Vol.
2, No. 6
GEOPHYSICALRESEARCHLETTERS
THE CHEMISTRY
OF ATMOSPHERIC
June 1975
BROMINE
StevenC. Wofsy,MichaelB. McElroyandYuk Ling Yung
Center
forEarthandPlanetary
Physics
HarvardUniversity,Cambridge,Massachusetts
Brominemay act asa catalystfor recombina- 1973). Usingavailabledatafor the gasoline
sourcestrength,
and empiricalvaluesfor the atmosphericlifetime of HBr,
oxide or chlorine. The lower atmospherecontains small we estimatea sourcestrengthfor total gaseousbromineof
Abstract.
tion of ozone and could be more efficient than either nitric
concentrations
of gaseous
bromineproducedin part by mar- 1.1x 106tonsyrq , of which
approximately
80%isderived
ine activity, in part by volatilizationof particulatematerial from marine aerosols,10% from CH3Br, and 10% from
releasedduringthe combustionof leadedgasoline,with an automobiles.•additional contribution due to the use of methyl bromide
as an agriculturalfumigant. Observations
by Lazruset. al.
(1975)indicate
smallconcentrations
of bromine,
~10'11
The importance
of brominein the stratosphere
relatesto
its potential role as a catalystfor recombinationof ozone.
Catalysiscan take placeeither through
(v/v), in the contemporarystratosphere
and appearto imply
a reductionof approximately0.3% in the globalbudgetof
03. Estimatesare givenfor future reductionsin 03 which
mightoccurif the useof CH3Brasan agriculturalfumigant
were to continue to grow at presentrates.
Br
+ 03 --> BrO + 02
BrO+ O -->Br + O2
(1)
or through
2(Br + 03 • BrO + O2)
The chemistryof atmosphericozonemay be influenced
to a remarkableextent by exceedinglysmallquantitiesof
select stratosphericgases.Attention has focusedin recent
years on nitric oxide releasedby SST's (Johnston, 1971; The sequence
(1) is equivalentto
Crutzen, 1972) and on chlorineproducedby photodecomO
+ 03 --> 202
(1')
position of chlorofluoromethanes
(Molina and Rowland,
1974). This paperis devotedto a discussion
of bromine.As The sequence
(2) is equivalentto
we shallsee,concentrationsof bromineas smallas 1 part in
0 3 q- 0 3 • 302
(2')
1011(v/v)canhaveanappreciable
influence
onozone.
There are indicationsthat the mixing ratio of gaseous
formsof acidicbromine,representedmostprobablyby HBr, Reaction(2)could evolveBr2 ratherthan Br. The efficiency
according
to thediscussion
below,mayapproach
10'11 in of the cyclewould not be greatlyaffectedby uncertainties
today's stratosphere(Lazrus et. al., 1975) with similar in the identity of the bromineproduct, sinceBr2 will
valuesobservedin the troposphere(seebelow). The mixing photolyserapidly in the presenceof sunlight.
Chlorineatom chainsare terminatedin the stratosphere
ratio of acidicbromineappearsto increasewith increasing
by reactionswith CH4 and H2, which tie up the gasin the
altitude between 15 and 19 km and there is evidence for
reactionsof
similar behavior in the height distribution of particulate relativelyunreactiveform HC1. The analogous
bromine (Lazrus et. al., 1975; Delany et. al., 1974). It is bromine with CH4 and H 2 are endothermic, and Watson
difficult to escapethe conclusionthat there must be a (1975) wasled to concludethat BrO mightbe the major
stratosphericsource for particulate and gaseousacidic form of gaseousbromine over an extensiveheight range.
bromine.-• Decompositionof methyl bromide offers a The efficiencyof the catalyticcycles(1) and (2) could'be
plausible explanation. Methyl bromide has been detected exceedinglylargein this caseand mixing ratiosof bromine
as10'12 wouldhaveappreciable
effects
on03.
recently in surface waters of the ocean and in antarctic assmall
It
seems
more
likely
however
that
HBr
should
dominate.
snow(R. Rasmussen,
private communication,1975). It is
by
usedextensivelyas a fumigantand theremay be additional Hydrogenbromidecouldbe produced
natural sources as discussed below.
Br + HO2 -->HBr + 02
(3)
Bromineis presentin both gasand particulatephasesin
the troposphere,with the gasphasedominantby a factor and
which rangestypically between4 and 20. The mixingratio
Br + H202 ->HBr + HO2
(4)
BrOq-BrO•2Brq-O2
(2)
of gaseous
bromineis about1.5 x 10'11 overHawaii
(Moyers and Duce, 1972) and it seemsprobablethat the and would be removedby
major constituentis HBr (see below). Measurements
near
OH + HBr -->H 20 + Br
(5)
the South Pole (Duce et. al., 1973) suggestvaluessmaller
with
an
additional
contribution
from
by about a factor of 5. There are indicationsthat the gaseous componentis derivedprimarily from volatilization of
hv
+ HBr --> H + Br.
(6)
bromine carried into the atmosphereas a componentof
marine aerosols(Duce et. al., 1963, 1965) with additional
contributionsfrom decomposition
of CH3Brandfrom particles releasedduring the combustionof leaded gasoline •The contributionfrom gasolineto the gaseousbrominebudgetof
(Moyers and Duce, 1972; Cadle, 1972; Martens et. al., the northern hemispherecouldbe as largeas 20%. In carryingout
the computationssummarizedhere we assumeda gasoline-related
bromine
source
of 1.8x 105tonsyr-1 (Klingman,
1974)andestimated that half of this bromine would be volatilized before removal
•'GaseousBrO could contribute to the "particulate" bromine found
fr6m the atmosphere.We adopteda lifetime for inorganicgaseous
bromine of 2 weeks, consistentwith the analysisby Moyers and
on neutral filters.
Duce (1972).
Copyright 1975 by the American Geophysical Union
215
216
Wofsy et al.'
Atmospheric
Bromine
The chemicalmodel adoptedfor this study,taken for the taken from Hunten (1975), but multipliedby 1.5 above
16 km to obtain agreementwith CH4 observations.
The
mostpart from Watson(1975), is givenin Table 1.
chemical schemeallowed for 95 reactions,including those
Methyl bromideis removedby
listedin Table 1. The flux of inorganicbromineat ground
OH + CH3Br --> H20 + CH2Br.
(7)
levelwastakenequalto 5.3 x 10'7 molecules
cm'2 sec'l•:
and the flux of CH3Br wasadjustedto agreewith Lazruset.
The bromine atom is subsequentlyreleased,and inorganic al.,,(1975).Thesurface
fluxof CH3Brwasestimated
in this
bromide(HBr q- Br q- BrO) is carriedout of the lower mannerat 3.0 x 106molecules
cm'2 sec
'• , approximately
atmosphere
by rain asdiscussed
earlier.
four timesthe valuewe inferredfrom Klingman(1974) for
The major uncertaintiesin the chemicalmodel relateto the net amount of bromine consumed in the manufacture
reactions.(3) and (7), for which we adopt rate constants of CH3Br during 1974. Plonka(private communication,
10q • cm3 sec
q and2 x 10'• 2 exp(-1200/T)respectively.1975) is of the opinion'thatonly 25%of the industrial
The rate for (3) wassuggested
by a consideration
of high CH3Bris releasedto the atmosphere,
mainlyasa byproduct
temperaturedatareportedby Day et. al., (197 !). Therate of fumigation.
for (7) is consistent
withpreliminary
measurements
for the
v'alue
isthefluxderived
fromacalculation
using
thewashout
reaction
at 300øK(Watson
a.nd
Davis,1975)andagrees
also :['This
with the hightemperaturedataby Wilson(1965).
profile
•ivenin Table
1, a mean
ground-level
mixing
ratioof
(1975)eddydiffusion
profile.
Heightprofilescomputedfor majorbrominespecies
are 1 x 10-1, andHunten's
givenin Figure1. The computations
summarized
by this •'Burresonet al. (1975) reported very recently that CHBr3 is an
and subsequent
figureswere carriedout usingprocedures important component of certain seaweeds.If releasedto the atmosdescribedelsewhere(McElroy et. al., 1974; Wofsyand phere, CHBr3 would play a role analogousto that attributed to
McElroy, 1974; Wofsyet. al., 1975), with the eddyprofile marineCH3Br in the presentcontext.
Table 1
Reference
RateExpression
a
Reaction
kl Br + 03 • BrO + 02
1.2x 10'l 2(300OK)
1x 10'• o exp(-1320/T)[A,C]
2.5x 10'l • exp(-900/T)[B]
Watson(1975)
k2 NO + BrO• NO2 + Br
2 x 10'•
Watson(1975)
k3 Br + HO2 • HBr + 02
1 x 10'• 0 [C]
1 x 10'• [B]
Day et. al., (1971)
(seetext)
O. [A]
k4 BrO +O•Br
+02
k sBrO+BrO•2Br+
BrO+he•
02
Br + O
k6 OH + HBr--> H20 + Br
k? OH + CH3Br -->H20 +CH2Br(c)
ks Br + H202 -->HBr + HO2
J1 HBr + hv • H + Br (b)
8x10 -•
Watson(1975)
6.4 x 10'12
Watson(1975)
rate unknownin the atmosphere,
probablyslowerthank 2 according
to Watson(1975)
3.7x 10'• • exp(-600/T)
2 x 10'• 2 exp(-1200/T)
5.5x 10'• 2 exp(-1900/T)
1 x 10'l • exp(-1200/T)
0, 0, 2.8x 10'9 4.9 x 10'6
Watson(1975)
(seetext)
(seetext)
(seetext)
Romandand Vodar (1948)
3.8x10 '6 6.7x10 '6
J2 CH3Br q-hv -->CH3 q-Br
0,0,2.2x10 '9 4.1x10'8
Davidson(1951)
7.3x10 '6 1.8x10 's
R• HBr q-rain, aerosol•
removedfrom atmosphere
1.7x10 '6
0 •< Z •< 4km
8.5x10 '?
4.3x10 '?
1.7x10 '?
4km<Z
6km<Z
8km<Z
0
kgOq-HBr• OHq-Br
10km<
•
•
•
(seetext)
6km
8km
10km
Z
4.0x 10'• 2 exp(-1360/T)
BrownandSmith(1975)
aunits
arecm3 sec
'1. Letters
A, B,Crefertocurves
inFigure
2.
bphotodissociation
rateisgiven
asafunction
ofaltitude
(inkm),units
ofsec
'1(0,10..... 50km).A24hour
average
isused
at30ølatitude,
0øsolardeclination.
CBratomassumed
released
duringsubsequent
chemistryof CH2Br.
Wofsy et al.:
Atmospheric
Bromine
217
•
' E7 D7
D15
Figure1: Heightprofilesfor Br, HBr, BrO, and(Br +
HBr + BrO):The data(A) arefrom Lazruset. al., (1975).
10
1980
Computationsallow for transportof N2O, NOx, CH4, CO,
I
1990
2000
TIME
2010
(YR)
2020
2030
Figure 3: Reductionin ozonedueto releaseof anthro03, (HBr + Br + BrO) and CH3Br, and accounted
for
completeHOx and NOx chemistry.The mean lifetime pogenicCH3Br.CurvesD andE differ in the valuesassumed
calculated
for CH3Br(columnconcentration/flux)
is 1.1yr. for k?: D, samerate as CH4;E , preliminarymeasurement
reportedby WatsonandDavis(1975). Otherratesarethe
Taken at face value, the analysisgivenhere impliesboth
natural and industrialsourcesof CH3Br. The sea is the
sameascurveB, Figure2. Sevenand fifteen per centannual
increasesare modelled by D7, E7, and D15, El5 respec-
most probablenatural source,providing75-95% of the
CH3Br shownin Figure1.? We cannothoweverexclude
the possibilityof additionalanthropogenic
contributions
to
stratospheric
bromine, associatedfor example with the
escapeof (CH•Br)• fromgasoline
storage
tanksor with the
productionof relativelyinsolublebrominespeciesduring
in 1974, approximately50% of the total globalproduction.
The releaserate estimatedby Plonka(seetext) wouldimply
a shift of the time axis by approximately10 yearsfor D7
and E7, 5 yearsfor D15 and E15.
tively.Theanthropogenic
release
wassetequalto 104 tons
the combustionof leaded gasoline.
release
fromautoexhausts.
Accordingto the model,the mixingratio of inorganic CH3Brand0.02%to bromine
in 03, computed
withvarious
valuesfor k • and
bromine decreases
by only a factor of 5 betweenground Reductions
and tropepause.This profile resultsfrom the choiceof k3, are shownin Figure 2. The resultsin Figure 2 were
heterogeneous
losscoefficient(cf. Wofsyet al., 1972),and derivedsubjectto the constraintof a heightindependent
is consistentwith observationaldata for other soluble spe- valuefor the mixingratio of HBr + Br + BrO, and illustrate
cies,HC1 (Farmer,1975), NH3 (GeorgiiandMuller,1974) the dependenceof the computedozone deficit on this
and HNO3 (Lazruset. al., 1974), and in agreementalso parameter.
with radioactivity removal rates summarizedby Junge
Figure3 givesseveralconceivable
time profilesfor the
ozone deficit associatedwith agriculturaluse of CH3Br.
(1963).
Theresults
in Figure1 implya reduction
of magnitudeCurves
D andE differin choice
of rateconstant
for reaction
0.3%in theglobalconcentration
of 03 dueto bromine
cat- of OHwithCH3Br.CurveE employs
thevaluesuggested
by
alysis,of whichapproximately
0.2%maybe attributedto analogy
withtheOH-CH3C1
reaction
andisconsistent
also
with the preliminary measurementby Watson and Davis.
INDUSTRIAL
RELEASE
OFCH3BR
(105METRIC
TONS/YR)
0
25
'
'
'
50
I
'
'
'
• 75
I
CurveD usesthe OH-CH4 reactionrate. Productionand use
of CH3Bris assumed
to growin curvesD7 andE7 at a rate
....
-
2
c
•-8
o
of 7% per annum, a value close to the historical trend.
Curves D15 and El5 envisagea somewhatlarger growth
rate, approximately15% per annum.A growthrate of this
magnitudemay not be unreasonablehowever,sincecost
factorscurrentlylimit fumigationto a very few high value
crops.Approximately10% of world bromineproductionis
used in the manufactureof CH3Br, with roughly 63%
employedin the productionof ethylenedibromide.
It seemsclearthat an unconstrained
growthin the useof
methyl bromide could causefuture problemsfor ozone.
Measurements
of brominecompoundsin the atmosphere,
togetherwith laboratorystudiesof severalkey reactions,
notablykl, k3 andk?, shouldbe pursuedin orderto clarify
-12
-14
these matters.
1 ;> 3 4 5 O•
ø 7 8 9 10
'31
MIXING RATIO (I
v/v)
Figure2: Ozonereduction
asa function
of (HBr-t-Br
-t-BrO).Differentcurves
referto different
combinations
of
rate expressions
for k3 andk l, asshown• Table1. Curve
Acknowledgements
B combines "best estimates" for these rates Curve A shows
We are indebted to C. Kolb, I. Chet and R. Mitchell for
modelwhereHBr formationdoesnot occur(k 3 andks set
valuablediscussions,
and to S. Coroniti and J. Plonkafor
useful communications.This work was supportedby the
to zero).CurveC illustrates
the effectof a choicefor k3
andk] designed
to minimizethe ozoneperturbation.
The NationalAeronauticsand SpaceAdministrationunderGrant
upperaxisshows
theapproximate
industriM
re]ease
rateof NASA NSG 2031 and by the Atmospheric SciencesDiviCH3Brwhichwou]dproducethegivenstratospheric
pertur- sion of the National Science Foundation under Grant
bation,
using
k?= 2 x ] 0'l 2 exp(-1;ZOO/T).
GA33990X to Harvard University.
218
Wofsy et al.:
Atmospheric Bromine
References
Klingman,
C.L.,u.S. Bureau
of MinesMineralYearbook,
U.S. Dept. of the Interior, 1974.
Brown, F.D.H. and I.W.M. Smith, Absoluterate constants Lazrus, A., B. Gandrud, R. Woodard, and W. Sedlacek,
for thereactions
of O(3p)atoms
withHC1andHBr,
Measurementsof stratospheric
halogensand nitric acid,
Int. J. Chem. Kin., 7, 301, 1975.
paperpresentedat CIAP Conference,Cambridge,Mass.,
Burreson,
B.J.andR.E. Moore,Haloforms
in essential
oil of
1975.
thealgaA.Taxiformis,Tetrahedron
Lett., 7, 473, 1975. Martens, C.S., J.J. Wesolowski,R. Kaifer, and W. John,
Catlie, R.D., Compositionof the stratospheric
"sulfate"
Lead and bromine particle size distribution in San
layer,Trans.of Am. Geophys.U., 53, p. 812, 1972.
Francisco Bay area, Atmospheric Environment, 7, p.
Crutzen,P.J., SST's,a threat to the earth'sozoneshield,
905, 1973.
Ambio, 1, p. 41, 1972.
McElroy, M.B., S.C.Wofsy,J.E. Penner,and J.C. McConnell,
AtmosphericOzone:possibleimpact of stratospheric
aviation,J. Atmos. $ci., 31, p. 287, 1974.
Chem.Soc.,73,p. 468, 1951.
Molina, M.J., and F.S. Rowland, Stratosphericsink for
Day, M.J., D.V. Stamp,K. Thompson,
andG. Dixon-Lewis,
chlorofluoromethanes:chlorineatom-catalyseddestrucInhibition of hydrogen-airand hydrogen-nitrous
oxide
tion of ozone,Nature, 249, p. 810, 1974.
flames by halogencompounds,
l Sth Comb. Syrup., Moyers,J.L. and R.A. Duce, Gaseousand particulateIodine
p. 705, 1971.
in the marine atmosphere,J. Geophys. Res., 77, p.
Delany, A.C, J.P. Shedlovsky,and W.H. Pollock,Strato5229, 1972.
spheric
aerosol:thecontribution
fromthe troposphere,Romand,J., and B. Vodar, Spectred'absorptionde l'acide
J. Geophys.Res., 79, p. 5646, 1974.
chlorohydriquegazeux dans la r•gion de Schumann,
Duce, R.A,, J.T. Wasson,J.W. Winchester,
and F. Burns,
C.R. (Acad. Sci.J, 226, p. 238, 1948.
Atmospheric
iodine,bromineandchlorine,J. Geophys. Watson, R.T., Chlorine, the chlorine oxides and other
Res., 68, p. 3943, 1963.
halogen species,CIAP Monograph, vol. 1, to be
Duce, R.A., J.W. Winchester,and T.W. Van Nahl, Iodine,
published.
bromine and chlorine-the
Hawaiian marine atmosWatson,R.T., and D.D. Davis,Meetingof the Am. Chemical
phere,J.Geophys.
Res., 70, p. 1775, 1965.
Soc.,Philadelphia,Pa., April 9, 1975.
Duce, R.A., W H. Zoller,andJ.L. Moyers,Particulateand Wilson, W.E., Structure, kinetics, and mechanismof a
gaseoushalogensin the Antarctic atmosphere,J.
methane-oxygen
flame inhibited with methyl bromide,
Geophys. Res., 78, p. 7802, 1972.
10th Comb.Symp., p. 47, 1965.
Farmer,C.B., CIAP Conference,Cambridge,
Mass.,1975.
Wofsy, S.C., J.C. McConnell and M.B. McElroy, AtmosGeorgii,H.W., and W.J. M/iller, On the distribution
of
pheric CH4, CO and CO2, J. Geophys.Res., 77,4477,
ammonia in the middle and lower troposphere,Tellus
1972.
XXVI 1-2, p. 180, 1974.
Wofsy, S.W.and M.B. McElroy, HOx, NOx, and C1Ox:their
Hunten, D.M., Philosophyof one-dimensional
modelling,
role in atmosphericphotochemistry,
Can.J. Chem.,52,
4th CIAP Conference,Cambridge,Mass.,1975.
p. 1582, 1974.
Johnston,H.S., Reductionof stratospheric
ozoneby nitro- Wofsy, S.W., M.B. McElroy, and N.D. Sze, Freon consumpgenoxidecatalysts
fromSSTexhaust,
Science,
173,
tion: implicationsfor atmosphericozone, Science,187,
p. 517, 1971.
p. 535, 1975.
Davidson,N., The ultraviolbtabsorptionspectraand the
refractive indicesof somefluorobromomethanes,J. Am.
Junge,C.E.,Air Chemistry
and Radioactivity,
p. 298ff,
AcademicPress,New York, New York, 1963.
(Received April 22, 1975;
accepted May 16, 1975)