Effect of organic matter removal on U-series signal in clay minerals

Abstract

Clay minerals hold some key information on time and rate of weathering processes that could be deciphered by U-series disequilibria. However a removal of organic matter is generally required prior to clay mineral chemical analyses, especially in soils. The impact of two protocols of organic matter removal on the U-series disequilibria in three different clay minerals was therefore investigated. Natural or commercial kaolinite, illite and montmorillonite were treated with hydrogen peroxide and by calcination. Both investigated treatments impact the U-series signature of clay minerals. Yet the extent of this impact strongly depends on the clay-type: while kaolinite remains almost insensitive to the treatments, illite shows some equivocal shifts since the activity ratios measured after calcination could describe more accurately the pure clay signal than initial activity ratios, and montmorillonite displays the larger reactivity. Overall, the H2O2 treatment seems to induce more disturbance than the calcination process, mostly on U isotopes in illite and montmorillonite. Due to the oxidizing power of H2O2, the redox state of U held in the clays, i.e.: U(IV) or U(VI), strongly influences the extent of U leaching. In addition, a preferential leaching of 234U relative to 238U was observed in montmorillonite sample only, resulting to a U isotopic ratio fractionation. Conversely, during calcination, the most pronounced effect arises for Th isotopes. The congruent feature of Th loss, i.e.: Th loss≈mass loss, affecting the three clay minerals suggests a loss of Th through volatilization during this treatment. U or Ra are also affected, but to a lesser extent, by this volatilization. Overall, calcination causes a two- or three-fold reduced change of the (230Th/234U) and (234U/238U) activity ratios relative to H2O2 treatment. The (226Ra/230Th) ratio was clearly the less sensitive to any treatment. It was only slightly affected in kaolinite after calcination, and the observed change suggests an improvement of the measured value due to a removal of a Th-bearing external matter. Albeit not perfect, calcination should therefore be preferred to H2O2 treatment for organic matter removal before U-series analyses.