Automatic Purification of Li Isotopes from Geological Samples by High-Pressure Ion Chromatography

Significance: In the global carbon cycle, chemical weathering of silicic rock is considered the main sedimentation mode of atmospheric carbon dioxide, and plays an important role in controlling global climate change and topography evolution^[1-3]. In the past decade, lithium isotopes have emerged as one of the most powerful and reliable indicators in weathering studies of silicic rocks^[10-11]. This is due to the significant relative mass difference of up to 16.7% between ⁶Li and ⁷Li, which makes lithium susceptible to substantial isotopic fractionation in geological processes. However, the relatively large mass increases the difficulty of analyzing Li isotope ratios by MC-ICP-MS because Li is prone to significant fractionation during the separation and purification process. This requires the Li recovery rate to be nearly 100% and to avoid interference from Na, Ca, Mg, and other matrix elements. Several scholars have proposed various traditional manual Li purification methods and have optimized them by cross-combining different types of eluent, resin types, resin particle sizes, column tube sizes, and resin volumes, but the process remains cumbersome and requires 1 to 2 days to complete. Due to single-step separation, shorter separation times, and the online quantification of each element, high-pressure ion chromatography (HP-IC) has become a new trend in isotope purification, applied in the purification of Sr^[39], Ca^[40], Li^[24], K^[41], Mg^[42], S^[43], F^[44], Cl^[44], Br^[43] and Rb^[45]. Unfortunately, the reported HP-IC separation studies generally confirm the purity and recovery rate of target isotopes based on the final accurate testing results from MC-ICP-MS. They also lacked systematic research on the elution process of HP-IC, which limits the application of this separation method. Here, several sets of conditional experiments were designed to explore the acid tolerance, sample-loading capacity, and matrix effect of the cationic chromatographic column configured in HP-IC. We systematically explored HP-IC, optimized the drenching conditions for the CS16 cation column, and successfully established an effective HP-IC separation method.

Methods: A high-pressure ion chromatograph (Dionex ICS-6000, Thermo Scientific) features an automatic sampler (AS-AP), an inorganic cation column (IonPac CS16, 5mm×250mm), a guard column (IonPac CG16, 5mm×50mm), a quadrupole pump (SP), an auxiliary pump (AXP), a regenerative suppressor (CERS 500, 4mm), a CD conductivity detector, and a fraction collector (ASX-280). The HP-IC is well-suited for separating and collecting alkali and alkaline earth metals. The temperature of the inorganic cation column can be set to 25℃, 40℃, or 60℃. The eluent was switched from the recommended methanesulfonic acid to high-purity nitric acid, and the sample loop volume was set at 1.7mL.

　　For sample preparation, 50mg of rock powders were digested by mixing 3mL of HF and 1mL of HNO₃ in Teflon vessels on a hotplate at 140℃, replenishing the dried residue with 3mL of aqua regia until the solutions became clear. The preliminary experiment demonstrated that geological samples dissolved by the standard method could not produce ideal peaks using HP-IC, which may be caused by three reasons: first, the acidity of the geological sample solution is too high; second, the sampling amount is too large; third, the matrix of the geological sample solution is too complex. To solve these problems, three sets of experiments were designed to identify the factors affecting the purification process of IC, including the acid tolerance test, sample-loading capacity test, matrix effect test, and so on. The eluting conditions were optimized to purify Li in the geological samples using HP-IC. The standard solutions used include GSB04-1767—2004 national multi-trace element solution and BWT20009-1000-W-50 K-Na-Ca-Mg solution. The natural geological standard samples used include GBW07103 (Granite), GBW07104 (Andesite), GBW07105 (Basalt), GBW07101 (Ultrabasic rock), GBW07107 (Shale), GBW07333 (Marine sediment of the Yellow Sea), GBW07159 (Rare earth ore), and GBW07180 (Bauxite). The recommended values of relevant elements in these standards are shown in Table 1.

Data and results: (1) Exploring acid tolerance. By fixing the eluting conditions (the eluent concentration: 40mmol/L; eluent flow rate: 1mL/min; column temperature: 60℃; injection volume: 500μL) and increasing the acidity of the sample solution (12.5mmol/L, 25mmol/L, 50mmol/L and 100mmol/L HNO₃), the maximum acidity of sample solution that can be tolerated by IC was determined by the leaching curves (<50mmol/L, Fig.1).

　　(2) Exploring sample-loading capacity. By fixing the eluting conditions (the eluent concentration: 40mmol/L; eluent flow rate: 1mL/min; column temperature: 60℃) and increasing injection volume of a certain sample solution (250μL, 500μL, 1000μL), the maximum sample amount that can be tolerated by HP-IC was determined by the leaching curves (>1.5mL 500ng/g multi-trace element standard solution+2.5μg/g K-Na-Ca-Mg standard solution, Fig.2).

　　(3) Exploring matrix effect. By adding 50μg/g K-Na-Ca-Mg, 500ng/g Fe, 500ng/g Al to 250ng/g multi-trace element standard solution in order and contrasting the peak position and peak height of alkali metal elements in these four solutions on leaching curves, the influence of matrix elements on Li separation was investigated (it had no effect on the peak position and height of Li, Fig.3a). In addition, the leaching curves of national geological standard samples GBW07103 (Granite), GBW07104 (andesite), GBW07105 (basalt) and GBW07101 (ultrabasic rock) with very different compositions were compared to further confirm that matrix effect is not obvious on the separation of Li isotopes (Fig.3b).

　　(4) Exploring optimal eluting condition. By fixing the type of sample solution (GSR-5, shale) and leaching conditions (the eluent flow rate: 1mL/min; column temperature: 60℃; injection volume: 200μL) and decreasing the eluent concentration (40mmol/L, 35mmol/L, 30mmol/L, 25mmol/L), the degree of separation for Li isotopes was compared to obtain the best eluent concentration (30mmol/L, Fig.4).

　　(5) Evaluating the purity and recovery rate of Li. Under optimal eluting conditions, the mixed standard solution was purified. Fractions were continuously collected and tested by ICP-MS. Moreover, 4 national geological standard samples were dissolved by the optimized method and eluted by HP-IC. The target Li fraction and the fractions before and behind the Li fraction were all collected and measured by ICP-MS and ICP-OES. By contrasting the amount of Li element injected to the HP-IC with the amount of Li element collected in the Li fractions, the recovery rates of Li element were calculated to be above 99.3% (Table 3), with no detectable matrix elements (e.g., Na, K, Mg, Ca) present, thus meeting the testing requirements of MC-ICP-MS.

　　In summary, the CS16 cationic chromatographic column can bear a volume of sample solution up to 1.5mL, and its matrix effect is not obvious, but its acidity tolerance for sample solution is low (less than 50mmol/L). By steaming normally dissolved sample solutions to a wet-salt stage and adding a small amount of ultra-pure water repeatedly, the acidity of the sample solution can be reduced from 320mmol/L (2% HNO₃) to ~30mmol/L, which meets the sampling-requirements of HP-IC, successfully extending the application of IC to insoluble geological samples with a complex matrix. By optimizing eluting conditions, Li fractions separated by HP-IC are pure and have a high recovery rate, which lays a foundation for the wide application of IC in the field of isotope purification.