Lecture 10: "Chemistry in Dense Molecular Clouds"

Outline

1. Observations of molecular clouds

2. Physics of dense clouds

3. Chemistry of dense clouds: C-, O-, N-chemistry

Types of molecular cloudsTypes of molecular clouds

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

• Translucent: Tkin~50–100 K, n ~102–103 cm-3

• Dark dense clouds: Tkin~10–100 K, n ~104–108 cm-3

CO (1-0) survey of Milky Way

Taurus-Auriga

Taurus molecular cloudPerseus

Auriga

Taurus

TMC-1

L1449

NGC1333B5

IC348

T Tau

Ungerechts & Thaddeus 1987

CO 1-0

Prestellar cores

Visible Infrared

Barnard 68

• Dust (1%), gas (99%)

• Typical mass ~10 – 103 Msun, size < 1 pc, n >104 cm-3, T ~10 K

• Dynamically "quiet", t ~ 1 – 10 Myr

• Clumpy

General scheme: physics• Bonnor-Ebert spheres:

hydrostatic equilibrium of isothermal clouds bound by external pressure

• Density profile: flat in the center, steep decrease toward the edge

• Many cores are highly asymmetric

• Some may have already pre-protostars/show signs of infall

Ward-Thompson et al. 1994, 1999

General scheme: physics

• Heating through cosmic rays/FUV

• Cooling through dust and molecular line radiation (C, CO, OH)

• Increasing temperature toward the edge

• Observed T is too low in the center: non-thermal pressure support (B-field) or ongoing collapse

Alves et al. 2001

Velocity structures

Barnard 68

L(NH3) ~ 5x1021 cm2 s-1 L(NH2D) ~ 3x1020 cm2 s-1

• Complex velocity pattern: rotation, infall, turbulence (outflows?)

• In some cores, angular momentum seems to be smaller on smaller scale ⇒ forming protostars?

Observations of dense clouds

Barnard 68

• Dust continuum mapping ⇒ accurate Tdust

• Tgas ~ 10 K ⇒ strongest molecular lines are at mm (rotation)

• nH ~ 104 cm-3 ⇒ higher J-lines can be subthermally excited

• Masses ~10 – 103 Msun ⇒ optically thick lines

• Line widths are thermal

• Intensevily observed since 70's, "classical" source: TMC-1S or TMC-1CP

• Sofar detected ~ 60 species, mostly with single-dish antennas

• Structures of individual cores are studied with radio-interferometers, e.g. B68, L1544, etc.

Molecules in dense clouds

Barnard 68

• Depletion of C-bearing species: CO, CCS, CS, ...

• Non-depletion of N2H+ and NH3:

- depletion of CO: CO + N2H+ → HCO+ + N2 does not work

- slow formation of N2

• Abundant carbon chains

• Abundant negative ions

• Abundant deuterated species

• Simple organics: HCOOH, CH3OH, ...

Ices in dense clouds

Barnard 68

• Ices are observed in absorption against background stars

• Dominated by H2O, CO, CO2,

• Complex ices: HCOOH, CH3OH,

• In starless clouds ~10–50% of heavy elements are in ices

• Up to 99% of the heavy elements may be frozen out in the center

CK 2 behindSerpens cloud

5

Knez et al. 2005Bergin et al. 2005

H2OHCOOH

NH4+

+ ?

Silicate CO2

- Probe ice composition prior to star formation, unaffected by heating and radiation from star- Features same as seen as for YSOs, including

6.0 and 6.8 m bands => do not require heating

Background sources: quiescent cloudsSpitzer IRS spectrum

Molecules as probes

Semenov et al. (2010)

Tracer Properties Quantity

12CO Optically thick linesTemperature

H2, NH3 Symmetric speciesTemperature

13CO, C18O, CS, CCS, H2CO Large dipole moment Density

HCO+, H13CO+, N2H+, C, C+ Charged species Ionization

H2CO, organics Complex species Surface processes

HDCS, D2CS, DCO+, DCN, NH2D, H2D+ Deuterated species Deuterium

fractionation

General scheme: chemistryC2S vs NH3 evolution

Suzuki et al. 1992Hirota et al. 2009Sakai et al. 2010 Suzuki et al. 1992

• Dense core: C+ is converted to C and CO

• Early times: CCS and HCnN are abundant

• Late times: N2H+, H2D+, NH3, CO is absent in the center

• CCS traces outer shell, NH3 traces central region

Dust and molecules in B68

Alves et al. 2002

• No traces of a protostar

• Two components

• CO is frozen in the center

• N2H+ is sitting there

• CS is concentrated toward a smaller clump ⇒ different 'chemical' age accross B68?

Barnard 68

Detection of carbon chains in TMC-1S

Langer et al. 1997

Long carbon chains in TMC-1

Negative ions in TMC-1

Barnard 68

Negative ions and saturated chainsNegative ions

McCarthy et al. 2006

Propene in TMC-1

Marcelino et al. 2007

Electron attachment rate increases with size of chains => search for large species

C6H + e- ⇒ C6H- + hν

Effective for molecules with large e- affinities

<10% of anion/neutral(predicted by Herbst 1981)

Discovered in clouds with predicted abundances(McCarthy et al. 2006, Bruencken et al. 2007)

k ≈ <10-7 cm3s-1

Deuterium chemistry

Barnard 68

Langer et al. 1980

DC3N/HC3N ~ 0.05-0.1

Schoerb et al. 1981

DC5N/HC5N ~ 0.02

HDCS/H2CS ~ 0.3D2CS/H2CS ~ 0.3

CCD/CCH ~ 0.01DCN/HCN ~ 0.02–0.1N2D+/N2H+ ~ 0.01–0.1D2CO/H2CO ~0.01–0.4

Efficient deuterium enrichment in cold cores: D/H in molecules > [D/H] ~ 10-5

Suzuki et al. 1992

• When C is available, CCS and HCnN are quickly produced

• At later times surface chemistry catches up: NH3 is formed, CO is frozen in the center

• CCS traces outer shell, NH3 traces central region where CO freezes out onto dust grains

Carbon chain surveysC2S vs HC3N C2S vs NH3

- Open symbols: cores with stars- Closed symbols: cores without stars

Suzuki  et  al.  ‘92

'Early' vs 'Late Time' molecules

Effect of external environment

Barnard 68

Foster et al. 2009

• CCS is more readily present in isolated, less evolved cores

• Effect of environment: Radiation or Temperature?

• Needs C to be synthesized ⇒ FUV/CRP destruction of CO?

Typical timescales

• Chemical time: >104–105 years

• Collision with dust grains: once in ~1 day at 10 K and 104 cm-3

• Life-time of a cloud: ~1–10 x 106 years

• Low-mass star formation: ~106 years

Chemistry is slow and can be affected by evolution

Early chemical models

• First gas-phase ion-molecule chemistry models: Herbst &

Klemperer 1973, Watson 1976, Dalgarno & Black 1977, Prasad

& Huntress 1980, Millar et al. 1991

• No photoprocesses, only H2 formation on grains

• Grain-surface chemistry gradually included by Allen &

Robinson 1978, Tielens & Hagen 1982, d’Hendecourt et al.

1985, Herbst & Hasegawa 1993

Modern gas-phase networks

• Most recent models contain ~5000 gas- phase reactions

between ~450 (up to 13 atoms):

- UMIST: http://www.udfa.net/

- Ohio State Univ.:http://www.physics.ohiostate.edu/~eric/

research.html

- Kinetic Database for Astrochemistry: http://kida.obs.u-

bordeaux1.fr/models

More detailed chemical models

• Models with evolving physics: nH and T vary with time

(e.g., Tarafdar et al. 1985, Aikawa et al. 2009)

• Depth-dependent models: FUV intensity is accurately

calculated (e.g. Lee et al. 1996)

• Chemo-dynamical models: cycling of gas parcels from inner

to outer to inner regions (e.g., Boland & de Jong 1982, Xie et

al. 1995, Willacy et al. 2002)

Chemical models with surface processes• Enough density in the center to build up icy mantles:

tc = [V(πr2nd)]-1 ~ 109/nH(cm-3) years ⇒ ~105 years if nH is 104 cm-3

• A grain accomodates ~ 1 species per day

•Migrating species: H, H2, D, but also O, C, N

• Low density ⇒ active surface hydrogenation: NH3, H2O, CH3OH,

etc.

• High density ⇒ O-, C-, N-chemistry on grains: hydrocarbons,

cyanopolyynes, etc.

• D’Hendecourt et al. 1985, Herbst & Hasegawa 1993, Shalabiea &

Greenberg 1994, Aikawa et al. 2005–09, Garrod & Herbst 2008,...

Problems of chemical models

HC3NC2S

Modeled abundances are uncertain by factors of >3

Vasyunin et al. (2004, 2007), Wakelam et al. (2005, 2006)

Pause

Oxygen chemistryI.P. of O > 13.6 eV ⇒ oxygen mostly neutral

Ionization provided by cosmic rays: H2 ⇒ H+ ,H2+, H3+ (rapid)

Then:

H+ + O → H + O+ (+227 K)

O+ + H2 → OH+ + H

H3+ +O → OH+ +H2

Once OH+ formed, rapid ion-molecule reactions lead to OH,

H2O and CO

Formation of water

H2 + CRP ⇒ H2+ + e-

H2+ + H2 ⇒ H3

+ + H

H3+ + O ⇒ OH+ + H2

OHn+ + H2 ⇒ OHn+1

+ + H

H3O+ + e- ⇒ H2O + H; OH + 2H, etc

Oxygen chemistry and its coupling with carbon

Oxygen chemistry

A problem of observed low O2Limits on O2

Goldsmith et al. 2000

- Models produce factor 100 more O2 than observed- Solution lies in freeze-out of O on grains (Bergin et al. 2000)

Goldsmith et al. 2000

• Factor >100 discrepancy between observed and modeled O2• Solution: Allow freeze-out of O on dust grains (Bergin et al. 2000) ⇒ converted to water ice rather than O2 ice

Carbon chemistryI.P. of C < 13.6 eV ⇒ carbon mostly C+

Reactions with H3+ are also important

C+ + H2 → CH2+ + hν possible at low T (initiating reaction)

CH2+ ⇒ rapid ion-molecule reactions lead to CH, C2, C2H,

C2H2, ...

C+ + H2 → CH+ + H: endothermic by 0.4 eV

Gas-phase formation of hydrocarbons

C+ + H2 ⇒ CH2+

CH2+ + H2 ⇒ CH3

+ + H

CH3+ + H2/O ⇒ CH5

+/HCO+ + H2

CH5+ + e- ⇒ CH3 + H2

CH3 + O ⇒ H2CO

CH3+ + H2O ⇒ CH3OH2

+

CH3OH2+ + e- ⇒ CH3OH + H

X (too low rate, Luca et al. 2002)

(3 ± 2%, Geppert et al. 2006)

Carbon chemistryCarbon chemistry and its coupling with oxygen

Nitrogen chemistryI.P. N > 13.6 eV ⇒ nitrogen mostly neutral

Nitrogen chemistry:

N + H3+ → NH2+ + H does not occur (barrier?)

N+ + H2 → NH+ + H (~+100 K)

So, it starts with neutral-neutral chemistry linked to carbon:

CH, C2 + N → CN + H, C

CH3+ + N → H2CN+ + H

H2CN+ + e → HCN or HNC + H

Nitrogen chemistry

Nitrogen chemistry and its coupling with oxygen

Sulfur chemistryProduction of C2S

C. Marka et al.: Tracing the evolutionary stage of Bok globules: CCS and NH3

Fig. 3. Abundance ratio NNH3 /NCCS versus evolutionary group; arrowsindicate lower limits and grey symbols sources with uncertain evolu-tionary group (slight horizontal o↵sets around the group positions aresolely for better visibility of individual datapoints).

which is likely because of comparable excitation conditionsand a similar telescope beam for both observations. Figure 3shows NNH3 /NCCS versus evolutionary group. Altogether, we de-rive abundance ratios from about 20 to 860, while the observ-able range is limited to ca. NNH3 /NCCS 2000 by our CCS de-tection limit.

Compared to samples of earlier papers, the NNH3 /NCCS ratiosderived for isolated Bok globules are on average similar to thoseof dense cores in the Perseus molecular cloud (Rosolowsky et al.2008), but a factor of two higher than those in the dark cloudsstudied by Suzuki et al. (1992). This corresponds to the fact thatdespite a comparable range of ammonia column densities in theSuzuki et al. (1992) sample and our globules, we do not findextremely high CCS column densities typical of carbon-chainproducing regions (e.g. Hirota et al. 2009, where carbon-chainproducing regions are defined as having NNH3 /NCCS 10).

The number of sources is too small to allow for a detailed sta-tistical analysis, but taking into account only sources with reli-able evolutionary stage, the following picture arises: within eachevolutionary group, a relatively large range of abundance ra-tios NNH3 /NCCS spanning about one order of magnitude is ob-served, but there is comparatively little variation of the valuesbetween the di↵erent groups. Fig. 3 possibly suggests a slightlydecreasing tendency of NNH3 /NCCS going from group 0 to group Iglobules, but this trend is only marginal. Thus, it can be con-cluded that the ratio NNH3 /NCCS is rather similar across the Bokglobules observed in this study, despite them harbouring YSOsin di↵erent evolutionary stages.

Specifically, the NNH3 /NCCS ratio does not increase goingfrom the presumably youngest objects of group �I to the mostevolved sources of group I. This finding does not easily fit intothe general picture of most chemical models, according to whichCCS as an early-phase molecule is expected to decrease in abun-dance rapidly around 106 yr, while the slowly forming ammoniareaches its maximum abundance in a later stage of the chemi-cal evolution – leading to the anticipation of steadily increasingNNH3 /NCCS ratio along with the evolutionary stage of the glob-ules. For comparison, we show in Fig. 4 the NNH3 /NCCS ratiowith respect to the chemical age of a cloud, calculated fromthe evolution of CCS and NH3 abundances from four chemi-cal models in the literature: Suzuki et al. (1992, dash-dottedline) and Scappini et al. (1998, dashed line) use pseudo-time-

Fig. 4. NH3/CCS ratio from chemical model calculations (see Sect. 5.2)of Suzuki et al. (1992, dash-dotted line), Scappini et al. (1998, dashedline), Bergin (2000, solid line) and Aikawa et al. (2001, dotted line),and this work (10 K warm group �I model marked with open cir-cles, 15 K group 0 model with filled circles and 25 K group I modelwith asterisks. The observed range of NNH3 /NCCS is designated by theshaded area.

dependent chemical models in which the gas density is constantwithin time. In contrast, Bergin (2000, solid line) and Aikawaet al. (2001, dotted line) take into account the dynamics of a col-lapsing core, as well as depletion of species from the gas phaseonto, and desorption from, dust grains. However, both modelsdo not include reactions on grain surfaces except of H2 forma-tion and ion-electron recombination in Aikawa et al. (2001). Asinitial conditions all models take hydrogen in molecular formand carbon and sulphur as ions, the initial abundances are takenas those typical of di↵use clouds or depleted by a certain factorfrom solar abundances; Bergin (2000) allow the cloud to evolvefor 1.5⇥105 yr at constant density and take the chemical compo-sition after this time as initial values for the dynamically evolv-ing core. All models consider the chemical evolution in a regionat a constant temperature of 10 K and shielded from external UVradiation. Aikawa et al. (2001) follow the chemical evolution ofan infalling fluid element in a collapse according to the Larson-Penston solution (we show here their result for a collapse sloweddown by a factor of 10). From Bergin (2000) we show the modelfor a collapse with ambipolar di↵usion and dust grains coveredwith a CO mantle. Both papers examine several variations oftheir models (e.g. grain properties, collapse timescales), whichresult in somewhat di↵erent time evolutions of the molecularabundances, but do not deviate significantly from the examplesshown in Fig. 4.

Despite the di↵erent approaches, all models agree upon arapid increase of the NNH3 /NCCS ratio by three orders of magni-tude at an evolutionary time of several 105 yr, caused by a fastdecrease of the CCS abundance due to depletion onto grains, de-struction by reactions and missing replenishment. The first broadpeak seen in the results of Scappini et al. (1998) and Aikawaet al. (2001) is due to a very slow formation and increase of theNH3 abundance, while CCS is still being formed e�ciently.

In general, a comparison of the absolute values observedwith those from modelling has to be considered with caution.The initial conditions assumed and the exact starting point forreactions defined in chemical models might not necessarily bein good agreement with real globules, and certain scatter has tobe expected due to the natural fluctuations of initial conditionsin the variety of globules. Nevertheless, comparing the measured

8

Marka et al. 2011

• Modeling supports the observed evidence that CCS/NH3 decreases with the age of the cloud

• CCS is produced when C is still in the gas, NH3 forms on dust grain surfaces later

H3+

HCO+, N2H+, OH+

H2D+

e-

HD

DCO+, HCO+, N2D+, N2H+, OD+, OH+

CO, N2, O

H2

H2, H HD, H2, D, H

D2H+ D3+

D2, D

DCO+, N2D+, OD+

• Mass difference between HD and H2 ⇒ fractionation via H2D+, D2H+, D3+, ~ 10 – 40 K

• At steady-state: H2D+/H3+ ~ 0.1 ⇒ D/H of e.g. N2D+, DCO+

• Freeze-out ⇒ DR of H3+ isotopologues ⇒ many D atoms ⇒ surface addition reactions

H2

HD HD

e- e- e-

Deuterium fractionation

CO, N2, O CO, N2, O CO, N2, O

Courtesy of H. Roberts

H2BarriersNo barriers

Surface chemistry: water

O O

H

H 2H

H

Grain surface chemistry

HO

OH

H2O

O2 O3

H2O2

H2O

OH

H2O2H

Tielens & Hagen 1982

Postulated 30 years ago, these reactions are now finally being tested in the laboratory

Surface chemistry: HCOOH and CH3OHExample: CH3OH and HCOOH

OH

See alsoWatanabe et al. 2005Fuchs et al. 2009Ioppolo et al. 2011a,b

Observations vs Theory: TMC1-S

About 75% agreement for 50 observed molecules in TMC1

The End