A CRITICAL REVIEW OF THE BORON ISOTOPE-PH PROXY

Published applications of the boron-pH technique are founded on a theoretical model of carbonate d¹¹B variation with pH that assumes the boron isotopic compositions of carbonates mirror the boron isotopic composition of the tetrahedrally-coordinated boron species in solution. The well constrained reconstruction of pH based on this model requires a precise understanding of the fractionation factor for the equilibrium isotope exchange between boric acid (trigonally-coordinated) and borate (tetrahedrally-coordinated) in solution (a_4-3), the equilibrium constant for the dissociation of boric acid (K^*_B), and the seawater d¹¹B value through time (d¹¹B_sw).  A review of the available data suggests the assumption that carbonate d¹¹B is identical to the d¹¹B of borate is not certain and that both the equilibrium fractionation factor and the history of d¹¹B_sw are poorly constrained. These conclusions reflect on the accuracy and precision of reconstructed pH values over long-time scales.

Evaluation of carbonate d¹¹B from inorganic precipitation experiments, assuming carbonate d¹¹B is isotopically identical to borate in solution, indicate that the most appropriate value of a_4-3 is ~0.974 rather than the theoretical value of 0.981 from Kakihana et al. [1977] as has been assumed in some publications.  Further, a compilation of inorganic and culture experiments suggest that either the value of a_4-3 is not constant with pH or the assumption that carbonate d¹¹B systematically reflects the isotopic composition of borate is faulty.   Nevertheless, if the fundamentals of the boron-pH technique are assumed correct and the empirical value of a_4-3 is applied to foraminifera d¹¹B values measured for the past 60 million years, revised pH reconstructions indicate little change in ocean pH over the Cenozoic.  Ambiguity in pH increases as vital effects, as well as various models for the evolution of d¹¹B_sw are considered.  Given the current understanding of boron systematics, pH values estimated by this technique likely have considerable uncertainty, particularly as reconstructions exceed the residence time of boron in the ocean.