Stability of Peroxide-Containing Uranyl Minerals

See allHide authors and affiliations

Science  14 Nov 2003:
Vol. 302, Issue 5648, pp. 1191-1193
DOI: 10.1126/science.1090259


Minerals containing peroxide are limited to studtite, (UO2)O2(H2O)4, and metastudtite, (UO2)O2(H2O)2. High-temperature oxide-melt solution calorimetry and solubility measurements for studtite (standard enthalpy of formation at 298 kelvin is –2344.7 ± 4.0 kilojoules per mole from the elements) establishes that these phases are stable in peroxide-bearing environments, even at low H2O2 concentrations. Natural radioactivity in a uranium deposit, or the radioactivity of nuclear waste, can create sufficient H2O2 by alpha radiolysis of water for studtite formation. Studtite and metastudtite may be important alteration phases of nuclear waste in a geological repository and of spent fuel under any long-term storage, possibly at the expense of the commonly expected uranyl oxide hydrates and uranyl silicates.

The minerals (UO2)O2(H2O)4, studtite, and (UO2)O2(H2O)2, metastudtite, are the only known peroxide-bearing minerals. Both are readily synthesized by adding H2O2 to a U-bearing solution (1, 2), but the conditions for their formation are uncertain. Studies of natural analogs (3) and many laboratory experiments (46) have been designed to simulate the alteration of spent fuel under conditions similar to those expected in the proposed radioactive material repository located in Yucca Mountain, Nevada, United States. These studies have emphasized uranyl oxide hydrates and uranyl silicates as dominant alteration products and have not reported the formation of uranyl peroxides. McNamara et al. (7) found studtite that had formed on the surface of spent nuclear fuel reacted at 298 K with deionized water for 1.5 years and proposed that it grew by incorporating peroxide created by the alpha radiolysis of water. Studtite is a major alteration phase associated with spent nuclear fuel contained in the K East Basins of the Hanford site, United States (8). Metastudtite formed on the surface of UO2 under irradiation with a 4He2+ (alpha-particle) beam, presumably incorporating peroxide formed by the alpha radiolysis of water (9). Studtite was also found on nuclear material (“lava”) after the Chernobyl Nuclear Plant accident (10). Thus, uranyl peroxides appear to be important alteration phases of nuclear waste. Here we examine the thermodynamic stability of natural studtite from the same specimen used by Burns and Hughes (11) for crystal structure determination.

High-temperature oxide-melt solution calorimetry (12) obtained the drop solution enthalpy of studtite (Table 1), that is, the enthalpy associated with dropping the sample from room temperature and dissolving it in a 3Na2O· 4MoO3 melt at 976 K (13, 14). Earlier experiments on UO2 and UO3 (15) confirm that U6+ is the stable oxidation state of uranium that is dissolved at low concentrations in 3Na2O· 4MoO3 at 976 K; UO2 is oxidized during dissolution, and UO3 is dissolved directly from an oxygen atmosphere. Thus, the dissolution of studtite similarly involves no oxidation-reduction. Samples dissolved readily and quickly. H2O was evolved into the gas air space above the solvent and swept out of the calorimeter by the flowing gas, as shown for other hydrous phases (16). The product of dissolution of studtite is dissolved UO3 in the melt and evolved H2O and O2 vapor.

Table 1.

Thermochemical cycles for studtite, (UO2)O2(H2O)4, in the calculation at 298 K of the enthalpy of formation from UO3, H2O, and O2, ΔHr-oxide; enthalpy of formation from UO3, H2O, and H2O2, ΔH °r-peroxide; enthalpy of formation from the elements, ΔH °f; enthalpy of formation from UO22+, H2O2, and H2O, ΔHr-aqueous; and enthalpy of formation from UO3(H2O)0.8, H2O, and H2O2, ΔHr-dehydrated schoepite. soln, solution.

Reaction ΔH (kj/mol) Ref.
1 ΔHds (studtite) (UO2)O2(H2O)4 (crystalline, 298 K) = UO3 (soln, 976 K) + 4H2O (g, 976 K) + 1/2O2 (g, 976 K) 273.2 ± 3.6 View inline
2 ΔHds (UO3) UO3 (crystalline, 298 K) = UO3 (soln, 976 K) 9.5 ± 1.5 View inline
3 ΔHhc (H2O) H2O (l, 298 K) = H2O (g, 976 K) 69.0 View inline
4 ΔHhc (O2) 1/2O2 (g, 298 K) = 1/2O2 (g, 976 K) 10.0 View inline
5 ΔH°f (UO3) U (crystalline, 298 K) + 3/2O2 (g, 289 K) = UO3 (crystalline, 298 K) -1223.8 ± 0.8 View inline
6 ΔH°f (H2O) H2 (g, 298 K) + 1/2O2 (g, 298 K) = H2O (l, 298 K) -285.8 ± 0.1 View inline
7 ΔH°f (H2O2) H2 (g, 298 K) + O2 (g, 298 K) = H2O2 (l, 298 K) -187.8 View inline
8 ΔH°f(dehydrated schoepite) U (crystalline, 298 K) + 1.9O2 (g, 298 K) + 0.8H2 (g, 298 K) = UO3(H2O)0.8 (crystalline, 298 K) -1469.9 ± 2.3 View inline
Thermodynamic cycles
ΔHr-oxide (298 K) = -ΔH(1) + ΔH(2) + 4ΔH(3) + ΔH(4)
UO3 (crystalline, 298 K) + 4H2O (l, 298 K) + 1/2O2 (g, 298 K) = (UO2)O2(H2O)4 (crystalline, 298 K) 22.3 ± 3.9
ΔH°f (298 K) = -ΔH(1) + ΔH(2) + 4ΔH(3) + ΔH(4) + ΔH(5) + 4ΔH(6)
U (crystalline, 298 K) + 4H2 (g, 298 K) + 4O2 (g, 298 K) = (UO2)O2(H2O)4 (crystalline, 298 K) -2344.7 ± 4.0
ΔH°r-peroxides (298 K) = ΔH°f - ΔH(5) - 3ΔH(6) - ΔH(7)
UO3 (crystalline, 298 K) + 3H2O (l, 298 K) + H2O2 (l, 298 K) = (UO2)O2(H2O)4 (crystalline, 298 K) -75.7 ± 4.1
ΔH°r-dehydrated schoepite = ΔH°f - ΔH(8) - 2.2ΔH(6) - ΔH(7)
UO3(H2O)0.8 (crystalline, 298 K) + 2.2H2O (l, 298 K) + H2O2 (l, 298 K) = (UO2)O2(H2O)4 (crystalline, 298 K) -58.2 ± 4.3

Measured drop-solution enthalpies were used in thermodynamic cycles to establish the enthalpy of formation of studtite from oxides (Table 1). Reference data were used for the standard enthalpies of formation of the binary oxides to calculate the standard enthalpy of formation of studtite from the elements.

The calorimetric data enable the calculation of enthalpy of formation (ΔHreaction) (Table 1) for the reaction (where l is liquid and g is gas): Embedded Image Embedded Image Embedded Image Embedded Image(1) Though the entropy change, ΔS°, for this reaction is not known, it must be strongly negative because of the consumption of O2 gas. Thus, relative to UO3, water, and oxygen, studtite is not stable at room temperature, and is likely to become even more unstable as temperature increases.

To explore whether studtite can be stable when H2O2 is produced and constantly replenished by radiolysis, we calculated the enthalpy of the reaction: Embedded Image Embedded Image Embedded Image Embedded Image(2) The data for H2O2 (17) refer to liquid H2O2.

In reality, hydrogen peroxide is present in the dilute aqueous solution. Nevertheless, the large exothermic enthalpy of formation for H2O2 is likely to persist even when infinitely dilute aqueous reference states are taken. Also, reaction 2 is likely to have a near zero ΔS because no gases are present in the products or the reactants, in contrast to reaction 1.

The enthalpy of formation of studtite from the elements Embedded Image Embedded Image Embedded Image(3) is –2344.7 ± 4.0 kJ mol–1. Previous estimates were based on the heat of formation value for the aqueous uranyl ion, UO22+, and the heat of precipitation of studtite (18). That value, –2394.9 kJ mol–1, suggests that studtite is stable in the absence of peroxide (–27.9 ± 4.0 kJ mol–1 calculated for reaction 1). However, because of the assumptions needed for that calculation, we consider our new value more direct and more reliable.

Studtite formation from the aqueous uranyl ion in nature may be described by: Embedded Image Embedded Image Embedded Image Embedded Image(4) We measured the solubility constant Ksp = [UO2+2][H2O2]/[H+]2 for reaction 4 in a glove box under continuously flowing N2 (Table 2). Each determination was done by weight using 100 g of ultrapure H2O with (UO2)2+ in solution ranging from 2.63 × 10–5 to 2.50 × 10–8 M. Titrations were done with H2O2 solution until a precipitate formed at about pH 3, which x-ray powder diffraction confirmed was studtite. Five independent determinations yielded Ksp, ranging from 1.32 × 10–3 to 1.37 × 10–3. As an example, studtite will form in a solution at 298 K that contains 25 μM (UO2)2+ at pH 3.1 when H2O2 levels reach 3.4 × 10–3 M. These results are in accord with the formation of studtite or metastudtite on UO2 wafers at H2O2 concentrations of 3.5 × 10–3 M at a pH of 3.8 and 298 K [measured for the uranyl ion in aqueous solution, as in reaction 4 (9)].

Table 2.

Calculation of the solubility constant, Ksp, for studtite, UO2O2(H2O)4.

Sample [UO2+2] (M) [H2O2] (M) pH Ksp
1 2.63 × 10-5 7.57 × 10-5 2.91 1.32 × 10-3
2 1.18 × 10-5 1.04 × 10-4 3.02 1.35 × 10-3
3 2.50 × 10-7 3.39 × 10-3 3.10 1.34 × 10-3
4 2.09 × 10-7 3.07 × 10-3 3.16 1.34 × 10-3
5 2.50 × 10-8 9.99 × 10-3 3.37 1.37 × 10-3

We contend that alpha radiolysis of water is the only feasible source of sufficient peroxide in nature for studtite formation, given that most studtite occurrences are found underground where no solar radiation reaches. Levels of radioactivity in U-rich rocks can be surprisingly high, owing to the importance of daughter products of U. The approximate alpha activity of 1 g of uraninite (88.1% U) is 2.45 μCi, which corresponds to 5.4 × 106 dpm (19).

Using single crystals of uraninite from the Topsham Mine in Maine, United States, formed during the Permian Period, we measured an alpha-particle dose at the surface of 35,000 dpm cm–2. The H2O2 yield for alpha irradiation in bulk neutral water is G = 0.985 molecules per 100 eV (20). If peroxide does not break down in solution over time, then for 1 cm3 of H2O in contact with 1 cm2 of Topsham uraninite, H2O2 concentrations would reach 3.5 × 10–3 M in 2100 years, assuming an average alpha-particle energy of 5.5 MeV. However, given that a 5.5-MeV alpha particle deposits all of its energy within 40 μm of water (9), H2O2 concentration would reach 3.5 × 10–3 M H2O2 in a 40-μm-thick layer of water in contact with 1 cm2 of uraninite in only 8 years. If such a layer of water were trapped between two uraninite crystals, only 4 years would be needed to accumulate that level of H2O2. The presence of certain ions in groundwater catalyzes the breakdown of H2O2 (21), but continued production of H2O2 will eventually overcome such effects. Even if a substantial fraction of H2O2 decomposes during this accumulation period, a time scale of years rather than millennia appears likely for the formation of studtite.

If natural radiation levels in U deposits can create sufficient peroxide for the formation of studtite or metastudtite, why are these minerals not much more common? Alpha radiolysis of water produces oxidizing reactants (such as H2O2 and ·OH) and reducing species (such as H2 and aqueous electrons) in equal proportions; the overall effect depends on the reactivity of the species (9). H2 is relatively inert at temperatures below 363 K; thus radiolysis under these conditions leads to oxidizing conditions locally, especially if the H2 can escape from the water. Peroxide concentration at sufficient levels for studtite precipitation is most likely to occur in thin films of water on mineral surfaces, especially if evaporation of water concentrates the peroxide further. In contrast, systems that contain larger volumes of water relative to the quantity of U in the rock, water that is rich in dissolved ions, and systems that involve water flow are less likely to attain sufficient peroxide concentration for the formation of uranyl peroxides. Studtite is thermodynamically unstable in systems with no peroxide. Thus, not only its formation but also its persistence requires specific geochemical conditions that continue to replenish aqueous peroxide. Such conditions, relatively rare in nature, are much more common at the surface of spent nuclear fuel.

The long-term alpha radiation associated with spent nuclear fuel in a geological repository is substantial, because of long-lived radionuclides such as 237Np, 235U, 238U, and 239Pu and their daughter products. Numerous studies have documented the importance of products of alpha radiolysis of water in enhancing spent-fuel oxidation and dissolution (2126). Recent studies also indicate that the formation of studtite or metastudtite as an alteration product of spent nuclear fuel in a geological repository is likely (7, 9).

To investigate whether studtite can be stable relative to common uranyl oxide hydrate alteration phases, we measured the ΔHds of dehydrated schoepite (12), (UO3)(H2O)0.8, an alteration phase found on spent nuclear fuel in laboratory studies (4). We calculated the enthalpy of the reaction (Table 1): Embedded Image Embedded Image Embedded Image Embedded Image(5) This indicates that studtite is the stable phase relative to dehydrated schoepite, H2O, and H2O2 in their standard states. ΔS° for this reaction is expected to be small given that no gas is evolved or consumed. Our calculations indicate that studtite can form in low peroxide concentrations (1.1 × 10–14 M H2O2, calculated for reaction 5). Whereas dehydrated schoepite is an alteration product that is often found in the paragenetic sequence involving uranyl oxide hydrates and uranyl silicates, studtite is thermodynamically the dominant phase where peroxide occurs.

After a few hundred years, the alpha activity of spent fuel declines only slightly for 10,000 years, and by a factor of about 10 by 100,000 years (25). It remains much higher than the radioactivity of natural uranium for millions of years. Therefore, studtite or metastudtite is likely to persist at the surface of spent nuclear fuel in contact with water in a nuclear waste repository. Uranyl peroxides may dominate at the expense of more common uranyl minerals, such as uranyl oxide hydrates and uranyl silicates, which have also been found as alteration phases on spent fuel in laboratory studies. Uranyl peroxides must be considered in assessing the impact of uranyl minerals on the release of radionuclides such as 237Np from nuclear waste in a repository.

Supporting Online Material

Materials and Methods

Tables S1 and S2

References and Notes

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article