A method has been developed to determine high-precision stable Nd isotopic compositions in geological samples by multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS) using sample-standard-bracketing (SSB) mode. Nd was pre-concentrated and purified through a two-column ion-exchange chromatographic procedure. The recovery for Nd is generally better than 96%, and the interferences, such as Ce and Sm can be efficiently removed, allowing negligible influence on the stable Nd isotopic composition during mass spectrometry measurement. The internal precision for the stable Nd isotopic compositions, ε¹⁴²Nd, ε¹⁴⁵Nd, ε¹⁴⁶Nd and ε¹⁴⁸Nd were generally better than ±0.2 (2 SEM), and the external precision was generally better than ±0.2 (1 SD) for ε¹⁴²Nd, ε¹⁴⁵Nd and ε¹⁴⁶Nd, and better than ±0.5 (1 SD) for ε¹⁴⁸Nd based on the long-term results of the Nd standard solutions, such as La Jolla, Nd-GIG and NIST 3135a. Both the internal and external precisions are comparable to those by double-spike method. The averages of our measured ε¹⁴²Nd, ε¹⁴⁵Nd, ε¹⁴⁶Nd and ε¹⁴⁸Nd results of La Jolla are identical to those by the double-spike method within analytical error. Thus, our method can provide comparable results with the double-spike method, but does not bring any inconvenience of calibrating double spikes, and the efficiency is greatly improved. This provides a more convenient way for studying stable Nd isotopes in geological processes. By using this method, the stable Nd isotopic compositions for a series of international rock standard references were measured.

The combined study of Ce and Nd stable isotopes is promising because Ce and Nd behave similarly in geological processes and distinctively in redox reactions. However, the sequential purification of Ce and Nd from other REEs is particularly challenging because of their similar physicochemical properties. In this contribution, we present a novel two-stage ion-exchange chromatography method to purify Ce and then Nd using Bio-Rad AG50W-X12 (200–400 mesh) and Eichrom TODGA resin (50–100 μm). The yields of Ce and Nd are both typically more than 99.3% after measurements of natural materials. The potential influence of the HCl content and rock type were thoroughly evaluated. Stable Ce and Nd isotopic ratios were measured by MC-ICPMS and calibrated with Sm and Eu internal standards, respectively. The long-term external reproducibility was better than ±0.04‰ (2SD) for δ^142/140Ce and ±0.03‰ (2SD) for δ^146/144Nd. The stable Ce and Nd isotopic ratios of ten commercially accessible geological rocks were measured here. The δ^142/140Ce values of four geological rocks were presented for the first time. Other results are in good accordance with reported values within analytical uncertainties. The similarity in the degree of Ce and Nd isotopic fractionation may be induced by diagenetic processes, whereas the difference may be accused by redox reactions. In addition, our method using a single sample aliquot can help to minimize the consumptions and heterogeneities of the samples. This procedure for a two-stage isolation of Ce and Nd and high-precision analysis of their isotopic ratios has important advantages in the study of Ce and Nd isotope geochemistry.

Removing interferences such as Ce on Nd isotopes during column chemistry is hard to achieve due to their similar properties. In this study, we established a simple but efficient Nd separation technique with the perfect separation of Nd from Ce and nearly 100% Nd yield using the TODGA resin (50–100 μm, Eichrom, USA) after actual measurements. All samples used here through the Nd separation technique have Ce/Nd ratios less than 0.003. The whole procedure takes about 10 h. The stable Nd isotopes were measured by MC-ICP-MS and calibrated with standard-sample-standard bracketing (SSB) combined Eu internal normalization. The potential influences of Eu and Ce on stable Nd isotopic analysis were strictly evaluated. The long-term analyses of the JNdi-1 standard produced a total average δ^142/144Nd of 0.000 ± 0.029‰ (2SD, n = 210), δ^145/144Nd of 0.000 ± 0.029‰ (2SD, n = 210), and δ^146/144Nd of 0.000 ± 0.027‰ (2SD, n = 210). Two pure Nd standards, La Jolla and Nd-GIG, produced δ^146/144Nd values (means ± 2SD, n = 20) of −0.227 ± 0.030‰ and 0.226 ± 0.021‰, respectively. δ^146/144Nd (means ± 2SD, n = 20) of seven geological reference materials (BCR-1, BHVO-2, AGV-2, GSP-2, NOD-P-1, NOD-A-1, and GBW07249) were also reported. The reproducibility of δ^146/144Nd was better than 0.030‰. These results are in good agreement with the published DS-TIMS methods. In addition, our method also allows the concurrent determination of the radiogenic ¹⁴³Nd/¹⁴⁴Nd ratio. All the ¹⁴³Nd/¹⁴⁴Nd ratios of these samples are identical with the preferred values published in the GeoReM Database. These results indicated that the method of Nd isotopic analysis established here can simultaneously measure high-precision δ^142/144Nd, δ^145/144Nd, δ^146/144Nd, and radiogenic ¹⁴³Nd/¹⁴⁴Nd ratio for geological materials.

A chromatographic procedure for Rb, K, Ba and Sr one-by-one separation from geological materials has been developed by using a single column with packing Sr-spec resin, followed by high-precision Rb isotope measurement by multi-collector inductively coupled plasma mass spectrometry (MC-ICP-MS). The main matrix elements (Ti, Al, Mg, Fe, Mn, Na and Ca) were removed with 4.4 mL of 3 M HNO₃ first; Rb and K were then sequentially eluted in different volumes with the same acid. After that, Ba was eluted with 8 M HNO₃, and finally Sr was eluted with Milli-Q water. This procedure enables us to collect pure Rb, K, Ba and Sr one by one in a single column with recovery close to 100%. The purified Rb was measured by using a standard-sample bracketing method via MC-ICP-MS. The short-term precision for δ⁸⁷Rb was better than ±0.03‰ (2SD, for 15 hours), and the long-term (more than 12 months) external precision was better than ±0.06‰ (2SD) based on repeated analysis of Rb standard solution NIST SRM 984. Matrix effects on Rb isotope measurement by MC-ICP-MS were tested by doping pure Rb standard with various amounts of matrix elements and were found to be insignificant when Na/Rb < 2, Ca/Rb < 4 and K/Rb < 25. We measured a set of geological reference standards by this method and found significant Rb isotopic variation. Our study suggests that Rb isotope geochemistry may be a promising tracer of various cosmochemical and geological processes.