arXiv:astro-ph/0407154v1 7 Jul 2004
Complex molecules in the hot core of the low mass protostar
S. Bottinelli1,2, C. Ceccarelli1, B. Lefloch1, J. P. Williams2, A. Castets3, E. Caux4, S.
Cazaux5, S. Maret1, B. Parise4, A. G. G. M. Tielens6
We report the detection of complex molecules (HCOOCH3, HCOOH and
CH3CN), signposts of a “hot core” like region, toward the low mass, Class 0
source NGC1333-IRAS4A. This is the second low mass protostar where such
complex molecules have been searched for and reported, the other source being
IRAS16293–2422. It is therefore likely that compact (few tens of AUs) regions
of dense and warm gas, where the chemistry is dominated by the evaporation
of grain mantles, and where complex molecules are found, are common in low
mass Class 0 sources.Given that the chemical formation timescale is much
shorter than the gas hot core crossing time, it is not clear whether the reported
complex molecules are formed on the grain surfaces (first generation molecules)
or in the warm gas by reactions involving the evaporated mantle constituents
(second generation molecules). We do not find evidence for large differences in
the molecular abundances, normalized to the formaldehyde abundance, between
the two solar type protostars, suggesting perhaps a common origin.
1Laboratoire d’Astrophysique de l’Observatoire de Grenoble, BP 53, 38041 Grenoble, Cedex 9, France.
2Institute for Astronomy, University of Hawai‘i, 2680 Woodlawn Drive, Honolulu HI 96822, USA.
3Observatoire de Bordeaux, 2 Rue de l’Observatoire, BP 89, 33270 Floirac, France.
4Centre d’Etude Spatiale des Rayonnements, CNRS-UPS, 9 Avenue du Colonel Roche, BP 4346, 31028
Toulouse, Cedex 4, France.
5INA Osservatorio Astrofisico d’Arcetri, 1 Ple Aldo Moro, Florence, Italy.
6Kapteyn Astronomical Institute, BO Box 800, 9700 AV Groningen, The Netherlands.
– 2 –
Subject headings: ISM: abundances — ISM: individual (IRAS4A) — ISM: molecules
— stars: formation
There is strong support — from the composition of cometary and meteoritic materials
— for the notion that the solar nebula, from which the planets formed, passed through a
phase of warm, dense gas with a rich chemistry. While much observational effort has been
dedicated to the study of such hot cores around massive protostars, hot cores around low
mass protostars have received little attention. Only very recently has the first hot core around
a solar-type protostar been discovered towards the typical Class 0 source, IRAS16293–2422
(hereafter IRAS16293), exhibiting all characteristics of such regions: warm temperatures
(> 100 K) and high densities (> 107cm−3: Ceccarelli et al. 2000a), high abundances
of hydrides (CH3OH, H2CO, H2O: Ceccarelli et al. 2000a, b; Sch¨ oier et al. 2002), high
deuteration levels (> 10%: Ceccarelli et al. 1998, 2001; Parise et al. 2002; Roberts et al.
2002), and complex molecules (HCOOCH3, HCOOH, CH3OCH3, CH3CN, C2H5CN: Cazaux
et al. 2003). The definition of “hot core” used for massive protostars implies the presence of
a relatively large amount of warm and dense gas, along with a complex chemistry triggered
by the grain mantle evaporation (e.g. Walmsley et al. 1992). In order to make clear that hot
cores of low and high mass protostars are, however, substantially different in the involved
amount of material, we will use hereinafter the term “hot corino” to identify the warm
inner regions of the envelope surrounding the low mass protostars.
The chemical composition of the (massive) hot cores is thought to reflect a variety of
sequential processes (Wamsley et al. 1992; Charnley et al. 1992; Caselli et al. 1993; Charn-
ley 1995; Rodgers & Charnley 2001, 2003). Specifically, in the pre-collapse cold cloud phase,
simple molecules form on grain surfaces by hydrogenation of CO and other heavy elements
(notably examples are H2CO, CH3OH and H2S). Upon heating by a newly formed star,
these molecules, called “first generation” or “parent” molecules, evaporate into the gas and
undergo fast neutral-neutral and ion-neutral reactions producing complex organic molecules,
i.e. “second generation” or “daughter” molecules. The first part of this sequence, i.e. the
formation of fully hydrogenated molecules on the grain surfaces, has been demonstrated to
occur in low mass protostars too, for example by studies of multiply deuterated molecules:
formaldehyde (Ceccarelli et al. 1998, Bacmann et al. 2003), methanol (Parise et al. 2002,
2004) and sulfide (Vastel et al. 2003). Evaporation from grain mantles of these first genera-
tion species (in particular H2CO and CH3OH) has been observed in IRAS16293 (Ceccarelli
– 3 –
et al. 2000b, Sch¨ oier et al. 2002) and in about a dozen low mass protostars (Maret et
al. 2004). However, since the timescale necessary to convert first generation molecules into
complex, second generation molecules (around 104−105yr; e.g. Charnley et al. 1992, 2001)
is much longer than the transit time of the gas in the hot corinos (few 100 yr; e.g. Sch¨ oier
et al. 2002), the formation in the gas of second generation molecules seems improbable (e.g.
Sch¨ oier et al. 2002). The detection of a high abundance of complex molecules in the hot
core of IRAS16293 (Cazaux et al. 2003) has evidently been a challenge to the simple theo-
retical sequence described above. The key question has shifted from “Is a hot core present
in low mass protostars?” to “What is the origin of molecular complexity in these sources?”
In particular, there may well be chemical pathways to complex molecules involving grain
surface networks (e.g. Charnley 1995). In order to answer this question, more observations
in other low mass protostars are necessary. This will allow the development of a solid ob-
servational framework within which we might search for clues to the formation of second
generation molecules. As remarked in previous studies, the question is far from being aca-
demic, since the molecules in the hot corinos constitute the material which will eventually
form the proto-planetary disk and, possibly, the planets of the forming Sun-like star.
In this Letter we present the first results of a survey we are carrying out on the sample
of Class 0 sources studied by Maret et al. (2004). Here we report the detection of complex,
second generation molecules in NGC1333-IRAS4A (hereafter IRAS4A), a well known Class
0 protostar, and a target of several studies of molecular emission (e.g. Blake et al. 1995).
IRAS4A is part of the binary system IRAS 4, located in the NGC1333 reflection nebula, in
the Perseus cloud. It is separated by 31′′from the other component, IRAS 4B, and was itself
resolved into two components with a separation of 2′′, by Lay et al. (1995). The distance to
the NGC1333 cloud is uncertain (see e.g. Maret et al. 2002), but assuming a value of 220 pc
(derived by˘Cernis 1990, for consistency with previous work), IRAS4A has a luminosity of
6 L⊙and an envelope mass of 3.5 M⊙(Sandell et al. 1991). IRAS4A is associated with a
very highly collimated outflow, detected in CO, CS, and SiO (Blake et al. 1995, Lefloch et
al. 1998). Infall motion was detected by Di Francesco et al. (2001) and Choi et al. (1999)
with an estimated accretion rate of 1.1×10−4M⊙yr−1, an inner mass of 0.7 M⊙and an age
of ∼6500 yr (see also Maret et al. 2002).
2.Observations and results
The observations were carried out in June 2003 with the IRAM 30-meter telescope. The
position used for pointing was α(2000) = 03h29m10.s3 and δ(2000) = 31◦13′31′′. Based on
the observations of IRAS16293 by Cazaux et al. (2003), we targeted the following complex
– 4 –
molecules: methyl formate, HCOOCH3(A and E), formic acid, HCOOH, dimethyl ether,
CH3OCH3, methyl cyanide, CH3CN, and ethyl cyanide, C2H5CN. Different telescope settings
were used in order to include as many transitions as possible for each molecule. All lines
were observed with a low resolution, 1 MHz filter bank of 4 × 256 channels split between
different receivers, providing a velocity resolution of ∼ 3, 2, and 1 km s−1at 3, 2, and 1 mm,
respectively. Each receiver was simultaneously connected to a unit of the autocorrelator,
with spectral resolutions of 20, 80 or 320 kHz and bandwidths between 40 and 240 MHz,
equivalent to a (unsmoothed) velocity resolution range of 0.1–0.4 km s−1. Typical system
temperatures were 100–200 K, 180–250 K and 500–1500 K, at 3, 2 and 1 mm respectively.
Two observation modes were used: position switching with the OFF position at an offset
of ∆α = –100′′, ∆δ = +300′′, and wobbler switching with a 110′′throw in azimuth. Point-
ing and focus were regularly checked using planets or strong quasars, providing a pointing
accuracy of 3′′. All intensities reported in this paper are expressed in units of main-beam
brightness temperature. At 3, 2 and 1 mm, the angular resolution is 24, 16 and 10′′and the
main beam efficiency is 76, 69 and 50%, respectively.
Fig. 1 shows two examples of low resolution spectra we obtained. Detected transitions
have been identified using the JPL molecular line catalog (Pickett et al. 1998) and are
reported in Table 1. We considered as good identifications only lines with a 3-σ detection
and a VLSR=6.8±0.3 km s−1. We detected three of the five targeted molecules: 10 transitions
for HCOOCH3(A and E), 2 for HCOOH and 9 for CH3CN. We also have a possible detection
for C2H5OH at 90.118 GHz; unfortunately, no other transition with a low enough energy and
high enough Einstein coefficient was contained within the frequency ranges we observed to
confirm the correct identification. No transitions of CH3OCH3and C2H5CN were detected
to a noise limit of 6 and 2 mK respectively. All detected lines have linewidths ∼ 2−3 km s−1,
with few exceptions, likely due to the presence of unresolved triplets or to the contamination
of unidentified lines. In order to derive the rotational temperature and column density,
we built rotational diagrams (Fig. 2) in which the observed fluxes were corrected for beam
dilution, assuming a source size of 0.′′5 (derived from a hot core radius of 53 AU and a
distance of 220 pc, as found by Maret et al. 2004). The assumption that the complex
molecules are confined to the hot corino is supported by a Plateau de Bure interferometric
study of IRAS16293 which shows localized emission in a region ∼ 1.′′4 around the protostar
(Bottinelli et al. in preparation).
The rotational temperatures, total column densities and abundances for the dectected
molecules are presented in Table 2. Note that the large errors in the HCOOH (this work) and
CH3OCH3(Cazaux et al. 2003) abundances are due to a poor constraint of the rotational
temperature, and hence column density, of these two molecules, even though each molecule
– 5 –
is clearly detected in each case.
3.Discussion and conclusion
The most important result of the present work is the detection of complex molecules in
the hot corino of IRAS4A, the second Class 0 in which those molecules have been searched
for, after IRAS16293 (Cazaux et al. 2003). This result demonstrates that as soon as a warm
region is created in the center of the envelope of low mass protostars, complex molecules are
readily formed and/or injected on timescales lower than the estimated Class 0 source ages
(∼ 5 × 104yr in IRAS16293 and ∼ 6500 yr in IRAS4A; e.g. Maret et al. 2002), and, most
importantly, shorter than the transit time in the hot corinos. The latter is ∼ 400 yr and
∼ 120 yr in IRAS16293 and IRAS4A respectively, based on the hot corino sizes quoted in
Maret et al. (2004) and assuming free-falling gas.
We compare the measured composition of the hot corino of IRAS4A to IRAS16293
(Cazaux et al. 2003) and the massive hot core of OMC-1 (Sutton et al. 1995) in Table 2.
Note that the latter abundances are derived from single dish measurements with a 14′′
beam, which encompasses several hot cores (Wright et al. 1996). Unfortunately, not all the
molecules considered here have interferometric measurements available, so that we can only
use these 14′′beam-averaged estimates of the abundances.
The first remark is that the absolute abundances of the observed molecules are one order
of magnitude smaller in IRAS4A than in IRAS16293, but their relative abundances with
respect to H2CO are quite similar, with the exception of methanol, which is underabundant
with respect to H2CO by about a factor 10 in IRAS4A (Fig. 3). There are two reasons
to consider abundances with respect to formaldehyde. The first one is observational: while
the IRAS16293 hot core has now been imaged with the Plateau de Bure interferometer
(Bottinelli et al. in preparation) and its size confirmed to be ∼ 1.′′4, the IRAS4A core size is
only indirectly estimated from dust continuum single dish (12′′) observations to be 0.′′5 and
no interferometric observations are available yet with such a high resolution. So the IRAS4A
core size might be wrong by up to a factor three (Maret et al. 2004) and the abundances
by up to a factor ten, i.e. the absolute abundances of IRAS4A could be comparable to
those of IRAS16293. Using abundance ratios allow us to remove this size uncertainty. The
second reason is theoretical: “standard” hot core models predict that molecules like methyl
formate or methyl cyanide are second generation molecules formed in the warm gas from the
evaporated grain mantle constituents (formaldehyde, ammonia and methanol: e.g. Charnley
et al. 1992; Caselli et al. 1993; Rodgers & Charnley 2003). It is therefore interesting to
compare the abundances of the complex molecules to those of one of these supposed parent
– 6 –
molecules. Formaldehyde was chosen because we only have an upper limit on the methanol
abundance (Maret et al. in preparation) and no measurements of the ammonia abundance
A possible interpretation for the similarity in the complex molecules’ relative abun-
dances, with respect to H2CO and not with respect to CH3OH, is that the former is the
mother molecule of the observed O-bearing species, e.g. likely the case of HCOOCH3(Charn-
ley et al. 1992), and that the chemical evolution timescale is shorter than the age of the
youngest source. Charnley et al. (1992) also predict that methanol is the mother molecule
of CH3OCH3, but we cannot say whether the available data confirm this hypothesis since
we only have an upper limit on the abundance of this molecule in IRAS4A and a large er-
ror in IRAS16293. Similarly, the N-bearing molecules CH3CN and C2H5CN could both be
daughters of the same mother molecule, probably ammonia. This would imply that the two
sources have a similar ammonia mantle abundance. Alternatively, (some of ?) the reported
molecules are possible mantle constituents themselves. This may be the case for formic acid,
as predicted by Tielens & Hagen (1982), and suggested by the observational study by Liu
et al. (2001). Moreover, the analysis of ISO absorption spectra towards the massive hot
core W33A (e.g. Schutte et al. 1997) is consistent with the presence of solid formic acid
and would also support the idea of this species being a mantle constituent. However, these
considerations do not take into account the evolutionary state of the objects and the funda-
mental question is: does the abundance of any of these complex molecules have anything to
do with the age and/or evolutionary stage of the protostar, or is it dominated by the initial
mantle composition? Evidently, two sources are not enough to answer this question, and
observations of more low mass sources are required.
Regarding the comparison with the massive hot core(s) in Orion, Fig. 3 would suggest
that, with respect to formaldehyde, there is a deficiency of methanol and of N-bearing
complex molecules in the low mass hot corinos. It is possible that these differences are
mostly due to a different grain mantle composition, i.e. to a different pre-collapse density.
However, recall that the abundance ratios of CH3CN and CH3OH in Figure 3 refer to the
14′′beam-averaged values around the OMC-1 hot core, which in fact includes several smaller
cores (Wright et al. 1996). Therefore, in order to make precise comparisons, higher resolution
observations of the OMC-1 hot core are needed. It is also worth noting that if we consider
for example the measurements by Wright et al. 1996 in the Compact Ridge component (a
region about 10′′away from the hot core central position, which is also a site of mantle
evaporation and of active gas phase chemistry; e.g. Charnley et al. 1992), the CH3CN and
CH3OH abundance ratios with respect to H2CO are (surprisingly) close to those found for
the hot corinos of IRAS16293 and IRAS4A. Hence, interferometric observations of a larger
number of massive hot cores are necessary to provide a significant comparison of the hot
– 7 –
corinos with their high mass counterparts.
In summary, although the present observations do not allow us to answer the questions
why and how complex molecules are formed, they do show that hot corinos, in the wide
definition of chemically enriched regions, are a common property of solar-type protostars in
the early stages. The evidence is that the types of complex molecules that are formed are
determined primarily by the composition of the grain mantles. At this stage, it is not clear
whether the evolutionary stage of the protostar plays any role at all, other than governing
the presence and size of the mantle evaporation region.
Bacmann, A., Lefloch, B., Ceccarelli, C., Steinacker, J., Castets, A., & Loinard, L. 2003,
ApJ, 585, L55
Blake, G. A., Sandell, G., van Dishoeck, E. F., Groesbeck, T. D., Mundy L. G., & Aspin, C.
1995, ApJ, 441, 689
Caselli, P., Hasegawa, T. I., & Herbst, Eric 1993, ApJ, 408, 548
Cazaux, S., Tielens, A. G. G. M., Ceccarelli, C., Castets, C. Wakelam, V., Caux, E., Parise,
B., & Teyssier, D. 2003, ApJ, 593, L51
Ceccarelli, C., Castets, A., Caux, E., Hollenbach, D., Loinard, L., Molinari, S., & Tielens,
A. G. G. M. 2000a, A&A, 355, 1129
Ceccarelli, C., Castets, A., Loinard, L., Caux, E., & Tielens, A. G. G. M. 1998, A&A, 338,
Ceccarelli, C., Loinard, L., Castets, A., Tielens, A. G. G. M., Caux, E., Lefloch, B., & Vastel,
C. 2001, A&A, 372, 998
Ceccarelli, C., Loinard, L., Castets, A., Tielens, A. G. G. M., & Caux, E. 2000b, A&A, 357,
˘Cernis, K. 1990, Ap&SS, 166, 315
Charnley, S. B. 1995, Ap&SS, 224, 251
Charnley, S. B., Rodgers, S. D. & Ehrenfreund, P. 2001, A&A, 378, 1024
Charnley, S. B., Tielens, A. G. G. M., & Millar, T. J. 1992, ApJ, 399, L71
Choi, M., Panis, J.-F., & Evans, N. J. 1999, ApJS, 122, 519
Di Francesco, J., Myers, P. C., Wilner, D. J., Ohashi, N., & Mardones, D. 2001, ApJ, 562,
Lay, O. P., Carlstrom, J. E., & Hills, R. E. 1995, ApJ, 452, L73
Lefloch, B., Castets, A., Cernicharo, J., & Loinard, L. 1998, ApJ, 504, L109
Liu, S.-Y., Mehringer, D. M., & Lewis, E. S. 2001, ApJ, 552, 654
– 8 –
Maret, S., Ceccarelli, C., Caux, E., Tielens, A. G. G. M., Jørgensen, J. K., van Dishoeck,
E., Bacmann, A., Castets, A., Lefloch, B., Loinard, L., Parise, B., & Sch¨ oier, F. L. 2004,
A&A 416, 577
Maret, S., Ceccarelli, C., Caux, E., Tielens, A. G. G. M. & Castets, A. 2002, A&A, 395, 573
Parise, B., Castets, A., Herbst, E., Caux, E., Ceccarelli, C., Mukhopadhyay, I., Tielens,
A.G.G.M. 2004, A&A 416, 159
Parise, B., Ceccarelli, C., Tielens, A. G. G. M., Herbst, E., Lefloch, B., Caux, E., Castets,
A., Mukhopadhyay, I., Pagani, L., & Loinard, L. 2002, A&A, 393, L49
Pickett, H. M., Poynter, R. L., Cohen, E. A., Delitsky, M. L., Pearson, J. C., & Muller, H. S.
P., 1998 “Submillimeter, Millimeter, and Microwave Spectral Line Catalog”, J. Quant.
Spectrosc. & Rad. Transfer 60, 883
Roberts, H., Fuller, G. A., Millar, T. J., Hatchell, J, & Breckle, J. V. 2002, A&A, 381, 1026
Rodgers, S. D., & Charnley, S. B. 2003, ApJ, 585, 355
Rodgers, S. D., & Charnley, S. B. 2001, ApJ, 546, 324
Sandell, G., Aspin, C., Duncan, W. D., Russell, A. P. G., & Robson, E. I. 1991, ApJ, 376,
Sch¨ oier, F. L., Jørgensen, J. K., van Dischoeck, E. F., & Blake, G. A. 2002, A&A, 390, 1001
Schutte, W. A., Greenberg, J. M., van Dishoeck, E. F., Tielens, A. G. G. M., Boogert, A.
C. A., & Whittet, D. C. B. 1997, Ap&SS 255, 61
Sutton, E. C. Peng, R., Danchi, W. C., Jaminet, P. A., Sandell, G., & Russell, A. P. G.
1995, ApJ, 97, 455
Tielens, A. G. G. M. & Hagen, W. 1982, A&A, 114, 245
Vastel, C., Phillips, T. G., Ceccarelli, C., & Pearson, J. 2003, ApJ, 593, L97
Walmsley, C. M., Cesaroni, R., Churchwell, E., & Hofner, P. 1992, Astron. Gesellschaft
Abstract Ser., 7, 93
Wright, M. C. H., Plambeck, R. L., & Wilner, D. J. 1996, ApJ, 469, 216
– 9 –
Fig. 1.— Two low resolution spectra obtained during our observations of IRAS4A. Lines
which are not labelled are unidentified. The rms noise level is 2 mK (top spectrum)
and 12 mK (bottom spectrum). The spectral resolution is 3.3 km s−1(top) and 1.2
km s−1(bottom). The VLSR is 7.0 km s−1. Known transitions are indicated but not all
of them are detections, e.g. HCOOCH3at 90.145 GHz is not considered as such, but the
upper limit derived from it is consistent with the rotational diagram of Fig. 2.
– 10 –
Fig. 2.— Rotational diagrams of the detected molecules, corrected for beam dilution. The
arrows show the upper limits for the transitions that have not been detected. Lines represent
the best fit to the data. Error bars are derived assuming a calibration uncertainty of 10%
on top of the statistical error. The excess of emission of the CH3CN transition at 210 K is
probably due to contamination from unknown line(s).
– 11 –
Fig. 3.— The abundances of the observed species (reported on the x-axis) normalized to
the H2CO abundances. Stars refer to the OMC-1 hot core, squares to the hot corino of
IRAS16293 and diamonds to the one of IRAS4A. Arrows represent upper limits in IRAS4A
derived from our observations. No errors were quoted by Cazaux et al. (2003) for the
HCOOH abundance, which was determined from two transitions only and is rather uncertain.
– 12 –
Table 1. Molecular lines detected toward IRAS4A.
Molecule Transition lineFrequency
(K km s−1)
(cm−1) (km s−1) (km s−1)
aSpectral resolution of the observation (when possible, the integrated intensity was derived from the high resolution data).
bWidth of the observed line.
crms computed over the linewidth.
dAll the CH3CN lines are (unresolved) triplets. The quoted signal is the integral over each triplet. Larger linewidths could
be due to the larger spacing between the components of the triplets.
– 13 – Download full-text
Table 2. Results from the rotational diagrams for IRAS4A, in comparison with
IRAS16293 and the massive hot core OMC-1.
5.5 ± 2.7(16)
5.8 ± 1.1(16)
2.6 ± 0.3(15)
3.4 ± 1.7(–8)
3.6 ± 0.7(–8)
4.6 ± 7.9(–9)
1.6 ± 0.2(–9)
1.7 ± 0.7 (–7)
2.3 ± 0.8 (–7)
1.0 ± 0.4 (–8)
36 ± 5
10 ± 6
27 ± 1
2.4 ± 3.7(–7)
1.2 ± 0.4(–8)
aAssuming an H2 column density in the hot corino of N(H2) = 1.6 × 1024cm−2(From
Maret et al. 2004).
bFrom Cazaux et al. 2003.
cFrom Sutton et al. 1995.
dTrot assumed to be similar to the one derived for HCOOCH3-E.
eTrot assumed to be similar to the one derived for CH3CN.
fFrom Maret et al. 2004, in prep.
gFrom Maret et al. 2004.
hFrom Ceccarelli et al. 2000b.