hide
Free keywords:
Sponge spicules, Silicon isotopes, Cambrian, Chert, Diagenesis
Abstract:
The early Phanerozoic has long been thought to have witnessed a decline in the ocean dissolved silicon inventory, but direct, quantitative investigations of dissolved silicon concentrations are missing. Silicon isotope fractionation by modern siliceous sponges is known to be related to ambient seawater silicon concentrations, opening a potential window onto reconstructions of silicon concentrations in geological time. However, whether the ratios of silicon stable isotopes (expressed as δ30Si) in spicule silica can be preserved during the diagenetic transformations from amorphous silica to (re)crystallized quartz is unclear.
Here, we use in situ analyses of δ30Si by femtosecond laser ablation multi-collector ICP mass spectrometry on sponge spicules and their matrix in Ediacaran-Cambrian cherts from the Yanziqiao section, South China. These cherts formed from seawater with little hydrothermal influence and span the proposed silicon inventory decline. All samples are pure cherts (SiO2 > 95%) dominated by microquartz. Under transmitted light, white spicules are easily distinguished from the darker matrix and usually exhibit monoaxial or multiaxial shape. There is only subtle difference between the δ30Si of matrix and sponge spicules (the difference between their average δ30Si values is within 0.4 ‰), implying alteration and homogenization of δ30Si during diagenesis. In situ δ30Si and Al/Si ratios in the matrix sometimes exhibit a positive relationship, which we explain by the influence of Al on the kinetics of silica diagenesis. Higher Al content can contribute to earlier transformation of silica phases under shallower burial and lower temperature, associated with a larger equilibrium fractionation factor and higher δ30Si. Overall, the original δ30Si at micrometer scales is most likely to be destroyed during diagenesis, and the reconstruction of dissolved silicon concentrations from in situ analyses of recrystallized spicules thus rendered unjustifiable. However, a reaction-advection–diffusion model shows that δ30Si of bulk chert can still record relatively pristine information of initial sediment δ30Si. Moreover, the range of in situ δ30Si in cherts provide hints to distinguish if they are dominated by equilibrium or kinetic effects, which is important if we wish to reconstruct seawater δ30Si from bulk chert δ30Si.
