United States
Environmental Protection
Agency
Atmospheric Sciences
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S3-85/017 Apr. 1985
Project Summary
Chemical Transformations in
Acid Rain: Volume II.
Investigation of Kinetics and
Mechanism of Aqueous-Phase
Peroxide Formation
Yin-Nan Lee
The aqueous-phase reaction kinetics
of dissolved O, with a number of at-
mospheric components was in-
vestigated. Special attention was
focused on the formation of H2O2 or
organic peroxide as reaction products.
The rate constants and peroxide yields
(y) determined for the specified
substrates were as follows:
H2O: k = 2.1 x 10 '4 s'1
(pH ~ 6), y s 0.5%
HjO,: k = 2.6 x 103 M'1 s'1
(pH ~ 6)
HCOjH: k = 4.3 x 10» M'1 f\
y < 0.5%
HjCO: k = 1.2 x 10'1 r\
y < 2%
C2H4: k = 3.0 x 10* M'1 s'1
PAN: k < 3 x 10> M-1 SM
NO2: k HNO,
atm'1 s
4 x 10-» M
NO2 + HCO2: k HNO, £ 6 atnr1 s"1
Using these data, the rates of aqueous-
phase peroxide production for these
reactions under typical atmospheric
conditions were calculated to be ~ 1 x
10"' M h~' or smaller. It was therefore
concluded that the reactions studied in
this work contribute insignificantly to
the formation of peroxides in at-
mospheric water.
This Project Summary was
developed by EPA's Atmospheric
Sciences Research Laboratory,
Research Triangle Park, NC, to an-
nounce key findings of the research
project that is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
It has been recognized that the chemical
reactions that produce strong acids can
take place either in the gas phase or in the
liquid phase. This notion was established
because convincing evidence had been col-
lected to indicate that the atmospheric ox-
idation of S02 was strongly affected by
aqueous-phase reactions, especially the
reaction of ozone with hydrogen peroxide.
In order to assess the importance of these
aqueous reactions, the atmospheric con-
centrations of 03 and H202 have to be
determined. Although the gas-phase con-
centration of ozone can be accurately
determined by various techniques such as
ethylene-chemiluminescence, no viable
method is currently available for the
measurement of gas-phase concentrations.
of H202. As a result, the major gas-phase
routes for peroxide generation, i.e., the
recombination of hydroperoxy radicals and
the photolysis of formaldehyde, cannot be
confidently employed in a numerical model
as the sole source for this species. Further-
more, recent attempts to determine the
gas-phase concentrations of peroxide using
bubbler series have revealed the existence
of in-situ production of artifact peroxide.
-------
This observation suggests that aqueous
pathways for peroxide formation might ex-
ist. Clearly, these pathways have to be
identified and characterized before it is
possible to accurately model the at-
mospheric budget of H202 and the rate of
S02 oxidation.
Among the various potential precursors
of aqueous peroxide, 03 might be a plausi-
ble candidate based on the following con-
siderations: (1) 03 produces peroxides upon
reaction with certain organic compounds
such as olefins, (2) aqueous-phase 03 reac-
tions involve free radicals derived from 03
decomposition that might serve as H202
precursors, and (3) bubbler series ex-
periments demonstrated that the levels of
artifact H202 do not diminish rapidly along
the bubbler train; this is consistent with the
presumption that the precursor species
might be present in relatively high concen-
tration and have a low aqueous solubility.
03 appears to fit the description.
Although the aqueous-phase reactions of
ozone have been the subject of numerous
studies for the past several decades, major
gaps exist in the understanding of the
detailed features of these reactions. No ma-
jor efforts have been directed to product
analysis. In this current laboratory research,
we have examined a series of aqueous-
phase reactions involving 03 and monitored
the formation of H202 and organic perox-
ides in an attempt to identify the direct
aqueous sources of peroxides. The
aqueous-phase reaction systems examined
included (1) 03 self-decomposition, and (2)
O3 reaction with formaldehyde (with and
without the presence of N02), formic acid,
ethylene, and peroxyacetyl nitrate (PAN).
Except for the 03-ethylene reaction, perox-
ide was not found as a reaction product.
Experimental
Two techniques were used to generate
ozone. For higher O3 output (up to - 200
ppm at 1 f/min flow rate) a 10-inch Pen-Ray
UV lamp was employed. This source was
used mainly for the preparation of saturated
03 solutions for batch-type reactions. For
lower 03 output, an AID Ozone Generator
(Model 565) equipped with a continuously
adjustable shutter for 03 output control
was employed (up to ~ 1 ppm at 1 l/min
flow). This source was used for the contin-
uous-flow reaction system. For both
generators, high purity 02 purchased from
Matheson was used in order to minimize
possible interferences from NOX and
organic impurities.
Gas-phase 03 concentrations were deter-
mined either by a Dasibi 03 monitor (Model
1008-PC) based on UV absorption or by a
Monitor Labs Ozone Analyzer (Model 8410)
based on the 03-ethylene chemilumi-
nescence; the Dasibi monitor was the
primary standard.
The concentration of aqueous-phase 03
was measured by a UV-vis spectropho-
tometer (Beckman Model DU-7) using
either a 10 cm or a 5 cm cylindrical optical
cell. Using the 10-inch Pen-Ray UV
ozonator at 02 flow rate of 70-90 cc
min"1, a typical aqueous-phase 03 concen-
tration of 1-2 x 10"5 M was obtained
after 40 min of bubbling. The limit of
detection for 03 with the 10-cm cell was 1
x 10-' M.
For batch-type experiments, 03 solu-
tions were prepared in a 2-1 bubbler
through which 03 was continuously bub-
bled at a total 02 flow of 70 cc min"1.
Each Pyrex 03 bubbler (volume ~ 2 0
was equipped with a coarse-sized frit for
the enhancement of mixing and with two
ports for liquid transfer. The plumbing
was constructed with parts made either of
stainless steel or Teflon for purity. One
bubbler was needed for the humidification
of the gas stream and the removal of any
soluble substances; a second bubbler sup-
plied saturated 03 solutions to be used for
the batch studies. For the study of some
continuous-flow reactions, a valve was
switched so that O3 would flow through
the gas-liquid reaction cell to initiate the
reactions.
Concentrations of aqueous-phase H202
and organic peroxides were determined by
a horseradish peroxidase-fluorescence
technique (HRPF). In our arrangement a
Perkin-Elmer fluorometer (Model 204S)
was employed in conjunction with a liquid
flow reaction system equipped with a
rotary injection valve (Altex, sample loop
size 0.5 m!). The limit of detection of the
HRPF technique was 1 x 10~7 M. Since
this technique did not distinguish organic
peroxides from inorganic H2O2, the deter-
mination of the concentration of organic
peroxides was achieved by a difference
method in which H202 was preferentially
destroyed (or inactivated) by the enzyme
catalase.
Two different types of kinetic methods
were used in this study. In the batch-type
method, reactions were initiated by mix-
ing the reagents with 03 solution in an
optical cell and the decrease of [03] ac-
companying the reactions was followed
spectrophotometrically at X = 260 nm. In
the continuous-flow method, a reagent
gas mixture containing constant concen-
trations of 03 and other gaseous reactants
was continuously bubbled through a solu-
tion contained in bulk-type gas-liquid
reactor. The kinetics of peroxide genera-
tion were followed by an aliquot method
in which peroxide concentrations were
determined by the HRPF technique.
Temperature of the reaction vessel was
maintained at 22.0 ± 0.1 °C for the latter
method, but for the spectrophotometric
technique the uncertainty was ± 2°C.
Approach
The peroxide production rate of an
aqueous-phase reaction of 03 can be ex-
pressed in a simplified fashion as
R(a) = -d[03]/dt = d[H202]/dt = (1)
k [03] [X],
where [03] and [X] are the aqueous con-
centrations of the reactants and k is the
effective second-order rate constant.
When such a reaction takes place in
cloudwater droplet, the mass transfer
rates pertinent to the small droplet sizes
may be sufficiently fast to allow the
establishment of gas-liquid equilibrium
with respect to the solutes. When this
happens, Eq. (1) can be rewritten as
R(a) = k Ho, Hx po, Px, (2)
where H is the Henry's law solubility and
p is the partial pressure. Given the values
of k and H, the rate of peroxide formation
in the cloud water can be calculated for
given reagent concentrations. The
aqueous-phase reaction rate can be
related to the gas-phase rate by
R(g) = R(a) LRT, (3)
where L is the liquid water concentration
of the space volume, R is the gas con-
stant and T is the absolute temperature.
The characteristic reaction time of, for ex-
ample, gas-phase 03 against the liquid-
phase reaction can then be calculated
from
TO, = [R(a) LRT/po,)-' = (4)
(k Ho, [H202] LRT)-1.
Equations (2) and (4) were used in this
work when appropriate to assess the at-
mospheric importance of a particular reac-
tion with respect to aqueous-phase perox-
ide production and gas-phase reagent
consumption.
Results
l
O3 Decomposition in
Optical Cell
Ozone in a 5-cm cylindrical cell under
neutral pH (no acid or base added) was
found to decay with a first-order rate con-
stant of 2.1 x 10'4 s'1. When the pH of (
-------
the solution was adjusted to 2 with HCI,
the rate constant of 03 decomposition
dropped to 1.4 x 10~4 s"1. The final
decomposition mixture of the 03 solutions
(containing initial [03] as high as 2.0 x
10~8 Mlwas analyzed for H202. The con-
centration of H202 was found to be below
the limit of detection of 1 x 10"' M. The
reaction time constant of 03 against
aqueous-phase decomposition was
estimated to be 5 x 10* h or longer for
solution of neutral or lower pH. Since the
H202 yield of this reaction was estimated
to be smaller than 5%, a maximum rate
of H202 production in atmospheric water
was calculated to be - 1 x 10"9 M fr1 at
Po, = 100 ppb.
Ot-HtOt Reaction
The kinetics of this reaction was
studied under pseudo-first-order condi-
tions, i.e., [H202]0 » [03]0. The
second-order rate constant, obtained by
dividing the pseudo-first-order rate con-
stants by [H202]0, which had been varied
from 8 x lO"8 M to 6.4 x 10"6 M, was
determined to be (2.6 ± 0.4) x 103 M'1
s'1 at neutral pH. The good agreement
obtained between the second-order rate
constants determined at widely different
H202 concentrations permitted the conclu-
sion that the reaction kinetics were also
first order with respect to [H202]. At
lower pH the reaction rate decreased
rapidly. A second-order rate constant of
38 ± 8 M'1 s"1 was determined at pH =
4.
The time constant for removal of at-
mospheric gas-phase 03 by the aqueous
03-H202 reaction is given by reaction (4).
For [H202] = 3 x 10~6 M, a represen-
tative summer cloudwater H202 concen-
tration and L = 10~6, TO, was calculated
to be 1.4 x 104 h and 1.0 x 10" h at
neutral pH and pH 4, respectively. Since
the aqueous-phase 03 concentration in at-
mospheric water is significantly smaller
than that of H2O2, the removal of the
dissolved atmospheric H2O2 by the
aqueous-phase 03-H202 reaction will be
extremely slow.
O3-HCOtH Reaction
The kinetic study of the reaction,
03 + HC02H — Products,
(5)
was made by monitoring the change of
103] spectrophotometrically at X = 260
nm. Since the rate of this reaction was
too fast to be studied under pseudo-first-
order conditions, initial concentrations of
p3 and HC02H were made approximately
iual at 2 x 10~5 M. Treating the reaction
with an overall second-order kinetics, the
rate constant was determined by
1/IOJ, - 1/[03]0 = k't (6)
Plots of [03]f' vs. time were found to be
linear for at least three half-lives. The rate
constants varied as a function of pH
where k' was (2.5 ± 0.2) x 102 M'1 s'1 at
pH = 3 and (3.3 ± 0.5) x 103 IvT1 s'1 at
pH = 4.6. The pH dependence of k'
could be fitted to a rate law that assumed
the rate determining step involved 03 and
the dissociated formate ion.
Rate = k [03][HCOr]
= k ( K. ) [03] [HC02H]T
[H+] + K8
= k' [03][HC02H]T,
(7)
where Ka is the acid dissociation constant
of formic acid and [HC02H]T is the total
analytical concentration of formic acid.
Fining Eq. (7) to the experimental data
allowed the values of k and Ka to be
determined; they were found to be 4.3 x
103 M'1 s-1 and 8.9 x 10"6 M'1, respec-
tively.
Final reaction mixtures of the
03-HC02H reaction were analyzed for
peroxide by the HRPF method. The level
of peroxide was found to be smaller than
the LOD of the instrument, namely, 1 x
10"7 M. Under the present reaction condi-
tions, e.g., [03]0 = [HC02H]0 = 2 x 10-*
M, a value of 0.5% was estimated as the
upper limit for the peroxide yield of the
03-HC02H reaction.
The rate of depletion of 03 and HC02H
from the gas-phase due to the aqueous
O3-HC02H reaction was also determined.
The depletion constants were calculated
to be 2.4 x 106 h and 4.8 x 104 h for 03
and HC02H, respectively. These values
are clearly too long to make this reaction
significant.
O3-H2CO Reaction
The kinetics, under conditions where in-
itial [H2CO] was in large excess of [03],
did not conform to a pseudo-first order
reaction. In fact, the effective second-
order rate constant appeared to increase
as [03] was decreasing, indicating a reac-
tion order of less than unity with respect
to 03. To determine the reaction order
with respect to each of the reagent con-
centrations, the following equation was
used:
Rate = - d[03]/dt = k[03]n[H2CO]m (8)
Under conditions where [H2CO]0 > >
[03]0, Eq. (8) can be reduced to
log (Rate/[H2CO]m) = log k (9)
+ n log [O3].
Fitting Eq. (9) to the experimental data
yielded a value for the slope n of - '/4.
Using the same approach, the analogous
equation,
log (Rateo/tO,!14) = log k (10)
+ m log
[H2CO]0,
was obtained. To determine m, the initial
rate was measured for three initial for-
maldehyde concentrations and Eq. (10)
was fit to the rate data. The value of m
was found to be close to % also. Eq. (8)
was then tentatively identified as
Rate = k[0,]* [H2C01* (11)
withk = (1.15±p.7)x1(TiVt.
Product analysis of the final reaction
mixtures showed that H202 was a minor
product. With [H2CO]0 = 8 x 1Q-5 M and
[03]o = 8 x 10'6 M, H202 at the end of
the reaction was found to be 1.5 x 10~7
M, corresponding to an H202 yield of
about 2%. The reaction time constants of
atmospheric 03 and H2CO for this reaction
were found to be 6 x 10* hours or
longer. Again, it may be concluded that
the aqueous-phase reaction of 03 and
H2CO has little effect on the atmospheric
life times of these species.
O3-C2H4 Reaction
Due to the low solubility of C2H4 at
25°C, the kinetics of this reaction was
studied using the continuous-flow
method. N2 was first allowed to flow
through the gas-liquid reaction cell con-
taining a known volume of liquid water to
remove the dissolved C02. When the con-
ductivity of the liquid water had stabilized,
C2H4 was added to the gas stream; the
concentration of C2H4 employed was 7-28
ppm. After the reaction was initiated by
the addition of 0.5 to 1.0 ppm 03, the
reaction mixture was analyzed for perox-
ide concentration at known time intervals.
The concentration of peroxide in the reac-
tion mixture was found to increase linearly
with time. Assuming that the aqueous-
phase reaction of 03 and C2H4 is first
order with respect to each reagent, perox-
ide formation can be expressed as
dtperoxidel/dt = kH0, HC,H. (12)
p03 pC2H4.
The values of k determined at various
reagent concentrations and pHs were
essentially the same; the average rate
constant over the pH range 3 to 7 was
found to be (3.0 ± 0.3) x 10" M-' s'\
-------
The fact that the rate constant remained
essentially unchanged even though p03
and pC2H4 were varied by a factor of 2
and 4, respectively, lends support to the
assumed overall second-order kinetics.
The fraction of organic peroxide formed
in the C2H4-03 reaction was determined
using an enzyme catalase technique in
which the H202 was preferentially
destroyed. The results showed that
>60% of the total peroxide was present
as organic peroxide (possibly CH302H),
which appeared to be reasonably stable.
The rate of depletion of 03 and C2H4
from the gas phase due to the 03-C2H4
reaction was estimated to be of the order
of 1 x 107 hours. These long reaction
time constants would have minor conse-
quences on the atmospheric residence
times of either 03 or C2H4. The aqueous-
phase rate of peroxide production from
this reaction was estimated to be only 6
x 10-" M Ir1 for p03 = 50 ppb and
pC2H4 = 20 ppb.
Aqueous-Phase Reactions
of PAN
Three reactions of PAN were examined:
ka
PAN - peroxide, (13)
kb
PAN + 02 - peroxide, (14)
PAN + 03 — peroxide.
(15)
The typical concentrations used for PAN,
02, and 03 were 100 ppb, 20% and 1.0
ppm, respectively. Due to the low
solubilities of all three species, these reac-
tions were again studied by the use of the
gas-liquid reaction cell described above.
Final reaction mixtures (with total reaction
times up to 1 h) were analyzed for perox-
ide. It was found that the levels of perox-
ide were all below the detection limit of
the HRPF technique. Using a general rate
expression
d[peroxide]/dt = k [PAN] [Ox], (16)
the upper limits for rate constants ka, kb,
and kc were estimated to be 2 x 10~* s~1,
3 x 102 M-1 s-', and 3 x 103 M'1 s'\
respectively. For typical ambient condi-
tions, the upper limits of the rate of
peroxide formation were estimated to be 4
x 10-", 1 x 10-", and 4 x 1Q-11 M h'1 for
PAN hydrolysis, O2-PAN, and 03-PAN
reactions, respectively. The atmospheric
residence times of PAN for the three reac-
tions were estimated to be 2 x 10*, 1 x
102, and 7 x 10' h, respectively. Since
these time constants are all significantly
longer than typical cloud lifetimes of ~1
h and are also longer than that for the
gas-phase PAN-NO reaction, it is conclud-
ed that the atmospheric lifetime of PAN is
not affected by these reactions.
O3-/VO2 Reaction
To conduct the N02 experiment, a 1-f
pre-reaction chamber was placed
upstream of the bubbler to allow the pro-
duction of N03 from the 03-N02 gas-
phase reaction. The experiments were
carried out in the dark in order to avoid
the photolysis of N03. After 40 min, the
reaction mixture in the gas-liquid reactor
was analyzed for H202. The concentration
of H202 was found to be ~1 x 10~7 M.
An upper limit rate constant for the
reaction
N03 + H20 - N02 + H202 (17)
can be estimated by equating the as-
sumed rate expression.
Rate = d[H202]/dt = k [N03], (18)
to the maximum possible rate estimated
from the limit of detection of the HRPF
method; that is, k[N03] < 1.2 x 1Q-10 M
s"1. At pN03 < 3 ppb, the atmospheric
contribution of reaction (17) to aqueous
peroxide production was estimated to be
less than 4 x 1(T8 M Ir1.
O3-NOt-HtCO Reaction
In our approach, N02 and 03 in N2 were
allowed to react in a pre-reaction chamber
for -30 s before entering the gas-liquid
bubbler that contained the H2CO solution.
The reaction extent of the formaldehyde
solution was followed by monitoring
peroxide concentration. Despite the long
reaction time (up to 100 min), no discern-
ible peroxide was detected as a product.
An upper limit to the rate of the
aqueous-phase reaction,
k
N03 + H2CO - Peroxide, (19)
was estimated from
d[Peroxide]/dt = k HNO, pN03 (20)
[H2COL
The product HNo, was calculated to be
smaller than 6 atnrr1 s'1. For the condition
where pN03 = 3 ppb and pH2CO = 20
ppb, the rate of peroxide formation in the
aqueous-phase was estimated to be 8 x
10~" M h"1. Again, this rate is too slow to
be of importance.
Summary
Aqueous-phase reactions of 03 with a
number of important atmospheric com-
ponents were examined. Reaction kinetics
and product analysis for hydrogen perox-
ide and organic peroxides were de-
termined for the following reaction
systems: (1) 03 - H20, (2) 03 - H202, (3)
03 - HC02H, (4) 03 - H2CO, (5) 03 - C2H4,
(6) 03 - PAN, (7) 03 - N02, and (8) 03
-N02 - H2CO. Calculations were made to
estimate the atmospheric importance of
these reactions to the removal of 03 and
the reagent species, as well as the pro-
duction of peroxides in the liquid phase.
For typical atmospheric conditions, it was
concluded that none of these aqueous-
phase reactions is important as a source
of aqueous-phase peroxides or a sink for
gaseous 03 and its corresponding
reagents.
-------
Yin-Nan Lee is with Brookhaven National Laboratory, Upton. NY 11973.
Marcia C. Dodge is the EPA Project Officer (see below).
The complete report, entitled "Chemical Transformations in A cidRain: Volume II.
Investigation of Kinetics and Mechanism of Aqueous-Phase Peroxide Forma-
tion," (Order No. PB 85-173 433/AS; Cost: $10.00, subject to change) will be
available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Atmospheric Sciences Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
<*U.S.Government Printing Office: 1985 — 559-111/10820
-------
United States Center for Environmental Research
Environmental Protection Information
Agency Cincinnati OH 45268
Official Business
Penalty for Private Use $300
OCOC329 PS
------- |