Astrochemistry: From Molecular Clouds to Planetary Systems

IAU Symposium, Vol. 197, 2000

Y. C. Minh and E.F. van Dishoeck, eds.

Chemistry in the Envelopes around Massive Young Stars

Ewine F. van Dishoeck and Floris F.S. van der Tak

Leiden Observatory, P.O. Box 9513, 2300 RA Leiden, The Netherlands
arXiv:astro-ph/0002098v1 4 Feb 2000

Abstract. Recent observational studies of intermediate- and high-mass

star-forming regions at submillimeter and infrared wavelengths are re-
viewed, and chemical diagnostics of the different physical components
associated with young stellar objects are summarized. Procedures for
determining the temperature, density and abundance profiles in the en-
velopes are outlined. A detailed study of a set of infrared-bright mas-
sive young stars reveals systematic increases in the gas/solid ratios, the
abundances of evaporated molecules, and the fraction of heated ices with
increasing temperature. Since these diverse phenomena involve a range of
temperatures from < 100 K to 1000 K, the enhanced temperatures must
be communicated to both the inner and outer parts of the envelopes.
This ‘global heating’ plausibly results from the gradual dispersion of the
envelopes with time. Similarities and differences with low-mass YSOs are
discussed. The availability of accurate physical models will allow chemi-
cal models of ice evaporation followed by ‘hot core’ chemistry to be tested
in detail.

1. Introduction

Massive star-forming regions have traditionally been prime targets for astro-
chemistry owing to their bright molecular lines (e.g., Johansson et al. 1984,
Cummins et al. 1986, Irvine et al. 1987, Ohishi 1997). Massive young stellar
objects (YSOs) have luminosities of ∼ 104 − 106 L⊙ and involve young O- and
B-type stars. Because their formation proceeds more rapidly than that of low-
mass stars and involves ionizing radiation, substantial chemical differences may
be expected. The formation of high mass stars is much less well understood
than that of low-mass stars. For example, observational phenomena such as
ultracompact H II regions, hot cores, masers and outflows have not yet been
linked into a single evolutionary picture. Chemistry may well be an important
diagnostic tool in establishing such a sequence.
Most of the early work on massive star-forming regions has centered on two
sources, Orion–KL and SgrB2. Numerous line surveys at millimeter (e.g., Blake
et al. 1987, Turner 1991) and submillimeter (Jewell et al. 1989, Sutton et al.
1991, 1995, Schilke et al. 1997) wavelengths have led to an extensive inventory
of molecules through identification of thousands of lines. In addition, the surveys
have shown strong chemical variations between different sources.
In recent years, new observational tools have allowed a more detailed and
systematic study of the envelopes of massive YSOs. Submillimeter observations
1
2

routinely sample smaller beams (typically 15′′ vs. 30′′ –1′ ) and higher critical
densities (≥ 106 vs. 104 cm−3 ) than the earlier work. Moreover, interferometers
at 3 and 1 millimeter provide maps with resolutions of 0.5′′ –5′′ . Finally, ground-
and space-based infrared observations allow both the gas and the ices to be
sampled (e.g., Evans et al. 1991, van Dishoeck et al. 1999). These observational
developments have led to a revival of the study of massive star formation within
the last few years. Recent overviews of the physical aspects of high-mass star
formation are found in Churchwell (1999) and Garay & Lizano (1999).
In this brief review, we will first summarize available observational diag-
nostics to study the different phases and physical components associated with
massive star formation. Subsequently, an overview of recent results on inter-
mediate mass YSOs is given, which are often better characterized than their
high-mass counterparts because of their closer distance. Subsequently, we will
discuss a specific sample of embedded massive YSOs which have been studied
through a combination of infrared and submillimeter data. After illustrating the
modeling techniques, we address the question how the observed chemical vari-
ations are related to evolutionary effects, different conditions in the envelope
(e.g., T , mass) or different luminosities of the YSOs. More extensive overviews
of the chemical evolution of star-forming regions are given by van Dishoeck &
Blake (1998), Hartquist et al. (1998), van Dishoeck & Hogerheijde (1999) and
Langer et al. (2000). Schilke et al. (this volume) present high spatial resolution
interferometer studies, whereas Macdonald & Thompson (this volume) focus on
submillimeter data of hot core/ultracompact H II regions. Ices are discussed by
Ehrenfreund & Schutte (this volume).

2. Submillimeter and Infrared Diagnostics

The majority of molecules are detected at (sub-)millimeter wavelengths, and

line surveys highlight the large variations in chemical composition between dif-
ferent YSOs, both within the same parent molecular cloud and between different
clouds. The recent 1–3 mm surveys of Sgr B2 (Nummelin et al. 1998, Ohishi
& Kaifu 1999) dramatically illustrate the strong variations between various po-
sitions (see Figure 1). The North position is typical of ‘hot core’-type spectra,
which are rich in lines of saturated organic molecules. This position has also
been named the ‘large molecule heimat’ (e.g., Kuan & Snyder 1994, Liu & Snyder
1999). The Middle position has strong SO and SO2 lines, whereas the North-
west position has a less-crowded spectrum with lines of ions and long carbon
chains. A similar differentiation has been observed for three positions in the
W 3 giant molecular cloud by Helmich & van Dishoeck (1997), who suggested
an evolutionary sequence based on the chemistry.
The availability of complete infrared spectra from 2.4–200 µm with the
Infrared Space Observatory (ISO) allows complementary variations in infrared
features to be studied. Figure 2 shows an example of ISO–SWS and LWS spectra
of two objects: Cep A (L ≈ 2.4 × 104 L⊙ ) and S 106 (L ≈ 4.2 × 104 L⊙ ). The
Cep A spectrum is characteristic of the deeply embedded phase, in which the
silicates and ices in the cold envelope are seen in absorption. The S 106 spectrum
is typical of a more evolved massive YSO, with strong atomic and ionic lines
in emission and prominent PAH features. A similar sequence has been shown
 3

Figure 1. Overview of the SEST 230 GHz line survey of SgrB2 at

the N, M and NW positions, illustrating the chemical differentiation in
high-mass star-forming regions (from: Nummelin et al. 1998).

by Ehrenfreund et al. (1998) for a set of southern massive young stars with
luminosities up to 4 × 105 L⊙ .
The most successful models for explaining these different chemical charac-
teristics involve accretion of species in an icy mantle during the (pre-)collapse
phase, followed by grain-surface chemistry and evaporation of ices once the
YSO has started to heat its surroundings (e.g., Millar 1997). The evaporated
molecules subsequently drive a rapid high-temperature gas-phase chemistry for
a period of ∼ 104 − 105 yr, resulting in complex, saturated organic molecules
(e.g., Charnley et al. 1992, 1995; Charnley 1997; Caselli et al. 1993, Viti &
Williams 1999). The abundance ratios of species such as CH3 OCH3 /CH3 OH
and SO2 /H2 S show strong variations with time, and may be used as ‘chemi-
cal clocks’ for a period of 5000–30,000 yr since evaporation. Once most of the
envelope has cleared, the ultraviolet radiation can escape and forms a photon-
dominated region (PDR) at the surrounding cloud material, in which molecules
are dissociated into radicals (e.g., HCN → CN) and PAH molecules excited to
produce infrared emission. The (ultra-)compact H II region gives rise to strong
ionic lines due to photoionization.
4

Figure 2. ISO-SWS spectra of two massive YSOs at different evolu-

tionary stages. Cep A (L ≈ 2.4 × 104 L⊙ ) is in the embedded stage,
whereas S 106 (L ≈ 4.2 × 104 L⊙ ) is in a more evolved stage (from:
van den Ancker et al. 2000a).

Table 1 summarizes the chemical characteristics of the various physical com-

ponents, together with the observational diagnostics at submillimeter and in-
frared wavelengths. Within the single-dish submillimeter and ISO beams, many
of these components are blended together and interferometer observations will
be essential to disentangle them. Nevertheless, the single-dish data are useful
because they encompass the entire envelope and highlight the dominant compo-
nent in the beam. Combined with the above chemical scenario, one may then
attempt to establish an evolutionary sequence of the sources.
The physical distinction between the ‘hot core’ and the warm inner envelope
listed in Table 1 is currently not clear: does the ‘hot core’ represent a separate
physical component or is it simply the inner warm envelope at a different stage
of chemical evolution? Even from an observational point of view, there appear to
be different types of ‘hot cores’: some of them are internally heated by the young
star (e.g., W 3(H2 O)), whereas others may just be dense clumps of gas heated
externally (e.g., the Orion compact ridge). This point will be further discussed
in §§4 and 5. Disks are not included in Table 1, because little is known about
their chemical characteristics, or even their existence, around high-mass YSOs
(see Norris, this volume).
 5

Table 1. Chemical characteristics of massive star-forming regions

Component Chemical Submillimeter Infrared Examples

characteristics diagnostics diagnostics
Dense cloud Low-T chemistry Ions, long-chains Simple ices SgrB2 (NW)
(HC5 N, ...) (H2 O, CO2 )
Cold envelope Low-T chemistry, Simple species Ices N7538 IRS9,
Heavy depletions (CS, H2 CO) (H2 O, CO2 , CH3 OH) W 33A
Inner warm Evaporation High Tex High gas/solid, High GL 2591,
envelope (CH3 OH) Tex , Heated ices GL 2136
(C2 H2 , H2 O, CO2 )
Hot core High-T chemistry Complex organics Hydrides Orion hot core,
(CH3 OCH3 , CH3 CN, (OH, H2 O) SgrB2(N),G34.3
vib. excited mol.) W 3(H2 O)
Outflow: Shock chemistry, Si- and S-species Atomic lines, Hydrides W 3 IRS5,
Direct impact Sputtering (SiO, SO2 ) ([S I], H2 O) SgrB2(M)
PDR, Compact Photodissociation, Ions, radicals Ionic lines, PAHs S 140,
H II regions Photoionization (CN/HCN, CO+ ) ([NeII], [CII]) W 3 IRS4

3. Intermediate-Mass YSOs

Intermediate-mass pre-main sequence stars, in particular the so-called Herbig

Ae/Be stars, have received increased observational attention in recent years (see
Waters & Waelkens 1998 for a review). These stars have spectral type A or
B and show infrared excesses due to circumstellar dust. Typical luminosities
are in the range 102 − 104 L⊙ , and several objects have been located within 1
kpc distance. Systematic mapping of CO and the submillimeter continuum of a
sample of objects has been performed by Fuente et al. (1998) and Henning et al.
(1998). The data show the dispersion of the envelope with time starting from the
deeply embedded phase (e.g., LkHα234) to the intermediate stage of a PDR (e.g.,
HD 200775 illuminating the reflection nebula NGC 7023) to the more evolved
stage where the molecular gas has disappeared completely (e.g., HD 52721).
The increasing importance of photodissociation in the chemistry is probed by
the increase in the CN/HCN abundance ratio. This ratio has been shown in
other high-mass sources to be an excellent tracer of PDRs (e.g., Simon et al.
1997, Jansen et al. 1995). Line surveys of these objects in selected frequency
ranges would be useful to investigate their chemical complexity, especially in the
embedded phase.
ISO-SWS observations of a large sample of Herbig Ae/Be stars have been
performed by van den Ancker et al. (2000b,c). In the embedded phase, shock
indicators such as [S I] 25.2 µm are strong, whereas in the later phases PDR in-
dicators such as PAHs are prominent. An excellent example of this evolutionary
sequence is provided by three Herbig Ae stars in the BD+40o 4124 region (d ≈ 1
kpc). The data suggest that in the early phases, the heating of the envelope
is dominated by shocks, whereas in later phases it is controlled by ultraviolet
photons.
ISO-LWS data have been obtained for a similar sample of Herbig Ae/Be
stars by Lorenzetti et al. (1999) and Giannini et al. (2000), and are summarized
by Saraceno et al. (1999). The [C II] 158 µm and [O I] 63 and 145 µm lines
6

78 80 82 84

130 140 150 160

160 170 180 190

Figure 3. Left: ISO-LWS spectra of the intermediate mass YSO

LkHα234, showing strong atomic fine-structure lines of [C II] 158 µm
and [O I] 63 µm, but weak molecular CO and OH lines. Right: CO ex-
citation diagram, indicating the presence of a warm (T ≈ 200− 900 K),
dense (n > 105 cm−3 ) region (from: Giannini et al. 2000).
are prominent in many objects and are due primarily to the PDR component
in the large LWS beam (∼ 80′′ ). High-J CO and OH far-infrared lines have
been detected in some objects and indicate the presence of a compact, high
temperature and density region of ∼1000 AU in size, presumably tracing the
inner warm envelope (see Figure 3). Far-infrared lines of H2 O are seen in low-
mass YSO spectra, but are weak or absent in those of intermediate- and high-
mass YSOs, with the exception of Orion-KL and SgrB2 (e.g., Harwit et al. 1998,
Cernicharo et al. 1997, Wright et al. 2000). The absence of H2 O lines in higher-
mass objects may be partly due to the larger distance of these objects, resulting
in substantial dilution in the LWS beam. However, photodissociation of H2 O to
OH and O by the enhanced ultraviolet radiation may also play a role.
In summary, both the submillimeter and infrared diagnostics reveal an evo-
lutionary sequence from the youngest ‘Group I’ objects to ‘Group III’ objects
(cf. classification by Fuente et al. 1998), in which the envelope is gradually dis-
persed. Such a sequence is analogous to the transition from embedded Class
0/I objects to more evolved Class II/III objects in the case of low-mass stars
(Adams et al. 1987). The ISO data provide insight into the relative importance
of the heating and removal mechanisms of the envelope. At the early stages of
intermediate-mass star formation, shocks due to outflows appear to dominate
whereas at later stages radiation is more important.

4. Embedded, Infrared-Bright Massive YSOs

4.1. Sample
The availability of complete, high quality ISO spectra for a significant sample
of massive young stars provides a unique opportunity to study these sources
through a combination of infrared and submillimeter spectroscopy, and further
develop these diagnostics. Van der Tak et al. (2000a) have selected a set of ∼10
deeply embedded massive YSOs which are bright at mid-infrared wavelengths
(12 µm flux > 100 Jy), have luminosities of 103 − 2 × 105 L⊙ and distances d ≤4
kpc. The sources are all in an early evolutionary state (comparable to the ‘Class
 7
DUST GAS
luminosity α i , n0(i); Xj
dust opacities collisional rate
αi coefficients

self-consistent Monte Carlo excitation

Tkin = Tdust ; gas / dust = 100

radiative transfer + radiative transfer

Tdust (r) level populations

l.o.s. integration, l.o.s. integration,

beam convolution beam convolution,
velocity integration

synthetic SED synthetic line fluxes

Compare with Compare with data

submm data (χ2 statistic)

n0 for each α best-fit α, X

i

Figure 4. Overview of method used for constraining the tempera-

ture and density structure of envelopes. A power-law density structure
with trial values for the exponent αi and molecular abundances Xj is
adopted (based on: van der Tak et al. 2000a).

0/I’ or ‘Group I’ stages of low- and intermediate-mass stars), as indicated by

their weak radio continuum emission and absence of ionic lines and PAH features.
In addition to ISO spectra, JCMT submillimeter data and OVRO interferometer
observations have been obtained. For most of the objects high spectral resolution
ground-based infrared data of CO, 13 CO and H+ 3 are available (Mitchell et al.
1990, Geballe & Oka 1996, McCall et al. 1999), and occasionally H2 (Lacy et
al. 1994, Kulesa et al. 1999). For comparison, 5 infrared-weak sources with
similar luminosities are studied at submillimeter wavelengths only. This latter
set includes hot cores and ultracompact H II regions such as W 3(H2 O), IRAS
20126+4104 (Cesaroni et al. 1997, 1999), and NGC 6334 IRS1.

4.2. Physical structure of the envelope

In order to derive molecular abundances from the observations, a good physi-
cal model of the envelope is a prerequisite. Van der Tak et al. (1999, 2000a)
outline the techniques used to constrain the temperature and density structure
(Figure 4). The total mass within the beam is derived from submillimeter pho-
tometry, whereas the size scale of the envelope is constrained from line and
continuum maps. The dust opacity has been taken from Ossenkopf & Henning
(1994) and yields values for the mass which are consistent with those derived
from C17 O for warm sources where CO is not depleted onto grains.
The temperature structure of the dust is calculated taking the observed
luminosity of the source, given a power-law density structure (see below). At
large distances from the star, the temperature follows the optically thin relation
8

Figure 5. Derived temperature and density structure for the envelope

around the massive YSO GL 2591 (L=2 × 104 L⊙ , d=1 kpc). The
critical densities of the CS 2–1 and 10–9 lines are indicated, as are the
typical beam sizes of the JCMT at 345 GHz and OVRO at 100 GHz
(based on: van der Tak et al. 1999).

∝ r −0.4 , whereas at smaller distances the dust becomes optically thick at infrared
wavelengths and the temperature increases more steeply (see Figure 5). It is
assumed that Tgas = Tdust , consistent with explicit calculations of the gas and
dust temperatures by, e.g., Doty & Neufeld (1997) for these high densities.
The continuum data are sensitive to temperature and column density, but
not to density. Observations of a molecule with a large dipole moment are needed
to subsequently constrain the density structure. One of the best choices is CS
and its isotope C34 S, for which lines ranging from J=2–1 to 10–9 have been
observed. Assuming a power-law density profile n(r) = no (r/ro )−α , values of α
can be determined from minimizing χ2 between the CS line data and excitation
models. The radiative transfer in the lines is treated through a Monte-Carlo
method. The best fit to the data on the infrared-bright sources is obtained for
α = 1.0 − 1.5, whereas the hot core/compact H II region sample requires higher
values, α ≈ 2. This derivation assumes that the CS abundance is constant
through the envelope; if it increases with higher temperatures, such as may be
the case for hot cores, the values of α are lowered. Note that the derived values of
α = 1.0 − 1.5 are lower than those found for deeply embedded low-mass objects,
where α ≈ 2 (e.g., Motte et al. 1998, Hogerheijde et al. 1999).
Figure 5 displays the derived temperature and density structure for the
source GL 2591, together with the sizes of the JCMT and OVRO beams. While
the submillimeter data are weighted toward the colder, outer envelope, the in-
frared absorption line observations sample a pencil-beam line of sight toward
the YSO and are more sensitive to the inner warm (∼ 1000 K) region. On these
small scales, the envelope structure deviates from a radial power law, which de-
 9

creases the optical depth at near-infrared wavelengths by a factor of ∼ 3 (van

der Tak et al. 1999).
For sources for which interferometer data are available, unresolved compact
continuum emission is detected on scales of a few thousand AU or less. This
emission is clearly enhanced compared with that expected from the inner “tip”
of the power-law envelope, and its spectral index indicates optically thick warm
dust, most likely in a dense circumstellar shell or disk. The presence of this shell
or disk is also indicated by the prevalence of blue-shifted outflowing dense gas
without a red-shifted counterpart on < 10′′ scales.

4.3. Chemical structure: infrared absorption lines

The ISO-SWS spectra of the infrared-bright sources show absorption by various
gas-phase molecules, in addition to strong features by ices. Molecules such as
CO2 (van Dishoeck et al. 1996, Boonman et al. 1999, 2000a), H2 O (van Dishoeck
& Helmich 1996, Boonman et al. 2000b), CH4 (Boogert et al. 1998), HCN and
C2 H2 (Lahuis & van Dishoeck 2000) have been detected (see also van Dishoeck
1998, Dartois et al. 1998).
In the infrared, absorption out of all J-levels is observed in a single spec-
trum. The excitation temperatures Tex of the various molecules, calculated
assuming LTE, range from < ∼ 100 to ∼ 1000 K between sources, giving direct
information on the physical component in which the molecules reside. While
CO is well-mixed throughout the envelopes, H2 O, HCN and C2 H2 are enhanced
at high temperatures. In contrast, CO2 seems to avoid the hottest gas. High
spectral resolution ground-based data of HCN and C2 H2 by Carr et al. (1995)
and Lacy et al. (1989) for a few objects suggest line widths of at most a few km
s−1 , excluding an origin in outflowing gas.
The abundances of H2 O, HCN and C2 H2 increase by factors of > ∼ 10 with
increasing Tex (see Figure 7). The warm H2 O must be limited to a < ∼ 1000 AU
region, since the pure rotational lines are generally not detected in the 80′′ ISO-
LWS beam (Wright et al. 1997). For CO2 , the abundance variation between
sources is less than a factor of 10, and no clear trend with Tex is found. For the
same sources, the H2 O and CO2 ice abundances show a decrease by an order of
magnitude, consistent with evaporation of the ices. However, the gas-phase H2 O
and CO2 abundances are factors of ∼ 10 lower than expected if all evaporated
molecules stayed in the gas phase, indicating that significant chemical processing
occurs after evaporation. More detailed modeling using the source structures
derived from submillimeter data is in progress.

4.4. Chemical structure: submillimeter emission

The JCMT data of the infrared-bright objects show strong lines, but lack the
typical crowded ‘hot core’ spectra observed for objects such as W 3(H2 O) and
NGC 6334 IRS1. Complex organics such as CH3 OCH3 and CH3 OCHO are de-
tected in some sources (e.g., GL 2591, NGC 7538 IRS1), but are not as prominent
as in the comparison sources. Yet warm gas is clearly present in these objects.
Is the ‘hot core’ still too small to be picked up by the single dish beams, or are
the abundances of these molecules not (yet) enhanced?
To investigate this question, van der Tak et al. (2000b) consider the analy-
sis of two species, H2 CO and CH3 OH. Both species have many lines throughout
10

Figure 6. Left: Abundance of HCN, N (HCN)/Ntot (H2 ), as a function

of the 13 CO excitation temperature measured by Mitchell et al. (1990).
Middle: Abundance of H2 O in the warm gas N (H2 O)/Nwarm (H2 ) as
a function of 13 CO excitation temperature; Right: Abundance of H2 O
ice as a function of 13 CO excitation temperature. The same trend is
found if the dust temperature, as derived from the 45/100 µm flux, is
used (based on: Lahuis & van Dishoeck 2000; van Dishoeck 1998).

the submillimeter originating from low- and high-lying energy levels. Given the
physical structure determined in §4.2, abundance profiles can be constrained.
Two extreme, but chemically plausible models are considered: (i) a model with
a constant abundance throughout the envelope. This model is motivated by the
fact that pure gas-phase reaction schemes do not show large variations in calcu-
lated abundances between 20 and 100 K; (ii) a model in which the abundance
‘jumps’ to a higher value at the ice evaporation temperature, Td ≈ 90 K. In this
model, the abundances in the outer envelope are set at those observed in cold
clouds, so that the only free parameter is the amount of abundance increase.
It is found that the H2 CO data can be well fit with a constant abundance of
a few ×10−9 throughout the envelope. However, the high J, K data for CH3 OH
require a jump in its abundance from ∼ 10−9 to ∼ 10−7 for the warmer sources.
This is consistent with the derived excitation temperatures: H2 CO has a rather
narrow range of Tex =50–90 K, whereas CH3 OH shows Tex =30–200 K. Moreover,
the interferometer maps of CH3 OH rule out constant abundance models. The
jump observed for CH3 OH is chemically plausible since this molecule is known
to be present in icy grain mantles with abundances of 5–40% with respect to
H2 O ice, i.e., ∼ 10−7 − 10−6 w.r.t. H2 . Similar increases in the abundances of
organic molecules (e.g., CH3 OH, C2 H3 CN, ....) are found with increasing Tdust
for a set of ‘hot–core’ objects by Ikeda & Ohishi (1999).
 11

4.5. Comparison with chemical models

Both the infrared and submillimeter data show increases in the abundances of
various molecules with increasing temperature. Four types of species can be
distinguished: (i) ‘passive’ molecules which are formed in the gas phase, freeze
out onto grains during the cold (pre-)collapse phase and are released during
warm-up without chemical modification (e.g., CO, C2 H2 ); (ii) molecules which
are formed on the grains during the cold phase by surface reactions and are
subsequently released into the warm gas (e.g., CH3 OH); (iii) molecules which are
formed in the warm gas by gas-phase reactions with evaporated molecules (e.g.,
CH3 OCH3 ); (iv) molecules which are formed in the hot gas by high temperature
reactions (e.g., HCN). These types of molecules are associated with characteristic
temperatures of (a) Tdust < 20 K, where CO is frozen out; the presence of CO
ice is thought to be essential for the formation of CH3 OH; (b) Tdust ≈ 90 K,
where all ices evaporate on a time scale of < 105 yr; and (c) Tgas > 230 K,
where gas-phase reactions drive the available atomic oxygen into water through
the reactions O + H2 → OH → H2 O (Ceccarelli et al. 1996, Charnley 1997).
Atomic oxygen is one of the main destroyers of radicals and carbon chains, so
that its absence leads to enhanced abundances of species like HCN and HC3 N
in hot gas.
Water is abundantly formed on the grains, but the fact that the H2 O abun-
dance in the hot gas is not as large as that of the ices suggests that H2 O is
broken down to O and OH after evaporation by reactions with H+ 3 . H2 O can
subsequently be reformed in warm gas at temperatures above ∼230 K, but Fig-
ure 7 indicates that not all available gas-phase oxygen is driven into H2 O, as the
models suggest. The low abundance of CO2 in the warm gas is still a puzzle,
since evaporation of abundant CO2 ice is observed. The molecule must be bro-
ken down rapidly in the warm gas, with no reformation through the CO + OH
→ CO2 reaction (see question by Minh).

4.6. Evolution?
The objects studied by van der Tak et al. (2000a) are all in an early stage of
evolution, when the young stars are still deeply embedded in their collapsing
envelope. Nevertheless, even within this narrow evolutionary range, there is
ample evidence for physical and chemical differentiation of the sources. This is
clearly traced by the increase in the gas/solid ratios, the increase in abundances
of several molecules, the decrease in the ice abundances, and the increase of the
amount of crystalline ice with increasing temperature (Boogert et al. 2000a).
The fact that the various indicators involve different characteristic tem-
peratures ranging from <50 K (evaporation of apolar ices) to 1000 K (Tex of
gas-phase molecules) indicates that the heating is not a local effect, but that
‘global warming’ occurs throughout the envelope. Moreover, it cannot be a
geometrical line-of-sight effect in the mid-infrared data, since the far–infrared
continuum (45/100 µm) and submillimeter line data (CH3 OH) show the same
trend. Shocks with different filling factors are excluded for the same reason.
Can we relate this ‘global warming’ of the envelope to an evolutionary effect,
or is it determined by other factors? The absence of a correlation of the above
indicators with luminosity or mass of the source argues against them being the
sole controlling factor. The only significant trend is found with the ratio of
12

envelope mass over stellar mass. The physical interpretation of such a relation
would be that with time, the envelope is dispersed by the star, resulting in a
higher temperature throughout the envelope.

5. Outstanding questions and future directions

The results discussed here suggest that the observed chemical abundance and
temperature variations can indeed be used to trace the evolution of the sources,
and that, as in the case of low- and intermediate-mass stars, the dispersion of
the envelope plays a crucial role. The combination of infrared and submillimeter
diagnostics is very important in the analysis. An important next step would be
to use these diagnostics to probe a much wider range of evolutionary stages
for high-mass stars, especially in the hot core and (ultra-)compact H II region
stages, to develop a more complete scenario of high-mass star formation. The
relation between the inner warm envelope and the ‘hot core’ is still uncertain:
several objects have been observed which clearly have hot gas and evaporated
ices (including CH3 OH) in their inner regions, but which do not show the typical
crowded ‘hot core’ submillimeter spectra. Are these objects just on their way to
the ‘hot core’ chemical phase? Or is the ‘hot core’ a separate physical component,
e.g., a dense shell at the edge of the expanding hyper-compact H II region due
to the pressure from the ionized gas, which is still too small to be picked up by
the single-dish beams? In either case, time or evolution plays a role and would
constrain the ages of the infrared-bright sources to less than a few ×104 yr since
evaporation. Interferometer data provide evidence for the presence of a separate
physical component in the inner 1000 AU, but lack the spatial resolution to
distinguish a shell from any remnant disk, for example on kinematic grounds.
An important difference between high- and low-mass objects may be the
mechanism for the heating and dispersion of their envelopes. For low-mass
YSOs, entrainment of material in outflows is the dominant process (Lada 1999).
For intermediate-mass stars, outflows are important in the early phase, but ul-
traviolet radiation becomes dominant in the later stages (see § 3). The situation
for high-mass stars is still unclear. The systematic increase in gas/solid ratios
and gas-phase abundances point to global heating of the gas and dust, consistent
with a radiative mechanism. However, a clear chemical signature of ultravio-
let radiation on gas-phase species and ices in the embedded phase has not yet
been identified, making it difficult to calibrate its effect. On the other hand,
high-mass stars are known to have powerful outflows and winds, but a quanti-
tative comparison between their effectiveness in heating an extended part of the
envelope and removing material is still lacking. Geometrical effects are more
important in less embedded systems, as is the case for low-mass stars, where the
circumstellar disk may shield part of the envelope from heating (Boogert et al.
2000b).
To what extent does the chemical evolution picture also apply to low-mass
stars? Many of the chemical processes and characteristics listed in Table 1 are
also known to occur for low-mass YSOs, but several important diagnostic tools
are still lacking. In particular, sensitive mid-infrared spectroscopy is urgently
needed to trace the evolution of the ices for low-mass YSOs and determine
gas/solid ratios. Also, molecules as complex as CH3 OCH3 and C2 H5 CN have
 13

not yet been detected toward low-mass YSOs, although the limits are not very
stringent (e.g., van Dishoeck et al. 1995). Evaporation of ices clearly occurs in
low-mass environments as evidenced by enhanced abundances of grain-surface
molecules in shocks (e.g., Bachiller & Pérez-Gutiérrez 1997), but whether a
similar ‘hot core’ chemistry ensues is not yet known. Differences in the H/H2
ratio and temperature structure in the (pre-)collapse phase may affect the grain-
surface chemistry and the ice composition, leading to different abundances of
solid CH3 OH, which is an essential ingredient for building complex molecules.
Future instrumentation with high spatial resolution (< 1′′ ) and high sen-
sitivity will be essential to make progress in our understanding of the earliest
phase of massive star formation, in particular the SMA and ALMA at submil-
limeter wavelengths, and SIRTF, SOFIA, FIRST and ultimately NGST at mid-
and far-infrared wavelengths.
Acknowledgments. The authors are grateful to G.A. Blake, A.C.A. Boogert,
A.M.S. Boonman, P. Ehrenfreund, N.J. Evans, T. Giannini, F. Lahuis, L.G.
Mundy, A. Nummelin, W.A. Schutte, A.G.G.M. Tielens, and M.E. van den An-
cker for discussions, collaborations and figures. This work was supported by
NWO grant 614.41.003

References

Adams, F. C., Lada, C. J. & Shu, F. H. 1987, ApJ, 312, 788

Bachiller, R. & Pérez-Gutiérrez, M. 1997, ApJ, 487, L93
Blake, G.A., Sutton, E.C., Masson, C.R., & Phillips, T.G. 1987, ApJ, 315, 621
Boogert, A.C.A. et al. 2000a, A&A, 353, 349
Boogert, A.C.A., et al. 2000b, A&A, submitted
Boogert, A.C.A., Helmich, F.P., van Dishoeck, E.F., Schutte, W.A., Tielens,
A.G.G.M., & Whittet, D.C.B. 1998, A&A, 336, 352
Boonman, A.M.S., Wright, C.M. & van Dishoeck, E.F. 1999, in The Physics and
Chemistry of the Interstellar Medium, ed. V. Ossenkopf et al. (Herdecke:
GCA), p. 275
Boonman, A.M.S. et al. 2000a, in preparation
Boonman, A.M.S. et al. 2000b, in preparation
Carr, J., Evans, N.J., Lacy, J.H., & Zhou, S. 1995, ApJ, 450, 667
Caselli, P., Hasegawa, T.I., & Herbst, E. 1993, ApJ, 408, 548
Ceccarelli, C. , Hollenbach, D. J. & Tielens, A. G. G. M. 1996, ApJ, 471, 400
Cernicharo, J., Lim, T., Cox, P. et al. 1997, A&A, 323, L25
Cesaroni, R., Felli, M., Testi, L., Walmsley, C., & Olmi, L. 1997, A&A, 325, 725
Cesaroni, R., Felli, M., Jenness, T., Neri, R., Olmi, L., Robberto, M., Testi, L.,
& Walmsley, C. M. 1999, A&A, 345, 949
Charnley, S.B. 1997, ApJ, 481, 396
Charnley, S.B. & Kaufman, M.J. 2000, ApJ, 529, L111
Charnley, S.B., Tielens, A.G.G.M., & Millar, T.J. 1992, ApJ, 399, L71
Charnley, S.B., Kress, M.E., Tielens, A.G.G.M., & Millar, T.J. 1995, ApJ, 448,
232
Churchwell, E.B. 1999, in The Origin of Stars and Planetary Systems, eds. C.J.
Lada and N.D. Kylafis (Dordrecht: Kluwer), p. 515
Cummins, S.E., Linke, R.A., & Thaddeus, P. 1986, ApJS, 69, 819
14

Dartois, E., d’Hendecourt, L., Boulanger, F., Jourdain de Muizon, M., Breitfell-
ner, M., Puget, J-L, & Habing, H.J. 1998, A&A, 331, 651
Doty, S.D. & Neufeld, D.A. 1997, ApJ, 489, 122
Ehrenfreund, P., van Dishoeck, E. F., Burgdorf, M., Cami, J., van Hoof, P.,
Tielens, A. G. G. M., Schutte, W. A. & Thi, W. F. 1998, Ap&SS, 255, 83
Evans, N.J., Lacy, J.H., & Carr, J.S. 1991, ApJ, 383, 674
Fuente, A., Martı́n-Pintado, J., Bachiller, R., Neri, R. & Palla, F. 1998, A&A,
334, 253
Garay, G. & Lizano, S. 1999, PASP, 111, 1049
Geballe, T.R. & Oka, T. 1996, Nature, 384, 334
Giannini, T. et al. 2000, A&A, 346, 617
Hartquist, T. W., Caselli, P., Rawlings, J. M. C., Ruffle, D. P. & Williams, D.
A. 1998, in The Molecular Astrophysics of Stars and Galaxies, eds. T.W.
Hartquist and D.A. Williams (Oxford: OUP), 101
Harwit, M., Neufeld, D.A., Melnick, G.J., & Kaufman, M.J. 1998, ApJ, 497,
L105
Helmich, F.P. & van Dishoeck, E.F. 1997, A&AS, 124, 205
Henning, Th., Burkert, A., Launhardt, R., Leinert, Ch., & Stecklum, B. 1998,
A&A, 336, 565
Hogerheijde, M. R., van Dishoeck, E. F., Salverda, J. M., & Blake, G. A. 1999,
ApJ, 513, 350
Ikeda, M. & Ohishi, M. 1999, in IAU Symposium 197 Abstract Book, Astro-
chemistry: from molecular clouds to planetary systems, p. 171
Irvine, W.M., Goldsmith, P.F., & Hjalmarson, Å. 1987, in Interstellar Processes,
ed. D. Hollenbach and H.A. Thronson (Kluwer: Dordrecht), p. 561
Jansen, D. J., Spaans, M., Hogerheijde, M. R. & van Dishoeck, E. F. 1995, A&A,
303, 541
Jewell, P.R., Hollis, J.M., Lovas, F.J., & Snyder, L.E. 1989, ApJS, 70, 833
Johansson, L.E.B., Andersson, C., Elldér, J., et al. 1984, A&A, 130, 227
Kuan, Y.J. & Snyder, L.E. 1994, ApJS, 94, 651
Kulesa, C. A., Black, J. H. & Walker, C. K. 1999, BAAS, 194, 4709
Lacy, J., J.H., Evans, N.J., Achtermann, J.M. et al. 1989, ApJ, 342, L43
Lacy, J.H., Knacke, R., Geballe, T.R., & Tokunaga, A.T. 1994, ApJ, 428, L69
Lada, C.J. 1999, in The Origin of Stars and Planetary Systems, eds. C.J. Lada
and N.D. Kylafis (Dordrecht: Kluwer), p. 143
Lahuis, F. & van Dishoeck, E.F., 2000, A&A, in press
Langer, W.D., van Dishoeck, E.F., Blake, G.A. et al. 2000, in Protostars &
Planets IV, eds. V. Mannings, A.P. Boss and S.S. Russell, (Tucson: Univ.
Arizona), in press
Liu, S.-Y. & Snyder, L. E. 1999, ApJ, 523, 683
Lorenzetti, D. et al. 1999, A&A, 346, 604
McCall, B.J., Geballe, T.R., Hinkle, K.H. & Oka, T. 1999, ApJ, 522, 338
Millar, T.J. 1997, in Molecules in Astrophysics: Probes and Processes, IAU
Symposium 178, ed. E.F. van Dishoeck (Dodrecht: Kluwer), p. 75
Mitchell, G.F., Maillard, J.–P., Allen, M., Beer, R., & Belcourt, K. 1990, ApJ,
363, 554
Motte, F., André, P. & Neri, R. 1998, A&A, 336, 150
Nummelin, A., Bergman, P., Hjalmarson, Å., Friberg, P., Irvine, W. M., Millar,
T. J., Ohishi, M. & Saito, S. 1998, ApJS, 117, 427
 15

Ohishi, M. 1997, in Molecules in Astrophysics: Probes and Processes, IAU Sym-

posium 178, ed. E.F. van Dishoeck (Dordrecht: Kluwer), p. 61
Ohishi, M. & Kaifu, N. 1999, in IAU Symposium 197 Abstract Book, Astro-
chemistry: from molecular clouds to planetary systems, p. 143
Ossenkopf, V. & Henning, Th. 1994, A&A, 291, 943
Saraceno, P., Benedettini, M., Di Giorgio, A.M., et al. 1999, in Physics and
Chemistry of the Interstellar Medium III, eds. V. Ossenkopf et al. (Berlin:
Springer), p. 279
Schilke, P., Groesbeck, T.D., Blake, G.A., & Phillips, T.G. 1997, ApJS, 108, 301
Simon, R., Stutzki, J., Sternberg, A., & Winnewisser, G. 1997, A&A, 327, L9
Sutton, E.C., Jaminet, P.A., Danchi, W.C., & Blake, G.A. 1991, ApJS, 77, 255
Sutton, E.C., Peng, R., Danchi, W.C., et al. 1995, ApJS, 97, 455
Turner, B.E. 1991, ApJS, 76, 617
van den Ancker, M. et al. 2000a, A&A, submitted
van den Ancker, M. et al. 2000b, A&A, in press
van den Ancker, M. et al. 2000c, A&A, submitted
van der Tak, F.F.S., van Dishoeck, E.F., Evans, N.J., Bakker, E.J., & Blake,
G.A. 1999, ApJ, 522, 991
van der Tak, F.F.S., van Dishoeck, E.F., Evans, N.J., & Blake, G.A. 2000a, ApJ,
in press
van der Tak, F.F.S. & van Dishoeck, E.F. 2000b, A&A, to be submitted
van Dishoeck, E.F. 1998, Far. Disc., 109, 31
van Dishoeck, E.F. & Blake, G.A. 1998, ARAA, 36, 317
van Dishoeck, E. F., Blake, G. A., Jansen, D. J. & Groesbeck, T. D. 1995, ApJ,
447, 760
van Dishoeck, E.F. & Helmich, F.P. 1996, A&A, 315, L177
van Dishoeck, E.F., Helmich, F.P., de Graauw, Th. et al. 1996, A&A, 315, L349
van Dishoeck, E.F. & Hogerheijde, M.R. 1999, in Origin of Stars and Planetary
Systems, eds. C.J. Lada and N. Kylafis (Dordrecht: Kluwer), p. 97
van Dishoeck, E.F. et al. 1999, in The Universe as seen by ISO, eds. P. Cox and
M.F. Kessler (Noordwijk: ESTEC), ESA SP-427, p. 437
Waters, L.B.F.M. & Waelkens, C. 1998, ARAA, 36, 233
Wright, C.M., van Dishoeck, E.F., Helmich, F.P., Lahuis, F., Boogert, A.C.A.,
& de Graauw, Th. 1997, in First ISO Workshop on Analytical Spectroscopy,
ESA SP-419, p. 37
Wright, C.M., van Dishoeck, E.F., Black, J.H., Feuchtgruber, H., Cernicharo,
J., González-Alfonso, E., & de Graauw, Th. 2000, A&A, in press

Discussion

M. Ohishi : I agree with the point you mentioned, that the chemical differences
among hot cores is due to a difference of evolutionary stage. Now we have sev-
eral well-known hot cores such as Orion KL/S, W 3 IRS5/H2 O/IRS4, SgrB2
N/M/NW etc. Can you give us your personal view on the evolutionary differ-
ences of these sources?
E.F. van Dishoeck : Van der Tak et al. (2000a) argue that the infrared-bright
objects such as W 3 IRS5 represent an earlier evolutionary phase than the hot
cores, on the basis of an anti-correlation with the radio continuum. The physical
16

picture is that the ionizing UV radiation and stellar winds push the hottest dust
in the inner regions further out, decreasing the temperature of the dust and thus
the near-infrared continuum. At the same time, the size of the region which can
be ionized is increased. The ‘erosion’ of the envelopes thus occurs from the inside
out. For other sources, infrared diagnostics are lacking, so that the situation is
less clear. It would be great if chemistry could help to tie down the time scales
of the various phases.
W. Irvine: How do you interpret the behavior of the PAH features as a function
of evolutionary stage in the sources that you discussed?
E.F. van Dishoeck : The absence of PAH features in the early embedded stage
can be due either to a lack of ultraviolet radiation to excite the features or to an
absence of the carriers. Manske & Henning (1999, A&A 349, 907) have argued
for the case of Herbig Ae/Be stars that there should be sufficient radiation to
excite PAHs in the envelope/disk system, so that the lack is likely due to the
absence of the PAHs themselves. Perhaps the PAHs have accreted into the
icy mantles at the high densities in the inner envelope and do not evaporate
and/or are chemically transformed into other more refractory species on grains.
Alternatively, the region producing ultraviolet radiation (H II region) may be
very small in these massive objects, and the photons may not reach the PAH-rich
material or have a very small beam filling factor. Once the envelope breaks up
and ultraviolet radiation can escape to the less dense outer envelope, the PAH
features from those regions will appear in spectra taken with large beams.
T. Geballe: You said that there is little evidence of the effect of ultraviolet
radiation on solid-state chemistry. Isn’t the 4.6 µm XCN feature a good example
of that influence?
E.F. van Dishoeck : The ‘XCN’ feature is indeed the best candidate for tracing
the ultraviolet processing of ices. If ascribed to OCN− , it likely involves HNCO
as a precursor. In the laboratory, HNCO is produced by photochemical reactions
of CO and NH3 , but in the interstellar medium grain surface chemistry is an
alternative possibility which does not necessarily involve ultraviolet radiation
(see Ehrenfreund & Schutte, this volume).
Y.C. Minh: Do you have an explanation of the low and constant abundances of
CO2 in the gas phase?
E.F. van Dishoeck : Charnley & Kaufman (2000, ApJ, 529, L111) argue that
the evaporated CO2 is destroyed by reactions with atomic hydrogen at high
temperatures in shocks. This is an interesting suggestion, but needs to be tested
against other species such as H2 O and H2 S which can be destroyed by reactions
with atomic hydrogen as well. Also, the amount of material in the envelope that
can be affected by shocks is not clear.