Samples for in-situ dissolved Fe(II) (dFe(II)) detection (see Supplemental Files: Table 1 and Figure 1):
\n157 samples were as soon as possible analysed from the Geotraces carousel from Stn 38-66;
\n5 ice samples from Stn43 /46\u00a0(operated by Ana Aguilar, Rob Rember);
\n3 small boat surface samples from Stan60,61,66\u00a0(operated by Ana Aguilar, Rob Rember);
\n1 Multicorer niskin sample from Stn66\u00a0(operated by Greg Cutter).
\n
Shipboard measurements were performed by Dr. Maija I. Heller for Fe(II) following a method by King et al. The photomultiplier (PMT) for analysis was borrowed from the FS Polarstern (Germany) at the North Pole from scientist Micha Rijkenberg, NIOZ, (Netherlands) since the PMT owned by UCSC (USA) was damaged at the beginning of the cruise, likely due to a bad power supply.
\n
Dissolved Fe(II) concentrations were determined using an automated flow injection analysis system (FeLume II Waterville Analytical) employing a luminol chemiluminescence based detection system (King et al, 1995). Luminol, prepared in 18.2 MO Milli-Q water and buffered to pH 10.3 with ultra pure ammonia (here used quartz distilled at UCSC), reacts with an Fe(II)- containing solution, resulting in luminol oxidation with concurrent chemiluminescent emission (Rose & Waite, 2001; Croot & Laan, 2002). The FeLume was fitted with a standard quartz flow cell and a Hamamatsu HC135 photon counter configured with the following settings: flow rate: 2.5 mL per minute; photon counter integration time: 200 ms; load time: 20\u201340 s. The mixing and reaction occur in a spiral flow cell positioned in front of a photomultiplier tube. The sample and luminol reagent were directly continually mixed in the flow cell by omitting the injection valve (Rose & Waite, 2001; Hopkinson & Barbeau, 2007; Roy et al, 2008). Once the signal was in steady-state, the mean of the last 50 data points was used to determine the signal. Fe(II) was quantified by 6-8 standard additions of Fe(II) (typically 0.025, 0.05, 0.1, 0.2, 0.4 and 0.8 nM) to samples of seawater which were stored in the dark for several days and showed low initial signals for Fe(II). A 0.01 M Fe(II) stock solution was prepared with ferrous sulfate (Sigma) in 0.2 M HCl. A working Fe(II) standard solution (100 nM) was prepared daily. Detection limits were determined for dark corrected aged seawater samples where ferrous Fe was negligible based on a standard 3r evaluation of the baseline signal (Moffett, et al. 2007; Kondo & Moffett, 2013; Heller et al., 2017).
\n
Instrumentation:
\nThe chemiluminescence was detected with a Hamamatsu HC135 Photon counter build into a flow injection system (FeLume, Waterville Analytical, US) borrowed from NIOZ (Netherlands) as described above (Rijkenberg et al., 2014).
\n
Problem Report:
\nThe required PMT for analysis was borrowed from the FS Polarstern (Germany) at the North Pole from NIOZ scientist Micha Rijkenberg, (Netherlands), since the PMT owned by UCSC (USA) was damaged at the very beginning of the cruise, most likely by a faulty power supply.
Dissolved Fe(II) concentrations were determined using an automated flow injection analysis system (FeLume II Waterville Analytical) employing a luminol chemiluminescence based detection system as described in King et\u00a0al. (1995) and Heller et al. (2017).
\n
From published and ongoing work in the Peru and Chile oxygen minimum zones an optimized methodology for Fe(II) detection by luminol has been described in Croot et al. (2019). This work details some of the challenges and limitations of using luminol for dFe(II) detection. It should be noted that as Fe(II) is a transient redox species with a short half life in oxygenated seawater that is dependent on dissolved oxygen\u00a0and pH, it is not possible to archive (preserve) samples for later analysis. Intercalibrations for Fe(II) have yet to be conducted in the context of GEOTRACES studies and should be pursued where logistically possible.\u00a0\u00a0
Quality Control:
\nSee Supplemental File \"Figure\u00a02\" for examples for standard additions of Fe(II) and blank calculation of the above described method.
\n
Quality Flags:
\nAll data have been assigned quality flags using the GTSPP convention and interpretation. Values below the detection limit were flagged as QF=6 in the GTSPP convention (also adopted by SeaDataNet and recommended by the GEOTRACES programme):
\n1 = good\u2014passed lab QC and oceanographically consistent;
\n2 = possibly good\u2014oceanographically consistent, but have minor sampling/instrumental issues;
\n3 = possibly bad\u2014not oceanographically consistent, or have major sampling/instrumental issues;
\n4 = bad\u2014failed lab QC or known issue with samples;
\n6 = below detection limit;
\n9 = data missing (including all \u201cnd\u201d).
\n
BCO-DMO Processing:
\nVersion history:
\n2020-05-15 - original data version;
\n2020-09-23 - made corrections to the following duplicate, erroneous sample and event numbers:
\n--\u00a0when {Event_ID}\u00a0== '6303', replaced {Sample_ID} 11891 with 11591;
\n-- when {Start_Time_UTC} == '2015-09-10T10:19:00.00Z', replaced {Event_ID} 6249 with 6253.