Dust Extinction in the Diffuse Interstellar Medium

Dust extinction measurements provide important constraints on the size ,composition, shape, and abundance of dust grains and an empirical model to account of the effects of extinction on astrophysical objects. For decades our understanding of dust grains was strongly biased by measurements in our Galaxy and the ultraviolet (UV). The UV bias is due to the extensive spectroscopic observations taken with the IUE satellite revealing the details of the 2175 A bump, far-UV rise, and underlying extinction continuum. I will discuss the results of a dedicated effort to expand our spectroscopic measurements of dust extinction to the far-UV, optical, near-infrared, and mid-infrared wavelength regimes. This work has revealed new optical extinction features, enabled the first combined combined study of UV and MIR extinction features, shown the possible presence of ice in the diffuse interstellar medium, and revealed an intriguing correlation between UV extinction and molecular hydrogen. Building on these works, a new R(V) dependent extinction relationship at spectroscopic resolution from 912 A to 32 microns has been determined. Moving out of our Galaxy, in progress work shows that the 2175 A bump is rare in an expanded sample of UV extinction curves and M31 and M33 show UV extinction curves quite similar to those seen our our Galaxy. Finally, prospects for future work especially with HST and JWST will be presented.

Transcript

1. Dust Extinction in the
   Diffuse Interstellar Medium
   Karl D. Gordon, Astronomer
   STScI, Baltimore, MD, USA
   Cardiff University
   School of Physics & Astronomy
   19 Apr 2023
   kgordon@stsci.edu
   @astrodon.social
   karllark@github
   “Have Dust – Will Study”
   Slides on speakerdeck
   

   View Slide

2. Summary
   ●
   Extinction measurements are fundamental to understanding dust
   ●
   Spectroscopic Extinction from FUV to MIR in Milky Way
   – Overall quite smooth
   – Few broad features including some new ones
   – One intriguing correlation, comforting non-correlations
   – One R(V) extinction relationship for all wavelengths
   ●
   Beyond Milky Way (if there is time)
   – Mostly like the the Milky Way, but not all
   ●
   Future w/ HST and JWST
   – More galaxies, more environments, larger samples
   

   View Slide

3. Extinction
   ●
   Critical clues to dust grain abundance, size, composition,
   and shape
   ●
   Clues in the continuum and features
   ●
   Straightforward measurement
   ●
   Focus on spectroscopic measurements
   ●
   Biased by my views, not comprehensive
   

   View Slide

4. Not Covered
   ●
   Dust scattering (e.g., albedo, scattering phase function)
   ●
   Dust emission
   ●
   Atomic composition of dust (e.g., depletions)
   ●
   Dust Polarization
   ●
   Dust grain models
   ●
   All important, but not the focus of this talk
   

   View Slide

5. Extinction (not Attenuation)
   ●
   Extinction
   – Absorption and scattering out of
   line-of-sight
   – Specific to a point source dust
   – Proportional to dust grain properties
   ●
   Attenuation
   – scattering into the line-of-sight
   – Varying extinction to stars
   – Applies to galaxies, circumstellar
   dust, etc.
   

   View Slide

6. Extinction vs
   Attenuation
   Github: karllark/dust_attenuation
   Github: karllark/dust_extinction
   

   View Slide

7. Pair Method
   Github: karllark/measure_extinction
   OB stars ideal targets
   Luminous at all wavelengths
   Fairly simple atmospheres
   

   View Slide

8. Basic measurement is the color “excess” versus wavelength
   (in magnitudes as we are astronomers)
   Normalize so measurements with
   different dust columns can be compared
   A(V) determined by extrapolating to
   infinite wavelength
   e.g.,
   

   View Slide

9. Slide from Dries Van De Puttte
   

   View Slide

10. Wavelength axis variations
    Allow full FUV-MIR extinction to be seen Wavelength scale is proportional to energy
    Emphasizes UV wavelengths
    Versus λ Versus 1/λ
    UV Opt NIR
    MIR
    

    View Slide

11. Diffuse vs Dense Dust
    ●
    Diffuse sightlines
    – Most detailed extinction studies
    – Generally A(V) values less than a few
    – No 3.0 μm H2
    0 ice feature
    ●
    Dense sightlines
    – Generally studied in the NIR/MIR only
    – Have 3.0 μm H2
    0 ice feature
    Decleir et al. (2022, ApJ, 930, 15)
    H20 Ice
    

    View Slide

12. Our Galaxy
    

    View Slide

13. Stecher 1969, ApJ, 157, 125
    

    View Slide

14. International Ultraviolet Explorer
    18 years! 100,000 UV spectra!
    

    View Slide

15. View Slide

16. Witt, Bohlin, & Stecher 1984, 279, 698
    

    View Slide

17. 2175 A Bump
    width → strong variation
    center → almost no variation
    Fitzpatrick & Massa (1986, ApJ, 307, 286)
    

    View Slide

18. Fitzpatrick & Massa 1988, ApJ, 328, 734
    

    View Slide

19. View Slide

20. Cardelli, Clayton, & Mathis 1988, ApJ, 329L, 33; 1989, ApJ, 345, 245
    

    View Slide

21. Valencic, Clayton, & Gordon 2004, ApJ, 616, 912
    

    View Slide

22. Far Ultraviolet
    Spectroscopic
    Explorer
    905 to 1195 Å
    

    View Slide

23. Pre-FUSE
    ORFEUS: Sasseen et al. 2002, ApJ, 566, 267
    Voyager: Snow et al. 1990, ApJ, 399, L23
    See also: Buss et al. 1994, France et al. 2004; Lewis et al. 2005
    

    View Slide

24. Gordon, Cartledge, & Clayton 2009, ApJ, 705, 1320
    See also: Sofia et al. 2005
    

    View Slide

25. FUV (& all UV) extinction smooth
    Residuals due to stellar or ISM HI
    FUV rise consistent with feature
    peaking at ~800 Å
    

    View Slide

26. ApJ, 944, 33
    Dries Van De Putte
    

    View Slide

27. FUV rise extinction component & H2
    column
    Very strong correlation & goes through zero!!!
    Grain responsible for FUV rise and H2
    co-spatial
    Not the case for the
    2175 Å bump
    

    View Slide

28. Hubble Space
    Telescope
    912 Å to 2.5 μm
    

    View Slide

29. Pre-Hubble
    ●
    Continuum measured mainly via
    photometry (until 2019)
    ●
    Except for Orion dust
    – Cadelli & Clayton (1988, AJ, 95, 516)
    ●
    Very Broad Structure with low
    resolution spectra
    – Next slide
    

    View Slide

30. Very Broad Structure
    = Deviations from linear w/ 1/lambda
    Whiteoak 1966, ApJ, 144, 305
    Hayes 1973, IAUS, 53, 83
    

    View Slide

31. View Slide

32. Broad Optical Features (origin unknown)
    Long known Very Broad Structure Explained!
    Massa, Fitzpatrick, Gordon, et al. 2020, ApJ, 891, 67
    Centers: 4370, 4870, & 6300 Å
    Widths: ~10%
    Two blue correlate w/ 2175 Å
    

    View Slide

33. NIR/MIR Extinction
    ●
    Continuum measured via photometry (until 2021)
    ●
    Features measured spectroscopically
    – Narrow wavelength ranges or towards “complicated” stars
    – 3.0 μm ice
    – 3.4 μm hydrocarbon
    – 10/20 μm silicate features
    

    View Slide

34. ●
    Many ice features
    ●
    3.0 μm H2
    0 the strongest
    ●
    Requires A(V) > 3
    Boogert, Gerakines, & Whittet
    

    View Slide

35. 3.4 μm
    Hydrocarbon grains
    Pendelton & Allamandola (2002, ApJS, 138, 75)
    

    View Slide

36. 10 & 20 μm
    Silicate
    Features
    Chiar & Tielens (2006, ApJ, 637, 774)
    

    View Slide

37. Crystalline Silicate Fraction
    ~0.2% (or zero)
    Kemper, Viernd, & Tielens (2004, ApJ, 609, 826)
    Amorphous Crystalline
    

    View Slide

38. NASA’s Infrared Telescope Facility
    0.8 to 20 μm
    

    View Slide

39. Marjorie Decleir
    

    View Slide

40. 3.0 μm ice feature measured in the diffuse average at
    A(ice)/A(V) = 0.0019 +/- 0.007
    Not a significant detection, but intriguing at the level predicted by
    Potapov et al. (2021) if ice is present in shadowed pits in silicate
    grains
    

    View Slide

41. Spitzer Space
    Telescope
    3 to 160 μm
    

    View Slide

42. View Slide

43. View Slide

44. Variations and 1st Direct Comparison of UV and MIR
    Strong variation
    Not correlated
    with grain size
    Silicates not correlated
    with 2175 A bump
    

    View Slide

45. Putting all the
    Wavelengths Together
    

    View Slide

46. ApJ, in press
    Based on spectroscopy
    at all wavelengths
    

    View Slide

47. Fitting a Line
    New penultimate technique*
    ●
    Fully accounts for correlated x &
    y uncertainties
    ●
    Allows for different correlations
    for different points
    ●
    Line integral for each point
    *The authors await knowledge of what paper to cite where this was first shown (likely from decades ago)
    

    View Slide

48. Average
    Variation with
    R(V)/grain size
    

    View Slide

49. Significant
    Deviations from
    Literature
    Curves at
    Select R(V)
    

    View Slide

50. MW Extinction Summary
    ● Spectroscopic from FUV to MIR
    – FUV smooth & consistent with feature peaking ~800 Å
    – H2
    correlates with FUV rise → co-spatial!
    – New optical features → unknown origin
    – NIR powerlaw & may contain H2
    0 ice
    – MIR versus UV features → 2175 Å not due to silicates
    – R(V) dependent extinction relationship → one
    relationship for all wavelengths
    

    View Slide

51. Beyond Milky Way
    

    View Slide

52. Clayton & Martin 1985, ApJ, 288, 558
    Add figure
    showing 30
    Dor/LMC2
    

    View Slide

53. Misselt et al. (1999, ApJ, 515, 128)
    

    View Slide

54. Spatial Distribution
    Misselt et al. (1999, ApJ, 515, 128)
    

    View Slide

55. Prevot et al. 1984, A&A, 132, 389
    

    View Slide

56. SMC Extinction Curves
    Milky Way-like!
    (2175 Å bump)
    4 similar curves are found in
    the star forming bar of the SMC!
    Ha image of the SMC
    Gordon & Clayton (1998, ApJ, 500, 816)
    Gordon et al. (2003, ApJ, 594, 279)
    STIS
    

    View Slide

57. SMC AzV 456
    SMC Bar
    LMC2
    LMC General
    MW
    Quiescent
    Processed
    Continuum of
    Properties
    Gordon et al. (2003, ApJ, 594, 279)
    Gordon et al. (2016, ApJ, 826, 104)
    

    View Slide

58. View Slide

59. Using MW HI 21cm
    

    View Slide

60. Spatial Distribution 2175 A bump
    Yes / No
    Image:
    MIPS 24um
    

    View Slide

61. Clayton et al. 2015, ApJ, 815, 14
    

    View Slide

62. M31/M33 Extinction
    ●
    All w/ HST photometry
    ●
    Similar to MW
    ●
    Radial, metallicity variations
    ●
    1st measurements in M33
    Petia Yanchulov Merica-Jones
    Clayton, G et al., in prep
    Yanchulov M-J, P. et al., in prep
    

    View Slide

63. Beyond the MW Summary
    ●
    Large Magellanic Cloud (½ solar)
    – Most of LMC similar to MW
    – Differences in LMC2 Supershell (near 30 Dor)
    ●
    Small Magellanic Cloud (1/5 solar)
    – Most very different from MW (very steep, no bump)
    – Small fraction, closer to MW (flatter w/ bump)
    ●
    M31 (~solar) and M33 (½ solar)
    – Similar to MW (preliminary)
    

    View Slide

64. Future
    

    View Slide

65. HST
    ●
    More galaxies
    – Investigate low metallicity galaxies
    ●
    SMC unique?
    – Expand samples in large galaxies
    ●
    Milky Way R(V) dependent extinction relationship works?
    ●
    High/low R(V)
    – Probe extremes of dust size, curvature in R(V) relationship?
    

    View Slide

66. JWST
    ●
    Two cycle 1 programs (PIs: Decleir & Zeegers)
    – 21 Milky Way sightlines from 2.4-28 μm
    – Continuum and 3.4, 10, & 20 μm features
    ●
    Map features photometrically
    ●
    LMC/SMC continuum & silicate measurements
    ●
    Other galaxies (sensitivity?)
    

    View Slide

67. Photometric Feature Measurements
    10 μm silicate
    3 μm ice
    3.4 μm
    hydrocarbon
    Burcu Günay
    

    View Slide

68. Summary
    ●
    Extinction measurements are fundamental to understanding dust
    ●
    Spectroscopic Extinction from FUV to MIR in Milky Way
    – Overall quite smooth, few broad features, DIBs (dust?) only narrow features
    – New broad features found in the optical
    – Intriguing correlations (or not) between features and gas tracers (esp. H2
    )
    – One R(V) extinction relationship for all wavelengths
    ●
    Beyond Milky Way
    – Most of LMC, small fraction of SMC, M31, & M33 → similar to Milky Way
    – SMC & LMC2 Supershell region → weaker/non-existent bumps, stronger UV slope
    ●
    Future w/ HST and JWST
    – More galaxies, more environments, larger samples
    

    View Slide

69. Thanks
    

    View Slide