Hatano Y., Katsumura Y., Mozumder A. (Eds.) Charged Particle and Photon Interactions with Matter - Recent Advances, Applications, and Interfaces

Подождите немного. Документ загружается.

1000 Charged Particle and Photon Interactions with Matter

alkanes such as solvent extraction diluents typically react at rates of ∼10

−1

. For these reasons,

radical redox reactions are probably less important than the nitrous acid concentration in establishing

actinide oxidation states in irradiated aqueous nitric acid.

It can be seen from the foregoing discussions that a multitude of radiolytically induced reactions

compete

to set the neptunium oxidation state in irradiated acidic solution. Vladimirova (1995) pro-

vided

the most comprehensive attempt to model the redox chemistry of neptunium in both alpha-

and gamma-irradiated acidic solution. Taking into account reactions with nitrous acid and the major

produced radicals, he calculated equilibrium concentrations of Np(IV), Np(V), and Np(VI) under

various conditions of dose rate, absorbed dose, and nitric acid concentration. High dose rates favored

the production of Np(IV), while high acidity resulted in rapid declines in Np(V) with increases in

Np(IV) and Np(VI). The reduction of neptunium to the tetravalent oxidation state was attributed

to disproportionation, and reaction with the

•

H atom. The rate constant of the reaction shown in

Equation 34.83 was calculated to increase by a factor of 100 from <5 × 10

−1

to 2 × 10

−1

with

an increase in nitric acid concentration from 1 to 6

NpO H H Np H O

+ + +

+ + → +

•

2 2

(34.83)

At a dose rate of 36kGy h

−1

and a nitric acid concentration of 3M, the equilibrium percentages of

Np(IV), Np(V), and Np(VI) were 8, 72, and 20, respectively. The equilibrium yields of the three

oxidation states were the same for both types of radiation. In additional work, this same author

included renements to account for the presence of organophosphorous extractants including TBP

(Vladimirova etal., 1996). It was concluded that the presence of the organic compounds favored

higher equilibrium concentrations of Np(IV) and Np(VI) at the expense of Np(V). Sensitivity analy-

sis using a mathematical model suggested that the rate constants for the reduction of Np(VI) by

nitrous acid, and the reaction of Np(IV) with Np(VI) to produce Np(V) were decreased, possibly

due

to complexation of Np(IV) and Np(VI) by the organic complexing agents.

34.4 geologiCal disposal oF high-level radioaCtive waste

34.4.1 introduction

Management and geological disposal of high-level radioactive waste are among the most critical

issues related to nuclear technology in recent years. High-level radioactive waste, which includes

spent nuclear fuel discharged from commercial power reactors or vitried waste arising from

table 34.8

rate

Constants for r

eactions

of a

ctinide

ons

reaction rate Constant k (m

−1

) references

•

OH → Np(IV)

4.00 × 10

Gogolev etal. (1988)

Np(IV)

•

OH → NpO

3.20 × 10

(pH 0) Shilov etal. (1982)

Np(V)

•

OH → ?

4.60 × 10

Lierse etal. (1988)

NpO

•

OH → NpO

4.30 × 10

(pH 0) Shilov etal. (1982)

•

H → Np

6.00 × 10

(pH 0) Gogolev etal. (1991)

•

H → ?

<1 × 10

(pH 0) Gogolev etal. (1991)

NpO

•

H → ?

<5 × 10

(pH 0) Shilov and Pikaev (1982)

NpO

•

H → NpO

<1 × 10

(pH 1–3) Schmidt etal. (1983)

+ e

−

→ Np(III)

6.00 × 10

(pH 0) Gogolev etal. (1991)

NpO

+ e

−

→ NpO2

2.40 × 10

(pH 3) Schmidt etal. (1980)

NpO

+ e

−

→ NpO

1.00 × 10

(pH 2.5) Schmidt etal. (1983)

NpO

•

→ NpO

8.10 × 10

(pH < 0) Gogolev etal. (1986)

Radiation Chemistry inNuclear Engineering 1001

reprocessing of power reactor fuel, will be disposed several hundred meters (e.g., more than 300 m

in Japan’s case) below sea level in isolation from humans and their environment. This is referred to

hereafter as geological disposal. The concept of geological disposal is being considered in many

countries and is based on a system of multiple passive barriers, consisting of engineered barri-

ers and the geological environment (natural barrier), as illustrated in Figure 34.26 (JNC, 2005;

SKB, 2006a). In general, the engineered barriers consist of three different types of barriers, as

described below.

The rst barrier is the waste form itself since radioactive nuclides cannot be released until after

being dissolved. The nature of the actual waste form will be dependent on whether the spent nuclear

fuel is directly disposed (once-through) or reprocessed to recover ssile and fertile materials in

order to provide fresh fuel for existing and future nuclear power plants (nuclear fuel cycle). In

the former case, uranium dioxide pellets are encapsulated in a cladding material such as zircaloy.

Cladding tube

Fuel pellet of

uranium dioxide(a)

Copper canister

with cast iron insert

Crystalline

bedrock

Underground portion of

deep repository

Spent nuclear fuel Bentonite clay Surface portion of deep repository

500 m

Buffer

Host rock

(b)

Geological environment

HLW borosilicate glass

Engineered barrier system

Stainless steel canister

Steel overpack

Bentonite/sand buffer

Figure 34.26 Conceptual design of geological disposal systems. (a) KBS-3 type spent nuclear fuel

disposal. (Adapted from SKB, Long-term safety for KBS-3 reositories at Forsmark and Laxemar—a rst

evaluation, Main report of the SR-Can project, SKB Technical Report TR-06-09, Swedish Nuclear Fuel

and Waste Management Co., Stockholm, Sweden, 2006a.) (b) HLW disposal in Japan. (Adapted from JNC,

H17: Development and management of the technical knowledge base for the geological disposal of HLW,

Knowledge Management Report, JNC-TN1400-2005-022, Japan Nuclear Cycle Development Institute,

Tokai-mura, Japan, 2005.)

1002 Charged Particle and Photon Interactions with Matter

In the latter case, radioactive elements in the vitried high-level radioactive waste are immobilized

by matrices such as borosilicate glass, which provides a low solubility matrix in which the radionu-

clides

are homogeneously distributed.

The

second barrier is a waste package, which encapsulates the waste form to ensure its con-

nement for a certain period of time, especially so as to prevent groundwater from contact

with the waste form. In some countries which are planning to implement direct disposal, spent

nuclear fuel is packed in a copper canister with a carbon steel insert, while vitried waste in a

stainless steel canister is encapsulated in a carbon steel package. Although any metal inevitably

corrodes even in the anoxic underground environment, a metallic waste package is expected to

provide a period of complete containment during the radiogenic thermal transient, when condi-

tions are rather dynamic and may be strongly coupled (e.g., thermal and water saturation proles

in the buffer).

The third barrier is a buffer material which surrounds the waste package. The buffer material is

required to have favorable physical and chemical characteristics, such as self-sealing and mechani-

cal buffering, good thermal conductivity, low hydraulic conductivity, colloid ltration function,

and chemical buffering capacity to retard radionuclide migration. Materials like clay-based ben-

tonites are typically used for the buffer due to their low hydraulic conductivity and large sorption

capacity.

At present, bentonite is the most commonly available material for the buffer, whose mineral

composition is shown in Table 34.9. Montmorillonite is the main component of bentonite and its

properties are essential for the application of bentonite as the buffer material, including low perme-

abilities and high sorption capacities, which strongly retard migration of any radionuclides released

from

the waste package.

Figure

34.27a shows the structural arrangement of montmorillonite (Bradbury and Baeyens,

2002). The unit crystal lattice of montmorillonite is built with two layers of tetrahedral silicate

units between which one layer of octahedral aluminate is sandwiched. A small part of Al

in the

octahedral layer is substituted with Mg

and Fe

or Fe

, which results in negative charges

in the layers. This charge deciency is compensated by interlayer cations. As a result, the actual

structure formula of montmorillonites from Wyoming, United States, and Tsukinuno, Japan, has

been

determined as follows:

table 34.9

mineralogical

Composition of b

entonites

kunigel v1

(

tsukinuno,

Japan)

mx-80

(wyoming,

the

united states)

Montmorillonite 46–49 75

Kaolinite <1

Mica <1

Quartz/Calcedony 29–38 15.2

Feldspar 2.7–5.5 5–8

Calcite 2.1–2.6 0.7

Dolomite 2.0–2.8

Analcime 3.0–3.5

Siderite 0.7

Pyrite 0.5–0.7 0.3

Organic

carbon 0.31–0.34 0.4

Ito etal. (1993).

Bradbury and Baeyens (2002).

Radiation Chemistry inNuclear Engineering 1003

Wyoming: Na

0.30

(Al

1.55

0.20

0.01

0.24

) (Si

3.96

0.04

(OH)

(Müller-Vonmoos and Kahr, 1983)

Tsukinuno:

(Ca

0.05

0.38

0.01

) (Al

1.55

0.09

0.01

0.34

0.01

) (Si

3.90

0.10

(OH)

(Suzuki etal., 1992)

Figure 34.27b shows a schematic illustration of compacted bentonite after water saturation

(Bradbury and Baeyens, 2002). The typical size of a sheet of montmorillonite is about several

hundred nanometers in diameter and 1 nm in thickness. Montmorillonite sheets form stacks and

such secondary particles are randomly oriented in the compacted state together with mineral

particles. Exchangeable cations in the stack are hydrated to some extent at ambient condition

and their further hydration induces very strong volume swelling. As can be seen from the

gure, any migration of ions and particles in this compacted bentonite would experience very

tortuous pathways. Chemical reactions such as ion exchange with interlayer cations and sur-

face complexation on edge sites of montmorillonite sheets are also important for retardation of

radionuclide migration.

After the repository closure,thebuffer materialis graduallysaturated with groundwater andchem-

ical reactions in the buffer material and groundwater would occur. It is very difcult to experimen-

tally determine chemical composition of porewater in bentonite, and therefore it has been estimated

from geochemical modeling calculations. Table 34.10 shows the highest estimated concentration of

the main chemical species in bentonite porewater (JNC, 2000). These results are summarized from

various conditions of groundwater (e.g., freshwater/seawater origin, oxygen concentration, pH) and

bentonite composition and reactions (Oda etal., 1999). Carbonates are commonly found in ground-

water and the chloride concentration is high particularly in the case of groundwater with seawater

origins.

High concentrations of S(-II) are due to the dissolution of pyrite (FeS

The bedrock surrounding the engineered barriers is expected to function as a natural barrier

system since it further retards migration of radionuclides, which may be released through the buffer

material.

The long-term performance of the multiple barrier system has been assessed in several coun-

tries in an attempt to demonstrate the technical feasibility of their disposal concepts. A number of

Exchangable cations

+ Water

Interlayer water and

exchangeable cations

Calcite/Quartz/Feldspar

Double layer water

[(Si,Al)]

OH/O

Al, Mg, Fe

Free water

(Bentonite porewater)

(a) (b)

Figure 34.27 Structure of (a) montmorillonite and (b) compacted bentonite. (Adapted from Bradbury, M.H.

and Baeyens, B., Pore water chemistry in compacted re-saturated MX-80 bentonite: Physicochemical charac-

terization and geochemical modelling, PSI Bericht 02–10, Villigen PSI and NTB 01– 08, Nagra, Wettingen,

Switzerland, 2002.)

1004 Charged Particle and Photon Interactions with Matter

scenarios with specic features, events and coupled thermal, hydraulic, mechanical, and chemical

processes have been considered. The source model includes the radioactivity of the waste form as

it causes heat generation and ionizing radiation, which lead to material alteration and radiolysis

of groundwater. Figure 34.28 shows changes in radioactivity with time for a typical spent nuclear

fuel from an LWR (Poinssot et al., 2005). In the early post-closure stage, radioactivity of ssion

products is dominant. Since half-lives of most of ssion products are shorter than those of actinides,

radioactivity of actinides becomes dominant around several hundred years after repository closure.

From the viewpoint of system evolution, the impact of ionizing radiation from the waste form

on the engineered barrier system can be categorized into the following two phases, as illustrated in

Figure 34.29: In the earlier post-closure stage, the radioactivity of the waste is extremely high and

the waste form itself is subjected to radiation damage including that of the damage by alpha recoil.

The buffer material would be subjected to gamma radiation, which penetrates through the waste

package. Pore uids in the buffer material would be subject to gamma radiolysis and radiolytic

products might result in oxidizing conditions in the porewater. This could lead to signicant impacts

table 34.10

range

of Chemical s

pecies

and p

in

orewater

of b

entonite

Concentration (mol dm

−3

)

HCO

−

/CO

2−

<7.3 × 10

−2

2−

<6.1 × 10

−2

−

S <9.2 × 10

−2

−

<5.9 × 10

−1

P (Total) <2.9 × 10

−6

−

0.0

<1.6 × 10

−4

<5.1 × 10

−3

B (Total) <1.7 × 10

−3

pH 5.9–8.4

1E+17

1E+16

1E+15

1E+14

1E+13

1E+12

1E+11

1 10 100 1000

Time (years)

Residual activity (Bq/tHM)

10,000 100,000 1,000,000

Beta

Alpha

Figure 34.28 Evolution of α, β and γ radiations as a function of time for a 55 GWd t

–1

UOX fuel.

(Reprinted from Poinssot, C., et al., J. Nucl. Mater, 346, 66, 2005. With permission.)

Radiation Chemistry inNuclear Engineering 1005

on the engineered barriers’ function. For example, the corrosion of the waste package would be

accelerated under oxidizing conditions, and if the extent is signicant the waste package could be

breached earlier than expected. The reactions of radiolytic products with the buffer materials might

deteriorate

their function of retarding radionuclide migration.

is assumed in the scenario that the waste package would eventually be breached even under

anoxic conditions due to metal corrosion and thus would fail more than several hundred years after

repository closure. Groundwater would eventually contact the surface of the waste form and radionu-

clides would be released by dissolution of the waste form. Around this period, most of beta/gamma

radionuclides would have decayed to a level several orders of magnitude lower than the initial level,

and alpha radiation from the surface of the waste would induce radiolysis of water in a very limited

region about 30μm from that surface. In the case of spent nuclear fuel pellets, it is anticipated that

the oxidative products of alpha-radiolysis of water would react with the uranium oxide matrix and

thus increase its dissolution rate signicantly. The radiolytic products might also react with the cor-

rosion products of the waste package. Those having escaped from the reactions in the waste package

might diffuse into the buffer material and also react with reductants such as Fe(II).

Those potential processes induced by the ionizing radiation in different phases of the repository

system

life are discussed in the following sections.

34.4.2 effect of gaMMa radiation on engineered barrier SySteMS

In the near term after repository closure, gamma radiation is the dominant source of radiation impact

on high-level waste forms and their immediate environment because alpha and beta radiation can be

shielded completely by the metal waste package. The major source of gamma radiation is

137

Cs with a

half-life is about 30 years. The impact of gamma radiation on the engineered barriers can be controlled

to some extent by the engineering design. For this purpose, it has been proposed to design the waste

package so as to reduce the dose rate at its outer surface to less than 1Gy h

−1

(Werme, 1998).

34.4.2.1 effect

on Clay m

inerals

As noted above, the engineered barrier system would receive gamma radiation from the waste forms

in the early phase after repository closure. Most of iron atoms exist as Fe(III) and they are reduced

after gamma irradiation, as probed by Mössbauer and ESR spectra (Gournis et al., 2000; Plotze et

al., 2003). Observed slight decreases in cation exchange capacities after long periods (22 months)

Overpack Overpack

Canister Canister

Anoxic

corrosion

HLW waste HLW waste

(a) (b)

Accelerated

corrosion?

Redox-front

migration?

Alpha-radiolysis on

HLW surface

Corrosion

products

(Fe

)

Gamma-radiolysis of

porewater

FeS

Bentonite Bentonite

Corrosion

products

(Fe

)

Figure 34.29 Effects of ionizing radiation on a repository. (a) Before the waste package failure and (b)After

the waste package failure. (Adapted from IRI, Radiation effects on barrier systems in high level radioactive waste

disposal, Research Project Annual Report, Institute of Research and Innovation, Tokyo, Japan, 2005. (in Japanese).)

1006 Charged Particle and Photon Interactions with Matter

may be attributed to dissipation of protons, lowering the layer charge (Plötze etal., 2003). No struc-

tural

change was observed after gamma-irradiation up to 2

MGy

(Negron etal., 2002).

Radiation-induced

defects formed in the framework of smectite clays were assigned as trapped

holes on oxygen atoms (Clozel etal., 1994, Allard and Calas, 2009). The relative amounts of the

defects are dependent on the atomic composition and arrangements and they resulted in isotope

exchange of hydroxyl ion and changes in specic surface area. However, those changes are small in

the

case of montmorillonite after irradiation up to 30

MGy

(Pushkareva etal., 2002).

Most

of the experiments summarized above were performed with powdered or dispersed smec-

tites. Pusch etal. performed experiments to simulate the actual condition of compacted benton-

ite contacting steel under high-temperature and gamma-radiation conditions (Pusch etal., 1992).

Commercial MX-80 was compacted to a dry density of 1.65g cm

−3

and saturated with simulated

groundwater. Irradiation was continued for 1 year and the total dose was 33MGy. Little change in

the framework structure and physicochemical properties were observed, although Fe(III) ions may

have

slightly been reduced.

These

results indicate physicochemical properties of montmorillonite as the buffer materials are

little affected by gamma-ray exposure. However, signicant damage leading to amorphization has

been observed under electron-beam irradiation in electron micrographs when the accumulated dose

is increased to the order of GGy (Gu etal., 2001, Sorieul etal., 2008). Threshold values for radia-

tion-induced amorphization were strongly dependent on the temperature and structure of smectites:

Threshold values are the largest at 300°C − 450°C, which coincides with the onset of dehydroxyl-

ation. At room temperature, thresholds for amorphization were over 10 GGy, which is comparable

to cumulative doses estimated in HLW repositories after 10

years for radionuclides adsorbed to the

bentonite

buffer.

Several

countries are planning to build repositories in clay formations and the effect of gamma

radiation on natural clays has been studied. In particular, various tests have been performed in the

underground research facility in the Boom clay in Belgium. Mössbauer spectra of the Boom clay

indicated that while pyrite was completely oxidized after gamma-irradiation up to 30MGy, Fe(II)

remained, as in the case of the compacted bentonite (Ladriere etal., 2009). An in situ experiment in

the underground research laboratory in the Boom clay formation was performed to develop a model

to predict transient behavior of the disposal system. In the CERBERUS experiment, a

Co gamma-

ray source and heaters were placed in the clay formation to simulate waste forms, and changes in E

and

pH were monitored (Zhang etal., 2008).

34.4.2.2

effect

on Corrosion of w

aste

ackage

In the initial phase after the repository closure, the buffer material is unsaturated and therefore the

gamma-radiolysis of air in pores of the buffers would generate nitrogen oxides, which eventually

form nitric acid which might affect the corrosion of the canister. The amount of nitrate in the vessel

is proportional to the gas volume against the liquid volume (Linacre and Marsh, 1981). However the

estimated

amount of generated nitric acid was negligible (King etal., 2001).

the buffer materials are gradually saturated with groundwater, gamma-radiolysis products

like H

or O

are generated. These radiolytic products would induce oxidative corrosion of metal

packages. Increased corrosion rates of carbon steel in saline water under gamma irradiation were

observed in the 1980s in the United Kingdom. However, the effect was negligible at dose rate of

about 3Gy h

−1

(Marsh etal., 1989). The effect was also observed in compacted bentonite saturated

with the synthetic groundwater. They have performed 1D diffusion reaction model calculation to

evaluate the effect of radiation by assuming cathodic reaction of radiolytically generated oxygen.

Burns etal. have incorporated surface reactions of short-lived radiolytic products in models to

calculate gas generation in gamma-irradiated steel vessels (Burns etal., 1983). A more elaborate

mixed potential model was applied for the corrosion of HLW canisters in which homogenous and

surface reactions of radiolytic products were fully taken into account (Macdonald and Urquidi-

Macdonald, 1990). Recently, the corrosion of the carbon steel inserts of spent nuclear fuel canisters

Radiation Chemistry inNuclear Engineering 1007

under gamma radiation was studied assuming cannsiter failure (Smart etal., 2008). The effect of

gamma radiation was observed at the dose rate of 300Gy h

−1

. The predominant corrosion products

were magnetite and oxyhydroxides (or maghemite), which are thought to be formed by anoxic cor-

rosion

of iron and subsequent dissolution of Fe

at pH < 7 (White etal., 1994).

3 4 4

2 4 2

Fe H O Fe O H

+ → + (34.84)

Fe O H -Fe O Fe H O

3 24 3

2+ → + +

+ +

(34.85)

Copper is employed for the material for the canister in several countries and its corrosion under

gamma radiation has also been studied (King etal., 2001). In the early stages of disposal, the oxygen

concentration

is still high and the corrosion in saline groundwater would proceed as follows:

Anodic Cu Cl CuCl e: + → +

− − −

(34.86)

Cathodic O e H O OH

: + + →

− −

4 2 4

(34.87)

The effect of gamma radiation on the corrosion of copper embedded in a montmorillonite–sand

mixture was studied (King etal., 1992; Shoesmith and King, 1999). The copper concentration in the

mixture signicantly decreased upon irradiation, presumably due to reduction of Cu(II) to Cu(I)

which forms a weakly sorbed anionic complex. It was expected that the oxidizing species produced

by groundwater radiolysis would accelerate the cathodic reaction. Although the observed reduction

of copper by gamma-radiolysis was the opposite of expected, the effect was similar to the case of the

reduction of iron in montmorillonite (Pusch etal., 1993). King etal. conjectured that electrons pro-

duced

from montmorillonite ionization might have reduced cupric ion in the porewater. It has been

proposed that the energy of ionizing radiation deposited in nanoparticles of metal oxides dispersed

in the aqueous phase might be transferred to water to generate the hydrated electron, and eventu-

ally molecular hydrogen (Schatz etal., 1998; LaVerne and Tonnais 2003; LaVerne, 2005). The same

reaction might have occurred in the porewater of montmorillonite under ionizing radiation as the

thickness of a single layer of montmorillonite is about 1 nm and the secondary electrons produced

might have easily escaped.

34.4.3 effect on diSSolution of waSte forMS

After the assumed failure of a waste package, groundwater would penetrate to the surface of the

waste form and eventually would receive alpha-radiolysis. In this section, the effects of water radi-

olysis on dissolution of the waste form are described. The dissolution rate of the waste form is a

key parameter for the performance assessment of the disposal system as it determines the release

rates of radionuclides which eventually diffuse into the engineered and natural barrier systems.

While various factors such as pH, solutes, temperature, etc., affect the dissolution rates, the effect of

ionizing radiation is particularly important in the case of geological disposal of spent nuclear fuel.

Groundwater radiolysis would generate oxidizing species in an otherwise reductive environment

and

may increase the dissolution rate several orders of magnitude.

34.4.3.1

dissolution

of s

pent

uclear

Fuel

the case of direct disposal, most of the radioactive nuclides are retained in a uranium dioxide

matrix and their release rates are determined by the dissolution of that matrix. Uranium dioxide is

stable under the anoxic conditions of the repository and its solubility is extremely low. Figure 34.30

summarizes chemical processes on the surface of uranium dioxide (Poinssot et al., 2005; Martínez

1008 Charged Particle and Photon Interactions with Matter

Esparza etal., 2005). Dissolution rates increase signicantly when UO

is oxidized to UO

2+x

(0 < x < 1) by dissolved oxygen or radiolytically generated oxidizing species. Dissolution of U(VI)

is facilitated by complexation with carbonate ions. Dissolved U(VI) precipitates on the UO

pellet

surface, although some amount would eventually migrate out of the canister to the engineered bar-

rier

systems.

Uranium

dioxide is a p-type semiconducting material and oxidative dissolution of uranium diox-

ide

can be regarded as corrosion with anodic and cathodic half reactions.

Anodic UO UO e

: → +

+ −

(34.88)

Cathodic H O e OH

2 2+ →

− −

(34.89)

O e H O OH

+ + →

− −

4 2 4

(34.87)

Under the condition in which fuel corrosion occurs, the redox potential of the groundwater (E

)

must be more positive than the equilibrium potential for fuel dissolution. When the groundwater

and the spent nuclear fuel are contacted, corrosion potential (E

corr

) is established at which the rates

of the anodic and cathodic reactions will be equal (Shoesmith, 2000, 2007)and is measured to monitor the

degree of oxidation of the sample (Johnson etal., 1983). Oxidation states of uranium on the surface

can

also be probed with x-ray photoelectron spectra (Sunder etal., 1990).

Gamma

radiation has been commonly employed to study the effect of water radiolysis on oxi-

dative dissolution of uranium oxide (Gromov, 1981; Sunder etal., 1992; Christensen and Sunder,

2000). However, alpha radiation is the main source of ionizing radiation after several hundred years

and various methods have been developed to study its effect on radiation-induced oxidative dissolu-

tion

of the fuel matrix.

The

most straightforward method is to study dissolution of spent nuclear fuels, although cur-

rently available spent nuclear fuels still have high beta/gamma radioactivity (Bruno et al., 1999,

2003; Cera et al., 2006). This activity affects water radiolysis with different G-values than alpha-

radiolysis. To eliminate the effect of beta/gamma radiation and complex chemical and morphologi-

cal changes in spent nuclear fuels, uranium dioxide doped with alpha emitters (

238

Pu,

233

U, etc.) has

been utilized to study the effect of alpha radiation (Rondinella etal., 2000; Cobos etal., 2002;

Cachoir and Lemmens, 2004; Jégou etal., 2004; Mennecart etal., 2004b; Muzeau etal., 2009).

Dissolution rates were proportional to the dose rate of the samples, which were controlled by the

1 Generation of oxidants by water radiolysis

Fuel oxidation by radiolytic oxidants

Aqueous ligands (HCO

–

, OH

–

...)

Precipitation of secondary phases

+RN dissolution

, OH , HO

, H

, O

Figure 34.30 Radiation-induced processes on the surface of uranium dioxide. (Reprinted from Poinssot, C.

et al., J. Nucl. Mater., 346, 66, 2005. With permission.)

Radiation Chemistry inNuclear Engineering 1009

addition of different amounts of alpha emitters. Asimulated high-burnup spent nuclear fuel called

SIMFUEL

(Lucuta etal., 1991) was also employed for dissolution studies (Ollila, 1992).

Not

only internal alpha sources but also external sources have been applied. Alpha emitter discs

(

241

Am and

210

Po) were employed in the rst study of alpha-radiolysis affects on UO

dissolution by

placing them close to the surface of UO

pellets (Bailey etal., 1985). In the presence of alpha-radiation,

the corrosion potential of UO

increased to the same level as in the presence of dissolved O

In recent years, helium ion beam irradiation has also been employed to simulate alpha-radiation.

This method is advantageous in that high dose rates can be achieved and ion energy can be con-

trolled to study the LET effect. Two different types of UO

are employed. The rst is colloidal par-

ticles dispersed in aqueous solution (Mennecart etal., 2004a, Suzuki etal., 2006), while the other

is UO

discs contacted with water on one face and being irradiated from the opposite face (Corbel

etal., 2001; Sattonnay etal., 2001). The dissolved U(VI) concentration due to ion beam irradiation

was higher than the value following the addition of the same amount of H

generated by the ion

beam,

indicating the effect of radical species is not negligible (Corbel etal., 2006).

Radiation-induced

oxidative dissolution of UO

was rst modeled by taking into account the

reactions of radical species with UO

(Christensen and Bjergbakke, 1987). Reaction rates at the

surface were estimated from the calculated number density of surface U(IV) atoms and homo-

geneous reaction rates for U(IV) in solution. The model was further rened (Christensen etal.,

1994) and one-dimensional diffusion–reaction calculations of UO

dissolution were performed by

setting different regions of water subjected to radiation (Christensen, 1998). The model was then

extended to the case of UO

dissolution in salt brine (Kelm and Bohnert, 2000a,b). Dissolution rates

of alpha-emitter-doped UO

were recently calculated (Christensen, 2006). The reaction rates of UO

and radical species were estimated to be about two orders of magnitude larger than for molecular

species. Difculties in setting appropriate values of surface-to-volume ratio were pointed out in the

calculation of the dissolution rates of UO

. One-dimensional diffusion–reaction model calculations

of the UO

surface were performed to simulate helium ion beam irradiation experiments assum-

ing alpha-dose rate proles from the surface (Poulesquen and Jégou, 2007). The calculated values

of U(VI) and H

were lower than the measured values in both aerated and deaerated cases.

Measured values remained relatively high even in the presence of hydrogen, in contrast to the case

of dissolution experiments under dissolved hydrogen. The authors attributed this to a high dose rate

used in the calculation, which may have overestimated radical–radical recombination to form H

Another model of radiolytic oxidative dissolution of UO

has been developed in which the reac-

tions of UO

were limited to those with H

and O

(Eriksen, 1996). The reaction of H

is the

most important and consumption rates of oxidants rapidly decreased possibly due to accumulated

, which scavenged OH

•

and suppressed H

formation. They also measured dissolution rates of

PWR spent nuclear fuel pellets (Eriksen etal., 1995). In the early stage of the experiments, mea-

sured concentrations of radiolytic products were lower than estimated by the model, and they were

therefore

assumed to have decomposed on the surface.

The

model has been further developed by Jonsson in recent years (Jonsson etal., 2007). Oxidative

dissolution was facilitated by U(VI) carbonate complex formation, and dissolution rates became

proportional to surface oxidation rates when [HCO

−

] > 10

−3

mol dm

−3

, typical of groundwater. One-

electron oxidation of UO

is considered to be the rate-determining step as the reaction rates of a

series of oxidants with UO

were found to be logarithmically proportional to one-electron reduction

potentials (Ekeroth and Jonsson, 2003). Based on UO

dissolution rates under gamma-irradiation

and reaction model calculations, the relative contributions of various oxidants to UO

oxidation by

water radiolysis were evaluated. It was concluded that contribution of H

was more than 99.9%

and other oxidant species were negligible, which supports the reaction model proposed by Eriksen

(1996)

and (Ekeroth etal., 2006).

One-dimensional

reaction–diffusion model calculations for the surface were also performed.

Dose rate distributions were provided from calculated inventories and geometrical considerations

on the dissipation probability of alpha particles (Nielsen and Jonsson, 2006). Steady-state H