FIELD-SCALE DETERMINATION OF URANIUM ISOTOPE FRACTIONATION INDUCED BY U(VI) ADSORPTION AND DESORPTION: THE SUPER 8 EXPERIMENT AT DOE'S RIFLE, COLORADO SITE

²³⁸U/²³⁵U shows promise as an indicator of U(VI) reduction in groundwater contaminated by natural U (e.g., from mining and milling sites). Like other stable isotope ratios which indicate redox reactions, ²³⁸U and ²³⁵U (radioactively stable on decadal time scales) can be fractionated during U(VI) reduction. Previous work on both field-scale bioremediation and ancient redox fronts preserved as ore deposits show ~1 per mil shift to lower ²³⁸U/²³⁵U in the water as U(IV) is removed as a solid. The sense of shift is opposite to that observed with many lighter elements, presumably because nuclear field shift effects dominate over mass-dependent effects. Identifying U(VI) reduction is critically important, both as a means of tracking bioremediation of uranium contamination in groundwater and as a possible signature of U redox cycling during Earth history.

Recent laboratory work, however, indicates small but significant ²³⁸U/²³⁵U shifts induced by adsorption of U(VI) onto manganese oxides, introducing a possible confounding effect for using ²³⁸U/²³⁵U variations to infer or quantify U(VI) reduction. The present study examines ²³⁸U/²³⁵U variation during a field experiment at the Rifle, CO, IFRC site in which bicarbonate was injected to de-sorb U(VI) from aquifer solids. Samples from the CU-03 well, 1 m downgradient from the bicarbonate injection gallery, showed an initial concentration of 150 ng/mL U(VI), followed by a desorption-induced increase to 300 ng/mL for ten days, then a resorption-induced decrease to 70 ng/mL over two weeks, and a return to 150 ng/mL over two weeks.

Thirty-three samples, taken over 67 days and covering all four phases of the system response, were analyzed. No significant shift in ²³⁸U/²³⁵U was observed; the standard deviation of all the ²³⁸U/²³⁵U measurements was 0.06 per mil, and no significant trends appeared in the data. This is encouraging, as it indicates that the ²³⁸U/²³⁵U variations observed during U(VI) reduction induced by acetate injection solely reflect the reduction process. This simplifies interpretation of previously observed reduction induced shifts. It is not clear why sorption-related ²³⁸U/²³⁵U shifts reported in recent laboratory experiments did not appear; experiments with sediments from this site are planned to address that issue.