Temperature dependence of the hydrogen peroxide production in the γ-radiolysis of water

Article abstract of DOI:10.1021/jp0131830

The radiation chemical yield of H₂O₂ was determined in the γ-radiolysis of water at neutral pH. Ethanol, methanol, and bromide were utilized as OH radical scavengers to explore the temporal dependence of H₂O₂ formation at elevated temperatures. OH radicals were the sole source of H₂O₂ at short times in the γ-radiolysis of water. Variation in dose rate by a factor of 20 revealed no effect on the yield. The yield of H₂O₂ decreased with increasing temperature. The scavenger concentration dependence of the yields at elevated temperatures was elucidated. The thermal decomposition reaction rate for H₂O₂ in aqueous solution was ～ 6.5 (± 0.2) × exp(-71 kJ/mole-RT)/sec, assuming an activation energy of 71 kJ/mole.

Full text of DOI:10.1021/jp0131830

J. Phys. Chem. A 2002, 106, 447-452
447
Temperature Dependence of the Hydrogen Peroxide Production in the γ-Radiolysis of
Water
Igor Sˇ tefani c´ and Jay A. LaVerne*
Radiation Laboratory, UniVersity of Notre Dame, Notre Dame, Indiana 46556
ReceiVed: August 16, 2001; In Final Form: October 25, 2001
The radiation chemical yield of hydrogen peroxide has been determined in the γ-radiolysis of water at neutral
pH. Methanol, ethanol, and bromide have been used as OH radical scavengers in order to explore the temporal
dependence of hydrogen peroxide formation at elevated temperatures. The scavenger results at all temperatures
confirm that OH radicals are the sole source of hydrogen peroxide at short times in the γ-radiolysis of water.
Variation in dose rate by a factor of 20 shows no influence on the yield. With increasing temperature the
yield of hydrogen peroxide is found to decrease, and the microsecond radiation chemical yields at low scavenger
-
3
concentrations are found to be G (H
scavenger, and G (H
) ) 0.74-2.40 × 10 T (°C) molecules/100 eV for bromide. The scavenger
concentration dependence of the yields at elevated temperatures is discussed. Assuming an activation energy
2
O
2
) ) 0.78-2.43 × 10 T (°C) molecules/100 eV for methanol as a
-3
2 2
O
-
1
of 71 kJ mol the thermal decomposition reaction rate constant for hydrogen peroxide in aqueous solution
5
-1
-1
is estimated to be 6.5 ((0.2) × 10 exp(-71 kJ mol /RT) s .
Introduction
technology.11 For instance, hydrogen peroxide has been found
in the core and in the recirculation lines of boiling water reactor
BWR) plants. A major problem in the assessment of H2O2
damage in the steady-state operation of BWRs is that it
undergoes thermal decomposition at higher temperatures. It is
also very difficult to measure radiation chemical yields directly
in the harsh environments of reactor cores. Therefore, reliable
measurements on the yields and kinetics of H2O2 in the
radiolysis of water give extremely useful data for reactor design
and management.
The main oxidizing species produced in the γ-radiolysis of
water are the OH radical and molecular H2O2, and a complete
understanding of their kinetics is important for many aspects
(
of fundamental and applied radiation effects.1
-3
It is usually
assumed that the primary mechanism to form H2O2 is the fast
combination reaction of OH radicals. These reactions occur in
the initial nonhomogeneous distribution of water decomposition
products, commonly called a spur, produced by the deposition
of energy by the γ-ray. No real-time studies on H2O2 formation
have been performed so its temporal dependence is usually
probed by varying the concentration of selective scavengers for
the OH radical.4 These scavenging experiments have sug-
gested that the sole precursor to H2O2 formation is the OH
radical and the studies have provided information on the
diffusion-kinetics of spurs at room temperature. Irradiation of
water at high temperatures leads to a decrease in H2O2 yields
due to an increase in the rate of relaxation of the spur. The
yields of H2O2 have been determined at high temperatures only
It this work, the production of H O was examined in the
2
2
γ-radiolysis of water at elevated temperatures. Methanol,
ethanol, and bromide were used as scavengers of the OH radicals
to give the temporal dependence of H2O2 formation. Temper-
atures were varied from 25 to 150 °C. Information was also
obtained on the thermal decomposition rate of H2O2 at tem-
peratures above 100 °C.
-7
Experimental Section
8-10
in the low scavenger concentration regime.
The correspond-
ing time is on the order of microseconds and the yields, known
as escape yields, represent the limiting value following the
complete relaxation of the spurs to a homogeneous distribution
of water products. The time scale for the formation of H2O2
has not been determined at elevated temperatures. In this work,
selective OH radical scavengers are used to probe the time
dependence of the formation of H2O2 at temperatures up to 150
The steady-state experiments were carried out using three
different Co γ-sources at the Radiation Laboratory of the
60
University of Notre Dame. Dose rates were 1.1, 3.9, and 19
4,7
krad/min as determined by Fricke dosimetry. All solutions
were freshly prepared before each experiment with water
purified by a Millipore Milli-Q UV system. Methanol, ethanol,
and NaBr were of the highest grade commercially available.
Hydrogen peroxide was determined by the Ghormley tri-iodide
°
C. The results give fundamental information on the diffusion-
kinetics of the main oxidizing species in the γ-radiolysis of water
at elevated temperatures.
-
method with KI, in which iodide is oxidized to I3 by the
7
,12,13
-
hydrogen peroxide.
Absorption measurements of I3 were
An important practical application for examining the radiolytic
formation of H2O2 at elevated temperatures is its role in nuclear
power reactors. Hydrogen peroxide formed in the radiolysis of
water is found to be the main corrosion product, and it is
involved in oxidation damage in almost every domain of nuclear
done at its peak maximum, λmax, taking into account that λmax
shifts with solvent composition. The absorbance of the dilute
aqueous solutions was measured at 350 nm using a diode array
spectrophotometer (Hewlett-Packard HP8453). The molar ex-
-
tinction coefficient of I3 at 350 nm, ꢀ350, was calibrated by a
*
Corresponding author.
standard solution of H2O2 (Fischer Scientific) and was found
1
0.1021/jp0131830 CCC: $22.00 © 2002 American Chemical Society
Published on Web 12/14/2001

448 J. Phys. Chem. A, Vol. 106, No. 2, 2002
Sˇ tefani c´ and LaVerne
-
1
-1
-
to be 25500 M cm . The equilibrium constant for I + I2
The OH radical scavengers used here are methanol, ethanol,
and bromide. All are efficient OH scavengers without producing
undesirable side effects.
-
-1
S I3 was found to be 800 M . To keep the equilibrium
quantitatively on the side of I3 the concentration of iodide was
-
kept at 0.12 M. For the solutions with high solute concentrations
•
OH + CH OH f •CH OH + H O
3 2 2
-
1
-1
the previously determined ꢀ350 value of 26000 M cm was
7
k ) 9.7 × 10 M s-1 (3a)
8
-1
used. All solutions contained 25 mM NaNO3 to inhibit
3
a
destruction of the H2O2 by hydrated electrons and H atoms.
-
5
The only exception was the lowest (10 M) concentration of
methanol, which was 2.5 mM in NaNO3 to minimize cooperative
effects, see below.
•OH + C H OH f CH C•HOH + H O
2 5 3 2
k ) 1.9 × 10 M s-1 (3b)
9
-1
3
b
-
-
k ) 1.1 × 10 M s-1
10
-1
Radiolysis at elevated temperatures was performed using a
specially designed oven that fits in the sample drawers of the
cobalt sources. The oven consisted of a 25.4 mm diameter brass
cylinder, 102 mm long, and with caps on each end. It was
insulated with a 25.4 mm ID, 50.8 mm OD, cylinder of
alumina-silica. The brass cylinder was bored axially to 10 mm
to accept a Pyrex sample cell. Each of the end caps contained
a 100-watt cartridge heater and temperature was maintained to
within 2% with an Omega Engineering temperature controller.
Examination of the cylinder with thermocouples showed that a
uniform temperature was maintained throughout the cylinder
interior after equilibrium was achieved, which took about 30
min. The main advantage of the oven design is that a sample
cell can be inserted and reach equilibrium within a few minutes
and then be quickly retrieved and cooled. The sample cells were
made from 10 mm OD Pyrex tubes sealed on one end and with
a slight restriction near to the other open end. Prior to irradiation
the sample cells were sealed with rubber septa, purged with
helium, and permanently sealed by heating and drawing the tube
closed at the restriction. Each sample cell contained about 5
mL of solution. Removing the cells from the source immediately
after irradiation and immersing them in cold water minimized
thermal decomposition of the H2O2.
•
OH + Br f •BrOH
(3c)
3
c
The scavenging capacity of the system is defined as the product
of the scavenger concentration and the appropriate rate constant.
The scavenging capacity is formally equivalent to the pseudo-
first-order rate constant and its inverse gives the lifetime for
the formation of H2O2 in neat water. As discussed below, there
are other more sophisticated methods using Laplace transform
techniques by which the scavenging capacity dependence of
H2O2 yields can give accurate temporal dependences.16
Hydrogen peroxide is determined in this study as an end
product. It therefore must be protected from the eaq and H
atoms escaping the spurs. At long times the following reactions
will considerably lower the observed H2O2 yields unless they
are prevented from occurring:
-
eaq- + H O f •OH + OH k ) 1.1 × 10 M s-1 (4)
-
10
-1
2
2
4
7
-1
•
H + H O f •OH + H O k ) 9.0 × 10 M s-1
(5)
2
2
2
5
In all of the experiments reported here, sodium nitrate was added
to scavenge hydrated electrons and hydrogen atoms.
eaq- + NO3 f NO32- k ) 9.7 × 10 M s-1 (6)
H + NO3 f HNO3- k ) 1.4 × 10 M s-1 (7)
-
9
-1
7
Error limits for H2O2 yields are estimated to be about (2%
for small scavenger concentrations and about (4% for high
scavenger concentrations because of the small yields measured.
-
6
-1
8
The concentration of sodium nitrate was 25 mM in all solutions
except one. At the higher solute concentrations this amount of
nitrate ensures that it will react with essentially all the hydrated
electrons and H atoms before they undergo any significant
reaction with hydrogen peroxide. At the lowest methanol
Results and Discussion
The γ-radiolysis of water leads to a series of isolated spurs
consisting of nonhomogeneous distributions of water decom-
-
5
1
-3
concentration (10 M) 25 mM of nitrate can scavenge the
hydrated electron before the alcohol scavenges the OH radical.
Since one of the major reaction of the hydrated electron is with
the OH radicals (reaction 2) scavenging it too soon will lead to
an increase in OH yields above that expected. Therefore, at the
position products.
The main oxidizing species is the OH
radical produced by proton transfer from the molecular water
cation to a neighbor water molecule.14 To a lesser extent O
atoms are also produced, and in γ-radiolysis they react im-
mediately with water to give two OH radicals. Other reactive
-5
lowest methanol concentration (10 M) the nitrate was 2.5 mM
in order not to interfere with the normal OH radical chemistry
in the spur. Previous work has shown that the increase in OH
radical yields is only about 5% over these nitrate concentra-
-
species also produced in water radiolysis are eaq and H atom.
All permutations of radical reactions are possible, but the
predominant reactions of the OH radicals as the nonhomoge-
neous distributions relax by diffusion are the following:
5,6
tions. Extensive studies examining the cooperative effect of
nitrate/methanol solutions on H2O2 formation are underway.
The dependence of H2O2 formation on scavenger concentra-
tion in the γ-radiolysis of water was first performed at room
temperature using a number of different scavengers. The purpose
of these experiments was to match previous data in order to
reveal any discrepancies and to verify that the scavenger systems
were clear of any undesirable side effects. Some scavengers
can produce secondary products that interfere in the normal
water chemistry, especially when they are added at high
concentrations. Figure 1 shows the results at room temperature
and a dose rate of 1.06 krad/min for the yields of H O as a
1
0
-1
•
OH + •OH f H O2 2k ) 1.1 × 10 M s-1 (1)
2
1
OH + eaq f OH- k ) 3.0 × 10 M s-1 (2)
-
10
-1
•
2
All room-temperature rate constants used here were taken from
Buxton et al. It has been shown that the temporal dependence
of H2O2 formation can be obtained by selective scavenging of
the OH radical. This procedure provides an easy method to
obtain temporal dependences of water products using end
product analysis.
1
5
1
6
2
2
function of the scavenging capacity of the system. This figure

Hydrogen Peroxide Production in the γ-Radiolysis of Water
J. Phys. Chem. A, Vol. 106, No. 2, 2002 449
TABLE 1: Parameters for the Model Fit to the Scavenger
Data
temperature
G
esc
R
ns
scavenger
methanol
°C
molecules/100 eV
25
0.733
0.668
0.605
0.551
0.494
0.429
0.737
0.701
0.488
0.439
0.408
0.721
0.642
0.616
0.883
0.875
0.562
0.569
0.606
0.706
0.405
0.277
6
8
0
0
1
1
1
00
20
50
25
25
00
20
50
ethanol
bromide
1
1
1
the G value is constant with dose up to about 100 krad
suggesting that possible side effects due to the reactions of
nitrate or products from the scavenging reactions are negligible.
It is sometimes important in many practical applications to
be able to predict H2O2 yields at a particular scavenging capacity
of the OH radical. Fundamental radiation chemistry studies often
require the time dependence of product formation in order to
compare with model calculation. Extensive studies on the
diffusion-kinetics within a spur suggest that the scavenger
capacity dependence of H2O2 yields can be well characterized
Figure 1. Hydrogen peroxide yields in γ-radiolysis as a function of
the scavenging capacity for •OH radicals: (b) methanol and 25 mM
sodium nitrate, this work; (9) ethanol and 25 mM sodium nitrate, this
work; ([) bromide and 25 mM sodium nitrate, this work; (]) aerated
bromide, ref 4; (O) methanol and 25 mM sodium nitrate, ref 7; (0)
ethanol and acetone, ref 5; (4) ethanol and sodium nitrate, ref 6. The
solid line is from eqs I and II using the parameters of Table 1 for
methanol at 25 °C.
16
using the following empirical relationship:
contains the results of the present work for methanol, ethanol,
and bromide, as well as previous results at neutral pH using
these same scavengers.4-7 The results throughout this work are
presented in terms of the radiation chemical yields with units
of molecules/100 eV of total energy absorbed by the entire
system (water and scavenger). It can be seen that the general
agreement between the results is very good. The unique
scavenging capacity dependence indicates that the scavenger
systems perform as expected. Some discrepancy between the
present results and that from the literature can be seen at very
high concentration of scavengers where the present yields are
distinctively lower. This difference is probably due to impurities
in the solutes used in the previous work. At high scavenger
concentrations these impurities make up a significant fraction
of the total system. It was found that using less than the best
quality methanol and ethanol available leads to increased H2O2
yields at the highest alcohol concentrations.
G (H O ) ) G (1 - F(s))
(I)
s
2
2
esc
1/2
1/2
F(s) ) ((Rs) + (Rs)/2)/(1 + (Rs) + (Rs)/2) (II)
Here, Gesc is the long time or escape yield, s is the scavenging
capacity of the system, and R is a time constant. Another
parameter representing the initial yield, G , is usually included
o
in eq I, but it is assumed to be zero in the case of H O . The
2
2
value of R depends on the reactant (OH radical in this work)
and the type of radiation (γ-rays in this work). The data of the
present work were fit using nonlinear least-squares methodology
to eqs I and II to give the results given in Table 1. The solid
line in Figure 1 shows the predicted yields using eqs I and II
with the constants for methanol at 25 °C. The form of eq II
was chosen because it well represents the track kinetics expected
for γ-rays and because an analytical form of its inverse Laplace
1
6
transform exists. The temporal dependence of H O can be
2
2
The yield of hydrogen peroxide at high scavenging capacities
is very small and difficult to measure. However, the relatively
higher yields in previous studies could be interpreted as due to
a unimolecular source of H2O2. The trend with the present results
obtained readily from an equation similar to that for the
scavenging capacity dependence using the same data param-
etrization values.
and the yield of only 0.07 molecules/100 eV at a scavenging
G
t
(H
2
O
2
) ) Gesc (1- F(t))
(III)
(IV)
capacity of about 1010
s
-1
strongly suggests that the limiting G
1
/2
(H2O2) value at high scavenger capacity is zero. Hydrogen
F(t) ) 2F ((4.0t/(Rπ)) )
f
peroxide production in γ-radiolysis at times less than ∼1 µs is
due almost exclusively to spur reactions of the OH radicals.
Low scavenger concentrations correspond to long times, and
the limiting yield represents the amount of product that escapes
into the bulk following the relaxation of the spur by diffusive
processes. At times longer than a few microseconds the
distribution of water decomposition products in γ-radiolysis is
essentially homogeneous. The escape yield represents the yield
found in continuous or long-time radiolysis conditions, which
is of significant practical importance. All the scavengers used
in this work give similar escape yields, Gesc (H2O2) ) 0.73 (
The auxiliary function for the Fresnel integrals, Ff, can be easily
1
7
evaluated using standard algorithms. Inverse Laplace trans-
forms can be problematical because they require a large (infinite)
amount of data. The use of an analytic formula greatly simplifies
data analysis without sacrificing precision.
All the experiments presented in Figure 1 were performed at
a relatively low dose rate of 1.06 krad/min to minimize radical-
radical termination reactions in the spur, which could affect the
escape yield. The effect is not expected to be significant with
primary radical reactions in the spur, but secondary reactions
due to the added solutes could cause problems. Secondary
reactions may especially be significant at higher temperatures.
Therefore, the effect of dose rate on the escape yield of hydrogen
peroxide has been measured with methanol and bromide as
0
.02 for methanol, 0.74 ( 0.02 for ethanol, and 0.70 ( 0.02
for bromide. The results agree nicely with previous results which
establish the escape yields of hydrogen peroxide with γ-rays to
7
be 0.70 ( 0.03 molecules/100 eV. Experiments also show that

450 J. Phys. Chem. A, Vol. 106, No. 2, 2002
Sˇ tefani c´ and LaVerne
Figure 3. Hydrogen peroxide yields as a function of the scavenging
capacity of methanol (25 mM nitrate, deaerated, pH neutral) at different
temperatures: (9) 25 °C; (b) 60 °C; (2) 80 °C; (1) 100 °C; ([) 120
°C; (f) 150 °C. The solid lines are from eqs I and II using the
parameters of Table 1 for methanol.
Figure 2. Hydrogen peroxide yields as a function of the scavenging
capacity of methanol and bromide at different dose rates. All solutions
deaerated with methanol or bromide and 25 mM sodium nitrate, pH
neutral: (9) methanol at 1.06 krad/min; (0) methanol at 3.79 krad/
min; (]) methanol at 19.03 krad/min; (b) bromide at 1.06 krad/min;
(O) bromide at 19.03 krad/min. The solid line is the same as in Figure
1
.
scavengers at the higher dose rates of 3.79 and 19.03 krad/min.
Other conditions of the system remain the same as in Figure 1,
i.e., deaerated and neutral pH aqueous solutions containing 25
mM NaNO3 and different concentrations of scavengers. Figure
2
shows the yields of H2O2 as a function of scavenging capacity
of methanol and bromide for the three different dose rates. It
can be seen that variation in the dose rate by a factor of 20
does not influence the escape yield of hydrogen peroxide,
irrespective of the type of scavenger. Sunaryo et al. examined
the products of the scavenging reactions and came to similar
conclusions that secondary reactions were not significant in
methanol solutions at high temperature.18 The products of the
scavenging reactions probably react with themselves or with
water to form stable products that do not decompose hydrogen
peroxide.
The scavenging capacity dependence of hydrogen peroxide
at elevated temperatures gives valuable new information on the
chemistry in spurs. Figure 3 shows the yields of hydrogen
peroxide as a function of the scavenging capacity of methanol
at 25, 60, 80, 100, 120, and 150 °C. The scavenging capacity
was calculated taking into account the changes in the density
of water with increasing temperature.19 For mixtures with high
methanol concentration the density is calculated assuming the
additivity of the different components, e.g., at the highest
methanol concentration the alcohol is 20 wt % of the solution.
The rate coefficients for the reaction of methanol and the
hydroxyl radical (reaction 3a) have been calculated by using
Figure 4. Hydrogen peroxide yields as a function of the scavenging
capacity of bromide (25 mM nitrate, deaerated, pH neutral) at different
temperatures: (9) 25 °C; (b) 100 °C; (2) 120 °C; (1) 150 °C; (3)
methanol at 150 °C. The solid lines are from eqs I and II using the
parameters of Table 1 for bromide and the dashed line is for methanol
at 150 °C.
H O . For instance, methanol could thermally decompose at high
2
2
temperature to give undesirable products that interfere with the
chemistry or there could be a significant change in methanol
concentration. Previous results suggest that aqueous methanol
2
1
solutions are stable to about 450 °C, but it was decided to
confirm this observation with the present experimental config-
uration. Additional experiments were performed using a bromide
as a scavenger with the other conditions unchanged from the
methanol solutions. The scavenging capacity for the hydroxyl
radical with bromide has been calculated on the same way as
-
1
the activation energy of 4.8 kJ mol , and the resulting rate
coefficients are expected to have errors of <10%.20 Such small
errors are not noticeable on the logarithmic scavenger capacity
scales used here. With increasing temperature the plateaus or
escape yields at low scavenger concentrations decrease. It
appears that the inflection points of the curves remain unchanged
-
1
for methanol. The activation energy of 18 kJ mol for the
reaction between bromide and the OH radical has been used.20
Figure 4 shows the yield of hydrogen peroxide as a function of
scavenging capacity of bromide at 25, 100, 120, and 150 °C.
Comparison between the results using the two scavengers shows
that the plateaus or low scavenger limiting yields with bromide
are slightly lower than for methanol at the high temperatures.
However, the differences are within the experimental errors.
The results for methanol at 150 °C are given in Figure 4 for
9
-1
with increasing temperature at a value of about 10 s . The
limiting yields at the highest scavenging capacities are nearly
independent of temperature and appear to approach zero. This
result suggests that the OH radical is the precursor to hydrogen
peroxide at all temperatures.
High temperatures could cause unknown, adverse effects on
the methanol and can lead to false radiation chemical yields of

Hydrogen Peroxide Production in the γ-Radiolysis of Water
J. Phys. Chem. A, Vol. 106, No. 2, 2002 451
Figure 5. Limiting hydrogen peroxide yields at low scavenger
concentration as a function of temperature: (b methanol, this work;
Figure 6. Concentration of hydrogen peroxide as a function of dose
for deaerated aqueous solutions of 1 mM methanol and 25 mM sodium
nitrate, pH neutral, at 120 °C: (9) 1.06 krad/min; (b) 19.03 krad/min.
(9) bromide, this work, (4) acrylamide, ref 9; (3) iodide and N
2
O, ref
1
0. The solid line is a linear fit through the methanol data.
the fast decomposition of H2O2, it was impossible to measure
its yield at higher temperature using the present experimental
configuration. Hydrogen peroxide thermally decomposes to give
two OH radicals.
comparison. To minimize the thermal decomposition of hydro-
gen peroxide all data presented in Figures 3 and 4 were made
at a dose rate of 19.6 krad/min. This dose rate used the least
amount of time and the reported G values were obtained by
extrapolating to zero dose. Dose rate effects will be discussed
below.
H O f •OH + •OH
(8)
2
2
Two other groups investigated the temperature dependence
of H2O2 production in γ-radiolysis at low scavenger capacities.
Elliot et al. measured hydrogen peroxide yields in light and
heavy water using 0.5 mM acrylamide as a scavenger to prevent
the primary radicals from destroying the peroxide. They also
used the Ghormley tri-iodide method with a dose range of 3-12
krad. Their G values extrapolated to zero dose at 20, 70, and
Takagi and Ishigure find the activation energy of reaction 8 to
-
1
5
be 71 kJ mol with a rate coefficient of k8 ) 6.4 × 10 exp
-1
-1 22
(
-71 kJ mol /RT) s . Thus, the thermal decomposition half-
life, τ1/2 ) ln(2)/k8, of H2O2 can be estimated to be 158, 49,
and 11 min at 100, 120, and 150 °C, respectively. At 200 °C
its half-life is only 77 s, which makes it difficult to perform
normal radiation chemical studies using end product analysis.
It was previously mentioned that the results presented here
in Figures 3 and 4 were obtained at the highest dose rate
available to prevent errors caused by the thermal decomposition
of hydrogen peroxide during the irradiation. Additional experi-
ments at lower dose rates showed no measurable changes in
yields of hydrogen peroxides. However, because of its thermal
decomposition, the dependence of the H2O2 concentration on
dose deviated significantly from linearity at low dose rates with
temperatures above 100 °C. The relatively longer irradiation
time leads to a greater destruction of H2O2. Figure 6 shows the
dose dependence for H2O2 concentration using methanol as a
scavenger irradiated at 1.06 and 19.06 krad/min and a temper-
ature of 120 °C. At the higher dose rate it takes only a couple
of minutes to reach a dose of 60 krad and the dose response is
linear. At the lower dose rate it takes a much longer irradiation
time to reach the same dose, and the thermal decomposition of
H2O2 is now significant. The difference in the dose response
can be used to measure the thermal decomposition half-life of
hydrogen peroxide at the higher temperatures using the follow-
ing equation:
9
1
00 °C are 0.69 ( 0.02, 0.61 ( 0.04, and 0.58 ( 0.03
molecules/100 eV, respectively. The correlation between the
temperature and the G value can be given by G (H2O2) ) 0.72-
-
3
1
.40 × 10 T (°C). Kent and Sims measured the yield of H2O2
over the temperature range 25-270 °C in a high temperature
and high pressure loop by measuring the oxygen production.
They used slightly alkaline solutions containing iodide ions as
scavengers and saturated with N2O.10 The reported G values
for H2O2 at 25, 145, and 270 °C are 0.75, 0.59, and 0.28
molecules/100 eV, respectively. From these results the correla-
tion between the temperature and the G value can be given by
G (H2O2) ) 0.82-1.92 × 10 T (°C). The low scavenging
capacity limiting yields for hydrogen peroxide measured in this
work are shown in Figure 5 with the results of the above data.
The correlation between the temperature and the G value for
methanol in the present work has been calculated to be G (H2O2)
-
3
-
3
)
0.78-2.43 × 10 T (°C), which is shown as a solid line in
Figure 5. For bromide this correlation was found to be G (H2O2)
-
3
)
0.74-2.40 × 10 T (°C). For both of the scavengers
examined here, these correlations give larger negative slopes
9
10
than those from Elliot et al. and Kent and Sims, i.e., the yield
of hydrogen peroxide in the present work decreases faster with
increasing temperature. The main discrepancy with Elliot et al.
is with the datum at the highest temperature, while the other
data agree well. Kent and Sims used N2O saturated solutions,
which converts hydrated electrons to OH radicals, thereby
making a direct comparison between the results improper.
The H2O2 yield has been measured at temperatures of 25,
-
1
-1
kobs ) k exp (-E kJ mol /RT) s
D
a
where kobs ) ln 2/τ , k is the preexponential factor, and E is
1/2
D
a
the activation energy. Using the activation energy of 71 kJ
-
1
mol , the estimated lifetimes for the thermal decomposition
of H2O2 at 100, 120, and 150 °C are 145, 47, and 9 min,
respectively, which are very close to those values of Takagi
6
0, 80, 100, 120, and 150 °C with methanol as a scavenger and
2
2
at 25, 100, 120, and 150 °C for bromide scavenger. Because of
and Ishigure. The calculated rate coefficient for the thermal

452 J. Phys. Chem. A, Vol. 106, No. 2, 2002
Sˇ tefani c´ and LaVerne
decomposition of hydrogen peroxide is estimated to be 6.5
(2) Draganic, I. G.; Draganic, Z. D. The Radiation Chemistry of Water;
Academic Press: New York, 1971.
5
-1
-1
(
(0.2) × 10 exp(-71 kJ mol /RT) s .
(3) Buxton, G. V. In Radiation Chemistry: Principles and Applications;
A recent paper by Croiset et al. reported an activation energy
Farhataziz, Rodgers, M. A. J., Eds.; VCH Publishers: New York, 1987; p
321.
(4) Allen, A. O.; Holroyd, R. A. J. Am. Chem. Soc. 1955, 77, 5852-
855.
-
1
of 182 kJ mol for the decomposition of H2O2 in supercritical
water and 46 kJ mol for water in the range of 150-450 °C.
Their reported rate coefficient for the thermal decomposition
-
1
22
5
(
5) Draganic, Z. D.; Draganic, I. G. J. Phys. Chem. 1969, 73, 2571-
2577.
(6) Draganic, Z. D.; Draganic, I. G. J. Phys. Chem. 1971, 75, 3950-
957.
7) Pastina, B.; LaVerne, J. A. J. Phys. Chem. A 1999, 103, 1592.
8) Elliot, A. J.; Chenier, M. P.; Ouellette, D. C. Can. J. Chem. 1990,
68, 712.
3.5(0.2
of H2O2 in the lower temperature regime is 10
exp(-46
-
1
-1
kJ mol /RT) s . Half-life values estimated from this equation
are 10, 5, and 2 min for 100, 120, and 150 °C, respectively.
These values are considerably lower than those of Takagi and
Ishigure and the present results. The observed differences may
be due to wall effects. The present work used Pyrex cells, while
3
(
(
22
(9) Elliot, A. J.; Chenier, M. P.; Ouellette, D. C. J. Chem. Soc., Faraday
Trans. 1993, 89, 1193.
10) Kent, M. C.; Sims, H. E. Harwell Research Report AEA-RS-2302,
1992; p 153.
(11) Cowan, R. L. Water Chemistry of Nuclear Reactor Systems 7;
British Nuclear Energy Society: London, 1996.
22
23
Takagi and Ishigure and Croiset et al. used stainless steel
and Inconel reactors, respectively. Further experiments examin-
ing the thermal decomposition of hydrogen peroxide are in
progress.
(
(
(
12) Ghormley, J. A.; Stewart, A. C. J. Am. Chem. Soc. 1956, 78, 2934.
13) Hochanadel, C. J. J. Phys. Chem. 1952, 56, 587.
Acknowledgment. The authors thank Drs. Simon M. Pim-
blott and Marie Begusova for sharing their unpublished results
of model calculations. The work described herein was supported
by Grant DE-FG03-99SF21923 of the Nuclear Energy Research
Initiative Program of the U.S. Department of Energy. This
contribution is NDRL-4327 from the Notre Dame Radiation
Laboratory, which is supported by the Office of Basic Energy
Sciences of the U.S. Department of Energy.
(14) LaVerne, J. A.; Pimblott, S. M. J. Phys. Chem. A 2000, 104, 9820.
(15) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J.
Phys. Chem. Ref. Data 1988, 17, 513.
(
(
16) Pimblott, S. M.; LaVerne, J. A. J. Phys. Chem. 1991, 95, 3196.
17) Abramowitz, M.; Stegun, I. A. Handbook of Mathematical Func-
tions; Dover: New York, 1970.
18) Sunaryo, G. R.; Katsumura, Y.; Hiroishi, D.; Ishigure, K. Radiat.
Phys. Chem. 1995, 45, 131.
19) Water: A ComprehensiVe Treatise, Vol. 1; Franks, F., Ed.; Plenum
Press: New York, 1972.
(
(
(
(
20) Elliot, A. J.; McCracken, D. R. Radiat. Phys. Chem. 1989, 33, 69.
21) Rice, S. F.; Hunter, T. B.; Ryden, A. C.; Hanush, R. G. Ind. Eng.
References and Notes
Chem. Res. 1996, 35, 2161.
(
1) Allen, A. O. The Radiation Chemistry of Water and Aqueous
(22) Takagi, J.; Ishigure, K. Nucl. Sci. Eng. 1985, 89, 177.
(23) Croiset, E.; Rice, S. F.; Hanush, R. G. AIChE J. 1997, 43, 2343.
Solutions; Van Nostrand: New York, 1961.