Analytical intercomparison between flow injection‐chemiluminescence and flow injection‐spectrophotometry for the determination of picomolar concentrations of iron in seawater

• Published in: Limnology and Oceanography: Methods
• Main Authors: Bowie, Andrew R., Sedwick, Peter N., Worsfold, Paul J.
• Format: Article in Journal/Newspaper
• Language: English
• Published: Wiley 2004
• Subjects: Southern Ocean
• Online Access: http://dx.doi.org/10.4319/lom.2004.2.42; https://api.wiley.com/onlinelibrary/tdm/v1/articles/10.4319%2Flom.2004.2.42; https://aslopubs.onlinelibrary.wiley.com/doi/pdf/10.4319/lom.2004.2.42

A lab‐ and ship‐based analytical intercomparison of two flow injection methods for the determination of iron in seawater was conducted, using three different sets of seawater samples collected from the Southern Ocean and South Atlantic. In one exercise, iron was determined in three different size‐fractions (<0.03 µm, <0.4 µm, and unfiltered) in an effort to better characterize the operational nature of each analytical technique with respect to filter size. Measured Fe concentrations were in the range 0.19 to 1.19 nM using flow injection with luminol chemiluminescence detection (FI‐CL), and 0.07 to 1.54 nM using flow injection with catalytic spectrophotometric detection with N,N ‐dimethyl‐ p ‐phenylenediamine dihydrochloride (FI‐DPD). The arithmetic mean for the FI‐CL method was higher (by 0.09 nM) than the FI‐DPD method for dissolved (<0.4 µm) Fe, a difference that is comparable to the analytical blanks, which were as high as 0.13 nM (CL) and 0.09 nM (DPD). There was generally good agreement between the FI‐CL determinations for the <0.03 µm size fraction and the FI‐DPD determinations for the <0.4 µm size fraction in freshly collected samples. Differences in total‐dissolvable (unfiltered) Fe concentrations determined by the two FI methods were more variable, reflecting the added complexity associated with the analysis of partially digested particulate material in these samples. Overall, however, the FI‐CL determinations were significantly ( P = 0.05) lower than the FI‐DPD determinations for the unfiltered samples. Our results suggest that the observed, systematic inter‐method differences reflect measurement of different physicochemical fractions of Fe present in seawater, such that colloidal and/or organic iron species are better determined by the FI‐CL method than the FI‐DPD method. This idea is supported by our observation that inter‐method differences were largest for freshly collected acidified seawater, which suggests extended storage (>6 months) of acidified samples as a possible protocol ...