Astrochemistry

I. Basic processes

Ewine F. van Dishoeck

Leiden/MPE
Spring 2009
 What is astrochemistry?
(or molecular astrophysics)

‘Formation, destruction and excitation of
molecules in astronomical environments and their
influence on the structure, dynamics and evolution
of astronomical objects’

‘Blending of astronomy and chemistry in which
each area enriches the other in a mutually
stimulating interaction’

‘Astrophysics is almost entirely applied atomic,
molecular and optical physics’
‘
Dalgarno 2008, ARA&A
 1.1 Introduction

Molecules are found throughout the universe

Molecular clouds, evolved stars, planetary nebulae, protoplanetary
disks, stellar and (exo-) planetary atmospheres, solar photosphere,
comets, galaxies (nearby to high z), …..

Some typical conditions

Diffuse clouds: Tkin~100 K, n~100 cm-3

Dense clouds: Tkin~10-100 K, n~104-108 cm-3

Hot cores: Tkin~100-1000 K, n~106-108 cm-3

Disk midplane: Tkin~10-1000 K, n~108-1013 cm-3

Compare atmosphere at sea level: Tkin~300 K, n~3 1019 cm-3

=> Conditions very different from those normally encountered

in lab on Earth: molecular physics
 Intro (cont’d)

Interstellar clouds are birthplaces of new stars

Evolution abundances molecules: astrochemistry

Molecules as physical diagnostics: astrophysics

Progress strongly driven by observations:
technology

=> Very interdisplinary topic!

 Birth and death of stars: astrochemical evolution

Based on Ehrenfreund & Charnley 2000

Lifecycle of gas and dust
 1.2 Composition clouds
Cosmic (solar) abundances elements

Element Abundance Element Abundance

H 1.00 Mg 4.2×10-5
He 0.075 Al 3.1×10-6
C 3.5×10-4 Si 4.3×10-5
N 8.5×10-5 S 1.7×10-5
O 5.5×10-4 Ca 2.2×10-6
Na 2.1×10-6 Fe 4.3×10-5

- Precise values for C, N and O under discussion because of revision of solar abundances
The Astronomers’ Periodic Table

0.1

1.0 by number
4 1 6 10-4

4 10-5 4 1 10-5
4 10-5

Dust grains: 10-12 by number B. McCall 2001

 Interstellar grains

- Small solid particles ~0.01-0.5 µm in size consisting of

silicates and carbonaceous material; ~10-12 by number w.r.t. H
- Most of Si, Mg, Fe incorporated in silicate cores;
~30% of O; ~60% of C in carbonaceous material
- Cold dense clouds (Tdust~10 K): gas-phase species condense on
grains forming an icy mantle
1.3. How do we observe molecules?

E=Eel + Evib + Erot

 Cold dense clouds

Opaque at visible and UV wavelengths =>
molecules shielded from dissociating UV radiation

Millimeter emission: rotational transitions

Limitation: molecule must have permanent dipole
moment => cannot observe H2, C2, N2, CH4, C2H2, …

Advantage: many molecules down to low abundances;
lines in emission => map

Infrared absorption: vibrational transitions

Limitation: need background IR source => only info
along line of sight

Advantage: symmetric molecules + solid state

Earth’s atmosphere prevents observations of key
molecules from ground: H2O, O2, CO2
Infrared: absorption gas and solids

Flux
107 cm-3
Continuum due to hot dust

104 cm-3 Absorption by cold dust

Wavelength

Vibrational transitions of gases and solids

1.4 Identified interstellar molecules

Herbst, priv. comm.

- ~150 total, not including isotopologues
- number of new molecules still increasing by ~3 per year
 Diversity of molecules

About 150 different molecules firmly identified

Ordinary molecules
NH3, H2O, H2CO, CH3CH2OH, ….

Exotic molecules
HCO+, N2H+, HCCCCCCCN, ….

⇒Unusual molecules
(rare on Earth but not in space)
Millimeter spectra: myriads of lines!

G327.3
Massive YSO

Gibb et al. 2000

- Ultra-high precision spectroscopy (accuracy > 1:106)

- Not all lines identified
 Some (recent) detections

H3+, H2D+, D2H+: cornerstones ion-molecule
chemistry

C4, C6H2, CH3CHCH2: new carbon chains

Cyclic C2H4O: fifth ring

C6H6: benzene: simplest PAH

C6H-, C8H-, C4H-: first negative ions!

D2CO, ND3, CD3OH: doubly + triply deuterated
molecules

NaCN, AlCN, SiN: metal-containing species

Not convincingly detected: O2, glycine, ….

 Negative ions
Propene in TMC-1
Negative ions

Marcelino et al. 2007

McCarthy et al. 2006

Larger chains have large electron

affinity (binding energy)
Herbst & Osamura 2008
Molecules at high redshift: z=6.4!

CO and [C II] in quasar SDSS J1148+5251 at z=6.4

Walter et al. 2003, Maiolino et al. 2005

 HCO+, CN at high z
HCO+ CN

Riechers et al. 2007a,b

Detection of HCO+ and CN toward Clover Leaf quasar at

z=2.56 (lensed system => signal enhanced)

Both lines require high densities 105-106 cm-3 for excitation
 1.5 Importance of molecules

Exotic chemistry: unique laboratory

Astrochemical evolution

Molecules as diagnostics of temperature Tkin,
density nH, velocity, …

Molecules as coolants

Radiation escapes from cloud => net kinetic
energy lost => cloud cools down
Collisions Radiation
CO (J=0) CO(J=1) CO(J=0)
 Questions addressed

What are chemical processes leading to formation
and destruction of molecules?

How well are basic molecular processes known
from experiments or theory

What is evolution of molecules in the universe,
from their creation at high redshifts to interstellar
clouds to incorporation in new solar systems

How can molecules be used as physical and
chemical diagnostics of physical structure,
evolution, cosmic-ray ionization, …
 1.6 Basic molecular processes:
gas phase

Because of low temperatures and densities in
clouds, chemistry is not in thermodynamic
equilibrium but controlled by two body reactions
=> abundances depend on physical conditions
(T,n,radiation field), history, …

Three body reactions do not become important
until n>1012 cm-3

Although models contain thousands of reactions,
only few different types of processes

Rate of reaction: k n(X) n(Y) cm-3 s-1

Rate coefficient in cm3 s-1

 Chemical network: example
Carbon chemistry and its coupling with oxygen

These networks look complicated but are built up from only a handful of
basic types of reactions
 Types of chemical reactions

Formation of bonds

Radiative association: X+ + Y → XY+ + hν

Associative detachment X- + Y → XY + e

Grain surface: X + Y:g → XY + g

Destruction of bonds

Photo-dissociation: XY + hν → X + Y

Dissociative recombination: XY+ + e → X + Y

Collisional dissociation: XY + M → X + Y + M

Rearrangement of bonds

Ion-molecule reactions: X+ + YZ → XY+ + Z

Charge-transfer reactions: X+ + YZ → X + YZ+

Neutral-neutral reactions: X + YZ → XY + Z
 1.7 Radiative association
τc τr

X + Y → XY* → XY + hν
τd

Energy conservation => photon must be

emitted, which is a very slow process

τr=10-2-10-3 s vibrational transition

τc,d=10-13 s collision time

⇒ Molecule formation occurs only

1:1010 collisions
 Radiative association (cont’d)

Process becomes more efficient if electronic states
available: τr smaller

C+ + H2 example

Also more efficient for larger molecules: τd longer

τr=10-8 s electronic transition

τc,d=10-13 s collision time

=> Efficiency increased to 1:105

 Radiative association (cont’d)

Radiative association is extremely difficult to

measure in laboratory because 3-body processes
dominate under most lab conditions.

Many rate coefficients are based on theory; overall
uncertainties 1-2 orders of magnitude

Exception: C+ + H2 → CH2+ + hν ν

k~10-15 cm3 s-1 within factor of 2-3

Initiates carbon chemistry
 1.8 Associative detachment

Usually not important in cold clouds, but
can play a role in partly ionized regions and
early universe

Form negative ions by radiative attachment
X + e → X- + hν slow process

Form molecule by associative detachment
X- + Y → XY + e fast process
 Associative detachment

fast

Examples: H + e → H- + hν
H- + H → H2 + e fast
Application: early universe chemistry

Black 1990
 First molecules in universe:
He2+ and HeH+

Latter 1989
He2+ and HeH+ formed by radiative association
 H chemistry in early universe

Hydrogen chemistry in early universe is
very different from that in the current era
due to the absence of dust => H2 must be
formed by slow gas-phase reactions

Direct formation by radiative association
H + H → H2 + hν is much too slow since H2
does not have a dipole moment => consider
other routes
 H+ route

H2 formation:

H + H+ → H2+ + hν

H2+ + H → H2 + H+

H2+ can be destroyed by photodissociation and
dissociative recombination

H2+ + hν → H + H+

H2+ + e → H + H
=> Formation of H2 only becomes effective when
TR<4000 K and photodissociation of H2+ ceases
 H- route

At later times (z~100), H2 can be formed
through H-

H + e → H- + hν

H- + H → H2 + e

H- is destroyed by photodetachment

H- + hν → H + e
with threshold of 0.75 eV => need TR <
1000 K before route becomes effective
 H2 chemistry

Both H+ and H- routes are catalytic, since
H+ and e returned

H2 destroyed by

H2 + H+ → H2+ + H

H2 + e → H + H-

Net result: f(H2) = n(H2)/nH ~10-6 as z → 0

Small molecular fraction in early universe

(but very important as coolants, allowing clouds to collapse
and setting the mass of the first stars)
H2 formation in early universe by
H+ and H- routes

Latter 1989
Sensitivity to cosmological
parameters

Hatched area covers range of variation of Ω0=0.1-1,

h=H0/100=0.3-1 and η10=baryon/photon=1-10
 1.9 Photodissociation

XY + hν → X + Y

Direct photodissociation Predissociation

OH, H2O, CH, CH2, … CO, NO, …..

Experiments available for stable molecules, but not for radicals or ions

Small molecules: theory works well: quantum chemical calculations of
potential surfaces of excited states + transition dipole moments,
followed by nuclear dynamics to obtain cross sections
 Examples experiments

H2O absorption followed by NO absorption (full) and

direct dissociation: accurate fluorescence (dashed); mostly
cross sections within 20% predissociation through discrete
transitions; large uncertainties
(order of magnitude)
H2 spontaneous radiative dissociation

90% of absorptions into B

and C states are followed
by emission back into bound
vibrational levels of the X state

10% of the absorptions
are followed by emission into
the unbound vibrational
continuum, leading to
dissociation

Both H2 and CO p.d. initiated

by line absorptions => self-shielding
 Interstellar radiation field

CH, OH
CO, k=∫σ(λ)I(λ)dλ
H2
cross radiation
section field

912 Å = 13.6 eV cutoff

Average radiation provided by early-type stars in solar neighborhood

Note that H2 and CO p.d. and C photoioniation only occur at 912-1100 Ǻ
C can be photoionized by ISRF, but not O and N
 Cosmic-ray induced radiation:
UV radiation inside dense clouds
H2 + CR => H2+ + e*

H2 + e* => H2* + e

H2* => H2 + hν
ν
p

750 1000 1250 1500 1750 Prasad & Tarafdar 1983

λ(Å) Gredel et al. 1987

-Detailed line + continuum spectrum peaking around 1600 Å and continuing below
912 Å
 Other radiation fields

Ly-α dominated

Shocks, …..

Stellar blackbodies Teff=4000-10000 K

Disks, (exo-)planets, cool PDRs, …

Solar radiation Teff=5500 K + Ly α

Comets

http://www.strw.leidenuniv.nl/~ewine/photo

See van Dishoeck et al. 2006

1.10 Dissociative recombination

Atomic ions: X+ + e → X + hν Radiative: slow

Molecular ions: XY+ + e → XY + hν Radiative: slow

→ X + Y Dissociative:
rapid at low T

Need curve crossing between XY+ and repulsive XY potential for

reaction to proceed fast

Most rates well known; branching ratios products major uncertainty
 Storage ring experiments

CRYRING, Stockholm

- Also Aarhus, Heidelberg

 H3+ + e DR rate

McCall et al.
2003

- Literature values range from <10-12 to 10-7 cm3 s-1 at 300 K over
last 25 years
- High rate coefficients now also reproduced by theory, even
without curve crossing
- Affects determination cosmic ray ionization rates in diffuse clouds
 1.11 Ion-molecule reactions

Ion induces dipole moment in molecule when it
approaches it => long-range attraction which goes
as ∝1/R4

Reaction is rapid even at low T if the reaction is
exothermic; rate can be readily computed by
classical capture theory developed by Langevin
1905
Energy
X+ + YZ XY+ + Z

XYZ+ Reaction path

 Classical capture theory

k independent of T
X+
k=√αe2/µ
bc YZ polarizability

- Reaction only occurs if impact parameter b small enough that X+

is ‘captured’, i.e., spends enough time near YZ for reaction to
take place
- Gas-phase reaction networks built on ion-molecule reactions;
initial ionization provided by cosmic rays (H+ and H3+) or photons (C+)
 Ion-molecule processes

X+ + YZ → XY+ + Z exchange
→ X + YZ+ charge transfer

Many experiments performed at room T,
some at low T. Most reactions (>90%)
indeed proceed at Langevin rate, but some
exceptions known
NH3+ + H2
→ NH4+ + H
 Experiments

M. Smith 1993

-Rate coefficients for ion-polar reactions may be factors of 10-100

larger than Langevin values at low T, because V(R)∝R-2
Example: C+ + OH → CO+ + H
 1.12 Neutral-neutral reactions

Long-range attraction weak: van der Waals
interaction ∝ 1/R6

Potential barriers may occur in entrance and
exit channels => reactions thought to be
slow at low T

Experiments: reactions can be fast at low T!
Energy X + YZ XY + Z

entrance exit Reaction path

XYZ
CN + C2H6: or why extrapolation is unreliable
T/K
1000 500 300
7
k / 10-11 cm3 molec-1 sec-1 6

1.0 2.0 3.3 4.0

103/(T/K)
I. Sims et al. Rennes/Birmingham
 CN + C2H6: or why extrapolation is unreliable
T/K
1000 500 300
7

k / 10-11 cm3 molec-1 sec-1

6

5
T/K
1000 300 200 4
100
10-10

k / cm3 molec-1 sec-1

3

1.0 2.0 3.3 4.0

10-11
T/K 103/(T/K)
1000 100 50 20
10-9
10-10
k / cm3 molec-1 sec-1

10-11 10-12
1.0 3.3 5.0 10.0
10-12
3
10-13 10 / (T/K)
10-14
10-15
10-16
10-17
10-18
1.0 10.0 20.0 50.0
103/ (T/K)
 CN + C2H6: or why extrapolation is unreliable
T/K
1000 500 300
7

k / 10-11 cm3 molec-1 sec-1

6

5
T/K
1000 300 200 4
100
10-10

k / cm3 molec-1 sec-1

3

1.0 2.0 3.3 4.0

10-11
T/K 103/(T/K)
1000 100 50 20
10-9
10-10
k / cm3 molec-1 sec-1

10-11 10-12
1.0 3.3 5.0 10.0
10-12
3
10-13 10 / (T/K)
10-14
10-15
10-16
10-17
10-18
1.0 10.0 20.0 50.0
103/ (T/K)
CN + C2H6: reaction stays rapid at low T!

T/K
1000 100 50 20
10-9
k / cm3 molec-1 sec-1 10-10
10-11
10-12
10-13
10-14
10-15
10-16 I. R. Sims, J. L. Queffelec, D. Travers, B. R.
Rowe, L. B. Herbert, J. Karthäuser, and I. W.
10-17 M. Smith, Chem. Phys. Lett. 211 (4-5), 461
(1993).
10-18

1.0 10.0 20.0 50.0

103/ (T/K)
 Isentropic expansion
50-100 slm carrier gas (He, Ar or N2) + and
CRESU technique
precursor + reagent
uniform supersonic
axisymmetric Laval nozzle flow
boundary layer sets
scale of apparatus

nozzle throat diameter

3 mm – 5 cm

uniform supersonic flow

T = 7 – 220 K
ρ = 1016 – 1018 cm–3
chamber pressure 0.1 – 0.25 mbar
pumping speed ∼ 30000 m3 hr-1
Laval nozzle and isentropic flow
Laval nozzles on rockets
Types of neutral-neutral reactions
EA/k V(R)

Molecule-molecule >104 K

E.g., H2+D2 → 2HD

Radical-saturated mol ~2000

E.g., OH + H2 → H2O + H

Radical-unsaturated mol ~0

E.g., OH + CO → CO2 + H

Radical-radical -100

E.g., O + OH → O2 + H

Not all neutral-neutral reactions are rapid

See also Smith 2006, Faraday Disc
 1.13 Pure gas-phase models

Most recent models contain nearly ∼4500 gas-
phase reactions between ∼450 species containing
up to 13 atoms. Publicly available on web

UMIST code

http://www.udfa.net/

Ohio state code

http://www.physics.ohiostate.edu/~eric/research.html

Most dark cloud models ignore depth dependence
⇒ solve chemical networks for given T, n at single
position and assume
N (AB) n(AB)
x (AB) = =
N ( H 2 ) n( H 2 )
observed calculated
 How important are reactions?
Sensitivity analysis gas-phase networks

- Take estimated uncertainties in rate coefficients into account

- Uncertainties become larger for large molecules
- Observed CH3OH cannot be produced by pure gas-phase reactions
- Use analysis to identify critical reactions for which rates need to
be better determined Wakelam et al. 2005
 1.14 Gas-grain chemistry

Evidence for gas-grain chemistry

H2 in interstellar clouds

NH in diffuse clouds

Abundances H2O, CO2, CH3OH, … in ices
higher than expected from freeze-out of gas
phase
 Formation mechanisms

Diffusive mechanism (Langmuir-Hinshelwood)

X + g:Y → X:g:Y → X-Y:g → XY + g
sticking diffusion + desorption
molecule formation

Direct mechanism (Eley-Rideal)

X + Y:g → X-Y:g →XY + g
direct desorption
reaction

Surface can be silicates, carbonaceous, ice, …

See Lecture 2 for more on grain surface processes
Simulations diffuse mechanism

Takahashi et al. 1998

 Summary lecture 1

Large variety of molecules observed in ISM

Basic processes for formation and destruction identified

Networks built for explaining abundances

Results completely dependent on thousands of input rates

Many not known under astrophysical conditions (but for many accurate
rates not needed)

Key: identify those reactions which are important to study well

Many experiments and theory on basic processes over last 30 years

Good, new chemical physics questions!

Significant progress in neutral-neutral reactions, surface reactions

Some processes now well understood, others take decades of hard
work to make just a little progress

Funding of lab astrophysics groups becoming major problem

Note: photoionization codes like CLOUDY got atomic input data thanks
to huge efforts in atomic and plasma physics