Actinides Q &A

392 Los Alamos ScienceNumber 26 2000

The Chemical Interactions of Actinides in the Environment

F rom a chemist’s perspective, the environment is a complex, partic- ularly elaborate system. Hundreds

of chemically active compounds and minerals reside within the earth’s formations, and every patch of rock and soil is composed of its own particular mix. The waters that flow upon and through these formations harbor a vari- ety of organic and inorganic ligands. Where the water meets the rock is a poorly understood interfacial region that exhibits its own chemical processes.

Likewise, from a chemist’s point of view, the actinides are complex elements. Interrupting the 6d transition elements in the last row of the periodic table, the actinides result as electrons fill the 5f orbitals. Compared with the 4f electrons in the lanthanide series, the 5f electrons extend farther from the nucleus and are relatively exposed. Consequently, many actinides exhibit multiple oxidation states and form dozens of behaviorally distinct molecu- lar species. To confound matters, the light actinides are likely to undergo reduction/oxidation (redox) reactions and thus may change their oxidation states under even the mildest of condi- tions. Uranium, neptunium, and espe- cially plutonium often display two or

more oxidation states simultaneously in the same solution. Accordingly, the light actinides exhibit some of the rich- est and most involved chemistry in the periodic table.

As a result, the chemical interactions of actinides in the environment are in- ordinately complex. To predict how an actinide might spread through the envi- ronment and how fast that transport might occur, we need to characterize all local conditions, including the nature of site-specific minerals, temperature and pressure profiles, and the local waters’ pH, redox potential (Eh), and ligand concentrations. We also need a quanti- tative knowledge of the competing geochemical processes that affect the actinide’s behavior, most of which are illustrated in Figure 1. Precipitation and dissolution of actinide-bearing solids limit the upper actinide concentration in solution, while complexation and redox reactions determine the species’ distrib- ution and stability. The interaction of a dissolved species with mineral and rock surfaces and/or colloids determines if and how it will migrate through the environment.

Understanding this dynamic inter- play between the actinides and the en- vironment is critical for accurately as-

sessing the feasibility of storing nuclear waste in geologic repositories. The actinides1 are radioactive, long-lived, and highly toxic. Over the last few decades, vast quantities of transuranic actinides (those with atomic numbers greater than that of uranium) have been produced inside the fuel rods of com- mercial nuclear reactors. Currently, most spent fuel rods are stored above- ground in interim storage facilities, but the plan in the United States and some European countries is to deposit this nuclear waste in repositories buried hundreds of meters underground. The Department of Energy (DOE) is study- ing the feasibility of building a reposi- tory for high-level waste in Yucca Mountain (Nevada) and has already licensed the Waste Isolation Pilot Plant (WIPP) in New Mexico as a repository for defense-related transuranic waste. Given the actinides’ long half-lives, these repositories must isolate nuclear waste for tens of thousands of years.

1Although “actinides” refers to the fourteen elements with atomic numbers 90 through 103 (thorium through lawrencium), in this article we limit the term to refer only to uranium, neptunium, plutonium, americium, and curium, unless stated otherwise. These five actinides are the only ones that pose significant environmental concerns.

Wolfgang Runde

Actinides typically form � large molecular complexes�

in solution. Neptunium assumes � the Np(V) species, NpO2(H2O)5

+, � in many natural waters.

Complexation with different ligands � can stabilize actinides in solution�

and enhance their transport� through the environment�

� �

Redox reactions change an actinide's � oxidation state and help to establish �

equilibrium between species

Microbes can facilitate actinide� redox processes, while sorption� or uptake by the microbes may�

be a potential transport or � immoblilzation mode.�

�

At low actinide concentrations, sorption onto particulates, � clays, or rock surfaces determines the environmental behavior. � Actinides can also diffuse into rock, or coprecipitate with � ��� natural ligands and become � ��� incorportated into minerals

Np(V) complex sorbed � to montmorillonite

Solvated actinide species can� precipitate, forming a solid that in turn� determines the upper concentration�

limit for the solvated species.

NaNpO2CO3(s)

The Np(IV) species� Np(H2O)8

4+

NpO2(CO3)3 5–�

C om

pl ex

at io

n

�

D issolution�

P recipitation

Actinides can migrate by � water transport or by sorption�

onto mobile particulates or colloids

Figure I. Overview of Actinide Behavior in the Environment

Transport

R edox

Bioavailability

S or

pt io

n

�

�

Number 26 2000 Los Alamos Science 393

Chemical Interactions of Actinides in the Environment

Over the course of millennia, however, it is possible that water seeping into a repository will eventually corrode the waste containers. The actinides will then have access to the environment.

Research at Los Alamos has focused on characterizing the behavior of actinides in the environments surround- ing those repository sites. (Only small amounts of transuranic elements are generally found in other environments, as discussed in the box “Actinides in Today’s Environment” on page 396.)

Our solubility studies, for example, confirm that neptunium is more than a thousand times more soluble in Yucca Mountain waters than plutonium. That is because plutonium in those waters favors the IV oxidation state while neptunium favors the far more soluble V state. Thus, for Yucca Mountain, neptunium is the actinide of primary concern. WIPP, however, is built within in a geologically deep salt formation. In the extremely salty brines found there, the focus shifts to plutonium, since

reactions with chloride ions are known to stabilize plutonium in the VI oxida- tion state. Plutonium(VI) species are much more soluble than Pu(IV) species, and plutonium mobility would be enhanced at WIPP unless a reducing environment can be maintained within or around the waste containers.

Unfortunately, very few studies of actinide geochemistry can be conducted in situ, so that we are forced to simulate environmental conditions in the laboratory. The concentrations of the actinides in natural waters are typically on the order of 10–6 molar (M) or lower. While those concentra- tions are high enough to be of environ- mental concern, they are too low to allow direct study with conventional spectroscopies. We have thus had to adapt advanced spectroscopic tech- niques, such as x-ray absorption fine structure (XAFS) and laser-induced fluorescence spectroscopies, to study these toxic, radioactive elements. As an example, recent XAFS investiga- tions of Pu(IV) colloids have yielded significant insights into the colloids’ structure, which for years has compli- cated the determination of Pu(IV) solubility. Scanning electron microscopy (SEM) has helped us eluci- date the confusing sorption behavior of U(VI) onto phosphate mineral surfaces. The box on page 412 high- lights a few examples of these spectroscopic studies.

The scope of actinide interactions in the environment is too broad to cover in a single article. We will concentrate therefore on the solubility and specia- tion of actinides in the presence of hydroxide and carbonate ions, which are the two most environmentally rele- vant ligands. We will then briefly dis- cuss some aspects of actinide sorption and actinide interactions with microor- ganisms. The solubility and sorption of actinides pose two key natural barriers to actinide transport away from a geologic repository, while microorgan- isms pose a less-studied but potential third barrier. While we are fully aware of additional geochemical reactions in

Chemical Interactions of Actinides in the Environment

394 Los Alamos ScienceNumber 26 2000

1.0

0.5

–0.5

0

R e

d o

x p

o te

n tia

l, E

h

(v o

lts )

0 2 4 6

pH

8 10 12 14

PuO2 +

PuF2 2+

Pu3+

PuO2CO3(aq)

PuO2(OH2)2(aq)

PuO2OH(aq)

Pu(OH)4(aq)

PuOH2 +

PuO2(CO3)2 2–

PuO2(CO3)3 4–

Groundwater

Rain/streams

Pu(III)

Pu(IV)

Pu(V)

Pu(VI)

natural waters

PuO2F +

PuO2CO3 –Ocean

Figure 2. Pourbaix Diagram for Plutonium This Eh-vs-pH diagram is calculated for plutonium in water containing hydroxide,

carbonate, and fluoride ions. (The ligand concentrations are comparable to those found

in water from well J-13 at Yucca Mountain, Nevada, while the plutonium concentration

is fixed at 10 –5 M). Specific complexes form within defined Eh/pH regions, while the

stable oxidation states (colors) follow broader trends. For example, more-oxidizing

conditions (higher Eh values) stabilize redox-sensitive actinides like plutonium in the

higher oxidation states V and VI. The red dots are triple points, where plutonium can

exist in three different oxidation states. The range of Eh/pH values found in natural

waters is bounded by the solid black outline. In ocean water or in groundwater,

plutonium is likely to be found as Pu(IV), while in rainwater or streams, plutonium can

assume the V state. Other natural environments will favor Pu(III) or Pu(VI) complexes.

The dashed lines define the area of water stability. Above the upper line, water is

thermodynamically unstable and is oxidized to oxygen; below the lower line, water

is reduced to hydrogen.

the environment, such as coprecipita- tion, mineralization, and diffusion processes (and have thus included them in Figure 1), we will not discuss their roles here.

Oxidation States and Redox Behavior

Water is the dominant transport medium for most elements in the envi- ronment. Compared with the pH values and ionic strengths that can be obtained in the laboratory, most natural waters are relatively mild. Typically, they are nearly neutral (pH 5 to 9), with a wide range of redox potentials (from –300 to +500 millivolts) and low salinities (ionic strengths ≤ 1 molal). The water conditions determine which actinide oxidation states predominate and which actinide species are stable.

For example, Figure 2 shows the Pourbaix diagram (Eh vs pH) for pluto- nium in water containing the two most environmentally relevant ligands, the hydroxide (OH–) and carbonate (CO3

2–) ions. Even in this simple aqueous system, plutonium can exist in four oxidation states: III, IV, V, and VI. (Plutonium is unique in this regard. The Pourbaix diagrams for the other actinides are simpler.) The normal range of natural waters is outlined in the figure and overlaps with the stability fields of plutonium in the III, IV, and V oxidation states. Within this range, plutonium exhibits two triple points,

where species in three different oxida- tion states are in equilibrium.

Because of intrinsic differences in redox potentials, each actinide will exhibit a different set of oxidation states for a given set of solution condi- tions. In contrast to plutonium, U(III) is unstable under most conditions and oxidizes easily to U(IV), while U(V) disproportionates easily to U(IV) and U(VI). Neptunium(III) and Np(VI) are on the edges of the water stability region and can exist only under strongly reducing or oxidizing conditions, respectively. Americium and curium will only be found in the III oxidation state under most conditions. Similarly, all actinides beyond curium are domi- nated by the lanthanide-like trivalent oxidation state.

As a rule of thumb, we expect to find U(VI), Np(V), Pu(IV), Am(III), and Cm(III) as the prevalent oxidation states in most ocean or groundwater environments. But in other aqueous en- vironments, including streams, brines, or bogs, U(IV), Np(IV), and Pu(III,V,VI) are also common and like- ly to be stable. Table 1 summarizes the oxidation states of the actinides and highlights the environmentally most relevant ones. Some of the states listed in the table, such as Pa(III) or Pu(VII), can be synthesized only under extreme conditions, far from those found in natural systems.

Additional chemical processes occur- ring in solution are likely to influence the actinide’s oxidation state stability.

The stability of Pu(V) in natural waters containing carbonate is an example. The plutonium will complex with the carbonate ligands, and if the plutonium concentration is low (less than about 10–6 M), radiolytically induced redox reactions are minimized. Consequently, the stability of Pu(V) is enhanced and its disproportionation to Pu(IV) and Pu(VI) is reduced. As discussed later, the solubilities of solids formed from Pu(V) complexes are orders of magni- tude greater than those of Pu(IV) solids, so the enhanced stability of Pu(V) would serve to increase the total pluto- nium concentration in solution.

Another example for the stabilization of actinides in solution is the oxidation of Am(III) and Pu(IV) through radiolytic formation of oxidizing species, such as peroxide (H2O2) or hypochlorite (ClO–). At high actinide concentrations and hence under the influence of (its own) alpha radiation, those normally stable oxidation states are oxidized to form Am(V) and Pu(VI), respectively, especially in concentrated chloride brines.

Despite the complexity of actinide redox behavior, however, we should emphasize that within a given oxidation state, actinides tend to behave similarly. For example, we can study a U(IV) complex and from it infer the behavior of the analogous Np(IV) and Pu(IV) complexes. In the following discus- sions, therefore, we often refer to “generic” actinides—e.g., An(IV) complexes—where An is shorthand for actinide.

Effective Charge. It has been known for many years that An(IV) forms the strongest, most stable complexes and An(V) the weakest. This behavior follows directly from the effective charges of the ion.

In the III and IV states, the actinides form hydrated An3+ and An4+ ions in solution, respectively. In contrast, the highly charged ions in the V and VI states are unstable in aqueous solution and hydrolyze instantly to form linear trans-dioxo cations, AnO2

+ and

Chemical Interactions of Actinides in the Environment

Number 26 2000 Los Alamos Science 395

Table I. Oxidation States of Light Actinides a

Th Pa U Np Pu Am Cm

III III III III III III III IV IV IV IV IV IV IV

V V V V V VI VI VI VI

VII VII

aThe environmentally most important oxidation states are bolded.

AnO2 2+, with overall charges of +1 and

+2, respectively. These cations are often referred to as the actinyl ions. The covalent bonding between the actinide and the two oxygen atoms in the actinyl ion (O=An=O)n+, where n = 1 or 2, enhances the effective charge of the central actinide ion to 2.3±0.2 for AnO2

+ and 3.3±0.1 for AnO2

2+ (Choppin 1983). A ligand approaching the actinyl ion sees this enhanced effective charge and bonds directly to the actinide in the equatorial plane of the linear actinyl ion. Thus, the preference for actinides to form com- plexes generally follows their effective ion charges, as shown below:

The second and third lines show the ion’s effective charge and formal oxida- tion states.

Recent measurements of bond lengths in the actinyl ions of V and VI complexes reflect the above sequence. A higher effective charge correlates with a stronger, hence shorter, bond to the coordinated ligands. Accordingly, the bond length of the actinyl bonds in- creases from about 1.75 angstroms (Å) in the VI oxidation state (U=O is 1.76 Å; Pu=O is 1.74 Å) to about 1.83 Å in the V oxidation state (Np=O is 1.85 Å; Pu=O is 1.81 Å).

Because An(IV) has the highest effective charge, it forms the most stable complexes in solution and also forms the most stable precipitates with the lowest solubility. Conversely, An(V) complexes are the weakest, and its solubility-controlling solids are the most soluble. An(V) species are there- fore the most likely to migrate through the environment. Table 1 shows that the only transuranics that favor the V state are neptunium and plutonium. But in the environments surrounding repository sites, as well as in most groundwaters and ocean environments, plutonium will most likely assume the IV state. Thus neptunium is predicted to be the most

soluble and easily transported actinide and is the actinide of major concern in assessing the performance of Yucca Mountain.

Coordination Number and Ionic Radii. Another fundamental aspect about the actinide cations is that they tend to act as hard Pearson “acids,” which means that they form strong

complexes with highly anionic ligands by electrostatic interactions. Because the actinide-ligand bond is substantially ionic, the number of bound ligands (the coordination number) and the relative positions of the ligands around the cation are determined primarily by steric and electrostatic principles. Consequently, for a given oxidation state, a range of coordination numbers

An4+ > AnO2 2+ ≥ An3+ > AnO2

+

+4 +3.3 +3 +2.3 IV VI III V

Chemical Interactions of Actinides in the Environment

396 Los Alamos ScienceNumber 26 2000

Actinides in Today’s Environment

In 1942, Enrico Fermi built mankind’s first fission reactor, and nuclear reactions within

that nuclear pile created the transuranic elements neptunium, plutonium, americium,

and curium. With the subsequent development of nuclear power, the inventory of

transuranics has increased dramatically. As of 1990, the estimated accumulations in

spent nuclear fuel were approximately 472 tons of plutonium, 28 of tons of neptunium,

6 tons americium, and 1.6 tons of curium.

Fortunately, the amount of radioactive material accidentally released from nuclear

power plants has been relatively small. Environmental contamination is severe in only a

few locations. Major releases were due to a fire in a reactor at Windscale (now

Sellafield) in the United Kingdom in 1957, an explosion at the Kyshtym nuclear waste

storage facility in the former USSR in 1957, and an explosion and fire at the Chernobyl

reactor in Russia in 1986. In addition, a number of DOE sites in the United States,

including the Hanford Site in Washington, Rocky Flats Environmental Technology Site

in Colorado, and Idaho National Engineering and Environmental Laboratory in Idaho,

contain high local concentrations of actinides.

The main source of transuranics in the environment has been nuclear weapons testing.

Since the first detonation of a nuclear device in 1945 at Trinity Site in Alamogordo,

New Mexico, more than 4 tonnes of plutonium have been released worldwide (account-

ing only for announced nuclear tests). Atmospheric and underground tests have also

injected about 95 kilograms of americium into the environment, along with tiny amounts

of the still heavier element curium.

Until they were banned by international treaty in 1963, aboveground nuclear tests

propelled the actinides directly into the atmosphere. There they dispersed and migrated

around the globe before settling back to earth. Little can be done to isolate or retrieve

this material. It is a permanent, albeit negligible, component of the earth’s soils and

oceans and poses little health risk to the general public.

Underground nuclear tests injected transuranics into highly localized regions surround-

ing the detonation sites. These concentrated actinide sources became mineralized and

integrated into the rock matrix once their areas cooled and solidified. For the most part,

the nuclear material is fixed in place. However, plutonium from a test conducted at the

Nevada Test Site in 1968 has been detected in a well more than a kilometer away. Its

remarkably fast migration rate (at least 40 meters per year) is likely due to colloidal

transport through the highly fractured, water-saturated rock surrounding the detonation site.

is allowed. Coordination numbers be- tween 6 and 12 have been reported for complexes of An(III) and An(IV) and between 2 and 8 for the actinyl ions.

For example, the aquo ions of the actinides, which have only water mole- cules in the first coordination sphere surrounding the ion, exhibit coordina- tion numbers that vary with oxidation state: 8–10 for An3+, 9–12 for An4+, 4–5 for AnO2

+ and 5–6 for AnO2 2+.

These coordination numbers have been determined from nuclear magentic reso- nance (NMR) and XAFS spectroscopy studies (see the article by Conradson et al. on page 422).

Generally, the stabilities of actinide complexes at a given oxidation state increase with atomic number. This increase in stability follows the trend of the actinide contraction, wherein the

ionic radii decrease with atomic number. The ionic radii for An(IV) ions with coordination numbers of 8 are reported to be 1.00, 0.98, 0.96, 0.95, and 0.95 Å for uranium, neptunium, plutonium, americium, and curium, re- spectively (Shannon 1976).

Solubility and Speciation

The solubility of an actinide is lim- ited primarily by two properties: the stability of the actinide-bearing solid (the solubility-controlling solid) and the stability of the complexed species in solution. An actinide species will precipitate when its dissolved concen- tration is above the thermodynamic (equilibrium) solubility of its solubility- controlling solid, or if the kinetics

favors the formation of a solid of high- er solubility. Formation of the solid sets an upper concentration limit for the actinide in solution.

Figure 3 illustrates these concepts for An(III), where An is either americium or curium. The data points in the lower graph show the measured total concen- tration of An(III) in solution. At low pH with increasing carbonate concen- tration, the hydrated carbonated solid An2(CO3)3•2–3H2O(s), where (s) indi- cates the solid phase, readily precipi- tates and hence lowers the concentration of An3+ in solution. But complexation of An3+ with carbonate stabilize the solution species and increases the over- all An(III) solubility. As the carbonate concentration increases and the different carbonate species form, the total actinide concentration in solution first

Chemical Interactions of Actinides in the Environment

Number 26 2000 Los Alamos Science 397

Figure 3. Total An(III) Concentration Versus Carbonate Concentration (a) The partial pressure of CO 2 in natural

waters spans about three orders of

magnitude (gray area). The stability

diagram for An(III) solid phases (at pH 7)

shows that the An(III) hydroxocarbonate

is favored in most natural waters. Lower

partial pressures or soluble carbonate

concentration translate to increased

stability of the An(III) hydroxide; higher

pressures mean increased stability for

the normal carbonate solids. (b) The red

curve is the total concentration of An(III)

in solution when the solubility-controlling

solid is An 2(CO3)3•2–3H2O(s). Calculated

concentrations for individual solution

species are indicated by the gray lines

that run tangent to the solubility curve,

or by the gray dashed lines. At low

carbonate concentrations, the An 3+ ion

dominates the solution species. The total

An(III) concentration drops rapidly with

higher carbonate concentrations as more

of the solid phase precipitates. But

carbonate complexation begins to stabi-

lize the amount of actinide in solution.

By the time the anionic bis- and triscar-

bonate complexes dominate, the solution

concentration increases.

10–7 10–6 10–5 10–4 10–3 10–2

An2(CO3)3 •2–3 H2O(s)

Carbonate concentration (M)

To ta

l s o lv

a te

d A

n (I

II )

co n ce

n tr

a tio

n (

M )

10–7

10–6

10–5

10–4

244Cm 241Am

An(CO3)3 3–

An(CO3)2 –

An(CO3) +

An(OH)2+

An3+

10–2

10–4

10–6

10–8

10–10

10–5 10–3 10–210–4 10–1 10

An3+(aq)A n (I

II )

co n ce

n tr

a tio

n (

M )

AnOHCO3(s)

CO2 Partial Pressure (atm)

An(OH)3(s)

An2(CO3)3(s)

(a)

(b)

reaches a minimum and then increases due to the formation of anionic An(III) complexes. The solubility would change if conditions favored the formation of a new solubility-controlling solid.

We obtain solubility data by performing experiments in the laboratory under well-controlled conditions in which the actinide concentration is measured while varying the ligand concentration. Such experiments enable us to measure the solubility product Ksp and the stability constant β. These thermodynamic parameters are the basis for modeling the solubility boundaries for actinides in natural waters, as discussed in the box on “Solubility and Speciation Parameters.”

We should note, however, that meaningful interpretation of solubility data requires a detailed knowledge of the composition, crystallinity, and solu- bility of the solubility-controlling solid, along with the steady-state concentra- tion and composition of the solution species. Steady state is assumed to be established when the actinide concen- tration remains stable for several weeks or longer. But the solubility-controlling solids of the actinides generally precipi- tate as an amorphous phase because radiolysis affects their crystallization. Actinide solids may become less soluble with time as they transform from their initially formed, disordered structures into ordered crystalline solids, thereby lowering their free energy. Such a change, however, may not appear for several years. Thus, the solids formed in laboratory experiments may not represent the most thermodynamically stable solids with the lowest solubility. Laboratory-based solubility studies therefore provide an upper limit to actinide concentrations in a potential release scenario from a nuclear waste repository.

Because of their omnipresence, hydroxide ions (OH–) and carbonate ions (CO3

2–) are the most important ligands that complex to actinides in the environment. While other ligands, such as phosphate, sulfate, and fluoride, can lower the actinide concentration

Chemical Interactions of Actinides in the Environment

398 Los Alamos ScienceNumber 26 2000

Solubility and Speciation Parameters

Solubility products (Ksp) of actinide-bearing solid phases are indispensable primary data

for predicting the concentration limits of actinides in the environment. The solubility

product describes the dissolution reaction of the solid into its ionic components. For

example, the solubility product of AmOHCO3(s), where (s) refers to the solid phase, is

given by the product of the concentration of the dissolved ions Am3+, OH–, and CO3 2–

according to the following reaction:

(1)

The solubility product is then given by

(2)

where the brackets indicate ion concentration. The formation of an actinide complex in

solution, for example,

(3)

is described by the complex’s formation constant, βn. For the forward reaction in Equation (3), the formation constant is

(4)

where n refers to the number of ligands. For the reverse reaction, βn changes sign and is known as the stability constant. The concentration of an actinide solution complex

can be calculated by using the known solubility product of the solubility-limiting solid

and the formation constant of the species of interest as follows:

(5)

Accordingly, the total actinide concentration in solution is given by the sum of the con-

centrations of species present in solution as follows:

(6)

Complexation of the dissolved actinide ion with ligands in solution (large βn values) generally increases the total actinide concentration in solution.

Am(III) = Am + +

= Ksp 1 + + .

3+

n n

n m

m

[ ] [ ]   [ ] [ ][ ]

× [ ] [  ∑ ∑β β

Am(CO3)n 3–2n

CO3 2–�

CO3 2–�

OH–� OH–�

Am(OH)m 3–m

�

m]

= Am CO = K CO

OH .n

3+ 3 2–� n

n sp 3 2–� n–1

–�  

  [ ] [ ]

[ ] [ ]

β βAm(CO3)n 3–2n

βn n= Am3+

,

 

 

[ ][ ] Am(CO3)n

3–2n

CO3 2–�

Am + 3+ nCO3 2–� Am(CO3)n

3–2n

K = [Am ][OH ][CO3 2–�] , sp

3+ –�

AmOHCO3(s) Am + OH + CO3 2–� . 3+ –� �

��

(because of the low solubility of the corresponding solid phase), their concentrations in natural waters are generally low. Consequently, they cannot compete successfully with hydroxide or carbonate complexation. However, organic biodegradation prod- ucts, such as humates or fulvates, are generally present in natural aquifers and can potentially play a role in actinide solubility and migration.

Hydrolysis. The interaction of actinide and hydroxide ions produces monomeric and—particularly for Pu(IV)—polymeric solution species. The solid-phase structures are oxides, oxyhydroxides, or hydroxides of low solubility.

In the absence of carbonate or other strong ligands (or with very low concentrations of them), trivalent actinides form the positively charged or neutral hydroxo complexes of general formula An(OH)m

3–m, where m = 1, 2, or 3, that is An(OH)2+, An(OH)2

+, and An(OH)3(aq). Similarly, An(IV) forms complexes in solution of general formula An(OH)n

4–n, where n = 1, 2, 3, or 4. The solid hydroxide

An(OH)n(s), where n = 3 for An(III) and 4 for An(IV) or the hydrated oxide AnO2•nH2O for An(IV) are expected to control the solubility of actinides in both oxidation states.

Understanding the An(IV) solution chemistry is still a formidable chal- lenge. The high charge allows the An4+

ions to easily hydrolyze and form mul- tiple species simultaneously, even at low pH. Plutonium(IV) is known to hydrolyze even in dilute acidic solutions and, at concentrations greater than 10–6 M, will undergo oligomeriza- tion and polymerization to form very stable intrinsic colloids.

Recent x-ray absorption fine-structure (XAFS) studies, discussed on page 432 in the article by Conradson, have given more insight into the structural details of Pu(IV) colloids. The sub-micron- scale colloid particles are nominally composed of Pu(IV) hydroxo polymers and contain structural characteristics of

plutonium oxide, PuO2. However, the XAFS data reveal that these particles possess multiple functional groups, such as terminal hydroxo and bridging oxo and hydroxo groups. Each colloid particle can be considered an amor- phous oxyhydroxide compound. The functional groups originate from the polymerization of monomeric species formed through hydrolysis, and the number of functional groups depends on the history of the solution prepara- tion and aging process. While the chemical “structure” of every Pu(IV) colloid may be similar, each colloid can have a unique distribution of functional groups depending on solution prepara- tion, particle size, and water content. Once they form, however, the colloids are inordinately stable in near-neutral solution, presumably because over time, the hydroxo groups that join the pluto- nium atoms together in the polymer are converted into chemically resistant oxy- gen bridges.

The generation of Pu(IV) colloids is one of the main processes that compli- cate determining thermodynamic constants and accurately predicting soluble plutonium concentrations in natural waters. Both the colloids and other amorphous Pu(IV) hydroxide solids span a wide range of structural features, crystallinities, and stabilities, and hence exhibit a wide range of solubilities. The presence of Pu(IV) colloids will therefore complicate any calculation of plutonium solubility, as will be evident when we review plutonium solubility and speciation in Yucca Mountain waters.

Solid hydroxides and oxides are also expected to limit the solubility of An(V) and An(VI) complexes. For example, Np2O5(s) has been determined to be the solubility-controlling solid in waters from Yucca Mountain, while schoepite, UO2(OH)2•nH2O(s) is most stable for U(VI). Interestingly, these solids exhibit both acidic and basic characteristics. With increasing pH, their solubilities tend to decrease in acidic media and increase in alkaline (basic) solutions because of the forma-

tion of anionic hydroxo complexes: AnO2(OH)2

– for An(V) and AnO2(OH)n

2–n (n = 3 and 4) for An(VI). The hydrolysis of U(VI) is complicated even more by the formation of polymeric hydroxide species at high U(VI) concentration (>10–5 M), i.e., (UO2)3(OH)5

+ or (UO2)4(OH)7

+. We are currently conducting experiments to see if Pu(VI) undergoes a similar polymerization.

Carbonate Complexation. Carbon- ate is ubiquitous in nature. In surface waters, such as oceans, lakes, or streams, its concentration ranges gener- ally between 10–5 and 10–3 M. In groundwaters, its concentration is enhanced (up to 10–2 M) because of the increased CO2 partial pressure, which is about 10–2 atm deep underground compared with the atmospheric partial pressure of about 3 × 10–4 atm. These relatively high carbonate concentrations influence the environmental chemistry of actinides in all oxidation states.

Carbonate generally bonds to actinides in a bidentate fashion, that is, two of the three oxygen atoms in the complex bond to the actinide. Thus, actinide carbonate complexes are very strong and highly stable both in solution and in the solid state.

We have investigated the structure of carbonate complexes involving the actinyl ions AnO2

+ and AnO2 2+. (These

complexes are more soluble and thus more easily studied.) As shown in Figure 4, three monomeric solution complexes form. Depending on the solution conditions, these complexes may coexist with each other and with hydroxo complexes. It is therefore diffi- cult to obtain clean, unambiguous struc- tural data about a single species.

We can make inroads into that prob- lem by investigating the stability and structure of the triscarbonato species, AnO2(CO3)3

m-6, shown in Figure 4(c). This is the limiting (fully coordinated) An(V,VI) carbonate complex in solu- tion. At carbonated concentrations above 10–3 M, this complex forms at the exclusion of the others. Using a

Chemical Interactions of Actinides in the Environment

Number 26 2000 Los Alamos Science 399

battery of spectroscopic techniques, including multinuclear NMR spec- troscopy, x-ray diffraction, and XAFS, we have obtained formation constants, bond lengths, and other structural parameters for the An(V) and (VI) car- bonate complexes (Clark et al. 1999). This information has helped us under- stand various aspects of the An(V,VI) carbonate system. (The article by Clark that begins on page 310 describes struc- tural investigations of the Pu(VI) tris-carbonate.)

Carbonate Solids.In many natural waters, carbonate effectively competes with hydrolysis reactions, and mixed hydroxocarbonate solids are likely to form. These solids have general formu- las An(OH)(CO3)(s) for An(III) and An(OH)2m(CO3)n(s) for An(IV), where An stands for uranium, neptunium, plu- tonium, or americium. (The equilibria in these systems are quite complicated, and neither the structure nor the values for n has been determined accurately for An(IV) compounds.)

In the laboratory, however, we can study actinide solids of different compositions by increasing the carbon- ate concentration. This increase reduces the extent of hydrolysis and allows the formation, for example, of pure An(III) carbonate solids, such as Am2(CO3)3•nH2O(s) or NaAm(CO3)2•nH2O(s), which are the solubility-controlling solids in solu- tions containing high concentrations of sodium and carbonate. These solids are analogous to the ones formed by the trivalent lanthanides Nd(III) and Eu(III).

Similarly, we have observed the formation of solid An(VI) and An(V) compounds in solutions containing high concentrations of alkali metals and carbonate. These An(V,VI) carbonate solids are inherently interesting. Together, they can be viewed as a structural series, as evident in Figure 5. Binary An(VI) solids have a general composition AnO2CO3(s) and are made from highly ordered actinyl carbonate layers. However, An(V) solids need to

incorporate a monovalent cation, such as Na+, K+, or NH4

+, into their struc- ture (otherwise they would have an overall charge). Thus, we observe only ternary An(V) solids with general com- positions M(2n–1)AnO2(CO3)n•mH2O(s), where n = 1 or 2, M is the monovalent cation, and An stands for neptunium, plutonium, or americium.

As the number of cations incorporated into the An(V,VI) carbonate solids in- creases (stoichiometrically from 0 to 1 to 3), the structures become increasing- ly open, and the solids generally become more soluble under conditions in which the uncomplexed actinide ion dominates solution speciation. Replace- ment of the cations with larger ones

enhances the solubility of the actinide by several orders of magnitude. We exploited that fact when we studied the neptunyl “double carbonate” solids M3NpO2(CO3)2. These solids exhibit such low solubilities in near neutral solutions (10–4 to 10–7 M) that the

solution species cannot be studied by many spectroscopic methods. To modi- fy the solubility and obtain more- concentrated neptunyl solutions, we examined the use of large, bulky cations in an attempt to generate an unfavorable fit of the complex into the stable lattices shown in Figures 5(b) and 5(c). We found that tetrabutyl- ammonium provided stable solutions of NpO2CO3

– and NpO2(CO3)2 3– in

Chemical Interactions of Actinides in the Environment

400 Los Alamos ScienceNumber 26 2000

Figure 4. Carbonate Complexes of the Actinyl Ions Actinides in the V and VI state form similar carbonate complexes in solution, although

the charge of the species depends on the oxidation state. Thus, in the three monomer-

ic complexes shown in (a)–(c), m = 1 for An(V) and m = 2 for An(VI). Actinides in the

V state have the lowest effective charge and form the weakest complexes. In these

complexes, Np(V) has the longest actinyl bond (An=O) length (1.86 Å). We have also

determined only slight changes in the equatorial An–O distances to the carbonate

ligands: they vary between 2.42 and 2.44 Å for An(VI) and between 2.48 and 2.53 Å for

Np(V). (d) Uranium(VI) may also polymerize at high carbonate and uranium concentra-

tions to form the polynuclear actinyl carbonate complex (UO 2)3(CO3)6 6–. If present,

this polynuclear complex stabilizes U(VI) in solution and increases the overall U(VI)

solubility in comparison with the solubility of its monomeric counterpart, AnO 2(CO3)2 2–.

(c) AnO2(CO3)3 m–6

(d) (UO2)3(CO3)6 6–

Terminal carbonate

Bridging carbonate

(b) AnO2(CO3)2 m–4(a) AnO2(CO3)

m–2

Oxygen

Carbon

Actinide

millimolar concentrations over a wide pH range. The enhanced solubility allowed us to study the temperature dependence of the carbonate complexa- tion reactions with near-infrared (NIR) absorption spectroscopy and to obtain additional structural information about the Np(V) solution carbonate complex- es with XAFS spectroscopy.

The compositions and stabilities of these actinyl carbonate solids depend on the ionic strength and carbonate concentration and are controlled by the actinide, carbonate, and sodium concen- trations in solution. For example, the stability of the Np(V) solids Na(2n–1)NpO2(CO3)n•mH2O(s) is given by the solubility product Ksp:

Ksp = [Na +]2n-1 [NpO2

+] [CO3 2–]n .

For the An(VI) solid AnO2CO3(s), n is equal to 1 and its solubility is nominally independant from the sodium concentration in solution. At the low ionic strengths of most natural waters, an unrealistically high NpO2

+ concen- tration would be needed to precipitate a ternary Np(V) carbonate solid. Thus, solid An(V) or An(VI) carbonates are in general not observed in natural wa- ters. However, because the ternary compounds of Np(V) have been observed to form in concentrated electrolyte solutions, these solids are important for solubility predictions in brines found at WIPP.

Solubility and Speciation in Yucca Mountain Waters

As a way of tying together many of the concepts presented in the previous sections, we can consider the solubility and speciation of neptunium and pluto- nium in Yucca Mountain waters. (The article by Eckhardt, starting on page 410, discusses more of the research conducted for Yucca Mountain.) As mentioned earlier, neptunium tends to exist in the V oxidation state in natural waters and because of the lower effec- tive charge of the AnO2

+ ion, forms weak complexes. Thus, neptunium has the distinction of being the most soluble and transportable actinide and is the

Chemical Interactions of Actinides in the Environment

Number 26 2000 Los Alamos Science 401

UO2CO3 KPuO2CO3 K3NpO2(CO3)2•nH2O

Figure 5. Actinide(V,VI) Alkali Metal/Carbonate Solids Actinide(V,VI) carbonate solids form a “structural series” that progresses from dense, highly ordered layered structures to mor e

open structures. A single layer is shown above a perspective drawing of each solid. (a) An(VI) will combine with carbonate to f orm

binary solids, such as UO 2CO3(s) or PuO 2CO3(s). Within a layer, each actinide is bonded to two carbonates in a bidentate fashion

and to two others in a monodentate fashion. (This is in contrast to solution species, in which carbonate always bonds to an An( VI)

in a bidentate fashion.) (b) Actinide(V) ions cannot form binary carbonate solids, since the structural unit would have an over all

charge. Solid An(V) carbonates therefore form ternary complexes in conjunction with monovalent metal cations (M), such as Na +,

K+, or NH 4 +. Within a layer, each actinyl ion binds to three carbonate ligands in a bidentate fashion. The metal ions sit between the

layers and form the ternary complex MAnO 2CO3(s). This opens the structure, in comparison with the An(VI) binary structure, and

increases solubility. (c) At higher metal ion concentrations, metal ions will be incorporated into each layer to form very open

structures of general formula M 3AnO 2(CO3)2(s).

one of greatest concern when assessing the environmental safety of the poten- tial underground storage site at Yucca Mountain. Plutonium is considered the most toxic actinide and is always of environmental concern.

Our studies were conducted in water from a particular borehole in Yucca Mountain (J-13 water). The sodium and carbonate concentrations are low (both around 2 to 3 mM), as is the ionic strength of J-13 water (0.003 molal).

However, groundwater conditions with- in and around Yucca Mountain vary. The carbonate concentration changes (because the CO2 partial pressure changes with depth), as does the pH, Eh, and temperature. We mimicked this variation in the natural waters by con- ducting our studies at various pH val- ues while maintaining a constant ionic strength. We also conducted our studies at different temperatures (25˚C, 60˚C, and 90˚C) to approximate the range of

temperatures expected at the repository. Not surprisingly, changing these basic parameters greatly affected the actinide behavior (Efurd et al. 1998). Figure 6 summarizes the results of our studies.

The controlling solids for the actinides in J-13 water—in all oxidation states—were mainly oxides and hydrox- ides. We observed only the formation of greenish brown Np(V) crystalline solids of general formula Np2O5•nH2O(s), green Pu(IV) solids of general formula PuO2•nH2O and/or amorphous Pu(IV) hydroxide/colloids. The solubility of neptunium was about 10–3 M at pH 6, but dropped to about 5 × 10–6 M at pH 8.5. The neptunium species distribution similarly changed with conditions. At pH 8.5, about 31 percent of Np(V) was present as the neptunyl ion NpO2

+. The remaining soluble Np(V) was com- plexed with either carbonate or hydrox- ide, with about 58 percent NpO2CO3

–

and 11 percent NpO2OH(aq). At pH 6, however, Np(V) hydrolysis and carbon- ate complexation reactions were mini- mized, and all of the neptunium was present as NpO2

+. By comparison, at all pH values

examined in our study, 99 percent of the soluble plutonium was calculated to be in the IV state as Pu(OH)4(aq). When using Pu(OH)4(s) as the solubility -limiting solid in our geochemical models, the plutonium concentration in solution was predicted to be about 10–9 M (using the thermodynamic data available in the open literature). As seen in Figure 6, experimental data are about one to two orders of magnitude higher. The discrepancy presumably arises because plutonium solubility is instead controlled by amorphous oxides/hydroxides or colloidal forms of Pu(IV). (Note that if crystalline PuO2(s) were the solubility-controlling solid, Pu(IV) solubility would be about 10–17

M, well below the Pu(IV) hydroxide solubility range.)

The J-13 water is almost neutral (pH ~7) and has a redox potential of about +430 millivolts. But different water conditions would dramatically affect neptunium solubility, as illustrated

Chemical Interactions of Actinides in the Environment

402 Los Alamos ScienceNumber 26 2000

Figure 6. Solubility of Neptunium(V) and Pu(IV) J-13 Water The figure shows our measured total soluble concentrations of neptunium (red circles)

and plutonium (red square) in J-13 water at a fixed carbonate concentration of 2.8 mM

(indicated by the dashed line), as well as earlier measurements conducted at a higher

ionic strength (black circle and square). The lines are theoretical predictions for differ-

ent solubility-controlling solids. The Np(V) concentration is about 100 times greater

than that of Pu(IV). Our measurements confirm that the controlling Np(V) solid is the

oxide Np 2O5(s), rather than the previously postulated sodium carbonate solid

NaNpO2CO3(s). The solubility of plutonium in J-13 water presumably is controlled by

amorphous oxides/hydroxides or colloidal forms of Pu(IV), since the measured Pu(IV)

solubilities are nearly an order of magnitude higher than predicted by Lemire and

Tremaine for the solubility product of Pu(OH) 4(s). Our data are consistent with the

range of Pu(IV) solubilities as calculated from Ksp data of Pu(IV) hydroxide or hydrous

oxide reported in the open literature (gray area).

0.01 0.1 1.0

S p e ci

e s

co n ce

n tr

a tio

n (

m o l k

g −1

)

Ionic strength (mol kg−1)J-13 water

NaNpO2CO3 •nH2O(s)

Np2O5 •nH2O(s)

Np undersaturated

Pu (Nitsche et al. 1993)

Np (Nitsche et al. 1993)

Pu oversaturated

Pu(OH)4(s) (Lemire and Tremaine 1980)

10−2

10−4

10−6

10−8

10−10

10−12

Np oversaturated

in Figure 7 (Kaszuba and Runde 1999). Under strong reducing conditions (below about –120 millivolts), the total soluble neptunium concentration is limited by a Np(IV) solid phase, Np(OH)

4 (s), and is conservatively

assumed to be about 10–8 M. Similar to plutonium, the presence of a more ordered solid oxyhydroxide or oxide, such as NpO2(s), would reduce that conservative boundary by several orders of magnitude. Under these reducing conditions, Np(IV) also dominates the solution speciation in the form of the neutral complex Np(OH)4(aq). The total neptunium concentration will rise as the water conditions become more oxidiz-

ing and Np(V) becomes more stable in solution. Eventually, the Np(V) concen- tration is great enough to allow a Np(V) solid phase, Np2O5(s), to precip- itate. Formation of this solid sets an upper limit to the soluble neptunium concentration. Within the pH and Eh ranges of most natural aquifers (between 4 and 9 and between –200 and +700 millivolts, respectively), that concentration may vary by more than four orders of magnitude.

When modeling actinide release from a radioactive waste repository, we must anticipate the conditions at the source (e.g., close to a spent nuclear fuel container) as well as those in the

far-field environment. Spent nuclear fuel contains mainly UO2 and actinides in low oxidation states (III and IV). If conditions at the disposal site are reducing—for example, because of microbial activities or the reduction capacity of corroding iron from storage containers—the actinides will be main- tained in their lower oxidation states and soluble actinide concentrations may be low. If local reducing conditions change because oxidizing waters infil- trate from the surroundings, actinides will enter their more-soluble oxidation states and begin to migrate into the environment. Once far from the reposi- tory, conditions may change drastically,

Chemical Interactions of Actinides in the Environment

Number 26 2000 Los Alamos Science 403

pH

pH 7

4 6 8 1210 −400 −200 2000 600400

400 mV

300 mV

200 mV

100 mV

Np(OH)4(s)

Np(OH)4(s)

Np2O5(s)

Np2O5(s)

−100 mV0 mV10−8

10−6

10−4

10−2

1

10−10

N p c

o n ce

n tr

a tio

n (

m o l k

g −1

)

N p c

o n ce

n tr

a tio

n (

m o l k

g −1

)

10−6

10−4

10−8

Eh (mV)

(b)(a)

−200 mV

Figure 7. Neptunium Solubility in Yucca Mountain J-13 Water Neptunium solubility varies tremendously depending on the Eh and pH of the water. (a) The graph shows the neptunium concentra-

tion as a function of pH. Calculated solubility boundaries at various Eh values are superimposed on the graph. The colored re gion

contains mainly Np(V) in solution, while the upper and lower bold lines are the solubility boundaries of the Np(V) solid phase,

Np2O5(s), and the Np(IV) solid phase, Np(OH) 4(s), respectively. Solutions with neptunium concentrations above these boundaries are

oversaturated and will precipitate the respective solid. The total neptunium concentration in solution may vary by more than s ix

orders of magnitude over the entire pH range under oxidizing conditions. (b) This graph shows the solubility as a function of Eh at

a fixed pH of 7, and corresponds to the Np solubility as one moves along the red line in the previous graph. Under reducing con di-

tions (below –120 mV), Np(IV) dominates the solution speciation. Its concentration in solution is limited to less than 10 –8 molal by

the solubility of Np(OH) 4(s). As conditions become more oxidizing, Np(V) becomes the redox-stable oxidation state and begins to

dominate the speciation. [The Eh determines the Np(IV)/Np(V) ratio in solution, while the overall Np(IV) concentration is main tained

by the solubility of Np(OH) 4(s).] The neptunium concentration rises with redox potential until it reaches approximately 10 –4 molal at

an Eh of ~250 mV. Solutions containing a higher neptunium concentration are oversaturated, and the Np(V) solid phase Np 2O5(s)

precipitates. The soluble neptunium concentration remains constant from then on. The shaded area spans the region where Eh an d

solubility of the Np(IV) solid control the neptunium concentration in solution. Outside this area, the solubilities of the Np( IV) and

Np(V) solids govern neptunium solution concentration, independent of Eh.

and the transport behavior of redox- sensitive elements such as plutonium or neptunium may change. Clearly, evalu- ation of actinide mobility in the geo- chemical environment around nuclear waste repositories must account for soluble actinide concentrations that are controlled by the redox potential of the transport medium and as well by the solubility of the actinides’ solid phases.

Complexation with Chloride and Implications for WIPP

In most natural waters, actinides are usually associated with hydroxide or carbonate ligands. However, WIPP, which is designed to hold defense- related transuranic waste, is carved out of a salt formation. Water at the WIPP site thus consists of concentrated brines

that are saturated with chloride salts and exhibit very high ionic strengths (between 4 and 8 molal). With increas- ing distance from the repository, the brines’ ionic strength decreases and may ultimately reach the lower values of most natural surface and ground- waters. This variation complicates our transport models, since the species that form within the WIPP site will not be those that control actinide behavior far from the site.

Chloride has been shown to affect the solubility and speciation of actinide compounds significantly. As mentioned earlier, in most natural waters plutonium favors the IV and V oxidation states, but radiolytic hypochlorite formation in concentrated chloride solution may result in the formation of Pu(VI). Furthermore, complexation by chloride may enhance stabilization of Pu(VI) with respect to reduction to Pu(V) and Pu(IV).

We have studied the complexation reactions of U(VI) and Pu(VI) with chloride using a variety of spectroscopic techniques, including ultraviolet, visible, and NIR absorption and XAFS spectroscopies (Runde et al. 1999). We could determine when U(VI) and Pu(VI) chloride species, such as PuO2Cl

+ and PuO2Cl2, formed as a function of chloride concentrations (see Figure 8). Interestingly, the tris- and tetrachloro complexes are not formed in significant concentrations in NaCl solutions (up to the point of NaCl saturation). However, we have observed spectroscopically UO2Cl4

2– and PuO2Cl4

2– at very high LiCl concentrations and obtained structural data from single crystals synthesized from concentrated chloride solutions.

The interaction of actinides with chloride and other weak anionic ligands significantly affects solubility predic- tions and must be considered in risk assessments of WIPP and other nuclear waste repositories in salt formations. With regard to neptunium, chloride enhances the Np(V) solubility because of a strong ion interaction between the Np(V) compounds and chloride ions

Chemical Interactions of Actinides in the Environment

404 Los Alamos ScienceNumber 26 2000

870860850840830820

Wavelength (nm)

PuO2 2+

PuO2Cl +

PuO2Cl2

PuO2Cl3 –

PuO2Cl4 2–

–10

–5

5

10 (a)

(b)

0 1 2 3 4

In te

n si

ty (

a rb

itr a ry

u n its

)

R – φ (Å)

1-M HClO4

Pu=O Pu–Oeq

Pu–Cl

15-M LiCl

4-M NaCl

1-M NaCl

[Cl] (M)

1 4 15

Pu=O (Å)

1.74 1.75 1.75

(Å)

2.40 2.43 2.49

Pu–Cl (Å)

— 2.75 2.70

Pu–Oeq

F o u ri e r

T ra

n sf

o rm

m a g n itu

d e

0

Figure 8. XAFS Data and NIR Spectra of Pu(VI) in NaCl Solutions (a) By comparing XAFS data for Pu(VI) in 1-M NaCl, 4-M NaCl, and 15-M LiCl, and non-

complexing 1-M HClO 4, we have obtained condition-specific structural information

about the solution species. Chloride appears to bond above 1-M chloride in solution.

Bonding is clearly evident by 4-M chloride. As seen in the table (inset), the plutonyl

bond length (Pu=O) increases slightly after the chlorine bonds and replaces water

molecules (indicated by the Pu–O eq bond) in the equatorial plane of the O=Pu=O moiety.

(b) The absorption bands at 838, 843, 849, and about 855 nm, plus the characteristic

band at 830 nm for the PuO 2 2+ ion, indicate the formation of inner-sphere Pu(VI) chlo-

ride complexes. These bands allowed us to determine the formation constant, β, of the most important Pu(VI) chloride complexes in NaCl solutions, PuO 2Cl

+ and PuO 2Cl2(aq).

(see Figure 9). Runde et al. (1996) simulated this interaction using the Pitzer ion-interaction model and devel- oped a Pitzer parameter set to predict the Np(V) solubility at different NaCl concentrations and in artificial brine solutions (Runde et al. 1999). Neglect- ing the chloride interaction leads to a predicted Np(V) solubility about an order of magnitude too low at a neutral pH. The change of water activity with ionic strength affects the chemical potential of the solid phase because of hydration water molecules in the solid.

Np(V) solubilities were measured from oversaturation in two artificial WIPP brines, AISinR and H-17. The solutions reached a steady state after 241 days. The measured Np(V) solubil- ities reported in the literature are 2 × 10–6 molal and 1 × 10–6 molal in AISinR and H-17, respectively, while our model predicts Np(V) solubilities of 1 × 10–5 and 3 × 10–6 molal, respec- tively. The prediction is acceptable for the H-17 brine but lacks accuracy for the AISinR brine. This is because the model accounts only for interactions between Np(V) and Na+ and Cl– ions, while the Pitzer parameters for electrolyte ions such as K+, Mg2+, Ca2+, and SO4

2– (other constituents in the brines) are unknown. Additional data in various electrolyte solutions are required to develop a more accurate model for the Np(V) solution chemistry in complex brines.

Sorption

The solubility of an actinide species provides an upper limit to the actinide concentration in solution and may be considered as the first barrier to actinide transport. Once an actinide has dissolved in water, its sorption onto the surrounding rock or mineral surfaces is expected to act as a secondary barrier. When the actinide concentrations are far below the solubility limits, as is likely in most surface and groundwater systems, water/rock interfacial processes are

expected to dominate actinide transport through the environment. (Simple dilu- tion will lower the actinide concentra- tion, typically to trace concentrations, once the actinide migrates away from the contaminant or waste source.)

Actinide sorption onto mineral surfaces depends on many factors, such as the actinide’s speciation and concen-

tration, the flow rate and chemistry of the water, and the composition of the surrounding rocks. In addition, different mechanisms, such as physical adsorp- tion, ion exchange, chemisorption, or even surface precipitation, can fix the actinide to a mineral/rock surface or in general inhibit its transport. Further- more, redox processes complicate the

Chemical Interactions of Actinides in the Environment

Number 26 2000 Los Alamos Science 405

10−7

10−6

10−5

10−4

10−3

10−7 10−6 10−5 10−4 10−3 10−2 10−1

A n

(V )

(m o

l k g

−1 )

S p

a tia

l d is

tr ib

u tio

n (

% )

Carbonate concentration (mol kg−1)

NaNpO2CO3•3.5H2O(s)

Na3NpO2(CO3)2•nH2O(s)

100

80

60

40

20

NpO2 +

NpO2(CO3)3 5−

NpO2(CO3)2 3−

NpO2(CO3) −

5-M NaClO4

5-M NaCl

Figure 9. Solubility of Np(V) in Concentrated Electrolyte Solutions The upper graph shows the distribution of predominant Np(V) solution species as a

function of carbonate concentrations, with environmentally relevant concentrations

shown in gray. The lower graph shows the soluble Np(V) concentrations in 5-M NaCl

solution—conditions that are relevant for WIPP—and noncomplexing 5-M NaClO 4 solution over the same range. Similar data are obtained for Am(V). The neptunium

solubility is about an order of magnitude higher in NaCl because of the stabilizing

effect of actinide-chloride complexation reactions. The solubility-controlling solid for

both curves at low carbonate is the hydrated ternary Np(V) carbonate,

NaNpO2CO3•3.5H2O(s). Interestingly, this solid initially forms at high carbonate

concentrations. While it is kinetically favored to precipitate, it is thermodynamically

unstable. Over time, it transforms to the thermodynamically stable equilibrium phase

Na3NpO2(CO3)2•nH2O(s) (dashed curve).

sorption behavior of neptunium and plutonium. Unfortunately, we are only beginning to understand such actinide sorption and transport mechanisms.

Generally, batch-sorption experi- ments are performed to investigate how actinides will be retained in the envi- ronment. An actinide-containing solu- tion is allowed to come in contact with the geomaterial of interest, and we measure the amount of actinide that

remains in solution. The difference be- tween the remaining and initial actinide concentrations equals the amount that sorbed to the sample. This difference allows us to calculate the batch-sorption distribution coefficient Kd, defined as the moles of radionuclide per gram of solid phase divided by the moles of ra- dionuclide per milliliter of solution and reported as milliliters per gram (mL/g). Low Kd values mean low sorption, and

high values imply effective immobiliza- tion due to sorption onto rock surfaces.

Batch-sorption experiments provide helpful information about the sorption capacity of a mineral or soil. But while they are easy to perform, they have some drawbacks. The experiments are fundamentally static and provide only partial information about the dynamic, nonequilibrium processes that occur in nature.2 Additionally, the results of batch-sorption experiments are very sample specific and difficult to extrapo- late to the different conditions found throughout a repository.

To apply sorption data accurately, we must obtain sorption coefficients from both individual, well-characterized minerals, such as zeolites, manganese oxides, or iron oxyhydroxides, and from mineral mixtures with varying but known compositions. Sorption coefficients from these studies can serve as input data into a transport model that will estimate the actinide retention along water flow paths of known mineral content. The model predictions can then be compared with experimental results obtained from dynamic experiments that use complex soils and rocks.

Studies using minerals (hematite, calcite, gibbsite, albite, and quartz), clays (montmorillonite), and zeolites (clinoptilolite) have shown that Am(III) and Pu(IV) sorb very rapidly to these solids in the following descending order: hematite, montmorillonite, clinoptilolite, and calcite, followed to a much lesser extent by gibbsite, albite, and quartz. But other factors, such as the concentration, solution speciation, and oxidation-state distribution of the actinide species, significantly affect the extent of surface sorption. As an exam- ple, K

d values for neptunium sorption

onto hematite range from 100 to

Chemical Interactions of Actinides in the Environment

406 Los Alamos ScienceNumber 26 2000

Figure 10. Uptake of Pu(V) by Colloids Lu et al. (1998) obtained data for Pu(V) adsorption onto colloids, including hematite

(Fe2O3), silica (SiO 2), and montmorillonite (Mont), in natural J-13 water and in a

synthentic (syn) J-13 water. The colloid concentration was maintained at 200 mg/L,

with the colloids being about 100 nm is size. The sorption clearly depends on the type

of mineral, but the sorption mechanism remains unknown. The time-dependence of

the sorption onto montmorillonite suggests a more complicated interaction than

occurs with oxide minerals.

Table II. Desorption of Plutonium from Colloids in J-13 Water after 150 Days

Colloid Material Amount Desorbed Pu(IV) Pu(V)

(%) (%)

Hematite (α-Fe2O3) 0.0 0.01 Goethite (α-FeOOH) 0.57 0.77 Montmorillonite 8.23 0.48 Silica (SiO2) 19.93 1.04

A d so

rb e d P

u (V

) (%

)

120

Fe2O3 (syn)

SiO2 (syn)

SiO2

Fe2O3

100

80

60

40

20

0 0 40 80 120

Time (h)

160 200 240

Mont (syn)

Mont

2Flow-through, or dynamic, column experiments, wherein the actinide solution is allowed to migrate through a short column of packed geomaterial, provide information about dynamic transport properties. Dynamic column experi- ments are difficult to perform and interpret, and most of the sorption studies on the actinides are still performed as batch experiments.

2000 mL/g, while plutonium sorption onto hematite results in K

d values

above 10,000 mL/g. This difference is mainly caused by the different stabili- ties of the actinide oxidation states and their complexation strengths. The low charge of the NpO2

+ ion results in a low sorption of neptunium, while Pu(IV) dominates the complexation and sorption processes of plutonium. Pluto- nium(V) is expected to follow the low sorption affinity of Np(V).

Lu et al. (1998) determined the sorption of Pu(V) onto oxide and clay colloids (Figure 10). Sorption onto iron oxides, such as hematite, is strong and irreversible, but only 50 percent of the Pu(V) is sorbed on the montmorillonite (clay). However, the sorption mecha- nisms are largely unknown. The time- dependence of the sorption onto the clay suggests a more complicated inter- action than occurs with oxide minerals.

Interestingly, the desorptionof Pu(V) does not follow this general pat- tern. Lu et. al. (1998) also discussed the desorption of Pu(IV) and Pu(V) from various colloids. Table II summarizes

their findings: Surprisingly, Pu(IV) exhibits a greater tendency to desorb than Pu(V). Particularly striking is the relatively high desorption of Pu(IV) from silica, which indicates a different sorption mechanism or different binding to the silica functional groups than to the Fe(III) minerals, including possible changes in plutonium oxidation states.

These investigations indicate the importance of understanding the oxida- tion-state stability of the actinide sorbed on a mineral surface. Redox-active minerals, such as manganese or iron oxide/hydroxides, are present in subsur- face environments and may alter the redox states of uranium, neptunium, and plutonium. At the Rocky Flats Environmental Technology Site, seasonal variations of plutonium con- centrations in settling ponds correlate astonishingly well with soluble manganese concentrations, indicating a potential release of plutonium by way of seasonal redox cycling.

It has been generally hypothesized that the environmentally mobile Pu(V)

is reduced to Pu(IV) through the formation of strong mineral-surface complexes. Quite surprisingly, Neu et al. (2000) showed that Pu(V) is not reduced when sorbed on MnO2. X-ray absorption near-edge structure indicated the presence of Pu(V) on MnO2, even though the starting material of the batch-sorption experiment was colloidal Pu(IV) hydroxide. Their data also indicated that the plutonium was oxidized to Pu(VI). The only time they did not observe the higher-oxidation- state surface species was when Pu(IV) was complexed with a strong ligand, such as nitrilo triacetic acid (NTA). These observations strongly support the hypothesis that plutonium undergoes a redox cycling that is enhanced by surface interactions. We need more data about the surface species, however, to model such interactions.

Obviously, batch experiments alone cannot elucidate sorption mechanisms. Specifically, we have to distinguish be- tween surface precipitation and interfa- cial molecular reactions in order to model sorption behavior. In the event

Chemical Interactions of Actinides in the Environment

Number 26 2000 Los Alamos Science 407

keV

0 1 2 3 4 5

Ca

P

Ca

O

Ca

Ca

O

P

U

keV

0 1 2 3 4 5

Figure 11. SEM Micrograph of Uranium Precipitate on Apatite A solution containing U(VI) at 10 –4 M was kept in contact with apatite for 48 hours. The central SEM micrograph (1100 × magnifica- tion) shows crystalline plates on the smooth apatite mineral surface. Analysis of these plates by energy-dispersive spectroscop y

(right graph) indicated the presence of calcium, oxygen, uranium, and phosphorus, consistent with crystalline uranium phosphate .

Similar analysis of the bare regions of apatite (left graph) showed signals corresponding to only calcium, oxygen, and phosphor us.

The precipitates were observed at uranium concentrations above 10 –5 M. Studies show that at lower concentrations, uranium sorbs,

rather than precipitates, to the mineral surface.

of a surface precipitation, the Kd value may represent the solubility product of the surface precipitate but will not reflect the sorption behavior of an actinide species.

For example, phosphate-containing minerals, such as apatite, are known to reduce thorium and uranium concentra- tions in contaminated solutions. Sorp- tion distribution coefficients for U(VI) and Th(IV) have been reported in the literature, but there are alsoreports of U(VI) and Th(IV) precipitates. Our SEM studies have identified microcrys- talline precipitates on the apatite when the initial U(VI) concentrations were above 10–5 M (see Figure 11). At lower U(VI) concentrations, no surface precipitation was observed. Thus, at the higher concentrations, the Kd values actually reflect a solubility product, while at the lower concentrations, the decrease of U(VI) in solution results from sorption of the actinide onto the mineral surface.

Complexation reactions can influ-

ence the actinide sorption processes that also depend strongly on pH (Figure 12). For the trivalent actinides Am(III) and Cm(III), the sorption reaches a peak under conditions of near-neutral waters. There, the cationic An(III) solution species predominate solution speciation. At lower pH, the An3+ ion is stable and does not participate in bonding with hydroxide or carbonate ligands. At higher pH, negatively charged An(III) species are formed that exhibit reduced interactions with chemically active groups at the mineral surface. Mineral or rock surfaces at high pH are generally negatively charged, and apparently the negatively charged actinide species are repelled from the anionic surface groups. As such, anionic actinide species behave as simple anionic lig- ands, such as SO4

2–, CrO4 2– or AsO4

3–. In addition to An(III) species, the same trend of enhanced stability and low surface retention has been observed for the An(VI) complex UO2(CO3)3

4–. Enhanced modeling efforts have

begun to address the fundamental mechanism of actinide adsorption onto mineral surfaces. Janecky and Sattel- berger (unpublished results) modeled the adsorption behavior of Np(V) on calcite by minimizing the free energy of the coordination. The structural parameters of NpO2CO3

–, which is the main solution complex in neutral and low-carbonate waters (such as J-13 water), were used to mimic the adsorp- tion. The resulting surface-complex modeling indicates a distance of about 4 Å between the neptunium center and the nearest calciumions in the calcite structure. Data from preliminary XAFS studies support this distance. Although this agreement between experiment and theory is encouraging, more detailed studies have to be performed to develop much more accurate transport models, including this surface complexation model.

We are also trying to better under- stand the sorption/desorption reactions of actinides with colloids and the

Chemical Interactions of Actinides in the Environment

408 Los Alamos ScienceNumber 26 2000

pH

S p e ci

e s

d is

tr ib

u tio

n (

% )

20

40

60

80

An(CO3)3 3–An(CO3)2

–An(CO3) +

An(OH)2+

An3+

4 5 7 96 8 10

100

80

(a)

(b)

60

40

20

0

0.1-M NaCl

1.0-M NaCl

S o rp

tio n (

% )

Figure 12. Correlation between An(III) Solution Speciation and Sorption onto Montmorillonite (a) Even with an order-of-magnitude

difference in NaCl concentration, sorption

of the trivalent actinides Am(III) and

Cm(III) peaks in near-neutral waters. At

lower pH, the An 3+ ion is stable and does

not participate in complexation reactions

with the most common ligands, hydroxide

or carbonate, no matter if the ligands are

in solution or on a surface . At higher pH,

negatively charged An(III) species are

formed that increase the species’

solubility and reduce their sorption

because they are repelled from the

negatively charged functional groups on

the mineral surface. (b) Calculations for

a weak sodium chloride solution

(0.1-M NaCl and CO 2 partial pressure of

0.01 atm) suggest that sorption is mainly

correlated with the presence of cationic

solution species. The gray area denotes

the pH range of natural waters.

actinides’ resulting transport character- istics. This area of environmental migration received heightened attention with the discovery of plutonium in a borehole at the Nevada Test Site (Kersting et al. 1999). The plutonium had evidently migrated 1.3 kilometers in only 30 years. (Each nuclear test is characterized by a unique 239Pu/240Pu isotope ratio, and that ratio can be used to identify the plutonium source.) As discussed in the article by Maureen McGraw, we now believe that colloidal transport was responsible for this remarkably fast movement of plutonium through the water-saturated rock. It is not clear, however, whether the transport was facilitated by intrinsic plutonium colloids or natural (clay or zeolite) colloids. What is clear is that transport models to date have underesti- mated the extent of colloidal transport on plutonium mobility.

Microbial Interactions

We have discussed the most promi- nent chemical processes involving actinides in the environment: dissolu- tion and precipitation of solids, com- plexation in solution, and sorption onto surfaces. But there is growing attention to the importance of actinide interac- tions with microorganisms. Great vari- eties of microorganisms exist in all aquifers and soils, and many survive under the most extreme conditions, such as around hydrothermal vents at the bottom of the ocean, within deep subterranean aquifers, or within the hypersalinity of geologic salt beds.

Microorganisms can accumulate on solid surfaces often via formation of biofilms, or they may be suspended in the aquifer. As such they act as mobile or even self-propelled colloids. The concern is that the actinides can be associated with microbes, either through surface binding or through metabolic uptake, and be transported by them through the environment. However, the microorganisms’ role in environ- mental transport behavior of radioactive

contaminants is still virtually unknown. Microorganisms, especially bacteria,

are also known to mediate redox processes. They may reduce elements from higher oxidation states and use the enthalpy of the redox reaction to grow and reproduce. Through metabolic activity, microbes can also establish reducing conditions by changing the pH and Eh of the local waters. While direct experimental evidence is still lacking, it is almost certain that some microbes can catalyze the transformation of uranium, neptunium, and plutonium into less-soluble forms. Several bacteria species have been reported in the litera- ture to effectively reduce the highly mobile U(VI) into the far less soluble and mobile U(IV).

Given these characteristics, microbes could pose a third natural barrier to actinide transport from geologic repositories and could also help remediate actinide-contaminated soil, groundwater, and waste streams. The ultimate goal is to use microorgan- isms as “living” backfill to immobilize actinides through processes such as bioreduction, biosorption, or bioprecip- itation and mineralization. At this time, however, actual applications of bacteria in remediation strategies and field demonstrations are rare.

Recent WIPP-related studies at Los Alamos have focused on the interaction of urnaium and plutonium with the bacteria’s cell walls or byproducts, such as extracellular biopolymers, and the toxicity and viability of microorganisms in the presence of the actinides. Francis et al. (1998) have shown that actinides are associated only at low levels with the halophilic bacteria isolated from muck pile salts around the WIPP site. The extent of association varied with bacterial culture, actinide species, pH, and the presence of organic or inorganic complexants. The radiolytic effects on the microbial viability were found to be negligible.

However, other microorganisms can concentrate metal ions to a great extent. Hersman et al. (1993) observed a strong uptake of plutonium by pseudomonas

sp. Within the first week of the experi- ment, the bacteria concentrated plutoni- um to levels that were nearly 10,000 times greater than a sterile control or- ganism. Lower uptake was observed for contact times of two weeks or longer. Another strain of this bacteria, pseudomonas mendocina, is known to remove lead from Fe(III) hydroxide/ oxides and was investigated for its potential to remove radionuclides from those environmentally abundant miner- als. In fact, Kraemer et al. (2000) found that 20 percent of the sorbed plutonium was removed by this aerobically grow- ing bacteria after about four days. The experimental conditions are presently optimized to remove the majority of the sorbed plutonium and to maximize its potential application in soil remediation strategies.

Leonard et al. (1999) also studied the inhibitory and toxic effects of mag- nesium oxide, MgO, on these WIPP-de- rived bacterial cultures. Magnesium oxide is of specific interest because it was chosen as backfill material at the WIPP repository to maintain the pH between 9 and 10, to reduce the amount of infiltrating water, and to limit the amount of CO2 that can accumulate through degradation of cellulose and other organic compounds in the wastes. Initial studies indicated an inhibited cell growth as the amount of MgO increased, as well as a decrease in microbially generated gas (N2O) by active halophilic bacteria. Ultimately, the toxic effect of MgO inhibits micro- bial growth, and consequently microor- ganisms may be unable to create a reducing environment at the repository that would stabilize actinides.

A more detailed study on the micro- bial association mechanism included analysis of the uptake of uranium and plutonium by sequestering agents and extracellular polymers (exopolymers). As an example, some exopolymers pro- duced by bacillus licheniformisconsist of repeating γ-polyglutamic acid (γ- PGA). Neu et al. (manuscript in prepa- ration) observed actinide hydroxide pre- cipitation at metal-to-glutamate ratios

Chemical Interactions of Actinides in the Environment

Number 26 2000 Los Alamos Science 409

higher than 1:20, but both U(VI) and Pu(IV) ions formed soluble complexes at ratios 1:100 and 1:200, respectively. Presumably at the higher PGA concen- trations, the actinide ion is encapsulated and isolated from hydroxo-colloid for- mation. The binding strength for U(VI) and Pu(IV) is lower than for other strong ligands and chelators, such as Tiron or NTA, but on the same order as for other exopolymers or humic acids.

An upper binding boundary for the polyglutamate exopolymer was deter- mined to be about 0.5 micromoles Pu(IV) per milligram of γ-PGA. Active cells of bacillus licheniformisbonded to more than 90 percent of the introduced Pu(IV)—up to 8.4 micromoles of Pu(IV) were added. As expected, other functional groups on the exterior cell surface (in addition to the PGA exo- ploymer) increase the number of bind- ing sites and enhance metal binding.

This research clearly shows that actinides bind with microorganisms and their metabolic byproducts. Depending on the static or dynamic nature of the microorganism, the interaction of pluto- nium and other actinides with extra- cellular ligands may affect the actinides’ mobility both away from storage containers and within the environment. Future studies will focus on understanding the mechanisms of actinide/microbial interactions and exploiting their potential for remedia- tion and immobilization efforts. Uptake studies of plutonium by means of sequestering agents are highlighted in the box on “Siderophore-Mediated Microbial Uptake of Plutonium” on page 416.

Concluding Remarks

From this overview of actinide inter- actions in the environment, it should be clear that such interactions are com- plex. Each local environment is unique, and the site-specific conditions deter- mine which actinide species predomi- nate as well as each species’ overall transport and migration characteristics.

In general, actinide solubilities are low in most natural waters—below micromolar concentrations. Actinides in the V oxidation state will have the highest solubilities; those in the IV oxidation state, the lowest. Conversely, An(V) species show low sorption onto mineral and rock surfaces, and An(IV) species tend to sorb strongly. Typically, neptunium is the only actinide that favors the V oxidation state, and as such it is the most soluble and most easily transportable actinide under envi- ronmental conditions. Plutonium favors the IV oxidation state in many natural waters, and normally its solubility is quite low. Once plutonium is in solu- tion, its migration is dominated by a variety of processes, including redox reactions in solution or on mineral surfaces, sorption, colloidal transport, simple diffusion into rock interfaces, coprecipitation, and mineralization. Furthermore, increasing research activites indicated that microbes can affect actinide migration, either through direct association or by helping to maintain a reducing environment.

To understand this complex set of interacting processes requires that we have a strong, multidisciplinary founda- tion in environmental chemistry, molec- ular science, and interfacial science and the requisite spectroscopies, subsurface testing, and laboratory experiments. But we also need to interpret experimental results within the context of the multi- component natural system.

A case in point is the recent identifi- cation of a higher oxide of plutonium of general formula PuO2+x(s), which is formed by water-catalyzed oxidation of PuO2(s). As discussed by Haschke in the article beginning on page 204, this Pu(IV) “superoxide” may also contain Pu(VI), which would increase plutoni- um solubility. In a recently published paper (Haschke et al. 2000), the authors speculated that the Pu(VI) could dissolve and be transported more easily from a nuclear waste storage site than previously estimated. Further, they conjectured that Pu(VI) might have caused the rapid migration of plutonium

at the Nevada Test Site. In these specu- lations, however, they failed to include influential environmental factors.

While we acknowledge that the potential presence of Pu(VI) in a PuO2(s) matrix may increase the dissolution kinetics, the chemical condi- tions of the local environment (water and rock compositions) determine the stability of plutonium oxidation states in the aqueous phase and the overall plutonium mobility. Under ambient conditions and in most aqueous envi- ronments, Pu(VI) is unlikely to be stable. Once in solution, it is typically reduced to lower oxidation states and, depending on the water redox potential, may eventually precipitate as Pu(OH)4(s) or PuO2(s). Consequently, the fact that the superoxide PuO2+x(s) may include Pu(VI) does not automati- cally imply enhanced plutonium solubil- ity or mobility in the natural environ- ment. Furthermore, given the results on colloidal transport, it is improbable that Pu(VI) is responsible for the plutonium migration at the Nevada Test Site.

As the inventory of actinides in interim storage facilities grows and storage containers age, increasing the risk of accidental releases, and as the safety of permanently storing nuclear waste in geologic repositories must be evaluated, it becomes increasingly important to understand how actinides will interact with the environment. More sophisticated models are needed to account for all the potential migration paths away from an actinide source. Theoretical and experimental scientists will be challenged for years by the demands of developing those models. �

Chemical Interactions of Actinides in the Environment

410 Los Alamos ScienceNumber 26 2000

Further Reading

Brown, G. E., V. E. Heinrich, W. H. Casey, D. L. Clark, C. Eggleston, A. Felmy, D. W. Goodman, et al. 1999. Chem. Rev.99: 77–174.

Clark, D. L., S. D. Conradson, D. W. Keogh, M. P. Neu, P. D. Palmer, W. Runde, et al. 1999. X-Ray Absorption and Diffraction Studies of Monomeric Actinide Tetra-, Penta-, and Hexavalent Carbonato Complexes. In Speciation, Techniques and Facilities for Radioactive Materials at Synchrotron Light Sources, OECD Nucl. Energy Agency. 121.

Choppin, G. R., and E. N. Rizkalla. 1994. Solution Chemistry of Actinides and Lanthanides. In Handbook on the Physics and Chemistry of Rare Earths. Edited by K. A. Gschneidner Jr. and L. Eyring. New York: North-Holland Publishing Co.

Choppin, G. R., J.-O. Liljenzin, and J. Rydberg. 1995. Behavior of Radionuclides in the Environment. In Radiochemistry and Nuclear Chemistry. Oxford: Butterworth-Heinemann.

Choppin, G. R., 1983. Radiochim. Acta32: 43.

Dozol, M., R. Hagemann, D. C. Hoffman, J. P. Adloff, H. R. Vongunten, J. Foos, et al. 1993. Pure and Appl. Chem.65: 1081.

Efurd, D. W., W. Runde, J. C. Banar, D. R. Janecky, J. P. Kaszuba, P. D. Palmer, et al. 1998. Environ. Sci. Tech.32: 3893.

Francis, A. J., J. B. Gillow, C. J. Dodge, M. Dunn, K. Mantione, B. A. Strietelmeier, et al. 1998. Radiochim. Acta82: 347.

Hascke, J., T. Allen, and L. Morales. 2000. Science287: 285.

Hersman, L.E., P. D. Palmer, and D. E. Hobart. 1993. Mater. Res. Soc. Proc.294: 765.

Kraemer, S. M., S. F. Cheah, R. Zapf, J. Xu, K. N. Raymond, and G. Sposito. 2000. Geochim. Cosmochim. Acta(accepted).

Kazuba, J. P., and Runde, W. 1999. Environ. Sci. Tech.33: 4427.

Kersting, A. B., D. W. Efurd, D. L. Finnegan, D. J. Rokop, D. K. Smith, J. L. Thompsom. 1999 Nature 397: 56.

Kim, J. I. 1986. Chemical Behavior of Transuranic Elements in Natural Aquatic Systems. In Handbook on the Physics and Chemistry of the Actinides. Edited by A. J. Freeman and G. H. Lander. New York: North-HollandPublishing Co.

Langmuir, D. 1997. Chap. 13. InAqueous Environmental Geochemistry. Upper Saddle River, NJ: Prentice Hall.

Leonard, P. A., B. A. Strietelmeier, L. Pansoy-Hjelvik, R. Villarreal. 1999. Microbial Characterization for the Source- Term Waste Test Program (STTP) at Los Alamos. Conference on Waste Management 99, Tucson, AZ.

Lemire, R. J. and P. R. Tremaine. 1980. J. Chem. Eng. Data.25: 361.

Lu, N., I. R. Triay, C. R. Cotter, H. D. Kitten, and J. Bentley. 1998. "Reversibility of sorption of plutonium-239 onto colloids of iron oxides, clay and silica." Abstract of Agronomy, p.31 ASA, CAA, SSSA Annual Meetings, Baltimore, MD.

Neu, M. P., W. Runde, D. M. Smith, S. D. Conradson, M. R. Liu, D. R. Janecky, and D. W. Efurd. 2000. EMSP Workshop, Atlanta, GA.

Nitsche, H., A. Muller, E. M. Standifer, R. S. Deinhammer, K. Becraft, T. Prussin, R. C. Gatti. 1993. Radiochima. Acta.62: 105.

Runde, W., S. D. Reilly, M. P. Neu. 1999. Geochim. Cosmochim. Acta. : 3443.

Runde, W., M. P. Neu, D. L. Clark. 1996. Geochim. Cosmochim. Acta60: 2065.

Shannon, R. D. 1976. Acta Cryst.32A: 751.

Chemical Interactions of Actinides in the Environment

Number 26 2000 Los Alamos Science 411

Wolfgang Runde earned B.S. and Ph.D. degrees in chemistry from the Technical Univer- sity of Munich, Germany, in 1990 and 1993, under the direction of Professor Dr. J. I. Kim. He then worked for two years as a scientific staff

member at the Institute for Radiochemistry, TU Munich. During this time, he was invited to work as a guest scientist at San- dia National Lab- oratories on tech- nical aspects of the actinide source term

model for WIPP. Wolfgang then spent a year as postdoctoral fellow at the G. T. Seaborg Institute for Transactinium Science at the Lawrence Livermore National Laboratory before joining the Chemical Science and Technology Division at Los Alamos National Laboratory in 1996. He is presently a staff member and leader of the Actinide Environmental Chemistry team in the Environmental Science and Waste Technology Division with a joint appointment in the Nuclear and Radiochemistry group in the Chemical Science and Technology Division. His research interests are in the areas of inorganic, environ- mental, and radiochemistry of f elements, including the actinides, with the focus on nuclear waste isolation and the environment. His research activities have been closely affiliated with the national nuclear waste programs in the US (WIPP and Yucca Mountain) and Germany (Gorleben).