contents

Fifty years of investigation

The path to PuO2+x

Nearly fifty years ago, J. Drummond and G. Welch reported the first preparations of plutonium dioxides from plutonium (III), (IV), and (VI) compounds by thermal decomposition of the hydride, nitride, carbide, oxalate, sulfate, chloride, fluoride, iodate, hydroxide, and nitrate. They reported apparent oxygen stoichiometries (the oxygen:plutonium ratio) that ranged from 2.04 to 2.09. These results were confirmed at the then-called Los Alamos Scientific Laboratory by G. Waterbury and others and later at Rocky Flats in Colorado by J. Moseley and R. Wing using thermogravimetric techniques.

In each case x-ray diffraction was performed on the oxide products to determine the change in lattice constant with oxygen:plutonium ratio. The x-ray analyses were insensitive to the small differences in the ratios observed in the various oxides. However, several researchers reported that the diffraction lines were broadened, and they also documented the descriptive chemistry of plutonium dioxide via the differing colors obtained from various starting materials and preparative methods (see photos on Page 2).

In contrast, the earliest work on plutonium oxides reported the existence of a monoxide (PuO), a sesquioxide (Pu2O3), and a dioxide (PuO2). The dioxide was found to be the highest-valent oxide based on W.H. Zachariasen's x-ray diffraction work (see Page 9), which showed a fluorite-type face-centered cubic (fcc) structure with a lattice constant of 5.396 angstroms (). E. Westrum Jr. found that the dioxide did not react further with high-pressure oxygen at 400 degrees Celsius or with ozone at 800 degrees Celsius; furthermore, J. Katz and D. Gruen showed that plutonium dioxide was similarly resistant to further oxidation by nitrogen dioxide.

Additional thermogravimetric studies by J. Stakebake and M. Dringman suggested that adsorbed oxygen might be playing a role in weight gains or losses. Furthermore, E. Jackson and M. Rand excluded interstitial lattice oxygen insertion from consideration on the basis of density measurements from plutonium dioxide solids (the expansion of the crystalline lattice was not considered). At the time, x-ray diffraction techniques were limited in their ability to detect slight changes in the oxygen:plutonium ratio, and it was also difficult to quantify and interpret the line broadening in the powder-diffraction photographs. Thermodynamic calculations, compiled by L. Brewer, revealed plutonium dioxide as the most stable phase in the plutonium-oxygen system and thus the continual quest for plutonium oxides higher than PuO2 was abandoned by the early 1970s.

However, a 1964 publication by K. Bagnall and J. Laidler, which contradicted the earlier Westrum study, noted that oxidation of plutonium(IV) hydroxide with ozone yielded PuO30.8H2O. The presence of plutonium(VI) was shown with ultraviolet-visible (UV-vis)spectroscopy, but x-ray diffraction data were compromised by the inability to obtain well-crystallized samples. Despite this singular observation of higher-valent plutonium, by the late 1980s the crystallography and the thermodynamics of the plutonium-oxygen system were well established with acceptance of the published plutonium-oxygen phase diagram. However, the descriptive chemistry of plutonium oxide continually revealed anomalies, such as the large variation in the solubility of PuO2 (103 variance) reported by various researchers.

The Bagnall and Laidler report, coupled with renewed interest in actinide-oxide behavior concerning migration in the environment, processing actinide materials for long-term storage, and pragmatic aspects of plutonium corrosion, sparked a resurgent interest in actinide research. These various drivers yielded several interesting developments. The first was in 1993 when Stakebake, D. Larson, and J. Haschke reported that the reaction of plutonium metal with water vapor gave a higher oxide formed at the gas-solid interface with a fluorite structure and contained hexavalent plutonium.

Plutonium-bearing solid-state storage in convenience (slip-lid) and crimp-seal containers. These types of storage vessels have noted leakage and contamination problems.

The second finding came in 1997 when O. Krikorian and coworkers reported that the dependence of the transpiration rate of plutonium on the oxygen pressure at high temperature could only be explained by the presence of PuO3 in the vapor phase. Additional mass spectroscopic studies monitoring PuO3 vapor in equilibrium above PuO2 by C. Ronchi and others in 2000 also revealed the possibility of preparing plutonium oxides with a valence state above (IV).

Independent of these research findings the Defense Nuclear Facilities Safety Board issued findings and recommendations following an incident investigation in which a worker was contaminated while handling a standard storage package (containing 2.5 kilograms of plutonium metal) that developed a breach as a result of radiolysis-driven decomposition of the packaging.

Partially as a result of the desire to ensure the safety of actinide workers, a new plutonium processing, packaging, and storage standard was developed. The DOE 3013 standard (in a number of versions) requires thermal stabilization of oxide materials prior to packaging (without the use of plastics) for storage. A full description of the standard and the fundamental understanding that underpins our technical basis for these procedures will be discussed in the next issue of ARQ.

The DOE 3013 standard states that the plutonium content of processed excess plutonium solids can vary from 88 weight percent (essentially pure PuO2) to about 30 weight percent. With such a range of composition, the plutonium-bearing solid can be in contact with a variety of other materials. Typically, these "impurities" include alkali metal chlorides, magnesium chloride (MgCl2), calcium chloride (CaCl2), magnesium oxide (MgO), ferric oxide (Fe2O3), and other materials that are not well characterized. In addition, these solid mixtures come into contact with atmospheric conditions where the moisture content may not be known or well controlled prior to packaging. Therefore, in the areas of excess plutonium disposition and storage, the issue of water and other small molecule interactions with pure or impure plutonium oxide materials and metal remains a continual concern.

Cutaway and exterior views of nested, welded 3013 containers for long-term storage of plutonium-bearing solids.

Before the DOE 3013 standard, incomplete process control over small molecule reactions in these types of systems led to changes in materials stoichiometry, containment breaches, and dispersal of material resulting from pressurization, corrosion of the container, and collapse of sealed containers because of the formation of partial vacuum. The exact nature of these reactions and the resulting storage implications were not entirely understood, although there had been studies that attempted to explain a large body of experimental observations. Some of these additional studies will also be described in the next issue of ARQ.

To provide a stronger technical basis for the new storage standard and explain the nature of these small-molecule reactions with plutonium metal and dioxide, we became involved in actinide research in 1995 with a follow-on study of the 1993 Stakebake results by investigating the interaction of plutonium dioxide powders with water from 373 to 623 kelvin (K). A suite of experimental techniques was used that included thermogravimetric analysis-some coupled to mass spectrometry, pressure-volume-temperature (PVT) methods, and x-ray and neutron diffraction of the processed solids. Reaction rates and oxide compositions were determined from changes in sample mass or pressure versus time. Gaseous and solid products were analyzed using mass spectrometry and diffraction methods, respectively. Oxide products have also been characterized by x-ray photoelectron spectroscopy (XPS) and are further discussed in the final article of this issue.

The plutonium-oxide specimens used in our studies were formed by oxidation of electrorefined alpha-phase metal containing approximately 100 parts per million americium as the major metallic impurity. The specific surface area of the oxide was 4.8 square meters per gram. The initial oxide stoichiometry was determined to be PuO1.97 based on the measured lattice parameter and data from the correlation of the cubic lattice parameter (a0) at fixed oxygen:plutonium ratios with temperature reported by E. Gardner and others.

Arrhenius plot for results for the plutonium dioxide-water reaction from 298 to 623 kelvin (K). Microbalance and pressure-volume-temperature (PVT) data are shown by filled circles and open ovals, respectively. The 298 K data point, represented by a filled triangle, is from an earlier study.

The reactivity of plutonium dioxide with water was examined using PVT and microbalance measurements at temperatures ranging from 473 to 623 K. These PVT and microbalance results show linear increases in pressure and mass as a function of time. This behavior indicates that water is irreversibly reactingwith the plutonium oxide. Mass spectrometric analysis of gas samples taken after termination of the tests show that only water (H2O) and hydrogen (H2) were present in the gas phase. Qualitative results were similar to those observed at 298 K and suggest the following reaction:

PuO2(s) + x H2O arrow PuO2+x (s) + H2(g) . (1)

This equation implies that a fraction of the plutonium is oxidized to an oxidation state greater than plutonium(IV) with a simultaneous increase in mass. Rate data for the reaction of plutonium dioxide and water versus temperature are shown in the figure below. This type of analysis, termed an Arrhenius plot, provides a measure of how difficult (or easy) the reaction proceeds as a function of temperature. The rate of reaction at 298 K is a value obtained from three independent kinetic measurements for the plutonium dioxide-plus-water reaction. The kinetic results from the microbalance and PVT measurements can be described by a single relationship describing the dependence of the rate on temperature:

ln[Rate of reaction] = - 6.441 - (4706/T). (2)

The activation energy (Eact; a quantitative indication of how hard it is to drive a reaction) for reaction 1 is 39.3 2.5 kJ M-1. The uncertainty in Eact results primarily from the uncertainty in the average rate at 298 K.

X-ray and neutron diffraction data show that the oxide product formed during reaction has a fluorite-related fcc structure derived from that of the dioxide. The results of eight measurements with calculated oxygen:plutonium ratios from 2.016 to 2.169 are shown in the figure on the next page. The lattice parameter of PuO2+x is a linear function of composition:

ao () = 5.3643 + 0.01764 O:Pu. (3)

When the PuO2+x product was heated above 673 K in subsequent thermogravimetric analysis (TGA) experiments, a mass loss was observed at approximately 633 K and the lattice constant of the resulting oxide returns to that of PuO2, indicating that PuO2+x is stable only up to 633 K. The oxygen:plutonium ratio calculated from the measured mass loss in the TGA experiments and the hydrogen generation from the PVT experiments are in excellent agreement.

Kinetic results for oxidation of plutonium dioxide by water show that the reaction has a prominent temperature dependence from 298 to 623 K. The temperature dependence observed for the rate demonstrates that one is observing a chemical reaction with water, and not the result of radiolysis of water. The rate of a purely radiolytic process is expected to be independent of temperature at a fixed water pressure. If formation of PuO2+x is promoted by radiolysis of water, the largest fractional contribution to the oxidation rate is anticipated at low temperature in a system with a high surface concentration of water.

Equation 3 describes the continuous variation of lattice parameter (a0) with stoichiometry. This behavior nominally follows Vegard's Law (see sidebar on Page 8) and clearly indicates that O in the PuO2+x structure exists as a solid-solution, not a new phase as previously conjectured. This slight lattice expansion, measured for the first time, suggests that additional oxygen is accommodated on interstitial sites in the fluorite lattice of plutonium dioxide. Whereas oxidation of plutonium(IV) on cationic sites of dioxide would tend to shrink the lattice, accommodation of oxide ions in interstitial lattice sites causes lattice expansion. The opposing changes are apparently of comparable magnitude, and the net effect is a small dependence of ao on the composition of PuO2+x. These findings were published by J. Haschke, T. Allen, and L. Morales in 2000.

Dependence of the cubic lattice parameter, a0, on oxide composition of PuO2x. Oxygen/plutonium data for PuO2-x from E. Gardner and others are shown by triangles. The composition of the starting oxide is given by the diamond. Values of a0 obtained from the microbalance and pressure-volume-temperature (PVT) experiments are given by circles and squares, respectively. In the molecular structure inset, oxygen atoms are shown in red, and plutonium in black.

Additional studies using x-ray absorption near-edge structure (XANES) and x-ray absorption fine structure (XAFS) were performed on PuO2+x samples in which x was carefully varied by monitoring the amount of hydrogen gas produced in reaction 1. The position of the XANES edge demonstrated that the oxidation state of the PuO2+x studied was a mixture of (IV) and (V), not (VI), as previously thought. These results were key in understanding how the oxygen was incorporated into the crystal lattice and were of enormous pragmatic use in interpreting environmental migration scenarios from real-world samples at Rocky Flats and Hanford in Washingon. A more complete description of the XAFS results and the local-range structure in PuO2+x is provided on Page 16.

Lattice constant for the binary solid solution of potassium bromide (KBr) and postassium chloride (KCl).

Vegard's Law

Crystalline solid-solution alloys have unit cell dimensions that are frequently linear with concentration. This property is referred to as Vegard's Law and has been generally assumed to imply that atomic volumes of the constituent elements are independent of the concentration or stoichiometry in a solid solution. For this model of solid solutions, an average lattice constant is assumed to arise from the close packing of hard-sphere atoms. Given this assumption, the average lattice parameter (Rav) is predicted to have a linear dependence on concentration

Rav = CARA + CBRB

where RA and RB refer to the atomic radii of the pure elements A and B at concentration CA and CB, respectively.

Taken together, the chemical, thermodynamic, crystallographic, and spectroscopic results form a self-consistent data set that provides a very strong case for the further oxidation of PuO2 to a hyperstoichiometric solid solution denoted as PuO2+x where the plutonium exists in a mixed-valence state.

Comparison of results for the plutonium-oxygen and uranium- oxygen systems suggests that the hyperstoichiometric regions above the dioxide compositions differ substantially. Although early work on the uranium system indicated the existence of a cubic UO2+x solid solution at oxygen:uranium ratios up to 2.33, subsequent studies show that UO2 coexists in equilibrium with the tetragonal U4O9 phase at temperatures below 573 K and that the UO2+x phase is stable only at elevated temperatures. Cubic lattice parameters measured for the UO2+x product suggest that ao decreases with increasing composition. Although UO2+x and PuO2+x apparently have similar structures with respect to the metal-atom lattice, it is evident that the manner in which oxygen-either as an oxide (O2-), a hydroxyl (OH-), or water (H2O)is accomodated in the host lattice is notably dofferent for UO2+x and PuO2+x. This difference prompts one to question the underlying reasons for this behavior and the potential impacts on our fundamental notions of electronic structure in the actinide oxides.

The stage is set for the next act in this unfolding adventure in actinide-materials science. Much work remains to be done to fully understand the relationships between the thermodynamic, crystallographic, and electronic properties of the actinide oxides. We are pursuing further studies with PuO2+x and the alkalai metal plutonates; however, the contrast between the uranium and plutonium oxides challenges our understanding of localized versus itinerant hybridization and the resulting consequences on crystal structure and physical properties. A systematic examination of all the actinide dioxides, with emphasis on uranium, neptunium, and plutonium (UO2, NpO2, and PuO2) and their ability to accommodate extra lattice oxygen, is needed.

This article was contributed by Luis Morales of the Nuclear Materials Technology Division.

Further reading

O.J. Wick, editor, Plutonium Handbook. A Guide to the Technology, American Nuclear Society, LaGrange Park, Ill., 1980.

J.J. Katz, G.T. Seaborg, and L.R. Morss, editors, The Chemistry of the Actinide Elements, 2nd edition, Chapman and Hall, New York, 1986.

H. Wriedt, "The O-Pu (Oxygen-Plutonium) System," Bulletin of Alloy Phase Diagrams, 11, No. 2, 184, 1990.

J. Stakebake, D. Larson, and J. Haschke, "Characterization of the Plutonium-Water Reaction II: Formation of a Binary Oxide Containing Pu(VI)," Journal of Alloys and Compounds, 202, 251, 1993.

O. Krikorian, A. Fontes, B. Ebbinghaus, and M. Adamson, "Transpiration Studies on the Volatilities of PuO3(g) and PuO2(OH)2(g) from PuO2(s) in the Presence of Steam and Oxygen and Application to Plutonium Volatility in Mixed-Waste Thermal Oxidation Processors," Journal of Nuclear Materials, 247, 161, 1997.

C. Ronchi, F. Capone, J. Colle, and J. Hiernaut, "Volatile Molecule PuO3 Observed from Subliming Plutonium Dioxide," Journal of Nuclear Materials, 280, 111, 2000.

J. Haschke and T. Ricketts, "Adsorption of Water on Plutonium Dioxide," Journal of Alloys and Compounds, 252, 148, 1997.

J. Haschke and T. Allen, Interactions of Plutonium Dioxide and Water and Oxygen-Hydrogen Mixtures, LA-13537-MS, January 1999.

J.M. Haschke, T.H. Allen, and L.A. Morales, "Reaction of Plutonium Dioxide with Water: Formation and Properties of PuO2+x" Science, 287, 2000.


NMT | LANL | DOE
Phone Book | Search | Help/Info

L O S  A L A M O S  N A T I O N A L   L A B O R A T O R Y
Operated by the University of California for the US Department of Energy

Questions? - Copyright © UC 1998-2000
- For conditions of use, see Disclaimer