IS HYDROGEN CYANIDE (HCN) A PROGENITOR OF ACETONITRILE (CH3CN) IN THE ATMOSPHERE?

GEOPHYSICAL RESEARCH LETTERS, VOL. 12, NO. 3, PAGES 117-120, MARCH 1985

IS HYDROGEN CYANIDE (HCN) A PROGENITOR OF ACETONITRILE (CH3CN) IN THE ATMOSPHERE?

G. Brasseur 1 R Zellner 2 A De Rudder 1 and E Arijs 1

l•elgian Institute for Space Aeronomy, B-1180 Brussels, Belgium.

-Institut fHr Physikalische Chemie, Universit•t GSttingen, FRG

Abstract. The possibility of a link between HCN residence time of about 1-5 years, that this

and CH3CN in the atmosphere has been suggested species is relatively well mixed in the tropo-

recentIy. A new chemical analysis of this problem sphere but that its mixing ratio decreases slowly as well as model calculations indicate that, most with altitude in the stratosphere.

probably, these gases are produced at the Earth's In order to test the suggestion of a link

surface and destroyed by oxidation in the middle between HCN and CH3CN , new model calculations

atmosphere. The strength of the photodestruction have been performed. These consider the latest

of these 2 molecules seems to be weak. It is field measurements of CH3CN as well as recent

unlikely that HCN is an atmospheric source of laboratory data for some rate constants.

CH3CN.

Brief Model Description Introduction

The model which is used in the present study is According to mass spectrometric observations one-dimensional and similar to the one used by

(Arnold et al., 1978; Arnold and Henschen, 1982; Brasseur et al (1983). The vertical transport is

one of the ligands of positive _' stratospheric (referred as K. in Brasseur et al, 1983). A large ₁ cluster ions. To validate this assumption, a number of reactions involving oxygen, hydrogen,

comprehensive understanding of the behavior of nitrogen, carbon and chlorine species are CH3CN is required. Brasseur et al (1983) have considered simultaneously (Brasseur et al., 1982)

suggested that acetonitrile could be produced at but only the processes directly related to the the Earth's surface, essentially by biomass and present study are listed in Tables 1 and 2. Two

industrial burning and destroyed in the tropo- cases are considered : in the first run (table sphere and the stratosphere by oxidation 1), CH•CN is assumed to be destroyed by OH, O(*D)

reactions, mainly with OH radicals. The overall and CI and, in the upper layers, by photo- budget of this molecule is nevertheless not yet dissociation process. The corresponding

completely understood, due to quantitative un- absorption cross section however has been certainties in the CH•CN release and the mixing measured (Zetzsch private communication) only in ratio in the lower troposphere. The two available a limited range of the spectrum, so that the observations of the acetonitrile surface mixing related photodissociation frequency can only be ratio (Becker and Ionescu, 1982; Snider and approximate and depends on interpolations, made

Dawson, 1984) differ by about a factor of 100, where data are missing. The precipitation although the measurements by Snider and Dawson scavenging of CHqCN in the troposphere appears to

(1984) seem to be more representative of a back- be small compared to other loss processes, except

Very recently, Murad et al (1984) suggested the low value of the Henry's law constant (Snider

that hydrogen cyanide (HCN) might be a source of and Dawson, 1984). Indeed, when following the

that the •orr.elsponding loss coefficient is lower

HCN + CH20 + M + CH2OH-CN + M. (1) than 10-- s TM at ground level and d•creases

rapidly with increasing altitude (Arijs and Another end-product of this mechanism would be Brasseur, in preparation). The corresponding CH•OH which also plays a rol e in the strato- average tropospheric lifetime is in reasonable

spheric ion chemistry (Henschen and Arnold, agreement with the recent estimation of Hamm et

1981a; Arijs et al, 1982). Moreover, Murad et al al (•984).

(1984) introduced the reaction with proton Hydrogen cyanide is assumed to be produced at

hydrates as a fast destruction process for CH•CN. the surface and destr?•ed in the atmosphere by

have examSned the reaction with OH and O(D) (Cicerone and Zellner,

Cicerone and Zellner (1983) • •

chemical and photochemical behavi,• • of HCN with- 1983). The photodissociation rate is not known

out considering the suggestion by • Murad, Their and will be considered as negligible (Jw• = O)

model calculations of the vertical distribution to obtain an upper limit of the concentr•ion or compare quite well with available observations as being equal to the photodissociation rate of

(Coffey et al, 1981; Carli et al, 1982; Rinsland hydrogen chloride (JH = J ). This last et al, 1982). Cicerone and Zellner (1983) derived assumption has been a•opted H• Cicerone and

that hydrogen cyanide has an atmospheric Zellner but their adopted value of the HC1 photo-

dissociation coefficient (J•c]) is significantly

lower than the currently •6cepted values for Copyright 1985 by the American Geophysical Union. J .... •1 Precipitation - of HCN in the troposphere is

considered to be a negligible sink.

Paper number 4L6392. The reactions considered in the second case, in 0094-8276/85/004L-6392503.00 addition to the chemistry used in case 1 are

117

118 Brasseur et al. : Hydrogen cyanide and acetonitri!e.

TABLE 1.- Reactions and corresponding rate constants adopted in case 1.

Reaction 3 -1

Rate constant (cm s ) (T in K; n(M) in cm ) -3

Reference

HCN + OH -- products

HCN + O(1D) -- OH + CN

HCN + hv --H + CN

CH3CN + OH -' products

CH3CN + O(1D) -- CH 2CN + OH

CH3CN + C1 -. CH 2CN + HC1 CH3CN + hv -' products CH3CN + wash-out

k n(M)

• o

k = 0.8

] + k n(M)/k

o oo

see below (k) •

1 x 10 -10

6.3 x 10 -13 exp(- 1030/T)

-10 1 x 10

8.0 x 10 -11 exp(- 3OOO/T)

see text

• = 4.5 x 10

-8 exp(- z2/]5.4)s

Cicerone and Ze]]ner (1983) Cicerone and Zellner (1983)

See text

Kurylo (1984) Working value

Olbregts et al. (1984) Zetsch, private communication

Calculated with method of Crutzen and Gidel (1983)

[1 + log 2 (k n(M)/k )]-1

o • Adopted from Cicerone

and Zellner (1983)

-31 6 -1

k = 1.5 x 10 exp(- 875/T) cm s

o -13 3 -1

k = 1.16 x 10 exp(- 4OO/T) cm s

given in Table 2. The important modifications are the link introduced between HCN and CH^CN as well

3 .

as the fast destruction process of acetonitrxle by proton hydrates.

For both molecules, a mixing ratio is prescribed as lower boundary condition at the

surface, namely 10 ppt and 170 ppt for CH3CN and

HCN respectively and a zero flux is assumed as upper boundary condition at 100 km altitude.

Model Results and Discussion

Figure 1 shows the vertical distribution of the

HCN and CHqCN mixing ratio obtained with the

reactions of Table 1 (case 1). These theoretical

profiles are compared with available observa- tions. As already indicated by Cicerone and Zellner (1983), the mixing ratio of hydrogen cyanide is almost constant in the troposphere and decreases progressively in the stratosphere. The corresponding slope however is steeper than in the paper by Cicerone and Zellner (1983) due to a

difference in the adopted J value. When JHCN is put equal to zero, the re•a•ive concentration

is considerably increased in the upper strato- sphere and is in much better agreement with the available observations. In both cases, the model suggests a residence time of about 2.7 yrs and an integrated loss in balance with an atmospheric release of the order of 0.26 MT/yr.

TABLE 2.- Additional reactions with their corresponding rate constants considered in case 2.

(cm 3 1

Reaction Rate constant s- ) Reference

HCN + CH20 - CH2OH-CN CH2OH-CN + hv - CH2CN + OH CH2OH-CN + hv --CH2OH + CN CH2OH + HO 2 --CH3OH + 02 CH2CN + HO 2 --CH3CN + 02

CH3OH + H+(H20)n -- H+.CH3OH. (H20) CH_CN ₊₃ + H +(H20) _n -' H + CH3CN _{' '} (H20)

H .CH30H

(H 20) + y- -- products

+ -

H . CH3CN. (H20) + y -- products

n-1 + H20 n-1 +H20

A : = 2 x 10 -31 n(M) B ß see below

yield ß 50%

yield ß 50%

Fast Fast 2 x 10 -9

-9 2xlO

-7 2 xlO 2x 10 -7

Murad et al (1984) Working value

Estimate; Murad et al (1984) Estimate; Murad et al (1984) Assumed, Murad et al (1984) Assumed, Murad et al (1984) Bohme et al (1979)

Estimate, Murad et al (1984) Smith and Adams (1982) Smith and Adams (1982)

-35 6 -1 -13

B = same expression as k (table 1) but k = 1 x 10 cm s and k = 5 x 10 exp(- 25OO/T)

o oo

Brasseur et al. : Hydrogen cyanide and acetonitrile. 119

5O ⁺

H (CH•CN)^(H•O) ions through reaction of aceto- _{mb • m}

nitrite with proton hydrates is irreversible, the latter is not an effective sink for CH3CN. This 40 can only be explained by the fact that CH3CN is

released again upon most of the recombinations of the H-(CH•CN)•(H•0) ions with negative ions.

• 30 Another •rg•men• •o reject a strong HCN/CH3CN

E coupling is based on chemical considerations.

• Reaction

• 20 CH2CN + HO 2 • CH3CN + 02 (2)

• which is assumed the final process for the

10 acetonitrile formation competes with reaction

CH2CN + 02 + M + OOCH2CN + M (3)

0

1•12 I•1 I0-10 1•9 The

rate constant

of this mechanism

is not known.

VOLUME MIXING RATIO for the CH + 0 rgact•on adopted

Fig. 1. Vertical distributions of HCN and CH3CN (2.6 x 10 -31 •/30• •3'• _ is

observational data. Case 1 refers to the In this case, reaction (1) still acts as a sink

chemistry described in Table 1. Two extreme for HCN but not as a source for CH3CN. The rate

values of the photodissociation frequency of HCN constant of reaction (1) however is most probably

are adopted (J-cN = 0 and J•CN(T•b•eHC« I. The lower than the value derived by Murad et al

dashed line refers to case when (1984) since the association of hydrogen cyanide

reaction rate B is adopted for the destruction of and formaldehyde is a 4- center process between

HCN by CH20. Other curves obtained for case 2 and closed shell molecules. Anticipating for such a

which are too far away from experimental data are reaction an activation energy of 5 kcal/mole for

not plotted. the high pressure rate constant and assuming in

the case of a reaction of such complexity a rate constant at 10 mbar which is 10-20 percent of the The calculated vertical distribution of CHACN high pressure limit, the rate constant labelled B is in fairly good agreement with the observed in Table 2 can be adopted as a working value. The

data derived from balloon borne ion mass spectro- loss rate in this case is reduced by 4 orders of

at the surface. The data obtained from high a dashed line (2B) in figure 1, does not depart

1983b) and rocket experiments (Arnold et al, 1977) reaction (1) is omitted. However, the lifetime of

loss mechanism in the upper stratosphere and in

the mesosphere. The calculated lifetime of CH3CN Conclusion

is 1.5 yrs and the ground level emission required

to balance the atmospheric sink is 0.045 MT/yr. Model calculations as well as an analysis of When considering the second reaction scheme the chemical scheme involved indicate that the

(case 2), the main loss process for HCN becomes link between HCN and CH3CN seems very unlikely.

its reaction with formaldehyde, if the ratg The best hypothesis for these two gases at

constant suggested by Murad et a! (2 x 10 -31 cm present is that they are produced at the Earth's

s -•) is adopted. Consequently, with this surface and destroyed in the atmosphere by

assumption, hydrogen cyanide whose lifetime in chemical reactions. A number of uncertainties the lower troposphere is as short as 20 seconds, however remain and should be elucidated in the is almost entirely destroyed near the ground and future. For example a quantitative analysis of cannot be involved as a source for CHACN in the all sources of these gases is required together stratosphere. With this chemical scheme, it thus with a study of the photodestruction and the seems difficult to reconcile model calculations heterogeneous removal of these species.

with the idea of HCN being a progenitor of CH3CN.

We have also considered the reaction of proton Acknowledgments. This work was partly supported

hydrates with CH3CN as a sink mechanism for by the Chemical Manufacturers Association under

acetonitrile, as suggested by Murad et al (1984). contract 83-468. R.Z. gratefully acknowledges a

from in situ measurements of Rosen and Hofman

(1981) below this altitude and a relative References

abundance of non proton hydrates compatible with

recent ion composition data (Arijs et al., 1983a Arijs, E., D. Nevejans, P. Frederick and J.

and b; Henschen and Arnold, l•lb), one obtains a Ingels, Positive ion composition measurements

mixing ratio of less than 10 - pptv above 20 km, between 33 and 20 km altitude, Annales Geo- in disagreement with the observations. physicae, •, 163-168, 1983a.

This indicates that, although the formation of Arijs, E., D. Nevejans and J. Ingels, Strato-

as

120 Brasseur et al. : Hydrogen cyanide and acetonitrile.

spheric positive ion composition measurements, Crutzen, P.J. and L.T. Gidel, A two-dimensional ion abundances and related trace gas detection,

J. Atmos. Terr. Phys., 44, 43-53, 1982.

Arijs, E., D. Nevejans and J. Ingels, Positive ion composition measurements and acetonitrile in the upper stratosphere, Nature, 383, 314- 316, 1983b.

Arnold, F., H. B•hringer and G. Henschen, Composition measurements of stratospheric positive ions, Geophys. Res. Lett., 5, 653-656,

1978.

Arnold, F. and G. Henschen, Positive ion

photochemical model of the atmosphere. 2. The tropospheric budgets on anthropogenic chloro-

carbons, CO, CH 4, CH3C1 and the effect of

various NO sources on tropospheric ozone,

J. Geophys XRes , 88, 6641-6661, 1983

Hamm, S., J. Hahn, G. Helas and P. Warneck,

A•etonitrile in the troposphere ß Residence

time due to rainout and uptake by the ocean, Geophys. Res. Lett., to be published, 1984.

Heaps, M.G., Parameterization of th• cosmic ray ion pair production rate above 18 kin, Planet.

composition measurements in the upper strato- Space Sci., 26, 513-517, 1978.

sphere - Evidence for an unknown aerosol Henschen, G. and F. Arnold, New positive ion component, Planet. Space Sci., 30, 101-108, species in the stratosphere, Nature, 291,

1982. 211-213, 1981a.

Arnold, D., D. Krankowsky and K.H. Marien, First Henschen, G. and F. Arnold, Extended positive ion mass spectrometric measurements of positive composition measurements in the stratosphere.

ions in the stratosphere, Nature , 267, 30-32, Implications for neutral trace gases, Geophys.

1977. Res. Lett., •, 999-1001, 198!b.

Baulch, D.L., R.A. Cox, P.J. Crutzen, R.F. Kurylo, M.J. and G.L. Knable, A kinetics Hampson Jr., J.A. Kerr, J. Troe and R.T. investigation o• the gas phase reaction of

Watson, Evaluated kinetic and photochemical Ci('P) and OH(X•) with CH3CN ß Atmospheric

data for atmospheric chemistry ß Supplement I, ß significance and evidence for decreased J. Phys. Chem., Ref. Data 11, 327-491, 1982. reactivity between strong electrophiles, •.

Becker, K.H. and A. Ionescu, Acetonitrile in the Phys. Chem., 88, 3305-3308, 1984.

lower troposphere, Geophys. Res. Lett., •, Murad, E., W. Swider, R.A. Moss and S. Toby,

1349-1351, 1982. Stratospheric sources of CH CN and CH3OH , B•hme, D.K., G.I. Mackay and S.D. Tanner, An Geophys. Res. Lett., 11, 147-1•, 1984.

experimental study of the •as phase kinetics of Olbregts, J., G. Brasseur and E. Arijs, Reaction reactions of hydrated H20' ions (n = 1-3) at of acetonitrile and chlorine atoms, J.Photo-

Brasseur, G,, E. Arijs, A. De Rudder, D. Nevejans Rinsland, C.P., M.A.H. Smith, P.L. Rinsland and and J. Ingels, Acetonitrile in the atmosphere, G.M. Stokes, Ground-based infrared spectro- Geophys. Res. Lett., 10, 725-728, 1983. scopic measurements of atmospheric hydrogen Brasseur, G., A. De Rudder and A. Roucour, The cyanide, J. Geophys. Res., 87, 11119-11125,

natural and perturbed ozonosphere, Proc. of 1982.

Int. Conf. on Environmental Pollution, Rosen, J.M. and D.J. Hofman, Balloon borne Thessaloniki (Greece), A. Anagnostopoulos ed., measurements of the small ion concentration,

p. 839-910, 1982. J. Geophys. Res., 86, 7399-7405, 1981.

Carli, B., F. Mencaraglia and A. Bonetti, New Snider, J.R. and G.A. Dawson, Surface aceto- assignments in the submillimeter emission

spectrum of the stratosphere, Int. J. IR.

Mill. Waves, •, 385-394, 1982.

Cicerone, R.J. and R. Zellner, The atmospheric chemistry of hydrogen cyanide (HCN), J.Geo- phys. Res., 88, 10689-10696, 1983.

Coffey, M.T., W.G. Mankin and R.J. Cicerone, Stratospheric detection of stratospheric hydrogen cyanide, Science, 214, 333-334, 1981.

nitrile near Tucson, Arizona, Geophys. Res.

Lett., 11, 241-242, 1984.

Smith+ D. and N.G. Adams, Ion recombination in the stratosphere, Geophys. Res. Lett., 9, 1085-1087, 1982.

(Received October 22, 1984 revised December 12, 1984 accepted December 18, 1984).