Dissolution Behavior of Uranium Aerosols in Simulated Lung Fluids: Effects of Particle Size

As uranium is a health hazard for nuclear fuel workers, correct dose assessments are necessary to ensure worker safety. Absorption parameters of uranium aerosols are important parameters for the dose assessments. In this study, size-fractionated uranium aerosols of uranium hexafluoride (UF6), ammonium uranyl carbonate (AUC), uranium dioxide (UO2), and uranium octoxide (U3O8) collected from four workshops at a nuclear fuel fabrication plant were used in two separate in vitro dissolution models, one with simulated lung-lining fluid (Gamble's fluid) and one with simulated phagolysosomal fluid (PSF). The dissolution characteristics in the different simulated lung environments were assessed. Lung absorption parameters: the rapid fraction (fr), rapid rate (sr, d-1), and slow rate (ss, d-1) were derived, and the dosimetric impact was evaluated. The conversion workshop has a known presence of uranium aerosols of both soluble (UF6, AUC) and relatively insoluble (UO2) compounds. Absorption parameters derived for dissolution in Gamble's fluid in comparison to dissolution in PSF showed a mixed pattern with impactor stages. Absorption parameters fr, sr, and ss were generally slightly lower in PSF. However, a higher slow rate in comparison to Gamble's fluid was derived in PSF for some impactor cut-points, indicative of the different solubilities of the mixed aerosols. The fr,sr, and ss derived in PSF ranged between 0.1-0.3, 4-16 d-1, and 0.004-0.015 d-1, respectively. At the powder preparations and pelletizing workshops, less soluble uranium aerosols UO2 and U3O8, are present. The absorption parameters derived for the pelletizing workshop from PSF were fr: 0.02-0.025, sr:4-6 d-1, and ss: 0.0007-0.001 d-1. For the powder preparations workshops, the fr ranged between 0.01-0.045, sr ranged between 4-32 d-1, and ss ranged between 0.002-0.004 d-1. The slow rates derived in PSF were several times higher than for Gamble's fluid, which resulted in ss: 0.00009-0.0004 d-1 and ss: 0.0001-0.001 d-1 for powder preparations and pelletizing workshops, respectively. The slow rate derived from dissolution in PSF showed good agreement with available in vivo data from studies performed at the pelletizing workshop. For the BA pelletizing workshop with UO2 and U3O8 aerosols that are blended with Gd2O3, fr, sr, and ss ranged between, 0.05-0.1, 1-7 d-1 and 0.002-0.005 d-1, respectively. No substantial differences in derived parameters for dissolution in PSF or Gamble's fluid were visible. The difference between dissolution patterns in the two fluids between aerosols from the BA pelletizing and from the other two workshops are believed to be related to differences in the specific surface area (SSA), differences in the mixture of uranium compounds, and the blending with Gd2O3. No significant correlation between absorption parameters and impactor stages was found in any of the solutions for the four workshops, with the exception of ss for powder preparation workshop in PSF (R2=0.76, p=0.03). The general conclusions, when evaluating absorption parameters derived for each workshop in PSF and in Gamble's fluid, were that the parameter most impacted by the type of lung fluid was the slow rate ss and that the effect was most significant for more insoluble uranium. The findings in this study demonstrate that in vitro dissolution in simulated phagolysomal fluid give more realistic estimates of absorption parameters for relatively insoluble material, such as UO2 and U3O8, that are more in line with in vivo data than in vitro dissolution in simulated lung lining fluid.