Authigenic clay formation and secular variability in the global Li cycle

Lithium isotopic (δ7Li) and elemental concentrations of pore fluids and carbonates from IODP Site U1338 Hole A (eastern equatorial Pacific Ocean) suggest that clay authigenesis (i.e., in situ precipitation) is a significant sink for Li in carbonate-rich sedimentary sections. Systematic variations in pore fluid δ7Li with depth in the section suggest that clay authigenesis can (i) strongly decrease pore fluid Li concentrations with depth and (ii) fractionate Li isotopically to a considerable degree (Δ ~ 5-21‰ relative to seawater). We hypothesize that clay authigenesis in carbonate-rich sections occurs due to the presence of reactive biogenic silica, and reactive trans- port modeling supports the contention that the pore fluid δ7Li depth profile at Site U1338 is best explained by faster authigenesis at depth. The significance of clay authigenesis in carbonate-rich sediments is two-fold: if global in scale, (i) it can generate sizeable output fluxes in the global Li cycle, and (ii) the evolution of the sedimentary system over time can markedly impact the isotopic composition of the global Li output flux. We compile ODP and IODP pore fluid Li data from 267 sites; of these, 207 have Li pore fluid concentration gradients in the upper 50-100 meters that indicate the sites as diffusive sinks of Li. We then estimate that clay authigenesis in carbonate-rich sediments could reasonably generate a Li output flux on the order of ~1.2·10^10 moles/year, which is comparable to the gross input fluxes in the modern Li cycle. A series of reactive transport simulations illustrate how clay authigenesis might impact the isotopic composition of the output flux of Li from the global ocean. The suggestion is that applying a constant fractionation factor from the global ocean over time is likely incorrect, and that secular changes in the δ7Li of the output flux will be driven by rates of authigenesis, burial rates, and the depth extent of authigenesis in the sedimentary section. Utilizing a time-dependent, depositional diagenetic model, the δ7Li values of bulk carbonate are shown to be a consequence not of recrystallization alone, but recrystallization in the presence of clay authigenesis. Further, our model results are used to illustrate how carbonate d7Li may be used to constrain the temporal evolution of clay authigenesis in the sedimentary section. Ultimately, this work suggests that the Li isotopic composition of bulk carbonates can be altered diagenetically. However, such alteration is not a detriment, but provides useful information on those diagenetic processes in the sedimentary column that impact the global Li cycle. Thus, Li isotopes in bulk carbonates have the potential to elucidate diagenetic controls on the global Li cycle over long time scales