Uranium and plutonium analysis of nuclear material samples by multi-collector thermal ionisation mass spectrometry: Quality control, measurement uncertainty, and metrological traceability

Metrological studies and recent improvements in multi-collector thermal ionisation mass spectrometry (MC-TIMS) of uranium and plutonium in bulk nuclear material samples are presented with a focus on nuclear safeguards. Using total evaporation and modified total evaporation methods, experimental data are presented for isotope ratio measurements in routine mode spanning a range of almost ten orders of magnitude, with n(U-236)/n(U-238) measurements as low as a few parts per billion. Based upon these data, measurement reproducibility, associated measurement uncertainties with comparison to International Target Values (ITV), an upper limit of the instrumental uranium memory effect and of the hydride formation, and process and instrumental blank levels are examined. A comparison of measurement performance between the latest and previous generation of instruments for the total evaporation method is conducted. In addition, the implementation of a quality control procedure including control charts is presented and, in this context, commercially available U or Pu certified reference materials for isotope ratio and isotope dilution mass spectrometry are surveyed. The implementation of the Guide to the expression of Uncertainty in Measurement (GUM) is discussed for the modified total evaporation method. To address the importance of metrological traceability of measurement results to the SI units, the agreement between the certified values of two independently produced series of uranium certified reference materials (CRMs) - NBL and IRMM U series - is probed. Using IRMM CRMs as calibration standards, a new set of values for selected NBL U series CRMs is presented with expanded uncertainties of about 0.035% for major isotope ratios and at an order of magnitude of 0.1-0.5% for minor ratios. (C) 2011 Elsevier B.V. All rights reserved.