vvEPA
United States
Environmental Protection
Agency
Environmental Sciences Research
Laboratory
Research Triangle Park, NC 27711
Research and Development EPA-600/D-83-070 Aug. 1983
ENVIRONMENTAL
RESEARCH BRIEF
An Assessment of Ozone as a Surrogate for Other Products of
Atmospheric Photochemical Reactions
A. P. Altshuller
In addition to ozone and nitrogen dioxide, many other
gaseous and aerosol products of atmospheric photo-
chemical reactions occur alone or in combination at
significant concentration levels compared to ozone in urban
areas. The more abundant of these other products include
formaldehyde, nitric acid, peroxyacetyl nitrate and sulfate.
Aldehydes are the most abundant group of such products.
At the time of maximum ozone concentration, these other
products in combination can constitute about half of the
ozone concentration in the South Coast Air Basin of
California. On a 24-hour average basis, these products in
combination can equal or exceed the.ozone concentration
in urban areas.
Ozone should serve directionally as a satisfactory surrogate
for these other products of atmospheric photochemical
reaction in the ambient atmosphere. However, laboratory
studies do indicate some circumstances in which ozone
may not be an adequate surrogate directionally. In addition,
both ambient air measurements and laboratory results
indicate that complex quantitative relationships exist
between ozone and other products, and between these
products and the precursor hydrocarbons and nitrogen
oxides.
Monitoring for aldehydes and peroxyacetyl nitrate has been
carried out at locations in the South Coast Air Basin and in
New Jersey. It appears practical to carry out monitoring
efforts for products other than ozone if necessary in a
limited number of critical urban areas.
Introduction
This assessment is intended as an evaluation of how well
ozone measurements can serve as a surrogate for the
variations in the concentrations of other products of
atmospheric photochemical reactions in the ambient air.
This work was done as a result of a request from staff in the
Off ice of Air Quality Planning and Standards to the Off ice of
Research and Development. This assessment is one of a
group of four assessments requested concerned with the
following topics: (1) natural sources of ozone, (2) transport
and reactions of ozone, and (3) products of photochemical
atmospheric reactions.
A scientific review has been prepared considering in detail
the other products of atmospheric photochemical reactions
as they relate to ozone. The mechanisms of formation, types
of products identified in laboratory studies, and the ambient
air concentrations and distributions are discussed in this
review. Because nitrogen dioxide has received much
scientific consideration and assessment elsewhere, nitro-
gen dioxide is not considered in detail within the review.
Qualitative Summary of Results of Studies
A substantial number of inorganic and organic products of
photochemical atmospheric reactions have been identified
and measured in laboratory studies. Many of these
products are gases or vapors, while others occur as
aerosols in the atmosphere. While several of these
products along with ozone are "oxidants," most of the
products do not respond as oxidants to the measurement
methods used in the laboratory and ambient air.
More products have been identified in laboratory investi-
gations than in the ambient air. Therefore, the lack of
identification in the ambient atmosphere of products
observed in the laboratory does not necessarily imply that
these products are not present in ambient air. Even in
-------
laboratory studies carbon and nitrogen balances between
reactants and products often are difficult to obtain. There-
fore, additional products not yet identified even in the
laboratory may be present in the ambient air.
The inorganic products of atmospheric photochemical
reactions identified in laboratory and/or ambient air
include carbon monoxide, carbon dioxide, nitrogen dioxide,
nitrogen trioxide, nitrous acid, nitric acid and hydrogen
peroxide. Most of the carbon monoxide and carbon dioxide
in ambient air is emitted by combustion sources rather than
by formation in atmospheric photochemical reactions.
Sulfate and nitrate aerosols in ambient air also are formed
in atmospheric photochemical reactions.
The organic vapors measured in the ambient air include a
number of aldehydes including but not limited to formal-
dehyde, acetaldehyde and acrolem, along with methyl ethyl
ketone, formic acid, peroxyacetyl nitrate and peroxy-
propionyl nitrate. A number of higher molecular weight
complex organic substances have been identified in
atmospheric aerosols. Based on comparisons with labora-
tory investigations, all of these substances are reasonably
assigned as products of atmospheric photochemical reac-
tions.
In the previous discussion, the assumption was made that
products measured in simple systems in the laboratory also
should be measurable in more complex laboratory systems
and in the ambient air. Within the scientific review, the
comparisons made do indicate satisfactory agreement in
most instances between predicted and measured yields in
experiments involving multicomponent hydrocarbon-nitro-
gen oxide systems. A number of products measurable as
major products in individual simple and multicomponent
hydrocarbon-nitrogen oxide laboratory systems also have
been measured as products occurring at substantial
loadings in the ambient atmosphere.
Finally, it should be noted that the biological effects of the
products of atmospheric photochemical reactions are
mentioned briefly but are not discussed in detail within the
review. This aspect was not included as part of the effort
requested.
Discussion of Issues
Three issues have been identified for consideration in this
assessment. To provide answers to these issues requires
use of both laboratory and ambient air measurements,
because the ambient air measurements alone either are
not complete or cannot be adequately interpreted at the
present time.
Issue 1. Are the other products of atmospheric photo-
chemical reactions, aside from nitrogen dioxide, present
singly or in combination at significant ambient air
concentrations compared to ozone?
As will be discussed in more detail below, several products
do occur in the atmosphere at concentrations equal to 5,10
or more percent of the ozone concentrations. These
products are considered to occur at significant concentra-
tion levels compared to ozone. However, there also are a
considerable number of products which occur individually
at a few percent or less of the ozone concentration. Because
of the large number of such products when summed they
add up to significant atmospheric concentration levels.
The largest group of appropriate air quality measurements
for making this evaluation has been obtained in California's
South Coast Air Basin. The measurements made by long
path Fourier transform infrared spectroscopy (LP-FTIR) are
particularly useful. Such measurements were made at sites
in Riverside, CA, Claremont, CA and Los Angeles, CA,
during periods in 1977 through 1980. The LP-FTIR measure-
ments provide ambient air concentrations of several
products concurrently. This technique avoids possible
sampling difficulties because of the open atmospheric path
used.
Of the twenty-eight days of measurement available, twenty
days were selected for the computations made because of
the completeness and duration of the measurements
obtained on those days. Concurrent measurements usually
were available for ozone, formaldehyde, nitric acid, peroxy-
acetyl nitrate and formic acid. The lengths of the periods of
measurement on these twenty days varied from six to
twenty-three hours.
The percentages of other products in the ozone concentra-
tions at the time of maximum ozone concentration will be
considered first. Averaging the concentrations of each
product at the time of maximum ozone concentration over
the twenty days, the percentages of each of the other
products in the ozone are as follows: formaldehyde, 11;
nitric acid, 6; peroxyacetyl nitrate, 5; formic acid, 2. During
several other air quality studies carried out in the South
Coast Air Basin, formaldehyde, total aldehydes, or other
individual aldehydes were measured. In these studies, the
sum of the other aldehydes on the average about equalled
the formaldehyde concentration. In some of the earlier
studies made during the 1950's and 1960's, the total
aldehyde concentrations were reported to equal or exceed
the ozone concentrations.
Assuming that the sum of other aldehydes did equal the
formaldehyde concentration during the LP-FTIR measure-
ments, the total aldehydes would be about twenty-two
percent of the ozone concentration during the more recent
measurement period. This assumption is supported by the
results of an independent study at one of the same sites
within the 1977 to 1980 period.
Sulfates, nitrates, and secondary organic carbon aerosols
also have been measured along with ozone in several other
air quality studies within communities in the South Coast
-------
Air Basin carried out between 1973 and 1979. Based on
these results, the percentages of these aerosol products in
the ozone concentrations at the time of maximum ozone
concentrations are as follows: sulfates, 9; nitrates, 5;
organic carbon, 4.
If these percentages for formaldehyde, other aldehydes,
nitric acid, peroxyacetyl nitrate, formic acid, sulfates,
particulate nitrates, and organic carbon aerosols are
totalled, their sum is over half of the ozone concentration. If
the earlier results for aldehydes were valid, these other
products may have totalled a much larger percentage of the
ozone concentration in earlier years.
The other products have diurnal concentration profiles
differing from that of ozone. Therefore, the relationships
between the other products and ozone on longer averaging
time bases, such as 24-hour averages, will be different than
at the time of maximum ozone concentration.
To explore the longer averaging time relationships, nine
days were selected from the LP-FTIR measurement periods
in 1977 to 1980. The products were measured for from ten
to twenty-three hours a day during these nine days.
Averaging the concentrations of each product over all of
these measurement periods, the percentage each of the
other products were of the ozone are as follows: formal-
dehyde, 31; nitric acid, 15; peroxyacetyl nitrate, 6; formic
acid, 4. Again, the sum of other aldehydes will be taken as
equal to formaldehyde based on other air qualities studies.
The same studies of aerosol composition in the South Coast
Air Basin discussed earlier also can be used to compute the
percentages of the aerosol products in the ozone concentra-
tion on a 24-hour average. These percentages are as
follows: sulfates, 15; nitrates, 11; organic carbon, 5. If these
percentages, the formaldehyde, other aldehydes, nitric
acid, peroxyacetyl nitrate, formic acid, sulfates, particulate
nitrates and secondary organic aerosols are totalled, their
sum exceeds the ozone concentration. There is about a
two-to-one ratio between the percentage which these
other products add up to on a longer time basis compared to
the percent they sum to at the time of maximum ozone
concentration.
In addition to the products discussed above, a number of
other gaseous products have been measured in the South
Coast Air Basin. However, the measurements for these
products are not adequate to estimate their percentages of
the ozone concentrations.
Similar air quality measurements are available for some of
the products measured at urban locations elsewhere in the
United States. The results are less complete than in the
South Coast Air Basin. The only approach that appears
feasible is to combine the measurements made at a number
of locations such as in Houston, TX; St. Louis, MO;
Columbus, OH; Atlanta, GA; and Hoboken, NJ. Even after
combining results, the measurements are inadequate to
estimate percentages at the time of maximum ozone
concentration, but they can be used to obtain a diurnal
basis. Based on this approach, the percentages the other
products were of the ozone concentration on a 24-hour
average basis are as follows: formaldehyde, 40; other
aldehydes, 38; peroxyacetyl nitrate, 5; sulfate, 15. Sincethe
percentages of these products sum to 98 percent of the
ozone concentration, it seems reasonable to expect that the
sum of all of the other products would somewhat exceed
the ozone concentration on a 24-hour average.
Even fewer measurements are available at nonurban sites
in the United States. Excepting sulfates, the concentrations
of other products appear to be lower relative to ozone at
nonurban than at urban locations.
The results discussed above indicate that the other
products of atmospheric photochemical reactions can
occur at significant concentration levels individually and at
substantial concentrations in total compared to ozone.
Additional health and welfare research results are needed
to adequately evaluate the significance of these air quality
results to criteria and standards development.
Issue 2. Can ozone through its air quality measure-
ments be used as a surrogate for the other products
observed in the ambient atmosphere?
It is important to decide what the term "surrogate" means
in the context of this issue. Two possible viewpoints are
suggested here. Firstly, the term surrogate can be inter-
preted to mean that other products display the same
directional characteristics as ozone. Secondly, the term
surrogate can be interpreted to mean that unique quanti-
tative relationships exist between ozone and other prod-
ucts. Both possibilities will be considered here. Two
approaches are available. One of these examines the
relationships that occur in laboratory studies between
hydrocarbons and nitrogen oxides precursors and the
photochemical reaction products. The second approach
examines the possible relationships indicated by concur-
rent measurements in the ambient atmosphere of ozone
compared to other products.
The laboratory results used are from three different smog
chamber irradiation studies. Measurements are available
from these studies for oxidant or ozone along with
peroxyacetyl nitrate and/or formaldehyde. The results
obtained were expressed either as concentrations of
products or dosages (concentration x irradiation time) of
products.
The experimental designs were such that three precursor-
product conditions were investigated: (1) the initial hydro-
carbon was varied (from experiment to experiment) over a
range of concentrations with the initial nitrogen oxide at a
fixed concentration, (2) both the initial hydrocarbon and the
nitrogen oxide concentrations were varied but with a
constant ratio of these precursors, and (3) the initial
nitrogen oxide was varied over a range of concentrations
with the initial hydrocarbon at a fixed concentration.
-------
The conclusions that can be obtained from the experi-
mental results relevant to the three conditions listed above
are as follows:
Condition 1. The concentrations and the dosages of
products vary in the same direction with changes in the
initial hydrocarbon concentration. The amounts of variation
of any one product with the changes in initial hydrocarbon
concentration are different from any other product.
Condition 2. The concentrations and dosages of products
under many experimental circumstances vary in the same
direction with the concurrent changes in initial hydro-
carbon and nitrogen oxide concentrations. At lower pre-
cursor ratios, the concentrations and dosages of other
products can decrease while the oxidant increases. The
amounts of variation of any one product with the con-
current changes in initial hydrocarbon and nitrogen oxide
concentration are different from any other product.
Condition 3. The concentrations and dosages of products
under many experimental circumstances vary in the same
direction with the changes in initial nitrogen oxide concen-
trations. In two of the three studies, the concentrations of
one of the other products did not vary in the same direction
as the oxidant or ozone with the changes in initial nitrogen
oxide concentrations. The amounts of variation of any one
product with the changes in initial nitrogen oxide concen-
tration are different from any other product.
These laboratory results indicate that other products under
many, but not all experimental circumstances, will vary in
the same direction as the oxidant or ozone with changes in
the concentration of precursors. The quantitative variations
in other products are not the same as for oxidant or ozone
with changes in the concentration of precursors.
In the second approach used, the ambient air relationships
were examined between ozone and several other products.
The measurements available by long path Fourier trans-
form infrared spectroscopy at several locations in the Los
Angeles Basin were used. In these ambient studies, pre-
cursor concentrations are only occasionally reported. Since
neither precursor emissions nor ambient air concentra-
tions upwind are reported in these studies, it is not possible
to relate products to precursors from these studies.
However, ozone concentration distributions can be com-
pared with those for peroxyacetyl nitrate, nitric acid, formalde-
hyde and formic acid.
The results for maximum diurnal ozone concentrations
were plotted relative to the concentrations at the same
times for each of the other products. Results obtained at
University of California, Riverside, CA in 1976 and 1977; at
Harvey Mudd College, Claremont, CA in 1978 and 1979;
and at California State College, Los Angeles, CA in 1980
were used. The conclusions which can be arrived at by
examining such plots are as follows:
(a) The concentrations of other products show the same
directional characteristics as the ozone concentration.
(b) These relationships are not linear. Significant curva-
tures are apparent in the plots, particularly at the
higher concentrations measured. Corresponding to a
fifty percent reduction in the concentration of ozone
from its highest measured diurnal maximum concen-
tration, the concentrations of the other products
decreased by more than 50 percent.
(c) There is considerable scatter in the measurement
values of each of the other products at a particular
ozone concentration. For example, in the 270-280 ppb
range of concentration for ozone, the formaldehyde
concentrations ranged from 12 ppb to 31 ppb.
(d) The scatter in measured values is not random. Based
on the ozone-formaldehyde plot in the 200-300 ppb
ozone range, the lowest concentration values of
formaldehyde were reported at Riverside, CA in 1 976
while those in Claremont, CA in 1978 are consistently
high with the results of the other two studies at these
locations falling in between these studies. Similar site-
specific effects can be identified in the other plots.
(e) The amount of scatter varied from plot-to-plot. The
results for the ozone-peroxyacetyl nitrate plot show the
least scatter. The variability observed is likely to be due
to a combination of factors including differing effects of
transport, varying contributions of local primary emis-
sions of products to the total loading of products, and
differences in precursor emission densities.
These ambient air results suggest that at least in the South
Coast Air Basin ozone serves as a surrogate directionally for
other products. Adequate measurements are not available
in other geographical areas relating ozone to a number of
other products.
Issue 3. Is it possible to conduct ambient air monitoring
for photochemical reaction products other than ozone?
Satisfactory "routine" monitoring techniques for other
products are needed if a separate standard should be
considered for products other than ozone.
Most of the results available are from research studies of
short duration. However, several monitoring efforts have
been carried out for other products. Both the Los Angeles
Air Pollution Control District during the 1950's, and the
state of New Jersey during the 1970's, monitored alde-
hydes or formaldehyde. Automated gas chromatographic
equipment has been utilized in Riverside, CA to obtain
several years of peroxyacetyl nitrate measurements. There-
fore, measurements for several products appear to be
possible on a routine basis. The available results indicate
that it should be possible, if desired, to conduct a monitoring
effort at least in a limited number of critical urban areas.
Conclusions and Recommendations
1. Other gaseous and aerosol products of atmospheric
photochemical reactions in combination occur at
substantial concentration levels in the ambient air
compared to ozone in urban areas. On a 24-hour
average basis, these other products in combination can
equal or exceed the ozone concentration. At the time of
maximum ozone concentrations, the other products
can constitute in combination about half of the ozone
concentration. Based on incomplete carbon and nitro-
gen balances obtained in laboratory measurements,
-------
additional laboratory and field studies are needed to
complete the identification of products.
2. Ozone may serve as a satisfactory surrogate direc-
tionally for other products in the ambient atmosphere.
Although the other products show similar qualitative
variations in the atmosphere as ozone, quantitative
relationships are much more complex. Laboratory
studies relating oxidant, formaldehyde and peroxy-
acetyl nitrate to precursor concentrations indicate
somewhat different relationships for each product.
3. In viewof concern from the air-quality viewpoint for the
other products, additional clinical and epidemiological
health studies are needed. If any of the other products
should prove to have a dose-response curve com-
parable to ozone, that other product might be included
in a standard by either developing an "equivalent"
ozone standard or estimating a margin of safety. By
"equivalent" ozone standard is meant a standard in
which the impact of the concentrations of other
products would be adjusted by equating them to
particular concentrations of ozone.
4. It appears practical to carry out monitoring efforts for
other products than ozone, if necessary, in a limited
number of critical urban areas.
. S. GOVERNMENT PRINTING OFFICE: 1983/659-095/0726
------- |