The chemical behavior of transuranium elements in natural aquifer systems is governed by a variety of geochemical reactions, such as dissolution reaction (solubility), hydrolysis, complexation with inorganics or organics, redox reaction, colloid formation, geochemical interaction with surfaces of various minerals, coprecipitation, mineralization, etc. This paper reviews the present state of knowledge on some of these particular reactions. The emphasis here is on how the individual reactions can be appraised for long-term prediction of the geochemical behavior of transuranium elements in the natural environment. of the various possible reactions, the primary thermodynamic processes are discussed with notable examples: dissolution of transuranium compounds in aquatic solution; complexation with important anions present in groundwater; and colloid generation. Various laser spectroscopic methods in use for chemical speciation are mentioned briefly for their spectroscopic capability, as well as for their applicability. The present paper attempts to better understand the migration behavior of transuranium elements in natural aquifer systems.

At low temperatures (<200°C), spent fuel from power reactors oxidizes from its UO2 lattice to a U4O9 lattice but with an oxygen-to-uranium (O/U) ratio of ∼2.4. Also, the weight gain time response has a plateau as the O/U approaches 2.4. Part of this response results from a geometrical dependency as a U4O9 oxidation front propagates into grain volumes of UO2. It may also be indicative of a metastable, non-stoichiometric U4O9 phase whose existence may inhibit the transition kinetics to the next expected phase of U3O8. To gain a mechanistic understanding and to plan future oxidation tests, lattice kinematic and thermodynamic models are developed for lattice deformations and energetics of lattice phase changes (UO2 → U4O9 → U3O7 → U3O8) that include zeroth order influences on oxidation kinetics due to interstitial oxygen atoms and vacancies plus interstitial and substitutional actinides and fission decay products in spent fuel.

The corrosion of SIMFUEL, simulated high-burnup CANDU (CANadian Deuterium Uranium) fuel, was investigated in 0.1 mol·dm−3 NaClO4 solution (pH ∼ 9.5) as a function of dissolved oxygen concentration using electrochemical techniques and X-ray photoelectron spectroscopy (XPS). Electrodes were constructed of SIMFUEL pellets with compositions close to those of a natural UO2 fuel that has undergone burnup of 3 at% and 6 at%. The XPS analysis of freshly polished SIMFUEL pellets showed that the extent of uranium oxidation in SIMFUEL was equivalent to that in stoichiometric UO2. The SIMFUEL electrodes showed higher conductivity and electrochemical reactivity than the pure UO2 electrodes. A comparison of the open circuit corrosion potentials of the SIMFUEL electrodes with that of a pure UO2 electrode in oxygenated solutions showed a much faster increase in the potential for the SIMFUEL electrodes at short times. This suggests that the initial stages of UO2 oxidation, i.e., UO2 → UO2+x → UO2.33, are facilitated by fission product impurities in the UO2 matrix. However, the “steady-state” oxidation of uranium in SIMFUEL by the dissolved O2 was similar to that observed in pure UO2 under similar conditions.

Two methods were used to measure grain-boundary inventories of 137Cs, 90Sr and 99Tc in used CANDU fuel, to corroborate source term estimates based on a fission gas release code. Used fuels were partially oxidized at 200°C in air to overall compositions of UO2+x (0.15≤ × ≤0.25) to expose UO2 grain boundaries, followed by leaching in aqueous solution. Only a fraction (2 to 18%) of the calculated gap + grain-boundary inventories for 37Cs was released. This suggests that the calculations overestimate Cs release or that oxidation does not expose all grain boundaries, or that Cs release from grain boundaries is slow. Release of 90Sr (0.01 to 0.7%) agreed reasonably well with the source term estimates (0.001 to 0.3%). Release of 99Tc (0.3 to 1.5%) suggests that the source term estimate for the upper boundary of 99Tc release (25%) may be too high. A second technique involved leaching of crushed and size-fractionated used fuel in either a static or dynamic system. A direct one-to-one correlation between calculated and measured gap + grain-boundary inventories for 137Cs was found for low- and medium-power fuels.

Dissolution rates for air-oxidized spent fuel were measured in flowthrough tests where U concentrations were kept well below the solubility limit. Results from two types of specimens, separated grains and coarse particles, both in oxidized (U4O9+x) and unoxidized (UO2) conditions indicated only minor effects of oxidation on the surface-area-normalized rates. Similar results were obtained for unirradiated specimens in three different oxidation states (UO2, U3O7, and U3O8). These observations have important practical implications for disposal of spent fuel in a geologic repository as well as implications regarding the oxidative dissolution mechanism of UO2 fuel.

Photothermal Deflection Spectroscopy (PDS) has been successfully applied to the study of uranium oxide electrochemistry. A brief description of PDS and preliminary results that demonstrate the technique are presented. Concentration gradients formed at the electrode surface are measured by this technique. The gradients give insight into the reaction mechanisms. There is some evidence of the initiation of non-electrochemical dissolution of the uranium oxide. Optical absorption by the uranium oxide is measured by PDS and the first results indicate that the absorption of the surface does not change during electrochemical experiments. This result is contrary to literature measurements of bulk samples that indicate that the optical absorption should be strongly changing.

The release of uranium from three samples of UO2(S) with different particle sizes (100–300 μm, 900–1100 μm, pellet) has been studied as a function of time. In all cases, the same pH, ionic medium, temperature, and oxygen partial pressure were used.

Two distinctive trends were observed in the three cases, with a relatively fast initial uranium release followed, after 10–15 days, by a slower dissolution rate. The experiments were continued during 200 days; no change in the second dissolution rate was noticed.

The uranium released as a function of time has been successfully fitted with a mathematical expression which combines an oxidation-dissolution mechanism.

The dissolution of unirradiated UO2(s), with a particle size of 1 mm, has been studied in MgCl2 brines at 298 K under both reducing and oxidizing conditions. Results obtained under reducing conditions (H atmosphere in the presence of a palladium catalyst) show an initial increase of the total uranium concentration in solution and a subsequent decrease until equilibrium (or steady state) values are reached. Results obtained under oxidizing conditions (nominal oxygen partial pressures of 0.05, 0.21 and 1 atm) show two different trends. A relatively fast initial dissolution rate and, after approximately two or three weeks, a slower dissolution rate. X-Ray Photoelectron Spectroscopy (XPS) has shown that the UO2 surface composition changes during the experiment.

The storage of high-level spent reactor fuel in a proposed national geologic repository will require the construction of containers to be placed in boreholes drilled into the host rock. Federal regulations require that the fuel be maintained subcritical under normal or accident conditions. This is determined through the calculation of a neutron multiplication factor, keff, that must remain below 0.95. Criticality will play an important role in the container design, the internal configuration of the fuel, and the selection of neutron poisons. An analysis of keff should be a normal step in the conceptualization of new waste container designs. Unlike thermal effects in a proposed repository, criticality will remain a problem long after the 10,000 year lifetime of the facility.

When immobilizing into borosilicate glass the radionuclides in the caustic high-level radioactive wastes stored in the USA, the soluble fission product Cs-137 has to be removed from supernates of the wastes. In the current processes zeolites or an organic precipitant will be used to remove the Cs. These are then treated further and mixed with the radioactive sludges and vitrified into a borosilicate glass. This paper describes the vitrification of a mixture resulting from using a new process to remove Cs from the caustic supernate. A resorcinol based organic ion exchange resin is used. This resin was then mixed with sludge and frit and vitrified. Using an organic ion exchange resin rather than zeolite or the organic precipitant has certain advantages. For example, use of the zeolite increases the amount of glass to be made and use of the organic precipitant produces benzene as a secondary waste stream. Results in the paper indicate that a mixture of the resin, sludge and frit can be successfully vitrified in a joule-heated, slurry fed melter. However, when resin is present in the feed, the glass becomes less durable due to the increased amount of Fe(II) caused by reduction of Fe(III) in the melt. Based on the durabilities of other waste glasses, this glass is still suitable as a canistered wasteform.

At a processing temperature of 1150°C for model West Valley glass compositions, the prescribed range of oxygen fugacities needed to achieve an [Fe2+]/[Fe3+] of 0.1 to 0.5 is 10−4 to 10−7 atm. Establishment of the Fe2+-Fe0 equilibrium, resulting in metal precipitation from the melt, occurs at oxygen fugacities lower than 10−11 atm at this temperature. The target processing range as defined by the iron redox ratio is equally valid at both lower and higher temperatures (±100°C). Elevations of the concentrations of redox-active components to I wt% Cr2O3, 1 wt% NiO, 1 wt% CuO, 1 wt% CeO2, and 4 wt% Mn2O3 in the waste glass will not affect the redox limits as established by the iron redox ratio of 0.1 to 0.5; these limits provide sufficiently large margins of safety to assure no stabilization of reduced or oxidized forms of these elements.

The pyrometallurgical processing of spent fuel from the Integral Fast Reactor (IFR) results in a waste of LiCI-KCI-NaCI salt containing approximately 10 wt% fission products, primarily CsCI and SrCI2. For disposal, this waste must be immobilized in a form that it is leach resistant. A salt-occluded zeolite has been identified as a potential waste form for the salt. Its leach resistance properties were investigated using powdered samples. The results were that strontium was not released and cesium had a low release, 0.056 g/m2 for the 56 day leach test. The initial release (within 7 days) of alkali metal cations was rapid and subsequent releases were much smaller. The releases of aluminum and silicon were 0.036 and 0.028 g/m2, respectively, and were constant. Neither alkali metal cation hydrolysis nor exchange between cations in the leachate and those in the zeolite was significant. Only sodium release followed t0.5 kinetics. Selected dissolution of the occluded salt was the primary release process. These results confirm that salt-occluded zeolite has promise as the waste form for IFR pyroprocess salt.

The calcination process used for sodium nitrate could be the most important part of the liquid-waste calcination process used in high-level-waste (HLW) disposal. A French process that involves conversion of the solution of nitrate fission products into a solid mixture of oxides and undecomposed nitrates. These products are then mixed with a glass-making mixture (called a frit) in an induction melting pot.

In 1984, Utake et.al., (1) suggested the use of the Sol-Gel technique to decrease the calcination temperature of NaNO3.

In this work, we used the thermal-analysis technique to study the catalytic effect of some glass network formers like SiO2, B2O3 and of some network modifiers like A12O3, Fe2O3, MgO, TiO2 ZnO. The best effect was achieved by adding 20% diatomaceous earth (Kieselguhr) and 20% B2O3 In that case, the calcination temperature onset decreased from 780°C to 565°C.

The advantages of this process are as follows:

1. The proposed additives (Kieselguhr and B2O3) are glass network formers that have to be added anyway in the next piocess step.

2. In the proposed process, no sodium is added, and the only process change suggested is that the glass formers (or part of them) should added at the calcination step, rather than at the glass-melting step.

3. The proposed modification is easily achieved and is expected to improve the homogenization of the final product.

Quantitative powder X-ray diffraction was used to analyse the phase assemblages of titanate nuclear waste ceramics doped with 244Cm and accumulating a-decay doses up to 1.24 × 1018 alpha-decay-events g−1. Among the Cm-bearing phases, perovskite is more susceptible to damage than zirconolite. Phases devoid of curium (hollandite and freudenbergite) experience a flux of ionizing radiation, and may also undergo radiation-induced structural degradation. The analysis demonstrated that at a dose of 1.24 × 1018 alpha-decay-events g−1 approximately 20 wt% of the waste phases remain aperiodic after annealing for 2 hours at 800°C.

Fundamental data are provided for the potential use of hydrous TiIV-si oxide (TS) and hydrous TiIV-Zr oxide (TZ) as the solidification media of radioactive Cs+ and Sr2+. The sorption capacity for Cs+ and Sr2+, at various conditions, and the structure and the composition of the exchangers in these ionic forms, at high temperature, were investigated, followed by measuring the leaching rate from the calcined exchangers into water at room temperature. TS uptakes more Cs+ and Sr2+ than TZ does. The leaching rate from calcined TS is much lower than that from calcined TZ, due to the formation of titanium silicates of Cs+ or Sr2+. Based on these results, TS is considered to be superior to TZ, in sorption capacity, and the fixation capability of Cs+ and Sr2+, on the exchanger matrix.

The complexation and thermal behavior of pyruvic acid oxime (PAO) with representative cations associated with nuclear reprocessing waste has been investigated. This ligand has the dual advantage that it precipitates most of the cations from solution and these complexes decompose at ca. 150°C to CO2, acetonitrile, and, initially, the metal hydroxide. Introduction of the appropriate elements required for Synroc phases to a simulated waste stream followed by NaPAO and disodium fumarate leads to precipitation of a metalorganic Synroc precursor in which the components have been intimately mixed at the molecular level. This approach circumvents incorporation of alkali metals in the final ceramic.

Improved AEM techniques were used to investigate three Synrocs containing 10 wt/ simulated HLW and a fourth sample with ∼18 wt/ simulated HLW. One of the 10 wt./ loaded Synrocs also contained an addition of 1.0 wt/ Na2O and another contained an addition of 2.0 wt/ Fe2O3. This work is part of a larger study initiated with the objective of determining if the bulk composition of Synroc affects the partitioning of elements between individual phases. Results from the four samples in this study show that, as expected, elemental partitioning is mainly controlled by the ionic radius criterion, with smaller Y, Gd, and U ions having a preference for zirconolite and the larger Ce and Nd ions favouring perovskite. Additions of Na and Fe lead to the formation of CAT and loveringite at the expense of rutile or Magneli phases, but only have minor effects on partitioning coefficients. Partitioning coefficients, DZ/P, for REE, Y, and U in the four Synrocs are the same (within experimental error).

Static leach tests were performed in both 304L stainless steel and Teflon vessels using a synthetic high-level waste glass with either deionized water (DIW) or a tuff groundwater solution as the leachant to assess the effects of the vessel and the initial leachant composition on the extent and nature of the glass reaction. The tests were performed using monolith samples at 340 m−1 and crushed samples at 2000 m−1 for times up to 1 year. The results show less silicon is released from the glass into the groundwater solution than into DIW at both high and low glass surface area/leachant volume ratios (SA/V), but the alkali metal and boron releases are not affected by the leachant used. Tests performed in a stainless steel vessel resulted in slightly lower leachate pH values, but similar reaction rates to those performed in a Teflon vessel, as measured by the boron release. Blank tests with DIW or EJ-13 in the vessels showed the Teflon vessels to release small amounts of fluoride (1 to 2 ppm) and to acidify the DIW slightly (4.0 < pH < 5.6). The pH values of blank tests with EJ-13 increased from 8.2 to about 8.6 in steel and to about 9.2 in Teflon vessels. The slightly higher pH values attained in Teflon vessels are attributed to outgassing of CO2 during the test.

The leaching behaviour of Pu, Am, Np and Tc from the high-level waste glasses of interest to Belgium has been studied. Corrosion tests were carried out on inactive glasses (the Cogéma R7T7 glass SON68 and the DWK Pamela glasses SM513 and SM527), doped with the radionuclides. Nuclide leaching was studied in contact with a slurry of Boom clay, diluted with claywater, at 90°C and 40°C, for durations of up to 4 years. The retention of Pu and Np in the glass surface layers was considerable for SON68 and SM513, but small for SM527. The retention of these elements decreased at lower temperature. Tc was not retained. The largest part of the leached activity is adsorbed onto the clay in an immobile form. Nuclide concentration in mobile form (smaller than 105 MWU) was found to be independent on duration for Pu and Am. Increasing concentrations were found for Np and Tc. The mobile concentrations were smaller in the presence of corrosion products.

The steady-state concentration of Pu was measured upon leaching Pu doped glass in synthetic interstitial claywater (SIC) and humic acid free interstitial solution (IS). The solubility of PuO2·xH2O(am) was determined in the same chemical condition. The results then were compared with the Pu solubility calculated by using a published solubility product and complex stability constants. The humic acid (HA) influence on the Pu leaching from the glass and the colloid formation were investigated.