Astron. Astrophys. 327, 337-341 (1997)

4. Discussion

What are the implications of our data for the C60 assignment? Foing and Ehrenfreund (1994) pointed out that the DIB positions are close to absorptions measured in the laboratory in matrix environment. For the two DIBs to be due to C60, the gas-phase C60 spectrum has to have slightly different band positions, presumably due to matrix interactions in the present laboratory data. Comparison of laboratory data with our measured band positions would imply a matrix shift of 2.6 4.0 and 9.4 3.0 Å, which are the same within measurement error, as expected. A more accurate laboratory position of the C60 absorptions can make this a more important discriminant.

The two bands behave in a similar way in a dense cloud environment. In general, DIBs of a similar broadness class tend to weaken together when some dense cloud matter is in the line of sight (Jenniskens & Désert 1995). Our DIBs are weaker than those measured by Foing& Ehrenfreund in lines of sight that represent the diffuse interstellar medium. This suggests that some amount of dense cloud matter is in front of HD 63804 and HD 80077. A constant line ratio would be expected if both lines are caused by the same molecule. Our result is consistent with this, but does not prove that both bands are due to the same molecule.

Our most striking result is that the relative intensity of the bands (1:0.9) does not agree with the results by Foing & Ehrenfreund (1994) (1:1.6 - with an error of 30% according to Ehrenfreund & Foing 1995), nor with the band intensity ratio measured for C60 in a neon matrix (1:1.5 - Fulara et al. 1993). Whether or not C60 can be ruled out completely on the basis of this result depends on the question whether the gas phase spectrum of C60 has a similar band intensity ratio. Fulara et al. pointed out that the relative intensity of the two band systems at 9580 and 9642 Å depends on experimental conditions: in the case of mass-selected ion deposition, 9580 was slightly more intense than in photoionization experiments. Relative intensities varied from 1:1.5 to 1:2.0 (Maier 1994). Those variations are outside the range of values observed for the two DIBs in our data. The question remains what is responsible for these band strength changes. Fulara et al. interpreted the two bands as due to two configurations of the molecule imposed by the matrix environment. Only if the gas-phase spectrum differs significantly from present laboratory data, would this interpretation be consistent with the two DIBs being due to gas phase C60.

The smaller band ratio that we find compared to Foing & Ehrenfreund (1994) might suggest that the bands are intrinsically variable, which would exclude C60 as a carrier of both. However, it is possible that the band ratio of Foing & Ehrenfreund is simply in error. From their graphs, it is clear that the equivalent width of the features are very uncertain and large formal errorbars contradict with a smooth increase of the band strength with reddening. Some residual stellar lines seem to be present in the 9632 feature, inspite of attempts to divide out the lines by adopting telluric standards with similar spectral type. For example, the positions of the sharp and narrow features of HD 190603 and HD 21389 differ by 1.0 0.2 Å. The doppler velocities of interstellar matter account for only 0.06 Å (Jenniskens & Désert 1994). If, on the other hand, one of the lines has a strong stellar line contribution, then the expected variation is 0.8 Å, whereby HD 190603 should be redshifted, as observed.

Indeed, recent new observations presented by Foing & Ehrenfreund (1997), and obtained independently from our results, also imply a much lower band ratio. Our results are in good agreement with data in the latter paper. However, note that that is fortuous, because the observing procedure used by Foing & Ehrenfreund (1997) does also not guarantee that stellar lines are correctly removed. In fact, the somewhat higher band ratio in HD 183143 (1.17 0.2) and the slightly more sharper band profile at the peak is likely due to residual stellar lines.

© European Southern Observatory (ESO) 1997

Online publication: April 8, 1998
helpdesk.link@springer.de