EXPLORATORY STUDIES OF 13C DIFFUSION INTO OLIVINE AND QUARTZ

An unresolved aspect of the terrestrial carbon budget concerns the capacity of nominally C-free minerals of the silicate Earth to incorporate C in their crystal lattices. Credible past studies of natural materials and experimental run products have reached conclusions on opposite sides of this issue (e.g., Freund et al. GCA 44, p. 1319; Keppler et al. Nature 424, p. 414). To provide additional insight, we have implemented new techniques aimed at characterizing diffusive uptake of ¹³C in common silicate minerals, initially olivine and quartz. Some of our experiments were conducted at P ~ 1 atm by sealing individual specimens of either San Carlos olivine or natural quartz (~1×3×3 mm, polished with colloidal silica) in evacuated SiO₂ glass ampoules with amorphous ¹³C. The ampoules were suspended in 1-atm furnaces at 500-850°C for periods of days to weeks. Additional experiments were run on olivine at 1 GPa in a piston-cylinder device by sealing the specimens in Ag containers with ¹³C glycine (¹³C₂H₅NO₂), which decomposes at run conditions to produce ¹³C-O-H-N fluid + ¹³C graphite. The quenched samples were depth-profiled for ¹³C using the ¹³C(p,γ)¹⁴N nuclear reaction. Experiments using both the amorphous ¹³C and glycine sources produced well-behaved diffusive-uptake profiles in San Carlos olivine, with surface concentrations exceeding 1000 ppm C and penetration depths of several hundred nanometers. The implied diffusivities are ~1E-20 and 8E-20 m²/s at 700° and 850°C, respectively. Quartz samples cut normal to c yield similar though shorter diffusive-uptake profiles yielding diffusivities of ~6E-22 and 4E-21 m²/s at 500° and 600°C, respectively. In the case of quartz, the implied Arrhenius relation is the same, within uncertainty, as that determined by Sharp et al. (EPSL 107, p. 339) for diffusion of oxygen into quartz from a CO₂ source, suggesting that C and O diffuse together in quartz as either CO₂ or CO. Early indications from our experiments are that 1000+ ppm of C can be incorporated into silicate minerals as diverse as quartz and olivine at modest P-T conditions—although as yet we have no information on C speciation or siting in the mineral structure. The apparent diffusivities are too low for effective grain-scale transport of C in a laboratory time frame, but geological equilibration is plausible for some scenarios.