Increasing the ionization yield for the detection of U-236 and U-233 by AMS

Full text: The limiting factor for actinide measurement by Accelerator Mass Spectrometry (AMS) is the total detection efficiency (TDE, sample to particle detector), which is strongly dependent on the ionization yield (IY) for anions in the ion source. Normally, uranium is extracted as UO^-, which results in an IY of about 10^-3. Fluorine-rich molecules with exceptionally high electron affinity produce strong anion currents by Cs-sputtering in the presence of excess fluorine reactants like PbF₂. Previous research suggested UF₅^- extraction from metal U mixed with PbF₂ powder [1]. At the Vienna Environmental Research Accelerator (VERA) we investigated whether and how much PbF admixture could enable sufficient U-fluoride anions extraction from a Fe₂O₃ matrix with embedded traces of U-oxides. For this, in-house U standard Vienna-KkU [2] solution was co-precipitated with Fe-hydroxide and calcined to UO in Fe₂O₃ (U:Fe = 1:30, weight ratio). Vienna-KkU-D30 is representative for environmental samples [3]. Three materials based on Vienna-KkU - each mixed with PbF₂ in variable ratios - were tested: I. Dry Vienna-KkU-D30; II. Fe solution (~200 μg Fe per sample, U:Fe = 1:30, weight ratio) added to Vienna-KkU solution and dried up, followed by calcination; III. Co-precipitation of UF₄ (Vienna-KkU solution) with NdF₃ (U:Nd = 1:18, weight ratio) [4]. UO F formation from each mixture and its IY was analyzed using VERA's first mass separator. The TDE of U, isotope ratios ²³⁶U/²³⁸U, ²³³U/²³⁸U and the fingerprint ²³³U/²³⁶U [5] of all mixtures were determined using the entire AMS set-up at VERA. UF^5- was the most efficiently extracted UO F ion for all the mixtures but the absolute IY for UF^5- was limited by other UO F ions in UO based materials. The optimal mixing ratio of Vienna-KkU-D30 powder with PbF₂ (I.) was 1:9 by weight ratio. It showed five times higher TDE for U than optimal UO^- extraction. Individually dried and calcined material (II.) with PbF₂ showed a TDE more than twice as much as by UO^- extraction. The UF₄ based mixture (III.) showed a TDE of up to ten-fold compared to UO^-, however, individual sub-milligram NdF₃ preparation was not yet possible. Methods I. and III. were checked for reproducibility by comparing the isotope ratios of all the mixtures to the Vienna-KkU consensus value [2]. Materials I and II (both U-oxides) showed ²³⁶U/²³⁸U within 1-σ of Vienna-KkU. Samples prepared as UF₄ (III.) showed surprisingly high ²³⁶U/²³⁸U, which might be explained by reagent contamination. To validate method II. an aliquot of an air filter sample analyzed previously by UO^- was used. ²³⁶U/²³⁸U ratios were confirmed using UF₅^- extraction. Besides higher IY, UF₅^- extraction also seems advantageous for isobaric background suppression (²³²ThH³⁺, ²³⁵UH³⁺). In conclusion, the extraction of U as UF₅^- combined with sub-milligram Fe preparation (high PbF₂ mixing ratios) compared to UO^- extraction allows shorter measurement times and/or lower statistical uncertainty or using smaller sample amounts.