LONG-TERM TECHNETIUM INTERACTIONS WITH REDUCING CEMENTITIOUS MATERIALS
Technetium is among the key risk drivers at the Saltstone Facility. The way that it is immobilized in this cementitious waste form is by converting its highly mobile Tc(VII) form to a much less mobile Tc(IV) form through reduction by the cement's blast furnace slag. This report includes a review of published data and experimental results dealing with Tc leaching from Portland cement waste forms. The objectives for the literature study were to document previous reports of Tc interactions with slag-containing cementitious materials. The objectives for the laboratory study were to measure Tc-saltstone Kd values under reducing conditions. From the literature it was concluded: (1) Spectroscopic evidence showed that when Tc(IV) in a slag-cement was exposed to an oxidizing environment, it will convert to the more mobile Tc(VII) species within a short time frame, 2.5 years. (2) SRS saltstone will reduce Tc(VII) in the absence of NaS or sodium dithionite in a reducing atmosphere. (3) Only trace concentrations of atmospheric oxygen (30 to 60 ppm O{sub 2}; Eh 120 mV) at the high pH levels of cementitious systems is required to maintain Tc as Tc(VII). (4) Experimental conditions must be responsible for wide variability of measured K{sub d} values, such that they are either very low, {approx}1 mL/g, or they are very high {approx}1000 mL/g, suggesting that Tc(VII) or Tc(IV) dominate the systems. Much of this variability appears to be the result of experimental conditions, especially direct controls of oxygen contact with the sample. (5) A field study conducted at SRS in the 1980s indicated that a slag-saltstone immobilized Tc for 2.5 years. Below background concentrations of Tc leached out of the slag-containing saltstone, whereas Tc leached out of the slag-free saltstone at the rate of nitrate loss. One possible explanation for the immobilization of Tc in this study was that the slag-saltstone maintained reducing conditions within the core of the 55-gallon sample, whereas in the small-scale lab experiments, where samples were crushed to <1mm, oxygen diffused through the particles and reoxidize the slag during the contact period. (6) Present site specific reduction capacity value of 820 {micro}eq/g is in the realm of literature values that were either measured or theoretically estimated based on thermodynamic calculations. (7) Almond and Kaplan (2011) measured desorption K{sub d} values from a Vault 4 saltstone core sample. Desorption leaching tests were conducted in a glovebag maintained at 30 to 60 ppm O2. A ground olive-colored saltstone sample, as compared to black monolith sample, was used in this study, indicating the sample had been exposed to O2, which is likely the cause for the lower then anticipated Kd value measured, 139 mL/g. Tc adsorption experiments were conducted under reducing conditions (<0.5 ppm O{sub 2}(g) -585 mV, 2% H{sub 2}, pH 11.66) and obtained K{sub d} values of {approx}1000 mL/g in a saltstone formulated with 45% slag (nominal concentration) and a K{sub d} of 10,000 mL/g when the saltstone contained 95% slag. The K{sub d} values logarithmically increased from 1 day to 56 days, with little sorption generally occurring in the first eight days. Steady state had not been achieved during the initial 56 days. However, the slag-free cement control samples also had K{sub d} values near 1000 mL/g and extremely low redox conditions, due to the 2% H{sub 2} atmosphere. A key concept that this literature review and the experimental results provide is that Tc immobilization is dependent on experimental conditions, specifically, the available oxygen that can oxidize technetium in a portland cement or saltstone-like monolith. The shrinking core model used in the saltstone performance assessment describes the existence of an oxidized outer layer of concrete surrounding a shrinking core of reducing intact saltstone. A sharp boundary between the two zones moves slowly inward, resulting in oxidation of Tc(IV). This work largely reinforced our conceptual model of the shrinking core model, but more importantly provided clarity regarding process kinetics, mechanisms, and input values for future detailed modeling. The shrinking core model in itself is devoid of geochemical mechanisms. This study provided more geochemical detail to this conceptual numerical model.