United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research-Triangle Park, NC 27711
Research and Development
EPA-600/S3-80-084 Dec, 1980.
Project Summary
Fate of Toxic and Hazardous
Materials in the Air
Environment
Larry T; Gupitt
Potential toxic/hazardous
chemicals are currently undergoing
Type I (preliminary) and Type II
(detailed) assessment by the Office of
Air Quality Planning and Standards.
This report evaluates the atmospheric
fate of these compounds, i.e., their
probable lifetime in the troposphere.
Emphasis has primarily been given to
the volatile chemicals undergoing
Type II Assessment, i.e., acrylonitrile.
ethylene dichloride, perchloroethyl-
ene, vinylidene chloride and benzo(a)-
pyrene (as a representative polycyclic
organic material).
Chemical and physical removal
processes are discussed. Chemical
removal processes which are
considered include photolytic trans-
formations and reactions with
hydroxyl radicals, ozone, and other
tropospheric species. Mathematical
descriptions of physical removal
mechanisms are developed and
applied to the volatile Type II Assess-
ment chemicals. Physical removal by
rainfall, by dry,deposition, and by
adsorption on aerosol particles are
generally demonstrated to be rather
inefficient.
Forty-six individual materials were
evaluated relative to their probable
fates and tropospheric lifetimes.
Known or theoretical rate constants
are listed for reaction with hydroxyl
radicals and ozone. The probability of
and of physical removal is
assessed, and residence lifetimes
assigned. Probable products of tropo-
spheric oxidation processes are also
tabulated.
Introduction
The Office of Air Quality Planning and
Standards (OAQPS).has recently distri-
buted lists of potentially toxic/
hazardous chemicals being assessed in
regard to possible regulatory action. The
list of "43 chemicals" under Type I (pre-
liminary) Assessment, and a second list
of chemicals under Type II (detailed)
Assessment, contain a variety of
species that may .be found in .the air
environment. Because of the, high
probability of harmful effects resulting "
from exposure to these chernjcalsy it-is
important to evaluate the atmospheric
fate of these compounds, particularly
those chemicals currently, under Type II
Assessment that may possibly be found
•as gaseous emissions, (e.g., acrylo-
nitrile, ethylene .dichlpride, perchloro-
ethylehe, vinylidene chloride, and
benzo(a)pyrene (BaP) as a representa-
tive polycyclic organic material).
Chemical Removal Processes
For a wide variety of molecules'/'ithe.,
most important, chemical' removal
process in the troposphere is reaction
with hydroxyl (OH) radicals. For those
-------
organic chemicals containing isolated
double bonds, ozonolysis is also avail-
able as a second reaction pathway. In
addition, reactions with minor, tropo-
" spheric species or with species known
to exist in the stratosphere (e.g., 0(3P),
0('D), HOI ROa, N03, 02 ('A g), etc.)
must be considered for. chemicals that
•are-relatively long-lived.
Reaction Wfth Hydroxyl
Radicals
Despite the fact that the average con-
centration of OH radicals is much lower
than'for many other reactive species,
reaction with hydroxyl radicals is often
so:rapid that it isthe predominanttrppo-
sphBrrc removal,mechanism for a wide
variety of organic molecules.
Atmospheric reactions with OH are(a
-combination of abstraction and addition
processes leading primarily to
oxygenated o/ganic compounds like
aldehydes, ketones, and dicarbonyls.
Halogenated organics tend to lose
halogen atoms in the form of halo-oxy
radicals. Systematic methods for
estimating OH reaction rates and the
reaction products resulting from the
atmospheric oxidation o.f organic
compounds have been published.
Reaction With Ozone
Organic, chemicals with isolated
double bonds may have a significant
atmospheric loss rate via reaction with
ozone (Q3). While the rate constant for
reaction of alkenes with Oa is much less
than the rate constant for OH reaction
(approximately 10"17 cm3 molecule"1
sec"' compared to approximately 10~11
cm3 molecule"1 sec"1), ozone concen-
trations are so much larger, than OH
concenicalions that tjie two loss pro-
cesses are competitive for many
alkenes. Afomatrc compounds may also
react with Os, but their ozonolysis
removal mechanism is usually slow
compared to.reaction with OH.
Ozone is believed to add across the
double-bond wit.hjsjo.bsequent cleavage
to form a carbonyl compound (aldehyde
or- k©torre)-ahd"a percarbonyl biradTcal.
The biradical may rearrange'or react to
form, a variety of products including;
organic acids, carbon.dioxide and a host
of ..organic radicals. Epoxidtes have also
•been suggested 'as minor products of
ozonolysis reactions.
Reaction With Other Radicals
Thorough understanding of labora-
tory experiments involving ozonolysis
and oxidation processes is clouded by
the formation, during reaction, of
reactive radicals that may interact with
the reagents, intermediates and
products. A wide variety of such reactive
species, including those produced
during ozonolysis and hydroxylation
reactions, exists in the ambient
atmosphere. With the exception of OH
and 03, however, their concentrations
are low, and their importance as
reactants with T/H M's is small. None-
theless, for certain relatively unreactive
chemicals, reaction with species other
than OH and 03 may provide the
predominant (albeit, slow) removal
process. Reactions with the minor
species need not be considered unless
all other chemical or physical removal
processes are ineffective.
Photolytic Transformations
Estimates of the magnitude of photo-
chemical processes are difficult to make
because of uncertainties in light inten-
sity, quantum yields, etc. Photolysis can
be an important removal process only
for chemicals which absorb strongly
within the solar radiation region, other-
wise reaction with OH or Oa is likely to
predominate the removal process. This
fact limits the compounds to be con-
sidered to those possessing a strongly
absorbing chromophore, like carbonyl
compounds, conjugated alkenes, nitro
and other nitrogen-containing
compounds and halides. Compounds
which may form from photolysis of
these absorbing groups can be
suggested although predictions of
efficiencies and specific yields are im-
practical.
Physical Removal Processes
Dissolution
Toxic hazardous materials in the gas-
phase may be removed from the air
environment by dissolving them into
cloud or rain droplets to be subsequent-
ly removed by rainfall. An estimate of
the efficacy of such a removal process
can be obtained by calculating the parti-
tioning of material between the
aqueous^and gas phases.
Henry's Law relates the equilibrium
vapor phase and liquid phase
concentrations of a dilute solution of
material i via the equation
i = xi K,
(Eq. 1,
where: pi = the equilibrium vapor
pressure of the solute
above the dilute solution
x, = the mole, fraction of the
•solute, and
Ki = the Henryfs Law constant.
For very dilute solutions, the concentra-
tion of solute, G (in g cm"3), is propor-
tional to xi. Assuming ideal gas behavior
at 298° K, the term ^.may also be
converted to a gas-phase>concentration
term, Ci9, in units of g cm*3 Equation
1 becomes
cig = 9.7 X 10"7 KiCi (Eq. 2)
where the Henry's Lawconstant, Ki, has
the units of torr..
Unfortunately, the. Henry's Law
constant has not been measured for the
chemicals under assessment and must
be approximated. For those materials
which are only slightly soluble in water,
it is reasonable to assume that Henry's
Law holds for all concentrations of i
from infinite dilution to the
point. At saturation,
are in equilibrium with the
solution and also in equilibrium with I
saturated vapor phase,. Therefore, at
saturation, Equation 1 .becomes
Pis =
i and
K, = PJS (Eq. 3)
where: pis = the saturation vapor
pressure of the solute,
and
xis =.the saturation mole
fraction of the solute
The value of the Henry's" Law constant is
now expressed in terms of two more
commonly measure'd values: the
saturation vapor pressure, pis, and the
mole fraction of i at saturation, XiS.
The ratio at equilibrium of the concen-
tration in the aqueous phase to the
concentration in the gas phase is a
dimensionless number a, describing
the partitioning of the material between
the two phases. Rearranging the
equations above, one- finds that
a = _£,_= 1.03 x-106 = 1 03 ;; 10s
c,g Ki (Eq. 4..
-------
The fi'aqtion of material i removed
t is.dissolved compound in rain
product of a times the ratio of the
of the average annual rainfall
to the volume of the well-mixed tropo-
sphere. Assuming an annual rainfall of
0.75 meter and a height of 8 km for the
homogeneous atmosphere, one
calculates the half-life as
7-1,3.= 739
(Eq. 5)
Table 1 lists the relevant physical
parameters and estimates Tt.z for some
of the compounds under Type II Assess-
ment. The (.calculated half-lives are in
general quite long.
The fate of the dissolved material is
subject to question. Hydrolysis, oxida-
tion by hydrogen peroxide or other
species, adsorption on particulates in
the rain drops or in contact with the
ground water, etc., may all remove or
alter the toxic materials. Volatilization of
the chemicals during evaporation of the
rainfall run-off will re-introduce the T/H
M into the atmosphere. Such changes
will naturally affect the half-lives of the
T/H M'S
orative losses from a solution
be directly proportional to the
u ^entration of the materials in the
gas-phase above the bulk solution. The
ratio of the equilibrium vapor-phase
concentration of i, Cig, to that of water,
cwg, divided by the similar ratio for the
aqueous phase can be considered as the
"relative volatility" of the compound.
Relative volatility = Ci2/c«g= £„ jE gj
Ci/Cw Qi"
For water .at 25°C, crw — .4.34 x 10".
Values of ffw/a have also oeen listed in
Table 1 and indicate that evaporation of
even small amounts of rain water are
likely to reintroduce the T/H M into the
air.
Estimation of actual evaporation rates
from physico-chemical parameters is
very complex and depends in large part
upon the model system selected. Even
in simple systems, laboratory
experiments have shown thatthe actual
evaporation rate is diffusion limited in
the liquid phase for species with high
relative volatilities. This result implies
that, while the evaporative losses are
still fast, they are not so fast as equili-
brium considerations would indicate.
Nonetheless, the data in Table 1
suggest that for those chemicals of
concern, dissolution into and removal
by raindrops is not a significant removal
mechanism.
Adsorption on Aerosol
Particles
Toxic chemicals in the vapor phase
may be adsorbed on aerosol particulates
and removed from the atmosphere with
the aerosol. Since the average tropo-
spheric lifetime of aerosol particles is
approximately seven days, adsorption
on natural aerosols may establish a limit
for the gas-phase lifetime of various
toxic compounds. If 0 is the fraction of a
compound attached to aerosol particles,
then the atmospheric lifetime of that
compound is 7/0.8
1.4
390.0
12.030.0
ITO.OUO'ti
aw/a*
4.8
8.O
2.300.0
71.000.0
670,000:0
3 The theory developed above does not directly apply to soluble chemicals like acrylonitri/e; however, substitution1 of
Raoult's Law for Henry's Law leads to an equivalent derivation, but with slightly different assumptions.
"Equation 4 defines a as 1.03 x JO6 (xis/p,s). Since x,s and p^ are both, temperature dependent, a is aJso temperature
dependent. The values quoted in the table assume a temperature of 25° C, whereas the average tropdspheric^empera-
jyre.may be substantially different. This means that all half-lives calculated from this data should be considered
approximations.
-------
This derivation is based upon assump-
tions which are valid only when p< 0.04
p0. For most pollutants, the requirement
that the ambient pressure be less than
4% of the saturation pressure is easily
met, especially on the global scale.
K can be approximated by
AT = 5.05/JJ. \
\M/
2/3
(Eq. 9)
cm
where: D = the bulk density of the
•absorbed chemical in
g/cm3, and
M = the gram molecular
weight
Table 2 lists values of M, D, K, p0, 0 and
lifetime for several chemicals currently
undergoing Type II evaluation.
Of the five compounds listed in Table
2 only B(a)P has a vapor pressure which
is sufficiently low to suggest a reason-
able value of 0. A particulate-to-
gaseous ratio of 7.47 has been mea-
sured for B(a)P in an urban sample. The
resulting value of 0 = 0.88 is in reason-
able agreement with the value of 0.99
predicted by Equation 8 when K = 0.151
andp0= 5.5x 10"9torr(seeTable2)fora
reasonable urban B value of 5 x .10~6
cmVcm3.
These results indicate that the
theoretical calculations, while filled
with approximations and assumptions,
can provide at least "order of magni-
tude" estimates of the adsorption of
toxic chemicals on atmospheric
aerosols. The results further suggest
that adsorption will be a reasonable
vapor-phase removal mechanism only
for materials with saturation vapor
pressures of-TO"7 torr or less.
Dry Deposition
Gaseous toxic/hazardous materials
will come in contact with soil and water
at the earth's surface and may be
removed by adsorption or absorption.
The ratio of the deposition flux divided
by the airborne concentration is defined
as the deposition velocity, Vd. The values
of Vd are highly variable, depending
upon the type of surface, meteorology,
and composition of the material settling
out. Values of v
-------
3. Fates and Residence Times of Some Chemicals Under Assessment
L v 1 n1^ *, y*o18
«OH X IU /Co, A 7O
cm3 cm3 Physical
molecule'1 molecule'1 Photolysis Removal
Compound sec'1 . sec"1 Probability Probability
TYPE II ASSESSMENT
Acrylonitrile
Arsenic
Cadmium3
Eth'ylene Dichloride
Perch/oroethylene
POM (Benzo(a)pyrenef
Vinylidene chloride
TYPE 1 ASSESSMENT
Acetaldehyde"
Acrolein
Ally! chloride
Benzyl chloride
Bis(Chloromethyl) Ether
^k
^J
'Carbon Tetrachloride
Chlorobenzene
Chloroform
Chloromethyl methyl ether
Chloroprene
o-, m-, p^Cresof
Dichlorobenzene'
Dimethyl nitrosamine
Dioxane
Dioxin
Epichlorohydrin
Ethylene Dibromide
^
^F(thylene Oxide
..formaldehyde0
Hexachlorocyclopentadiene
2
0.22
0.17
4"
16.
44"
28°
3"
4*
< 0.001
0.4°
0.1
3"
46*
55.
0.3*
39°
3b
2b
0.25
2"
10.
59*
< 0.05 '
- -
— —
Possible
0.002 Possible
Possible
0.04 Possible
Probable
4* Probable
18.3 Possible
0.004" Possible
Possible
- -
< 0.00005° Possible
- -
— Possible
8° Probable
0.6
< 0.00005* Possible
Probable
- -
— Probable
Possible
— Possible
< 0.000002 Probable
8* Probable
Unlikely
Possible
Probable
Unlikely
Unlikely
Probable
Unlikely
Unlikely
Unlikely
Unlikely
Unlikely
Probable9
Unlikely
Unlikely
Unlikely
Probable9
Unlikely
Unlikely
Unlikely
—
Unlikely
—
Unlikely
Unlikely
Unlikely
Unlikely
Atmospheric
Residence
Time
Days Anticipated Products
5.6
—
~ 7
53
67
~ 8
2.9
0.03-0.7°
0.2
0.3
3.9
0.02-2.9
> 1 1000
28
120
0.004-3.9
0.2
0.2
39
<0.3
3.9
—
5.8
45
5.8
0. 1-1.2"
0.2
HzCO. HC(0)CN. HCOOH.
CN-
—
—
CIHCHO. HzCCICOCI. HzCO.
HzCCICHO
ClzCO. ClzC(OH)COCI. Cl-
B(a)P-1 ,6-quinone
HzCO, ClzCO. HCOOH
HzCO. COz
OCH-CHO. HzCO. HCOOH.
COz
HCOOH. HzCO, CICHzCHO,
chlorinated hydroxy
cardonyls. CICHzCOOH
CHO. Cl: chloromethyl-
phenols. ring cleavage
products
Decomposition products
(HCI + HzCO\ chloro-
methylformate. CIHCO
ClzCO. Cl-
Chlorophenols. ring
cleavage products
ClzCO, Cl-
Decomposition products.
chloromethyl and methyl
formate. CIHCO
HzCO. HzC = CCICHO,
OHCCHO. CICOCHO.
HzCCHCCIO. chlorohydroxy
acids, aldehydes
Hydroxynitrotoluenes,: ring
cleavage products
Chlorinated phenols, ring
cleavage products, nitro
compounds
Photolysis products.
aldehydes. NO
OHCOCHzCHzOCHO.
OHCOCHO. oxygenated
formates
—
HzCO. OHCOCHO.
CICHzO(0)OHCO
Br; BrCHzCHO. HzCO
BrHCO
OHCOCHO
CO. COz
ClzCO. diacylch/orides.
ketones. Cl-
-------
Table 3. (continued)
i
(onX 7012
cm3
molecule'*
Compound
TYPE 1 A SSESSMENT (cont 'd)
MaJeic Anhydride
Manganese3
Methyl chloroform
Methyl chloride
Methyl Iodide
Nickel* "' "• '
Nitrobenzene
2-Nitropropane
N-Nitrosodiethylamine
Nitrosoeth ylurea
Nitrosometh ylurea
Nitrosomorpholine
Phenols
Phosgene*
Polychlorinated Biphenyls
Propylene oxide
Toluene
Trich/oroeth ylene
o-, m-, p-Xy/ene'
sec"1
60°
0.012
0.14
0.004*
—
0.06"
55*
26*
13*
20*
28°
17*
0
< 1*
1.3
6.
2.2
~ 16.
it y rn18
t\Q /\ / U
cm3
molecule'* Photolysis
sec"1 Probability
160* Possible
- - .
— Possible
— Possible
— Possible
— —
< 0.00005* Possible
— Possible
Probable
— Possible
Possible
Possible
1*
— —
0.00005* Possible
—
0.0003
0.006 Possible
-0.001
Atmospheric
Physical Residence
Removal
Probability
Possible
Probable
Unlikely
Unlikely
Unlikely
Probable
Unlikely
Unlikely
—
—
—
—
Possible
Possible
Unlikely
Unlikely
Unlikely
Unlikely
Unlikely
Time
Days
0.1
' /
370
53
2900
~ 7
190
0.2
<0.4
<0.9
<0.6
<0.4
0.6
> 11
8.9
1.9
5.2
0.7
Anticipated Products
COz, CO; acids, aldehydes
and esters which should
photo/yze
H2CO, CI2CO, Cl-
ChCO, CO. CIHCO, Cl-
H2CO. I; IHCO. CO
Nitrophenols, ring cleavage
products
HzCO. CH3CHO .
Photolysis products.
aldehydes, nitramines
Photolysis products.
aldehydes, nitramines
Photolysis products.
aldehydes, nitramines
Photolysis products.
a/dehydic ethers
Dihydroxybenzenes, nitro-
phenols, ring cleavage
products
COz. Cl; HCI
Hydroxy PCB's, ring
cleavage products
CH^C(O)OCHO,
CHiC(0)CHO. HzCO.
HC(O)OCHO
Benzaldehyde, cresols, ring
cleavage products, nitro
compounds
ChCO. CIHCO. CO. Cl-
Substituted benzaldehydes.
hydroxy xylenes, ring
cleavage products, nitro
compounds
"Material is not expected to exist in vapor phase at normal temperatures. Residence time calculation assumes the
chemical is substantially adsorbed on aerosol particles and that the aerosols have a residence time of approximately 7
days.
*Rate constant calculated theoretically.
^Reaction with O('D) is possible: k = 3.6 x 70"'° cm3 molecule'' sec"1, and (O^D)] = 0.2 molecules cm'
tropospheric lifetime of 440 years. In addition, slow hydrolysis is expected.
"The shorter residence time includes a photolysis rate.
implies a
-------
rat< d no higher than "possible."
^Materials with saturation vapor
pressures of less than 10"7 torr are
likely to be adsorbed on
atmospheric aerosols, and they
were given a "probable" rating.
Species, like bis(chloromethyl)-
ether, which have been reported to
decompose rather quickly in the
environment, were also rated as
"probable." A priori prediction of
'decomposition is difficult, and it
was considered as a physical
removal mechanism only in cases
where decomposition had been
reported.
Atmospheric residence time—This
number represents the estimated
time in days, required for a quantity
of the individual chemical to be
reduced to l/e of its original value.
As it is not a concentration lifetime,
it does not include a dilution term.
For species with likely photolysis or
physical removal mechanisms, the
residence time was expressed as a
range, and a comment about the
smaller lifetime value was included
in the footnotes to the table. In cal-
culating the numbers listed in the
able, two major assumptions were
ade:
The room temperature rate
constants for OH and 03 are
valid for the ambient atmos-
phere.
a.
b. Background concentrations of
OH and O3 were assumed to be
constant, with values of 1 x 1 O6
and 1 x 1012 molecules cm"3
respectively.
Anticipated -products—The last
column lists some of the products
likely to result from the photo-
chemical oxidatton of dilute
quantities of the specific compound
in the atmosphere. The product list
is not intended to be all-inclusive
but is suggestive of the kinds of
materials likely to be produced.
Many of the products can be
expected to react further, producing
still other chemical species. Often
reactive species like radicals are
listed as products in order to
suggest that whatever reaction
scheme is available in their
immediate locale will dictate the
specific products. In addition, all the
reaction schemes proceed through
RO- and ROO- radicals which may
add nitrogen oxides to form a wide
variety of nitrogen-containing
species. These species are not
tabulated, ' but they should be
considered as possible products for
every reactive material.
Conclusions
Quantitative description of the atmos-
phere degradation or photolytic trans-
formation of the toxic/hazardous
material (T/H M) must be on a chemical-
by-chemical basis and requires a great
deal of experimental evidence that may
not yet be available. Nonetheless, esti-
mates of reasonable residence times
are feasible on the basis of the normally
predominant chemical removal
mechanisms. Based upon .current
knowledge of atmospheric oxidation
mechanisms, possible tropospheric
reaction products may be anticipated.
For the volatile chemicals currently
undergoing Type II Assessment (i.e.,
acrylonitrile, vinylidene chloride, ethyl-
ene dichloride, perchloroethylene and
benzo(a)pyrene), chemical removal
residence times, based upon reaction
with hydroxyl radicals and ozone, are
estimated to range between approxi-
mately 3 to approximately 70 days.
Estimates can also be made of the
physical removal rates for the volatile
Type II chemicals. Removal by dissolu-
tion into rain droplets yields estimated
half-lives of from 0.8 to greater than
100,000 years. Dry deposition rates
imply half-lives of approximately 25
years, and adsorption on aerosols is
demonstrated to be a reasonable
removal mechanism only for materials
with saturation vapor pressures of 10~7
torr or less. Only in the case of B(a)P,
where adsorption on aerosol particles
suggested a lifetime of about 8 days,
was a physical removal mechanism
significant.
The author of this Project Summary is Larry T. Cupitt. who is also the EPA
Project Officer (see below).
The complete report, entitled "Fate of Toxic and Hazardous Materials in the Air
Environment," (Order No. PB 80-221948; Cost: $6.50subject to change) will
be available from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
-.- U.S. GOVERNMENT PRINTING OFFICE: 1981 -757-064/0204
------- |