 684Miao Y., Mehringer D. M., Kuan Y.-J., Snyder L. E., 1995,ApJ, 445, L59Minh Y. C., Irvine W. M., Ziurys L. M., 1988, ApJ, 334,175Minh Y. C., Haikala L., Hjalmarson A., Irvine W. M., 1998,ApJ, 498, 261Nummelin A., Bergman P., Hjalmarson A., Friberg P.,Irvine W. M., Millar T. J., Ohishi M., Saito S., 1998,ApJS, 117, 427Nummelin A., Bergman P., Hjalmarson ˚A., Friberg P.,Irvine W. M., Millar T. J., Ohishi M., Saito S., 2000,ApJS, 128, 213Ohishi M., Ishikawa S.-I., Amano T., Oka H., Irvine W. M.,Dickens J. E., Ziurys L. M., Apponi A. J., 1996, ApJ, 471,L61Oka T., Hasegawa T., Sato F., Tsuboi M., Miyazaki A.,1998, ApJS, 118, 455Palmer P., Goss W. M., 1996, GCNew, 2, 3Piddington J. H., Minnett H. C., 1951, AuSRA, 4, 459Pierce-Price D., et al., 2000, ApJ, 545, L121Price S. D., Egan M. P., Carey S. J., Mizuno D. R., KucharT. A., 2001, AJ, 121, 2819Reid M. J., 1993, ARA&A, 31, 345Reid M. J., Schneps M. H., Moran J. M., Gwinn C. R.,Genzel R., Downes D., Roennaeng B., 1988, ApJ, 330,809Requena-Torres M. A., Mart´ın-Pintado J., Rodr´ıguez-Franco A., Mart´ın S., Rodr´ıguez-Fern´andez N. J., de Vi-cente P., 2006, A&A, 455, 971Sato F., Hasegawa T., Whiteoak J. B., Miyawaki R., 2000,ApJ, 535, 857Sawada T., Hasegawa T., Handa T., Cohen R. J., 2004,MNRAS, 349, 1167Snyder L. E., Kuan Y.-J., Miao Y., 1994, in The Structureand Content of Molecular Clouds, ed. T. L. Wilson, K. J.Johnston (Berlin: Springer), 187Turner B. E., 1989, ApJS, 70, 539Tsuboi M., Handa T., Ukita N., 1999, ApJS, 120, 1Vrtilek J. M., Gottlieb C. A., Thaddeus P., 1987, ApJ, 314,716Wilson T. L., Snyder L. E., Comoretto G., Jewell P. R.,Henkel C., 1996, A&A, 314, 909Yusef-Zadeh F., Uchida K. I., Mehringer D. M., Roberts D., Nyman L.-A., Casement S., Lindqvist M., 1996, ASPC,102, 60This paper has been typeset from a TEX/ L A TEX file preparedby the author. c (cid:13)000