Carriers of 4964 and 6196 diffuse interstellar bands and environments dominated by either CH or CH + molecules
T. Weselak, G.A. Galazutdinov, O. Sergeev, V. Godunova, R. Kołos, J. Krełowski
aa r X i v : . [ a s t r o - ph . GA ] N ov ACTA ASTRONOMICA
Vol. (0) pp. 0–0 Carriers of 4964 and 6196 diffuse interstellar bands and environmentsdominated by either CH or CH + molecules Weselak , T., Galazutdinov , , G.A., Sergeev , O., Godunova , V., Kołos ,R., Krełowski , J. Institute of Physics, Kazimierz Wielki University, Weyssenhoffa 11, 85-072 Bydgoszcz,Poland e-mail: [email protected] Instituto de Astronomia, Universidad Catolica del Norte, Av. Angamos 0610,Antofagasta, 1270709 Chile Pulkovo Observatory, Pulkovskoe Shosse 65, Saint-Petersburg 196140, Russiae-mail: [email protected] International Center for Astronomical, Medical and Ecological ResearchPostal code and address: 27 Zabolotnoho Str., Kyiv 03680 Ukrainee-email: [email protected]; [email protected] Institute of Physical Chemistry, Polish Academy of SciencesKasprzaka 44, 01-224 Warsaw, Polande-mail: [email protected] Center for Astronomy, Nicolaus Copernicus University, Grudzia¸dzka 5, Pl-87-100Toru´n, Polande-mail: [email protected]
Received Month Day, Year
ABSTRACTThe analysis of radial velocities of interstellar spectral features: CH, CH + as well as 4964 and6196 diffuse interstellar bands, seen in spectra of HD 151932 and 152233, suggests that carrier ofthe former is spatially correlated with CH while that of the latter – with CH + . A further analysis,done in this paper and based on the sample of 106 reddened OB stars, partly confirms this suggestion,showing that the CH column density correlates indeed much better with the equivalent width of the4964 DIB than with that of the 6196 DIB. However, the strengths of the 6196 DIB correlate onlymarginally better with CH + than with CH. Key words:
ISM: molecules
Introduction
Diffuse interstellar bands (DIBs) are numerous narrow to broad absorption fea-tures mostly found in the visual and near infrared (4,000-10,000Å) spectra of earlytype OB–stars. Despite many observational efforts (Herbig 1995, Galazutdinov et ol. 0 + has been recently reported by Krełowski et al. (2010).However, this identification, as well as the recent assignment of two broad diffusebands near 4882 and 5450 Å to the propadienylidene ( l -C H ) molecule (Maier etal., 2011a) have been disputed (Krełowski et al. 2011; Maier et al. 2011b) and re-main uncertain. Recent results of Ehrenfreund et al. (2002), Cox and Patat (2008)proved the existence of DIBs and molecular features in spectra of LMC/SMC andM100 stars, respectively. However, measurements of weak interstellar features arehindered by relatively low signal–to–noise (S / N) inside their profiles.The molecular origin of at least some DIBs seems likely, their profiles beingoften complex and reminiscent of rotational envelopes in molecular spectra (Sarreet al. 1995a, Kerr et al. 1998). The 6614 DIB contains three main componentsthat have the appearance of unresolved P, Q and R branches of an electronic tran-sition of a large molecule (see Herzberg 1950). A more or less similar patternis observed also in the case of 5797 DIB, the strength of which correlates tightlywith the CH column density (Weselak et al. 2008b). This correlation supportsalso the idea that carriers of some DIBs are molecules. Another important pieceof evidence is that the width of certain DIB can depend on the temperature withinan interstellar cloud - as shown by Ka´zmierczak et al. (2009) who have found acorrelation between the FWHM of the 6196 DIB and the rotational temperature ofthe homonuclear C molecule. An important conjecture here is that some of theDIBs, namely those sensitive to temperature of the environment, may originate incentrosymmetric molecules, as the lack of electric dipole moment slows down therotational relaxation (see Bernath 2005).Methylidyne (CH), first identified in the ISM by McKellar (1940), is closelyrelated to molecular hydrogen (H ), as already shown by Mattila (1986) and Wese-lak et al. (2004) and also with OH (Weselak et al. 2009b, 2010b). The abundancesof CH molecule are also well correlated with those of NH (Weselak et al. 2009c).On the other hand, the column density of the corresponding cation, CH + , corre-lates very poorly with that of H - indicating no relation between the abundances ofthese two molecules (Weselak et al. 2008a). The formation and existence of CH + in the ISM remains an unsolved problem (van Dishoeck and Black 1989, Gredel etal. 1993, Sheffer et al. 2008 and references therein). One can argue that environ-ments dominated by the CH molecule (i.e. regions where it is much more abundantthan CH + ), and those where CH + dominates, are well separated in space. Thisconclusion is grounded in the publication of Allen (1994) which demonstrates thatradial velocities of CH and CH + may differ by as much as 7.3 km/s.Spectral observations of highly-reddened early type stars offer an excellent op- A. A. portunity to probe large column densities of gas along diverse lines of sight, toderive physical parameters of corresponding environments and eventually unveilthe nature of DIB carriers. Identified and unidentified interstellar spectral featuresshould share the radial velocities if they originate in the same environments. Insuch a case, their intensities should be correlated as well.The aim of this work is to present the relations between the column densities ofsimple diatomic molecules CH and CH + and the equivalent widths of two narrowDIBs: 4964 and 6196 Å. It is certainly important to specify the physical parametersof environments which host the DIB carriers. It may shed some light on the con-ditions facilitating the formation / preservation of DIB carriers. This is particularlyinteresting in the context of interstellar CH and CH + synthesis paths, presented byFederman et al. (1982) and by Zsargó and Federman (2003).
1. The observational data
The observations <
20 mÅ), the following relation were applied (Weselak et al.2009a) to obtain the column density (in cm − ): N = . × (cid:0) EW / f l (cid:1) ,where EW is equivalent width of the line, l its wavelength (both in Å), and f –its oscillator strength.We calculated the column density of the CH + cation adopting, for an usaturated4232 Å band, the oscillator strength f= Based on observations made with ESO Telescopes at the La Silla or Paranal Observatories un-der programs 71.C-0367(A), 073.D-0609(A), 074.D-0300(A), 075.D-0369(A), 076.C-0431(B) and082.C-0566(A). ol. 0 + feature at 3957 Å were performed, using its recently obtained f -value of 0.00342(Weselak et al. 2009a). When estimating the column density of CH + , we onlyused the bands of the total equivalent width lower than 20mÅ. In the case of twoobjects where Doppler–spliting is evident (HD 152235 and 168607), we also madeuse of the 3957 Å band to estimate the column densities. In the case of HD 34078,we relied on the unsaturated CH + (2, 0) band at 3745Å, with its recently published f -value of 0.00172 (Weselak et al. 2009a).In the case of CH molecule we calculated column density based on unsaturated CHA–X band at 4300 Å adopting recent f -value of 0.00506 (Weselak et al. 2014).When the 4300 Å feature was saturated we used both CH B–X bands at 3886 and3890 Å ( f -values equal to 0.00320 and 0.00213, respectively – see Weselak et al.2014). The CH B–X bands at 3886 and 3890 Å were used in the case of HD’s:34078, 147889, 154368 and 204827 respectively.
2. Results
For this project we choose a sample of 106 reddened OB stars, for which wecollected the echelle spectra featuring CH and CH + bands, as well as the DIBsat 4964 and 6196 Å. For each line of sight. Tables 1 and 2 list a stellar HD/ BDnumber (observed by b – BOAO, f – FEROS, H – HARPS, U – UVES), spec-tral type, luminosity class, colour excess E(B-V) (in magnitudes), as well as thecolumn densities of CH + and CH molecules (in 10 cm − units, together withcorresponding error estimates), and finally the equivalent widths (W l ’s) of diffusebands, with their standard deviations. In Figure 1 we present, in the radial velocityscale, the CH and CH + features, as seen in a very high S/N ratio (2,500) spectrumof HD 149757 ( z Oph). It is evident that radial velocities of both species are identi-cal. At the bottom of this figure we also show the two narrow DIBs at 4963.85 and6195.97 Å in the spectrum of HD 23180, taken from the survey of Galazutdinovet al. (2000). A weak feature at 6194.76 Å can be seen in the vicinity of 6196DIB. Also the radial velocities of 4964 and 6196 DIBs do not differ, as far as theirasymmetric profiles allow for central wavelengths measurements.To determine the DIB radial velocities we have used the rest wavelength veloc-ity frame based on the KI line at 7698.974 Å (Galazutdinov et al. 2000) towardHD 23180 where radial velocities of all identified interstellar features are the same.This method allows us to measure radial velocities of diffuse bands and of narrowinterstellar features with a very low uncertainty, smaller than 0.3 km/s (see Fig. 1).HD 149757 is a moderately reddened, popularly observed object, with no Dopplersplittings of interstellar features. The above statement is supported by the obser-vations of CH B-X lines (centered near 3886 and 3890 Å) and of the feature near3957 Å, belonging to the CH + cation, which proved useful for column density A. A. derivation when saturation affected stronger bands (Weselak et al. 2008a, 2008b).We observed z Oph using UVES and HARPS spectrographs, and the result is thesame in both cases: no Doppler shift of any feature, relative to any other one, isobserved toward this target.It is well known that the CH + cation needs different interstellar conditionsthan CH to be formed (Federman 1982). In many cases CH + features are Doppler-splitted, while those of CH are not (see Pan et al. 2004). However, there are alsolines of sight where we observed the splitting in both CH and CH + bands. In Fig-ure 2 we present a fragment of HD 152233 spectrum, measured with HARPS, withCH and CH + (4300 and 4232 Å, respectively) features plotted in the radial velocityscale. The CH + line is apparently Doppler splitted, and contains two components;conversely, only one component is detected for the neutral CH molecule. At thebottom of Fig. 2 we present the spectral region of 4964 and 6196 DIBs, in the samevelocity scale. Intrinsic DIB wavelengths were those of Galazutdinov et al. (2008).In the case of broader DIBs (see Fig. 1), the error of radial velocity measurementsis not larger than 1-2 km/s. It is evident that the 6196 DIB shares its radial veloc-ity with the blue-shifted (-7.6 ± + line, while the4964 DIB shares the radial velocity with CH. A similar effect can be observed inthe UVES spectrum of HD 151932, with the blue–shift of CH + equal to -6.1 ± ol. 0 -100 0 1000.9810.9900.999 DIB4964 R e l a t i v e i n t en s i t y Radial velocity (km/s)
DIB6196 OphUVESR=80,000
CH 4300 A CH + Figure 1: Top panel: CH A-X and CH + A-X lines centered near 4300 and 4232 Å,respectively, as seen in the spectrum of HD 149757. Common radial velocity scale.Bottom panel presents the DIBs at 4963.85 and 6195.97 Å (Galazutdinov et al.2000), in the same radial velocity scale. A weak DIB at 6194.76 Å can be discerned.Difference in radial velocities is not higher than 0.3 km/s.
A. A. -100 0 1000.9800.9870.9941.001
DHARPSR=115,000 DIB6196 DIB4964 R e l a t i v e i n t en s i t y Radial velocity (km/s)
CH 4300A CH + Figure 2: The 6196 DIB is shown sharing the velocity of the second componentof CH + , while 4964 DIB remains at the position of CH. The depth of the 6196DIB is scaled down by a factor of 2. It is evident that the 6196 DIB shares itsradial velocity with the blue-shifted (-7.6 ± + line. ol. 0 -100 0 1000.970.980.991.00 DUVESR=80,000 DIB4964DIB6196 R e l a t i v e i n t en s i t y Radial velocity (km/s) CH + Figure 3: Same as Fig. 2 but for a different target observed with another instru-ment. The depths of 4300 Å CH line and of 6196 DIB are scaled down by a factorof 2. The 4964 DIB shares the radial velocity with CH in the UVES spectrum ofHD 151932, with the blue–shift of CH + equal to -6.1 ± A. A.
Such telltale Doppler shift analyses can at present, unfortunately, not be accom-plished for a statistically meaningful sample of cases, given the fact that velocitydifferences between CH and CH + are usually very small. Therefore, the only reli-able possibility of investigating the spacial relationships between DIB carriers andsimple molecules is to correlate DIB strengths with molecular column densities de-rived for an extensive sample of sight lines. Figure 4 shows such relations for CHversus either 4964 or 6196 DIBs. It is clear that the 4964 DIB is very well corre-lated with the column density of CH, while for 6196 DIB the similar correlatationis moderate (correlation coefficients equal 0.86 and 0.48, respectively). E W ( ) [ m A ] E W ( ) [ m A ] N(CH) [10 cm -2 ]R=0.86 (0.03) R=0.48 (0.08) Figure 4: Equivalent widths of 4964 and 6196 DIBs plotted vs. the column densityof CH (correlation coefficients equal 0.86 and 0.48, respectively). Very good rela-tion between equivalent widths of 4964 DIB and column densities of CH moleculeis presented with dotted line.Very good relation with correlation coefficient equal to 0.86, seen in Fig. 4proves that the carrier of 4964 DIB occupies the same regions of translucent cloudsas the neutral methylidyne molecule. It also means that 74 % (0.86x0.86) of thetotal variation in equivalent width of the 4964 DIB can be explained by the linearrelationship between equivalent widths of the 4964 DIB and column densities ofCH molecule seen in Fig. 4. This molecule, on the other hand, does not seemto share the environment with a 6196 DIB carrier. Fig. 5 shows another pieceof evidence: the column density of CH + correlates weakly with the strength of4964 DIB (correlation coefficient equal to 0.47), and only marginally better withthat of 6196 DIB with the calculated correlation coefficient equal to 0.68. Of note,since CH and CH + are not well mixed within the interstellar clouds, their columndensity ratio cannot serve as a useful physical parameter characterizing individual ol. 0 R=0.47 (0.08) E W ( ) [ m A ] E W ( ) [ m A ] N(CH+) [10 cm -2 ]R=0.68 (0.05) Figure 5: Correlation of both 4964 and 6196 DIBs with column density of CH + (correlation coefficients equal to 0.47 and 0.68, respectively). Note better relationbetween equivalent widths of 6196 DIB and column densities of CH + moleculesuggested by radial velocity shift presented in Figs. 2 and 3. R=0.72 (0.05) N ( CH ) [ c m - ] N ( CH + ) [ c m - ] E(B-V) [mag]
R=0.75 (0.05)
Figure 6: Column densities of both CH and CH + correlate to a similar degree withE(B-V) (in magnitudes). Moderate relations with correlation coefficients equal to0.72 and 0.75, respectively.lines of sight. It seems also that a predictive value of the colour excess E(B-V), inthe discussed contexts, is very doubtful (Fig. 6). Both CH and CH + are correlated0 A. A. with E(B-V) to a similar, moderate, degree with correlation coefficients equal to0.72 and 0.75, respectively.
3. Conclusions
The above considerations lead us to infer the following conclusions:1. The 4964 DIB carrier is spatially correlated with neutral methylidyne; this isconfirmed by both the analysis of Doppler velocities and by the correlationof N(CH) vs. EW(4964). The 6196 DIB, which seemed related to CH + based on Doppler shifts, correlates only marginally better with this speciesthan with 4964, in terms of N(CH + ) vs. EW(6196).2. The equivalent widths of discussed diffuse bands, as well as the molecularcolumn densities derived here, all correlate reasonably well with E(B-V),which makes the latter parameter not very useful in case of finding DIB car-riers originating in different interstellar environments.The good correlation of CH with the 4964 DIB suggests the existence of other,similar connections between simple molecules and DIBs. However, in order toshow the division of DIBs between different environments, one needs the spectraof very high S/N ratio. It could be very rewarding to check for the relations be-tween the strong DIBs and the interstellar molecules such as C or CO - based onnew samples, coming from homogeneous measurements. However, a majority ofDIBs show the profiles too broad to allow for a Doppler shift analysis. It is cer-tainly important to collect more spectra of the high signal-to-noise ratio, offeringthe correct column density values for simple diatomics, and thus indicating the pos-sible relations of observed molecules to DIBs carriers. In any case it seems to be ofbasic importance to relate the origin of diffuse bands to certain interstellar environ-ments, defined by specific physical parameters, the latter available from analyzesof identified atomic and/or molecular features. Acknowledgements.
JK and TW acknowledge the financial support pro-vided by the Polish National Center for Science for the period 2012 - 2015 (grantUMO-2011/01/BST2/05399). The authors benefited from the funds of the Polish-Ukrainian PAS/NASU joint resarch project No. 21(2009-2011). We are gratefulfor the assistance of the ESO and BOAO observatories staff members.REFERENCES
Allen, M.M., 1994, ApJ, 424, 754Bagnulo, S., Jehin, E., Ledoux, C., Cabanac, R., Melo, C., Gilmozzi, R., The ESO Paranal ScienceOperations Team, 2003, Msngr, 114, 10Bernath, P.: 2005, in: Spectra of Atoms and Molecules (Oxford Univ. Press) ol. 0 Cox, N.L.J. and Patat, F., 2008, A&A, 485, 9Crane, P., Lambert, D.L., Sheffer, Y., 1995, ApJS, 99, 107Ehrenfreund, P., Cami, J., Jiménez–Vincente, J., et al., 2002, A&A, 576, 117Federman, S.R., 1982, ApJ, 257, 125Galazutdinov, G.A., 1992, Prep. Spets. Astrof. Obs., No 92.Galazutdinov, G.A., Musaev, F.A., Krełowski, J., Walker, G.A.H., 2000, PASP, 112, 648Galazutdinov, G. A., Lo Curto, G., Krełowski, J., 2008, ApJ, 682, 1076.Gredel, R., van Dishoeck, E.F., Black, J.H., 1993, A&A, 269, 477Heger, M.L., 1922, Lick Obs. Bull. 10, 146Herbig, G. H., 1975, ApJ, 196, 129Herbig, G. H., 1995, ARAA, 33, 19Herzberg, G.: 1950, Molecular Spectra and Molecular Structure, Vol. 1: Spectra of DiatomicMolecules (New York: D. Van Nostrand Co.)Hobbs, L. M., York, D. G., Snow, T. P., Oka, T., Thorburn, J. A., Bishof, M., Friedman, S. D., McCall,B. J., Rachford, B., Sonnentrucker, P., Welty, D. E., 2008, ApJ, 680, 1256Jenniskens, P., Desert, F, -X., 1994, A&AS, 106, 39Ka´zmierczak, M., Gnaci´nski, P., Schmidt, M. R., Galazutdinov, G., Bondar, A., Krełowski, J., 2009,498, 785Kerr, T.H., Hibbins, R.E., Fossey, S.J., Miles, J.R., Sarre, P.J., 1998, ApJ, 495, 941Krełowski J., Beletsky Y., Galazutdinov G.A., Kołos R., Gronowski M., LoCurto G., 2010, ApJ, 714,L64Krełowski, J., Ehrenfreund, P., Foing, B.H., Snow, T.P., Weselak, T., Tuairisg, S.Ó, Galazutdinov,G.A., Musaev, F.A., 1999, A&A, 347, 235Krełowski, J., Galazutdinov, G., Kołos, R., 2011, ApJ, 735, 124Larsson M., Siegbahn P.E.M., 1983, Chem. Phys. 76, 175Maier, J.P., Walker, G.A.H., Bohlender, D.A., Mazzotti, F.J., Raghunandan, R., Fulara, J., Garkusha,I., & Nagy, A., 2011, ApJ, 726, 41Maier, J.P., Chakrabarty, S., Mazzotti, F. J., Rice, C. A., Dietsche, R., Walker, G. A. H., Bohlender,D. A., 2011, ApJ, 729, 20Mattila, K., 1986, A&A, 160, 157McKellar, A., 1940, PASP, 52, 187Pan, K., Federman, S.R., Cunha, K., Smith, V.V., Welty, D.E., 2004, ApJS, 151, 313Sarre, P.J., Miles, J.R., Kerr, T.H., Hibbins, R.E., Fossey, S.J., Sommerville, W.B., 1995a, MNRAS,277, 41Sheffer, Y., Rogers, M., Federman, S.R., Abel, N.P., Gredel, R., Lambert, D.L., Shaw, G., 2008, ApJ,687, 1075Thorburn, J. A., Hobbs, L. M., McCall, B. J., Oka, T., Welty, D. E., Friedman, S. D., Snow, T. P.,Sonnentrucker, P., York, D. G., 2003, ApJ, 584, 339van Dishoeck, E.F., & Black, J.H., 1989, ApJ, 340, 273Weselak, T., Schmidt, M., Krełowski, J., 2000, A&AS, 142, 239Weselak, T., Galazutdinov, G.A., Musaev, F.A., Krełowski, J., 2004, A&A, 414, 949Weselak, T., Galazutdinov, G.A., Musaev, F.A., Krełowski, J., 2008a, A&A, 479, 149Weselak, T., Galazutdinov, G.A., Musaev, F.A., Krełowski, J., 2008b, A&A, 484, 381Weselak, T., Galazutdinov, G.A., Musaev, F.A., Beletsky, Y., Krełowski, J., 2009a, A&A, 495, 189Weselak, T., Galazutdinov, G.A., Beletsky, Y., Krełowski, J.: 2009b, A&A, 499, 783Weselak, T., Galazutdinov, G.A., Beletsky, Y., Krełowski, J.: 2009c, MNRAS, 400, 392Weselak, T., Galazutdinov, G.A., Beletsky, Y., Krełowski, J.: 2010b, MNRAS, 402, 1991Weselak, T., Galazutdinov, G.A., Gnaci´nski, P., Krełowski, J.: 2014, AcA, 64, 277Zsargó, J., Federman, S.R., 2003, ApJ, 589, 319 A. A.
Table 1: Observational and measurement data. Are given: HD/BD number (b– BOAO, f – FEROS), Spectral type / Luminosity class, E(B-V) (in magnitudes),column densities (in 10 cm − ) of the CH + and CH molecules with error in eachcase, equivalent widths (in mÅ) of the 4964 and 6196 DIBs with error in each case. HD/BD Sp/L E(B-V) N( CH + ) s N( CH ) s EW(4964) s EW(6196) s ol. 0 cm − ) of the CH + and CH moleculeswith error in each case, equivalent widths (in mÅ) of the 4964 and 6196 DIBs witherror in each case. HD/BD Sp/L E(B-V) N( CH + ) s N( CH ) s EW(4964) s EW(6196) ss