Atmospheric Persistence of
Eight Air Toxics
(U.S.) Environmental Protection Agency
Research Triangle Park, NC
Jan 87
PB87-145306
J
-------
PB87-H5306
EPA/600/3-87/004
January 1987
ATMOSPHERIC PERSISTENCE .OF EIGHT AIR TOXICS
by
Larry T. Cupitt
Atmospheric Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
ATMOSPHERIC SCIENCES RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
REPRODUCED BY
U.S. DEPARTMENT OF COMMERCE
NATIONAL TECHNICAL
INFORMATION SERVICE
SPRINGFIELD. VA. 22161
-------
TECHNICAL REPORT DATA
(Httu rMrf liutmeiioiu on Mr nuru ft/o
1. REPORT NO.
EPA/600/3-87/004
ItCIPIENTI ACCESSION NO.
PB87 145806/AS
4. TITLE AND SUBTITLE
ATMOSPHERIC PERSISTENCE OF EIGHT AIR TOXICS
». REPORT OATI
January 1987
B.PERPORMINO ORGANIZATION CODE
7. AUTHOR!*)
L. T. Cupltt
I. PERFORMING ORGANIZATION REPORT NO.
9, PERFORMING ORGANIZATION NAME AND ADDRESS
(Same as Box 12)
lO-'PROORAMTELlMENT NO.
AlOl/C/72/01 -. 0016(FY-87)
11. CONTRACT/GRANT NO.
12. SPONSORING AOENCV NAME AND ADDRESS
Atmospheric Sciences Research Laboratory-RTP, NC
Office of Research and Development
U. S. Environmental Protection Agency
Research Triangle Park. NC 27711
13. TYPE Of REPORT AND PERIOD COVERED
Final
14. SPONSORING AOENCV CODE
EPA/600/09
16. SUPPLEMENTARY NOTES
6. ABSTRAC I
Atmospheric lifetimes Mere estimated for eight air toxic chemicals
(i.e., methylene chloride, chloroform, carbon tetrachloride, ethylene
dichloride, trichloroethylene, perchloroethylene, 1,3-butadiene, and
ethylene oxide), which were designated in 19BS in "Intent-to-List"
notices by EPA in the Federal Register. Seven of the eight chemicals
were removed from the atmosphere primarily by reaction with OH radicals.
Because of the importance of OH radical chemistry to the estimation of
the atmospheric lifetimes of many air toxics, recommendations were made
•for the "average" conditions to use in estimating the lifetime of air
toxics over the continental U.S. due to OH reactions. These conditions
were then applied to the designated air toxics to derive estimates of the
"average" atmospheric lifetimes. Lifetimes of the seven chemicals
primarily removed by OH reaction ranged from around 4 hours to around IB
.months. In the case of eighth chemical,, carbon tetrachloride, the
primary removal mechanism is unknown. Carbon tetrachloride has a very
long atmospheric lifetime, estimated in the literature as around 50
years.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.tDENTIFIERS/OPEN ENDED TERMS C. COS AT I Field/Group
IB. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
18. SECURITY CLASS (ThaRtportl
UNCLASSIFIED
21. NO. OF PACES
67
20. SECURITY CLASS (Thitpagtl
UNCLASSIFIED
22. PRICE
£PA F«nn 7220-1 (R.». 4-77) PHKVIOUI KDITION II OBSOLETE
-------
NOTICE
The Information In this document has been funded by
the United States Environmental Protection Agency. i
It has been subject to the Agency's peer and admin- 1
1strat1ve review, and It has been approved for
publication as an EPA document. Mention of trade
names or commercial products does not constitute
endorsement or recommendation for use. j
n
-------
ABSTRACT
The concept of the "atmospheric lifetime" of an air toxic chemical
is defined, and methods are described for estimating the lifetimes of
such chemicals in the atmosphere. For many air toxics, the primary
removal mechanism in the air is its reaction with hydroxyl radicals.
Because hydroxyl radical chemistry is so important in determining the
atmospheric lifetime of many chemicals, reconmendations are made for the
"average" conditions to use in estimating the lifetime of air toxics over
the continental U.S. These methods and conditions are applied and
estimates of the "average" atmospheric lifetimes are derived for eight
volatile air toxic chemicals which EPA had identified in "Intent-to-List"
notifications during 1985. The eight chemicals for which lifetimes are
derived are methylene chloride, chloroform, carbon tetrachloride,
ethylene dichloride, trichloroethylene, perchloroethylene, 1,3-butadiene,
and ethylene oxide. Seven of the eight chemicals are removed frcm the
atmosphere primarily by reaction with OH radicals. The lifetimes of the
seven chemicals primarily removed by OH reaction ranged frcm around four
hours to around 18 months. The eighth chemical, carbon tetrachloride,
has such a long atmospheric lifetime (ca. 50 years) that the primary
removal mechanism is not identifiable.
-------
Abstract. ....... ..... . .......... ^ ....... ill
.1. Introduction .......................... 1
2. Conclusions ............ . ............. 3
3. factors Affectiny the Atmospneric Lifetime ........... b
Chemicals Involved ............... . ....... 5
Atmospheric Lifetimes .................... 6
Chemical Reaction Processes ................. 12
Reactions with Hydroxyl Radicals .......... ..... 13
Atmospheric Mixing and OH Removal ......... . . .15
Selection of Temperatures ................ 19
Selection ot Hydroxyl Radical Concentrations ...... 21
Estimation of Atmospheric Lifetimes and Removal by OH. .27
Reaction with other Species ................. 28
Other Removal Processes ......... .......... 28
4. Atmospheric Persistence ot Eight Air TDXICS ........ . .31)
Methylene Chloride ...................... 30
Chloroform ........................... 32
Carbon Tetrachloride ..... ................ 34
Ethylene Dichloride ..................... 35
Trichloroethylene ...................... 40
Perchloroethylene ........ .' ............. 41
1,3-Butadiene ...................... . .42
Ethylene Oxide. ............. .......... 44
References
-------
SECTION 1
INTRODUCTION
During 1985, the Environnental Protection Agency (EPA) published
intent to list decisions for eight organic chemicals (i.e., methylene
chloride, chloroform, carbon tetrachloride, ethylene dichloride, tri-
chloroethylene, perchloroethylene, and ethylene oxide). Since that time,
the Agency has continued to develop data on each of these potentially
hazardous air pollutants in order to assess better the exposure and risk
that these species pose in the environment. During the regulatory
decision making process, a large quantity of information is collected and
evaluated to develop a comprehensive understanding of the environmental
impact of such air toxics. Important aspects to be considered include
human health effects, the concentration levels to which people are
exposed, routes of exposure, the environmental distribution, removal
processes and the fate of the chemical, the strengths and locations of
the emission sources, and the global impacts of the air toxics.
The environmental concentrations to which humans may be exposed are
the result of a dynamic balance between emissions of the air toxic
chemical and the processes which remove or modify that chemical. Once a
quantity of an air toxic is introduced into the environment, its life-
time, distribution, and ultimate fate is determined by a variety of
factors which act to transport, remove, or redistribute the chemical.
These factors control the magnitude and nature of any potential exposure
to the air toxic chemical. A wide range of processes may be involved:
they may he chemical in nature, like reactions with ozone or hydroxyl
radicals, or physicochemical processes like photolytic decomposition,
-------
or physical mechanisms like washout, dry deposition, transport to the
stratosphere, etc.
This report will focus on describing the probable range of ambient
concentrations and the atmospheric processes which control the environ-
mental lifetime and distribution of the eight air toxic chemicals named
above. - - ..-'••'.
-------
SECTION 2
CONCLUSIONS
Relationships have been developed which describe the atmospheric
lifeti.-nes of potentially hazardous chemicals in terms of-their probable
removal mechanisms. These relationships have been applied to eight air
toxic chemicals identified by EPA in Intent to List notifications. The
eight chemicals and their estimated atmospheric lifetimes are tabulated
below.
TABLE 1. ESTIMATED ATMOSPHERIC LIFETIMES OF EIGHT AIR TOXICS
Chenical Name
Methylene chloride
Chloroform
Carbon tetrachloride
Ethylene dichloride
Tr ichloroethylene
Perchloroethylene
1,3-butadiene
Ethylene oxide
Atmospheric Lifetime
131 days
181 to 373 days
50 years
46 to 184 days
4 days
119 to 251 days
4 hours
217 to 578 days
===^==============================
For all the chemicals except carbon tetrachloride, the dominant
removal mechanism was reaction with hydroxyl (OH) radicals. Removal
rates for carbon tetrachloride were so slow that the dominant removal
mechanisn could not be determined, and the lifetime given is one reported
elsewherel based upon modeling. Average tropospheric conditions for OH
-------
reactions were defined and utilized as the basis of the predicted
atmospheric lifetimes for the other chemicals. In the case of ethylene
oxide, questions regarding the atmospheric stability of the chemical and
the lack of ambient data were addressed and resolved.
-------
SECTION 3
FACTORS AFFECTING THE ATMOSPHERIC LIFETIME
CHEMICALS INVOLVED
Right potentially hazardous chemicals are under consideration oy 'T>A
for possible regulation under the Clean Air Act. Those chemicals, for
which intent to list decisions have been published, are identified in
Table 2 below. Also listed are the name and registry number used by
Chemical Abstracts Service (CAS) to identify uniquely the specific
chemicals.2
TABLE 2: ORGANIC CHEMICAL NAMED IN INTUMT TO LIST NOTIFICATIONS
==========
CHEMICAL
Methylene Chloride
Chloroform
Carbon Tetrachloride
Ethylene Dichloride
Trichloroethylene
Perchloroethylene
Butadiene
Ethylene Oxide
====================================
CAS INDEX NAME
Dichloromethane
Trichlorcmethane
Tetrachloronethane
1,2-Dichloroethane
Trichloroetnene
Tetrachloroethene
'»
1,3-Butadiene
Oxirane
REGISTRY NO.
75-09-2
67-66-3
56-23-5
107-06-2
70-01-6
127-18-4
106-99-0
75-21-8
-------
ATMOSPHERIC LIFETIMES
Once a potentially hazardous air pollutant, like those under
consideration in this report, is released into the atmosphere, its
concentration and fate is determined by a variety of chemical or physical
processes. The emitted material immediately begins to mix in the
atmosphere and to dilute, reducing the exposure levels and the associated
risk. The total mass of the chemical in the environment is not reduced,
however, until seme process acts to remove or transform the chemical. The
rate at "which physical or chemical processes remove or transform the
target compound determines the atmospheric lifetime of the chemical.
Within any environmental compartment (e.g., a plume from an emission
source, the troposphere, the total atmosphere, etc.), the total quantity
(mass), 0, of an air toxic chemical can be accounted for. (In the
preceding sentence, the term "environmental compartment" could be
replaced by "atmospheric air parcel", "environmental reservoir", or
"microenvironment": the concept is that of an identifiable region of the
atmosphere or environment for which it is co;:-'enient to consider the fate
of the air toxic.) The time rate of change for the total mass in the
compartment, dQ/dt, is described by Equation (1).
dQ/dt = (P + I) - (R + 0) = G - D , (1)
where p is the total mass production rate oc the air toxic compound
within the compartment, I is the total mass influx rate, R is the total
mass removal rate, and O is the total mass outflow rate. G is the sum, p
plus I, and represents the total growth rate. Similarly, D represents
the total decay rate. All of the parameters on the right side of
Equation (1) are expressed in units of mass time™!.
The production rate term, P, would be important to consider in the case
of a pollutant like formaldehyde, which is both directly emitted and is
also formed in the troposphere by secondary photochemical reactions. The
-------
influx rate term represents both additional source emissions and the
inflow of material fron other environmental compartments. Stratospheric
intrusion of ozone into the troposphere is an example of the flux of a
pollutant from one compartment into another. The removal rate and
outflow terms include a variety of processes which bring about chemical
reactions, photolytic decomposition, or the transport of the pollutant
from one environmental compartment to another. Trie actual processes
which should be included in these terms may vary considerably, depending
upon the particular chemical involved and the environmental compartment
being discussed. For air toxics, a wide variety of processes may need to
be considered in describing the atmospheric fate, lifetime and distribu-
tion of these chemicals: the processes to be considered include dry
deposition, washout and rainout, changes in phase distribution between
aerosol-bound and gaseous pollutants, transport into and reaction within
the stratosphere, etc.3
Equation (1) represents a dynamic situation in which the factors
which cause the total mass loading of an air toxic to grow are counter-
acted by factors which cause the total mass to decay. The total quantity
of a chemical in the compartment, and the resultant concentrations to
which people may be exposed, can clearly be expected to change with time
in response to variations in the growth and decay rates.
In describing the transit of an air toxic through the environment,
it is useful to define several characteristic time parameters associated
with Equation (1). The characteristic decay time may be defined as
't
TD = 0/(R + O) = Q/D . (2)
Since 0 has units of mass and D has units of mass time"*, TD is expressed
in units of time. Similarly, a characteristic growth time, TQ, may be
defined as Q/G.
-------
TD is not necessarily a constant, however, since it is clearly a
function of Q, and according to Equation (1), 0 way change. In addition,
the decay terms, R, O, and D, may be arbitrary functions of Q, making TD
even more complex. Once a release scenario is defined, TD may be
determined if all of the factors comprising D are known. Since 0 is a
function.of time, TD will also vary with time as the chemical is removed
from the compartment. .•'.'.
The concept of a characteristic decay time would be most useful if
it were really a constant. This can be accomplished if one constrains 0
to be invariant. D, which is some arbitrary function of Q, also becomes
invariant, and Equation (2) becomes constant. Since 0 is constant, the
time rate of change of Q as given in Equation (1) must equal zero. This
implies that a "steady state" condition is achieved where the growth rate
is exactly equal to the decay rate, and TD equals TQ. It is only under
these "steady state" conditions that various authors have defined the
"residence time", T.4~^
= TD = TG
(3)
The requirement of a "steady state" condition is both very stringent
and, probably, unrealistic, sampling measurements at renote locations
have shown that the concentrations of many trace air toxics are slowly
increasing with time.8,9 These chemicals are clearly not at a "steady
state" condition, although, in some cases, their growth may be so slow as
to approximate "steady state".
It is not practical to depend upon a definition which is applicable
only under unrealistic conditions, especially when the requirement in
Equation (3) that the growth rate equal the decay rate is not reasonable
for many air toxics. The emission rate of many of the potentially
hazardous air pollutants can be highly variable, depending on production
volumes and market pressures, accidental releases, or even the imposition
of regulations. Fran an exposure assessment perspective, EPA is often
8
-------
primarily concerned with how long an air toxic chemical is likely to
remain in the air, once it has been released into the environment. A
reasonable question to ask in any exposure assessment is, "Does the
chemical persist in the atmosphere long enough for it to constitute an
exposure threat?" These concerns lead to a focus on the decay rate as
the atmospheric lifetime.
The characteristic decay time, TQ, will also be a constant whenever
D can be represented as a linear function of Q. If D = kp Q, then
Equation (3) simplifies to
(4)
The assumption that the decay rate can be exprecsed, or at least
approximated, as a linear function of 0 is reasonable for many of the
processes which remove air toxics from the atmosphere. Neglecting the
input terms, Equation (1) can be rewritten as a series of first order
terms which represent the processes which remove or transform the
chemical. The time rate of change in 0 is given by the equation,
-dQ/dt = kj-xn 0 + kpnot 0 + ... + kdrydep 0 + ktrans 0 , (5)
where the k's represent first-order removal rate constants for the
various processes involved. Equation (5) can be summarized as
-dQ/0 = kD dt , . . (6)
where kD is the combined removal rate constant defined above.
Integrating from time t=0 to t=t, one finds
In QQ ~ ln 0 = *D t , (7)
\
-------
where QQ is the total mass of 0 in the compartment at time = 0. Rearrang-
ing Equation (7), one finds
t - (In QO - In Q)/ ko • (8)
Letting (In Qg-ln 0) - 1» Equation (8) becomes identical to Equation (4),
t - 1 / kD s TD . (9)
Throughout the rest of this report, the characteristic decay tune,
TD, will be referred to as the "atmospheric lifetime", and the subscript,
D, will be dropped for convenience.
Clearly, the discussion of atmospheric lifetimes in this report
focuses on the removal processes and ignores any additional infl'yc of
material. In this sense, T represents the decay time for the air toxic
in the environment and differs from other discuss ions4~7 in the litera-
ture where T is used to represent the "atmospheric residence time" or the
"turn-over time" or the "average transit time" through the environment.
From Equations (8) and (9), it is clear that T is the time at which
an initial injection of a chemical has decayed until 0 = Oo/e» where e is
the natural number 2.718... . After one lifetime, about 37% of the
starting concentration is still present. After a second lifetime, only
13.5% of the chemical remains; After three lifetimes, less than 5% of
the starting material remains.
't
Similarly, one may calculate the half-life of a chemical by setting
the mass at time t equal to one-half the starting mass. The half-life,
tnen is given by
t1/2 = In 2/kD - 0.693/kD - 0.693 T . (10)
10
-------
Note also that, by combining Equations (5) and (9), one concludes
1AD • KD • krxn '+ kphot * ••• " _ l/*rxn .* lAphot * ••• »
where Trxn' etc. are defined analogously to TQ. The characteristic time
constants for the various decay processes combine in the same way that
parallel resistances combine in an electrical circuit. And like an
electrical circuit, the total resistance (atmospheric lifetime) is
dominated by the paths of least resistance, i.e., the processes with the
shortest lifetimes. This has practical significance in assessing the
atmospheric lifetime of an air toxic, because one need only consider the
processes which are most affective at removing the chemical. It is safe
to ignore processes which can be shown to be relatively slow, without
wasting valuable time and effort in measuring those processes precisely.
In any environmental compartment, the average concentration of
chemical A in the compartment is simply determined by dividing the total
mass, Of by the total volume, V, of the compartment.
[A] = Q/V, (12)
If chemical A is well mixed in the compartment (that is, the atmospheric
lifetime of A in the compartment is long compared to the mixing time in
the compartment), then any individual observation" of (A) will be very
close to the average value, indeed, Junge^ has demonstrated that, for a
wide variety of chemicals which at least approximate a "steady state"
condition, there appears to be an inverse relationship between the
tropospheric residence time and the relative standard deviation, o, of at
least one year's worth of measurements taken at enough sites to represent
a tropospheric average. The estimated lifetime is given by the empiri-
cally-derived equation
T = 0.14/a years . (13)
11
-------
Obviously, Equation (13) is not applicable for non-steady state condi-
tions, since the chemical concentration is continually changing during
the sampling period under consideration.
Localized sources or sinks'may cause the observations of the concen-
tration of A, [A], at any particular site to vary widely and make o be
large. Such inhomogeneities in concentration are to be expected for many
chemicals. For example, air toxics adsorbed on atmospheric aerosols are
likely to be removed effectively by rain out and washout.5 Since rain
events are often highly localized phenomena, they will introduce inhomo-
geneities in the concentration of the aerosol-bound species.
These inhomogeneities do not invalidate any of the equations above.
However, it may be useful to give careful consideration to the data sets,
the environmental compartment, and the averaging times involved. Equa-
tions (5) through (12) require only that the removal processes be repre-
sented as first-order functions of 0 (or [A]), not that chemical A be
well-mixed in the compartment. Practically, however, thorough mixing can
make the estimation of the removal rates easier, even though it is not
mathematically necessary. In the real atmosphere, inhomoyeneities are
likely to arise frcm non-uniform mixing effects or from significant
differences in the removal rates. To avoid difficulties in estimating
the removal rates because of inhcmogeneities, it may became useful to
consider the fate of an air toxic as it traverses through a series of
distinctive compartments.
CHEMICAL REACTION PROCESSES
A wide range of reactive species are known to be present in both the
troposphere and the stratosphere which may react with the emitted air
toxics to transform them into other compounds. The estimated concentra-
tions of these trace species in the troposphere are listed in Table 3.
12
-------
TABLE 3. ESTIMATED CONCENTRATIONS OP REACTIVE
SPECIES IN THE LOWER TROPOSPHERE*
CHEMICAL
NAME
OH
°3i
©2 (3I)
0 (3p)
0 (ID)
HOj
R02
CONCENTRATION
molecule atf-3
1 x 106
1 x 1012
5 x 108
7 x 102
5 x 104
2 x 10-1
5 x 109
5 x 109
2 x 103
PP™
4 x 10-8
4 x 10'2
2 x 10"5
3 x 10~U
2 x lO-9
8 x 10-16
2 x 10~4
2 x 10~4
8 x 10~6
* Concentrations are taken from References 10-13 and from sources cited
therein.
For many air pollutants, the predominant tropospheric reaction is
with t.he hydroxyl radical, OH, despite its relatively low concentra-
tion.3'10'13 Reactions with OH may occur via abstraction, addition, or
both. For chemicals with isolated double bonds, a significant removal
pathway may occur via reaction with ozone.10,14 por a fey, chemicals,
reaction with the NO-j radical also appears to be significant.^'^ The
other reactive species do not play a major role in*removing air pollu-
tants from the troposphere, although in sane cases the actual removal
mechanism may be in doubt (e.g., carbon tetrachloride).
REACTIONS WITH HYDROXYL RADICALS
Hydroxyl radicals are ubiquitously present throughout the tropo-
sphere, being formed from the normal photochemical processes which occur
13
-------
even in "clean" air. The radicals are so reactive that their own
atmospheric lifetime is very short, and their concentration never becomes
very large. They are consistently present in the atmosphere during
daylight hours, however, because their concentration results from the
dynamic balance between the photochemical processes which produce them
and the reactions which remove them. 16 Because of their importance to
atmospheric and combustion chemistry, the reaction rates for. OH and many
organic species have been measured. A large data base of kinetic
information is already available^ which may be applied to the problem of
estimating atmospheric lifetimes. It should be relatively simple,
therefore, to apply Equation (9) to chemical A and estimate the atmo-
spheric lifetime via
TA " VkD - I/OCQH lOH]) . (14)
Unfortunately, it is not always simple to assign values to kg^ and
the hydroxyl radical concentration, [OH]. The rate constant is a func-
tion of tenperature, and the choice of the value to use can sometimes be
difficult, since the temperature can change with the time of day, the
season, the latitude, how high in the troposphere the reactions are
taking place, etc. In addition, the [OH] depends upon a balance of
competing chemical reactions. Seme of the reactions are photochemical in
nature, while others depend upon both the temperature and the concentra-
tions of the many other chemicals involved. Profiles of lOH] in the
troposphere can, therefore, be expected to vary as a function of light
intensity, temperature, latitude, altitude, the presence of pollutants,
whether the air mass is over land or water, etc.***
The difficulties in estimating kg^ and [OH] are not really so great
as one may infer from the tone of the previous paragraph. As physical
phenomena, both the temperature and the OH concentration can be described
as continuous, albeit highly variable, functions of location across any
environmental compartment, e.g. the troposphere. The temperature dependent
function, k^, and the decay rate constant, kD => (kQH [OH]), are also
14
-------
continuous functions. The mean value theorem of calculus17 states that
the integrated value of a continuous function across a finite region can
be represented as the product of sane average value of the function times
the size of the finite region. This means that it is not necessary to
describe k^ and [OH] at all locations throughout the environmental com-
partment. Only reasonable estimates of the average values, for kQj and
[OH] are necessary to calculate the decay rate and the subsequent atmo-
! spheric lifetime. As one attempts to determine the "mean values" to use
i in Equation (14), it is useful to consider the structure of the atmosphere
i and a polVjtant's transport through it.
j Atmospheric Mixing and OH Removal
I The troposphere is that portion of the atmosphere nearest the earth.
It contains about 75-80% of the mass of the atmosphere and extends frcm
the surface to a height of about 8 to 18 km, depending upon the latitude
and the season.I8 The name, troposphere, derives frcm the Greek word
"tropos" which means "turning". The troposphere is a well-mixed region
of continuous turbulence. Above the troposphere is the stratosphere.
This portion of the atmosphere extends roughly frcm 11 km to 50 km. It
is a region with little turbulence or mixing. The name derives from the
Latin "stratum" meaning "layer". Most of the remaining 20-25% of the
atmosphere's mass is in this zone. The characteristic feature which
distinguishes these regions is the manner in which temperature varies
with height. In the troposphere, the temperature normally decreases
] linearly with increasing altitude, while in the stratosphere, the
temperature is essentially constant or increases with height. The
boundary between these two zones is called the tropopause.
The troposphere itself is often described as being divided into
layers. The planetary boundary layer extends frcm the surface to a few
thousand meters altitude. Within this zone, frictional forces between
the atmosphere and the earth's surface strongly influence mixing and
transport.19 Mixing within the boundary layer can be very efficient and
15
-------
intenset under normal meteorological daytime conditions, the boundary
layer can be considered to be well-mixed within a few hours. At the
bottom of the planetary boundary layer is a layer of essentially constant
shearing stress, called the surface layer. Frictional forces between
wind fields and the earth are greatest in this layer. It extends roughly
from 0 to 30*100 ra5'19. During inversions, winds in this layer tend to
die out because of frictional losses of energy to the earth's surface.
Because man's activities and emissions of air pollutants occur primarily
in the surface layer and boundary layer, it is important to understand
the transport and fate of air toxics in these reservoirs.
The third layer in the troposphere is called the geostrophic layer,
or the free troposphere. *9 It extends from from the top of the boundary
layer to the top of the troposphere, winds in this region are not
strongly influenced by the earth's surface, and mixing in the free
troposphere is, in general, less efficient than mixing within the
boundary layer.W Exchange between the boundary layer and the free
troposphere can also be sanewhat restricted. Siinn5 has estimated that
the time constant for vertical mixing in the troposphere is about 1 week.
Horizontal mixing (north-south mixing across latitudes) is estimated to
have a period of about 1 year.5 Exchange between the Northern and
Southern Hemispheres has an estimated time constant of 1.2 years.9
The eight air toxics under consideration in this paper are man-made
pollutants. Their release, therefore, is most likely to occur where man
inhabits the earth, namely near the surface, in temperate zones, and into
polluted air over the continental land masses. These factors suggest
that such a released pollutant will initially encounter an environment
which is likely to be warmer, and more reactive, than the average tropo-
sphere.16 As the pollutant is mixed, first within the boundary layer, then
vertically in the troposphere, and finally horizontally throughout the
troposphere, it encounters continually changing "average" temperatures
and reactant species. If it persists long enough, it becomes well-mixed
throughout the atmosphere, and the "average* conditions no longer change.
16
-------
Clearly, average tropospheric values for temperature and (CH) are actually
only applicable for species which are sufficiently long-lived to become
evenly dispersed throughout the troposphere. As Altshuller^O points out,
"For compounds with lifetimes substantially less than one year, seasonal
variability in OH concentration and temperature is important." It may be
necessary, therefore, to consider the temperature and [OHj in each of the
tropospheric layers as the chemical diffuses.
Under normal meteorological conditions, the surface layer exists as
a distinctive atmospheric reservoir only during nocturnal inversions.
Since the hydroxyl radical is formed from photochemical processes, its
nighttime concentration effectively drops to zero, and no removal by
hydroxyl radical takes place within the nocturnal surface layer. This
means that a hazardous chemical, which may be rapidly removed by OH
during the daytime, may persist all night long when trapped in this
layer. Because the mixing height is so low and the resultant mixing
volume is so small, emissions of air toxics into this layer often result
in the highest observed concentration levels. Singh et al.21*22 have
documented numerous occasions when diurnal profiles of hazardous air
pollutants reached a maximum at night and a minimum in the early after-
noon.
After sunrise /heating of the air at the earth's surface begins to
form convective currents which penetrate the nocturnal inversion and
establish a well-mixed layer averaging about 1 km in depth5 throughout
the daytime period. Because sunlight must be present for thorough mixing
to occur, conditions are also conducive for the photoc^iemical formation
of OH. Since decay of the air toxics by OH .occurs only during the
daytime, the appropriate temperature to use in determining the rate
constant is the daytime average. Air toxics which mix into this layer
during the day, remain at night and are transported and mixed by the
normal meteorological processes. Once again, chemical removal by OH at
night can be considered to be non-existent.
17
-------
Stable chemicals injected into the boundary layer will subsequently
mix into the free troposphere. Nixing is not uniform in all directions,
however, vertical mixing has a time constant of about 1 week.5 Air
toxics with lifetimes longer than a few weeks can therefore be considered
to be thoroughly mixed throughout the troposphere in the vertical
direction. Horizontal mixing, specifically north-south mixing across
latitudes, has a time constant of about one year.5 A similar time
constant applies to the exchange of air in the troposphere between
northern and southern hemispheres.? Air toxics with lifetimes longer
than several years become distributed globally.
While the stratosphere is quite stable, with inefficient vertical
mixing, there is sane exchange of air between the troposphere and the
lower stratosphere. This exchange can transport extremely stable air
toxic chemicals into the stratosphere where a wide range of chemical
processes, other than reaction with OH, may occur. Junye23 estimates
that the time constant for exchange fron the troposphere to the strato-
sphere is 4 years, while exchange from the stratosphere to the tropo-
sphere has a time constant of 1 year. The difference in time constants
arises from the fact that the troposphere contains roughly four times as
much mass as does the stratosphere. For extremely long-lived air toxics,
stratospheric removal processes may need to be characterized.
The previous discussions suggest that the.atmosphere can be split
into several distinctive reservoirs which should be considered when
estimating chemical removal by OH. The first reservoir is the boundary
layer. The [OH] is that for a polluted air mass at or near the surface,
and the temperature of choice is a daytime average. The second reservoir
is the vertically well-mixed troposphere. The choices for [OH] and
temperature are dependent upon the latitude selected. The choice oi"
temperature is once again influenced by the fact that OH removal occurs
only during the daytime, but since the full troposphere is being consid-
ered the impact is much less. The third reservoir is the hemispheric or
global troposphere. Here concentrations and temperatures should be
18
-------
averaged over both altitudes and latitudes. Finally, transfer into the
stratosphere may need to be included for very stable pollutants.
Selection of Temperatures
The choice of temperature to use in each of the domains to be
considered depends upon icany factors. Obviously, temperatures vary with
time of day, fran day to day, fran location to location, from season to
season. Any choice of any single temperature to represent an average in
the various reservoirs will require some compromises, for specific
chemicals or release scenarios, the fate may need to be determined on a
case-by-case basis. This paper will attempt to establish some starting
values which may reasonably represent annual averages across the United
States. Because Alaska and Hawaii may represent significantly different
climates, the following discussion will focus only on the contiguous
forty-eight states.
The average temperature is strongly influenced by the latitude which
is selected to represent the nation. A variety of approaches can be used
to estimate the appropriate latitude. The middle of the latitudes for
the northernmost and southernmost locations24 in the contiguous forty-
eight states is around 37° N. The center of the total U. S. population
in the 1980 census was found to be at 38° 8' N.25 Since it is reasonable
to expect that the location of air toxics emissions may roughly correlate
with the location of the population, a choice between the 37th and 38th
parallels seems reasonable.
The U.S. Standard Atmosphere Supplements, 19662& uses long-term
averages of temperature and pressure observations to describe profiles
for the atmosphere at 30° N and 45° N during the months of January and
July. January and July normally represent the extremes in temperature at
any location. Averaging those two monthly periods and interpolating
across latitudes, one can derive a reasonable, observation-based approx-
imation of the annual average tropospheric conditions for the contiguous
19
-------
United States. The definitions of standard atmospheres suggest an
average sea-level temperature of 289 K at the mid-latitude of 37.5° N,
and an average lapse rate of about -4.6 K per kilometer of height. If
one assumes that the continental land mass has approximately an altitude
of 500 meters above sea level, then an annual mean temperature of about
286.7 K is estimated. This value is very close to the mean temperature
of 286.9 K estimated for the same latitude from a regression fit of
thirty-year averages of the annual mean temperature25 for 58 continental
U. S. cities located between 30° N and 45° N latitude. The estimate is
also very close to the average temperature derived for the U. S. of 287.0
K, which is obtained by weighting the ambient temperature by the nunber
of vehicular miles traveled at that temperature.27 if vehicular use is
indicative of human activity and, therefore, of air toxic emissions, then
it is reasonable that the estimates should agree.
r.
For use with estimates of OH removal, one is not interested in the
mean temperature at the surface, but rather in the average daytime
temperature across the mixing layer. A regression fit of the thirty-year
averages of the normal daily high temperatures for the same 58 cities
mentioned above suggests that the annual average high temperature is 5.9
K higher than the mean at this latitude. If the average daytime tempera-
ture is taken to be mid-way between the annual mean and the annual high
temp^r-ture, then the estimated temperature must be increased by about
2.9° to account for the effect of daylight. But the mixing layer is
about 1 km thick and the descriptions of the standard atmospheres suggest
an average lapse rate of about -4.6° per kilometer, so the estimated
temperature must also be decreased by about 2.3 K. The net result is
that the estimated mixing layer temperature for the contiguous forty-
eight States is approximately 287.5 K. This is somewhat lower than the
values of 291-293 K suggested by Heicklen.28,29
I;
The second domain of interest is the vertically mixed troposphere at
the same latitude. TO estimate this value, the temperature profiles fron
the four standard atmospheres (two seasons at two latitudes) were
20
-------
weighted at each altitude by the density of the air and averaged. The
temperature values were weighted by density to account for the fact that,
as altitude increases both the temperature and the mass of gas at that
temperature decrease. The four calculated temperatures were averaged to
yield a value at 37.5° N of 263 K.
The third reservoir of interest is the global troposphere, Altshul-
ler20 states that a temperature of 265 K is very close to the average
annual tropospheric temperature weighted for distribution of species with
altitude, and that it corresponds within a few degrees to the annual
600-mb, 4-km altitude value derived by two methods, involving averaging
600-mb temperatures over latitudes or over the monthly average tempera-
ture. The density-weighted temperature profiles for the standard atmo-
spheres26 at 45° N, mid-way between the equator and the pole, yield an
average tropospheric temperature of 260 K. The density-weighted, standard
atmosphere based on traditional definitions also yields an average tropo-
spheric tenperature of 260 K.30
Activation energies for many OH reactions seem to fall within the
range of -1000 cal mole"1 to +4000 cal mole'1 (cf. Reference 13). With
this range of activation energies, the rate constant is not likely to
vary more than a tactor of 3 across the range from room temperature to
260 K. Across an even smaller tanperature range, the difference in rate
constant is proportionately small, e.g., from 287.5 K to 293 K, the rate
constant will vary not r.ore than 15%.
Selection of Hydroxyl Radical Concentrations.
Over the years, a wide variety of attempts have been made to char-
acterize the hydroxyl radical concentration in the troposphere. Un-
fortunately, the modeling efforts have often disagreed and direct experi-
mental measurement is still difficult. The various modeling approaches
have predicted average [OH] in the troposphere across a widely divergent
range.16'31 The instrumental measurements are not yet sufficiently
21
-------
sensitive to measure the ambient OH concentration routinely and
accurately, although values in excess of several million radicals per
cubic centijneter have been reported in seme polluted situations.32
The concentration of OH in the boundary layer has been estimated
experimentally by using both direct and indirect measurements. The
reported values by direct measurement of [OH] vary widely, ranging from
0.3 to 150 x 10§ molecules cnT3. Problems with sensitivity, calibration,
and interferences pose serious questions about the accuracy-of the
experimental observations, especially the very large values determined
using laser induced fluorescence (LIF). A set of measurements does seem
to be emerging, however, which is consistent within itself and which also
agrees with the most comprehensive of current tropospheric models.
Perner et al. used a long path optical absorbtion technique to measure OH
in the boundary layer at about 51° N in Germany. Their early measure-
ments33 usually remained below their detection limit of 4 x 106, with
occasional excursions up to 7 x 10*> molecules cm"3. More recent measure-
ments with an improved instrument32 have yielded values of 0.8 to 4 x
106, with a simple average of 1.9 x 106 molecules cm"3. The measurements
by Perner et al. have been made between April and October in different
years. Given the scatter in the data, there is no discernable effect of
season upon the OH concentration. Campbell et al. monitor the oxidation
of isotopically labeled 00 to deduce the concentration of OH.3^ They
report four values (2.0, 3.3, 3.4, 3.4 x 106) for [OH] in non-pristine
air over the North American continent. Campbell's data from New Zealand
show lower [OH], which is to be expected in a comparatively unpolluted
site dominated by a marine environment.16 D. Davis'et al.35»36 have
reported [OH] using an LIF technique. Most of their measurements have
been made from an aircraft at high altitudes, but one report3** from an
airplane at the top of the boundary layer gave [OH] of 8 to 11 x 10^.
Wang and L. I. Davis, who first reported using LIF techniques to measure
atmospheric concentrations of OH,3? have recently made considerable
advances in overcoming the problems inherent in the method. Their latest
measurements at ground level in Dearborn, MI show an average [OH] cf
i •
22
-------
1.9 x 106 for two days, almost a year apart.38 Hjorth et al.39 estimated
[OH] by collecting bag samples of ambient air at a semi-rural site in
Italy, then adding isotopically labeled GO, and irradiating the bags with
natural sunlight. They monitored the oxidation rate using Fourier trans-
form infrared spectroscopy and, based upon thirteen experiments fron May
to July, deduced an average OH concentration of 2 x 106 molecules cm"3.
There have been several estimates of [OH] using measurements of
other species, frcm which the [OH] can be deduced. Calvert40 used the
rate of disappearance of a variety of alkanes and alkenes relative to
acetylene to estimate the OH concentration in the morning hours of
November 5, 1973, in Los Angeles as 2.5 ± 2.0 x 106 molecules cm'3.
Singh et al.4l used the same technique on more recent data frcm the Los
Angeles area to estimate a mean OH daytime concentration of 2.9 ± 1.9 x
106 molecules cnT3 during the months of June through September. In a
subsequent study, Singh and coworkers22'42 applied the technique to a
series of aromatic species to estimate a lower limit for the OH concen-
tration in the Los Angeles area in (February. The calculated values of
I OH] represented an average for the time period frcm 7:30 a.m. to 1:30
p.m. and resulted in a postulated OH concentration of 2.6 ± 0.6 x 106
molecules on-3. Similarly, Anderson43 used the rate of disappearance of
a light alkene relative to acetylene to deduce the OH concentrations
present during an experiment in November and December in Denver. The
estimated OH concentrations ranged frcm 0.6 to 6 x 10*> molecules on~3,
with an average of 1.8 x
Several modeling efforts have also attempted to predict OH concen-
trations throughout the troposphere. TVo of the more comprehensive
models are those by Logan et al.16 and by Crutzen and Fishman.44 The
Logan et al. model predicts that diurnal ly-averaged (24 hr. ) concentra-
tions of OH at the continental land surface, in the Northern Hemisphere
across the latitudes of the continental U.S., should be about 0.8-0.9 x
106 molecules cnT3 annually. (See Figures 18, 27a, and 27b of the paper,
Reference 16.) Suntnertime diurnally-ave raged concentrations at the
23
-------
surface ace probably near 1.3 x 106. The Crutzen and Fishraan model
suggests an annual average surface concentration of about 0.5-0.6 x 106
across the same latitudes. Crutzen and Fishman predict summertime
average concentrations of about 1.4 x 106 molecules cffi-3.
»
Assuming equal periods of daylight and darkness, and diurnal concen-
tration profiles similar to those predicted by Logan, one finds that the
average concentration for a 24-hour day should be a factor of 2 to 4 less
than the daytime maximum values. The experimentally measured (or
derived) concentrations suggest that the average daytime [OH] in the
boundary layer should be about 2 to 3 x 10$ molecules on"3. The daytime
maximum in concentration can be even larger. A reasonable 24-hr, average
for (OH] in the boundary layer, based upon experimental data, therefore,
is 1 x 106 molecules on'3. This estimate is consistent with current
models in that it exceeds, only slightly, the model predictions of Logan
et al. and of Crutzen and fishman for an annual average concentration at
the earth's surface over the continental U.S.
The average OH concentration in the vertically-mixed troposphere at
about 38° N is less clear. While the models of Logan et al. and Crutzen
and Fishman predict roughly the same surface level concentrations of OH,
the vertical profiles differ remarkably. In general, the Logan model
predicts an [OH] maximum at an altitude of 2 to 6 kn, while the Crutzen
and Fishman model indicates a maximum concentration at the surface,
dropping off to an average concentration of around 0.1 x 10*> at altitudes
of 6 to 10 km. The choice of which vertical profile to use could be made
easier by inspection of the measured values of [OH] at altitude. The
experimental data on [OH] at various altitudes,are highly variable,
probably because of instrumental difficulties. A variety of tropospheric
measurements, t&Xen between 30° N and 45° N, are listed in Table 4.
24
-------
mmm-mmmmm
Latitude
0 N
.37
37
37
37
~ 37
32
32
- 37
~ 37
34
37
37
37
37
34
34
The
suggest!
10 km.
more sea
106 are
model pr
From the
may esti
for the
TABLE 4. EXPERIMENTALLY MEASURED
IN THE FREE TROPOSPHERE
Altitude [OH]
tan 106 raolec on"3
6
6
6
6
6.9
7
7
10
10
10
10
10
10
10
10.7
11.9
= S=SS3S=SSS33SSC333S=:
experimental data
ng a value of 2 'to
The results obtain*
ttered, but they dc
present. Clearly,
edict ions of Logan
vertical profiles
mate a diurnal ly a\
continental U. S.
1.9
2.1
1.5
3.5
3.3
< 2
3.5 * 2.3
Ot5
1*3
20 t 6
< -5
< .5
0.8
1.9
10 ± 4
2 ± .5
OH CONCENTRATIONS
Analytical
Method
Spin Trap
Spin Trap
Spin Trap
Spin Trap
LIF
LIF
LIF
LIF
LIF
LIF
Spin Trap
Spin Trap
Spin Trap
Spin Trap
LIF
LIF
reference
Watanabe4*
Watanabe4*
Watanabe4*
Watanabe4*
i
Davis36 ;
Davis3*
Davis35
Wang46
Wang46
Wang46
Watanabe4*
Watanabe4*
Watanabe45
Watanabe4*
Wang46
Wang46
~|
of Watanabe et al.45 are self-consistent,
3 x 106 at 6 km and a somewhat lower value at
sd by laser-induced fluorescence35,36,46 are
> suggest that concentrations in excess of 1 x
the experimental data agree better with the
et al. than with those of Crutzen and Pishman.
of CH concentration in the Logan paper, one
reraged [OH] of around 1 x 106 molecules cm**3
25
-------
The annual average [OHJ for the global troposphere and/or for the
Northern and Southern Hemispheres has been estimated by several authors.
Crutzen and Pishman estimate an average [OHJ in the Northern Hemisphere
(NH) of 0.25 x 106. Lagan et al. estimate a value of 1 x 106 for the
Northern Hemisphere, other researchers have attempted to use atmospheric
distributions of man-made pollutants to estimate the global concentration
of OH through a tropospheric budget model. Because production quantities
of synthetic organic pollutants are often reasonably well known, tneir
emissions can be balanced against the observed tropospheric distributions
to estimate their tropospheric lifetime. Since the primary removal
mechanism for many of these chemicals is from reaction with OH, the
lifetime calculated from source strengths may be used to deduce an
average tropospheric OH concentration.
Several authors have deduced average global and hemispheric OH
concentrations from the distribution of methyl chloroform (1,1,1-tri-
chloroethane). Singh*7*48 was the first to apply a global budget
estijnate to methyl chloroform. He calculated Northern Hemispheric and
global values of OH concentration as 0.21 - 0.31 x 106 and 0.33 - 0.51 x
106, respectively. Neely and Plonka49, from Dow Chemical Co., used a
smaller emissions budget in a similar tropospheric model to estimate NH
and global OH concentrations of 0.48 and 1.1 x 106. Unfortunately,
these investigators used a rate constant which was subsequently found to
be in error.50'51 Use of the correct rate constant requires that all the
estimates of (OH] be increased by a factor of 1.3 to 2.16 singh has
subsequently reapplied the model with the correct rate constant^ and has
deduced a global average [OH] of 0.6 x 106 molecules cm~3 Using four
other man-made chemicals, Singh estimates the global average [OH] of
0.25, 0.37, 0.51 and 0.52 x 106. Logan et al.16 also applied their
calculated OH profiles to the atmospheric accumulation of methyl chloro-
form. They concluded that the data in both hemispheres fit their pro-
files to within a factor of two. In the NH, the best fit appeared to be
at the lower limit value, implying an average [OH] of 0.5 x 106. Also,
Volz et al.51 recently deduced an average tropospheric OH concentration
26
-------
of 0.65 x 10* molecule cra-3 from consideration of the oxidation of 1400.
A reasonable estimate of the hemispheric or global [OH], therefore, seems
to be around 0.5 x 106 molecules on'3.
Estimation'of 'Atmospheric'Lifetimes and Removal by OH
In order to estimate the removal rate of air toxics due to reaction
with OH radicals, it is convenient to divide the troposphere into three
compartmentst the boundary layer* the vertically-mixed troposphere at a
specific latitude, and the horizontally-mixed global or hemispheric
troposphere. The average temperature and (OH] in the appropriate
compartments can then be used to estimate the lifetime of an air toxic.
Average values in each compartment for continental air at around 37.5° N
latitude are shown in Table 5. The actual choice of the compartment to
u«e depends upon the estimated lifetime of the chemical and the mixing
times between reservoirs. For species with lifetimes of less than 2 or 3
days, the estimates for the boundary layer should be used. For chemi-
cals with lifetimes of about 3 weeks to 5 months, the values for the
vertically-mixed troposphere should be used. For very stable chemicals
with lifetimes in excess of 3 years, the global values should be used.
For species with lifetimes intermediate to the times described above,
lifetimes should be estimated for the two bracketing regimes, and the
lifetime expressed as a range.
TABLE 5. SELECTED AVERAGE VALUES OF TEMPERATURE AND (OH)
IN THREE REGIMES OF THE TROPOSPHERE AT 37.5° N LATITUDE
Temperature [OH]
Regime
Boundary Layer
Vertically-Mixed Troposphere
Hemispheric/Global Troposphere
K
288
263
260
106 molec cra~3
1.0
1.0
0.5
1333==========
Applicable
Lifetimes
< 3 days
3 wk to 5 mo
> 3 yrs
27
-------
REACTION WITH OTHER SPECIES
Air toxic pollutants may also react with other chemical species in
the atmosphere, primarily 03 and N03. Ozone reactions are an important
atmospheric removal pathway for alkenes. The kinetics and mechanisms of
ozone reactions under atmospheric conditions have recently been reviewed
by Atkinson and Carter.14 NC>3 radicals may react with a variety of
compounds.13'15 Data on individual chemicals must be consulted to deter-
mine the relative importance of these removal processes. In some cases,
reaction with other pollutants, like nitric acid, or stratospheric
reactants may need to be considered.
OTHER REMOVAL PROCESSES
Air toxics may also be removed from the atmosphere by a variety of
other processes. Previous papers3*23 have discussed these processes in
detail, so they will only be described briefly here.
Solar radiation in the troposphere may be absorbed by specific air
toxics, causing them to degrade. This process of photolytic degradation
is likely to be important only for those chemicals which absorb strongly
within the solar radiation region. This limits the applicable compounds
to those possessing a strongly absorbing chronophore, like carbonyl
compounds, conjugated alkenes, alkyl nitrites, nitro compounds, etc.
Air toxics may also interact with objects at the earth's surface and
be adsorbed and removed through a process called dry deposition. The
deposition velocity, which is expressed as the speed (length time'1) at
which chemicals deposit, has been determined for only a few gases.
Except for a few reactive gases, the available data53 suggest that this
removal route is not very effective for gas-phase chemicals. Two of the
chemicals under consideration, carbon tetrachloride and tetrachloro-
ethene, have had dry deposition velocities measured or estimated.3'54 in
both cases, the deposition rate was so slow that lifetimes in excess of
28
-------
25 years are estimated from this removal mechanism. Dry deposition is an
important process for other pollutants, like polychlorinated biphenyls,
and can constitute a major input source for such pollutants into lakes
and streams.55
Toxic pollutants which are emitted as gases nay adsorb on small
airborne particles, or aerosols, in the atmosphere. Adsorption does not
immediately transform an air toxic or remove it from the air medium. The
interaction of the air toxic with the surface of the particle may,
however, change the nature of the chemical and photolytic removal proces-
ses which affect the pollutant. Consideration of the phase distribution,
i.e., the quantity of the air toxic in the vapor phase compared to the
aerosol-bound phase, can be important in risk assessment, however, since
the route of exposure and possible uptake of the pollutant by plants and
animals may be considerably different for the two phases. In addition,
the average tropospheric lifetime of an ambient aerosol particle is
approximately seven days.23 Any air toxic adsorbed on aerosol particles
will then have only a short lifetime in the atmosphere before it enters
the soil or water media.
Finally, gases which are water soluble may dissolve into water
droplets present in the atmosphere in the fora of rain, snow, fog, or
aqueous aerosols. Deposition of these droplets may remove the pollutant
£ron the air environment. While the processes of rainout and washout are
effective for removing atmospheric particles, they are not generally
considered to be effective for gaseous pollutants23»56 because of subse-
quent evaporation and revolatilization of the pollutant.3 This process
could be important, however, if transformations occurred rapidly enough
in the aqueous phase for the droplets to constitute a sink for the air
toxic.
29
-------
SECTION 4
ATMOSPHERIC PERSISTENCE OP EIGHT AIR TOXICS
METHYLENE CHLORIDE
Methylone chloride (dichlorcrae thane) is a high volute, commonly used
solvent which is a reutagen and suspect carcinogen. 22 it is used as a do-
greaser, cleaner, paint solvent, aerosol propellant, laboratory solvent,
etc. Most of its uses result in large and rapid losses bo the atmo-
sphere.20 The chemical is of anthropogenic origin, with no known natural
sources.22 Its background concentration at 40° N is about 50 parts per
trillion (ppt).22 The global average background concentration is 29 ppt,
with the concentration in the Southern Hemisphere being only about
one-half that in the Northern Hemisphere.9 Its concentration in urban
areas of the U. S. is highly variable,57 probably due in part to the many
sources resulting from its frequent use. Average urban concentrations
are often ten to one hundred times as large as 'the gecchemical background
concentration.57
Methylene chloride reacts with OH radicals in the atmosphere at a
moderate rate. Atkinson13 has recently reviewed the literature on the
reaction of OH and methylene chloride. The rates have been measured by
at least five different investigators and are in reasonably good agree-
ment. Atkinson recommends the modified Arrhenius expression
k(CH2Cl2) - 8.54 x 10~18 T2 exp(-500/T) cm3 molecule"1 s"1 (15)
30
-------
for estimating the hydroxyl radical rate constant. Reccnnended values of
the rate constant at -.trious relevant temperatures are tabulated below,
together with estimates of the atmospheric lifetime due to OH reactions.
TABLE 6. METHYLENE CHLORIDE REACTION RATE CONSTANTS AND LIFETIMES
Temperature
K
288
263
260
OH Reaction Rate Constant
10-14 on3 raolec-1 a'1
12.5
8.8
8.4
Assumed [OH]
106 nolec cnr3
1.0
1.0
0.5
Lifetime
days
93
131
274
Reactions are not expected to occur with ozone or other cannon
atmospheric pollutants. Dry deposition and rainout are also expected to
be very slow, since other removal mechanisms are not likely to be
effective, they can be ignored in estimating the total removal rate. The
atmospheric lifetime of methylene chloride is, therefore, simply that
predicted from hydroxyl radical removal. Table 6 lists three choices of
atmospheric lifetime based upon hydroxyl radical, however. The three
choices reflect the three domains described previously. (See Table 5.)
The first line, with the temperature of 288 K, represents an average
boundary layer condition; The second line lists conditions in the
representative, vertically well-mixed troposphere. The last estimate is
for the hemispheric or global troposphere. Since'the first estimate of
93 days exceeds the applicable range for the boundary layer domain, that
value must be neglected. The estimate of 131 days, however, is within
the applicable range for the vertically well-mixed troposphere (i.e., 3
weeks to 5 months) so that value is the one selected as the estimate of
the atmospheric lifetime. This means that, on average, a quantity of
31
-------
roethylene chloride emitted into the atmosphere above the continental U.S.
will persist in the troposphere for several months. Even after four
months, 37% of the emitted material will still remain in the atmosphere.
The estimated half life is about 69% of the lifetime quoted above, or
about 91 days.
Singh et al.9 recently used a two-box model to estimate the global
lifetime of methylene chloride. The model uses the background concentra-
tion, the distribution of the chemical between the Northern and southern
Hemispheres, and estimates of production and release to calculate a
global lifetime. His data suggested a lifetime of 0.9 ± 0.3 years, a
value two and one-half times as large as that calculated above. Consid-
ering the uncertainties in the emission strength, contributions from
Europe and elsewhere, and the scatter in the experimental data, the
agreement is not bad. It is reassuring that the global modol lifetime
exceeds the estimate from hydroxyl removal alone, since that fact
suggests that no major loss mechanism has been overlooked. It seems,
therefore, that the estimate of 131 days is reasonable.
CHLOROFORM
Chloroform (trichlorcroethane) is a mutagen and suspect carcinogen
which is ubiquitously .present in the atmosphere. It has received much
attention recently because of its high concentrations in chlorinated
drinking water.22 Unit risk numbers, which estimate the probability of
death due to cancer arising from a lifetime of exposure to the chemical,
h?ve been developed for chloroform.20
•me chemical is manufactured for use as a solvent, a cleaning
agent,58 and as an intermediate in the manufacture of chlorofluorocarbon-
22.22 Geochemical background concentrations are around 16 ppt, with
slightly higher values in the Northern Hemisphere.9 Data on chloroform
reported between 1970 and 1980 were recently critically evaluated.57
32
-------
While the results were highly variable and much of the data taken with
Tenax had to be rejected because it consistently reported concentrations
below the known'(jeochemical background levels, an interesting concentra-
tion pattern did emerge. Rural and remote areas of the U.S. had concen-
trations roar 40 ppt; urban and suburban areas had concentrations around
270 ppt; and source'dominated areas had still larger concentrations.
More recent measurements22 have reported average urban concentrations of
60 to 250 ppt, with maximum values being many times higher. World wide
emissions of chloroform do not seem to. account for the levels observed:
the sources of chloroform, both in urban and remote areas, appear to be
ill defined.22
The reaction rate constant for chloroform with OH has been reported
by at least three investigators, and the data are in excellent agree-
ment. 13 Recommended values of the OH rate constant are given in Table 7.
TABLE 7. CHLOROFORM REACTION RATE CONSTANTS AND LIFETIMES
Temperature OH Reaction Rate Constant Assumed [OH] Lifetime
K 10~14 cm3 molec-1 s'1 106 molec cm"3 days
288
263
260
9.08
6.41
6.13
1.0
1.0
0.5
127
181
378
For chloroform, reaction with other tropospheric pollutants and
removal by physicochemical processes is not expected to be very large.
The atmospheric lifetime is equated with the lifetime due to hydroxyl
radical reaction. None of the estimated lifetimes listed in Table 7 fall
within the applicable periods for the three dcraains. It is best, there-
fore, to express the lifetime as a range. Because the lifetime appears
to fall most appropriately between the applicable periods for the
vertically-mixed troposphere and the hemispheric or global troposphere,
33
-------
it is appropriate to assign chloroform an atmospheric lifetime of between
181 and 378 days (or 0.5 to 1.0 years).
CARBON TETRACHLORIDB
Carbon tetrachloride (tetrachloronethane) is a suspect carcinogen^
which is nearly uniformly distributed around the globe.1 This uniform
distribution means that the chemical is well mixed in the troposphere and
suggests that it has a very long lifetime compared to the atmospheric
mixing processes.
Carbon tetrachloride has seen wide usage as a solvent and as an
intermediate in the production of other compounds.58 Large natural
sources of carbon tetrachloride have been postulated, but none have been
found.22 Indeed, the most comprehensive analyses of the carbon tetra-
chloride budget have concluded that all of the ambient concentrations can
easily be accounted for from man-made production and emissions.22
The measurement of carbon tetrachloride in the ambient atmosphere
has proven difficult over the years, with investigators reporting
significantly different values.57 since the detailed efforts of the
Atmospheric Lifetime Experiment,1»59 however, a consistent set of data on
carbon tetrachloride has begun to emerge. Many of the data currently in
the literature involved the use of Tenax, which seems to have severe
problems in collecting this- chemical. Since there are no known large
sinks for this chemical, any data set which consistently reports concen-
trations significantly below the geochemical background concentrations
must be considered suspect.57 The Atmospheric Lifetime Experiment1
established the geochemical background u>ncentrations in 1978-1981 to be
around 118 ppt. Concentrations of the chemical were found to be increas-
ing at about 2 ppt per year. Urban measurements, reported by investiga-
tors whose data are consistent with the geochemical background, show that
urban concentrations are a factor of two to three higher than background.
34
-------
Maximum excursions in urban areas can be in excess of 1000 ppt.57 A
lifetime cancer risk number for carbon tetrachloride has been pub-
lished. 20
The reaction of carbon tetrachloride with UH radical is so slow that
the rate constant has not been measured. Upper limits for the rate
constant have been determined, and they suggest that the atmospheric
lifetime due to hydroxyl radical reaction is longer than 50 years.3*13
In the troposphere, photolyMc decomposition and other chemical removal
procesjes seem to be equally slow. An inferred dry deposition velocity
has been used to suggest a lifetime in excess of 25 years.3 With such a
long tropospheric lifetime, stratospheric removal processes must be
considered. In addition to possible chemical reactions and photolysis in
the stratosphere, a number of removal processes have been postulated
including hydrolysis in the ocean, ion-molecule reactions, and photolysis
on sand. None of these processes has a very large effect, however. Most
of them result in lifetime estimates ef around 50 years.1 There are no
known removal processes which can be used to estimate a lifetime of
carbon tetrachloride reliably.
Two modeling approaches were applied to the carbon tetrachloride
data set of the Atmospheric Lifetime Experiment.1 A trend analysis
technique and an inventory analysis method yielded most probable life-
times of 50 to 57 years, respectively. Vtiile these modeling approaches
did produce lifetime estimates with substantial uncertainties, they are
consistent with the known chemical and physical processes involving
carbon tetrachloride. A lifetime estimate of around 50 years is the best
that can currently be done.
ETHYLENt: DICHLORIDE
Ethylene dichloride (1,2-dichloroethane) is another high-volume man-
made pollutant which has become ubiquitous in the atmosphere. The
material is a mutagen and suspected carcinogen, and a lifetime unit risk
-------
number has been developed. Ethylene dichloride (EEC) is primarily used
as an intermediate in the production of other chemicals, particularly
vinyl chloride monomer, other minor uses include use as a gasoline
additive and as a solvent.20'22'58
Singh et al.9 report a global background concentration of EDC of 25
ppt in December 1981. The Northern Hemisphere average was 37 ppt, very
close to the value of 40 ppt calculated by Altshuller from production
values. A critical assessment and compilation of data on EDC for values
reported from 1970-1980 gave a median urban/suburban concentration of 120
ppt.57 Those data were strongly influenced, however, by a large data set
from a single investigator in which all of the values, except two, were
below the detection limit. Actual urban concentrations may be even
higher, and are likely strongly influenced by local users.
The reaction rate of EDC with OH radicals has not been so thoroughly
investigated as the reactions of many other chlorinated alkanes. In
fact, Atkinsonl3 lists only one rate constant, by Howard and Lvenson^Q,
and does not give an activation energy. There is, however, one addi-
tional literature value for the rate constant which Atkinson does not
mention. Snelson et al.61 have also reported a value for the rate
constant which is a factor of 3 below the value of Howard and tvenson.
Both research groups measured the rate constant at room temperature
(about 298 K), and no measurement or the activation energy of the
reaction was made. The measurement by Howard and Evenson was made in a
relatively simple, or "clean", kinetic system, while Snelson et al. used
a much more 'complex photochemical arrangement. Snelson et al. argue that
their method, with oxygen and nitrogen also present(in the reactor, was
more representative of actual tropospheric conditions than bimolecular
reaction studies. That is not true in this case. The reaction proceeds
via abstraction, and there will not be a significant "pressure effect",
as there is for an addition reaction. (See, for example, References 13
and 62.) The addition of other reactive gases, in this instance, only
makes the reaction scheme more complex and makes extraction of kinetic
36
-------
rate data more difficult. Indeed, Snelson et al. had to assure a
reaction mechanism and deduce an effective chain length without compre-
hensive product analysis. Such an assumption is always precarious, but it
is especially precarious when it involves photolysis of chlorinated
species. In any case, a pressure effect should increase the reaction
rate (by making any addition mechanism more effective), but the rate
constant calculated by Snelson et al. is actually lower than that
measured at low pressure by Howard and Bvenson. Snelson et al. report
that they had difficulty separating the EEC effects from the background
reactivity, and they actually estimated an upper limit to the rate
constant which was almost a factor of 4.5 larger than their "most
probable" value. Snelson et al. demonstrate little confidence in their
own value of 6.5 X 10~14 on3 molecule""1 sec'1 since they selected a value
of 19.1 X 10'14 cm3 molecule'1 sec'1 as the 298 K rate constant for use
in calculating the tropospheric residence time of EDC. They arrived at
that value by averaging their value of 6.5 X 10~14 cm3 molecule"1 sec'1
with that of Howard and Evenson, weighting each by its probable accuracy.
Obviously, the Howard and Evenson value of 22 X 10~14 cm3 molecule'1
sec'1 is relatively unchanged.
The v*3rk by Howard and Evenson, on the other hand, is more straight-
forward, and their measured value of 22.0 ± 5.0 X 10~14 cm3 molecule'1
sec"1 is much more consistent with the reaction rates of similar chlorin-
ated ethanes than is the value of Snelson et al.. Table 8 below shows
the reported roan temperature rate constants for ethane and a number of
chlorinated ethanes, including EDC. It seems clear that a value of 22.0
X 10~14 cm3 molecule'1 sec'1 is much more consistent with the other
measured values than is a value of 6.5 X 10~14 cm3 molecule'l sec'1.
Once a room temperature rate constant is selected from the experi-
mental data, one still needs to estimate the rate constant at cooler
temperatures in order to estimate the lifetime. In estimating the
lifetime of EDC, Altshuller20 chose a temperature of 265 K, stating that
it is very close to the average annual tropospheric temperature weighted
37
-------
for distribution of species with altitude. He then uses rates at this
temperature to represent an average tropospheric reaction rate. When the
rate constant at 265 K is not known, Altshuller estimates it by dividing
TABLE 8. HYDROXYL RADICAL RATE CONSTANTS FOR
ETHANE AND SEVERAL CHLORINATED ETHANES
«aaaa«»a«a»a««aa«»»«««aaa««a«aa«a«a-»«»»«m«at«jaa«3aM»a«»»»«aa3»aa«a33a3a»
Chemical Rate Constant .
Name Units of 10~14 cm3 molecule'1 sec"1
CH3CH3 2713
CH3CH2a 3960, 3963, 4461
CH3CHC12 2660
CH2aCH2a (EDC) 2260, 6.561
CH2ClCHa2 3313, 3360
the 298 K rate constant by 1.75. This is equivalent to an activation
energy of 2.66 Real mol"1. This relationship is strictly empirical and is
applied only to saturated organics. He estimates the EDC rate constant
as 22/1.75= 12.6 X 10~14 on3 molecule-1 sec'1. Snelson et al.61 calcu-
late a rate of 12.5 X 10"14 cm3 molecule"1 sec"1 at 265 K and use that
value to estimate the tropospheric lifetime of EDC. They calculate the
265 K rate constant by using an assumed activation energy of 2.07 kcal
mol"1, which they calculated by averaging seven values for the activation
energy of chlorinated methanes and ethanes taken from the literature.
(In two instances, E/V. values were used. The actual average of the
activation energies is 2.49 kcal mol"1.) Atkinson has apparently
reccmnended to Singh9 that the best activation energy to assume is that
of methylene chloride. Atkinson's reconnended expression for the
temperature dependence of methylene chloride, given previously, is
equivalent to an activation energy of 2.1 kcal mol"1. In light of the
small number of activation energies which have been measured for analo-
gous species, Atkinson's recommendation to use the activation energy of
38
-------
I .« • • «T> »« ^. * i~S \*f
methylene chloride seems roost appropriate for EDC. An activation energy
of 2.1 Real mol-1 is assured for the lifetime calculations in Table 9.
TABLE 9. ETHYLENE DICHLORIDE REACTION RATE CONSTANTS AND LIFETIMES
Temperature OH Reaction Rate Constant Assumed [OH] Lifetime
K 10~14 cm3 roolec"1 s"1 106 roolec cm"3 •. days
288 19.0 1.0 61
263 12.6 1.0 92
260 11.9 0.5 195
Other removal processes are, once again, not likely to be effective
in removing this air toxic. For example/ the atmospheric lifetime
resulting from rain out processes has been estimated to be around 400
years.3 From Table 9, one would select the value of 92 days as the
lifetime, since this falls within the appropriate range for the verti-
cally well-mixed troposphere. Given the uncertainty in the activation
energy however, the value of 92 days should only be viewed as the most
probable value. If one selects a factor of two as a margin of safety,
then one estimates that the EDC residence time is between 0.13 and 0.50
years. The upper limit value is close to the residence time of 0.6 ± 0.2
years recently estimated by Singh et al.9 from ambient concentration
measurements and a global budget model. Singh's calculated residence
time does not depend upon an assumed [OH]: it is based, however, on a
sparse set of ambient data. There are only 16 measurements, 8 in each
hemisphere (based upon a preprint copy of reference 9 kindly supplied by
Dr. Singh). All samples were collected over the eastern Pacific Ocean in
December 1981. Variations in the release rate for EDC may have a sub-
stantial etfect on the calculated residence time.
39
-------
TRICHLOROE7IHYLENE
Trichlotoethylene (trichloroetnene) is a chlorinated alkene which
has found use as a degreaser and solvent.58 It, too, is a mutagen and a
suspect carcinogen. 21 The Northern Hemispheric background concentrations
in the Eastern Pacific Ocean were reported as 12 ppt.9 Brodzinsky and
Singh57 reported median urban concentrations of 150 ppt for measurements
reported between 1970 and 1980. A two-hour average concentration of 900
ppt was recently reported for stagnant conditions in San Jose.22
The available rate data for reactions with OH radicals are in
reasonable agreement. Atkinson^-3 reccnntends calculating the temperature
dependent rate constants using the Arrhenius expression:
k(CHClCCl2) = 5.63 x 10"13 exp(427/T) on3 molecule'1 s"1 . (16)
Notice that the exponential function in Equation (16) is raised to a
positive power. This means that the activation energy is actually
negative and that the rate constant for this chemical will increase as
the temperature cools. A negative activation energy is consistent with
an addition reaction, as is often seen in alkenes.13 One of the antici-
pated products from this chemicals decomposition in the atmosphere is
phosgene.3'^4 The recommended rate constants for reaction with OH
radicals are given in Table 10.
TABLE 10. TRICHLOROETHYLENE REACTION RATE CONSTANTS AND LIFETIMES
Temperature OH Reaction Rate Constant Assumed [OH] Lifetime
K 10~14 cm3 molec'1 s"1 106 molec cm~3 days
288 248 1.0 4.7
263 286 1.0 4.1
260 291 0.5 8.0
40
-------
ft-'
Although trichloroethylene is an alkene, with a double bond which
may be subject to attack by ozone, the only reported rate constant for
reaction with ozone is very low. 14 Ozonolysis, therefore, plays no
significant role in the removal of this compound. Other removal
mechanisms are also expected to be ineffective. The lifetime, then, is
computed solely from the hydroxyl removal rate. The lifetime for the
boundary layer of 4.7 days is at the upper end of the applicable range.
The lifetime computed for the vertically mixed troposphere is even
shorter, due to the negative activation energy exhibited by this com-
pound. The lifetime of trichloroethylene is estimated as slightly over
four days.
PERCHLOROETHYLENE
Perchioroethylene (tetrachloroethene) is a suspect carcinogen21 for
which lifetime risk estimate for carcinogenesis has been published.20
The chemical is quite volatile and is commonly used as a degreaser and
solvent.58 It is emitted in significant amounts from dry cleaning opera-
tions.20 The chemical is consistently found in urban areas of the U.S.
The data compilation of Brodzinsky and Singh57 reported a median concen-
tration of 340 ppt in urban and suburban areas of the U.S. During a
stagnant period in San Jose in December 1985, a two-hour integrated
sample yielded a value of 6639 ppt.22
The smog chamber data for perchloroethylene was recently reviewed by
Dunitriades et al.65 They concluded that much of the data reported in
the literature on mechanisms and product distribution in smog chamber
irradiations was influenced by chlorine atom chain reactions. Atkinson
recently reviewed the data on reactions between OH radicals and per-
chloroethylene. 13 prom a consistent set of kinetic rates, he derived the
following Arrhenius expression for the rate constant:
41
-------
k(CCl2CCl2) " 9.64 x 10~12 exp(-1209/T) cm3 molecule"1 s"1 . (17)
Unlike the case of the other chlorinated alkene, trichloroethylene,
above, the rate constant for perchloroethylene decreases as the tempera-
ture cools. The reconmended rate constants are tabulated below, one of
the reaction products which results from the atmospheric photooxidation
of this compound is phosgene.3
TABLE 11. PERCHLOROETHYLENE REACTION RATE CONSTANTS AND LIFETIMES
Temperature
K
288
263
260
OH Reaction Rate Constant
10~14 on3 molec'1 sT1
14.5
9.7
9.2
Assuned [OH]
10^ molec cm"3
1.0
1.0
0.5
Lifetime
days
80
119
251
The only reported value for the rate constant of the reaction of
perchloroethylene and ozone14 is very small, implying that ozone will not
effectively remove the chemical from the atmosphere. Dry deposition
rates of perchloroethylene to seme cannon surface materials found in
urban areas were recently measured.56 The rates were so slow that they
were often indistinguishable from zero. Dry deposition, therefore, does
not appear to be a significant removal pathway. The lifetime estimated
from hydroxyl removal rates, taken from Table 11, should bo expressed as
ranging fron 119 to 251 days. This value is completely consistent with
the estimate of Singh et al.9 of 146 to 292 days, based upon a two-box
model.
1,3-BUTADIQJE
1,3-butadiene is a very reactive alkene with two double bonds, it
is connonly used as a component in the synthesis of rubber and many other
42
-------
diverse compounds.58 Data on its occurrence in urban areas is not very
extensive, but a median value of 1500 ppt was reported by Brodzinsky and
Singh. 57 Host of the reported data points were from only one city,
however.
Atkinson has recently reviewed the reactions of butadiene with both
hydroxyl radicals and ozone.13»14 Much of the data on hydroxyl reactions
is in excellent agreement. There was only one reported temperature
dependence with an Arrhenius activation energy of -0.93 kcal mol'1. The
negative activation energy implies that the rate should increase as the
temperature decreases. The recommended GH reaction rate constant is
given by the expression:
k(CH2-CHCH»CH2) - 6.68 x 10~U exp(468/T) era3 molecule'1 s"1 . (18)
The recommended ozonolysis rate constant at room temperature is 8.1 x
10~18 cm3 molecule'1 s"1.14 The reaction with ozone has an activation
energy of around 5.5 kcal mol~l. The resultant rate constant at 288 K is
5.9 x 10~18 cm3 molecule'1 sr1. The recommended rate constants and the
computed lifetimes are shown in Table 12. Unlike the similar previous
tables, the lifetimes in Table 12 are given in minutes. In calculating
the ozone removal rate, an estimated concentration of 40 ppt (1 x 1012
molecules cm~3) was assumed.10»20
TABLE 12. 1,3-BUTADIENE REACTION RATE CONSTANTS AND LIFETIMES
Temperature Reaction Bate Constant Assumed Cone. Lifetime
K 10-14 an3 molec-1 s'1 106 roolec cm'3 mins
288 7060 1.0 236
263 8240 1.0 202
260 8410 0.5 396
288 Ozone: 0.00059 Ozone: 1,000,000 Ozone: 2800
==============
43
-------
The estimated lifetime of butadiene is very shorn. The tabulated
lifetimes were calculated for "average* conditions as *ere described
above, obviously, the actual lifetime of such a reactive compound will
depend upon the specific conditions at the time of release. Because
the estimated lifetime is so short, the actual degradation of any real
emissions is very dependent upon time of day, sunlight intensity, actual
temperature, etc. Vtiile the lifetime during the middle of the day in the
summer under polluted conditions could be much shorter than the' estimated
4 hours, the lifetime of emissions at night could be essentially infi-
nite. After sunset, there will be no hydroxyl radicals generated and the
small amounts of residual ozone present in the evening will have little
effect on the butadiene concentrations. On average then, 1,3-butadiene
has an estimated lifetime of around 4 hours.
ETHYLENE OXIDE
Ethylene oxide (oxirane) is the smallest possible organic epoxide.
The nature of the chemical structure induces a high strain energy in the
three-membered ring, and this strain energy influences the reaction
kinetics and products.66 Ethylene oxide finds its use as an intermediate
in the synthesis of ethylene glycol and as a sterilant or pesticide.58
The chemical is a mutagen and suspect carcinogen, having been classified
as being probably carcinogenic to humans by EPA's Carcinogen Assessment
Group.67
While ethylene oxide has been monitored in the workplace, data on
ambient concentrations of ethylene oxide is very sparse. Brodzinsky and
Singh57 did not report finding any measurements published during the
period 1970-80. At this point, no ambient measurements are known.
Two investigators have recently measured the reaction rate constant
for ethylene oxide with OH radicals. 13*66 xhe experimentally determined
room temperature rate constants were 5.3 and 8.0 x 10~14 cm3 molecule"!
s"1. An activation energy of 2.9 kcal mol~l was measured by one of the
44
-------
investigators across the temperature range 297 K to 435 K. ttiile one
should be careful in extrapolating the temperature dependency outside the
measured range, it is still reasonable to assume a similar activation
energy across the small range to 260 K. Such an assumption does cause
greater uncertainty in any lifetime estimates, however. The larger rate
constant and the reported activation energy were used to estimate the
rate constants shown in Table 13. The choice of the larger rate constant
means that the estimated lifetimes are actually on the low side of the
possible values*
TABLE 13. ETOYLENE OXIDE REACTION RATE CONSTANTS AND LIFETIMES
mmmmmmmmmmmmmmmm
Temperature OH
K
288
263
260
Reaction Rate Constant
10-14 0*3 moiec-1 s'l
6.9
4.3
4.0
Assumed (OH)
106 molec on~3
1.0
1.0
0.5
Lifetime
days
167
217
578
Of the eiyht chemicals named in "Intent to List" notifications,
ethylene oxide is the most soluble in water. When the concentrations of
a chemical in wacer and air are expressed in the same units (e.g., moles
liter"1',, the ratio of the aqueous phase concentration to the vapor phase
concentration is defined as the dimensionless solubility parameter, a.
This value has recently been measured at 288 K and found to be 6.2.5*
This means that ethylene oxide does distribute preferentially into the
aqueous phase. Even without considering revolatiliaption of the chemi-
cal, however, rain out will still not be effective in removing the
chemical fron the environment: the estimated lifetime due to rain out is
hundreds of years. Experimental measurements and theoretical modeling of
rain out effects have demonstrated little impact from rain out for gases
which are even far more soluble than ethylene oxide.56
45
-------
Nor should the chemical distribute into th> aqueous phase on ambient
aerosols and be removed by deposition of the aerosol. Even a worst case
condition of 150 pg nT? of aerosol, all of it being water, will reduce
the gas phase concentration by only one part in one billion. If rapid
hydrolysis reactions were to occur to the ethylene oxide dissolved in the
aqueous aerosol, that chemical process could increase significantly the
loss by this mechanism. Half lives of ethylene oxide in the aqueous
phase have been reported^8 to fall between 200 and 400 hours for a wide
variety of types of water (e.g., sterile distilled water, sea water,
fresh water, sterile and non-sterile river water). Such long lifetimes
suggest that hydrolysis reactions in aqueous aerosols are also not likely
to be fast.
No other removal process are known which can rapidly deplete the
ethylene oxide from the air. Results fron smog chamber irradiations69'70
in both natural sunlight and artificial illumination (private communica-
tion, Dr. E. Bdney, Northrop Services, Inc., Research Triangle Park,
North Carolina) are consistent with a slowly reacting organic chemical:
they suggest that t ire is not seme overlooked chemical or photolytic
process occurring to remove ethylene oxide. The estimated lifetime,
therefore, can be calculated simply from the OH radical removal rate.
Fran Table 13, one estimates the lifetime as 217 to 578 days.
This estimate of lifetime is in disagreement with a previous EPA
report by Bogyo et al.7* and a monograph72 by SRI International for the
National Cancer Institute. Those references conclude that "ethylene
oxide is highly reactive and does not persist in the environment" and
that epoxides like ethylene oxide are "expected to degrade rapidly" in
the environment. The SRI conclusion is based upon an extrapolation of
the work by Bogyo et al. Bogyo's conclusions are based upon a few liquid
phase experiments which may not be applicable and upon a single 1976
publication by Darnall et al.7^ in which the "reactivity" of various
organics was ranked according to their reactivity with OH radicals.
Citing the Darnall reference, Bogyo et al. state "ethers as a class
46
-------
(epoxides are a type of ether) have been classified among the most
reactive hydrocarbons." That sentence implies that the conclusion of
•rapid* removal is based upon a doubtful analogy with unstrained ethers.
Darnall et al. do not rank any ethers at alll The only reference to
ethers is found in a table copied from an earlier publication7* which
concluded that ethers were capable of producing significant quantities of
ozone. It also illustrates a misinterpretation of the literature: the
word "reactivity" used in the paper cited by Darnall et al. referred to
the ozone forming potr>tial, and not necessarily the rate of removal.
The conclusion that ethylene oxide "does not persist" is not warranted,
in light of the recent kinetic data.
It is interesting that ethylene oxide, with an estimated lifetime as
long as 1.5 years, has not been observed in the ambient atmosphere.57 A
study75 of breakthrough volumes in Tenax concluded that there was no safe
sampling volume for ethylene oxide when using Tenax. It is not surpris-
ing, therefore, that previous data from Tenax measurements did not
include ethylene oxide. Although Singh et al.21»22 have carried out a
great deal of ambient measurements using a different technique, ethylene
oxide was never one of the chemicals which they attempted to measure. In
a recent EPA field study using samples collected in polished stainless
steel canisters, all attempts to measure ethylene oxide were confounded
by an interference from the water peak (private ccmnunication, T. A.
Hartlage, U.S. EPA, Research Triangle Park, NC). -A variety of other
methods have been reported in the literature for use in analyzing for
ethylene oxide.67»76~79 These methods all have reported sensitivities
frcm 0.05 parts per million to greater than 3 parts per million. The
1982 estimate80 for production in the U.S. was 5000 million pounds, or
2270 million kilograns. Assuming that the total Northern Hemispheric
production is twice that of the U.S. and that one-fourth of all the
material produced is vented to the atmosphere, one calculates an annual
input to the Northern Hemisphere of 1.14 x 10*2 grams. If the OH radical
decay rate is taken as 1/1.5 year~l and the transfer rate to the southern
Hemisphere is assumed9 to be 1/1.2 year~lr one estimates a background
47
-------
concentration in the Northern Hemisphere to be around 240 ppt. This
value is a factor of 200 to 12,000 below the quoted analytical detection
limits. Even if excursions of factors of 10 to 100 above geochemical
background were to occur in urban areas (analogous to sane of the
observations of Singh et al,2l»22 for other pollutants), the concentra-
tions would still likely be below the detection limit.
It is not surprising then, that no ambient data on ethylene oxide
have been reported. Nor does the lack of ambient data argue that there
must be seme rapid, but unknown, removal mechanism, until additional
data arrive to modify these conclusions, it is appropriate to assign
ethylene oxide an atmospheric lifetime of fron 0.6 to 1.5 years.
48
-------
REFERENCES
1. Sinroonds, P. G., P. N. Alyea, C. A. Cartelino, A. J. Crawford, D. M.
Cunnold, B. C. Lane, J. E. Lovelock, R. G. Prinn, R. G. and R. A.
Rasmussen. The Atmospheric Lifetime Experiment. 6. Results for
Carbon Tetrachloride Based on 3 Years Data, journal of Geophysical
Research. 88: 8427-8441, 1983.
2. The American Chemical Society. Cnemical Abstracts Service Registry
Handbook, Number Section, 1965-1971. (ISSN 0093-058X). Chemical
Abstracts Service, The Ohio State University, Columbus, Ohio, 1974.
2028 pp.
3. Cupitt, L. T. Fate of Toxic and Hazardous Materials in the Air
Environment. EPA-600/3-80-084, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1980, 29 pp.
4. Junge, C. E. Residence time and variability of tropospheric gases.
Tellus XXVK4); 477-488, 1974.
5. Slinn, W. G. N. Relationships Between Removal Processes and
Residence Times for Atmospheric Pollutants. (NTIS CONF-780611-3,
March 1978) In: AIChE Symposium Series, No. 196, Vol. 76, 1980. pp.
185-203.
6. aolin, B. and H. Rodhe. A note on the concepts of age distribution
and transit time in natural reservoirs. Tellus XXV(l); 58-62, 1973.
49
-------
7. Lyman, W. J. Atmospheric Residence Time. In- Handbook of Chemical
Property Estimation Methods: Environmental Behavior of Organic
Compounds, W. j. Lyman, W. P. Reehl, and D. H. Rosenblatt, eds.
McGraw-Hill Book Company, New York, New York, 1982. p. 10-1.
8. Prinn, R. G., P. G. Simmonds, R. A. Rasmussen, R. D. Rosen, F. N.
Alyea, C. A. Cardelino, A. J. Crawford, D. M. Cunnold, P. J. Fraser,
and J. E. Lovelock. The Atmospheric Lifetime Experiment: 1.
Introduction, Instrumentation and Overview. Journal of Geophysical
Research. 88: 8353-8367, 1983.
9. Singh, H. B., L. J. Salas, and R. E. Stiles. Selected Man-made
Halogenated Chemicals in the Air and Oceanic Environment. Journaj.
of Geophysical Research. 88: 3675-3683, 1983.
10. Hendry, D. G. and R. A. Kenley. Atmospheric Reaction Products of
Organic Compounds. EPA-560/12-79-001, U.S. Environmental Protection
Agency, Washington, D.C., 1979, 81 pp.
11. Sprung, J. L. Tropospheric Oxidation of f^S. In: Advances in
Environmental Science and Technology, vol. 7, J. N. Pitts, Jr., and
R. L. Metcalf, eds. John Wiley and Sons, New York, New York, 1977.
pp. 263-278.
12. Graedel, T. E. Chemical Compounds in the Atmosphere. Academic
Press, New York, New York, 1978. 440 pp.
13. Atkinson, R. Kinetics and Mechanisms of the Gas-Phase Reactions of
the Hydroxyl Radical with Organic Compounds under Atmospheric
Conditions. Chem. Rev. 85: 69-201, 1985.
50
-------
14. Atkinson, R. and W. p. L. Carter. Kinetics and Mechanisms of the
Gas-Phase Reactions of Ozone with Organic Compounds under Atmo-
spheric Conditions. Chem. Rev. 84s 437-470,1984.
15. Carter, W. P. L., A. M. Winer, and J. N. Pitts, Jr. Major Atmo-
spheric Sink for Phenol and the Cresols: Reaction with the Nitrate
Radical. Environmental'Science & Technology. 15: 829-831, 1981.
16. Logan, J. A., M. J. Prather, S. C. WOfsy, and M. B. McElroy.
Tropospheric Chemistry: A Global Perspective. Journal of Geophys-
ical Research. 86: 7210-7254, 1981.
17. Sherwood, G. E. F. and A. E. Taylor. Calculus, Third Edition.
Prentice-Hall, Inc., Englewood Cliffs, New Jersey, 1954. 579 pp.
18. Sanders, H. J. Chemistry and the Atmosphere: This Most Excellent
Canopy, the Air. In: The Earth's Atmosphere, W. W. Vaughn and L.
DeVries, eds. American Institute of Aeronautics and Astronautics,
New York, New York, 1972. pp. 69-93.
19. Peters, L. K. and G. R. Carmichael. Modeling of Transport and
Chemical Processes that Affect Regional and Global Distributions of
Trace Species in the Tropospherei, In: Trace Atmospheric Constitu-
ents: Properties, Transformations, and Fates. Advances in Environ-
mental Science and Technology, Volume 12, S. E. Schwartz, ed. John
Wiley & Sons, New York, New York, 1982. p. 493-538.
20. Altshuller, A. P. Lifetimes of organic Molecules in the Troposphere
and Lower Stratosphere. In: Advances in Environmental Science and
Technology, Volume 10, J. N. Pitts, Jr., R. L. Metealf and D.
Grosjean, eds. John Wiley and Sons, New York, New York, 1980. pp.
181-219.
51
-------
21. Singh, H. B., L. J. Salas, R. Stiles, and H. Shigeishi. Measure- <
ntents of Hazardous Organic Chemicals in the Ambient Atmosphere.
EPA-600/3-83-OU2, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, 1983, 91 pp. ,
i
22. Singh, H. B., R. J. Ferek, L. J. Sal as, and K. C. Nitx. Toxic
Chemicals in the Environment i p. program of Field Measurements. j.
EPA/600/3-86/047, U. S. Environmental Protection Agency, Research ;
Triangle Park, North Carolina, 1986, 92 pp.
.4
I
'
;
23. Junge, C. E. Bade Considerations About Trace Constituents in the '
Atmosphere as Helated to the Fate of Global Pollutants. In: Fate of ; .
t
Pollutants in the Air and Water Environments: Part 1, Mechanism of •
Interaction Between Environments and Mathematical Modeling and The :
Physical Fate of Pollutants. Advances in Environmental Science and
Technology, Volune 8, I. H. Suffet, ed. John Wiley & Sons, New
York, New York, 1975. pp. 7-25.
24. Lane, H. U., ed. The World Almanac and Book of Facts 1986.
Newspaper Enterprise Association, Inc., New York, New York, 1985, :
928 pp. ;
I
25. U.S. Department of Commerce. Statistical Abstract of the United ;
States 1986. Washington, D.C., 1985. 985 pp. j
I
26. Environmental Science Services Adrainstration, National Aeronautics i
and space Administration, and the United States Air Force. U.S. ;
Standard Atmosphere Supplements, 1966. U.S. Government Printing
Office, Washington, D.C., 1966. 289 pp.
27. Murrell, D. Passenger Car Fuel Economy: EPA and Road. EPA-460/3-
80-010, U.S. Environmental Protection Agency, Ann Arbor, Michigan,
1980, 292 pp.
52
-------
28. Heicklen, J. Atmospheric Chemistry. Academic Press, New York, New
York, 1976. 406 pp.
29. Heicklen, J. Atmospheric Lifetimes of Pollutants. Atmospheric
Environment. 16I 821-823. 1982.
30. National Oceanic and Atmospheric Administration, National Aero-
nautics and Space Administration, and the United States Air Force.
U.S. Standard Atmosphere, 1976. U.S. Government printing Office,
Washington, D.C., 1976. 227 pp.
31. Derwent, R. G. On the Comparison of Global, Hemispheric, One-
Dimensional and Two-Dimensional Model Formulations of Halocarbon
Oxidation by OH Radicals in the Troposphere. Atmospheric Environ-
ment. 16: 551-561, 1982.
32. Perner, D. and G. HQbler. Experimental Detection of OH in the
Troposphere. In: Chemistry of the Unpolluted and Polluted Tropo-
sphere, H. W. Georgii and W. Jaeschke, eds. D. Reidel Publishing
Company, Dordrecht, Holland, 1982. pp. 267-294.
33. Perner, D., D. H. Ehhalt, H. W. Patz, U. Platt, E. P. R&th and A.
Volz. OH-Radicals in the Lower Troposphere. Geophysical Research
Letters. 3: 466-468, 1976.
34. Campbell, M. J., J. C. Sheppard and B. F. Au. Measurement of
Hydroxyl Concentration in Boundary Layer Air by Monitoring CO
Oxidation. Geophysical Research Letters. 6: 175-178, 1979.
35. Davis, D. D., W. Heaps, and T. McGee. Direct Measurements of
Natural Tropospheric Levels of OH via an Aircraft Borne Tunable Dye
Laser. Geophysical Research'Letters^ 3: 331-333, 1976.
53
-------
V
36. Davis, D. D., W. Heaps, D. Philen and T. McGee. Boundary Layer
Measurements of the OH Radical in the Vicinity of an Isolated Power
Plant Plume: 802 and N02 'Chemical Conversion Times. Atmospheric
Environment. 13: 1197-1203, 1979.
37. Wang, C. C. and L. I. Davis, Jr. Measurement of Hydroxyl Concentra-
tions in Air using a Tunable uv Laser Beam. Physical Review
Letters. 32t 349-351, 1974.
38. Bakalyar, D. M., L. I. Davis, Jr., C. Quo, j. V. James, S. Kakos, P.
T. Morris and C. C. Wang. Shot Noise Limited Detection of Hydroxyl
Using the Technique of Laser-Induced Fluorescence. Applied Optics.
23: 4076-4082, 1984.
39. Hjorth, J., G. Ottobrini, F. Cappellani, G. Restelli, H. Staryl and
C. Lohse. Hydroxyl Radical Concentration in Ambient Air at a
Semirural Site Estimated From Carbon-13 Monoxide oxidation. In:
Com. Eur. Communities, EUR 9436, Physical-Chemical Behavior of
Atmospheric Pollutants, 1984. pp 216-226.
40. Calvert, J. G. Hydrocarbon Involvement in Photochemical Smog
Formation in Los Angeles Atmosphere. Environmental Science &
technology.. 10: 256-262, 1976.
41. Singh, H. B., J. R. Martinez, D. G. Hendry, R. J. Jaffe, and W. B.
Johnson. Assessment of the Oxidant-Forming potential of Light
t Saturated Hydrocarbons in the Atmosphere. Environmental Science &
Technology. 15: 113-119, 1981.
42. Singh, H. B., L. J. Salas, B. K. Cantrell and R. M. Redmond.
Distribution of Aromatic Hydrocarbons in the Ambient Air. Atmo-
spheric Environment. 19: 1911-1919, 1985.
I
54
-------
•^**J*^^
43. Anderson, L. G. Pate of Nitrogen oxides in Urban Atmospheres, in:
Trace Atmospheric constituents! Properties, Transformations, and
Pates. Advances in Environmental Science and Technology, vol. 12, S.
E. Schwartz, ed. John Wiley 6 Sons, New York, New York, 1983. pp.
371-409.
44. Crutzen, P. J. and J. Pishman. Average Concentrations of OH in the
Troposphere, and the Budgets of 014, GO, H2, and 01303.3. Geophys-
ical Research Letters. 4: 321-324, 1977.
45. Watanabe, T., M. Yoshida, S. Pujiwara, K. Abe, A. Onoe, M. Hirota,
and S. Igarashi. Spin Trapping of Hydroxyl Radical in the Tropo-
sphere for Determination by Electron Spin Resonance and Gas Chrona-
tography/Mass Spectronetry. Analytical Chemistry. 54i 2470-2474,
1982.
46. wang, C. C., L. I. Davis, jr., P. M. Selzer, and R. Nunoz. improved
Airborne Measurements of OH in the Atmosphere Using the Technique of
Laser-Induced Fluorescence, journal of Geophysical Research. 86:
1181-1186, 1981.
47. Singh, H. B. Atmospheric Halocarbons: Evidence in Favor of Reduced
Average Hydroxyl Radical Concentration in the Troposphere. Geophys-
ical Research Letters. 4: 101-104, 1977.
48. Singh H. B. Preliminary Estimation of Average Tropospheric HO
Concentrations in the Northern and Southern Hemispheres. Geophys-
ical Research Letters^ 4: 453-456, 1977.
49. Neely, W. B. and J. H. Plonka. Estimation of Time-Averaged Hydroxyl
Radical Concentration in the Troposphere. Environmental Science &
Technology. 12: 317-321, 1978.
55
-------
it'ti^t:. ; •»• - ;• >_; • .'••..-.. ' • • • '
50. Jeong, K. and F. Kaufiman. Rates of the Reactions of 1,1,1-Tri-
chloroethane (Methyl Chlorofoiro) and 1,1,2-Trichloroe thane with OH.
(Seophysical'Research'tefcters. 6: 757-759, 1979.
51. Kurylo, M. J. , P. C. Anderson, and 0. Klais. A Flash Photolysis
Resonance Fluorescence Investigation of the Reaction of OH *
Geophysical' Research 'Letters. 6: 760-762, 1979.
52. Volz, A. , D. H. Ehhalt, and R. G. Derwent. Seasonal and Latitudinal
Variation of l^CO and the Tropospheric Concentration of OH Radicals.
Journal of Geophysical 'Research. 86: 5163-5171, 1981.
53. Sehmel, G. A. Particle and Gas Dry Deposition: A Review. Atmo-
spheric Environment. 14: 983-1011, 1980.
54. Sehmel, G. A., R. N. Lee, and T. W. Horst. Hazardous Air Pollu-
tants: Dry-Deposition Phenomena. EPA-600/3-84-114 , U.S. Environ-
mental Protection Agency, Research Triangle Park, North Carolina,
1984, 71 pp.
55. Eisenreich, S. J. , B. B. Looney, and J. D. Thornton. Airborne
organic contaminants in the Great Lakes ecosystem. Environmental
Science & Technology. 15: 30-38, 1981.
56. Dana, M. T. , R. N. Lee, and J. M. Hales. Hazardous Air Pollutants:
Wet Removal Rates and Mechanisms. EPA-600/3-84-113. U.S. Environ-
mental Protection Agency, Research Triangle Park, North Carolina,
1984, 98 pp.
57. Brodzinsky, R. and H. B. Singh. Volatile Organic Chemicals in the
Atmosphere: An Assessment of Available Data. EPA-600/3-83-027(a).
U.S. Environmental Protection Agency, Research Triangle Park, North
Carolina, 1983, 199 pp.
56
-------
58. Windholz, M.r ed. The Merck Index, Ninth edition. Merck 6 Co.,
Inc., Rahway, New Jersey, 1976. 1313 pp.
59. Rasnussen, R. A. and J. E. lovelock. The Atmospheric Lifetime
Experiments 2. Calibration. Journal 'of 'Geophysical' Research. 88:
8369-8378, 1983.
60. Howard, C. J. and K. M. Evenson. Rate constants f.or the reaction of
OH with ethane and sane halogen substituted ethanes at 296 K. The
Journal of Chemical Physics. 64s 4303-430*, 1976.
61. snelson, A., R. Butler, and P. Jarke. study of Removal Processes
for Halogenated Air Pollutants. EPAH500/3-78-058, U.S. Environ-
mental protection Agency, Research Triangle Park, North Carolina,
1978, 107 pp.
62. Atkinson, R., K. R. Darnall, A. C. Lloyd, A. M. Winer, and J. N.
Pitts, Jr. Kinetics and Mechanisms of the Reactions of the Hydroxyl
Radical with Organic Compounds in the Gas Phase. Ini Advances in
Photochemistry, Vol. 11, J. N. Pitts, jr., G. S. Hanmond, K.
Gollnick, and D. Grosjean, eds. John Wiley & Sons, New York, New
York, 1979. pp. 375-488.
63. Paraskevopoulos, G., D. L. Singleton, and R. S. Irwin. Rates of OH
Radical Reactions. 8. Reactions with O^FCl, Otf^Cl, CHFC12,
CH3CF2C1, CH3C1 and C2H5C1 at 297 K. journal of Physical Chemistry.
85: 561-564, 1981.
64. EHney, E., S. Mitchell, and J. J. Bufalini. Atmospheric Chemistry
of Several Toxic Compounds. EPA-600/3-82-092, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina, 1983.
109 pp.
57
-------
65. Dimitriades, B., B. W. Gay, Jr., R. R. Arnts, and R. L. Seila.
Photochemical Reactivity of Perchloroethylenei A New Appraisal.
Journal of the' Air Pbllutioh'Confcrol Asabciattoh. 331 575-587, 1983.
66. Fritz, B., K. Lorenz, N. Steinert, and R* Zellner. Laboratory
Kinetic Investigations of the Tropospheric oxidations of Selected
Industrial Emissions, ins Coon. Bur. Connunltles, EUR 7624,
Physical-Chemical Behavior of Atmospheric Pollutants, 1982. pp.
192-202.
67. Office of Health and Environmental Assessment. Health Assessment
Document on Ethylene Oxide. EPA-600/8-84-009(f), U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina, 1985.
68. Conway, R. A., G. T. Waggy, M. H. speigel, and R. L. Berglurr*,.
Environmental fate and effects of ethylene oxide. Environmental
Science t Technology. 17i 107-112, 1983.
69. sickles, J. B. II, R. S. Wright, C. R. Sutcliff, A. L. Blackard and
D. P. Dayton. Snog Chamber Studies of the Reactivity of Volatile
Organic Compounds. Proc. Ann. Meet. Air Pollut. Control Associa-
tion, Paper 80-501, 1980. 16 pp.
70. spicer, C. W., R. N. Riggin, M. W. Holdren, F. L. DeRoos, and R. N.
Lee. Atmospheric Reaction Products from Hazardous Air Pollutant
Degradation. EPV600/3-85/028, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1985. 79 pp.
71. Bogyo, D. A., S. S. Lande, W. M. Meylan, P. H. Howard, and J.
Santodonato. Investigation of Selected Potential Environmental
Contaminantsi Epoxldes. EPA-560/11-80-005, U.S. Environmental
Protection Agency, Washington, D.C., 1980. 201 pp.
58
-------
72. SRI International. Monographs on Organic Air pollutants. Prepared
for the National Canon: Institute under contract N01-CP-26004-02,
1983.
73. Darnall, K. R., A. C. Lloyd, A. M. Winer, and J. N. Pitts, Jr.
Reactivity scale for Atmospheric Hydrocarbons Based on Reaction with
Hydroxyl Radicals. Environmehfcal 'Science't 'Technology. 101 692-
696, 1976.
74. Chemistry and Physics Laboratory. Proceedings of the Solvent
Reactivity Conference. EPA-650/3-74-010, U.S. Environmental Protec-
tion Agency, Research Triangle Park, North Carolina, 1974. 41 pp.
75. Brown, R. H. and C. J. Purnell. Collection and analysis of trace
organic vapor pollutants in ambient atmospheres. The performance of
a Tenax-QC adsorbent tube. Journal'of Chromatography. 178s 79-90,
1979.
76. Dmitriev, H. T. and V. A. Mishchikhin. Gas chronatographic deter-
mination of ethylene oxide in the atmosphere. (Abstract only). Gig.
Sanit. 4i 65-68, 1982.
77. Kapila, S., R. K. Malhotra, and C. R. Vogt. A versatile test
atmosphere generation and sampling system. In: Chemical Hazards in
the workplace: Measurement and Control, ACS Symposium Series, Volume
149, 1981. pp. 533-542.
78. Syrjala, R. J. Quantitative analysis of atmospheric pollutants using
a microcomputer-controlled single beam infrared spectrometer, in:
Environmental Analysis (Pap. Annu. Meet. Fed. Anal. Chan. Spectrosc.
Sec.), 3rd, 1977. pp. 111-125.
59
-------
^^^•^^^^
79. Mouilleaeaux, A., A. N. Laurent, M. Fabre, M. Jcxian, and B. Festy.
Atmospheric oonoantration of ethylene oxide in the occupational
environment of disinfection and sterilization facilities. (Abstract
only.) Arch: "Mai: 'Prof; 'ted: 'tray.' 'Secur. 'tec; 44s 1-14, 1983.
80. American Chemical society. Output down for chemicals, related
products. Chemical"fc Engineering'tews. 61s 28-34, 1983.
60
------- |