20 Applications of Interstellar Chemistry 1. General Gas Phase Chemistry 2. Ionization Theory for Clouds 3. Molecular Ions 4. The Discovery of Interstellar H 3 + Appendix A. Experiment on Dissociative Recombination of H 3 + Appendix B. Deuterium Fractionation General References van Dishoeck, UCB Lectures Notes (2000) van Dishoeck, ARAA, 42 119 2004
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
20 Applications of Interstellar Chemistry1. General Gas Phase Chemistry2. Ionization Theory for Clouds3. Molecular Ions 4. The Discovery of Interstellar H3
+
Appendix A. Experiment on DissociativeRecombination of H3
From Lec.19: Chemically recall that chemically active ions are generated from neutral atomic O:
O+ (by charge exchange of H+)OH+ (by proton transfer of H3
+).These ions are then hydrogenated by H2, producing the series OHn
+
with n=1,2,3, but mainly OH3+ (if the abundance of H2 is large).
Recombination of OH3+ leads to O, OH & H2O; CO is made from
the neutral reaction: C + OH → CO + H
or by ionic reactions: C+ + OH → CO+ + HCO+ + H2 → HCO+ + H followed by: HCO+ + e → CO + HC+ + H2O → HCO+ + H
Once CO is formed, HCO+ can be produced directly from CO with H3+:
H3+ + CO → HCO+ + H2.
Since H3+ and H3O+ are spherical rotors & HCO+ is linear (with a large
dipole moment), HCO+ is considered a strong signature of ion-molecule interstellar chemistry.
Extension of the CO ChemistryQuestions immediately arise from the discussion of CO chemistry, e.g., Can other heavy reactive hydrides analogous to OH be generated?Can other heavy diatomics analogous to CO be generated from OH?
A. More Complex Oxygen Chemistry - OH is the precursor of many molecules via moderately fast radical reactions e.g., O2 and CO2:
O + OH O2 + H CO + OH CO2 + H
Chemical models had predicted full association of oxygen into H2O & O2 in shielded regions, a conclusion not supported by SWAS & ISO.
SWAS (4' beam) did not detect O2 at 487 GHz (616 microns), givingupper abundance limits of 10-6. It did detect the lowest rotational transition of H2O at 547 GHz (538 microns), giving H2O/H2 ~10-8, & 10 higher for Orion & the Sgr B2 cloud (galactic center). ISO (1.3‘ beam) detected a variety of molecules in emission & absorption,giving a large range of H2O abundances: H2O/H2 ~ 10-8 - 10-4.
Chemical Spectroscopy with ISOVan Dishoeck, ARAA 42 119 2004
CO2 H2O
There is much lessCO2 in the gas thanin the solid phase.
Conclusions on Oxygen ChemistryObservations of O2 and H2O suggest that the dominant gas phase oxygen species is often atomic oxygen, consistent with limited observations of the OI fine-structure emission lines at 63 & 145 µm.
SWAS may have emphasized cool regions where H2O is frozen out onto grains or where the OH radical chemistry is inoperative due to activation energies (Ta ~ 3000 K).
ISO detections of H2O in warm regions probably arise from thermally desorbed H2O or OH radical chemistry, with heating by outflow shocks
Grains are important: In addition to incorporating O (as oxides ofSi etc.), volatile gases freeze out on grains (as shown by ISO); grains may also catalyze chemical reactions.
Grain chemistry is poorly understood, but it is generally accepted that grains can hydrogenate the abundant heavy elements and synthesize H2O, CH4, NH3, and CH3OH (see van Dishoeck 2000 notes for furtherdiscussion.
B. More Complex Carbon Chemistry
1. Gas-phase Chemistry of Mono-carbon Species (CHn n=1-4, CH+)
Problem #1: C+ + H2 CH+ + H is endothermic by 0.4 eV. OnePossibility is to invoke radiative association
C+ + H CH+ + hν k ~ 10-17 cm3 s-1
C+ + H2 CH2+ + hν k ~ 7x10-16 cm3 s-1
Problem #2: The sequence of abstraction reactions is broken at CH3
+.
CHn+ + H2 CHn+1 + H
Again radiative association has been invoked:
CH3+ + H2 CH5
+ + hν
N.B. Radiative association is difficult to measure in lab (de-excitation of initial complex by 3rd body); theory is unreliable by 1-2 dex.
CH4 is probably formed by hydrogenation on grains.
2. Gas-phase Chemistry of Complex Carbon Species
More than half of the interstellar molecules are polyatomic hydrocarbons.
Most promising route is “insertion reactions” (c.f. early work by Suzuki), e.g. C2, C3, C2H2 & C3H2 can be made with gas-phase reactions:
C+ + CH C2+ + H C+ + CH4 C2H3
+ + HC2
+ + H2 C2H+ + H C2H3+ + e C2H2 + H
C2H+ + e C2 + H
C+ + C2H2 C3H+ + HC3H+ + e C3 + HC3H+ + H2 C3H3
+ + hvC3H3
+ + e C3H2 + H
N.B. Not all routes and branches have been shown.
Compared to the mono-carbon species (CH, CH+, CH4), the gas phase chemistry of multi-carbon species is quite promising.
C. More Complex Chemistry: Nitrogen
IP(N) = 14.5 eV: The most abundant nitrogen species in diffuse regions is atomic N.
Some reactions may have barriers at low T, e.g.,
H3+ + N → NH+ + H2, NH2
+ + HN+ + H2 → NH+ + H
NH3+ and NH4
+ interact weakly or not at all: standard ion-molecule sequence ends at NH2.
Ion-molecule chemistry doesn't work as well for N as for C & O.
NH3 is probably made on grains.
The CN radical can be made by neutral reactions from carbon radicals, e.g.,
CH + N → CN + HC2 + N → CN + C
Both CN & HCN can also be made by ion-molecule reactions, e.g.,
C+ + NH → CN+ + HCN+ + H2 → HCN+ + H
H3+ + CN → HCN+ + H2
HCN+ + H2 → HCNH+ + HHCNH+ + e → HCN, HNC + HHCN+ + e → CN + H2
Polyacetylene chains can also be built up along these lines, e.g.,
C2H2+ + HCN --> → HC3N2+ + H2.
This type of chemistry is particularly appropriate to the circumstellar envelopes of AGB stars such as IRC 10216.
2. Ionization of Interstellar Clouds
Promising as it is, ion-molecule chemistry is difficult to test because chemical models are extremely complex. For example, the UMIST program (www.rate99.co.uk) has ~500 species and ~3500 reactions, few of which have been measured in the laboratory. Furthermore,the chemistry depends on the temperature and density, so the results are sensitive to the dynamical and thermal model.
Testing ion-molecule chemistry naturally hinges on the problem ofthe ionization of interstellar clouds, i.e., signature molecular ions.
In earlier lectures we developed simple closed-form theories for theionization fraction, e.g., in Lecture 7, we gave the theory for photo and collisionally ionized atomic hydrogen:
(α2 + kc)xe2 + (G/nH –kc)ne – G/nH = 0.
In the next we generalize this to include heavy atomic and molecular ions.
A molecular ion AH+ will usually easily transfer Its proton to another molecule B if it gets better bound. The determining quantity is the proton affinity which, aside from the fact that its defined in terms of the enthalpy per particle, is essentially the binding energy. Thus the energy yield (difference between initial & final energies) of a proton transfer reaction is,
∆E(AH+ + B → BH+ + A) = [pa(B) - pa(A)]
In the first entry in the table for H, one finds the binding energy of the H2
+ molecule, 2.65 eV. From the entry for H2, one sees that H3
+ has a binding similar to H2 itself. Exothermicproton transfers go from bottom to top.
3. The Discovery of Interstellar and H3+
H2+ is the first ion produced by cosmic rays in molecular
gas, but its abundance is very low because it is destroyed rapidly by both atomic and molecular hydrogen. In the latter case, it produces H3
+, which is much more abundant since it is not destroyed by either atomic or molecular hydrogen. However it is destroyed by many neutral species, especially by proton transfer, thereby generating a rich ion-molecule chemistry. HCO+, the first molecular ion discovered in dense regions (Snyder et al. 1970) is produced this way,
H3+ + CO → HCO+ + H,
and was the surrogate for H3+ and the signature for
interstellar ion-molecule chemistry until H3+ was
discovered in the ISM in 1995. Why did this take 25 years?
Discovery of H3+
J. J. Thomson Phil. Mag. 21 225 1911
H+
H2+
H3+
H3+ in Laboratory Hydrogen Plasmas
A. J. Dempster (discoverer of 235U) established that H3
+ is abundant in lab plasmas (Phil Mag 31 438 1916)
H3+ is produced by the reaction
H2+ + H2 → H3
+ + HHogness & Lunn (Phys Rev 26 44 1925)
The exothermic nature and rate of this reaction had been known since the 1930’s, but appreciation of its possible interstellar role had to wait 30 years for a short note in ApJ by Martin et al. (next page);10 more years for serious consideration; and 25more years for it to be detected in the ISM.
ApJ 134 1012 1961
Early Work on Interstellar H3+
• Martin et al. (ApJ 134 1012 1961) suggest significant interstellar abundance of H3
Energy levels are those of a symmetric top:E = B J(J+1) + (C-B) K2
B = 43.56 cm-1 C = 20.61 cm-1
Observed ν2 = 1-0 transitions
J=K=0 is forbidden by the exclusion principle,so the J = 1, K = 1 state of para H3
+ is the lowest rotational state; the J = 1, K = 0 of orthoH3+ is only E/k = 32.9 K higher. These arethe only levels populated for T ≈ 5 - 50 K.
They have ~ equal populations since the higher statistical weight of ortho (I = 3/2) compared to para (I =1/2) compensates approximately for the Boltzmann factor.
Discovery of Extraterrestrial H3+
First detection of H3+ in space
was in the aurora of the Jovian planets (Trafton et al. ApJ 343 L17 1989).
First interstellar detections towards massive, IR-bright, deeply embedded YSOs, W33A and AFGL2136, in the ortho-para doublet. N(H3
+ ) ≈ 4 x 1014 cm-2,T = 30-35 K Geballe & Oka Nature 384 334 1996
Subsequent Detections of H3+
Geballe, Hinkle, McCall, & Oka have detected H3+ along
many lines of sight, originally towards distant IR-luminous embedded YSOs and more recently towards some close bright stars for which extensive UV observations are available.
•McCall et al. Nature 279 1910 1998 Cyg OB2 #12 – “diffuse”•Geballe et al. ApJ 510 251 1999 Galactic center – “diffuse” and “dense” (also Cyg OB2 #12)•McCall et al. ApJ 522 338 199 luminous embedded YS0s – “dense”•McCall et al. ApJ 567 391 2002 – diffuse clouds•McCall et al. Nature 422 500 2003 ζ Per – diffuse cloud
Quotation marks have been used for some of the diffuse & dense designations where the lines of sight are long and very likely involve inhomogeneous gas.
H3+ in “Dense” Clouds
1.02
1.00
0.98
0.96
0.94
36700366803666036640 3717037150
AFGL 2136
AFGL 2591
R(1,1)u R(1,0) R(1,1)l
Wavelength (Å)
N(H3+) = 1–5×1014 cm-2
McCall, Geballe, Hinkle, & OkaApJ 522, 338 (1999)
Diffuse Clouds
8
6
4
2
0
H3+ C
olum
n D
ensi
ty (1
014cm
-2)
6543210
E(B-V) (mag)
ζ OphP Cygni
HD 183143
WR 118
Cyg OB2 12
WR 104
Cyg OB2 5
WR 121
HD 168607
HD 194279
GC IRS 3
χ2 Ori
HD 20041
1.01
1.00
0.99
0.98
0.973.7173.7163.7153.6693.6683.667
Wavelength (µm)
R(1,1)u
R(1,1)l
R(1,0)
HD 183143
McCall, et al.ApJ 567, 391 (2002)
Cygnus OB2 12 and ζ Per
N(H3+) ≈ 3×1014 cm-2
similar to columns for “dense” clouds
N(H3+ ) ≈ 8 x 1013 cm-2
ζ Per is a well studied cloud and nicely exemplifies theconundrum posed by the observations of H3
+
Confronting Theory & Observation for ζ PerEarlier we derived the following expression for the local abundance of H3
+
Oe
2
H
23
)H()H(kxx
xn
x+
=+
βς
and showed that, for a dense cloud with nH =2500 cm-3, the H3+
abundance is ~10-8. Thus if the cloud has a linear dimension of 1-2 pc, the H3
+ column is ~ 1014 cm-2 – the same order as observed for “dense” and “diffuse” clouds. The problem arises when the case of the well studied diffuse cloud towards ζ Per is considered.
From the UV observations (Copernicus, Savage et al. ApJ 216 291 1977; HST Cardelli et al. ApJ 467 334 1966), we have:
If we further assume that β is roughly constant, we can solve for the CR ionization rate of H2 in terms of the measured H3
+ column
All of these results are average quantities and need to be converted into local quantities with some realistic model. In the absence of a model, and to illustrate principles, we integrate the above expression for x(H3
+) to obtain
1
23 )H()H(−+
≈xNβς
e2 xL
Discussion of H3+ in ζ Per
But there is a problem in that the cloud average of the abundance ratio of H2 to electrons is variable, one increasing and the other decreasing, so that the ratiorapidly increases going into a cloud (or into a dense in-homogeneity of the cloud). What McCall et al. (and everyone else) do is replace these abundance by column density ratios:
H2
H32 /)H(
/)C()H(NNNN
LN ++
≈βς
Substituting the measured column densities leads to:115
2 s10 −−≈ς
H3+ in ζ Per (concluded)
This value is 10 times larger than that obtained from:(1) demodulated measurements of the local CR intensity(2) chemical modeling of OH using ion-molecule reactions.
This is a challenge to interstellar chemistry. Some of the explanations that come to mind:a. ζ CR is larger for the ζ Per l.o.s. due to cosmic ray sources in the Per OB associationb. very low-energy CRs (< 2 MeV) are screened from thicker clouds since the range of a 2-MeV proton is 4x1021 cm-2.c. some other ionizing radiation operates in diffuse cloudsd. spatial in homogeneities may increase <n(H2)/ne> above 3,000.
If the large ζ CR is sustained, it has important implications for the heating and ion-molecule chemistry of diffuse clouds, the ones most amenable to detailed study.
20 Years of Uncertainty About the Dissociative Recombination Rate
Measured values of β (H3+ ) have fluctuated, and the
theory has been uncertain. Lab experiments on the afterglow of decaying H plasma ranged from < 10-11 to 2x10-7 cm3 s-1 (at 300 K). The small rate coeffcientsoriginally led McCall & Oka to suspect that the large ζwas really an incorrect large β.
There always were concernsthat the measurements werewrong because the ions werewere in excited rovib states. Ion storage ring experimentsconsistently give large values.
Cryogenic Ion Storage Ring Results
T-1/2
β for rotationally cold H3+ vs. electron temperature
β(30 K) = 2.6 x 10-7 cm3 s-1
β(300 K) = 6.8 x 10-8 cm3 s-1
For details of McCall’s experiments, see Appendix A of these notes, or McCall et al. Nature 400 500 2003and Phys Rev A70 052716 2004
New Chemical Model for ζ PerLe Petit, Roueff, & Herbst (A&A 417 993 2004) use a 3-component model to calculate the abundances of 18 measured species, plus the rotational population of H2. There are 2 uniform phases, with these parameters,
plus an arbitrary number of MHD shocks. The dense phase is tiny (100 AU) and the diffuse phase has L = 5 pc.
The shocks and the dense phase are introduced to solve other problems (the old one of CH+ and C2, C3) & not the H3
+ abundance.
Chemical Model for ζ PerBy varying the temperature (which affects the H+ + O charge exchange rate), they get factor of 3-4 agreement with all measurements using an intermediate value of the CR ionization rate, 2.5x10-16 s-1. This is at the expense of under-predicting the H3
+ abundance by 3 and over-predicting many others.
Thus their CR ionization rate would be only 30% smallerthan McCall’s if they fit the H3
+ measurement exactly (andof course get other things wrong). Therefore, this model does not really solve the problem raised by the H3
+
observations. But it does suggest that somewhat larger CR ionization rates may be tolerated in ion-molecule chemistry of diffuse clouds.
Future Work on H3+
• Search for H3+ in more UV-accessible sightlines
– comparisons with ζ Per– meaning of non-detections in ο Per & ζ Oph
• Better observations and modeling of ζ Per cloud and nearby lines of sight
• Observations of H3+ in more reddened sources
– decrease in ionization rate
Closing Thought
Appendix A: Cryogenic Ion Storage Ring Experiment to Measure Dissociative Recombination of H3
+
• CRYRING ion storage ring, Manne Siegbahn Laboratory in Stockholm– Cold H3
+ ions injected from the Berkeley supersonic expansion ion source
– In a straight section of the storage ring, the ions passed collinearly through an approximately homogeneous electron beam
– Electron energy is chosen so that ion–electron velocity matching is achieved
– H3+ ions are lost to dissociative recombination with electrons and
collisions with residual gas molecules– Total number of neutral reaction products (H + H + H and H2 + H)
are measured– Electron energy varied by changing the cathode voltage in the
electron cooler, and the neutral fragment signal recorded as a function of the longitudinal energy difference between electronsand ions (‘detuning energy’)
+ – e- impact energy-Rotationally hot ions produced-No rotational cooling in ring
Cryogenic Ion Storage Ring ResultMcCall et al. Nature 422 500 3003
Note structure (resonances) in the cross-section.
Recent Theoretical Work
Significantly better than earlier attempts, but not without problems
Appendix B: D Isotope FractionationThe ion-molecule reactions are all fast in this problem (except when endothermic). Reactions not discussed here: very fast dissociative recombination, photodissociation of HD (little self-shielding), and neutral destruction of molecular ions.
Transforming D into HD:H+ + D --> D+ + H endothermic by ~41 KD+ + H2 --> HD + H+
Fractionating H3+:
H3+ + HD --> H2D+ + H2 exothermic
H2D+ + H2 --> H3+ + HD endothermic by ~ 150 K
Passing on the fractionationH2D+ + CO --> DCO+ + H2HCO+ + HD --> DCO+ + H2
Observations:a. FUV: absorption lines of D and HD in diffuse clouds, b. mm: DCO+/HCO+ ratios up to ~0.1 in dense clouds plus many other deuterated molecules, e.g., D3H
Importance of Low TemperaturesD HD, DCO+ etc., is favored by low T because reversing fractionation is endothermic.