Methodology of Lithium Analytical Chemistry and Isotopic Measurements

Lithiumwas discovered as an elemental species in 1817 by J.A.Arfvedson and isolated as ametal one year later by Sir H. Davy. The existence of two naturally occurring isotopes of lithium, Li and Li, was unambiguously proven by A.J. Dempster (1921) and the atomic weight of Li and the individual isotope masses were first determined with reasonable accuracy and precision by F.W. Aston (1932) (few available data are referenced therein) at 6.928 ± 0.008, 6.012, and 7.012 atomic mass units, respectively (atomic mass unit, amu, is defined as 1.660538921 ± 73 × 10 kg and corresponds to 1/12 ofmass of unbound neutral C in nuclear and electronic ground state; modern physics uses the term “unified atomic mass” abbreviated as ‘u’ or dalton). In the decades that followed this, however, only a few attempts to refine the absolute atomic weight as well as its uncertainty (see summary in Svec and Anderson 1965) were implemented. The actual atomic weights of Li and Li are 6.0151223 ± 5 and 7.0160040 ± 5 u, respectively. The actual atomic weight of Li, currently accepted by IUPAC, is 6.941 ± 0.002 (Wieser 2006) and it may well be that this level of precision will not improve significantly in the near future due to natural variations in Li/Li exceeding*80‰ (Tomascak 2004, this volume). The natural variations of Li/Li were not measured extensively or with clearly established reproducibility until late 1980s (Chan 1987). This article reported a method for chemical separation and precise isotopic measurements with *2.5 ‰ external uncertainty. Although earlier measurements which used mass spectrometry were performed on a range of terrestrial and extraterrestrial samples (e.g., Balsiger et al. 1968; Brown et al. 1977; Eugster and Bernas 1971; Krankowsky and Müller 1967; Michiels and De Bièvre 1983), data reported in these studies were usually expressed in absolute Li/Li ratios with errors exceeding several per mil. This severely hampered data comparison and the overall utility of Li isotopes in geochemistry. A major problem of early reports on Li isotopic compositions was the lack of a widely distributed reference material that would allow for reliable inter-laboratory comparison of measured isotopic ratios. In 1973, Flesch and co-workers isolated Li from “virgin” spodumene-rich ores from the granitic pegmatite at Kings Mountain, North Carolina, USA, and determined its isotopic composition at Li/Li = 12.0192 ± 0.0002 (Flesch et al. 1973). The isolated Li was made available by the National Bureau of Standards (now NIST) as L-SVEC (since reassigned as SRM 8545). The availability of an external standard permitted calibration of measured Li/Li and presentation of data relative to L-SVEC in permil deviations. Chan (1987) and several papers that followed employed the normalized Li/Li (δLi). Since that time, though, the use of the δLi value has prevailed, as it is consistent with the other major