Calcium Isotope Analytical Method And Its Geological Applications

As one of the main petrogenetic elements and the important nutrient elements in organisms, calcium is widely distributed in all spheres of Earth systems. There is a 4‰ variation in the Ca isotopic compositions(expressed as δ^44/40Ca) of various Ca reservoirs on Earth, suggesting Ca isotopes as a promising tracer of many geological processes. However, because of the big abundance differentiation of Ca isotopes, large isotopic fractionation during analysis and matrix effect, the accuracy and precision of Ca isotope data are not really good enough. In addition, most of the studies related to Ca isotopes are focused on low-temperature processes. There is also lack of studies that are focused on Ca isotopic behaviors during high-temperature geological processes.In this thesis, we systematically review the previous studies of Ca isotopes, including its analytical method and applications. To solve the main problems existed at present, we provide a method to measure Ca isotopic compositions for natural rock samples by thermal ionization mass spertrometry（TIMS） through a ⁴²Ca-⁴³Ca double-spike technique. Although the error on the fractionation factor α is relatively big, the average mass of the two isotopes in ⁴²Ca-⁴³Ca double-spike is only 0.5 mass unit different from target ⁴⁴Ca/⁴⁰Ca ratio, and ⁴²Ca and ⁴³Ca can be easily collected simultaneously on TIMS without significant ion optic effect, which enhances the data quality collected. A chemical procedure of separating Ca isotopes from geological samples, adopting AG MP-50（100-200 mesh） cation resin as ion exchange media and 1.6 N HCl as rinsing acid, is carefully calibrated through detailed evaluations on a series of factors, such as resin types and volumes, column types, acid types and concentrations, the samples loading amount and the lithology, etc. Compared with previous procedures, our procedure could obtain nearly complete recovery and good separation of Ca from matrix elements at the same time. Before TIMS measurement, 5-10 μg purified Ca is loaded onto each out-gassed 99.995% Ta filament with 1 μl 10% H₃PO₄ as an activator. The instrument is operated in positive ionization mode, and the cup configuration contains two sequences. To monitor isobaric interference of ⁴⁰ K on ⁴⁰Ca, ⁴¹ K is measured, and an online correction is carried out using a value for the ⁴⁰K/⁴¹ K ratio of 1.7384×10-3. Instrumental fractionation is corrected by a ⁴²Ca-⁴³Ca double-spike technique using an offline iterative routine with the exponential law. The Ca isotopic compositions of NIST SRM 915 a and IAPSO seawater based on long-term analysis are 0.01±0.01（2SE, n=233） and 1.83±0.01（2SE, n=85）, respectively. To further assess the accuracy and precision of this method, Ca isotopic compositions of 19 reference materials, including silicate, carbonate and manganese nodules, are measured. Long-term external reproducibility for δ^44/40Ca is better than ±0.1‰（2SD）. Such precision allows us to investigate Ca isotopic fractionation in high-temperature geological processes.Due to the large relative mass differences of Ca isotopes（Δm/m≈20%）, significant Ca isotopic fractionation can occur during analysis. To make proper correction, Ca isotopic fractionation during column chemistry and determination by TIMS is well investigated in this study. The δ^44/40Ca of ISPSO seawater in different Ca cuts on column chemistry ranged from 4‰ to 0‰. The more Ca is eluted, the lower δ^44/40Ca of the elution is found. The isotopic fractionation of Ca on the column appears to follow the exponential law. However, TIMS instrumental fractionation during Ca runs does not always follow the exponential law due to mixing effects from sample reservoirs on the filament. Our results show that errors could be caused if the instrumental fractionation deviates from the exponential law, especially when the fractionation degree is large. To improve the measurement accuracy and precision, a model is proposed to check the behavior and degree of instrumental fractionation, which will provide a quick and reasonable verdict on the data quality of TIMS runs.As one of the major elements in carbonates, Ca isotopic compositions of igneous rocks derived from the mantle can be applied to trace recycled carbonates. But we need to make sure whether there is Ca isotopic fractionation during partial melting. Mid-ocean ridge basalts（MORB） are formed directly from the partial melts of the normal upper mantle and are less affected by crustal contamination during eruption. This study report Ca isotopic compositions of N-MORB and D-MORB samples from the East Pacific Ocean to investigate the behaviors of Ca isotopes during partial melting. δ^44/40Ca of these samples varies from 0.75 to 0.88. Compared to the δ^44/40Ca in the upper mantle of 1.05±0.04, the Ca isotopic compositions of studied MORB samples are 0.15-0.30‰ lighter. As Ca O content in olivine is quite low, and δ^44/40Ca in plagioclase is equal to or lower than that of the whole rock, the low δ^44/40Ca in studied MORB samples cannot be attributed to fractional crystallization of olivine or plagioclase. Since δ^44/40Ca in studied MORB samples is negatively correlated with La/Yb and Nb/Y, indicating that up to 0.3‰ Ca isotopic fractionation could occur during partial melting. Thus, partial melting is probably one of the essential high-temperature processes that cuase the heterogeneous Ca isotopic compositions in Earth’s mantle, and Ca isotope may shed light on the accretion history of the Earth.In contrast to the Pacific and Atlantic MORB samples, the Indian MORB samples have enriched Sr, Nd and Pb isotopic compositions. To evaluate the influence of recycled marine carbonates to the enriched Indian MORB mantle, Ca-Mg-Sr-Nd isotopic studies are conducted on the Indian MORB samples. Compared to the upper mantle, the MORB glasses from the Central Indian Ridge（CIR MORB） have lower δ^44/40Ca（0.70-0.89） and ¹⁴³Nd/¹⁴⁴Nd（0.5129-0.5131）, and higher ⁸⁷Sr/⁸⁶Sr（0.7029-0.7040）. The δ^44/40Ca variation of CIR MORB samples is not correlated with La/Yb or Nb/Y. But the δ^44/40Ca variation of CIR MORB samples is negatively correlated with ⁸⁷Sr/⁸⁶ Sr, and positively correlated with ¹⁴³Nd/¹⁴⁴ Nd, which suggest the involvement of recycled marine carbonates in the mantle source. However, CIR MORB samples show the mantle like δ²⁶Mg values, indicating that recycled marine carbonates are mainly composed of calcium carbonate. In the diagrams of Ca-Sr and Ca-Nd isotopes, Indian-type MORB samples from the Australian-Antarctia Discordance fall on the trends of CIR MORB samples, while Pacific-type MORB samples are similar to N-MORB and D-MORB samples from the Pacific Ocean. Therefore, recycled marine carbonates might exist in the mantle source of Indian-type MORB samples. The 0.25‰ Ca isotopic fractionation between Pacific-type MORB and the upper mantle might be caused by partial melting. Partial melting and recycling of marine carbonates in the mantle could both cause the heterogeneous Ca isotopes in Earth’s mantle, and could be effectively recognized from each other by Ca-Mg-Sr-Nd isotopic studies.