Fig. 1: P(V) and P(III) sorption patterns from hydrous ferric oxide co-precipitation experiments.
From: Bioavailable phosphite in the surface ocean during the Great Oxidation Event
A compiled P(V) sorption data from previous studies28 are compared with new data (black and red). SW stands for seawater. ‘Natural SW’ is the low-nutrient Sargasso seawater whereas ‘Artificial SW’ is artificially prepared containing NaCl, Ca2+, and Mg2+ and representing the Archean calcitic seawater. Kads is the coefficient of sorption. Individual datapoints for artificial seawater with 2.2 mM Si were not available, therefore the trend line (orange line) is reconstructed from the Kads value28. Data for the natural SW28 show the effect of Si on sorption while the other three datasets show the effect of salinity. B, C P(III) adsorption data generated in this study. P(III) adsorption is limited compared to P(V) irrespective of solution chemistry. Error bars in (A–C) represent standard deviations of the means. For a direct comparison with P(V) sorption trend lines in (A), those in (B) are not forced to go through the origin. Two datapoints shown by hollow circles in (B) are not considered for producing the trendline because of experimental error. D Kads values used for P(V) and P(III) estimates in seawater around the Neoarchean-Paleoproterozoic boundary. As we used the reported Kads value to produce the trend line for P(V), it went through the origin. For a direct comparison, the trendline for P(III) is forced to go through the origin.
