THE ASTROPHYSICAL JOURNAL, 471 : L61–L64, 1996 November 1
q 1996. The American Astronomical Society. All rights reserved. Printed in U.S. A.
DETECTION OF A NEW INTERSTELLAR MOLECULAR ION, H 2 COH 1 (PROTONATED FORMALDEHYDE)
MASATOSHI OHISHI
AND
SHIN-ICHI ISHIKAWA
Nobeyama Radio Observatory,1 Nobeyama, Minamimaki, Mimanisaku, Nagano 384-13, Japan
TAKAYOSHI AMANO
AND
HIDEHIKO OKA
Institute for Astrophysics and Planetary Science, Ibaraki University, 2-1-1 Bunkyo, Mito, Ibaraki 310, Japan
WILLIAM M. IRVINE
AND
JAMES E. DICKENS
Five College Radio Astronomy Observatory, 619 Lederle GRC, University of Massachusetts, Amherst, MA 01003
AND
L. M. ZIURYS
AND
A. J. APPONI
Department of Chemistry, Arizona State University, Box 871604, Tempe, AZ 85287-1604
Received 1996 June 7; accepted 1996 August 21
ABSTRACT
A new interstellar molecular ion, H 2 COH (protonated formaldehyde), has been detected toward Sgr B2,
Orion KL, W51, and possibly in NGC 7538 and DR21(OH). Six transitions were detected in Sgr B2(M). The
1 1,0 –1 0,1 transition was detected in all sources listed above. Searches were also made toward the cold, dark clouds
TMC-1 and L134N, Orion (3N, 1E), and a red giant, IRC 110216, without success. The excitation temperatures
of H 2 COH 1 are calculated to be 60 –110 K, and the column densities are on the order of 10 12 –10 14 cm 22 in Sgr
B2, Orion KL, and W51. The fractional abundance of H 2 COH 1 is on the order of 10 211 to 10 29 , and the ratio
of H 2 COH 1 to H 2 CO is in the range 0.001– 0.5 in these objects. The values in Orion KL seem to be consistent
with the “early time” values of recent model calculations by Lee, Bettens, & Herbst, but they appear to be higher
than the model values in Sgr B2 and W51 even if we take the large uncertainties of column densities of H 2 CO
into account. We suggest production routes starting from CH 3 OH may play an important role in the formation
of H 2 COH 1 .
Subject headings: ISM: abundances — ISM: molecules
1
of the Nobeyama Radio Observatory as a part of the molecular
spectral survey toward Sgr B2 (Ohishi et al. 1996). We used
two SIS receivers covering the 7 mm and 3 mm regions, with
system temperatures 1200 and 1300 K, respectively. The back
end was a set of acousto-optical spectrometers with a frequency resolution of 250 kHz. The pointing was checked by
observing the SiO maser in VX Sgr. The beam sizes and
main-beam efficiencies are summarized in Table 1.
Additional observations were performed using the NRAO2
12 m telescope at Kitt Peak in 1995 November at 2 mm. The
front end was the dual-channel, single-sideband SIS receiver,
and the system temperatures were between 300 and 1100 K
depending on weather, elevation, and frequency. As back ends
we used both the filter-bank spectrometers and the hybrid
spectrometers. The pointing was checked by using the planets.
1. INTRODUCTION
Formaldehyde (H 2 CO) was the first interstellar organic
molecule to be detected, and it has been found in a wide
variety of molecular clouds. H 2 CO is thought to form through
both gas-phase reactions and surface reactions on dust grains.
Most chemical models of the formation of interstellar molecules predict that H 3 CO 1 (protonated formaldehyde) plays an
important role in the depletion of H 2 CO. It has been shown
that the lowest energy isomer of H 3 CO 1 has the proton
attached to the oxygen atom, i.e., H 2 COH 1 (Nobes, Radom,
& Rodwell 1980 and references therein).
Laboratory experiments on H 2 COH 1 were made for the
first time using IR spectroscopy by Amano & Warner (1989).
Based on the IR measurement, Minh, Irvine, & McGonagle
(1993) tried to detect H 2 COH 1 in Orion KL, Orion (3N, 1E),
TMC-1, and L134N, but without success. In 1994, microwave
spectroscopy of H 2 COH 1 was carried out by Chomiak et al.
(1994), resulting in accurate transition frequencies for
searches in interstellar space (Table 1). In this Letter we
report the detection of H 2 COH 1 in several high-mass, starforming regions and discuss its chemical significance in order
to clarify the chemistry of H 2 COH 1 and H 2 CO.
3. RESULTS
At Nobeyama, we observed transitions of H 2 COH 1 in the
7 mm and 3 mm bands toward Sgr B2(M) and Sgr B2(N). We
detected two unblended lines, the 3 0,3 –2 1,2 transition at
31914.617 MHz and the 1 1,1 –2 0,2 transition at 36299.952 MHz,
and one blended line, the 4 0,4 –3 1,3 transition at
102065.846 MHz. These lines are shown in Figure 1. The last
transition is blended with the NH 2 CHO 5 1,5 – 4 1,4 line. The first
two lines have radial velocities between 60 and 67 km s 21 in
Sgr B2(M) and about 64 km s 21 in Sgr B2(N), which are typical
for molecular emission so far observed at these positions. We
2. OBSERVATIONS
The initial observations were made in 1994 May, 1994
November, and 1995 February using the 45 m radio telescope
1 Nobeyama Radio Observatory, National Astronomical Observatory of
Japan, is open for researchers in the field of astrophysics and astrochemistry.
2 The National Radio Astronomy Observatory is operated by Associated
Universities, Inc., under contract with the National Science Foundation.
L61
L62
OHISHI ET AL.
Vol. 471
TABLE 1
TRANSITION PARAMETERS
OF
H 2 COH 1 a
Frequency
(MHz)
J9K921 K911 –J 0K 021 K 011
Sb
Eu
(K)
Telescope
31914.617. . . . .
36299.952. . . . .
102065.846. . . . .
132219.699. . . . .
168401.143. . . . .
173766.877. . . . .
3 0,3 –2 1,2
1 1,1 –2 0,2
4 0,4 –3 1,3
2 1,1 –1 1,0
1 1,0 –1 0,1
5 0,5 –4 1,4
1.079
0.524
1.676
1.500
1.500
2.326
18.3
10.9
30.4
17.5
11.1
45.5
NRO
NRO
NRO
NRAO
NRAO
NRAO
Beam Size
(arcsec)
hB
(%)
55
50
18
45
35
34
70
70
55
78
66
66
NOTES.— a Chomiak et al. (1994).
b
Intrinsic line strength.
also observed but did not detect an a-type transition at
31062.844 MHz (3 1,2 –3 1,3 ), which is in the bandpass of the
3 0,3 –2 1,2 line.
The 12 m telescope was subsequently used to confirm the
existence of H 2 COH 1 via several 2 mm lines. We detected
three additional transitions, listed in Table 2, making the
detection of H 2 COH 1 in Sgr B2 secure. The 1 1,0 –1 0,1 transition at 168 GHz was observed in absorption, as was expected
from the Nobeyama results. This clearly indicates that
H 2 COH 1 exists in the cold clouds in front of Sgr B2. We note
that the 2 1,1 –1 1,0 line at 132 GHz is the only a-type transition
detected and the other detected lines are all b-type. This is
curious when we consider that the calculated values of the two
components of the dipole moment, m a 5 1.44 debye and
m b 5 1.77 debye, are comparable (P. Botchwina 1995, private
communication; T. Hirano 1995, private communication).
Along with confirming observations at Kitt Peak, we made a
survey of H 2 COH 1 via its 1 1,0 –1 0,1 transition. We detected
H 2 COH 1 in Orion KL, W51, and possibly in NGC 7538 and
DR21(OH) (Fig. 2). No H 2 COH 1 lines were detected in
Orion (3N, 1E), TMC-1, L134N, or IRC 110216. The ob-
served line parameters are summarized in Table 2. The line
parameters of H 2 COH 1 in Orion KL are typical for the
extended ridge or the compact ridge (cf. Johansson et al. 1984
and Blake et al. 1987), except for the weakest transition at
132 GHz. The radial velocity of H 2 COH 1 in W51 coincides
with that of the 52 km s 21 feature observed in CO (Mufson &
Liszt 1979) and HCO 1 (Cox et al. 1987), and indicates that the
H 2 COH 1 cloud is associated with a cool, low-density envelope
surrounding the dense molecular core in W51 MAIN.
4. DISCUSSION
The excitation temperatures, Tex , and the column densities
of H 2 COH 1 , N(H 2 COH 1), were derived from the observed
lines by adjusting the excitation temperature and the column
density by a least-squares fit (Ohishi et al. 1995). This procedure is similar to the well-known rotation diagram method
(e.g., Turner 1991), but is more general, since we do not
assume that the observed lines are optically thin. Intrinsic line
strength, S, is given in Table 1.
In Orion KL, we obtained Tex 1 110 K and N(H 2 COH 1 )
FIG. 1.—H 2 COH 1 spectra detected in Sgr B2(M). (a) 3 0,3 –2 1,2 , (b) 1 1,1 –2 0,2 , (c) 4 0,4 –3 1,3 , (d) 2 1,1 –1 1,0 , (e) 1 1,0 –1 0,1 , and ( f ) 5 0,5 – 4 1,4 .
No. 1, 1996
DETECTION OF A NEW INTERSTELLAR MOLECULAR ION
L63
FIG. 2.—Sample spectra of H 2 COH 1 1 1,0 –1 0,1 detected in Sgr B2(N), W51, Orion KL, NGC 7538, and DR21 (OH).
1 2.3 3 10 13 cm 22 . The excitation temperature is close to that
derived for CH 3 OH in the compact ridge (Blake et al. 1987),
indicating that H 2 COH 1 is in this source component. This is
reasonable in view of the fact that the column density of H 2 CO
in the compact ridge is an order of magnitude higher than that
in the extended ridge (Irvine, Goldsmith, & Hjalmarson 1987),
given the formation mechanisms for H 2 COH 1 starting from
H 2 CO described later. The column density of H 2 COH 1
corresponds to a fractional abundance relative to H 2 of
f (H 2 COH 1) 1 8 3 10 211 for N(H 2 ) 5 3 3 10 23 cm 22 (Blake et
al. 1987). If we use N(H 2 CO) 1 1.6 3 10 15 cm 22 (Mangum &
Wootten 1993) or 13.0 3 10 16 cm 22 (Irvine et al. 1987), we
obtain [H 2 COH 1 ]y[H 2 CO] 5 0.014 – 0.0008. This abundance
seems to be consistent with an unpublished predicted value at
t 5 104 –105 yr of f (H 2 COH 1) 5 (2–3) 3 10 211 (P. Caselli
1996, private communication; although she overestimated the
abundance of H 2 CO by 2 to 3 orders of magnitude compared
with the observed value) and with the new standard model
values at t 5 105 yr (Lee, Bettens, & Herbst 1996); i.e.,
f (H 2 COH 1) 1 6 3 10 211 and [H 2 COH 1 ]y[H 2 CO] 5 0.001–
0.002 (if we adopt the value in Irvine et al. 1987 as the H 2 CO
column density). The fractional abundance of H 2 COH 1 is
TABLE 2
OBSERVED LINE PARAMETERS
OF
H 2 COH 1
Source
J9K921 K911 –J 0K 021 K 011
TB (rms)
(mK)
VLSR
(km s 21)
Dv
(km s 21)
Resolution
(kHz)
Remarks
Sgr B2(M) . . . . .
3 0,3 –2 1,2
1 1,1 –2 0,2
4 0,4 –3 1,3
2 1,1 –1 0,1
1 1,0 –1 0,1
5 0,5 –4 1,4
3 0,3 –2 1,2
1 1,1 –2 0,2
1 1,0 –1 0,1
2 1,1 –1 0,1
1 1,0 –1 0,1
5 0,5 –4 1,4
2 1,1 –1 0,1
1 1,0 –1 0,1
5 0,5 –4 1,4
1 1,0 –1 0,1
1 1,0 –1 0,1
97 (14)
2123 (16)
398 (47)
55 (19)
2191 (21)
192 (30)
93 (14)
291 (14)
280 (15)
22 (5)
57 (8)
70 (29)
. . . (9)
39 (13)
61 (15)
45 (15)
42 (13)
66.4
64.3
59.6
56.0
67.3
57.9
64.0
62.2
73.2
10.9
8.4
8.9
...
49.7
51.8
257.9
23.0
13.4
15.1
10.0
6.3
10.3
8.8
8.1
13.8
18.9
11.7
3.7
3.5
...
9.6
10.0
2.0
3.7
250
250
250
1000
1000
1000
250
250
1000
1000
1000
1000
1000
1000
1000
250
1000
...
...
blended
...
...
...
two comp.
two comp.
two comp.
...
...
...
...
...
...
...
...
Sgr B2(N) . . . . . .
Orion KL . . . . . .
W51 . . . . . . . . . . . .
NGC 7538. . . . . .
DR21(OH). . . . .
NOTES.—Line parameters were derived by Gaussian fitting. T B 5 T*A /h B for NRO, and 5T*R /h B for Kitt Peak.
Negative results are as follows: the rms’s of the 1 1,0 –1 0,1 line were; 20, 20, 17, and 5 mK for Orion (3N, 1E), TMC-1,
L134N, and IRC 110216 respectively, and the rms of the 5 0,5 – 4 1,4 line for IRC 110216 was 9 mK.
L64
OHISHI ET AL.
much higher than that at chemical equilibrium (16 3 10 213),
suggesting that even in high-mass, star-forming regions the
chemistry is not equilibrated.
For Sgr B2, single-temperature fittings did not reproduce
the observed line intensities; no absorption lines were reproduced. Instead, two-temperature models gave better results,
i.e., we applied the above method to a warm core and to a cold
envelope separately, referring to, for example, Hüttemeister et
al. (1995). In these calculations, we assumed that the background continuum temperature at 36 GHz is 20 K based on
the report by Akabane et al. (1988) and estimated the temperature at 168 GHz using the flux distribution of Fn 22.1 . We
note that we did not include the contribution to the background continuum temperature due to the dust emission. We
also fixed the excitation temperature in the envelope, Tex
(envelope), to be 2.7 K because no emission or absorption was
observed at about 19 northwest of Sgr B2(M) where no
continuum background emission other than the cosmic background radiation exists.
Then, in Sgr B2(M), we obtained Tex (core) 1 65 K with
N(H 2 COH 1)(core) 1 1.5 3 10 14 cm 22 and N(H 2 COH 1)
(envelope) 1 3 3 10 14 cm 22 . Therefore, the total column
density of H 2 COH 1 toward Sgr B2(M) is 15 3 10 14 cm 22 ,
which corresponds to f (H 2 COH 1) 1 2 3 10 210 for N(H 2 ) 5
2.6 3 10 24 cm 22 (Lis & Goldsmith 1990). This value is close
to the calculated one, [H 2 COH 1 ]y[H 2 CO] 1 0.14 (Lee et
al. 1996), if we use N(H 2 CO) 5 3.5 3 10 15 cm 22 (Sutton et
al. 1991). This ratio is significantly higher than the value the
gas-phase model calculations predict, although the estimated column density of H 2 CO may be too low (Sutton et
al. 1991). In Sgr B2(N), we fixed the excitation temperatures
to be the same as those in Sgr B2(M), and calculated the
total column density of H 2 COH 1 to be almost equal to that
toward Sgr B2(M).
For the other sources, we did not detect a sufficient number
of lines of H 2 COH 1 to utilize the least-squares fitting. Hence,
we assumed that the energy-level population follows a Boltzmann distribution at a given excitation temperature, all lines
are optically thin, and the beam-filling factor is unity. For W51,
the lower limit of Tex (HCO 1) is given as 10 K and the upper
limit as 50 K (Cox et al. 1987). By considering a difference
of dipole moment between HCO 1 and H 2 COH 1 and the
physical condition in the H 2 COH 1 cloud, we assumed
Tex (H 2 COH 1) 5 10 –30 K, resulting in N(H 2 COH 1) 5
(7.3–3.0) 3 10 12 cm 22 . This value corresponds to
f (H 2 COH 1) 5 (4.6 –1.9) 3 10 29 if we use N(H 2 ) 5 1.6 3
10 21 cm 22 for the 52 km s 21 feature (Arnal & Goss 1985).
Consequently, the ratio [H 2 COH 1 ]y[H 2 CO] is 0.23– 0.56
for N(H 2 CO) 5 1.3 3 10 13 cm 22 (Arnal & Goss 1985).
Because N(H 2 CO) was estimated using the K-type doubling
transition at 6 cm, it may contain a large uncertainty.
For NGC 7538 and DR21(OH) the detection of H 2 COH 1 is
tentative. Therefore, we derive only the column densities.
McGonagle (1995) reports Tex (NS) 5 8 –18 K for NGC
7538 and 14 –34 K for DR21(OH), respectively. The dipole
moment of NS (1.81 debye) is very close to the b-dipole of
H 2 COH 1 , so we adopt these values as Tex (H 2 COH 1). As a
result, we get N(H 2 COH 1) 5 (3.2–5.0) 3 10 12 cm 22 for NGC
7538 and (3.8 –9.1) 3 10 12 cm 22 for DR21(OH).
The principal formation pathway for H 2 COH 1 has been
thought to be proton transfer, primarily from H 13 , HCO 1 , and
H 3 O 1 to H 2 CO (Lee et al. 1996; Herbst & Lee 1996). If we
recall that H 2 COH 1 is detected only in high-mass, starforming regions where CH 3 OH is quite abundant, and that the
fractional abundance of H 2 COH 1 is higher than the predicted
value from chemical models, the role of the following exothermic formation route to H 2 COH 1 must be checked:
CH3 OH 1 H 13 3 CH3 OH 12 1 H2 3 H2 COH1 1 2H2 . In the
model by Lee et al. 1996, they estimate a contribution of the
above reaction in the formation of H 2 COH 1 to be less than
0.5%. But the observed CH 3 OH abundance in the compact
ridge component, (1–10) 3 10 27 (Menten et al. 1988), is 2–3
orders of magnitude higher than the calculated value,
6.5 3 10 210 . This is also the case for Sgr B2(M); observed
CH 3 OH abundance is 2.4 3 10 28 (Sutton et al. 1991). If we
take these facts into account, we conclude that the proton
transfer reactions to CH 3 OH may play comparable roles in
forming H 2 COH 1 with those to H 2 CO.
We thank all the staff of the Nobeyama Radio Observatory
and the 12 m telescope of NRAO for their assistance with our
observations. We acknowledge Peter Botchwina and Tsuneo
Hirano for providing us the dipole moment values prior to
publication, and Ho-Hsin Lee and Eric Herbst for providing us
their results from their new chemical model. M. O. is grateful
to Kaori Fukuzawa and Yoshihiro Osamura for discussion of
the possible formation routes of H 2 COH 1 . T. A. acknowledges partial support by a grant-in-aid for scientific research
(A) from the Ministry of Education, Science, and Culture (No.
06403005). W. M. I. and J. E. D. acknowledge support from
NASA grant NAGW-434. L. M. Z. and A. J. A. acknowledge
support from NSF grant AST-95-03274. We are grateful to an
anonymous referee for her/his critical reading of the manuscript.
REFERENCES
Akabane, K., Sofue, Y., Hirabayashi, H., Morimoto, M., & Inoue, M. 1988,
PASJ, 40, 809
Amano, T., & Warner, H. E. 1989, ApJ, 342, L99
Arnal, E. M., & Goss, W. M. 1985, A&A, 145, 369
Blake, G. A., Sutton, E. C., Masson, C. R., & Phillips, T. G. 1987, ApJ, 315, 621
Chomiak, D., Taleb-Bendiab, A., Civis, S., & Amano, T. 1994, Canadian J.
Phys., 72, 1078
Cox, M. J., Scott, P. F., Andersson, M., & Russel, A. P. G. 1987, MNRAS, 226,
703
Herbst, E., & Lee, H.-H. 1996, MNRAS, submitted
Hoffman, M. R., & Schaefer, H. F. 1981, ApJ, 249, 563
Hüttemeister, S., Wilson, T. L., Mauersberger, R., Lemme, C., Dahmen, G., &
Henkel, C. 1995, A&A, 294, 667
Irvine, W. M., Goldsmith, P. F., & Hjalmarson, Å. 1987, in Interstellar
Processes, ed. D. J. Hollenbach, & D. Thronson (Dordrecht: Reidel), 561
Johansson, L. E. B., et al. 1984, A&A, 130, 227
Lee, H.-H., Bettens, R. P. A., & Herbst, E. 1996, A&A, in press
Lis, D. C., & Goldsmith, P. F. 1990, ApJ, 356, 195
McGonagle, D. 1995, Ph.D. thesis, Univ. Massachusetts
Mangum, J. G., & Wootten, A. 1993, ApJS, 89, 123
Martı́n-Pintado, J., Wilson, T. L., Gardner, F. F., & Henken, C. 1985, A&A,
142, 131
Menten, K. M., Walmsley, C. M., Henkel, C., & Wilson, T. L. 1988, A&A, 198,
253
Minh, Y. C., Irvine, W. M., & McGonagle, D. 1993, J. Korean Astron. Soc., 26,
99
Mufson, S. L., & Liszt, H. S. 1979, ApJ, 232, 451
Nobes, R. H., Radom, L., & Rodwell, W. R. 1980, Chem. Phys. Lett., 74, 269
Ohishi, M., Ishikawa, S., Yamamoto, S., Saito, S., & Amano, T. 1995, ApJ, 446,
L43
Ohishi, M., Ishikawa, S., Noumaru, C., & Kaifu, N. 1996, in preparation
Sutton, E. C., Jaminet, P. A., Danchi, W. C., & Blake, G. A. 1991, ApJS, 77, 255
Turner, B. E. 1991, ApJS, 76, 617