United States
Environmental Protection
Agency
Atmosheric Research and
Exposure Assessment Laboratory
Research Triangle Park, NC 27711
Research and Development
EPA/600/SR-92/222 February 1993
EPA Project Summary
Atmospheric Chemistry and
Physical Fate of HCFCs and
HFCs and Their Degradation
Products
E.G. Edney
Laboratory experiments were con-
ducted to determine the fate of the pro-
posed CFC substitutes HCFC-22, HCFC-
123, HCFC-124, HCFC-141b, HCFC-142b,
HFC-125, HFC-134a, and HFC-152a. The
laboratory program consisted of (1)
photochemical oxidation experiments
to identify stable oxidation products
and measure their yields; (2) deposi-
tion studies to measure the extent oxi-
dation products are absorbed into aque-
ous media; and (3) experiments to de-
termine the fate of hydrolysis products
during evaporation of aqueous media.
An elementary multimedia model was
used as a tool for discussing the impli-
cations of the laboratory findings.
The product studies showed that acid
halides, including COF2, COFCI,
CF3CFO, HFCO, and CF3CCIO, are ma-
jor oxidation products of HCFCs and
HFCs. The deposition studies were
compatible with significant uptake of
the acid halides to aqueous media, fol-
lowed by hydrolysis reactions leading
to the formation of HF, HCI, and
CF3COOH. Model results, obtained us-
ing laboratory derived lower limits esti-
mates for aqueous deposition veloci-
ties and assuming a well mixed atmo-
sphere, suggest the uptakes rates to
cloudwater and oceans are sufficiently
fast such that significant buildup of
gas phase products is unlikely. How-
ever, for a global emission rate of 106
tonnes yr1, precipitation concentrations
as high as 20 pmol ml'1 of the appar-
ently long lived species CF3COO" are
possible. The laboratory studies sug-
gest that product accumulations in
aqueous media could be affected by
losses during evaporation. Direct loss
by evaporation of halogenated acids
and/or production of volatile com-
pounds after further reactions of the
dissolved acids could return haloge-
nated compounds to the atmosphere.
This Project Summary was developed
by EPA's Atmospheric Research and
Exposure Assessment Laboratory, Re-
search Triangle Park, NC, to announce
key findings of the research project
that is fully documented in a separate
report of the same title (see Project
Report ordering information at back).
Introduction
There is little doubt that chlorofluorocar-
bons (CFCs) released into the environ-
ment contribute significantly to destruction
of the stratospheric O3 layer. Any thinning
of the layer is apt to increase adverse
effects associated with exposure to UV-B.
The first international agreement aimed at
reducing CFC emissions, The Montreal
Protocol on Substances that Deplete the
Ozone Layer, was agreed to in 1987. How-
ever, by the time the treaty was to be
implemented, O3 depletion in the Antarc-
tica by Cl and Br compounds was of suffi-
cient concern to amend the treaty to ac-
celerate the phase-out time tables. Under
the 1990 London Amendments, CFCs,
halons, and carbon tetrachloride are
scheduled to be completely phased out
by the year 2000, with methyl chloroform
phased out five years later. The 1990
Clean Air Act Amendments (CAAAs) also
established similar phase-out schedules
Printed on Recycled Paper
-------
for Oa depleting compounds, with the ex-
ception of more rapid phase-out time tables
for hydrochlorofluorocarbons (HCFCs) and
methyl chloroform. However, recent O3
depletion measurements have led the U.S
to consider phasing out O3 depleting com-
pounds at an even more rapid rate. At
present, discussions are underway to
implement options in the 1990 CAAAs to
phase out CFCs by the end of 1995.
As it became clear that CFCs and
batons would be phased out, industry initi-
ated programs aimed at finding substitute
compounds. The focus of replacement
strategies was largely on developing H
atom bearing compounds that react in the
troposphere. The compounds that have
received the most attention are the HCFCs
and the hydrofluorocarbons (HFCs). Al-
though the removal of HCFCs and HFCs
by reaction with OH will decrease their
tropospheric concentrations and reduce the
possibility of the compounds reaching the
stratosphere, they do not eliminate the
possibility their release could affect strato-
spheric O3 depletion, global climate
change, human health, damage to the eco-
system, and tropospheric air and water
quality. Rather it shifts such questions to
the impact of the oxidation products. To
establish the information necessary to de-
termine the life cycles of the proposed
substitutes, a multimedia research pro-
gram was initiated. The program consisted
of photochemical oxidation experiments,
product deposition studies, evaporation
experiments, and multimedia modeling
studies.
Procedure
Laboratory Experiments
Laboratory experiments were conducted
to identify and determine the yields of the
major stable gas phase products of the
following CFC substitutes: HCFC-22,.
HCFC-123, HCFC-124, HFC-125, HFC-
134a, HCFC-141b, HCFC-142b, and HFC-
152a, The laboratory experiments were
conducted by irradiating ppm levels of the
compounds in the presence of CI2 in dry
air and identifying the products by long
path Fourier transform spectroscopy. Cl
atoms, formed by photolysis of CI2, mimic
the action of OH removing H atoms from
the compounds being studied. The major
products detected and their room tem-
perature yields in 700 Torr of air (e) were:
COF,, e - 1.11 ± 0.06, (HCFC-22);
CFjCCIO, e- 1.01, (HCFC-123); CF3CFO,
e - 1.00 ± 0.04, (HCFC-124); COF2, e =
1.09 ±0.05, (HFC-125); CF3CFO, E « 0.16
± 0.03, HFCO, e - 0.83 ± 0.22, COF2, e -
0.23 ± 0.02, (HFC-134a); COFCI, E =
0.94, (HCFC-141b); COF2, e « 0.98 ± 0.03,
(HCFC-142b); and COF2, e = 1.00 ± 0.05,
(HFC-152a).
The critical step in determining the oxi-
dation products was the decomposition of
the halogenated alkoxy radical generated
after H atom abstraction from the HCFC
or HFC. Alkoxy radical decomposition
modes included C-C bond cleavage, Cl
elimination, and H atom abstraction by
reaction with O2. The decomposition path-
way for HCFC-123, HCFC-124, and
HCFC-22 halogenated alkoxy radicals was
Cl elimination. In the case of HFC-125, C-
C bond cleavage was the major decom-
position route, whereas for HFC-134a, both
C-C bond cleavage and-reaction with O2
contributed. Secondary Cl reactions in the
HCFC-141D, HCFC-142D, and HFC-152a
experiments prevented unambiguous de-
terminations of the decomposition modes
of the associated halogenated alkoxy radi-
cals. The data were compatible with both
C-C bond scission and Cl reactions with
halogenated aldehydes producing COF2
from HCFC-142D, and HFC-152a and
COFCI fram HCFC-141 b.
The detected oxidation products were
in general resistant to further gas phase
oxidation. In order to determine whether
uptake to aqueous media could serve as
a sink for oxidation products of HCFCs
and HFCs, product deposition studies were
carried out. The experimental program in-
cluded single component exposure experi-
ments where ppm levels of the oxidation
product were exposed to quiescent aque-
ous solutions. The chemical composition
of the aqueous solution was monitored as
a function of exposure time using ion chro-
matography. Single component experi-
ments consisted of exposing CF3CCIO,
CFgCFO, and COF2 to deionized water
and to acidic and alkaline solutions. The
second phase of the program consisted of
first irradiating HCFC/CI2 and HFC/CI2 mix-
tures in clean air and then exposing the
irradiated mixture to deionized water. Irra-
diation/deposition experiments were con-
ducted using HCFC-22, HFC-41, HCFC-
123, HCFC-124, HFC-125, HFC-134a,
HCFC-141D, HCFC-142b, and HFC-152a.
The exposed aqueous solutions were ana-
lyzed for HCOO-, F-, and CF3COO' and
the anionic concentrations were then used
to calculate the effective deposition ve-
locities to bulk aqueous solutions.
The deposition results showed that un-
der laboratory conditions the products were
taken up into the aqueous solutions at
measurable rates. The uptake rates were
consistent with the formation of haloge-
nated acids by hydrolysis. The effective
deposition velocities in cm sec'1, for T =
295 ± 3 °K, were: 0.100 (COF2); 0.051
(COFCI); 0.068 (CF3CFO); 0.049
(CF3CCIO); and 0.052 (HFCO). Although
the deposition measurements were con-
sonant with removal of HCFC and HFC
oxidation products by aqueous media, the
data were not sufficient for extrapolating
to ambient conditions. However, by as-
suming the total resistance to deposition
was due to aqueous processes, a lower
limit of 0.05 cm sec-I was obtained for the
reciprocal of the aqueous resistance. For
an aqueous diffusion coefficient of 1 x 10"
5 cm2 sec'1 and T = 295 °K, the lower limit
was consistent with HVk,, = 0.65 M atnr1
sec-1/2, where H is the Henry's law con-
stant and kh is the first order hydrolysis
constant.
For aqueous media such as rivers,
streams, and oceans, deposition of an
acid halide is likely to be a permanent
sink because of the solubility of'the'hy-"
drolysis products in these large aquatic
bodies. However, for other tropospheric
aqueous media, including cloudwater, dew
and precipitation, evaporation can play a
major role in determining the fate of the
hydrolysis products. Whether uptake to
these media constitutes a permanent at-
mospheric sink depends on the fate of
hydrolysis products during evaporation.
Evaporation effects were investigated
by conducting two laboratory experiments
where aqueous films containing NH4NO3,
H2O2, and either sodium fluoride (NaF) or
sodium trifluoroacetate (NaCF3COO) were
exposed for five hours to irradiated smog
mixtures of o-xylene/NOx/SO2. The expo-
sure experiments were carried out by ex-
posing on stainless steel panels films of
aqueous solutions containing equal con-
centrations of NH4NO3 and either NaF or
NaCF3COO, but different H2O2 levels. At
the end of the exposure, the films were
collected and an aliquot was analyzed.
The exposed films consisted of solutions
containing dissolved oxidants, acids, and
salts in the presence of either F' or
CF3COO-, with the films differentiated by
and large by the H2SO4 levels. The un-
used portions of the aqueous films were
then returned to the panels and allowed
to evaporate overnight at room tempera-
ture, after which they were rinsed with Dl
water. This rinse was then chemically ana-
lyzed to determine the composition of ma-
terial left after evaporation and by infer-
ence what material was re-emitted into
the atmosphere.
The laboratory data agreed with the ex-
pected result that CF3COO' and F- do not
react with constituents in the bulk films.
However, chemical analyses of the post
evaporation Dl rinses revealed significant
losses for HCOO", CH3COO-, NO2', NO3',
F-, and CF3COO-, thus suggesting that
acid halide deposition to aqueous media
-------
that undergo evaporation may not serve
as a permanent sink under all circum-
stances. The NO3-, NO2-, HCOO', CH3COO'
, and P losses were congruous with vola-
tilization of the corresponding acids. How-
ever, the evaporation data were insuffi-
cient to establish a loss mechanism for
CF3COO\ Possible loss mechanisms in-
cluded direct evaporation of the corre-
sponding carboxylic acid and/or oxidative
decarboxylation reactions of CF3COOH
with (1) dissolved metals ions from acid
dissolution of the stainless steel panels or
(2) other oxidants. If oxidative decarboxy-
lation reactions occur, then the fate of CF3
could become an important issue. How-
ever, whether ambient hydrometeors un-
dergo decarboxylation reactions' has yet
to be established. Nonetheless, even in
the absence of suchTeactions, it appears
likely that, under acidic conditions, evapo-
rating ambient hydrometeors including
dew, cloudwater, and precipitation could
lose CF3COO' by volatilization of
CF3COOH.
Results and Discussion
To determine the risk associated with
the release of HCFGs and HFCs the labo-
ratory results were integrated into an el-
ementary multimedia model. The model
takes into account photochemical oxida-
tion of the emitted compound and uptake
of the products into cloudwater and
oceans. The atmosphere is assumed to
be completely mixed. The oceanic depo-
sition process is described in terms of the
transport through and reactions within the
following three layers: a thin stagnant air
layer directly above the ocean-air inter-
face; a thin chemically inert stagnant aque-
ous layer just below the ocean-air inter-
face; and the ocean mixed layer that cov-
ers the portion of the ocean from the bot-
tom of the stagnant layer down to the
thermocline. In the cloud model, the_ hy-
drolysis rate is assumed to be sufficiently
slow and aqueous diffusion within the drop-
let is rapid enough so the dissolved prod-
uct concentration remains in equilibrium
with the gas phase concentration.
The modeling results showed that for H
= 1 M atrrr1, a liquid water content of 2 x
10'7; an effective scale height of 10 km;
and T = 295 °K, the criterion k,, > 10'7 seer1
must be satisfied so the aqueous removal
lifetime is less than two years. For the low
k,, region, oceanic uptake was governed
largely by the resistance of the mixed
oceanic layer, whereas for k,, > 10"6 seer1,
removal by the oceans was determined
by the aqueous stagnant layer and atmo-
spheric resistances. Uptake to cloudwater
was significant only if k^ exceeded
sea1.
For these same conditions and a global
emission rate of 106 tonnes yr1, the model
results showed that for an atmospheric
lifetime (t) of one year the steady state
concentration of the emitted compound
quickly reached 56 pptv, whereas for a
lifetime of 20 years it took about 50 years
to reach the steady concentration of 1120
pptv. The model simulations further
showed that, for T = 1 year and kh < 10'7
sec*1, the oxidation products were the
dominant atmospheric species. Between
k,, = 10'9 and 10'7 sec'1, the product re-
moval rate was controlled by the resis-
tance of the mixed oceanic layer and prod-
uct concentrations decreased an order of
magnitude to a level similar to that of the
reactants. Between 10'7 and 10'3 sec'1, the
product concentration decreased another
order of magnitude, with the hydrolysis
rate becoming sufficiently fast such that
oceanic uptake of the product was even-
tually controlled by the resistances of the
atmosphere and the aqueous stagnant
layer. For k,, > 10'3 sec'1, the already low
product levels decreased further. In this
regime, the products hydrolyzed fast
enough to be absorbed into cloudwater.
For an atmospheric lifetime of 20 years,
the reactant concentrations exceeded the
product concentrations for all k,,.
The model was also used to predict
product concentrations in precipitation. For
a precipitation rate of 100 cm yr1, a global
emission rate of 106 tonnes yr1, H = 1 M
atrrr1, and T = 1 year, the condition kh >
10'3 sec'1 was required for substantial up-
take to precipitation. The resulting maxi-
mum precipitation concentration was 19
pmol ml'1.
Conclusions and
Recommendations
Buildup of HCFCs and HFCs will be
limited by their OH based atmospheric
lifetimes that range from about 1 to 30
years. For an emission rate of 106 tonnes
yr1, the corresponding steady state con-
centrations vary from 56 to 1680 pptv.
Acid halides will constitute a major portion
of the oxidation products. The laboratory
deposition data for COF2. COFCI, HFCO,
CF3CFO, and CF3CCIO agree with the
proposition that acid halides will be taken
up into aqueous media and hydrolyze,
forming halogenated acids. The labora-
tory derived lower limit estimate for vd,
while adequate for estimating a room tem-
perature limit for HVkh, was not sufficient
for estimating individual values for k,, and
H, both required input parameters for the
multimedia model. However, for H = 0.1,
1, and 10 M atnr1 and the corresponding
values for !<„ that were consistent with the
lower limit for HVk,,, the atmospheric prod-
uct concentrations were all less than a
few percent of the total concentrations.
The limit data were not sufficient to estab-
lish the partitioning between product up-
take to cloudwater versus oceans.
The partitioning of acid halides among
aqueous media is an important issue be-
cause of concerns surrounding the pos-
sible toxicity of the long lived species
CF3COO', a common hydrolysis product
of CF3CFO and CF3CCIO. The model
simulations showed that for an emission
rate of 10e tonnes yr1 and precipitation
levels of 100 cm yr1, CF3COO precipita-
tion concentrations as high as 20 pmol ml'
1 are possible. However, this value is prob-
ably an upper limit because the CF3COO'
producingi HCFCs and HFC represent only
a portion of total emissions and,the aver-
age atmospheric yield from HFC-134a, the
compound that is likely to have the most
widespread use, is only 0.27. Furthermore,
the possibility remains that a portion of
CF3COO" precursors are taken up into
oceans and other large aquatic bodies,
thus reducing precipitation concentrations,
but also increasing the accumulations in
other possibly sensitive aquatic bodies.
The impact of evaporation should be
considered in assessing the risks associ-
ated with the release of CFC substitutes.
Most aqueous droplets that come in con-
tact with ecological surfaces, for instance,
plant leaves, undergo evaporation. The
laboratory evaporation data suggest that
loss of hydrolysis products can occur di-
rectly by evaporation of the corresponding
acid and/or by volatilization of a reaction
product from further reactions of the hy-
drolysis products. Thus, clouds could serve
as processing centers where COF2 is con-
verted to HF and CO2, COFCI to HF, HCI
and CO2, etc.
Further investigations are needed to es-
tablish the life cycles of HCFCs and HFCs.
Chemical kinetic studies are needed to
further clarify the degradation mechanisms
for HCFC-141b, HCFC-142b, and HFC-
152a, with special emphasis placed on
establishing the gas and aqueous phase
chemistry of halogenated PANs. Research
is needed to determine the fate of CF3O2
in the presence of NOX compounds. Fur-
ther refinements of the deposition studies
are needed to measure the individual val-
ues for k,, and H. The evaporation studies
should be expanded to include studies to
establish the key chemical and physical
parameters controlling the fate of com-
pounds during droplet evaporation. Em-
phasis should be placed on determining
the fate of CF3COO' during evaporation.
Finally, research is required to address
general issues related to the development,
implementation, and validation of multi-
•U.S.Government Printing Office: 1993—750-071/60188
-------
madia models. In particular, both gas and
aqueous transport should be addressed
in a rigorous fashion and the method used
to parameterize aqueous uptake should
ba improved.
The EPA author E.O. Edney (alsatheEPA Project Offic9ri^ee,below)js with the_
Atmospheric Research and Exposure Assesment Laboratory, Research Triangle
Park, NC 27711.
The complete report, entitled "Atmospheric Chemistry and Physical Fate of HCFCs
and MFCs and Their Degradation Products, "(Order No. PB93-131449/AS; Cost:
$17.50; subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Atmospheric Research and Exposure Assessment Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States
Environmental Protection Agency
Canter for Environmental Research Information
Cincinnati, OH 45268
Official Business
Penalty for Private Use $300
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT No. G-35
EPA/600/SR-92/222
------- |