Long-term γ-radiolysis kinetics of NO3- and NO2- solutions

Article abstract of DOI:10.1021/jp200262c

Radiolysis kinetics in NO₃^- and NO₂ ^- solutions during γ-irradiation were studied at an absorbed dose rate of 2.1 Gy·s¹ at room temperature. Air- or argon-saturated nitrate or nitrite solutions at pH 6.0 and 10.6 were irradiated, and the aqueous concentrations of molecular water decomposition products, H₂ and H₂O₂, and the variation in the concentrations of NO₃^- and NO₂^- were measured as a function of irradiation time. The experimental data were compared with computer simulations using a comprehensive radiolysis kinetic model to aid in interpretation of the experimental results. The effect of nitrate and nitrite, present at concentrations below 10^-3 M, on water radiolysis processes occurs through reactions with the radical species generated by water radiolysis, ^?e_aq^-, ^?O ₂^-, and ^?OH. The changes in H₂ and H₂O₂ concentrations observed in the presence of nitrate and nitrite under a variety of conditions can be explained by a reduction in the radical concentrations. The kinetic analysis shows that the main loss pathway for H₂ is the reaction with ^?OH and the main loss pathways for H₂O₂ are reactions with ^?e_aq^- and ^?OH. Nitrate and nitrite compete for the radicals leading to an increase in the concentrations of H₂ and H₂O₂. Post-irradiation measurements of H₂, H₂O₂, NO₂^- and NO₃^- concentrations can be used to calculate the radical concentrations and provide information on the redox conditions of the irradiated aqueous solutions.

Full text of DOI:10.1021/jp200262c

ARTICLE
pubs.acs.org/JPCA
ꢀ
ꢀ
Long-Term γ-Radiolysis Kinetics of NO3 and NO2 Solutions
Pamela A. Yakabuskie,† Jiju M. Joseph,† Craig R. Stuart,‡ and J. Clara Wren*,†
†Department of Chemistry, the University of Western Ontario, London, Ontario, Canada, N6A 5B7
‡Reactor Chemistry and Corrosion Branch, Atomic Energy of Canada Ltd., Chalk River, Ontario, Canada, K0J 1P0
ABSTRACT: Radiolysis kinetics in NO3ꢀ and NO2ꢀ solutions during
γ-irradiation were studied at an absorbed dose rate of 2.1 Gy sꢀ1 at
3
room temperature. Air- or argon-saturated nitrate or nitrite solutions at
pH 6.0 and 10.6 were irradiated, and the aqueous concentrations of
molecular water decomposition produꢀcts, H2 and H2O2, and the
ꢀ
variation in the concentrations of NO3 and NO2 were measured
as a function of irradiation time. The experimental data were compared
with computer simulations using a comprehensive radiolysis kinetic
model to aid in interpretation of the experimental results. The effect of
nitrate and nitrite, present at concentrations below 10ꢀ3 M, on water radiolysis processes occurs through reactions with the radical
species generated by water radiolysis, •eaqꢀ, •O2ꢀ, and •OH. The changes in H2 and H2O2 concentrations observed in the presence
of nitrate and nitrite under a variety of conditions can be explained by a reduction in the radical concentrations. The kinetic analysis
shows that the main loss pathway for H2 is the reaction with •OH and the main loss pathways for H2O2 are reactions with •eaqꢀ and
•OH. Nitrate and nitrite compete ꢀfor the radicals leading to an increase in the concentrations of H2 and H2O2. Post-irradiation
ꢀ
measurements of H2, H2O2, NO2 and NO3 concentrations can be used to calculate the radical concentrations and provide
information on the redox conditions of the irradiated aqueous solutions.
chemistry driven by water radiolysis during long duration irradiation.
Exposed to ionizing radiation, water decomposes to form a range of
chemically reactive species from highly oxidizing (•OH, H2O2) to
highly reducing (•eaqꢀ, •H, •O2ꢀ). The speciation of water radiolysis
products changes rapidly (on a microsecond scale) following an
initial interaction of the radiation particle (or photon) with the water
molecules.4 Under continuous irradiation, water decomposition
products are continuously produced and these products undergo
reactions with each other, with hydrolysis products (such as Hþ and
OHꢀ), and, if present, with solutes and surfaces. Due to fast kinetics
of the radical reactions, a system reaches a pseudo-steady state
(maintained by the continuous radiation source) on a time scale on
the order of milliseconds (note that this is well beyond the
submicrosecond time scale typically employed in pulse radiolysis)
and then more slowly reaches a true steady state controlled by the
presence of slow reactions and interfacial mass transport rates. The
steady-state concentrations of the radiolytically produced species
determine the aqueous redox conditions that control corrosion or
colloid formation from dissolved corrosion products.2,5ꢀ7
1. INTRODUCTION
Many of the current operational and safety issues associated with
nuclear power plants are related to materials degradation. Issues
include the potential corrosion of the steel piping used in the reactor
coolant system or other equipment used to contain radioactive
materials. An ability to accurately monitor and control the redox
state of the coolant could support strategies to reduce corrosion
rates and increase the lifetime of the piping. For example, careful
control of the redox conditions of the primary heat transport system
during plant outages and start-up to remove oxidants prior to raising
the coolant temperature to above 150 °C minimizes the risk of
degradation of system materials.1
Associated with corrosion is the formation and transport of
dissolved or particulate corrosion products. As the reactor coolant
circulates, temperature changes can cause dissolved metal ions and
particulates to deposit on system surfaces. Corrosion products
transported into the reactor core, or formed in situ in the case of
zirconium alloys, can be neutron activated and converted to radio-
active isotopes (e.g., 54Mn, 58Co, 59Fe, 60Co, and 95Zr). Recent
studies on the radiolysis of slightly acidic aqueous solutions containing
dissolved ferrous ions, Fe2þ(aq), have found that uniform-sized
colloidal γ-FeOOH particles can be formed.2 These colloidal particles
can aggregate to form larger particulates. The sizes of the primary
particles and the aggregates will depend on the chemical composition
and phase structure of the particles and the aqueous redox
conditions.3 The production of particulates has important implica-
tions for the design of water filtration/purification systems, particle
deposition on surfaces, and radioactivity transport in reactor systems.
Optimizing aqueous chemistry in environments where ionizing
radiation fields are present requires a detailed understanding of the
The steady-state concentrations that develop during continuous
γ-irradiation are strongly affected by pH and the presence of
dissolved species (even at impurity levels less than 10ꢀ3 M) if the
dissolved species can react readily with water radiolysis products in a
pseudocatalytic manner. Due to the many potential oxidation states
þ
ꢀ
of N, ranging from ꢀ3 in NH3/NH4 to þ5 in NO3 under
ambient conditions, nitrogen-containing compounds have a
Received: January 10, 2011
Revised:
February 28, 2011
Published: April 06, 2011
r
2011 American Chemical Society
4270
dx.doi.org/10.1021/jp200262c J. Phys. Chem. A 2011, 115, 4270–4278
|

The Journal of Physical Chemistry A
ARTICLE
potential to interact pseudocatalytically with water radiolysis pro-
ducts. For example, NO3ꢀ can react with reducing species (such as
•eaqꢀ, •H, and •O2ꢀ) to form NO2ꢀ and the NO2ꢀ can then react
with oxidizing species (such as •OH and H2O2) to regenerate the
NO3ꢀ. Such a series of reactions can affect the balance of the redox-
active species in solution and result in significant changes in their
steady-state concentrations.
Nitric acid is used in spent reactor fuel processing and waste
treatment, and hydrazine and ammonia are used as pH control-
ling agents or corrosion inhibitors in nuclear facilities. Degrada-
tion of ion exchange resins used for water purification in coolant
systems or unintended air ingress may inadvertently introduce
nitrogen compounds into the coolant. When such an event
occurs, measurement of the nitrogen species present may be
used to monitor the coolant redox state.
Table 1. G Value, Reactions and Rate Constants, and
Equilibria Used in the Model Calculations
G Values (Units of Molecules/100 eV)
reaction no.
formation reaction
G value
ꢀ
R1a
R1b
R1c
R1d
R1e
R1f
H2O f •eaq
2.6
2.6
H2O f Hþ
H2O f H•
H2O f •OH
0.6
2.7
H2O f H2
0.45
0.7
H2O f H2O2
Reactions and Rate Constants
(mꢀ1 ꢀ1 or sꢀ1
s )
3
Our understanding of the radiolysis of nitrate and nitrite in
water in a continuous radiation flux is not clear, even after several
years of investigation.8ꢀ12 Better knowledge of the radiation-
induced decomposition of nitrate and nitrite is important
because this process occurs not only in nuclear systems but also
in a wide range of fields such as atmospheric chemistry, UV
photolysis, and high-temperature combustion.13 In particular,
studies of high-flux, continuous radiolysis behavior are very rare.
The focus of this work is to identify the set of reactions that
control the nitrogen and water radiolysis product speciation
during long-term irradiation and to determine how the speciation
is affected by the initial solution pH.
chemical reaction
rate constant (25 °C)
•
R2
•eaqꢀ þ eaqꢀ þ 2H2O f H2 þ 2OHꢀ
5.5 ꢁ 109
2.5 ꢁ 1010
3.0 ꢁ 1010
2.22 ꢁ 1010
1.6 ꢁ 1010
1.3 ꢁ 1010
3.5 ꢁ 109
2.2 ꢁ 1010
7ꢁ 109
•
R3
•eaqꢀ þ H þ H2O f H2 þ OHꢀ
•
R4
•eaqꢀ þ OH f OHꢀ
ꢀ
R5
•eaqꢀ þ O2 f •O2
•
R6
•eaqꢀ þ H2O2 f OHꢀ þ OH
•eaqꢀ þ HO2• f HO2
ꢀ
R7
R8
•eaqꢀ þ HO2ꢀ f •Oꢀ þ OHꢀ
•
R9
•eaqꢀ þ Oꢀ þ H2O f OHꢀ þ OHꢀ
•
R10
R11
R12
R13
R14
R15
R16
R17
R18
R19
R20
R21
R22
R23
R24
R25
R26
R27
R28
R29
R30
R31
R32
R33
R34
R35
•H þ OH f H2O
•H þ • H f H2
7.75 ꢁ 109
2.1 ꢁ 1010
1.0 ꢁ 1010
9.0 ꢁ 107
2.0 ꢁ 1010
1.1 ꢁ 101[H2O]
5.5 ꢁ 109
2.7 ꢁ 107
4.2 ꢁ 107
8ꢁ 109
•H þ O2 f HO2
•
•H þ HO2• f H2O2
2. EXPERIMENTAL SECTION
•H þ H2O2 f •OH þ H2O
•
ꢀ
•H þ O2ꢀ f HO2
All solutions were freshly prepared before each experiment
•
•H þ H2O f H2 þ OH
with water purified to a resistivity of 18.2 MΩ cm using a
3
•
•OH þ OH f H2O2
NANOpure Diamond UV ultrapure water system. The experi-
ments at pH 6.0 were performed without the addition of a buffer
and those at pH 10.6 used a phosphate buffer (10ꢀ3 M). The pH
of the solution was measured prior to, and at the end of, the
irradiation period using a pH meter (Accumet). High-purity
sodium nitrate and sodium nitrite were obtained from Sigma-
Aldrich (purity g99%) and used without further purification.
Deaerated solutions were prepared by purging the bulk solution
with ultra-high-purity argon (Praxair, total impurities <0.001%) for
more than one hour. The solution was then transferred into
individual 20 mL vials (Agilent Technologies) and sealed, using
aluminum crimp caps with PTFE silicone septa, in an argon-filled
glovebox where the oxygen level was maintained below 1000 ppm.
The irradiation was carried out in a 60Co gamma cell (MDS
Nordion) which provided the irradiation chamber with a uniform
•
•OH þ H2O2 f HO2 þ H2O
•OH þ H2 f H• þ H2O
•
ꢀ
•OH þ O2ꢀ f OHꢀ þ O2
•OH þ HO2• f H2O þ O2
6ꢁ 109
•OH þ HO2ꢀ f HO2 þ OHꢀ
7.5 ꢁ 109
2.0 ꢁ 1010
2ꢁ 108
•
•
ꢀ
•OH þ Oꢀ f HO2
•Oꢀ þ H2O2 f •O2ꢀ þ H2O
•Oꢀ þ H2f H• þ OHꢀ
8.0 ꢁ 107
4.0 ꢁ 108
6.0 ꢁ 108
3.8 ꢁ 109
3.0 ꢁ 102
1.6 ꢁ 106
8.9 ꢁ 105
2.5 ꢁ 105
8.9 ꢁ 107
2.0 ꢁ 106
2.3 ꢁ 10ꢀ7
•Oꢀ þ HO2ꢀ f •O2ꢀ þ OHꢀ
•Oꢀ þ O2ꢀ f 2OHꢀ þ O2
•Oꢀ þ O2 f •O3
ꢀ
•
•O3ꢀ f O2 þ Oꢀ
•O3ꢀ þ H2O2ꢀ f •O2ꢀ þ O2 þ (H2O)
•O3ꢀ þ HO2ꢀf •O2ꢀ þ O2 þ OHꢀ
•O3ꢀ þ H2 f O2 þ H• þ OHꢀ
absorption dose rate of 2.1 Gy sꢀ1 determined using Fricke
3
dosimetry.4,14 Individual vials were removed from the irradiation
chamber at regular time intervals to perform analysis.
•
•
HO2 þ O2ꢀ f HO2ꢀ þ O2
HO2 þ HO2• f H2O2 þ O2
•
•
H2O2 f •OH þ OH
The gaseous H2 concentration was determined using a gas
chromatograph (GC-MS, 6580 Agilent Technologies) equipped
with a thermal conductivity detector (TCD), technique detailed in
ref 15. With this method, the detection limit for the aqueous H2
concentration was 1.0 ꢁ 10ꢀ5 M and the uncertainties in the
measurement arising from sampling and instrumental errors were
estimated to be (50% at the low end of the measured concentration
range and (0.005% at the high end of the concentration range.
All spectrophotometric measurements were performed using a
diode array UV spectrophotometer (BioLogic Science Instruments).
The concentration of hydrogen peroxide was determined by the
Ghormley triiodide method.16,17 From a calibration curve, the
Related Equilibria and the
Equilibrium Rate Constants
equilibria
Keq
R36
R37
R38
R39
R40
H
þ þ OHꢀ / H2O
1.78 ꢁ 10ꢀ16
1.6 ꢁ 10ꢀ12
H
þ þ HO2ꢀ / H2O2
•
H
þ þ O2ꢀ / HO2•
1.5 ꢁ 10ꢀ5
•
H
þ þ Oꢀ / •OH
1.6 ꢁ 10ꢀ12
•
H
þ þ eaqꢀ / •H
k40f = 2.25 ꢁ 1010
k40b = 0.63 ꢁ 101
4271
dx.doi.org/10.1021/jp200262c |J. Phys. Chem. A 2011, 115, 4270–4278

The Journal of Physical Chemistry A
ARTICLE
Table 2. Reactions of Nitrate and Nitrite Species with Water Radiolysis Products and Their Rate Constants Used in the Model
Calculations
chemical reaction
rate constant (25 °C)
reference
•eaqꢀ þ NO3ꢀ f NO3
9.7 ꢁ 109
5.5ꢁ 104
3.5 ꢁ 109
4.3ꢁ 104
5.0 ꢁ 109
1.0 ꢁ 106
6.5 ꢁ 107
1.0 ꢁ 1010
2.0 ꢁ 108
5.0 ꢁ 106
1.0 ꢁ 1010
6.7 ꢁ 109
2.3 ꢁ 1010
22
2ꢀ
R41
R42
R43
R44
R45
R46
R47
R48
R49
R50
R51
R52
R53
NO32ꢀ þ H2O f •NO2 þ 2OHꢀ
22, 23
24, 27
29
•eaqꢀ þ NO2ꢀ f NO2
2ꢀ
NO22ꢀ þ H2O f •NO þ 2OHꢀ
•OH þ NO2ꢀ f •NO2 þ OHꢀ
27
•O2ꢀ þ NO3ꢀ f NO32ꢀ þ O2
this worka
2•NO2 þ H2O f NO3ꢀ þ NO2ꢀ þ 2Hþ
21
•
ꢀ
•eaqꢀ þ NO2 f NO2
this workb
29
•
•O2ꢀ þ NO2 f NO2ꢀ þ O2
•O2ꢀ þ NO2ꢀ f NO22ꢀ þ O2
30
•
•OH þ NO f HNO2
27
•
•O2ꢀ þ NO f ONOOꢀ
25
•
•eaqꢀ þ NO f NOꢀ
26, 27
equilibria
pKa
reference
R54
R55
R56
R57
R58
H
H
þ þ NO3ꢀ f HNO3
þ þ NO2ꢀ f HNO2ꢀ
ꢀ1.6
3.4
28
28
22
24
28
H
H
þ þ NO32ꢀ f HNO3ꢀ
7.5
þ þ NO22ꢀ f HNO2
7.7
H
þ þ ONOOꢀ f ONOOH
6.8
a This reaction rate constant was not available in literature and was estimated to be on the order of 106 based on other available rate constants for
reactions of •O2ꢀ with molecular ionic species (e.g., R50). b This reaction rate constant was not available in literature and was estimated to be on the
order of 1010 based on other available rate constants for reactions of •eaqꢀ with radical species (e.g., R53).
detection limit for [H2O2] was determined to be 3 ꢁ 10ꢀ6 M. The
H2O2 analysis for samples containing NO2ꢀ was performed using a
reagent where the concentration of the KI was reduced from 0.40 to
0.078 M to minimize any potential interference caused by nitrite
species, as described in the literature.18
Table 1). This model has been shown to reproduce the steady-state
radiolysis data of liquid water over a wide range of conditions.15,22ꢀ24
The effects of dissolved nitrate and nitrite on the steady-state
concentrations of radiolysis products have been modeled by
incorporating an additional simplified set of reactions of nitrate
and nitrite species (13 reactions) with radiolysis products, and
associated equilibria (see Table 2). The rate constants for these
reactions were taken from the literature.25ꢀ34 The evolution of
these coupled reactions was solved using the software package
FACSIMILE. The nitrate/nitrite reaction set included in the
model spans nitrogen oxidation states from NO3ꢀ to NOꢀ. The
reactions of NOꢀ and further reactions that are associated with
forming nitrogen species with an oxidation state lower than þ2
are not included in the reaction set. These reactions occur at
much longer time scales than the durations of the experimental
study and, hence, will not significantly influence the radiolysis
kinetics observed in our experiments. Omission of these reac-
tions is supported by our ability to measure a mass balance of
stable nitrogen species (nitrate and nitrite) at nearly 100%
without including the lower oxidation state nitrogen species.
The concentration of dissolved nitrite was determined by the
Griess method.19,20 The sample was first treated with a diazotizing
reagent, sulfanilamide, in an acidic medium to form the intermediate
diazonium salt. The diazonium salt then reacts with the coupling
reagent N-(1-napthyl)-ethylenediamine to form the stable azo
compound which has a maximum UV absorption peak at 540 nm
with a molar extinction coefficient of 4.0 ꢁ 104 Mꢀ1 cmꢀ1 20
.
3
Calibration of this analysis method with known nitrite solutions
yielded a detection limit for [NO2ꢀ] of 3 ꢁ 10ꢀ6 M.
The concentration of nitrate in the sample solution was
determined by reducing the nitrate to nitrite using copper-coated
cadmium pellets.2ꢀ0,21 The total nitrite concentration formed after
reduction (NO3 þ NO2ꢀ) was determined by the above
method. The initial nitrate concentration was then calculated
as the difference between the nitrite concentration determined
following the reduction step reaction and the nitrite concentra-
tion determined prior to the reduction step.
4. RESULTS
3. RADIOLYSIS MODEL
4.1. Experimental Results. The concentrations of the molec-
ular products (H2 and H2O2) of ꢀradiolytic decomposition of water
and the nitrogen species (NO3 and NO2ꢀ) as a function of
irradiation time were measured in aqueous solutions initially con-
taining 10ꢀ3 M NO3ꢀ or 10ꢀ3 M NO2ꢀ. The results are shown in
Figure 1 for pH 6.0/deaerated water and Figure 2 for pH
10.6/deaerated water. For comparison, the results for pure water
free of nitrate and nitrite are shown in Figure 3. Also shown in these
figures are the results of computer simulations performed using the
The radiolysis kinetic model used in this study incorporates a
comprehensive reaction set for γ-radiolysis of pure water that has
been reported previously.15 This reaction set includes the primary
radiolysis production reactions (whose rates are determined by the G
values multiplied by the radiation dose rate), all of the possible
reactions of the radiolysis products with each other (∼ 40 elemen-
tary homogeneous reactions with well-defined rate constants),
hydrolysis reactions, and related acidꢀbase equilibria (see
4272
dx.doi.org/10.1021/jp200262c |J. Phys. Chem. A 2011, 115, 4270–4278

The Journal of Physical Chemistry A
ARTICLE
Figure 1. The time-dependent speciation of NO3ꢀ (red solid triangle down, red solid line) and NO2ꢀ (black solid triangle up, black solid line), the
overall mass balance of nitrogen species (blue open diamond, blue dash line), and the time-dependent concentrations of H2 (black solid square, black
solid line), H2O2 (red solid circle, red dash line), and O2 (green dotted line) at pH 6.0 deaerated condition are shown for aqueous solutions containing
(a) 1 ꢁ 10ꢀ3 M [NO3ꢀ]0 and (b) 1 ꢁ 10ꢀ3 M [NO2ꢀ]0. Data points represent the experimental data, and the lines are the model simulation results. The
total absorbed dose (irradiation time ꢁ DR) for the 5 h irradiation period was 37.8 kGy.
Figure 2. The time-dependent speciation of NO3ꢀ (red solid triangle down, red solid line) and NO2ꢀ (black solid triangle up, black solid line), the
overall mass balance of nitrogen species (blue open diamond, blue dash line), and the time-dependent concentrations of H2 (black solid square, black
solid line), H2O2 (red solid circle, red dash line), and O2 (green dotted line) at pH 10.6 deaerated condition are shown for aqueous solutions containing
(a) 1 ꢁ 10ꢀ3 M [NO3ꢀ]0 and (b) 1 ꢁ 10ꢀ3 M [NO2ꢀ]0. Data points represent the experimental data and the lines are the model simulation results. The
total absorbed dose (irradiation time ꢁ DR) for the 5 h irradiation period was 37.8 kGy.
radiolysis kinetics model described above. For the pure water case at
pH 6.0, the concentrations of H2O2 and H2 were below the
detection limit and therefore only the model predictions are shown
in the figure. The pH values of the sample solutions were also
measured before and after irradiation. Following the 5 h irradiation
period, the change in pH was less than half a pH unit for all samples.
4273
dx.doi.org/10.1021/jp200262c |J. Phys. Chem. A 2011, 115, 4270–4278

The Journal of Physical Chemistry A
ARTICLE
(particularly concerning the nitrite fitting results) could be attrib-
uted to the assumptions made during the modeling process. First,
the model fits were performed on a reduced overall reaction set. The
most reduced nitrogen species assumed to form in this model is
NOꢀ. Although a near 100% nitrogen mass balance was measured
for most experiments, which validates the use of a simplified reaction
set, it is possible that even small contributions from additional
reactions involving species of lower nitrogen oxidation state may
change the rate of achieving steady state and affect the model fits.
Second, the modeling calculations were also performed by holding
the pH of the solution constant throughout the irradiation period;
hence, the pH was set as a constant parameter. This was deemed to
be an appropriate assumption based on the consistency of the
experimentally measured pH values ((0.5 pH unit) before and after
irradiation. However, some variations observed between the model
simulations and the experimental data may be partially attributed to
the ignorance of slight pH changes during the modeling.
Figure 3. The time-dependent concentrations of H2 (black solid
square, black solid line), H2O2 (red solid circle, red dash line), and O2
(green dotted line) under deaerated condition at pH 6.0 (bottom) and
pH 10.6 (top). Data points represent the experimental data and the lines
are the model simulation results. At pH 6.0, the measured concentra-
tions of H2 and H2O2 were below the detection limits of 1 ꢁ 10ꢀ5 and 3
ꢁ 10ꢀ6 M, respectively. The total absorbed dose (irradiation time ꢁ
DR) for the 5 h irradiation period was 37.8 kGy.
5. DISCUSSION
5.1. Radiolysis in the Absence of Nitrate/Nitrite. Previous
studies have established the key reactions that determine the
concentrations of the molecular water radiolysis products, H2
and H2O2.5,15,22 Amongthe ∼40 elementary reactions in Table 1,
the key aqueous reactions that control the concentrations of H2
and H2O2 are
ꢀ
For all of the conditions studied, the addition of NO3 or
NO2ꢀ increases the radiolytic production of both reducing (H2)
and oxidizing (H2O2 and O2) molecular products, relative to the
pure water cases. The addition of nitrate ions increases the H2O2
production more than the addition of nitrite ions, while the
addition of nitrite increases the H2 production more than the
addition of nitrate. However, the differences between the mo-
lecular product concentrations in nitrite and nitrate solutions are
small; less than a factor of 2 for all of the tested conditions. In
pure water, the concentrations of H2 and H2O2 reached at longer
times are a strong function of pH.5,15,22 On the other hand, in a
solution containing a given nitrogen compound (nitrate or
nitrite), the concentrations are nearly independent of pH.
The nitrogen speciation (Figures 1 and 2) shows very different
kinetic behavior depending on whether the initial nitrogen species is
NO3ꢀ or NO2ꢀ. Nevertheless, nitrogen speciation at longer times is
nearly independent of the initial nitrogen species but strongly
dependent upon pH. An increase in the initial level of dissolved
oxygen appears to accelerate the rate of conversion between the
nitrate and nitrite (i.e., steady state is reached faster) but has a
negligible impact on the resulting steady-state speciation.
For H2
H2O f H2
ðR1eÞ
ðR19Þ
•
•
H2 þ OH f H2O þ H
For H2O2
H2O f H2O2
ðR1fÞ
ðR6Þ
ꢀ
•
•
H2O2 þ eaq f OH þ OHꢀ
•
•
H2O2 þ OH f H2O þ HO2
ðR18Þ
where (R1e) and (R1f) represent the primary homogeneous
production by radiolysis and the other reactions are aqueous
chemical reactions (reaction numbers are taken from Table 1).
Except at very short time scales (less than milliseconds), the
net changes in the concentrations of these species are relatively
small. The steady-state approximation can be thus used to obtain
analytical solutions for the rate equations. Using only the main
reactions listed above, the analytical solutions provide the con-
centrations for the molecular species as a function of those of the
radical species15
4.2. Model Simulation Results. The main objective of the
model simulation is to follow the kinetic behavior and concen-
trations of radical species that are difficult (practically im-
possible) to monitor directly. The results of model simulations
are used in interpreting the overall radiolysis kinetics.
The computer simulation results are in good agreement with the
experimental data over a wide range of conditions; the model
predictions are within a factor of 2 of the observed data for both the
time to reach steady state and the steady-state concentrations. Such
agreement is remarkable considering that the radiolysis kinetic
model consists of such a large number of different reactions for
the radiolysis of pure water plus an additional 13 reactions of water
radiolysis products with nitrogen compounds, and we have limited
confidence in the rate constants for some of the reactions of
nitrogen species. This is particularly true for those slower reactions
of nitrogen species with the more “stable” radicals in the system such
as •O2ꢀ. We plan to use the results from this work and future studies
to further refine the range of some of those rate constants.
H2
rad
k
½H2ꢂt ꢃ
ðE1Þ
ðE2Þ
kR19½•OHꢂ
t
H2O2
rad
k
½H2O2ꢂt ꢃ
kR6½•eaqꢀꢂ þ kR18½•OHꢂ
t
t
H2 O2
rad
where kHrad2 ≈ CRGH DRFH O and k
≈ CRGH O DRFH O
represent the zeroth 2order 2radiolytic production rates for H2
and H2O2. These rates are determined by the primary radiolysis
yields (the G values), multiplied by the radiation dose rate (DR)
2
2
2
Furthermore, the slight discrepancies between the model-pre-
dicted concentrations and the experimentally measured values
and the density of water (FH O) and using a unit conversion
2
4274
dx.doi.org/10.1021/jp200262c |J. Phys. Chem. A 2011, 115, 4270–4278

The Journal of Physical Chemistry A
ARTICLE
factor (CR). The other rate constants, kR6, kR18, and kR19, are the
second-order rate constants.
than that of the hydrated electron, since the reaction rate constants
for the reactions of nitrate and nitrite with •H are approximately 2
orders of magnitude smaller than those with •eaqꢀ. Furthermore, the
These analytical solutions state that the concentrations of H2
and H2O2 are inversely proportional to the radical concentra-
tions. These inverse relationships were confirmed by the com-
putational results, and good agreement with the data also
confirms that the main reactions that control [H2] and [H2O2]
are (R1e) and (R19) and (R1f), (R6), and (R18), respectively.15
For the radical radiolysis products, primary radiolysis (R1) is
also the main production pathway but the main removal paths are
reactions with molecular species:
pseudo-steady-state concentration of •H is not high enough to drive
ꢀ 15
the reactions to comparable rates as those involving •eaq
.
Introduction of other chemical species into water that can react
with the radical species can upset the chemistry by providing
additional removal pathways for those radicals.35 If the added species
can be regenerated by back reactions such that they can establish a
(semi-) pseudocatalytic cycle with the radical species before they are
totally consumed, their influence on radiolysis product concentra-
tions can be significant and last for a long time. Aqueous phase
reactions of a chemical additive with the molecular water radiolysis
products are typically too slow to be important. However, dissolved
additives can affect the molecular product concentrations indirectly,
via their interactions with radical species.
For •OH
•
H2O f OH
ðR1dÞ
ðR18Þ
ðR20Þ
•
•
H2O2 þ OH f H2O þ HO2
•O2ꢀ þ OH f O2 þ OHꢀ
5.2. Radiolysis of Aqueous Solutions Initially Containing
Nitrate or Nitrite.
•
5.2.1. Prꢀoduction ofꢀH2 and H2O2. Under all conditions studied,
both NO3 and NO2 increase the radiolytic production of H2
and H2O2 compared to that seen for pure water (Figures 1ꢀ3).
Dissolved nitrogen species can provide several additional reaction
pathways for radical radiolysis products, with the key reactions being
For •eaq
ꢀ
•
ꢀ
H2O f eaq
ðR1aÞ
ðR40Þ
ðR6Þ
•
ꢀ
•
Hþ þ eaq a H
•eaqꢀ þ NO3 f ½•NO32ꢀ þ H2Oꢂ f NO þ 2OHꢀ
ꢀ
•
2
•
ꢀ
•
H2O2 þ eaq f OH þ OHꢀ
ðR41, R42Þ
ꢀ
ꢀ
•
•
þ NO2 f ½•NO22ꢀ þ H2Oꢂ f NO þ 2OHꢀ
ꢀ
•eaq
ꢀ
•
O2 þ eaq f O2
ðR5Þ
Note that although (R19) is an important removal path for H2, it is
ðR43, R44Þ
not important for the removal of •OH because reactions R18 and
•OH þ NO2 f NO2 þ OHꢀ
ðR45Þ
ꢀ
•
R20 are faster. Reaction R19 is only important if H2 is present at a
ꢀ
high concentration prior to the buildup of H2O2 and •O2
.
The interaction between the molecular and the radical radiolysis
products leads to regeneration of some species that can establish a
pseudocatalytic reaction cycle with the radical species. This changes
the relative contributions of different reactions to the net removal of
the radicals as water radiolysis progresses, making it difficult to
predict the net radiolysis production at longer times based on the
rates of elementary reactions.5,15 For example, for initially deaerated
pure water at pH 6.0, the concentrations of the molecular products
are below the detection limits and the model simulations also predict
values lower than the detection limits as shown in Figure 3. The
molecular concentrations remain low at pH 8.5 (results not
shown).15 However, at pH 10.6 the molecular product concentra-
tions are about 3 orders of magnitude larger (Figure 3). These
dramatic changes in the molecular product concentrations with pH
occur in the pH range 9ꢀ11. At pH 10.6, a pH higher than the pKa of
•H of 9.7 (•H a •eaqꢀ þ Hþ), (R40) is very slow and the reaction
with secondary radiolysis product, O2 (reaction R5) becomes an
•O2ꢀ þ NO3 f ½•NO32ꢀ þ H2Oꢂ f NO þ 2OHꢀ þ O2
ꢀ
•
2
ðR46, R42Þ
2•NO2 þ H2O f NO3ꢀ þ NO2ꢀ þ 2Hþ
ðR47Þ
Nitrate and nitrite ion reactions (R41ꢀR45) can compete
•
with (R18) and (R20) for OH and with (R5), (R6), and
R40 for •eaqꢀ and, at sufficiently high concentrations, can become
the dominant removal paths for the radicals. Addition of nitrate/
•
ꢀ
•
nitrite will lower the net concentrations of eaq and OH,
thereby reducing the decomposition rate of H2 by (R19) and the
decomposition rate of H2O2 by (R6) and (R18). This can be clearꢀly
seen in Figure 4, where the predicted concentrations of •OH, •eaq
,
and •O2ꢀ at pH 6.0 and pH 10.6 in solutions with initially dissolved
NO3ꢀ and NO2ꢀ are compared with those in pure water. At a given
pH, the concentrations of all three radicals are substantially lower in
the presence of either nitrate or nitrite than those in pure water. The
lowered radical concentrations lead to reduced decomposition rates
for molecular water radiolysis products and cause the observed
increase in the net radiolytic production of H2 and H2O2 in the
presence of nitrogen species.
•
•
ꢀ
important removal path for eaqꢀ. The product of (R5), O2
,
however, reacts with •OH (reaction R20) to re-form O2. Once this
pseudocatalytic cycle is established, the secondary radiolysis product
O2 can build up to a significant level, while it can continuously react
with the radicals. This results in the reduction of the decomposition
rates of molecular products, thus, increasing their concentrations (E1
and E2). The significant reduction in the radical concentrations and
increase in the molecular concentrations, in pure water at pH 10.6
compared to pH 6.0, are further supported by the computer
simulation predictions shown in Figure 4. While a similar cycle can
The increases in [H2] and [H2O2] due to the addition of nitrate
or nitrite, compared to pure water cases, are larger at pH 6.0 than at
pH 10.6. At pH values >9.7, the pKa of •H (•H a •eaqꢀ þ Hþ), the
reaction of •eaqꢀ with Hþ (R40) is significantly restricted and no
longer a main loss path for •eaqꢀ. In pure water, this leads to the
establishment of the pseudocatalytic cycle of (R5) and (R20) as
described earlier in section 5.1. This substantially increases not only
•
•
be established between H and O2 (leading to HO2 which can
dissociate to Hþ and •O2ꢀ), the H atom cycling will be less effective
4275
dx.doi.org/10.1021/jp200262c |J. Phys. Chem. A 2011, 115, 4270–4278

The Journal of Physical Chemistry A
ARTICLE
Figure 4. The model simulation results for the time-dependent concentrations of •OH, •eaqꢀ, and •O2ꢀ achieved in the radiolysis of deaerated solutions
of pure water (blue solid line), [NO3ꢀ]0 = 1 ꢁ 10ꢀ3 M (red dash line), and [NO2ꢀ]0 = 1 ꢁ 10ꢀ3 M (black dotted line) at (a) pH 6.0 and (b) pH 10.6.
The total absorbed dose (irradiation time ꢁ DR) for the 5 h irradiation period was 37.8 kGy.
[O2] but also [H2] and [H2O2] at longer times while decreasing the
Similarly, nitrite solutions would be expected to yield a higher [H2]
concentrations of •OH and •eaqꢀ. Nitrate and nitrite will compeꢀte
than nitrate solutions, since NO2 is a good scavenger for OH,
whereas no reaction of •OH with NO3ꢀ has been observed.8 The
computer model indeed predicts initially 2 orders of magnitude
higher [•OH] and a factor of 2 lower [•eaqꢀ] in nitrate solution than
those in nitrite solution at both pH values, Figure 4.
ꢀ
•
•
for the radicals, mainly with Hþ at pH less than 9.7 (for eaq
,
reaction R40), and with the molecular radiolysis products O2 as well
as H2 and H2O2 at pH greater than 9.7 (for •eaqꢀ and •OH, (R5),
(R6), and (R18)). Due to significant net radiolysis production of
the molecular species at the higher pH, the additional nitrate or
nitrite has less influence on the radical concentrations. In pure water,
the loss paths for the radicals change as pH changes around the pKa
of •H. However, in the presence of 10ꢀ3 M nitrate or nitrite, their
main loss paths are the reactions with nitrate and nitrite, and the net
water radiolysis production does not vary significantly with pH in
the presence of nitrogen-containing species.
The effect of the initial nitrogen species on the net radiolytic
production of H2 and H2O2, however, diminishes with irradiation
time, and this can be explained by the radiolytically induced
interconversion between nitrate and nitrite. Nitrate and nitrite ions
can both participate in a series of redox reactions with reducing
(e.g., •eaqꢀ, •O2ꢀ) and oxidizing (e.g., •OH, HO2 ) species, (R41)
•
to (R46). More importantly, one of their inꢀtermediate products,
•
ꢀ
At a given pH, the addition of nitrate ions increased H2O2
production more than the addition of nitrite ions, while the addition
of nitrite increased the H2 production more than the addition of
nitrate. These differences can be partially attributed to the different
rate constants of the reactions of NO3ꢀ and NO2ꢀ with the radical
water radiolysis products, (R41) to (R45). However, the dif-
ferences in [H2] and [H2O2] are smaller than those expected
based on the elementary reaction rate constants. The reaction of
NO2 , disproportionates to NO3 and NO2 at a fast rate. Thꢀis
series of reactions establishes a pseudocatalytic cycle between NO3
and NO2ꢀ. Under continuous irradiation, this will result in a slow
net conversion between nitrate and nitrite, while significantly
affecting the water radiolysis product concentrations. This expꢀlains
that even though there is very little net conversion of NO3 to
ꢀ
NO2 the net radiolytic productions of H2, H2O2, and O2 are
significant, Figures 1a and 2a.
The competitiveness of nitrate and nitrite for the radicals under a
ꢀ
•
ꢀ
NO3ꢀ with eaq has a larger rate constant than the reaction of
NO2 with eaqꢀ, and a larger increase in [H2O2] would be
given set of conditions determines the interconversion rate toward a
•
•
•
ꢀ
expected for radiolysis of nitrate solutions than nitrite solutions.
steady state. The concentrations of OH and O2 increase with
4276
dx.doi.org/10.1021/jp200262c |J. Phys. Chem. A 2011, 115, 4270–4278

The Journal of Physical Chemistry A
ARTICLE
time in the nitrite solution while the concentration of •OH decreases
and the concentration of •O2ꢀ is nearly constant with time in the
nitrate solution. This results in nearly the same steady-state
concentrations of the radicals and, hence, the molecular products
independent of the initial nitrogen speciation at a given pH. The
observed nitrogen speciation at this steady state, where the relative
changes in concentrations are small, is thus also nearly inꢀdependent
of whether the initial nitrogen species is NO3ꢀ or NO2
.
5.2.2. Nitrogen Speciation, [NO3ꢀ] and [NO2ꢀ]. The pseu-
docatalytic reactions of the nitrogen compounds with the radicals
and their competition with the reactions between radical and
molecular water radiolysis products are responsible for the net
production of H2 and H2O2 observed under different experi-
mental conditions. Although the nitrogen speciation reached at
longer times is nearly independent of the initially dissolved
nitrogen species, it is strongly dependent on pH.
ꢀ
Figure 5. The speciation of NO3ꢀ (red solid triangle down) and NO2
Nitrate will initially react with reducing radical species (•eaq
ꢀ
(black solid triangle up), in the end shield cooling system (pH 10) at
Point Lepreau Generating Station.
and •O2ꢀ) to form NO2ꢀ, but NO2ꢀ will quickly react with the
oxidizing radical species (•OH) to re-form NO3ꢀ resulting in a
slow evolution toward steady-state speciation. Approximate
analytical solutions for the concentrations using the steady-state
approximation have been developed and compared with the
computer predictions. These kinetic analyses have prꢀovided an
approximate solution for the ratio of NO2ꢀ and NO3
agreement with the results of our work which predict the same
trend in speciation, as shown in Figures 1 and 2. The radiolytic
production of deuterium (it is a heavy water system), with air
ingress, is also elevated in this system, in further agreement with
the predictions of our models. We expect that further improve-
ments in our knowledge of the radiation chemistry of nitrogen
compounds will support analysis of real reactor chemistry
conditions and understanding of the consequences and mitiga-
tion of abnormal reactor chemical conditions in the future.
kR41½•eaqꢀꢂ þ kR46½•O2ꢀꢂt
½NO2ꢀꢂt
t
ꢃ
ðE3Þ
½NO3ꢀꢂt
kR45½•OHꢂ
t
The ratio of NO2ꢀ to NO3ꢀ is a function of the concentrations of
•
ꢀ
,
•O2ꢀ, and •OH, and the pH
three radical species, eaq
dependence of the ratio arises from the dependence of the
radical concentrations on pH. From Figure 4 ꢀit is seen that the
6. CONCLUSIONS
The kinetics of nitrate/nitrite solution radiolysis was followed by
monitoring the concentrations of molecular water radiolysis pro-
ducts, H2 and H2O2, as well as NO2ꢀ and NO3ꢀ, as a function of
irradiation time. The evolution of the concentrations as a function of
irradiation time is described well by reactions of nitrogen species
with the radicals generated by the ionizing radiation. The complete
reaction kinetics model for radiolysis of nitrate/nitrite solutions was
•
ꢀ
•
steady-state concentrations of eaq and O2 do not change
when the pH is changed from 6.0 to 10.6. However, the
concentration of •OH is about 10 times lower at pH 10.6 thaꢀn
at pH 6.0, resulting in a change in the ratio of NO2ꢀ to NO3
observed as a function of pH.
The combination of (E1), (E2), and (E3) suggests that the
approximate concentrations of the radical species in solution can
be determined indirectly from the measurements of H2, H2O2,
and the steady-state ratio of [NO2ꢀ] and [NO3ꢀ], providing a
complete picture of water radiolysis product distribution. This is
a valuable tool for monitoring water chemistry since a combina-
tion of the concentrations of the redox active radical species and
molecular species determines the aqueous redox potential that
controls corrosion behavior.
5.2.3. Assessment of Nuclear Reactor Data. The ultimate
objective of studying the chemistry of water systems in a
radiation field is to be able to predict the chemical behavior
and develop effective means to mitigate the damaging conse-
quences of corrosion. We have access to some actual data from
operating nuclear power reactors to test our understanding of
water radiolysis chemistry. Nitrogen compounds are present in
nuclear reactor water systems as a result of either air ingress or
deliberate additions of nitrogen-containing compounds. Some
CANDU plants experience periodic, inadvertent air ingress into
the end shield cooling system. The water in this system (which is
maintained at pH 10) is exposed to high fluxes of γ-radiation
during reactor operation. Chemistry data for this system at the
Point Lepreau Generating Station are shown in Figure 5. The
data show that the speciation of nitrogen compounds in the
system has more nitrite than nitrate present. This finding is in
ꢀ
reduced to include only those reactions with highly reducing (•eaq
•
and O2ꢀ) and highly oxidizing (•OH) radicals, and the dispro-
portionation reaction of •NO2. Analysis using this reduced reaction
set demonstrated that it was sufficient to match the observed system
chemistry well. Using this reaction set, the ratio of steady-state
concentrations of NO2ꢀ and NO3ꢀ can be expressed as a simple
function of the concentrations of the key water radicals.
This study, which combines experiments and kinetic model
analysis, has established a reaction mechanism for the steady-state
radiolysis of nitrate and nitrite solutions that can satisfactorily explain
data from a nuclear power reactor station as well as laboratory scale
experiments. For water systems containing nitrate or nitrite, the
concentrations of the redox radical species arising during water
radiolysis can be determined fꢀrom measurements of H2, H2O2,ꢀand
ꢀ
the ratio of NO2 and NO3 in solution. The ratio of NO2 to
ꢀ
NO3 at steady state is shown to be independent of the initial
nitrogen species, but a strong function of pH; at pH values <9.7, the
pKa of •H, nitrate is the dominant species at steady state whereas
nitrite is dominant at higher pH values.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: jcwren@uwo.ca.
4277
dx.doi.org/10.1021/jp200262c |J. Phys. Chem. A 2011, 115, 4270–4278

The Journal of Physical Chemistry A
ARTICLE
’ ACKNOWLEDGMENT
(27) Logager, T.; Sehested, K. J. Phys. Chem. 1993, 97, 6664–6669.
(28) Lymar, S. V.; Schwarz, H. A.; Czapski, G. J. Phys. Chem. A 2002,
106, 7245–7250.
(29) Goldstein, S.; Czapski, G. Inorg. Chem. 1996, 35, 5935–5940.
(30) Seddon, W. A.; Young, M. J. Can. J. Chem. 1970, 48, 393–394.
(31) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B.
J. Phys. Chem. Ref. Data 1988, 17, 513–886.
(32) Tretyakov, V. P.; Minko, L. A.; Rudakov, E. S.; Alekhina, T. A.
Theor. Exp. Chem. 2003, 39, 36–40.
(33) Goldstein, S.; Czapski, G. Free Radical Biol. Med. 1995,
19, 505–510.
The work described herein was funded under the NSERC/
AECL (Natural Science and Engineering Research Council of
Canada/Atomic Energy of Canada Limited) Industrial Research
Chair in Radiation Induced Processes. Support from the Canada
Foundation for Innovation New Opportunity grant is greatly
acknowledged for the purchase of the Gamma-cell, GC-MSD/
TCD, and UVꢀvis absorption spectrophotometer. The authors
thank the NB Power (New Brunswick Power) staff for providing
reactor data for the end shield cooling system.
(34) Bielski, B. H.; Cabelli, D. E.; Arudi, R. L. J. Phys. Chem. Ref. Data
1985, 14, 1041–1100.
(35) Allen, A. O.; Hochanadel, C. J.; Ghormley, J. A.; Davis, T. W.
J. Phys. Chem. 1952, 56, 575–586.
’ REFERENCES
(1) Stuart, C. R.; Reid, V.; Turner, C. W. Optimization of Chemistry
Control in the Primary Heat Transport System of CANDU Reactors.
Proceedings of the International Conference on Water Chemistry of Nuclear
Reactor Systems, Jeju Island, Korea, Oct 23ꢀ26, 2006.
(2) Yakabuskie, P. A.; Joseph, J. M.; Keech, P; Botton, G. A.;
Guzonas, D; Wren, J. C. Phys. Chem. Chem. Phys. 2011, 13, 7198ꢀ7206.
(3) Fu, D; Keech, P. G.; Sun, X; Wren, J. C. Phys. Chem. Chem. Phys.
2010, Submitted.
(4) Spinks, J. W. T., Woods, R. J. An Introduction to Radiation
Chemistry; Wiley: New York, 1990.
(5) Wren, J. C. Steady-state radiolysis: effects of dissolved additives.
In Nuclear Energy and the Environment; Chien, W., Ed.; ACS Symposium
Series, Vol. 1046, American Chemical Society: Washington, DC, 2010;
Chapter 22.
(6) Daub, K.; Zhang, X.; No€el, J. J.; Wren, J. C. Corros. Sci. 2011,
53, 11–16.
(7) Daub, K.; Zhang, X.; No€el, J. J.; Wren, J. C. Electrochim. Acta
2010, 55, 2767–2776.
(8) Mezyk, S. P.; Bartels, D. M. J. Phys. Chem. A 1997,
101, 6233–6237.
(9) Buxton, G. V.; Stuart, C. R. Chem. J. Soc., Faraday Trans. 1996,
92, 1519–1525.
(10) Buxton, G. V.; Stuart, C. R. Chem. J. Soc. Faraday Trans. 1997,
93, 1535–1538.
(11) Buxton, G. V.; Lynch, D. A.; Stuart, C. R. In Water Chemistry of
Nuclear Reactor Systems 7. Proceedings of the British Nuclear Energy
Society Conference; Bournemouth, United Kingdom, Oct 13ꢀ17, 1996;
Thomas Telford Publishing: London, 1996; pp 150ꢀ152.
(12) Buxton, G. V.; Lynch, D. A. J. Chem. Soc., Faraday Trans. 1998,
94, 3271–3274.
(13) Mezyk, S. P.; Bartels, D. M. J. Phys. Chem. A 1997,
101, 6233–6237.
(14) Upadhyay, S. N.; Ray, N. K.; Goel, H. C. Indian J. Nucl. Med.
2002, 17, 35–41.
(15) Joseph, J. M.; Choi, B.-S.; Yakabusie, P.; Wren, J. C. Radiat.
Phys. Chem. 2008, 77 (9), 1009–1020.
(16) Hochanadel, C. J. J. Phys. Chem. 1952, 56, 587–594.
(17) Stefanic, I.; LaVerne, J. A. J. Phys. Chem. A. 2002, 106 (2),
447–452.
(18) Schwarz, H. A.; Salzman, A. J. Radiat. Res. 1958, 9, 502–508.
(19) Slack, P. T. Analytical Methods Manual; Leatherhead Food R.A.:
London, 1987.
(20) Sun, J.; Zhang, X.; Broderick, M.; Fein, H. Sensors 2003,
3, 276–284.
(21) Cortas, N. K.; Wakid, N. W. Clin. Chem. 1990, 36, 1440–1443.
(22) Yakabuskie, P. A.; Joseph, J. M.; Wren, J. C. Radiat. Phys. Chem.
2010, 79, 777–785.
(23) Wren, J. C.; Ball, J. M. Radiat. Phys. Chem. 2001, 60, 577–596.
(24) Wren, J. C.; Glowa, G. A. Radiat. Phys. Chem. 2000,
58, 341–356.
(25) Goldstein, S.; Lind, J.; Merꢀenyi, G. Chem. Rev. 2005,
105, 2457–2470.
(26) Cook, A. R.; Dimitrijevic, N.; Benjamin, W. J. Phys. Chem. A
2001, 105, 3658–3666.
4278
dx.doi.org/10.1021/jp200262c |J. Phys. Chem. A 2011, 115, 4270–4278