High resolution image_of_a_cometary_globule_in_helix_nebula

The Astrophysical Journal, 573:L55–L58, 2002 July 1
2002. The American Astronomical Society. All rights reserved. Printed in U.S.A.

HIGH-RESOLUTION CO AND H2 MOLECULAR LINE IMAGING OF A COMETARY GLOBULE
IN THE HELIX NEBULA1
P. J. Huggins,2 T. Forveille,3,4 R. Bachiller,5 P. Cox,6 N. Ageorges,7 and J. R. Walsh8
Received 2002 April 14; accepted 2002 May 29; published 2002 June 7

ABSTRACT
We report high-resolution imaging of a prominent cometary globule in the Helix Nebula in the CO J p 1–0
(2.6 mm) and H2 v p 1–0 S(1) (2.12 mm) lines. The observations confirm that globules consist of dense condensations
of molecular gas embedded in the ionized nebula. The head of the globule is seen as a peak in the CO emission
with an extremely narrow line width (0.5 km s 1) and is outlined by a limb-brightened surface of H2 emission
facing the central star and lying within the photoionized halo. The emission from both molecular species extends
into the tail region. The presence of this extended molecular emission provides new constraints on the structure of
the tails and on the origin and evolution of the globules.
Subject headings: planetary nebulae: general — planetary nebulae: individual (NGC 7293) —
stars: AGB and post-AGB

1. INTRODUCTION lecular gas and the structure of the globules revealed by optical
images, we have made new observations of the Helix Nebula
The cometary globules in the Helix Nebula (NGC 7293) are in both CO and H2 with significantly better resolution than
among the most remarkable structures seen in planetary nebulae previous observations. In this Letter, we report results on the
(PNe). They occur in large numbers in the lower ionization molecular gas in a single cometary globule that resolve its head-
regions of the nebula and appear in high-resolution optical tail structure.
images as small (∼1 ), convex, photoionized surfaces facing
the central star, with comet-like tails extending in the opposite 2. OBSERVATIONS
direction (e.g., Meaburn et al. 1992; O’Dell & Handron 1996).
These structures are seen in other PNe (e.g., O’Dell et al. 2002) The globule observed is a prominent feature lying within
and are probably quite common, but they are best seen in the the ionized nebula to the north of the central star at offsets
Helix Nebula because it is the nearest example, at a distance ( 10 , 135 ). It is designated C1 by Huggins et al. (1992),
of D ∼ 200 pc (parallax p 4.70 0.75 mas; Harris et al. and its location is shown in Figure 3a of Meaburn et al. (1998),
1997). where it is labeled “1.” Optical images of the globule from
A key step in understanding the nature of the Helix globules Walsh & Meaburn (1993) in Ha [N ii] l6584 and in [O iii]
has been their detection in CO by Huggins et al. (1992). With l5007, where it is seen in absorption against the nebular emis-
a resolution of ∼12 , the CO observations did not resolve their sion, are shown in the top right panels of Figure 1.
structure but demonstrated that globules contain a major mass The CO observations were made in the 2.6 mm (115 GHz)
component of molecular gas, consistent with observations of J p 1–0 line using the IRAM interferometer at Plateau de
dust, seen in absorption against the nebula emission by Mea- Bure, France, in 1999 May. The array consisted of five 15 m
burn et al. (1992). The molecular gas places important con- antennas, equipped with SIS heterodyne receivers. The obser-
straints on the origin and evolution of the globules and provides vations were made with the D configuration of the array, with
a direct connection with the massive shell of neutral gas that maximum baselines of ∼147 m. The primary beam size of the
surrounds the ionized nebula (Forveille & Huggins 1991; interferometer is 44 at 2.6 mm, and the effective velocity
Young et al. 1999; Rodrıguez, Goss, & Williams 2002). This
´ resolution for the analysis is 0.2 km s 1. The u-v data were
connection is underscored by wide-field imaging of the nebula Fourier-transformed and CLEANed, using the Clark algorithm,
in the infrared lines of H2 by Kastner et al. (1996), Cox et al. and the restored Gaussian clean beam is 7.9 # 3.8 at a position
(1998), and Speck et al. (2002), at resolutions from 8 to 2 , angle of 14 . The results are shown in Figures 1 and 2.
that show a highly fragmented envelope. Images of the H2 v p 1–0 S(1) emission in the northern quad-
In order to determine the detailed relation between the mo- rant of the Helix Nebula were obtained with the SOFI infrared
camera on the ESO New Technology Telescope in 2001 June.
1
The instrument has a 10242 HAWAII array and was used with
Based on observations carried out with the IRAM Plateau de Bure Inter-
ferometer. IRAM is supported by INSU/CNRS (France), MPG (Germany), and
an image scale of 0.24 pixel 1. The seeing was 1.2. The ob-
IGN (Spain). servations were made with an H2 filter of width 0.028 mm,
2
Physics Department, New York University, 4 Washington Place, New York, centered at 2.124 mm, well separated from the He i 2 1P0–2 1S
NY 10003. 2.058 mm and H i 2.166 mm lines, the latter having a transmission
3
Canada-France-Hawaii Telescope, P.O. Box 1597, Kamuela, HI 96743. of a few percent in the filter passband. Fifteen 1 minute exposures
4
Observatoire de Grenoble, B.P. 53X, 38041 Grenoble Cedex, France.
5
IGN Observatorio Astronomico Nacional, Apartado 1143, E-28800 Alcala
´ ´ were made, using sky offset positions at 5 north, 5 east, and
de Henares, Spain. 7 northeast. The individual images were sky-cleaned, flat-
6
Institut d’Astrophysique Spatiale, Universite de Paris XI, 91405 Orsay,
´ fielded, and combined with shift-and-add registration using the
France. brightest stellar images.
7
European Southern Observatory, Alonso de Cordova 3107, Vitacurea, Cas-
´
illa 19001, Santiago 19, Chile.
Many globules are detected in the full H2 image, and the
8
Space Telescope Coordinating Facility, ESO, Karl-Schwarzschild-Strasse region around globule C1 is shown in the top center left panel
2, D-85748 Garching bei Munchen, Germany.
¨ in Figure 1. An approximate calibration has been made by
L55

L56 CO AND H2 IN HELIX NEBULA Vol. 573

Fig. 1.—Observations of the Helix globule C1. Top row, from left to right: CO (1–0) integrated intensity, H2 v p 1–0 S(1) , Ha [N ii] l6584, and dust absorption
seen against the nebula emission in [O iii] l5007. Bottom row: CO (1–0) channel maps. The channels are 0.2 km s 1 wide and are centered at the velocities given
in the upper right of each panel. The contour intervals are 0.4 K km s 1 for the CO integrated intensity map and 0.9 K for the channel maps. The ellipse in the
top left panel shows the beam size for the CO observations. For each panel, the offsets are relative to a field center of R.A. p 22h29m37.78, decl. p s

20 48 02.00 (J2000.0), which is used for all images in this Letter.

comparing the smoothed, full image with the data of Speck et the deconvolved source size, assuming a Gaussian distribution,
al. (2002). The intensity of the emission at the head of the is ∼2 # 10 (1 corresponds to 3 # 10 15 cm at 200 pc). The
globule is ∼10 4 ergs s 1 cm 2 sr 1. Astrometry of the H2 and emission is marginally resolved in right ascension but is re-
optical images was carried out using stars in the USNO catalog, solved in declination and extends roughly along the head-tail
and the registration between the images is ∼0.1 rms. The ac- axis of the globule. The peak intensity is 2.6 K km s 1, and
curacy of the absolute positions for comparison with the CO the total flux is 166 K km s 1 arcsec2.
interferometry is ∼0.3–0.5. The CO velocity strip (Fig. 2) shows that to the south, toward
the head of the globule, the line width is extremely narrow,
3. PROPERTIES OF THE GLOBULE DV p 0.5 km s 1 (FWHM), and it broadens out to ∼0.8 km s 1
3.1. Overview farther north into the tail region. The CO radial velocity is pre-
cisely determined to be Vlsr p 27.9 km s 1 (the correction to
The observations presented in Figure 1 provide comple- Vhel is 2.9 km s 1), consistent with 28.7 2 km s 1 measured
mentary views of the Helix globule. The CO (1–0) line, with in [N ii] l6584 by Meaburn et al. (1998). The systemic velocity
an upper level of Eu p 5.5 K above the ground state, shows of the whole envelope of the Helix Nebula is 23 km s 1 (Young
the overall distribution and kinematics of the cool molecular et al. 1999), so the globule is blueshifted by ∼5 km s 1. For a
gas; the high-lying (Eu ∼ 7000 K) H2 v p 1–0 S(1) line traces
radial expansion velocity of the globule system of ∼20 km s 1
excited molecular gas; the image in [O iii] shows the distri-
(Young et al. 1999), the globule lies on the near side of the
bution of dust, seen in absorption against the nebular emission;
and the image in Ha [N ii] shows the photoionized surfaces nebula, on a radius vector from the star inclined to the line of
facing the central star. sight by ∼75 . The head-tail axis, assumed radial, is also seen
The observations immediately confirm the molecular nature at this angle.
of the globule and reveal some important details of its structure. The CO channel maps (Fig. 1) show the structure in the
molecular gas. There are two peaks, offset 3 north and 8 north
from the field center. The first is just north of the maximum
3.2. CO Structure and Kinematics
absorption in the [O iii] image (centered at 1.5 north), and we
The CO (1–0) velocity-integrated intensity map (Fig. 1) is identify this molecular emission with the head of the globule
seen to be extended with respect to the telescope beam, and whose photoionized surface toward the central star is seen in

No. 1, 2002 HUGGINS ET AL. L57

Fig. 3.—Close-up of the head of the globule. Contours of the emission in
H2 v p 1–0 S(1) (left) and Ha [N ii] (right) are superposed on the [O iii]
image. The H2 data are smoothed with a Gaussian 0.6 (FWHM). The contours
are 0.3, 0.5, 0.7, and 0.9 of the peak emission.

accord with the expectations of H2 excitation in a thin photo-
dissociation region (PDR) at the surface of the molecular gas.
The observed intensity of the line is probably consistent with
this interpretation (Natta & Hollenbach 1998; Speck et al.
2002), although detailed PDR models of the globules have not
Fig. 2.—CO (1–0) velocity strip map of the globule along the major axis.
The contour intervals are 0.5 K.
yet been developed. The possibility that the observed distri-
bution is caused by shocks is unlikely, in view of the small
the Ha [N ii] image. Substantial CO emission, however, ex- crossing time ( 400 yr) for shocks (vs 5 km s 1) able to
tends away from the head into the tail region, and the second excite the v p 1–0 S(1) H2 line (e.g., Burton, Hollenbach, &
CO peak lies close to a second, weak maximum in the dust Tielens 1992); shocks in the bulk of the molecular gas are ruled
absorption image (at 8 north). This extended CO emission, out by the absence of any disturbance in the CO emission with
together with the line broadening into the tail, accounts for the velocities larger than ∼0.5 km s 1.
offset of the peak in the CO integrated intensity map from the
head of the globule. 3.4. Physical Properties of the Molecular Gas
These observations demonstrate that the molecular gas in Using the CO (2–1) observations of the globule by Huggins
the cometary globule is not a compact spheroidal bullet but et al. (1992), we find the (2–1)/(1–0) flux ratio to be ∼2–3
has a substantial component extending into the tail region. [assuming significant (1–0) flux is not resolved out on the
shortest baselines], which suggests that the lines are at least
3.3. H2 Distribution and Excitation partly optically thin, with an excitation temperature ∼18–40 K
in the thin limit. For a representative value of 25 K and a CO
The distribution of the H2 emission in the globule is strikingly
different from that of CO (see Fig. 1). The strongest H2 emis- abundance of 3 # 10 4, the CO (1–0) flux gives a mass of
molecular gas in the globule of Mm 1 # 10 5 M,. The cor-
sion occurs at the face of the globule toward the central star.
There is little emission directly behind the globule, but ob- responding average density in a volume with projected dimen-
servable emission extends to large distances ( 24 ) along the sions of 2 # 10 , which includes the head and the tail seen
in CO, is n H 2 # 10 4 cm 3.
tail. Remarkably, the H2 distribution most closely follows that
of the ionized gas seen in Ha [N ii]. These values are consistent with a mass of ∼2 # 10 5 M,
The high spatial resolution of our observations and the well- and n H ∼ 4 # 10 5 cm 3 determined for the head of the globule
defined geometry of the globule provide a unique perspective by Meaburn et al. (1992) from the dust absorption seen in
on the question of the excitation of the H2 (Cox et al. 1998), [O iii], corrected to a distance of 200 pc. The typical mass of
which affects its observed distribution. The bright H2 emission photoionized gas at the surface of a globule, ∼10 9 M, (e.g.,
clearly arises in a thin surface layer in the molecular gas. Fig- O’Dell & Handron 1996), is negligible in comparison.
ure 3 shows a close-up of the globule head, comparing the H2
4. ORIGINS AND EVOLUTION
and Ha [N ii] emission with [O iii]. The H2 emission forms
a limb-brightened cap on the globule facing the central star The structure and kinematics of the molecular gas in the
and lies just inside (0.5–1 ) the halo of photoionized gas seen head and in the tail of the globule provide basic constraints on
in Ha [N ii]. In the tail region too (Fig. 1), the H2 is enhanced their origin and evolution.
near two Ha [N ii] peaks, at 8 north and 12 north on the One scenario for the origin of globules is that they form in
east side, which are directly illuminated by the star. More de- the atmosphere of the progenitor star and are carried out in the
tailed observations are needed to determine whether these are expanding circumstellar envelope (Dyson et al. 1989). In this
separate, small globules along the line of sight or substructures case, the windswept appearance of the globules suggests a model
of the main tail. in which material from the head is swept into the tail by a radially
The observed distribution of the H2 emission is in complete directed wind, and Dyson, Hartquist, & Biro (1993) have shown

L58 CO AND H2 IN HELIX NEBULA Vol. 573

that the wind needs to be subsonic to form a narrow tail. Our occur, and the consequent reduction of the CO and H2 photo-
observations constrain this model in showing no evidence for a dissociation rates in the shadowed regions must play a key role
windswept flow pattern at the present time. There is no difference in preserving the molecular gas in the extended tails.
in velocity between the CO emission in the head and the tail of If globules originate close to the star as proposed by Dyson
the globule, and from the CO strip map (Fig. 2) the differential et al. (1989), the most crucial phase in their evolution occurs
motion is 0.2 km s 1 along the line of sight, or 0.8 km s 1 when they are overrun by the ionization front of the nebula.
in a radial direction, corrected for the inclination. This would At this stage, the shadowing of preexisting molecules or hy-
produce a tail of length 4 in 5000 yr, which is likely the drodynamic flows, or both, could lead to globules with mo-
maximum time available for such a process. The tails could have lecular tails. An alternative scenario is one in which the glob-
been fully formed at an earlier epoch, or it might be that ablation ules and their tails originate simultaneously in instabilities at
occurs only from the ionized edges of the head of the globule, the ionization front (e.g., Capriotti 1973). In this model, shad-
but this would not account for the presence of molecular gas owing and possibly hydrodynamic flows would also likely play
seen in H2 along the whole length of the tail. a role, and the fingers of gas that result from the instability
A different view of the gasdynamics around the globule at could contribute directly to the formation of the tails. The later
the present time is provided by the photoevaporation model, development of these two models will likely be similar and
which has been discussed in the context of the Helix globules can plausibly lead to the globules with substantial molecular
by Lopez-Martın et al. (2001). In this model, photoionization
´ ´ tails that we now observe. Further studies should allow us to
of the neutral globule, whose molecular core is unambiguously discriminate between them.
confirmed by our observations, produces an ionized outflow As for the fate of the globules, Meaburn et al. (1998) esti-
from the surface. Given that the density of the ionized gas at mated an ablation rate for globule C1 of 3 # 10 8 M, yr 1
the head of the globule is ∼103 cm 3, and that of the ambient based on densities inferred from the [O iii] image and an as-
gas is much lower, ∼50 cm 3 (e.g., O’Dell & Handron 1996), sumed ablation flow velocity of 10 km s 1. This implies an
any subsonic flow of the ambient gas around the globule is improbably short lifetime of ∼500 yr, which led Meaburn et
unlikely to have important dynamical effects in shaping the al. to consider ad hoc, restricted flow models around the glob-
gas in the tail. ule. The absence of an observable ablation flow in the globule
An alternative mechanism for growing a tail on a preexisting reported here for CO ( 0.8 km s 1) completely resolves this
globule is by shadowing (Canto et al. 1998), which can form
´ problem by increasing the original estimate of the ablation
a tail behind the globule as the result of underpressure of the timescale by at least an order of magnitude. In fact, the current
gas in the shadowed region. For the simplest case of a neutral mass-loss rate of the globule is likely dominated by photo-
globule in ionized gas, this model produces only modest density evaporation (Lopez-Martın et al. 2001), and the corresponding
´ ´
increases in the tail and is unlikely to lead to the substantial timescale is ∼104 yr. Thus, the globules are long-lived and may
molecular tails revealed by the present observations. Striking even escape the nebula.
effects of shadowing of the ionizing radiation are, however,
seen in optical images of the Helix (e.g., Henry, Kwitter, &
Dufour 1999), where long, radial plumes appear as extensions We thank the staff of the Plateau de Bure interferometer for
to globules and their tails. Similar shadowing of the stellar making the observations. This work has been supported in part
radiation at wavelengths longer than the Lyman limit must also by NSF grant AST 99-86159 (to P. J. H.).
REFERENCES

Burton, M. G., Hollenbach, D. J., & Tielens, A. G. G. M. 1992, ApJ, 399, Lopez-Martın, L., Raga, A. C., Mellema, G., Henney, W. J., & Canto, J. 2001,
´ ´ ´
563 ApJ, 548, 288
Canto, J., Raga, A., Steffen, W., & Shapiro, P. R. 1998, ApJ, 502, 695
´ Meaburn, J., Clayton, C. A., Bryce, M., Walsh, J. R., Holloway, A. J., &
Capriotti, E. R. 1973, ApJ, 179, 495 Steffen, W. 1998, MNRAS, 294, 201
Cox, P., et al. 1998, ApJ, 495, L23 Meaburn, J., Walsh, J. R., Clegg, R. E. S., Walton, N. A., Taylor, D., & Berry,
Dyson, J. E., Hartquist, T. W., & Biro, S. 1993, MNRAS, 261, 430 D. S. 1992, MNRAS, 255, 177
Dyson, J. E., Hartquist, T. W., Pettini, M., & Smith, L. J. 1989, MNRAS, 241, Natta, A., & Hollenbach, D. 1998, A&A, 337, 517
625 O’Dell, C. R., Balick, B., Hajian, A. R., Henney, W. J., & Burkert, A. 2002,
Forveille, T., & Huggins, P. J. 1991, A&A, 248, 599 AJ, in press
Harris, H. C., Dahn, C. C., Monet, D. G., & Pier, J. R. 1997, in IAU Symp. O’Dell, C. R., & Handron, K. D. 1996, AJ, 111, 1630
180, Planetary Nebulae, ed. H. J. Habing & H. J. G. L. M. Lamers (Dor- Rodrıguez, L. F., Goss, W. M., & Williams, R. 2002, ApJ, in press
´
drecht: Kluwer), 40 Speck, A. K., Meixner, M., Fong, D., McCullough, P. R., Moser, D. E., &
Henry, R. B. C., Kwitter, K. B., & Dufour, R. J. 1999, ApJ, 517, 782 Ueta, T. 2002, AJ, 123, 346
Huggins, P. J., Bachiller, R., Cox, P., & Forveille, T. 1992, ApJ, 401, L43 Walsh, J. R., & Meaburn, J. 1993, Messenger, 73, 35
Kastner, J. H., Weintraub, D. A., Gatley, I., Merrill, K. M., & Probst, R. G. Young, K., Cox, P., Huggins, P. J., Forveille, T., & Bachiller, R. 1999, ApJ,
1996, ApJ, 462, 777 522, 387

High resolution image_of_a_cometary_globule_in_helix_nebula

More Related Content

What's hot

Similar to High resolution image_of_a_cometary_globule_in_helix_nebula

More from Sérgio Sacani

Recently uploaded

High resolution image_of_a_cometary_globule_in_helix_nebula