H 3 + : A Case Study for the Importance of Molecular Laboratory Astrophysics Ben McCall Dept. of ChemistryDept. of Astronomy

H 3 + : Cornerstone of Interstellar Chemistry

H 3 + in Dense Clouds R(1,1) u R(1,0) R(1,1) l Wavelength (Å) N(H 3 + ) ~ 3  cm -2 Consistent with expectations McCall, Geballe, Hinkle, & Oka ApJ 522, 338 (1999) UKIRTKitt Peak Nature 384, 334 (1996)

Role of Laboratory Astrophysics Four and a half years – much of it assembling the IR laser system and discharge cell Scanned from: –6/12-8/3 (1978) –12/18-1/26 ( ) –4/24-12/18 (1980) Success on April 25, 1980

Surprise: H 3 + in Diffuse Clouds! observed at UKIRT observed at Kitt Peak ~ 4 × cm -2 ! Similar column density to dense clouds!! B. J. McCall, T. R. Geballe, K. H. Hinkle & T. Oka, Science 279, 1910 (1998) Cygnus OB2 12

Rate = k e [H 3 + ] [e - ]  [H 2 ] Diffuse Cloud H 3 + Chemistry H 2 H e - H 2 + H 2 +  H H cosmic ray H e -  H + H 2 or 3H Rate = Formation Destruction [H 3 + ]  = keke [e - ] Steady State [H 2 ] = (3  s -1 ) (5  cm 3 s -1 )  (2400) = cm -3 Density Independent If L ~ 3 pc ~ cm, expect only N(H 3 + ) ~ cm -2 !

H 3 + in Lots of Diffuse Clouds! HD McCall, et al. ApJ 567, 391 (2002) Cygnus OB2 12

Big Problem with the Chemistry! Steady State: [H 3 + ]  = keke [e - ] [H 2 ] To increase the value of [H 3 + ], we need: Smaller electron fraction [e - ]/[H 2 ] Smaller recombination rate constant k e Higher ionization rate  >1 order of magnitude!!

 Persei H 3 + toward  Persei Cardelli et al. ApJ 467, 334 (1996)Savage et al. ApJ 216, 291 (1977) N(H 2 ) from Copernicus N(C + ) from HST [e - ]/[H 2 ] not to blame McCall, et al. Nature 422, 500 (2003)

Big Problem with the Chemistry! Steady State: [H 3 + ]  = keke [e - ] [H 2 ] To increase the value of [H 3 + ], we need: Smaller electron fraction [e - ]/[H 2 ] Smaller recombination rate constant k e Higher ionization rate 

Enigma of H 3 + Recombination Laboratory values of k e have varied by 4 orders of magnitude! Problem: not measuring H 3 + in ground states

Ion Storage Ring Measurements 20 ns 45 ns electron beam H3+H3+ H, H 2 +Very simple experiment +Complete vibrational relaxation +Control H 3 + – e - impact energy +Rotationally cold ions from supersonic expansion source CRYRING 30 kV 900 keV 12.1 MeV

CRYRING Results Considerable amount of structure (resonances) in the cross-section k e = 2.6  cm 3 s -1 Factor of two smaller McCall, et al. Phys. Rev. A 70, (2004)

Agreement with Other Work Reasonable agreement between: –CRYRING Supersonic expansion –TSR 22-pole trap –Theory S.F. dos Santos, V. Kokoouline and C. H. Greene, J. Chem. Phys 127 (2007)

Big Problem with the Chemistry! Steady State: [H 3 + ]  = keke [e - ] [H 2 ] To increase the value of [H 3 + ], we need: Smaller electron fraction [e - ]/[H 2 ] Smaller recombination rate constant k e Higher ionization rate 

(7  cm -2 ) Implications for  Persei [H 3 + ]  L =  keke N(e - ) N(H 2 ) = = L N(H 3 + ) N(H 2 ) N(e - )  L = 5300 cm s -1 keke N(H 3 + ) (1.6  cm 3 s -1 ) (4.7  ) Adopt  =3  s -1 Adopt  n  = 215 cm -3 → L=2.4 pc L = 60 pc  n  = 9 cm -3  =7.4  s -1 (25x higher!) (firm) (dense cloud value) Similar results in many other sightlines N. Indriolo, T. R. Geballe, T. Oka, & B. J. McCall, ApJ 671, 1736 (2007)

Surprise → Conventional Wisdom Higher  in diffuse (vs. dense) clouds initially greeted with “skepticism” Incorporated into models without incident Now generally accepted (but not understood!)

Low Energy CRs? Could there be a large flux of low energy cosmic rays? M.D. Stage, G. E. Allen, J. C. Houck, J. E. Davis, Nat. Phys. 2, 614 (2006) T. E. Cravens & A. Dalgarno, ApJ 219, 750 (1978) AVAV 1 MeV 2 MeV 10 MeV 20 MeV 50 MeV (diffuse) (dense)

Cosmic Ray Observations W.R.Webber, ApJ 506, 329 (1998) Inferred interstellar

Theoretical Spectra

Consider This Spectrum

Inferred Ionization Rate  dense ~6× s -1

Inferred Ionization Rate  dense ~6× s -1  diffuse ~3.1× s -1 See poster (Nick Indriolo)

Summary H 3 + surprisingly abundant in diffuse clouds –Enabled by laboratory spectroscopy H 3 + is now a direct probe of ionization rate –Enabled by storage ring measurements & theory Ionization rate ~10× higher than thought –only in diffuse clouds –would still be unknown if not for laboratory astrophysics work Two proposed explanations –MHD self-confinement (Padoan & Scalo) –high flux of low energy cosmic rays

Acknowledgments Nick Indriolo (U. Illinois) Brian Fields (U. Illinois) Tom Geballe (Gemini) Takeshi Oka (U. Chicago) NASA Laboratory Astrophysics NSF Divisions of Chemistry & Astronomy

Astronomer's Periodic Table H He CN O Ne Mg Fe SiS Ar

Observing Interstellar H 3 + Equilateral triangle No rotational spectrum No electronic spectrum Vibrational spectrum is only probe Absorption spectroscopy against background or embedded star 1 2

Interstellar Cloud Classification* Diffuse clouds: H ↔ H 2 C  C + n(H 2 ) ~ 10 1 –10 3 cm -3 –[~ atm] T ~ 50 K  Persei Photo: Jose Fernandez Garcia Diffuse atomic clouds –H 2 << 10% Diffuse molecular clouds –H 2 > 10% (self-shielded) * Snow & McCall, ARAA, 44, 367 (2006) Barnard 68 (courtesy João Alves, ESO) Dense molecular clouds: H  H 2 C  CO n(H 2 ) ~ 10 4 –10 6 cm -3 T ~ 20 K

 [H 2 ] Dense Cloud H 3 + Chemistry H 2 H e - H 2 + H 2 +  H H cosmic ray H CO  HCO + + H 2 Rate = Formation Destruction Rate = k [H 3 + ] [CO]  [H 3 + ] = k [CO] Steady State = (3  s -1 ) (2  cm 3 s -1 ) [H 2 ]  (6700) = cm -3 Density Independent! (fast) McCall, Geballe, Hinkle, & Oka ApJ 522, 338 (1999)

H 3 + as a Probe of Dense Clouds Given n(H 3 + ) from model, and N(H 3 + ) from infrared observations: –path length L = N/n ~ 3  cm ~ 1 pc –density  n(H 2 )  = N(H 2 )/L ~ 6  10 4 cm -3 –temperature T ~ 30 K Unique probe of clouds Consistent with expectations –confirms dense cloud chemistry (forbidden) 32.9 K K

R(1,0) R(1,1) u R(2,2) l 33 K 151 K (0,0) (2,0) (J,G) probe of temperature not detected H 3 + Energy Level Structure

Spectroscopy of H 3 + Source Confirmed that H 3 + produced is rotationally cold, as in interstellar medium Infrared Cavity Ringdown Laser Absorption Spectroscopy McCall, et al. Nature 422, 500 (2003)

Theoretical Calculations

TSR Results H. Kreckel, et al. Phys. Rev. Lett. 95, (2005) CRYRINGTSR Ion source Supersonic expansion RF 22-pole ion 13 K Electron target kT  ~ 2 meV n e ~6.3×10 6 cm -3 kT  ~ 500 µeV n e ~4.5×10 5 cm -3 Electron cooler (same as target) kT  ~ 10.5 meV n e ~1.6×10 7 cm -3 Beam energy 12.1 MeV5.25 MeV CRYRING TSR theory Good agreement Different ion production Different conditions Experiments likely “right”!

Supersonic Expansion Ion Source Similar to sources used for laboratory spectroscopy Pulsed nozzle design Supersonic expansion leads to rapid cooling Discharge from ring electrode downstream Spectroscopy used to characterize ions H2H2 Gas inlet 2 atm Solenoid valve -900 V ring electrode Pinhole flange/ground electrode H3+H3+

Observational Consequences? Energy required for acceleration –about 0.2 × ergs/century Heating of diffuse clouds –about 1/10 of photoelectric heating Production of LiBeB (spallation) –roughly consistent with observed abundances γ -ray line production (nuclear excitation) –below detectable limits our spectrum is not excluded by observations!

The Future (The Dream?) Improved precision in  determinations –improved density estimates –more sophisticated cloud models Measure H 3 + in wider range of sightlines –diffuse, translucent, dense clouds Infer  (A V ) → cosmic ray spectrum –information on acceleration mechanism(s) –information on galactic propagation

Recent Astronomical Results Range of ζ from  s -1 Biggest uncertainty is in adopted  n  N. Indriolo, T. R. Geballe, T. Oka, & B. J. McCall, ApJ 671, 1736 (2007)