Effect of molecular hydrogen on hydrogen peroxide in water radiolysis

Article abstract of DOI:10.1021/jp012245j

The long-time chemistry occurring in the radiolysis of water with different types of radiation has been examined. Radiolytic processes were probed by determining the influence of added molecular hydrogen on the formation of hydrogen peroxide in the radiol

Full text of DOI:10.1021/jp012245j

9316
J. Phys. Chem. A 2001, 105, 9316-9322
Effect of Molecular Hydrogen on Hydrogen Peroxide in Water Radiolysis
Barbara Pastina† and Jay A. LaVerne*,‡
National Research Council Board on RadioactiVe Waste Management, Washington, D.C. 20418, and
Radiation Laboratory, UniVersity of Notre Dame, Notre Dame, Indiana 46556
ReceiVed: June 14, 2001; In Final Form: August 7, 2001
The long-time chemistry occurring in the radiolysis of water with different types of radiation has been examined.
Radiolytic processes were probed by determining the influence of added molecular hydrogen on the formation
of hydrogen peroxide in the radiolysis of water with γ rays, 2 and 10 MeV protons, and 5 MeV helium ions.
Homogeneous model calculations were used to obtain quantitative information about the yields of radicals
and molecular products escaping the heavy ion tracks. The results show that the yields of radicals escaping
from the tracks of 10 MeV protons is significant, whereas the corresponding yields with 2 MeV protons and
5 MeV alpha particles are much lower. The addition of molecular hydrogen has a negligible effect on the
formation or consumption of hydrogen peroxide in the radiolysis of water using heavy ions with a high linear
energy transfer rate, LET. This result is contrary to the predictions of a homogeneous model and suggests
that the long-time chemistry of water is not well-known or that a homogeneous model cannot be applied to
high LET radiation. There is also the possibility that a significant yield of an oxidizing species is produced
at high LET.
Introduction
by 1 µs in high-energy electron radiolysis, and the subsequent
reactions are dominated by the interconversion of oxidizing and
reducing radicals and the reaction of radicals with molecular
products to re-form water. The interesting reaction to be
considered in this work is the reaction of OH radicals with H2.
This reaction and its complement (the H atom reaction with
H2O2) make up a chain reaction for the reformation of water
molecules. In this study, the effects of the addition of H2 on
the yields of H2O2 are reported for different types of radiation.
Chain reactions are difficult to examine because even small
amounts of impurities can interfere. The many competing factors
make it necessary to perform experiments on the long-time
radiolysis of water, especially with heavy ions.
The long-time effects of water radiolysis are important for
both fundamental and practical reasons, and yet, almost no basic
experimental studies have been performed, especially with
incident heavy ions. Most radiation chemistry studies of water
use conventional radiation such as high-energy electrons or γ
rays and probe the intratrack chemistry of the water decomposi-
tion products with pulse radiolysis or selective scavenger
techniques.1-3 This chemistry ranges from the water dissociation
following the passage of the radiation to about 1 µs, the time at
which the spatially nonhomogeneous distribution of reactive
species has relaxed. Monte Carlo modeling techniques are able
to correlate well the different product yields and give mecha-
nistic information on the fast physical and chemical processes
occurring on this time scale.4 Models employing the yields from
the fast stage of water radiolysis and appropriate chemical
reactions are straightforward to develop because the kinetics is
homogeneous.5,6 However, even small errors in the input
parameters can lead to incorrect predicted yields after extra-
polating to long periods of time. Experimental yields for only
a few of the radicals and molecular species escaping the tracks
of heavy ions are known, so it is difficult to determine the
accuracy of any model of the long-time radiation chemistry of
water with heavy ions.7,8 Selected experiments that examine the
chemistry of water radiolysis in the time domain beyond a few
microseconds can greatly aid the development of track models
for understanding water radiolysis and for application to practical
situations.
One very important practical application for examining the
long-time radiolysis of water is the use of molecular hydrogen
dissolved in the cooling water of pressurized water reactors
(PWRs) and boiling water reactors (BWRs) to mitigate corro-
sion. Hydrogen is known to reduce the concentration of
oxidizing species such as O2 and H2O2 and lower the electro-
chemical corrosion potential (ECP) of metals in the reactor’s
internal components, which is associated with cracking.9 The
beneficial effect of molecular hydrogen has been known for
many years,10 and justification for its use has been based on a
mechanism proposed by Allen et al. in 1952.11 Very little further
fundamental research has been performed on the radiolysis of
aqueous hydrogen solutions. The influence of molecular hy-
drogen on the formation of oxygen and hydrogen peroxide has
been examined recently in an experimental reactor as a function
of the linear energy transfer (LET ) stopping power, -dE/dx)
of the radiation, of the temperature, and in the presence of
different impurities.8,12 Hydrogen overpressure effects on prod-
uct formation in the recoil radiolysis of LiOH salts have also
been examined.7 Water in a nuclear reactor is exposed to many
types of radiation depending on the distance from the core. Much
is known about the general effects of LET on water radiolysis,
but the specific yields of all of the water decomposition products
Within a few milliseconds following the passage of ionizing
radiation in water, the main reducing species are H atoms and
H2 and the main oxidizing species are OH radicals and H2O2.
Most of the track reactions between the sibling radicals are over
* To whom correspondence should be addressed.
† National Research Council Board on Radioactive Waste Management.
‡ University of Notre Dame.
10.1021/jp012245j CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/12/2001

Effect of Molecular Hydrogen on Hydrogen Peroxide
J. Phys. Chem. A, Vol. 105, No. 40, 2001 9317
TABLE 1: Characteristics of the Radiation Used in This
Worka
Radiolysis with γ rays was performed using a Gammacell-
220 60Co source at the Radiation Laboratory of the University
of Notre Dame. The dose rate was 1.5 krad/min (15 Gy/min)
as determined using the Fricke dosimeter.17
energy
(MeV)
integral LET
(eV/nm)
range
dose rate
(rad/s)
particle
(mg/cm2)
4 × 105
25
20
100
25
The solutions were prepared using water from a Millipore
Milli-Q UV system. Degassed solutions were saturated with
ultrahigh purity N2 or Ar. No difference in yields was observed
between saturated solutions of Ar and N2, probably because
relatively low doses were used and N2 was not converted to
acid in any significant quantity. Selected mixtures of H2 were
prepared by saturating the solution with pure H2 or with mixtures
of 10 or 1% H2 in Ar to give solutions of 800, 80, and 8 µM,
respectively. Some of the experiments are performed with an
initial H2O2 concentration of 50 µM in order to examine its
disappearance because of reactions with radical species. The
pH of the solutions was near neutral (pH ≈ 6.5). The hydrogen
peroxide concentrations were measured using the iodide method
of Ghormley.17-19 The absorbance of the solution was measured
at 350 nm using a diode array spectrophotometer (Hewlett-
γ ray
0.23
34.8
13.8
1H+
2
10
5
7.31
119
3.56
4He2+
156
a The energy of the particle is the initial energy of the particle in
the sample. The integral LET is the LET averaged over the entire range
of the particle.
following the fast chemistry is not known for most high LET
particles.13 It has long been observed that continuous irradiation
by low LET radiation leads to steady-state concentrations of
H2 and H2O2 that are too low to be observable, whereas high
LET radiation can produce significant quantities of these
products.2 Furthermore, the transition from nonhomogeneous
to homogeneous chemistry takes place over a much longer time
scale, if at all, in the track of a heavy ion compared to a high-
energy electron track. This transition gives rise to uncertainties
in the application of homogeneous kinetic models to heavy ion
radiolysis. The experiments reported here were designed to give
fundamental information on the long-time reactions in the
radiolysis of water and to obtain estimates of the yields of
radicals and molecular products escaping heavy ion tracks.
-
Packard HP8453). The molar extinction coefficient for the I3
was measured at 350 nm with a commercial solution of H2O2
(Fischer Scientific). It was found that ꢀ350 ) 25 850 M-1 cm-1
,
in good agreement with previous data.17,19
Results and Discussion
The passage of ionizing radiation in water leads to a number
of ionic and excited states that further decompose to give radical
In this work, the influence of added molecular hydrogen on
the formation of hydrogen peroxide was examined in the γ
radiolysis and the heavy ion radiolysis of water. The heavy ions
consisted of protons (2 or 10 MeV) and 5 MeV helium ions
(alpha particles). Pure water or dilute hydrogen peroxide
solutions were irradiated with various concentrations of dis-
solved molecular hydrogen. The purpose of these experiments
is to enhance the existing knowledge on the escape yields of
radicals with high LET radiation and to examine the effects of
molecular hydrogen on the ultimate concentrations of hydrogen
peroxide. A homogeneous model calculation was used to predict
product concentrations in the long-time radiolysis of water with
various amounts of additives. This model is also used to estimate
radical and molecular product yields escaping the tracks of heavy
ions.
and molecular species. Within a few picoseconds of the initial
-
energy deposition event, the water decomposition products eaq
,
H, OH, and H2 can be found in clusters along the particle
path.1-3 The initial geometry of the nonhomogeneous spatial
distributions of water decomposition products is dependent on
the physics of the energy deposition by the incident particle,
but it is currently believed that the initial yields are independent
of particle type.13 Track chemistry is defined here as the fast
reactions occurring while the nonhomogeneous distributions
relax. Because the distributions of reactive species are non-
homogeneous, kinetic models for the track chemistry must
include diffusion. In principle, all permutations of reactions of
the initial species can occur, but in practice, only 10 or so of
the fastest reactions are needed to account for the observed
yields.20 Relaxation of the track occurs within a few micro-
seconds for high-energy electrons, but it can extend to longer
times for high LET heavy ions. Secondary reactions involving
radicals and product molecules become important as well as
acid-base equilibrium reactions at the longer times.5,6 The
complex interactions of radicals and molecular products are
difficult to predict, and homogeneous kinetic models are very
useful for examining the chemical events occurring in constant
radiation fields at long times.
Homogeneous Kinetic Model. Although the short time
radiation chemistry (<1 µs) of water can be described with only
a few equations, examination of the long-time kinetics requires
the contributions of many more reactions. The complex inter-
action of radical species as they interconvert and react with
molecular products is very difficult to follow without the aid
of kinetic models. The model chosen for this work is essentially
the same as previously developed by Elliot and McCracken,
and the appropriate reactions are listed in Table 2.5 Basically,
the yields of water decomposition products following the end
of the nonhomogeneous phase, called the escape yields, are
allowed to react homogeneously according to the set of reactions
given in Table 2. During the irradiation period, water decom-
position products were formed continuously in conjunction with
Experimental Section
The heavy ion radiolysis experiments were performed using
the facilities of the Nuclear Structure Laboratory of the
University of Notre Dame Physics Department. Protons and
4He ions were produced and accelerated using a 10 MeV FN
Tandem Van de Graaff. The window assembly and the irradia-
tion procedure were the same as reported earlier.14,15 After
acceleration, the ions were energy and charge-state selected
magnetically and the energies of the particles are known with
a precision of about 10 kV. Energy loss of the particles in
passing through all windows was determined from a standard
stopping power compilation.16 The characteristics of the particles
used in this experiment are given in Table 1. The solutions were
irradiated with completely stripped particles at a beam current
of about 2 nA of charge into 20 mL samples, which were
vigorously stirred. Absolute dosimetry was obtained from the
product of the integrated beam current and the particle energy.
The ranges of the heavy ions (Table 1) are smaller than the
sample thickness, so the ions are completely stopped in the
solution. The radiation chemical yields represent all processes
from the initial particle energy to zero and are therefore track
averaged yields.

9318 J. Phys. Chem. A, Vol. 105, No. 40, 2001
Pastina and LaVerne
TABLE 2: Reactions and Rate Constants Used in the Model Calculations
Formation Reaction
H2O
1
f
e-aq, H, OH, H2, H2O2, HO2
Equilibria
H2O
H2O2
OH
HO2
H
pKa
2
3
4
5
6
T
T
T
T
T
H+ + OH-
13.999
11.65
11.9
4.57
9.77
H+ + HO2
-
H+ + O-
H+ + O2
-
H+ + e-
aq
Acid - Base Reactions
Rate coefficients
(M-1s-1 or s-1
1.4 × 1011
)
7
8
9
H+ + OH-
H2O
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
H2O
H+ + OH-
k7 × K2/[H2O]
k10 × K3
-
H2O2
H+ + HO2
H2O2
-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
H+ + HO2
5.0 × 1010
H2O2 + OH-
HO2- + H2O
e-aq + H2O
H + OH-
H
HO2- + H2O
H2O2 + OH-
H + OH-
e-aq + H2O
e-aq + H+
H
1.3 × 1010
k11 × K2/ K3 × [H2O]
1.9 × 101
2.2 × 107
k16 × K6
e-aq + H+
OH + OH-
O- + H2O
OH
2.3 × 1010
O- + H2O
OH + OH-
O- + H+
OH
1.3 × 1010
k17 × K2/ K4 × [H2O]
k20 × K4
O- + H+
HO2
1.0 × 1011
O2- + H+
HO2
k22 × K5
O2- + H+
HO2 + OH-
O2- + H2O
5.0 × 1010
O2- + H2O
HO2 + OH-
5.0 × 1010
k23 × K2/ K5 × [H2O]
Chemical Reactions
e-aq + OH
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
f
OH-
3.0 × 1010
1.1 × 1010
1.3 × 1010/[H2O]
2.0 × 1010
1.9 × 1010
5.5 × 109/[H2O]
2.5 × 1010/[H2O]
3.5 × 109
e-aq + H2O2
OH + OH-
e-aq + O2- + H2O
e-aq + HO2
HO2- + OH-
-
HO2
O2
e-aq + O2
-
e-aq + e-aq + 2H2O
e-aq + H + H2O
H2 + 2OH-
H2 + OH-
-
e-aq + HO2
O- + OH-
e-aq + O- + H2O
e-aq + O3- + H2O
e-aq + O3
OH- + OH-
O2 + OH- + OH-
2.2 × 1010/[H2O]
1.6 × 1010/[H2O]
3.6 × 1010
1.1 × 101
-
O3
H + H2O
H2 + OH
H + O-
OH-
1.0 × 1010
9.0 × 107
H + HO2
OH + OH-
OH- + O2
H2
-
H + O3
1.0 × 1010
7.8 × 109
-
H + H
H + OH
H + H2O2
H + O2
H2O
OH + H2O
HO2
7.0 × 109
9.0 × 107
2.1 × 1010
1.8 × 1010
1.8 × 1010
3.8 × 1010
3.6 × 109
H + HO2
H2O2
-
-
H + O2
H + O3
HO2
HO3
H2O2
OH + OH
OH + HO2
H2O + O2
6.0 × 109
OH + O2
OH- + O2
8.2 × 109
-
OH + H2
H + H2O
4.3 × 107
OH + H2O2
HO2 + H2O
HO2-
2.7 × 107
OH + O-
2.5 × 1010
7.5 × 109
OH + HO2
HO2 + OH-
O3 + OH-
-
-
OH + O3
2.6 × 109
OH + O3
O2- + O2- + H+
HO2 + O2
6.0 × 109
-
OH + O3
1.1 × 108
HO2 + O2
HO2- + O2
H2O2 + O2
O2 + OH-
8.0 × 107
-
HO2 + HO2
7.0 × 105
HO2 + O-
6.0 × 109
HO2 + H2O2
OH + O2 + H2O
OH + O2 + OH-
O2 + O2 + OH-
HO3 + O2
5.0 × 10-1
5.0 × 10-1
6.0 × 109
-
HO2 + HO2
-
HO2 + O3
HO2 + O3
5.0 × 108
O2- + O2- + 2H2O
O2- + O- + H2O
O2- + H2O2
H2O2 + O2 + 2OH-
O2 + 2OH-
OH + O2 + OH-
1.0 × 102/2[H2O]
6.0 × 108/[H2O]
1.3 × 10-1

Effect of Molecular Hydrogen on Hydrogen Peroxide
J. Phys. Chem. A, Vol. 105, No. 40, 2001 9319
TABLE 2: (Continued)
Chemical Reactions
-
67
68
69
70
71
72
73
74
75
76
77
78
79
O2- + HO2
O2- + O3- + H2O
O2- + O3
f
f
f
f
f
f
f
f
f
f
f
f
f
O- + O2 + OH-
1.3 × 10-1
1.0 × 104/[H2O]
1.5 × 109
O2 + O2 + 2OH-
O3- + O2
O- + O- + H2O
O- + O2
HO2- + OH-
1.0 × 109/[H2O]
3.6 × 109
-
O3
O- + H2
H + OH-
8.0 × 107
O- + H2O2
O2- + H2O
O2- + OH-
5.0 × 108
O- + HO2
4.0 × 108
-
O- + O3
O2- + O2
7.0 × 108
-
-
O- + O3
O2- + O2
O2 + O-
O2 + OH
O2 + OH
5.0 × 109
O3
3.3 × 103
-
O3- + H+
HO3
9.0 × 1010
1.1 × 105
TABLE 3: G Values Used in Model Calculations (Units of
Molecules/100 eV)
species
G(e-
G(H)
γ rays
10 MeV H
2 MeV H
5 MeV He
)
aq
2.60
0.66
0.45
2.70
0.70
0.02
3.10
0.50
0.90
0.57
0.64
1.18
0.74
0.03
1.10
0.20
0.30
0.20
0.90
0.63
0.76
0.05
0.36
0.06
0.15
0.10
1.20
0.35
1.00
0.10
0.18
0.03
G(H2)
G(OH)
G(H2O2)
G(HO2)
G(H+)
G(OH-)
their reaction until the desired dose was received by the system.
After that time, the production of water decomposition products
was terminated, and the reactions were allowed to continue until
all radical species were gone. The escape yields are given in
Table 3, and the dose rates appropriate to the heavy ion
radiolysis of this work are given in Table 1. The dose rate for
γ radiolysis was 25 krad/s (1 krad ) 10 Gy) as determined by
the Fricke dosimeter. The dose rates for heavy ion radiolysis
are for the experimental conditions of 2 nA beam currents into
20 mL of water. The rate coefficients used throughout this work
are taken from refs 7 and 21-23. There is very little disagree-
ment on the values of most of the rate coefficients because they
have been measured independently and used extensively in
nonhomogeneous model calculations.4 The major uncertainties
in the rate coefficients are with the very slow reactions and will
be discussed below. The simulation of the reactions was
performed using the Facsimile code.24,25
Figure 1. Temporal dependence of water decomposition product
concentrations for a 1 h γ radiolysis (25 rad/s) of neat water.
(25 rad), and then the major reactions are with H2O2 (reaction
26) and O2 (reaction 29) as their concentrations rise. The main
reaction of OH radicals is with OH- (reaction 17) at all times,
but it does react to some extent at times above 10 s (250 rad)
-
with O2 (reaction 49) and H2 (reaction 50). The major
additional formation of H2O2 above that escaping from the track
is the combination of OH radicals (reaction 47), although this
reaction is never the dominant reaction of OH radicals at long
times. An additional source of H2 above that escaping from the
track is the reaction of H with water (reaction 36).
The addition of H2 or H2O2 to the solution will lead to
significant scavenging of the radical species by these molecules.
With sufficient concentrations of H2 and H2O2, there is a net
interconversion of OH radicals and H atoms by reactions 42
and 50. The equivalent stoichiometric equation is H2 + H2O2
) 2H2O, resulting in a chain reaction for the net addition of H
atoms and OH radicals to re-form water. With a constant source
of radicals, this reaction will consume all of the H2 or H2O2,
whichever is the lesser. There must be a sufficient supply of
radicals to continue the chain propagation, a situation not
necessarily found in heavy ion radiolysis. The termination of
the chain reaction can be caused by the accumulation of O2,
which scavenges the H atoms (reaction 43). A large excess of
hydrogen peroxide can also break the chain reaction because it
consumes part of the OH radicals before reaction with H2 can
occur and because it leads to the formation of O2.
The recombination process for radicals to re-form water is
effective when the length of the propagation step in the chain
reaction is high, i.e., when the concentration of the radical
species is high. Therefore, the chain reaction is the most effective
in the presence of low LET radiation, such as γ radiation,
because of the high radical yields (H, OH, and e-aq) and low
molecular yields (H2O2, H2, and HO2). A heavy ion beam, i.e.,
a high LET radiation, will lead to either a shortening of the
The predicted results for the γ radiolysis of water are shown
in Figure 1. It can be seen that the concentrations of eaq-, H,
OH, H2O2, and H2 initially increase at a constant rate because
of the radiolytic source terms (product of escape yields and dose
rate). Within a few milliseconds of the start of the irradiation,
-
the eaq concentration plateaus at a value below that of the H
-
atom. Conversion of the eaq to H atom occurs mainly by
reactions 13 and 16. Within about 100 seconds (2500 rad), all
products from water radiolysis have reached their limiting
values, which are nearly independent of dose rate. It can be
seen that the major radicals are OH, H, and the O2-/HO2
conjugate pair, whereas the dominant molecular species are
H2O2, O2, and H2. All concentrations are below 1 µM, which
is about the lowest detection limit for most analytical techniques.
Therefore, the predicted results agree with the general assess-
ment that there is no net decomposition of water in the γ
radiolysis of neat water.
The dominant long-time reactions of the radicals produced
in water radiolysis are with molecular species and not with each
other. H atoms mainly react with water (reaction 36) up to about
10 s (250 rad) when the O2 concentration becomes large enough
-
to scavenge them (reaction 43). The eaq mainly reacts with
H+ (reaction 16) and with water (reaction 13) up to about 1 s

9320 J. Phys. Chem. A, Vol. 105, No. 40, 2001
Pastina and LaVerne
Figure 2. H2O2 concentration dependence on dose in the γ radiolysis
of aqueous 50 µM H2O2 solutions with various concentrations of added
molecular hydrogen: (9) no H2, (b) 8 µM H2, (2) 80 µM H2, ([) 800
µM H2. The solid lines are the homogeneous model predictions.
Figure 3. H2O2 concentration dependence on dose in the heavy ion
radiolysis of water with no H2 (filled symbols) and H2 saturated (800
µM; open symbols): ([) 5 MeV He, (9) 2 MeV H, and (b) 10 MeV
H. The solid lines are the homogeneous model predictions for solutions
with no H2, and the dashed lines are the estimated concentrations based
on the escape yields (decreasing in the order He, 2 MeV H, and 10
MeV H).
chain process or completely stop it with formation of stable
decomposition products from water radiolysis (O2, H2O2, and
H2). It has been shown that water decomposition by radiolysis
in the presence of H2 is a threshold phenomenon as a function
of the LET of the radiation.8,12 With low LET radiation, no stable
decomposition products (O2 and H2O2) are detectable. A
progressive increase of the LET leads to an increase in the
formation of O2 and H2O2 indicating that the chain reaction
has stopped or is at least inhibited. The threshold depends on
different factors such as the H2 concentration, the temperature
of water, the presence of impurities, and their nature.
Heavy Ion Radiolysis of Water. H2O2 production was
determined in the 5 MeV helium and 10 and 2 MeV proton
radiolysis of water. Water saturated with an inert gas, Ar or
N2, and solutions saturated with H2 were examined. No
difference in yields was observed between saturated solutions
of Ar and N2, probably because the doses were too low to
convert much of the N2 to acid. Figure 3 shows the H2O2
formation as a function of the absorbed dose in heavy ion
radiolysis of water without added H2O2. It can be seen that the
formation of H2O2 increases with increasing LET of the incident
particle. The H2O2 concentration dependence found with protons
is not linear as a function of the absorbed dose, indicating that
this product is reacting at long times with radicals produced in
the radiolysis. The predicted yields based on the estimated
escape yields are shown as the dashed lines in Figure 3. Only
in the case of helium ion radiolysis is the observed yield even
close to the value predicted from the escape yield. With
increasing LET of the heavy ion, the escape yields of radicals
decrease. In the 5 MeV helium ion radiolysis of water, the
radical yields have dropped sufficiently that decomposition of
the H2O2 occurs to only a small extent at the longer times.
The results found on the addition of H2 in the heavy ion
radiolysis of water are shown as the open symbols in Figure 3.
It can be seen that with increasing LET there is a proportionately
less effect of H2 on the formation of H2O2. Virtually no effect
on H2O2 formation is found in H2 saturated solutions irradiated
with 5 MeV helium ions. This result is in agreement with the
predictions of the neat water systems based on the escape yields.
At the other extreme, low LET ions such as 10 MeV protons
behave much like γ radiolysis, with H2O2 showing a large
dependence on H2 concentration.
γ Radiolysis. The extensive literature on γ radiolysis helps
in predicting the long-time radiolysis of water. However, the
exact effects of H2 on H2O2 production have not been
determined. Mixed pile irradiations showed that the addition
of H2 to solutions of H2O2 would decrease the concentration of
the latter considerably.8 A series of γ irradiations were
performed in order to directly compare the results with those
obtained with heavy ions and to check the predictive ability of
the homogeneous model calculations. Figure 2 shows the
depletion of H2O2 in the γ radiolysis of water in the presence
and in the absence of added H2. Model calculations, see Figure
1, show that the limiting concentration of H2O2 in the γ
radiolysis of water is below 1 µM, which is below the lowest
detectable limit with the experimental techniques used in this
work. Therefore, to examine the effect of dissolved H2 in γ
radiolysis it is necessary to add a small amount of initial H2O2
(50 µM) so that it can react with the radical species H, OH,
and e-aq. The results in Figure 2 show that there is a net decrease
of H2O2 in γ radiolysis and more is consumed in the presence
of H2 than in neat water. This result is due to the scavenging of
the OH radicals by H2 and propagation of the chain reaction.
Simulations of the γ radiolysis of water were performed using
the homogeneous model with the escape yields given in Table
3. It can be seen that the model can reproduce well the results
with and without added H2. In only the lowest H2 concentration
is the high dose limit of H2O2 concentration great enough to be
experimentally determined. The limiting values for H2O2 with
the other two concentrations of H2 are too low to be measured,
so the accuracy of the model in those situations cannot be
examined. The model predicts that the initial H2O2 concentration
would have to be increased by about 2 orders of magnitude
before the low limit of H2O2 with H2 saturated solutions can be
measured using the present experimental technique. Such a large
H2O2 concentration will perturb the track chemistry and was
not examined.
Heavy ion radiolysis experiments were also performed with
added H2O2 in order to examine its decomposition due to
radicals escaping the track. Figure 4 shows the results for the
observed H2O2 concentration as a function of dose for solutions
initially 50 µM in H2O2. The two highest LET particles, 5 MeV
helium ions and 2 MeV protons, show increasing H2O2
concentrations with increasing dose. For both of these ions, the
rate of increase is less than expected from the escape yields,
but clearly, the H2O2 is being formed at a greater rate than it is
being consumed by radical reactions. The 10 MeV proton results
show a slight decrease in H2O2 concentrations with increasing
dose, which more closely resembles that observed with γ

Effect of Molecular Hydrogen on Hydrogen Peroxide
J. Phys. Chem. A, Vol. 105, No. 40, 2001 9321
the application of Monte Carlo track codes may lead to a
solution to this problem.
The addition of H2 gives a noticeable decrease in the
experimental values for H2O2 concentrations at the lowest LET.
However, the simulations predict a much larger effect for all
of the particles. The dotted line in Figure 4 shows the predicted
H2O2 dependence for the helium ion radiolysis of solutions
saturated with H2. The observed results are far from the
predicted ones. No reason for this discrepancy is known.
Previous results using tritium recoil particles showed little effect
on H2O2 concentrations up to about 3 atm of H2 overpressure.7
Those experiments involved the neutron radiolysis of LiOH
solutions to give the tritium recoil. The results were interpreted
in terms of oxidation/reduction reactions of a metal impurity in
the LiOH. No metals are expected in the systems examined here.
It is equally unclear how an impurity would affect the heavy
ion results and not the γ radiolysis.
Figure 4. H2O2 concentration dependence on dose in the heavy ion
radiolysis of 50 µM H2O2 solutions with no H2 (closed symbols) and
H2 saturated (800 µM; open symbols): ([) 5 MeV He, (9) 2 MeV H,
and (b) 10 MeV H. The solid lines are the homogeneous model
predictions for solutions with no H2, and the dashed line is for H2
saturated helium ion radiolysis.
Equilibrium conditions, i.e., the long time steady-state
concentrations, are reached with doses of about 50 krad in the
γ radiolysis of H2O2 solutions. On the other hand, hundreds of
Mrad of energy are required to reach the steady state limits with
the heavy ions. Unfortunately, only the steady-state results at
1% H2 (8 µM) were examined in γ radiolysis because of the
lower limit of detection of H2O2. It is possible that several
reactions are leading to a coupled equilibrium in which one or
more reactions are incorrect in the model. The buildup of an
oxidizing species, O2 for instance, can also give the observed
effects. High concentrations of an oxidizing species will
scavenge H atoms and terminate the chain reaction. No
experimental determination of a large O2 yield in heavy ion
radiolysis has been made, but future experiments will examine
this possibility. Of course, the results could be due to an
experimental artifact caused by the depletion of the H2.
However, the H2 effect is occurring in the homogeneous regime
and the bulk H2 adsorbed in the water is not likely to be depleted
in the present experimental configuration because H2 is bubbled
through the solution continuously during the irradiation. De-
creasing the beam current by over 1 magnitude also had no
effect on the observed results. It is also possible that a
homogeneous model cannot be used for predicting long-time
chemistry with some irradiations because of the overlap with
the track chemistry regime. Further detailed Monte Carlo
calculations are in progress to examine this possibility. Clearly,
the homogeneous model must be used with caution in practical
applications involving high LET radiations.
radiolysis. The relative decrease in H2O2 concentrations with
increasing dose is much less for 10 MeV protons (36%) than
with γ radiolysis (81%). The track average LET with 10 MeV
protons is about 14 eV/nm, and sufficient amounts of radicals
are escaping the track to lead to a net decrease in H2O2
concentration. The track average LET of a 2 MeV proton is
only about 35 eV/nm, which is above the threshold where H2O2
is formed at a rate greater than it is consumed. The LET at
which H2O2 production is about the same as consumption is
estimated to be about 20 eV/nm, which corresponds to a 5 MeV
proton. This is a critical LET, and below it, there are insufficient
amounts of radicals escaping the track to carry the propagation
step of the chain reaction.
The simulations of the results obtained with heavy ions are
shown as the solid lines in Figures 3 and 4. The homogeneous
model and the escape yields listed in Table 3 were used. A
first estimation of the escape yields was made using radiation
chemical yields in the literature and material balance.5,17,26-30
The escape yields were varied until the predicted results matched
the measured values for neat water. Figure 3 shows that a very
good agreement to the data is obtained with the escape yields
listed in Table 3. For those species that have been measured,
the escape yields used here are within a few percent of the values
given in the literature. The agreement with those values suggests
that the escape yields used in the model calculations well
represent the true yields. The predictions of the model are
reasonably sensitive to the escape yields chosen, but the use of
a deterministic model may be in doubt, see below. Therefore,
experiments measuring the escape yield for more ions at
different energies are desirable.
Good agreement between the model predictions and experi-
ments is found for the helium ion radiolysis of water with added
H2O2. However, the agreement is poor for the proton results
with aqueous solutions of H2O2, especially at the higher doses.
The same escape yields for each of the heavy ions was used in
the calculations with and without added H2O2. The fastest
reaction of H2O2 is with the hydrated electron, and at 50 µM,
the lifetime of this reaction is about 2 µs, which is too long to
have a significant effect on the track chemistry. It appears that
there is a problem matching the long-time chemistry of protons.
The problem could be in the application of a homogeneous
model to a system that has a long transition from the track
chemistry to the homogeneous phase. It is also possible
equilibrium processes that are not accurately described by the
homogeneous model are dominating the long-time chemistry
of protons. Selective examination of a number of reactions did
not reveal any particular one in error. Further experiments and
The major reactions that may be causing problems in the
model calculations are the slow reactions, especially in the
solutions with added H2. The only reaction of H2 is with OH
radicals (reaction 50). The H atom produced from this reaction
is then required to react with H2O2 (reaction 42) to propagate
the chain. This latter reaction is in competition with H atom
reaction with water (reaction 36). The rate coefficient for the
later reaction is very low. Its value was obtained by a model fit
to a series of reactions and could be in significant error.31
However, it is not understood why this reaction would cause a
problem with the heavy ion radiolysis and not the γ radiolysis.
The good results with modeling the γ radiolysis suggests that
the kinetics assumed in the model is basically sound.
One of the goals of this work was to provide information on
the radiation chemistry that pertains to reactors, which primarily
involves neutrons and γ rays. In most reactors, the neutron
energy distribution, and the subsequent proton recoils, is lower
in energy than the 10 MeV protons examined here. The lowest
practical proton energy capable with the present facility is about

9322 J. Phys. Chem. A, Vol. 105, No. 40, 2001
Pastina and LaVerne
2 MeV because of excessive energy straggling at lower energies.
Interpolation between the results with 2 and 10 MeV protons
can give predictions of the radiation chemistry at other proton
energies. However, any realistic application to reactors will
involve the use of models, and the results here are ideally suited
for model development.
Acknowledgment. The authors thank Professor J. J. Kolata
for making the facilities of the Notre Dame Nuclear Structure
Laboratory available. The latter is funded by the National
Science Foundation. 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-4310 from the Notre Dame Radiation
Laboratory, which is supported by the Office of Basic Energy
Sciences of the U. S. Department of Energy.
In the presence of a mixed field radiation, such as that in the
cooling circuit of a power reactor, the effects of low and high
LET radiation are simultaneous. The γ component of the
radiation may supply the necessary radical concentrations to
carry the chain reaction and destroy the molecular products
formed by the high LET component. Under the present
conditions, the critical LET corresponding to the threshold of
water decomposition appears to be near 20 eV/nm, which
corresponds to 5 MeV protons. Below this threshold, essentially
all of the products formed by radiolysis, including hydrogen
peroxide, are recombined by the radicals through the chain
reaction. This chain reaction is accelerated by the dissolved
molecular hydrogen. Above this critical LET, the water decom-
poses to form H2, H2O2, and O2 because not enough radicals
are escaping into the bulk solution. Obviously, the relative
fraction of γ ray to high LET radiation in the primary coolant
of nuclear reactors, as well as other factors such as temperature
and impurities, will determine if water is protected by decom-
position under normal reactor operating conditions. In case of
an accident or other mishap resulting in a local accumulation
of H2O2 or O2, the water reformation would stop even in the
References and Notes
(1) Allen, A. O. The Radiation Chemistry of Water and Aqueous
Solutions; Van Nostrand: New York, 1961.
(2) Draganic, I. G.; Draganic, Z. D. The Radiation Chemistry of Waters
Academic Press: New York, 1971.
(3) Buxton, G. V. In Radiation Chemistry: Principles and Applications;
Farhataziz., Rodgers, M. A. J., Eds.; VCH Publishers: New York, 1987; p
321.
(4) Pimblott, S. M.; LaVerne, J. A. J. Phys. Chem. A 1997, 101, 5828.
(5) Elliot, A. J.; McCracken, D. R. Fusion Eng. Des. 1990, 13, 21.
(6) Christensen, H. Nucl. Technol. 1995, 109, 373.
(7) Elliot, A. J.; Chenier, M. P. J. Nucl. Mater. 1992, 187, 230.
(8) Pastina, B.; Isabey, J.; Hickel, B. J. Nucl. Mater. 1999, 264, 309.
(9) Cowan, R. L. Water Chemistry of Nuclear Reactor Systems 7;
British Nuclear Energy Society: London, 1996.
(10) BWR Water Chemistry Guidelines, 1993 ReVision: Normal and
Hydrogen Water Chemistry, Electric Power Research Institute, Report TR-
103515, 1994.
(11) Allen, A. O.; Hochanadel, J.; Ghormley, J.; Davis, T. W. J. Phys.
Chem. 1952, 56, 575.
(12) Pastina, B. Ph.D. Thesis, Universite´ Paris-Sud, Paris, France, 1997;
no. 4862.
1
presence of γ radiation only. An excess of H2 will minimize
the effects with γ rays but not the high LET radiation
component.
(13) LaVerne, J. A. Radiat. Res. 2000, 153, 487.
(14) LaVerne, J. A.; Schuler, R. H. J. Phys. Chem. 1987, 91, 5770.
(15) LaVerne, J. A.; Schuler, R. H. J. Phys. Chem. 1987, 91, 6560.
(16) Ziegler, J. F.; Biersack, J. P.; Littmark, U. The Stopping Power
and Range of Ions in Solids; Pergamon: New York, 1985.
(17) Pastina, B.; LaVerne, J. A. J. Phys. Chem. A 1999, 103, 1592.
(18) Ghormley, J. A.; Stewart, A. C. J. Am. Chem. Soc. 1956, 78, 2934.
(19) Hochanadel, C. J. J. Phys. Chem. 1952, 56, 587.
(20) Schwarz, H. A. J. Phys. Chem. 1969, 73, 1928.
(21) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J.
Phys. Chem. Ref. Data 1988, 17, 513.
(22) Elliot, A. J.; Buxton, G. V. J. Chem. Soc., Faraday Trans. 1992,
88, 2465.
(23) Mezyk, S. P.; Bartels, D. M. J. Chem. Soc., Faraday Trans. 1995,
91, 3127.
(24) Chance, E. M.; Curtis, A. R.; Jones, I. P.; Kirby, C. R. Report
AERE-R 8775; AERE: Harwell, 1977.
(25) Burns, W. G.; Sims, H. E.; Goodall, J. A. B. Radiat. Phys. Chem.
1984, 23, 143.
(26) Schwarz, H. A.; Caffrey, J. M.; Scholes, G. J. Am. Chem. Soc.
1959, 81, 1801.
(27) LaVerne, J. A.; Schuler, R. H.; Burns, W. G. J. Phys. Chem. 1986,
90, 3238.
(28) LaVerne, J. A. Radiat. Res. 1989, 118, 201.
(29) LaVerne, J. A.; Yoshida, H. J. Phys. Chem. 1993, 97, 10720.
(30) LaVerne, J. A.; Pimblott, S. M. J. Phys. Chem. A 2000, 104, 9820.
(31) Hartig, K. T.; Getoff, N. J. Photochem. 1982, 18, 29.
Conclusions
A quantitative insight into the effects of the track structure
of different heavy ions on the yields of radicals and molecular
products from the decomposition of water has been obtained.
These experiments involved the radiolysis of a number of
aqueous solutions containing various concentrations of added
H2 and H2O2. H2 is an OH radical scavenger and is used to
probe the escape yields of radicals in the presence of an initial
small concentration of H2O2. From the experimental results, it
appears that the track structure of 10 MeV protons is somewhat
similar to that for γ radiation with a slightly smaller fraction of
radicals escaping into the bulk water. Higher LET particles have
very low radical escape yields in pure water leading to a net
decomposition of bulk water. The threshold for this process is
about 20 eV/nm (a 5 MeV proton). The addition of H2 promotes
the recombination of radicals to water for the lowest LET
radiation. However, experimentally, no effect due to added H2
is observed at the highest LET. These results are contradictory
to homogeneous model predictions.