Diffuse Interstellar Bands Toward Hd 62542 Abstract

Draft version December 3, 2003 Preprint typeset using LATEX style emulateapj v. 3/25/03

DIFFUSE INTERSTELLAR BANDS TOWARD HD 62542 Mat´ e´ Ad´ amkovics,´ 1 Geoffrey A. Blake, 2 and Benjamin J. McCall1,3 Draft version December 3, 2003

ABSTRACT Diffuse interstellar bands (DIBs) have been observed along the peculiar translucent line of sight toward HD 62542. Of the 15 DIBs observed, 10 are relatively weak features known to correlate with the carbon chain molecules C2 and C3. The DIBs λλ5780 and 5797 are detected but are significantly weaker than towards other lines of sight with similar reddening. The other commonly observed DIBs such as λλ4430, 6270, and 6284 remain noticeably absent. We discuss how these observations support the suggestion that the line of sight toward HD 62542 crosses the core of a diffuse cloud. A comparison of CH, CN, C2, and C3 column densities and C2 DIB strengths toward HD 62542, HD 204827, and HD 172028 suggests that the line of sight toward HD 204827 also passes through a diffuse cloud core. Subject headings: astrochemistry — ISM: lines and bands — ISM:molecules

1. introduction of atomic hydrogen, N(H i). Recently, a broad DIB sur- A great deal of the fascination with the diffuse inter- vey at Apache Point Observatory has shown that a group stellar bands (DIBs) arises from their ubiquity towards of 18 relatively weak DIBs correlate with C2 (Thorburn lightly reddened sightlines and the challenge of identify- et al. 2003), which also correlates with C3 (Oka et al. ing their molecular carriers (for recent reviews see Her- 2003; Adámkovics´ et al. 2003). These results provide big (1995), Tielens & Snow (1995) and Snow (2001)). support for the suggestion that some DIBs arise from car- Much like the total column densities of the interstellar bon chain molecules (Douglas 1977) or molecules closely molecules C2,C3, CN and CH, the strengths (i.e. equiv- chemically related to carbon chains. alent widths, Wλ(DIB)) of individual DIBs generally in- In diffuse clouds the H i concentration falls towards crease with extinction yet can be highly variable among the center of the cloud as the density increases and H i is sightlines with the same color excess, E(B − V ). This converted to H2. Likewise, one would expect λ5780 to be- indicates that the strengths of these features relative to come weaker toward the center of a diffuse cloud, so that one another are sensitive to some varying physical con- lines of sight probing the dense regions of diffuse clouds dition(s) in the diffuse interstellar medium (ISM) in ad- will have weaker λ5780 for a given color excess. This dition to simply the total amount of material along the was first emphasized by Snow & Cohen (1974) and has line of sight. An understanding of the mechanisms that been studied more thoroughly for λλ5780, 5797, 5850, determine the relative line strengths of the DIBs may 6196, 6203, 6269, 6284 in the Taurus clouds (Adamson turn the enigmatic DIBs into a widely applicable probe et al. 1994, 1991). The propensity for increased DIB con- of diffuse ISM properties. centration in the outer regions of diffuse clouds—known Correlations between individual DIB strengths have as the ‘skin’ phenomenon—is the opposite of the mod- been used to identify features that respond similarly to eled behavior for small carbon species such as C2, CH, differences in the diffuse cloud environment (see reviews and CN (van Dishoeck & Black 1986), which increase in above and references therein). The fundamental idea is concentration toward the center of clouds. that any group of features arising from a particular car- The line of sight toward HD 62542 offers a unique rier, or set of chemically related carriers, must maintain opportunity to study the molecular abundances in the the same relative intensities in all lines of sight. Identify- dense (n∼500 cm−3, Gredel et al. (1993)) central re- ing such correlating subsets of DIBs provides a valuable gion of a diffuse cloud. The diffuse cloud toward this spectroscopic constraint on the identity of the carrier(s). star has apparently been stripped of the lower density This method has the advantage of directly comparing outer layer(s)—that is, its ‘skin’—by stellar winds and DIB equivalent widths without the potentially compli- radiation pressure (Cardelli et al. 1990). Evidence for cating factor of normalizing individual DIB equivalent the missing outer layer(s) includes the unexpectedly high widths by some measure of the extinction along any par- molecular abundances for a line of sight with only 1 mag ticular line of sight. of visual extinction (Cardelli et al. 1990), the extreme There have been a few successful measurements of the far-UV extinction (Cardelli & Savage 1988), and recent correlation between DIB strengths and the column den- observations showing that the DIBs are conspicuously sity of identified species. Herbig (1993) showed that missing from this sightline (Snow et al. 2002). Further Wλ(5780) depends strongly on the total column density intriguing characteristics of the cloud toward HD 62542 include the very small column of CH+ (Cardelli et al. 1 Department of Chemistry, University of California, Berkeley, 1990; Gredel et al. 1993) and the unusually high ratio of CA, 94720; [email protected] ´ 2 Division of Geological and Planetary Sciences and Division of N(C3)/N(C2)(Adámkovics et al. 2003). Chemistry and Chemical Engineering, M/S 150-21, California In- One would expect that unlike the generally strong stitute of Technology, Pasadena, CA 91125. DIBs (e.g. λλ5780, 5797, and 6284), those that corre- 3 Astronomy Department, 601 Campbell Hall, University of Cal- late with carbon chain molecules (Thorburn et al. 2003) ifornia, Berkeley, CA 94720 2 AD´ AMKOVICS´ ET AL. would be located predominantly in the central regions Table 1. Measurements of HD 62542, HD 204827 and HD 172028 of diffuse clouds (diffuse cloud ‘cores’) where the C2 concentration increases. The C2 DIBs should therefore be observable in a sightline with a diffuse cloud core. 62542 204827 172028a 204827b Here we use the high sensitivity (S/N∼700-1000) spec- no ‘core’ tra of HD 62542 obtained during our measurements of C3 Type B5 V B0 V B2 V ··· (Adámkovics´ et al. 2003) to test this assumption, while V 8.04 7.94 7.83 ··· E(B−V ) 0.35 1.11 0.79 0.76 searching for the unusually weak DIBs in this sight line. AV 0.99 2.62 2.20 1.62 We report the detection toward HD 62542 of 10 of the 18 DIBs that correlate with C2 and C3, as well as the DIBs C2 DIB Equivalent Widths (mA)˚ λλ4727, 5494, 5544, 5780, and 5797. 4363.77 5.5 ± 1 15 ± 1 ··· 9.5 ± 1.4 2. observations and results 4726.27 blend blend ······ 4734.73 5 ± 0.5 17.8 ± 1 12.3 ± 0.8 12.8 ± 1.1 Observations were performed using the High Resolu- 4963.96 6 ± 1 50.2 ± 2 50 ± 1 44.2 ± 2.2 tion Echelle Spectrometer (HIRES) (Vogt et al. 1994) on 4969.12 3.7 ± 1.5 11.8 ± 2 ··· 8.1 ± 2.5 the 10-m Keck I telescope atop Mauna Kea on 2002 De- 4979.58 2.6 ± 0.5 14.7 ± 1 ··· 12.1 ± 1.1 4984.73 5.6 ± 0.5 30.8 ± 1 26 ± 1 25.2 ± 1.1 cember 14, with details presented in Adámkovics´ et al. 5003.62 <2 16.9 ± 1 ··· 15.9 ± 2.2 (2003). Briefly, spectra in the 3985-6420 A˚ range were 5170.44 <2 11.3 ± 1 ··· 10.3 ± 2.2 recorded with small gaps in the coverage starting at 5175.99 3.7 ± 0.5 32.4 ± 1 23 ± 2 28.8 ± 1.1 5418.91 4.5 ± 0.5 52.6 ± 2 ··· 48.1 ± 1.6 ∼5000 A˚ and increasing to ∼30 A˚ at 6315 A.˚ The re- 5512.62 <2 24.0 ± 1 26 ± 1 23.0 ± 2.2 solving power was R∼67,000 and the reduced data have 5541.77 <2 25.3 ± 2 ··· 24.3 ± 2.8 a dispersion ranging from d=0.0287 A˚ pixel−1 at 4050 A˚ 5546.46 <2 17.2 ± 1 ··· 16.2 ± 2.2 ˚ −1 ˚ 5762.74 <2 18.3 ± 2 ··· 17.3 ± 2.8 to d=0.0431 A pixel at 6265 A. 5769.08 3.3 ± 1 17.6 ± 1 17 ± 1.5 14.3 ± 1.4 The data were reduced using standard procedures in 5793.14 <2 22.4 ± 2 ··· 21.4 ± 2.8 the echelle package in NOAO’s Image Reduction and Analysis Facility (IRAF). Calibration and target images Other DIB Equivalent Widths (mA)˚ were bias corrected and combined, before target images were flat fielded and echelle apertures extracted. Wave- 4727.06 blend blend ······ length calibration was performed with standard routines 5494.14 5.4 ± 1 22.9 ± 1 ··· 17.5 ± 1.4 5544.97 4.1 ± 1 25.8 ± 1 26 ± 1 21.7 ± 1.4 using spectra of a ThAr lamp, and one-dimensional spec- 5766.25 <2 28 ± 1 ··· 27.0 ± 2.2 tra of the targets were extracted for further analysis. 5780.59 52.8 ± 4 266 ± 5 256 ± 8 213 ± 6 Figure 1 shows the spectrum of HD 62542 in the regions 5797.11 8.5 ± 1 158 ± 4 217 ± 5 150 ± 4 6270.06 <2 80 ± 3 45 ± 5 79.0 ± 3.6 where DIBs have been detected, along with the spectrum 6284.31 <8 518 ± 60 450 ± 60 514 ± 61 of HD 204827 for comparison. Equivalent widths are measured by numerical integration and are reported in Column Densities (×1012cm−2) Table 1 with 1σ errors and 2σ upper limits. The detec- + c d tion of CH toward HD 204827 and tentative detection N(C2) 80 ± 20 440 ± 29 290 ± 26 360 ± 35 e e f toward HD 62542 are shown in Figure 2. Since we have N(C3) 10.5 ± 0.5 11.5 ± 0.8 3.6 ± 0.6 1.0 ± 1.0 observed only the strong 4232 A˚ line of CH+ toward HD N(CH) 38 ± 2g 80 ± 10d 53 ± 6 42 ± 10 204827 that is likely saturated, we report only a lower N(CN) 41.7 ± 14.7g 55 ± 5d 28 ± 5 13 ± 15 + limit for N(CH+) toward HD 204827. N(CH ) <1.9 >33.6 24 ± 2.2 >31.7

a 3. discussion All values for HD 172028 from Thorburn et al. (2003) unless otherwise noted. b 3.1. C2 DIBs and the Skin Phenomenon Shown in this column are the DIB equivalent widths and molecular column densities measured toward HD 62542, representing a The implications of the weakness of the generally central core, subtracted from the same quantities measured toward strong DIBs λλ5780, 5797, 6270 and 6284 in HD 62542 HD 204827. have already been articulated by Snow et al. (2002). In cGredel et al. (1993) d short, the strength of these DIBs is consistent with the Thorburn et al. (2003) e ´ idea that most DIB carriers exist in the outer ‘skin’ re- Adámkovics et al. (2003) fOka et al. (2003) gions of diffuse clouds, which in the HD 62542 sightline gCardelli et al. (1990) have been stripped away by stellar winds and radiation pressure. We further emphasize their results by plac- ing upper limits on λλ6270 and 6284 that are factors of 6 and 15, respectively, lower than previously reported. of C2 DIBs toward HD 62542 differentiate the chemical It is clear that the conditions which favor the presence environment in which the C2 DIB carriers exist from that of the carriers of λλ6270 and 6284 toward most diffuse of other DIBs. Whereas the DIB carriers had previously clouds do not exist toward HD 62542. Our detection been thought to exist on the ‘skin’ of diffuse clouds, the of predominantly those DIBs that correlate with carbon present results suggest that some of the C2 DIB carriers chain molecules remains consistent with the picture of a can survive in the central high density regions. However, missing ‘skin’. these results do not rule out the presence of these carriers These results further support grouping of the C2 DIBs in lower density regions as well. together as features arising from a single carrier or a Comparison with HD 204827 shows that, per unit set of closely chemically related carriers. Measurements E(B −V ), the C2 DIBs are in general a factor of a few

DIBS TOWARD HD 62542 3

132K476987:6

(+/

()+*

¢J

;3<>=?A@9BC=9D

# %

¦I¢ ¨¤ ¦I¢ £§ ¦I£ £¤ ¢ ¨¢¤ ¢ ¨¦§ ¢ £I£¤ ¢ £I¢§ ¢¡¢§£§ ¢¡¢ ¨¤ ¢¡¢ £§ ¢¡£ £¤ ¢¡£ ¢§ ¢¡¦¥©¤ ¢¡¦¥¨§ ¢¡£¡£¤

LNMOQP RSPTU©VXW_Y [`]

1325476987:A6

(+/

- .

()+*

¦F

¢G

# %

;3<>=?A@9BC=9D

©H

§¨¤£¤¨¤ §¨¤£¤£§ §¨¤EA¤ § £¤ § ¢§ § 9¥©¤ § ¢¢¤ § ¢¦§ §¨§¨¤¢§ §¨§¤ §¨§§ §¨§ § §¨§£§¨¤

1325476987:6

(+/

- .

()+*

# $%

;3<>=?A@9BC=9D

§© ¢ ¨¤ §© ¢ £§ §© £ £¤ §© £ ¢§ §© ¦¥©¤ §© ¦¥¨§ §© £¡£¤ §© £¡¢§ ©¢ £§ ©£ £¤ ©£ ¢§ ©¦¥©¤ ©¦¥¨§ ©£¡£¤ ©£¡¢§ LNMOQP RSPTU©VXWZY\[^]

Fig. 1.— Detections of diffuse bands toward HD 62542 along with spectra of HD 204827 for comparison. Arrows indicate the locations of DIBs, with the C2 DIBs highlighted with ‘C’s, and stellar lines in HD 62542 identified by ‘*’s. Note that the vertical scales are different for the two stars. stronger toward HD 204827 than HD 62542. λλ4364, mated using molecular proxies. 4734 and 4969 are exceptions where the Wλ(DIB)/E(B− We use the relationship between Wλ(5780) and V ) ratio is nearly the same in both stars. The broad N(H i) measured for a large number of diffuse sight velocity dispersion of C3 (Adámkovics´ et al. 2003) and lines (York et al. 2004, in prep.) and calculate multiple velocity components of CH+ (Figure 2) toward N(H i)=3.6×1020cm−2 toward HD 62542. Likewise, one 21 −2 HD 204827 indicate absorption through multiple diffuse can calculate N(H2)=1.3×10 cm from N(CH) yield- clouds, in contrast with the HD 62542 line of sight. If ing the very high f(H2)=0.88, which is consistent with the line of sight toward HD 62542 is an environment like the outer layer(s), rich in H i, having been stripped the central core of a cloud, the line of sight toward HD away from a core that is predominantly H2. Using the 20 −2 204827 could cross both the ‘skin’ and central core ma- observed N(H2)=6.5×10 cm (Rachford et al. 2002) terial for multiple clouds. with N(H i) derived from Wλ(5780) yields f(H2)=0.78. Since C2 correlates with C2 DIBs, it is not surpris- In either case, the molecular fraction reported here is ing that N(C2)/E(B − V ) is greater in HD 204827 by significantly higher than f(H2)=0.60 reported by Rach- nearly a factor of two as well, however this contrasts ford et al. (2002) and discussed by Snow et al. (2002), with the column densities of CH, CN, and C3, which per which was determined with measurements of H2 and an unit E(B −V ) are stronger toward HD 62542. This is estimate of NH from E(B−V ). discussed further below, and raises the possibility that Our analysis can also be performed for stars that are Wλ(C2 DIB)/E(B−V ) is smaller toward HD 62542 be- too faint in the UV for H2 and H i measurements, such as cause C , and likewise the C DIBs, have reacted to form HD 204827, for which we infer N(H i)=1.8×1021cm−2, 2 2 21 −2 other species (e.g. CN and C3). N(H2)=2.6×10 cm , and f(H2)=0.74. This value is also higher than for typical diffuse clouds and suggests 3.2. Molecular Column Densities that HD 204827 may also have a dense core rich in H2. The column densities of many molecular species to- Cardelli et al. (1990) have shown that HD 62542 and ward HD 62542 and HD 204827 have been previously re- HD 204827 have remarkably similar FUV extinction ported (Table 1) and to these measurements we add our curves—the main difference being a weaker 2200 A˚ bump + observations of CH (Figure 2). Since Wλ(5780) corre- toward HD 62542. Both have exceptionally steep FUV lates with N(H i) (Herbig 1993), as well as N(H2) with extinction when compared to typical diffuse clouds such N(CH)(Cardelli & Wallerstein 1986; Federman 1982; Fe- as those toward ζ Per. If, as Cardelli et al. (1990) sug- derman et al. 1984; Rachford et al. 2002), the fractional gest, the high column densities of CH and CN in HD abundance of molecular hydrogen f(H2), may be esti- 62542 are due to the decreased photodissociation rates 4 AD´ AMKOVICS´ ET AL.

outer layer(s), the question arises as to whether the ma- 1.00 + terial along this sightline is analogous to cores that may * CH exist in diﬀuse clouds with ‘skins’, or if the exposure of 0.99 the central core has signiﬁcantly altered molecular abun- HD 62542 dances. To test this question we have subtracted the DIB 0.98 * line strengths and molecular column densities observed 0.97 toward HD 62652 from the same quantities measured toward HD 204827; in the absence of chemical alteration of 1.00

Normalized Flux HD 62542 material due to exposure, the calculated dif- 0.95 ference should then be representative of a diffuse cloud without a core. 0.90 The calculated value for a cloud without a core can HD 204827 0.85 be compared with measurements of HD 172028 from the Apache Point DIB survey (Thorburn et al. 2003). This 4231 4232 4233 4234 4235 sightline was selected simply because the E(B−V ) is close Wavelength [Å] to the difference in E(B−V ) between HD 204827 and HD 62542, and measurements of molecular column densities Fig. 2.— CH+ toward HD 204827 and tentative detection to- have been made. Table 1 shows that the strengths of the ward HD 62542. Column densities of CH+ have been calculated using the oscillator strength f=5.45×10−3. The intricate and un- DIBs and molecular column densities toward our hypo- usual structure of the stellar lines (marked by ‘*’s) in HD 62542 thetical ‘diffuse cloud without a core’ are similar to those complicates the measurement of the CH+ equivalent width. The in HD 172028. The agreement becomes slightly better CH+ line is likely blended with a stellar line and we report only in many cases when taking into account the somewhat an estimated upper limit (assuming the FWHM of the CH+ fea- ˚ larger E(B−V ) toward HD 172028. ture would be ∼0.1 A) by integrating the total observed absorption There is less C and more C toward HD 172028 than between 4232.44 A˚ and 4232.64 A.˚ 2 3 our fabricated cloud, which would be consistent with the destruction of C2 and formation of C3 toward HD 62542. Conversely the cloud toward HD 172028 may have of these molecules from the attenuated flux shortward of a ‘core-like’ component indicated by the lower than ex- ∼3000 A,˚ then similar column densities per unit color expected N(C2) and higher than expected N(C3), N(CH), cess should be observed toward HD 204827. However, the and N(CN). The general agreement in molecular column measurements in Table 1 show that this is not the case; densities and DIB lines strengths is striking and suggests while E(B−V ) and AV are larger toward HD 204827 than that the sightline toward HD 62542 may be similar to the HD 62542 by factors of 3.2 and 2.6, N(CN) and N(C3) environment found in the center of other diffuse clouds. ratios in the two stars are nearly the same and N(CH) is larger in HD 204827 by a factor of only 2.1. This, and the order of magnitude difference in N(CH+), indicate that the chemistry in these sightlines cannot be explained by We acknowledge helpful conversations with D. E. Welty, the shape of the extinction curves alone. Perhaps it is T. P. Snow and D. G. York. The data presented herein important that the color excess toward HD 204827 arises were obtained at the W.M. Keck Observatory, which is oper- from multiple clouds. ated as a scientific partnership among the California Institute of Technology, the University of California and the National 3.3. Diffuse cloud cores Aeronautics and Space Administration, and was made pos- sible by the generous financial support of the W.M. Keck Since many observations are consistent with the dif- Foundation. BJM is supported by the Miller Institute for fuse cloud toward HD 62542 having been stripped of its Basic Research in Science.

REFERENCES Adámkovics,´ M., Blake, G. A., & McCall, B. J. 2003, ApJ, 595, Rachford, B. L., Snow, T. P., Tumlinson, J., Shull, J. M., Blair, 235 W. P., Ferlet, R., Friedman, S. D., Gry, C., Jenkins, E. B., Adamson, A. J., Kerr, T. H., Whittet, D. C. B., & Duley, W. W. Morton, D. C., Savage, B. D., Sonnentrucker, P., Vidal-Madjar, 1994, MNRAS, 268, 705 A., Welty, D. E., & York, D. G. 2002, ApJ, 577, 221 Adamson, A. J., Whittet, D. C. B., & Duley, W. W. 1991, MNRAS, Snow, T. P. 2001, Spectrochimica Acta, 57, 615 252, 234 Snow, T. P., & Cohen, J. G. 1974, ApJ, 194, 313 Cardelli, J. A., Edgar, R. J., Savage, B. D., & Suntzeff, N. B. 1990, Snow, T. P., Welty, D. E., Thorburn, J., Hobbs, L. M., McCall, ApJ, 362, 551 B. J., Sonnentrucker, P., & York, D. G. 2002, ApJ, 573, 670 Cardelli, J. A., & Savage, B. D. 1988, ApJ, 325, 864 Thorburn, J. A., Hobbs, L. M., McCall, B. J., Oka, T., Welty, Cardelli, J. A., & Wallerstein, G. 1986, ApJ, 302, 492 D. E., Friedman, S. D., Snow, T. P., Sonnentrucker, P., & York, Douglas, A. E. 1977, Nature, 269, 130 D. G. 2003, ApJ, 584, 339 Federman, S. R. 1982, ApJ, 257, 125 Tielens, A. G. G. M., & Snow, T. P., eds. 1995, The Diffuse Federman, S. R., Danks, A. C., & Lambert, D. L. 1984, ApJ, 287, Interstellar Bands (Dordrecht: Kluwer) 219 van Dishoeck, E. F., & Black, J. H. 1986, ApJS, 62, 109 Gredel, R., van Dishoeck, E. F., & Black, J. H. 1993, A&A, 269, Vogt et al. 1994, Proc. Soc. Photo-Opt. Instr. Eng., 2198, p.362 477 Herbig, G. H. 1993, ApJ, 407, 142 —. 1995, ARA&A, 33, 19 Oka, T., Thorburn, J. A., McCall, B. J., Friedman, S. D., Hobbs, L. M., Sonnentrucker, P., Welty, D. E., & York, D. G. 2003, ApJ, 582, 823