5. Comparison with molecular abundances in the interstellar medium
Comets have not undergone significant thermal and chemical processing since their formation. Therefore, the composition of their nuclear ices should provide clues to the composition of the outer regions of the Solar Nebula where they formed - presumably in the Uranus-Neptune range for Oort cloud comets such as Hale-Bopp.
Cometary ices contain both oxygenated and hydrogenated species (e.g., H2S coexists with SO2, CH4 with CO), and both saturated and unsaturated species (e.g., hydrocarbons CH4, C2H6 versus C2H2). This specificity provides key diagnostics to the chemical processes that led to the formation of such material.
Several scenarios have been proposed for the origin of cometary ices, ranging from unprocessed interstellar ices (Greenberg 1982) to purely solar condensates (Lewis & Prinn 1980), including intermediate scenarios (Lunine et al. 1991). More recently, it has been argued that cometary ices could reflect the low-temperature chemistry that occurred in the outer regions of the protoplanetary disk during its accretion phase (Aikawa et al. 1999).
As pointed out earlier (Mumma et al. 1993), models invoking the formation of cometary material from the condensation of the initially hot Solar Nebula (Lewis & Prinn 1980; Fegley 1993 and references therein), those used to explain the composition of chondritic meteorites and planets, fail in reproducing the CH4 to CO ratio, the HCN and NH3 abundances and the presence of CH3OH and other organics in comets. To resolve these discrepancies in the framework of Solar Nebula thermochemistry requires a wealth of new hypotheses, such as formation in the dense subnebulae of the giant planets and via lightning-induced shock chemistry (Fegley 1993). Thus the formation of cometary material must be considered in another framework. The high D/H ratios measured in cometary H2O and HCN show that these species have kept, at least partially, the interstellar signature of the ion-molecule isotopic exchanges that occurred in the protosolar cloud, prior to its collapse (Balsiger et al. 1995; Eberhardt et al. 1995; Bockelée-Morvan et al. 1998b; Meier et al. 1998a, 1998b).
Comprehensive studies of the composition on interstellar ices are now available, from infrared observations of deeply embedded protostars either from the ground or with ISO. The analogy between interstellar and cometary ices is strengthened. Admittedly, there is some scatter in the abundances among the various interstellar sources which were studied. However, it is striking that the major constituents (H2O, CO, CO2, CH3OH, H2CO, CH4, as well as OCS and total amount of nitriles X-CN) have approximately the same abundances relative to water in both kinds of material (Whittet et al. 1996; Ehrenfreund et al. 1997; Crovisier 1998, 1999). Unfortunately, this comparison has limitations and cannot presently be extended to other species with minor abundances because (i) infrared observations of interstellar ices are not as sensitive as radio observations in the gas phase; (ii) spectroscopic infrared features of ices are not as narrow as those of radio lines and, in some cases, do not provide unambiguous interpretation.
It is however possible to go further into the origin of cometary ices by making a comparison with hot molecular cores and bipolar flows in the interstellar medium (ISM), whose composition can be investigated in great detail by radio spectroscopy.
Hot cores are dense gas clumps heated to temperatures above 100 K by UV radiation and shocks from nearby high-mass stars that have just formed from the gravitational collapse of a cold molecular cloud. It is believed that some species observed in the gas phase in these regions result from the sublimation of icy grain mantles. Indeed, high levels of deuteration are seen in many molecules (e.g. HDO, NH2D, HDCO, CH3OD), which can only be explained by deuterium fractionation in an earlier cooler phase. Chemically, hot cores are characterized by very large abundances (when compared to dark clouds) of hydrogenated molecules such as H2O, NH3, CH3OH and H2S, as found in ices around protostars and comets, and contain also complex organic species. In these regions of the ISM, cometary species such as HNCO, HCOOH, NH2CHO, HCOOCH3 and CH3CN are predominantly found. These species were possibly synthesized on grains in the earlier cold collapse phase, either by grain-surface reactions or by subsequent UV or cosmic ray processing, and then evaporated. The former mechanism is invoked for the formation of simple hydrogenated species such as H2, H2O, CH4, NH3 or H2S, and could also be responsible for the formation of CH3OH, H2CO and CO2 on grains (Tielens & Whittet 1997). H, C, N, O atom additions to the accreted CO has also been proposed for the formation of HCOOH, NH2CHO, and HNCO on the grain mantles (Charnley 1997a). There is some debate on the exact mechanisms from which these species formed on the grain mantles (see the review of Tielens & Whittet 1997). Laboratory studies show that species like NH2CHO, HCOOH, HCOOCH3 can form from the UV irradiation of simple ices (Cottin et al. 1999, and references therein).
Models of the gas-phase chemistry in hot cores show that, after their evaporation, most species survive for time-scales of at least 103-104 yr, which correspond to the estimated age of hot cores (Hatchell et al. 1998a; Charnley 1997b). This is because of their high activation barriers for reactions with atomic hydrogen (one exception is H2S which is rapidly destroyed, as discussed later) and the relatively low fractional ionization in these dense regions. However, chemical processing can be efficient for some species, so that second-generation species, created by gas-phase chemistry in the hot cores, exist as well (c.f. Millar 1997; see also the review of van Dishoeck & Blake 1998). An active point of discussion is the discrimination between first-generation or second-generation species. This study is done by comparing molecular abundances predicted by chemical modelling to observations, but is hampered by the poorly known structure of molecular hot cores, the incomplete knowledge of the composition of icy mantles and the lack of laboratory or theoretical data on some reaction rates. As discussed later in this paper, there is a consensus on the origin of some species, while there is a debate for others.
Like hot cores, bipolar flows produced by young protostars provide means to probe the composition of icy mantles disrupted by sublimation and/or sputtering, but this time in the envelopes of low-mass stars. Species present in interstellar ices (CH3OH, H2CO) and/or in hot cores (e.g. NH3) show large enhancements in the region where the outflow interacts with the surrounding molecular cloud (Bachiller & Pérez Gutiérrez 1997). The high enhancement in CH3OH reported in several outflows is thought to be caused by direct desorption from grains, so that other desorbed species could be present as well (Bachiller & Pérez Gutiérrez 1997and references therein). On the other hand, high-temperature shocks also drive a specific gas-phase chemistry where endothermic reactions can occur. The chemistry of sulphur in particular is severely affected by shocks (Pineau des Forêts et al. 1993).
We have restricted our comparative study to a few, well studied, hot cores, where a large number of molecules have been detected, namely: W3(H2O), Orion KL (Hot Core and Compact Ridge) and G34.3+0.15. Table 5 lists the molecular abundances with respect to H2 that we use and the corresponding references. Most abundances in W3(H2O) come from the spectral line survey of Helmich & van Dishoeck (1997) performed at the James Clerk Maxwell Telescope (JCMT). Most abundances in G34.3+0.15 were also determined from spectral surveys made at the JCMT (Macdonald et al. 1996; Hatchell et al. 1998a, 1998b, 1998c). Such spectral surveys allow to probe the physical conditions in the sources (in particular rotational temperatures from multi-line studies) and provide consistent sets of abundances. The determination of molecular abundances in the hot cores of Orion KL is more delicate due to the complexity of this source. High spatial resolution and/or a deconvolution of the line profiles into velocity components are required to disentangle the different regions. For the N-bearing species in the Orion Hot Core, we used, when available, the abundances inferred by Wright et al. (1996) from observations with the BIMA array. A consistent set of CS, SO2, OCS and H2CS abundances in the Orion Hot Core is given by Sutton et al. (1995) from a spectral line survey performed at the JCMT. The Orion Compact Ridge has considerable amounts of oxygen-rich species, when compared to the Orion Hot Core: molecular abundances from Sutton et al. (1995) were used. The comparison with bipolar flows was restricted to the best studied flow L1157 and abundances inferred in its B1 position were taken (Bachiller & Pérez Gutiérrez 1997). It should be noted that the molecular column densities (and abundances relative to H2) in these sources are uncertain (possibly by up to an order of magnitude), as they rely on assumptions on, e.g., the sources size and structure and the excitation conditions of the molecules.
Table 5. Molecular abundances with respect to H2 in the bipolar flow L1157 and in hot cores.
Fig. 3 shows the comparison between abundances measured in comet Hale-Bopp and in these sources. The abundances are given relative to HCN for N- and S-bearing species (Figs. 3a and 3c), and relative to CH3OH for CHO-bearing species (Fig. 3b). Note that HCN presents a high deuterium fractionation in hot cores, as does CH3OH, which suggests that most of the HCN molecules in hot cores are directly evaporated from grains (Hatchell et al. 1998b). Note that the [CH3OH]/[HCN] ratios in L1157, W3(H2O) and G34.3+0.15 agree within a factor of 3 with the ratio observed in comet Hale-Bopp and within a factor of 5-10 for the Orion hot cores. Thus a correlation similar to that seen in Fig. 3b is observed when comparing abundances with respect to HCN, although with a larger scatter. From Fig. 3, we see that the molecular abundances measured in the different hot cores, when normalized to HCN or CH3OH, generally agree within about one order of magnitude. This scatter is small, when compared to the range of molecular abundances investigated here, which spreads over 6 orders of magnitude. There are however a few exceptions: NH3, whose abundance in G34.3+0.15 is much higher than in the other hot cores, HNCO and, more dramatically, H2CS among the S-bearing species. Hatchell et al. (1998a) observed less scatter when observing sulphur species in several hot cores including G34.3+0.15. Molecular abundances in L1157 are within the range of measured values in hot cores, except for CO. But the CO column densities in these sources trace the ambient gas, rather than CO sublimation from grains.
We find a good correlation between the abundances of N-bearing and CHO-bearing species measured in the hot cores and the bipolar flow and those measured in comets (Figs 3a, 3b). The agreement is particularly good for CHO-species. In details, some differences are seen, such as the [H2CO]/[CH3OH] ratio, 0.04 in the bipolar flow and W3(H2O) and 0.2-0.5 in comets. There is, however, a much stronger disagreement with dark clouds where the high [H2CO]/[CH3OH] ratio () conveys the low efficiency of gas-phase reactions in forming CH3OH. In addition, the H2CO abundance measured in comets does not refer to the H2CO content in the nucleus, which is much smaller (see Sect. 4.3). We also note a net tendency for the [HC3N]/[CH3CN] ratio to be significantly higher in comet Hale-Bopp ( 1, Table 4) than in hot cores (0.1-0.6), although lower than in dark clouds ( 10). CH3CN is believed to be a first-generation species in hot cores and to form onto grains (Millar et al. 1997). In contrast, HC3N is produced efficiently by ion-molecule reactions in cold dark clouds before condensing onto grains (Millar et al. 1997). In hot cores, it is rather considered as a second-generation species as it can be easily produced from evaporated C2H2 (Millar et al. 1997). The [HC3N]/[CH3CN] ratio in comets could result from the incorporation of species formed both in gas phase and by grain-surface reactions. The good agreement for HCOOCH3 abundances in hot cores and comets might be surprising, as this species is usually considered as a daughter species in hot cores. However, this has been questioned recently from the inability of chemical models to reproduce the high abundances observed in several hot cores (Millar et al. 1997; Hatchell et al. 1998c). Proposed production routes for HCOOCH3 in hot cores are via the reaction of protonated methanol with H2CO (Millar et al. 1997) or with HCOOH (Charnley 1997a). It will be interesting to investigate whether such reactions in the coma are efficient enough to produce the observed cometary HCOOCH3 abundance.
We find a larger scatter when comparing the cometary and ISM abundances of S-bearing species (Fig. 3c). H2S is thought to be the main sulphuretted molecule in interstellar ices, as found in comets, and to form on grains as its gas-phase formation routes involve large energy barriers. But only in the Orion Hot Core is H2S much more abundant than SO2. However, H2S is rapidly destroyed by hot gas chemistry after evaporation from grain mantles to form SO and SO2 (Hatchell et al. 1998a) and, to a smaller extent, OCS and H2CS. Thus, the high [SO2]/[H2S] ratios found in hot cores and bipolar flows, and more generally the abundances of the S-bearing species in these sources, are believed not to be directly indicative of initial grain composition. Concerning OCS and H2CS, models of hot core chemistry require them to be ejected from grains, in order to reproduce their observed abundances relative to H2 (Charnley 1997b; Hatchell et al. 1998a). Indeed, OCS is directly observed in interstellar ices with abundances consistent with cometary ices. The average [H2CS]/[OCS] ratio in the hot cores observed by Hatchell et al. (1998a) is 0.15, within a factor of 3 from the cometary value. Therefore, there is a good indication that sulphur-bearing compounds in comets are also closely related to those present in interstellar ices.
Bockelée-Morvan et al. (1998b) and Irvine et al. (2000) point out the similarity between the (D/H) and (D/H) ratios measured in comets and in hot cores. However, it should be noted that these ratios in hot cores are significantly lower than those measured or expected in interstellar ices, suggesting that the originally higher deuterium fractionation in icy grain mantles might have been altered in the hot post-evaporation phase (Hatchell et al. 1998b; Teixeira et al. 1999). Independently, the relatively low deuterium enrichment found in cometary ices with respect to that measured in carbonaceous meteorites and interstellar ices suggests that the water contained in the interstellar grains of the protosolar cloud was partly mixed with water reprocessed in the warm inner part of the Solar Nebula by turbulent diffusion (Drouart et al. 1999; Bockelée-Morvan et al. 1998b). Therefore, the strong similarity between the levels of deuteration in comets and hot cores might be partly fortuitous. A different point of view is presented by Meier et al. (1998a,b) and Hatchell et al. (1999) who argue that the D/H ratios observed in comets and hot cores reflect the moderately high temperatures (50-80 K) at which their ices formed.
© European Southern Observatory (ESO) 2000
Online publication: January 18, 2000