BACKGROUND Uranium-series nuclides are one of the three major radioactive decay systems, which are suitable for studying various geological processes at different time scales. In addition, ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd isotopes (Sr-Nd isotopes for short) are two commonly used isotopes, for rock dating, chemical weathering assessment and tracing sediment sources. In this case, the combination of Sr-Nd-U isotopes can provide more comprehensively knowledge of the element cycle on the earth's surface and deepen our understanding of the sediment "Source to Sink"processes.Most previous studies involving the Sr-Nd-U isotopes have been established by separating Sr-Nd and U isotopes respectively. In this way, the digestion operation must be performed twice, one for separating Sr-Nd and the other for separating U. Alternatively, if only one sample is digested for measuring all three isotopes, between each element separation, the residue must be dried and dissolved in another solution so as to start a new column work. The former increases the amount of samples, which is not conducive to analysis of precious and trace samples; the latter adds additional drying operation, which is time consuming and increases the risk of sample loss and contamination.

OBJECTIVES  To establish a new Sr-Nd-U combined separation scheme. In this method, only one sample is dissolved, avoiding solution transfer between each separation, so as to reduce sample amount and improve the efficiency of separation and purification of Sr-Nd-U isotopes.

METHODS  A new chromatographic scheme of separating Sr-Nd-U with one sample digestion using a tandem column scheme is presented. Three columns were overlain sequentially to separate Sr in Sr Spec column, Nd in AG50W-X8 column and U in UTEVA column. 3mol/L HNO₃ was used to pre-condition, load the sample, and rinse the matrix. After rinsing the matrix, the tandem column was separated to 3 independent columns to elute the target elements (Sr, Nd, U) respectively.As the connection sequence of different resin columns may interfere with the recovery of target elements, two different chromatographic schemes were compared. In Scheme 1, U column was placed on top of Sr column, while in Scheme 2 the positions of U column and Sr resin column were exchanged. In both schemes, the AG50W-X8 resin column was set at the bottom as the cationic resin can adsorb the most complex elements.All separated elution was tested for element concentration using inductively coupled plasma-mass spectrometry (ICP-MS). The basalt standard sample (BCR-2) was used to examine the behavior and recovery of each element in the separation procedure.

RESULTS  A total of 10 fractions were recovered from the tandem column scheme, among which fraction 1 represented the leachate recovered by loading samples and rinsing matrix from the tandem three columns. Fraction 2, 5 and 8 represented the leachate recovered by Sr Spec resin, AG50W-X8 resin and UTEVA resin respectively rinsing matrix after separation of the three columns. Fraction 3, 6 and 9 represented the leachate recovered by Sr, REE and U columns, which was Sr, Nd and U collection. Fraction 4, 7 and 10 represented the leachate from each resin recycle stage.In either Scheme 1 or Scheme 2, most matrix elements (high content of K, Ca, Na, Mg, Al, Fe, Ti and P and low content of Rb, Hf and Th) were mainly concentrated in fraction 1. The elution rate of Na, Ti, Rb and Hf was up to 99%. The elution rate of K and Ca was slightly lower at about 85% and the elution rate of Fe was about 56%. Sr was mainly concentrated in fraction 3, which contained only a small amount of P and Ba. Nd was mainly concentrated in fraction 6, which also contained both Sm and Ce. U was mainly concentrated in fraction 9, which only contained a very small amount of P and Pb. The column recovery was almost 99.9% for U, 90% for Sr and over 80% for Nd.The removal rate of major matrix elements (K, Ca, Na, Ba, Fe, Rb, etc.) exceeded 99%, which reduced interference with high-precision isotope analysis of Sr, Nd, and U. The recovery and purity of Sr, Nd, U were all quite high. A very small amount of P and Ba in fraction 3 had no interference with Sr isotopes (⁸⁷Sr/⁸⁶Sr). The Rb which was isomorphism of Sr was removed completely. With regard to Sm and Ce in fraction 6, previous studies had shown that ¹⁴²Ce could not interfere with Nd isotope (¹⁴³Nd/¹⁴⁴Nd), and Sm could be further separated by Ln resin, so as not to affect Nd isotopic test. Fraction 9 contained nearly 100% U with no other elements.The sequence of resin column splicing is a crucial consideration which may impact the element separation. Hence, the position of Sr Spec column and UTEVA column was exchanged to compare the influence of different column sequences on eluting target elements. Both Scheme 1 and Scheme 2 can effectively wash off most of the matrix elements, and the target elements Sr, Nd and U can be efficiently adsorbed on the resin. There is no significant difference on target element separation between the two different column sequences. This indicates that Sr Spec and UTEVA resins do not interfere with each other on the target elements.

CONCLUSIONS  The new chromatographic scheme of separating Sr-Nd-U with one sample digestion using a tandem column scheme can be used to quickly and efficiently separate Sr, Nd and U elements from silicate rock samples. The recovery rate for U, Sr and Nd is 99.9%, 92.5% and 82.1%, respectively, which meet the requirements of subsequent isotope analysis. This Sr-Nd-U combined separation method can be used to reduce the sample consumption by about 50%, which is beneficial to the analysis of precious and trace samples. Meanwhile, as no solution transfer is needed between each column separation, this method can also save time for column work and increase the efficiency of chemical separation. A new idea for Sr-Nd-U multi-isotope separation is provided. If the recovery of Pb in the fraction 4 of this chromatographic scheme can be improved in further studies, the application of this new method may be expanded to more fields in the future.