Astron. Astrophys. 322, 962-974 (1997)

4. Chemical abundances in the gas around Oph

The first suggestions that a complex chemistry might be visible in the gas around Oph were the CS observations of Drdla et al. (1989) and the detection of HCO emission by Liszt and Lucas (1994); this latter work was apparently the first time that polyatomic molecules were detected in a classical diffuse cloud. The column density of HCO , N(HCO ) , is found to be 20 (van Dishoeck and Black 1986) to 1000 (Federman et al. 1996, Kopp et al. 1996) times higher than predicted by conventional quiescent chemical models, even when the abundances of CO, OH, and/or CH are reproduced.

To pursue the matter further we mapped the HCO emission somewhat better and searched for other species at the peak of the CO and HCO emission distribution to the South of the star. The point is that neither N(H) nor is necessarily very much higher to the South than toward the star and even N(CO) and N(HCO ) may be constant within a factor two or so, with the emission pattern determined largely by excitation conditions. If so, the chemical patterns seen to the South may be interesting not only in their own right, but as guidelines to more complex phenomenon toward Oph as well.

As shown in Fig. 5 and Table 1 the HCO emission roughly follows that of CO, but it is relatively strong toward the star and weak to the North. The CO J=2-1/J=1-0 intensity ratio is noticeably larger at the position South, which is a clear indication that the excitation is stronger there. The weaker HCO lines are really at the limit of what one can expect to map with current technology.

4.1. Emission and column densities at the HCO emission peak

Emission profiles at the HCO peak are shown in Fig. 5 and summarized in Table 3. Toward L134, and at many scattered positions along the inner regions of the northern galactic plane at (Liszt 1995), HCO , HCN, CS (J=2-1), and H emission profiles have similar brightnesses (accounting for hyperfine structure in species like H); the same is true for CN in L134 as well. Typically such species are about 10 times weaker than ¹³ CO (J=1-0) and 50 times weaker than ¹² CO. For HCO the lines seen around Oph seem to follow this latter rule of thumb; they are 1%-2% as strong as ¹² CO (see Table 2).

This common similarity of the emission brightness in so many species seems a remarkable coincidence but it is certainly not repeated South of Oph. There, HCO is at least an order of magnitude stronger than CS and more than three times stronger than the main HCN line. The bump in the CN spectra is noise, unless it is supposed that the velocity of CN is unique; this is demonstrably not so toward the star (in optical spectra at least).

To interpret these line profiles in terms of molecular column densities we follow the n solution locus determined from ¹² CO and ¹³ CO profiles of the J=2-1 and J=1-0 lines by Liszt (1993), shown in Table 4. As noted by Liszt (1993) the CO column density is well-determined as long as the gas is not at the high-density, low-temperature limit; the data of Kopp et al. (1996) yield very similar values of N(CO). In any case, even the maximum possible CO column density is small compared to N(C ) toward the star ( ; Cardelli et al. 1993). It is not even necessarily large compared to the CO column seen there, log N(CO) = 15.4.

Table 4. Logarithmic column densities S of Oph

The optically determined electron fraction, measured relative to is X(e) = n(e)/n() (Cardelli et al. 1993). Because the maximum allowed CO column density is still quite small compared to the total amount of carbon toward the star, it seems unavoidable that the majority of the carbon is ionized and the electron fraction is near the relative abundance of carbon itself. Thus only the largest of the paramterized choices of X(e) is really relevant and the uncertainty in derived column densities is not as great as the full range of entries in the tables might otherwise suggest. The calculations with negligible X(e) are included to illustrate the effect of neglecting the electron excitation when it is important.

For the species having higher dipole moments, we have performed excitation calculations using the electron-ion and electron-neutral rate constants of Bhattacharayya, Bhattacharayya, and Narayan (1981) and Dickinson and Flower (1981). We used the -molecule excitation rates of Green and Chapman (1978) and Turner et al. (1992) for CS and those of Monteiro (1985) for HCO . For CN, following the recommendations of Turner et al. 1992), we used the same rate constants as for CO, from Flower and Launay (1985). For CN and HCN, where the lines are split, the calculations were done for the strongest component only since line overlap does not occur and the neutral cross-sections are not always known in detail anyway, especially for the admixture of atomic hydrogen which may be present.

The CO solution locus asymptotically but quickly reaches conditions of constant thermal pressure and column density when . It is this phenomenon, easily applied when the CO column density and excitation temperature are known, which first suggested that the density in the neutral-bearing clouds was small toward the star (Smith, Krishna Swamy and Stecher 1978; Liszt 1979). Species having higher dipole moments do not follow this behaviour if electrons provide significant excitation, as expected here. Unlike the CO solutions, those for the other species require higher molecular abundances at higher , for two reasons; the density indicated by the CO solution locus declines at higher temperature and fixed pressure, lowering the density of electrons, and the electron excitation is itself less effective owing to an inverse-square root functional dependence on the kinetic temperature.

At the highest and most relevant electron fraction, the abundances or upper limits for HCO , HCN, CS, and CN vary by a factor of five or so. The CO column density changes little for K and the lowest-temperature (10 K) highest-pressure solutions probably can be excluded by a comparison of the ¹² CO and ¹³ CO linewidths (Sect. 5; the ¹² CO is not saturated). It is clear that HCN is at most only slightly more abundant than HCO , and that CS is not as abundant as HCO or HCN. Furthermore, the column density of CN is at most only slightly larger than toward the star (log N(CN) = 12.45; van Dishoeck and Black 1989), although it might be the most abundant of the mm-wave species observed here. The range of HCO column densities in Table 3 is wider but fully overlapped with that inferred by Liszt and Lucas (1994) toward the star.

Drdla et al. (1989) presented a detection of CS J=2-1 emission toward Oph with km s^-1 and a peak intensity K; their quoted CS column density N(CS) = 0.7- 5.0 is substantially larger than the limits we set at the molecular emission peak to the South (Table 3). Their line profile occupies the range -3 km s^-1 v -1 km s^-1, which does not overlap with emission from any other species. Our CS J=2-1 spectrum toward Oph is featureless. Over the velocity range corresponding to the CO J=1-0 emission profile, our data have rms noise K.

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de