An intercalibration between the GEOTRACES GO‐FLO and the MITESS/Vanes sampling systems for dissolved iron concentration analyses (and a closer look at adsorption effects)

Abstract

An intercalibration of dissolved iron (dFe) concentrations was conducted from samples collected on the GEOTRACES Pacific Intercalibration cruise using two different sampling devices: the GEOTRACES GO‐FLO rosette system and MITESS/Vane samplers. At each depth, the dFe concentrations were identical within analytical error, except at 500 m where contamination in one bottle is suspected. dFe adsorption kinetics to bottle walls was also investigated. Over 29 h, 18% of the dFe adsorbed to the walls of 1 L bottles, whereas over 15 h, 19% adsorbed to the walls of 250 mL bottles, suggesting a relationship between dFe adsorption and sample bottle surface area to volume ratio. Contrary to expectations that refrigeration would slow adsorption, cold 250 mL bottles demonstrated a 29% dFe loss over 15 h compared to 19% loss at room temperature. Finally, we tested the hypothesis that the decreasing dFe observed in successive sub‐sampled bottles from the (unacidified) SAFe D1 tank was due not only to adsorption but also to pH‐dependent Fe solubility changes resulting from carbon dioxide outgassing to the headspace of the 500L SAFe D1 tank. Filtered, low‐pH seawater collected at 1000 m in the North Pacific was placed into bottles with variable headspace for 15–17 h. The pH rose with increasing headspace, demonstrating that carbon dioxide outgassed, and dFe decreased in magnitude similar to the SAFe D1 sample. Fe size fractionation results did not conclusively reveal an Fe loss mechanism, but estimates of wall adsorption predicted from our adsorption experiments suggest that the decrease in dFe was more than can be expected by simple adsorption.