UC Irvine

Faculty Publications

Title
Interaction of ozone and hydrogen peroxide in water: Implications for analysis of H
2
O
2
in air

Permalink
https://escholarship.org/uc/item/7j7454z7

Journal
Geophysical Research Letters, 9(3)

ISSN
00948276

Authors
Zika, R. G
Saltzman, E. S

Publication Date
1982-03-01

DOI
10.1029/GL009i003p00231

License
https://creativecommons.org/licenses/by/4.0/ 4.0

Peer reviewed

eScholarship.org Powered by the California Digital Library

University of California
 GEOPHYSICLA
RESEARCH
LETTERS,VOL. 9, NO. 3, PAGES
231-234 , MARCH
1982

INTERACTION OF OZONE AND HYDROGEN PEROXIDE IN WATER:

IMPLICATIONS
FORANALYSIS
oFH202INAIR
R.G. Zika and E.S. Saltzman

Division of Marine and Atmospheric Chemistry, University of Miami,

Miami, Florida 331#9

Abstract. We have attempted to measure gaseous Analytical Methods

H202 in air usingan aqueoustrappingmethod. With
continuousbubbling,H20 2 levels in the traps reacheda a. Hydrogen Peroxide. Hydrogen peroxide in aqueous
plateau, indicating that a state of dynamic equilibrium solution was measured using a modified fluorescence
involving H202 destrbction was established. We decay technique [Perschke and Broda, 1976; Zika and
attribute this behavior to the interaction of ozone and its Zelmer, 1982]. The method involved the addition of a
decomposition products (OH,O[) withH202inacld:•ous known amot•at of scopoletin (6-methyl-7-hydroxyl-i,2-
solution. This hypothesis was investigated by replacing benzopyrone)to a pH 7.0 phosphatebuffered. sample.
the air stream with a mixture of N2, 02 and 0 3. The The sample was prepared bY diluting an aliquot of the
results Of this experiment show that H O was both reaction solution to 20 mls with low contaminant
producedand destroyedin the traps. Theseresultshave distilledwater. Thesizeof the reactionsolutionaliqu0t
led us to question the validity of techniques which wassuchthattheconcentr.•tion
of H202øfinhorseradish
samplewaslessthan2x 10- M. Addition
the20ml
employaqueoustraps to measureH 20 2 in air.
peroxidase
(HRP)phenolmixturecatalyzedoxidationof
Introduction the scopoletin by H202 resulting in a decrease in
fluorescence of the sample. Fluorescence measurements
Theoccurrence
of H202 in theatmosphere
has.been were madeon a TurnerDesignsModel10 fluorometer
explained exclusively through photochemical gas phase With excitationand emissionwaveler•gths
centeredat
reaction mechanismS. Model calculations suggest that 36.• and #90 nm, respectively.
the lower troposphericconcentrationof H20 2 produi:ed Calibration curves were prepared by .analyzing a
via these processes Should be of the order 1-#.ppbv series of solutions Of known peroxide concentrations
[Levy, 1973]. Measurementsof gas phase 1'1202 prepared
by dilutionof a I x 10-2 M H202 stock
concentrationsgive valuesOf 2-180 ppbv [Bufalini et al., solution. By varying the amount of scopoletin added,
1972• Kok et al., 1978 ]. The higher values in this range peroxide
concentrations
of 2 x 10-9 M to .• x 10-7 could
have been recorded for polluted atmospheric beaccuratelydetermined
witha precision
of + 2%.
environments and the lower levels are probably more Low contaminant distilled water was t•ed for all
typical of ambient concentrations•andwouldtherefore be reagents preparations, dilutions.andin. the traps. It was
in agreement with the model predicted values. . prepared by slow distillation from a neutral
The data basefor environmentalH 20 2 measurements permanganate-Milli-Qwater solution. The water Was
is limited and little information exists for true ambient distilled througha 55 cm packedcolumn.andcollected in
background values. The Complete lack of such a reservoir which was isolated from the atmosphere by
measurements for the southeastern United States and our means of an activated carbon filter. Water prepared in
current interest in determining the atmospheric input of thiswaywasfoundto belowin organic,transitionmetal
H202 tO the oceansled usto make H202 measurements (i.e., Cu•Fe andMn),andresidualoxidantCOntaminants.
in ground level air in South Flori•da and the Bahamas. b. OzOne.Ozonein the gasstreamwasdetermined
The air measurementsappeared to be in good agreement bY plac!nga 296neutralKI solutionin the first fritted
with those determined elsewhere by other analytical gas washing trap. After sampling a known volume of the
methods [Kok et al., 1978 ] which, like this study, equilibrated gas stream the absorbance of the trapping
employ aqueous trapping of gas phase H20 2. Further solution was measured at 362 rim. At neutral pH the
evaluation of our experimental procedure indicated that reaction,
the validity of the air measurements was in question
because of the anomalous behavior of H202 in the 03+3I +H20+I 3 +O2+2OH
aqueoustraps. It is possiblethat the observedanomalies
are the resultof aqueous
reac•:ions
of mixturesof H2 O2 yields
onemole
ofI• f.oreach
m0ieofozone
consumed
and ozone. Although the chemistry of aqueousozone is [Schechter,1973]. The absorbances of standard!3_-
not usually recognized in the atmospheric literature, solutions
were Usedto calibratethe method, The 13
there are numerous studies on its reactivity in water in solution was standardized by titration with standard
the chemical literature [ Hoigne and Bader, 197.5;Peleg, thiosulfate solution.
1976; Weinstein and Bielstd, 1979]. Although much of While I• does react with H202, the kinetics are
this information is inconsistent, it is proposedthat ozone much slower than those of reaction with Ozone. Based on
this criteria there was no evidence for the occurrence of
decaysto produceH202 [Hoigneand Bader, 1976] via
the following pathway: H 202 in the KI trap solutions.

Experimental Procedures
203+OH-+OH+O
2 +202 (1)
a. H202 in Air. The apparatus used to measure
(2) H202 in air consists of two .500ml fritted gas washing
202 +2H++H202+O
2 traps, each containing .500ml Of water, a vacuum pump,
and a flowmeter (Figure 1). Teflon tubing was usedto
Copyright 1982 by the American Geophysical Union. connect the various components,and to draw air into the

Paper number 1L1876. 231

0094-8276/82/001L-1876501.00
232
Zika andSaltzman:
Interactionof Ozone
andH202in Water

Ozonizer . increase was again observedas was the plateau. Figure 2

showsa casewherethe secondtrap plateauoccurredat a
lower concentration ot• H20 2 than the first. In some
Washing
cases there was virtually no difference between the
measured plateau H20 2 Concentrations in the two traps.
In view of H202's low vapor pressure and infinite
solubility in water, the first trap should have removed all
of the H20 2 from the air stream. This is contrary to
what was found by Pilz and Johann [1974] who reported
O3 that H 202 is very difficult to quantitatively extract
outflow fromair withwatertraps. Thispossibility
wastestedby
outside
air
(• I'" ....'.....J
replacing the air in the scrubbing experiments with a
mixture of oxidantfree 99.999%N2 andH 202 (g). This
experiment conclusively demonstrated that the first trap

Flow
Meter
was > 99% efficient in extracting H 202 and that the
concentrationincreasedlinearly with gas volumeover a
Fritted Gas Vacuum rangeof < 10-8Mto > 10-aM. It is likelythatPilz and
• WashingTraps Pump johann observedthe samephenomenon
that we did in our
air sampling methods.
Fig.1. Schematic
diagramof the apparatus
usedto The only plausible explanation we have for these
collect peroxide from air, and for ozone experiments. results is that some partially soluble and reactive
component(s) in air is generatingH20 2 in the second
trap. Since the dual trap experiments gave similar
laUratoryapparatus
fromtheoutside.Thetrapswere results on the open ocean as in the laboratory, the
filled with distilledwater and air was drawnthrough component(s)is not restricted to near shore or city air.
them at a controlled flow rate. At variousintervals the The H202 precursor must be present in substantial
flow was briefly interrupted, and samples were concentrations to generate the peroxide levels observed
Withdrawnfrom the traps for H2 02 analysis. The in the two traps. The only possible candidate known to
VOlumeof sampleremovedwas replacedwith distilled us is ozone. During the experiment, shown in Figure 2,
water; this caused only a small dilution effect of the the O3 concentration in the air at the experimental site
large volumein the traps. Prior to analysis,samples on Virginia Key varied between 35 and 45 ppb as
were degassealwith high purity helium to remove any monitored by Dade Environmental Resources
ozonethat mayhavebeenpresent. Management.
The efficiency of this system for the extraction of In dual trapping experiments with outside air, we
H202fromairwastested
bypassing
a controlled
stream observedan inductionperiodfor the formationof H2 02
of nitrogen over 30% H 202 and then through the traps.
The first trap wasfoundto be in excessof 99%efficient 10.0 .........
in removing
H202 fromthegasstreamandfor therange
of H202 concentrations
tested,10-Uto 10-3 molar,no
H 20 2 carry over could be detected in the secondtrap.
}---analytical
error
b. AqueousOzone Reactions. The effect of aqueous
ozone chemistry on the H202 levels in the collection o o
8.0
traps was investigated by replacing the air used in the
Previous experiment with a gas stream of N2, 02 and
O3. The collectors were again filled with water and the
H20 2 concentrations were measured at various intervals
during the experiment. Ozone was introduced into the
N2 stream by flowing 02 through a Supelco micro 6.0 o
ozOnizer. The gas stream was bubbledthrough two in- o
line wash traps, the first containing .1 molar HCIO• and
the second, water to remove any H202 that may have
been produced in the ozonizer. These wash traps were
equilibrated with the ozone mixture prior to the start of
the experiment. The N2 flow was 250 ml/min and the
O2 flow was 25 ml/min. After 30 minutes the 02 flow
was reduced to 5 ml/min. This effectively reduced the
ozone level to 1/6 of its original value in the flow o
system. o

2.0
Results and Discussion o
o
o
Experiments were conducted by scrubbing measured o

volumes of outside air in aqueous gas washing bottles o ß

(Figure1). The trap solutionswere analyzedfor H20 2

with the HRP-phenol method at intervals during the 0•----"' i i i i i i i i

0 40 80 120 160 200

Course of scrubbing procedure. The validity of this
approach for the determination of H 202 in the gas phase
TIME (minutes)
was questionedwhen it was discoveredthat the H20 2
increased steadily with increasing air volume scrubbed, Fig. 2. Hydrogen peroxide concentrations in aqueous
but only to a point where it reached a plateau. When a bubbletraps with 2 l/rain flow of outsideair. O = first
secondtrap was placedin series with the first the H20 2 trap, ß = secondtrap.
 233
Zika andSaltzman:Interactionof Ozone
andH202in Water

HaO2 to be a simple linear function of ozone

3.0 - concentration. In the experiments discussed here,
elaborate measures were taken to clean the glassware
and to purify the water used in the experiments. It is,
however, virtually impossibleto eliminate all traces of
I transition metals and organic materials from the water.
I This coupledwith the fact that ozoneis ubiquitousand
I
variable in concentration in the atmosphere raises
I ß
I
serious questions about methods which attempt to
I measureatmosphericH2 0 2 levels by aqueoustrapping.
I o Furthermore, these experimentsindicate that interaction
I between H20 2 and 0 3 may be important to the
I chemistry of these substancesin cloud and rain water
I
I
[ Zika et al., 1981].
I o

I
Conclusions
ß o
I
I 1. Experiments show that ozone and/or its
i
decompositionproductscan both produceand destroy
I
HaO2 in aqueoussolution.
I OZONE 2. Therefore, techniques which employ aqueous traps to
I CONCENTRATION
REDUCED measure H202 in air are subject to large
interferences from 03- H202 reactions.
iI i--analytical
error • %
Acknowledgements.
This researchwas supportedby
I I I I I NSF Grants OCE78-25628 and ATM 79-09239. We wish
10 20 30 40 50 60 to acknowledge Victor Rossinsky for translations of
TIME (minutes) foreign language manuscripts and DaleAspy for
providing atmospheric ozone data.
Fig. 3. The effect of ozone decompositionon hydrogen
peroxide concentrations in aqueousbubblingtraps using a
References
N2-O2-O3 gas stream. The ozone flow rate was
constant for the first 30 minutes. At 30 minutes the
ozone flow rate was reduced to 1/6 of its original value. Bufalini, 3.3., B.W. Gay, Jr., and K.L. Brubaker,
O = first trap, ß = second trap. Hydrogen peroxide formation from formaldehyde
photooxidationand its presencein urban atmospheres,
Environ.Sci. Technol.,6, 816-821, 1972.
in the second trap. This delay probable reflects the Hoigne,3., and H. Bader, Ozonationof water: role of
saturation of the first trap with ozone and some initial hydroxyl radicals as oxidizing intermediates, Science,
consumptionof ozone by reaction with impurities in the 190, 782-780, 1975.
secondtrap. A long induction period was not observedin Hoigne,3., and H. Bader, The role of hydroxylradical
laboratory experiments usingmuch higher concentrations reactions in ozonation processesin aqueoussolutions,
of ozonein the gasstream (Figure 3). Water Res., 10, 377-386, 1976.
To test the possibility that aqueous ozone reactions Kilpatrick, M.L., C.C. Herrick, and M. Kilpatrick, The
were producingH202 in the traps, a NZ+ Oz+O3 decomposition of ozone in aqueous solution, 3. Am.
mixture (100:10:06 by volume) was passed through the Chem. Soc.,7_•8,178#-1789, 1956.
dual trap system. The concentration of H202 was Kok, G.L., K.R. Darnall, A.M. Winer, 3.N. Pitts, Jr., and
measuredin both traps with time. The results(Figure 3) B.W. Gay, Ambient air measurements of hydrogen
clearly show the accumulation of H202 in both traps peroxide in the California South Coast air basin,
and the approach to a steady state concentration Environ.Sci. Technol.,1_•2,1077-1080,1978.
near3 x 10-6Min thefirst trapafterabout30minutes. Levy II, H., Photochemistryof minor constituentsin the
At this time the concentration of ozone in the gas was troposphere,Planet. SpaceSci., 2__1,575-591, 1973.
reducedto 1/6 its original value, or •.$ ppm, as measured Peleg, M., The chemistryof ozone in the treatment of
with the neutral KI method. The H2 02 concentration in water, Water Res., 10, 361-365, 1976.
the traps respondedby declining to a new, lower steady Perschke, H., and E. Broda, Determination of very small
state. amounts of hydrogenperoxide, Nature., 190, 257-258,
To explain these observations it is necessary to 1961.
invoke a mechanism involving both formation and decay Pilz, w., and I. Johann, Die Bestimmung Kleinster
reactions of H202. The observed concentration of Mengert yon Wasserstoff peroxyde in Luft, Intern. 3.
H20 2 is ultimately controlledby thosepropertiesof the Environ. Anal. Chem., 3, 257-270, 1970.
system which affect the ozone decompositionrate. Schechter,H., Spectrophotometric method for
The mechanism proposed in equations I and 2 for determination of ozone in aqueous solutions, Water
H202 formation involve both strong oxidizing and Res., 7, 729-739, 1973.
reducingfree radicals (i.e., OH and O2-) which are Tau---•,IT., and W.C. Bray, Chain reactions in aqueous
reactivewith ozoneand H202. Solutionpropertiessuch solutions containing ozone, hydrogen peroxide and
as pH and the presence of free radical initiators or acid, O. Am. Chem. Soc., 6_•2,
3357-3373, 19#0.
scavengersmay have a dramatic affect on the rate of Weinstein, 3., and B.3. Bielski, Kinetics. of the
decay of ozone and on the rate of formation and interaction of HO2 and O2 radicals with hydrogen
preservationof H2 02 in aqueoussolutions [Taube and peroxide. The Haber-Weiss reaction, 3. Am. Chem.
Bray, 19#0;Kilpatrick et al., 1956]. As a result of these Soc., 101.,58-62, 1979.
factors we would not expect the rate of formation of Zika, R.G., E. Saltzman, W.L. Chameides, and
234 Zika andSaltzman:Interaction of OzoneandH202in Water

D.D. Davis, H202 levels in rainwater collected in peroxide in natural waters, Anal. Chem., submitted,
South Florida and the Bahama Islands, 3. GeoDhys. 1982.
Res., submitted, 1981.
Zika, R.G., and P. Zelmer, An evaluation of the HRP- (Received September 28, 1981;
Scopoletinmethodfor the measurementof hydrogen accepted December 7, 1981.)