ERIC HERBST

DEPARTMENTS OF PHYSICS, CHEMISTRY AND ASTRONOMY

THE OHIO STATE UNIVERSITY

Low-Temperature Gas-Phase & Surface Reactions in Interstellar Clouds

Molecules seen in IR absorption and radio emission

10 K

10(4) cm-3

H2 dominant

sites of star formation

Dense Interstellar Cloud Cores

Cosmic rays create weak plasma

Fractional ionization < 10(-7)

Cosmic Elemental Abundances

• H = 1• He = 6.3(-2)• O = 7.4(-4) 1.8(-4)• C = 4.0(-4) 7.3(-5)• N = 9.3(-5) 2.1(-5)• S = 2.6(-5) 8.0(-8)• Si = 3.5(-5) 8.0(-9)• Fe = 3.2(-5) 3.0(-9)

• Dust/gas = 1% by mass

• Gas-phase abundances of heavy elements in clouds reduced.

GAS PHASE INTERSTELLAR/CIRCUMSTELLAR MOLECULES - HIGH RESOLUTION (12/03) _____________________________________________________________________________________________ H2 KCl HNC NH3 C3S C5 C6H CH3 HC4CN CH AlCl HCO H3O+ CH4 CH3OH C7H, C6H2 C8H CH+ AlF HCO+ H2CO SiH4 CH3SH HCOOCH3 CH3COOH NH PN HOC+ H2CS CH2NH C2H4 CH3C2CN H2C6(lin) OH SiN HN2+ HCCH H2C3(lin) CH3CN C6H2 H2COHCHO C2 SiO HNO HCNH+ c-C3H2 CH3NC C2H5OH (CH3)2O

CN SiS HCS+ H2CN CH2CN HC2CHO C2H5CN CO CO+ C3 C3H(lin) NH2CN NH2CHO CH3C4H

CSi SO+ C2O c-C3H CH2CO HC3NH+ HC6CN CO2 C4H2 (CH2OH)2 CP H3

+ C2S HCCN HCOOH H2C4(lin) (CH3)2CO AlNC CS CH2 SiC2 HNCO C4H C5H CH3C4CN? HF SiCN SiC3 C5N NO NH2 SO2 HOCO+ HC2CN CH3NH2 NH2CH2COOH? NaCN CH2CHOH NS H2O OCS HNCS HCCNC CH3CCH HC8CN SO H2S MgNC C2CN HNCCC CH3CHO c-C6H6 HCl C2H MgCN C3O C4Si CH2CHCN HC10CN NaCl HCN N2O NaCN H2COH+ c-CH2OCH2 + ISOTOPOMERS c-CH2SCH2

Some Fractional Abundances in TMC-1

• CO 1(-4)• HCN 2(-8)• C4H 9(-8)• HCO+ 8(-9)• c-C3H2 1(-8)• HC9N 5(-10)

• OH 2(-7)• NH3 2(-8)• HC3N 2(-8)• N2H+ 4(-10)• HNC 2(-8)• O2 < 8(-8)

AN INTERSTELLAR GRAIN

0.1

silicates & carbonaceous material

ices

+ small grains and PAH’s

Water ice = 10(-4) of

Gas density

Water, CO, CO2

H

H

H2

Formation of Hydrogen

dust particle

O

Cosmic ray

H2+ + e

Efficient Low T Gas-PhaseReactions

1. Ion-molecule reactions

2. Radiative association reactions

3. Dissociative recombination reactions

4. Radical-radical reactions

5. Radical-stable reactions

Ea = 0

In areas of star formation, reactions with barriers occur.

Exothermic

Ion-Molecule Reactions

• Experimental evidence down to a few K• Rate coefficients explained by classical

“capture” models in most but not all instances.

• ion-non polar (Langevin case)

91022 ekL cm3 s-1

DCBA

Ion-mol. Rx. (cont)

• Ion-polar

]4767.062.0[ xkk LTS 13710]

2/12

1[ scmxLk

LDk

2/1

2 T

Tkx

B

D

+ more complex state-specific models

Remaining Questions

1) Why are some reactions slow?

223 HDHHDH

2) Is there a quantum limit?

2

2sin)12(

4

k

Radiative Association

hABABBA *

12

1

1 10 ;)(

skkTKkk

kk rrrra

2/)()( BA rrTTK

What is the 0 K limit?

Few ion trap measurements by Gerlich, Dunn down to 10 K

What about competitive channels?

, size, bond engy

Dissociative Recombination Reactions

BAeAB

nTATk )300/()(

Studied in storage rings down to “zero” relative

energy; products measured for approx.10 systems

n=0.5, 1.5

Some systems studied: H3+, HN2

+, HCNH+, H3O+, NH4

+, CH5+ ,CnHm

+

13710 scmA

QUESTION

• How large must ions be before the dominant process becomes radiative recombination? “statistical trapping”

• Answer via statistical theories (RRKM): 20-30 atoms?????

OH + O(3P) O2 + Hpresent experimental results

T / K050100150200250300k /

10-1

1 cm

3 mo

lec-1

s-1

024681012 Davidsson & Stenholm, EL, 1990

Clary, ACCSA, 1984Harding & Troe, cS SACM + CT, 2000Howard & Smith, 1980Lewis & Watson, 1980Howard & Smith, 1981Stewart & Smith, 1994Current astro chemical modelRobertson & GP Smith, 2002this work

Radical-radical Reactions

Detailed capture models by Clary, Troe

RADICAL-NEUTRAL RX RADICAL-NEUTRAL RX (CONT)(CONT)

CN + C2H2 HCCCN + H

CCH + HCN HCCCN + H

YES

NO

C + C2H2 C3H + H YES

Barrier cannot be guessed!!

AttachmentAttachment

hAAe

If enough large molecules with large electron affinities present, electrons may not exist! Note no competitive fragmentation channels.

FORMATION OF GASEOUS WATER

H2 + COSMIC RAYS H2+ + e

Elemental abundances: C,O,N = 10(-4); C<O Elemental abundances: C,O,N = 10(-4); C<O

H2+ + H2 H3

+ + HH3

+ + O OH+ + H2

OHn+ + H2 OHn+1

+ + HH3O+ + e H2O + H; OH + 2H, etc

FORMATION OF HYDROCARBONS

H3+ + C CH+ + H2

CHn+ + H2 CHn+1

+ + H; n=1,2

CH3+ + H2 CH5

+ + h

CH5+ + e CH4

+ H (5%) CH3 + 2H (70%)

CH5+ + CO CH4

+ HCO+

CURRENT GAS-PHASE MODEL NETWORKS

4,000 reactions; 10-20% "studied"; 400 species through 13 atoms in size

elements: H, He, N, O, C, S, Si, Fe, Na, Mg, P, Cl

Latest network – osu.2003 – contains over 300 rapid neutral-neutral reactions. Rate coefficients estimated by Ian Smith and others for many of these. Verification needed!!

Solved kinetically; thermodynamics useless!

t=0; atoms except for H2

GAS-PHASE MODELS OF DENSE CLOUD CORES

"SUCCESSES"

1. IONS ( H3+, HCO+, HC3NH+)

2. ISOMERS (HNC) & RADICALS (OH) HCNH+ + e ----> HCN + H; HNC + H 3. ISOTOPIC FRACTIONATION H3+ + HD <====> H2D+ + H2 4. UNSATURATED MOLECULES A+ + H2 -------> No Reaction 5. ORDER-OF-MAGNITUDE AGREEMENT WITH AT BEST 80% OF MOLECULES

Chemistry imperfect!!

Nature of Solution for a homogeneous, time-independent cloud

Time (Myr)

fi

0.1 10Small species (CO)

Large species (HC9N)

“early time if O- rich”

Nature of Solution for a homogeneous, time-independent cloud

Time (Myr)

fi

0.1 10Small species (CO)

Large species (HC9N)

accretion

Found in pre-stellar cores

“early time if O- rich”

Low Temperature Surface Chemistry on Amorphous Surfaces

• 1) Mechanisms (diffusive [Langmuir-Hinshelwood], Eley-Rideal, hot atom, impurity site)

• 2) Dependence on size, mantle, fluffy nature, energy parameters

• 3) Rate equations vs. stochastic treatments

• 4) non-thermal desorption (cosmic rays)

(diffusion)“physisorption”

Ediff

Edes

Desorption & Diffusion

*0)/exp()( 1 TkEsk Bdesdes

desdiffBdiffhop EETkEsk 30.0);/exp()( 1

For H, tunneling can occur as well.

kdiff = khop/N; N is the number of binding sites

H diffuses the fastest and dominates the chemistry.

Desorption via evaporation and cosmic-ray heating.

for heavies

TYPES OF SURFACE REACTIONS 

REACTANTS: MAINLY MOBILE ATOMS AND RADICALS

A + B AB associationH + H H2

 H + X XH (X = O, C, N, CO,

etc.) WHICH CONVERTS  

O OH H2O 

C CH CH2 CH3 CH4

 N NH NH2 NH3

 CO HCO H2CO H3CO CH3OH 

X + Y XY (CO + O CO2) ?????????? 

Experiments on cold surfaces

• Vidali et al. Formation of H2 on silicates, carbon, and amorphous ice; LH mechanism characterized and energies obtained; formation of CO2; energy partitioning of hydrogen product (also UCL group)

• Ediff(H, olivine) = 287 K; Ediff(H, carbon) = 511 K• But whole analysis of data has been questioned by others,

who feel that both tunneling and some chemisorption sites are involved!!!!!

• Hiraoka et al. Formation of ices (CH4, H2O,NH3, H2CO)• Watanabe et al. Formation of methanol• Danish group formation of H2

MODELLING DIFFUSIVE SURFACE CHEMISTRY

)()()()()(

HNHNKHNkHnkdt

HdNHHdesacc

)()( HkHkK diffdiffHH

Rate Equations

The rate coefficient is obtained by

Method accurate if N>1 Biham et al. 2001

NBNkK hopBA /)(

STOCHASTIC METHODS

Based on solution of master equation, which is a kinetic-type equation in which one calculates not abundances but probabilities that certain numbers of species are present. Can solve directly (Hartquist, Biham) or via Monte Carlo realization (Charnley).

MASTER EQUATION

replaced when N(H) << 1 by a series of coupled equations for Pn(H):

<N(H)> = n Pn(H) dP0(H)/dt = ……….

)()()()()(

HNHNKHNkHnkdt

HdNHHdesacc

Stochastic States

• Unfortunately, with more than one reactive surface species, one must compute joint probabilities ...),,( 321 nnnP so that

the computations require significant computing power. It is necessary to impose cutoffs on the ni and the total number of surface species considered.

More simple fix: modified rate method

New Gas-Grain Stochastic-Deterministic Model

• Stantcheva & Herbst (2004)

• Gas-phase chemistry solved by deterministic rate equations, while surface chemistry solved by solution of master equation. Some results quite different from total deterministic approach.

RESULTS: surfaces

• From observations of grain mantles, the dominant species in the ice are water, CO, CO2, and occasionally methanol.

• The models at 10 K and a gas density of 10(4) cm-3 are able to reproduce the high abundance of water, seem to convert CO into methanol too efficiently, and tend to underestimate the amount of CO2. Results sensitive to density.

• The modified rate method reproduces the master equation approach at 10 K, but the normal rate method can be in error.

Results from Stantcheva & Herbst (2004)

CO

% Agreement in TMC-1

Roberts & Herbst 2002

Gas-phase species

Some Conclusions

• 1) Low-temperature chemistry in interstellar clouds (both gas-phase and surface) partially understood only.

• 2) Chemistry gives us many insights into the current state and history of sources

• 3) More work on “cold chemistry” is clearly needed to make our mirror into the cosmos more transparent.