High-Temperature Processes: Is it Time for Lithium Isotopes? Microphotograph of high- pressure fluid vein (right side) consisting of garnet Horst R. Marschall1,2 and Ming Tang3,4 (red) and quartz (colorless) crosscutting an eclog- itic host rock from New 1811-5209/20/0016-0247$2.50 DOI: 10.2138/gselements.16.4.247 Caledonia (field of view 5 cm) (Taetz et al. 2018). he field of high-temperature Li isotope geochemistry has been rattled by samples did display Li isotopic major paradigm changes. The idea that Li isotopes could be used to trace excursions, supporting a model in which subducted materials are Tthe sources of fluids, rocks, and magmas had to be largely abandoned, heterogeneous in Li concentration because Li diffusion causes its isotopes to fractionate at metamorphic and (noted here as [Li]); specifically, it magmatic temperatures. However, diffusive fractionation of Li isotopes can be was suggested that Li isotopes, and the average bulk subducted slab, used to determine timescales of geologic processes using arrested diffusion introduces isotopically heavy Li profiles. High diffusivity and strong kinetic isotope fractionation favors Li (i.e., high 7Li/6Li) into the mantle isotopes as a tool to constrain the durations of fast processes in the crust and to create high-δ7Li domains (Chan mantle, where other geochronometers fall short. Time may be the parameter et al. 2002a, see “Lithium Facts Box” for the definition ofδ 7Li). that high-temperature Li isotope studies will be able to shed much light on. The situation became more KEYWORDS: lithium isotopes, diffusion, geochronometry, timescales, complex when the first high- isotope fractionation pressure metamorphic rocks (i.e., eclogites and garnet mica schists) INTRODUCTION were investigated for their Li isotopic composition with Elements at the low-mass end of the periodic table show the aim of constraining the composition of subducted, large equilibrium isotope fractionation, especially at the dehydrated slabs. These rocks displayed a large spread in low temperatures prevailing at the Earth’s surface where δ7Li with some very negative values that were interpreted rocks interact with the atmosphere and hydrosphere. to represent strong equilibrium Li isotope fractionation For these reasons, the lightest of all metals, lithium (see between minerals and fluids during slab dehydration “Lithium Facts Box”), would seem to be most suitable in (Zack et al. 2003). It was concluded that subducting slabs tracing the recycling of surface materials into the deep would be driven to very low δ7Li values during metamor- mantle. Early work on Li isotopes, pioneered by Lui-Heung phic dehydration and, thus, introduce isotopically light Li Chan, focused on the characterization of the marine into the mantle (Zack et al. 2003). This model was rapidly environment, including seawater, sediments, and fresh and widely accepted and boosted the interest of geochem- and altered oceanic crust (e.g., Chan and Edmond 1988). ists in applying Li isotopes to a range of high-temperature Efforts then moved to identify the Li isotopic composition processes in the geosciences (Elliott et al. 2004). of the depleted upper mantle, as defined by mid-ocean- ridge basalts and mantle samples, and the differences between the depleted mantle and subduction-related LITHIUM FACTS BOX and ocean-island volcanoes (Chan et al. 1992, 2002b; Tomascak et al. 2002). Disappointingly, the vast majority Element: Lithium, Li (alkali metal) of samples from these settings revealed Li isotopic composi- • Atomic number: 3 −1 tions that are indistinguishable from the depleted upper • Molar mass: 6.94 g mol mantle, despite the unmistakable contribution of recycled Two stable isotopes (natural abundances): materials to their magma sources, as evidenced by various • 6Li (7.4% –7.9%) other trace elements and isotope systems (e.g., Chan et • 7Li (92.1%–92.6%) al. 2002b; Tomascak et al. 2002). Nonetheless, some rock All radioisotopes of Li have half-lives <1 s The delta notation for stable isotopes: 1 Institut für Geowissenschaften & FIERCE 76 7 Li Lisample Goethe Universität δ Li = – 1 76Li Li Altenhöferallee 1 standard 60438 Frankfurt am Main, Germany This value is multiplied by 1,000 to express it in per mil (‰) E-mail: [email protected] The standard for Li isotopes is NIST-RM8545, also called 7 6 2 Department of Geology & Geophysics “L-SVEC”, a Li2CO3 with Li/ Li = 12.019 Woods Hole Oceanographic Institution 7 Woods Hole, MA 02543, USA δ Li values (in parts per mil) in the geologic framework: • Continental crust: +1.7‰ ± 1.0‰ 3 School of Earth and Space Sciences • Modern seawater: +31.2‰ ± 0.3‰ Peking University, Beijing 100871, China E-mail: [email protected] • Mantle (mid-ocean-ridge basalts): +3.5‰ ± 1.0‰ • Highest bulk sample (continental brine): +46.9‰ 4 Department of Earth, Environmental and Planetary Sciences Rice University • Lowest bulk sample (eclogite): −21.9‰ 6100 Main Street • Natural variation among rocks and fluids: ~70‰ Houston, TX 77005, USA ELEMENTS, VOL. 16, PP. 247–252 247 AUGUST 2020 Two natural examples of very strong kinetic Li FIGURE 1 isotope fractionation on different length scales. A Xenolith Opx 2 (A) A millimeter-scale traverse through a mineral grain (orthopy- (San Carlos) roxene) in a mantle xenolith from San Carlos (Arizona, USA). This revealed a total range of 37‰ in δ7Li within less than a millimeter, 1 mm which is comparable to the total range of all bulk rock, sediment, and soil values from Earth and the Moon reported to date (rectangle at left). The mineral analysis spot map (orthopyroxene 1) Orthopyroxene 1 shows the locations of the measurement spots. Opx = orthopy- Clinopyroxene roxene. AFTER JEFFCOATE ET AL. (2007). (B) Diffusion processes on the meter-scale of an outcrop produced a range of 27.5‰ in δ7Li (circles) within a distance of 10 m from the Tin Mountain pegmatite dike (South Dakota, USA) into its surrounding country rocks. [Li] = Li concentration (brown curve). Starred symbol represents a distant sample of country rock. δ7Li = see Lithium Facts Box. AFTER TENG ET AL. (2006). SIMS measurement spots +20 +10 rock of a pegmatite dike within a distance of approximately Earth 10 m—a range in δ7Li that nearly encapsulates all conti- mantle 3 0 nental crustal and eclogite bulk rocks analyzed to date Li range of Earth 7 (FIG. 1). In another example, Jeffcoate et al. (2007) found δ -10 an even larger range of 37‰ in the space of less than δ7Li Li (‰) total 7 2 and Moon rocks soils a millimeter within a single mineral grain in a mantle -20 δ xenolith (FIG. 1). -30 The lithium isotope systematics of subducting slabs and their [Li] (µg/g) 1 [Li] likely effect on mantle heterogeneity was revisited in the -40 light of this understanding. It was concluded that dehydra- Orthopyroxene 1 tion does not change the Li isotopic composition of the slab 0 -50 to a large degree (Marschall et al. 2007). The subducting 0 0.5 1.0 1.5 2.0 2.5 slab after (near) complete dehydration would, thus, carry a mix of isotopically heavy Li in eclogites and isotopi- Distance (mm) cally lighter Li in metasediments and would on average B Pegmatite dyke be very similar in δ7Li to the ambient mantle (Marschall 7 (Tin Mountain) et al. 2007). Very low δ Li values observed in eclogites, 5 m accompanied in many cases by high Li abundances, have been reinterpreted as the result of diffusive influx of Li from surrounding rocks; in this case, they are the conse- quence of local diffusive redistribution of Li on the scale δ7Li +10 of minerals, hand-specimens, and outcrops and are not country rock representative of the bulk chemical flux between crust and (amphibolite & mantle (Marschall et al. 2007). 500 mica schist) 0 [Li] The realization that Li isotopes could, therefore, no Li (‰) longer be employed unambiguously as tracers of recycled 400 7 δ surface materials in the deep mantle disappointed many a crustal rocks pegmatite dyke 300 Li range of continental -10 7 geochemist. Any sample of igneous or metamorphic rock δ could be suspected of having had diffusive Li exchange [Li] (µg/g) 200 total with other solids or fluids it might have encountered in its 100 -20 environment. Therefore, a sample’s Li isotopic composition 0 can only be interpreted in terms of larger-scale geodynamic processes after all possible diffusive effects have been 0 5 10 300 positively excluded—a condition that is rarely fulfilled. Distance (m) However, the notion that this novel geochemical tracer IS IT TIME FOR LITHIUM? could be used to study deep-mantle recycling was dealt Prof. Jerry Wasserburg, the eminent geochemist, cosmo- a blow when the world of geochemistry became aware chemist, and winner of the Crafoord Prize, once claimed, of the fact that Li diffuses very rapidly at temperatures when referring to novel geochemical tools, that “sometimes that are typical of magmatic and high-grade metamorphic science advances because of a new shovel and the urge to processes and that Li isotopes are kinetically fractionated dig a hole somewhere” (Wasserburg 2000). It seems that, in during diffusion (e.g., Richter et al. 2003; Teng et al. 2006; the field of high-temperature geochemistry, we have been Parkinson et al. 2007). This mechanism of isotope fraction- using Li isotopes to dig entirely in the wrong place. Our Li ation had been well established in other disciplines (e.g., shovel was blunt and unsuited to tracing recycled materials Arnikar 1959), but it took a number of detailed studies in the mantle. But, if we use lithium’s unique combina- of rocks, minerals, and high-temperature experiments tion of fast diffusion and kinetic isotope fractionation we to convince the geochemical community that kinetic could develop a powerful geochronometer to use in a range fractionation of Li isotopes was a mechanism that really of geologic settings.