Methodology:
Particulate acid volatile sulfide (pAVS) was determined at sea using the Radford-Knoery and Cutter (1993) method. This method entails placing the frozen Supor filter into a gas stripping vessel with DDI water and purging with helium, acidification to 1 M hydrochloric acid, and then gas stripping and cryogenically trapping the hydrogen sulfide and subsequent quantification using a gas chromatograph coupled with a flame photometric detector. The particulate chromium reducible sulfide (pCRS) was determined back at the ODU laboratory using the Radford-Knoery and Cutter (1993) method like pAVS but acidifying the QMA filter to 4 M hydrochloric acid with the addition of 1 M chromium (II) chloride.

Sampling and analytical procedures:
Particulate sulfide samples were collected using modified McLane in-situ pumps with two mini-MULVFS filter holders (Bishop et al., 2012). One of the holders contained a 51 μm polyester mesh prefilter and particles were collected on a 0.8 μm polyethersulfone Supor membrane filter (0.8 – 51 μm size fraction). The Supor filters had between 100 and 1,100 L filtered through them with an average of 400 L. The other filter holder contained the same prefilter but collected particles on a Whatman QMA quartz fiber filter (1 – 51 μm size fraction) and filtered between 300 and 1,540 L with an average of 1,100 L filtration volume. Excess seawater was removed using a vacuum pump within an hour of pump recovery. Both the Supor and QMA filters were then subsampled in the trace metal clean lab by P. Lam’s (UCSD) pump team and placed in cryovials that were subsequently stored frozen at -80°C until analysis. A subfraction of the Supor and QMA filters for particulate sulfide analysis were analyzed, yielding 12.5% and 2.3% of each filter with average filter volumes of 49.5 L and 23.4 L, respectively.

The pAVS samples were analyzed at sea when time allowed. Upon analysis, the frozen Supor filter was placed in a 50 mL gas stripping vessel with 10 mL of DDI water and purged with helium (100 mL/min) for 2 minutes. After 2 minutes has passed, the cryogenic trap was immersed in liquid nitrogen and 1 M hydrochloric acid was injected through the Teflon septum and stripped/trapped the gases for 15 minutes before quantifying pAVS using a gas chromatograph coupled with a flame photometric detector.

The pCRS samples were analyzed back at the ODU laboratory by the end of November 2019. Upon analysis, the frozen QMA filter was placed in the 50 mL gas stripping vessel with 10 mL of DDI water and purged with helium (100 mL/min) for 2 minutes. After 2 minutes has passed, the cryogenic trap was immersed in liquid nitrogen and 4 M hydrochloric acid and 1 M chromium (II) chloride were injected through the Teflon septum and stripped/trapped for 20 minutes before quantifying pCRS using a gas chromatography coupled with a flame photometric detector. Each day prior to analysis, chromium (II) chloride blanks were recorded and later applied as corrections to the pCRS concentration due to the addition of the chromium (II) chloride as a reducing agent.

This method has particle detection limits of 0.2 pmol S for acid volatile sulfide and 2 pmol S for chromium reducible sulfide (Radford-Knoery & Cutter, 1993) which corresponds to relative detection limits of 0.004 pmol L-1 pAVS and 0.085 pmol L-1 pCRS with average filtered volumes of 49.5 L and 23.4 L, respectively. Both pAVS and pCRS samples were analyzed mostly in single analyses, however, previous studies report pAVS and pCRS precision near 10% RSD (Cutter & Kluckhohn, 1999; Cutter & Radford-Knoery, 1991). To ensure accuracy, the H2S and OCS gases are calibrated using permeation tubes whose permeation rates have been gravimetrically measured for 2-4 years. By trapping and measuring known amounts of permeated H2S and OCS over a range of times, linear calibration curves for each gas were assembled daily and applied to the unknown samples.

Instruments:
These samples were processed following the analytical apparatus first stated by Radford-Knoery and Cutter (1993). All details can be found in that publication.

A Hewlett Packard (HP) 5890 Series II gas chromatograph coupled with a HP flame photometric detector (model 19256A) was used in the quantification of pAVS and pCRS. The output signal from the detector was processed using a PeakSimple Chromatography Data System (model 333). Hydrogen (130 mL/min) and air (130 mL/min) are used for the flame photometric detector’s flame while ultra-high purity helium is used as the carrier gas (30 mL/min) and the stripping gas (100 mL/min). While cylinders were used as the source of air and ultra-high purity helium, a VWR hydrogen generator (model H2PEM-165) was used for the hydrogen source. A VWR circulating water bath (model 1130S) was used to maintain 40°C for H2S and OCS permeation devices (Metronics) which were used to calibrate the instrument for H2S and OCS.

Known Problems/Issues:
Occasionally, there were technological errors where the connection between the detector and the PeakSimple data system became loose and resulted in peaks that were cut off and thus could no longer be quantified. These data have been flagged 9 according to the SeaDataNet scheme.

Quality Flags:
The SeaDataNet scheme was used to assign data quality flags to samples. More information can be found at https://www.seadatanet.org/Standards/Data-Quality-Control.

The reported codes for flagged data are:
0 = no quality control
1 = good value
2 = probably good value
3 = probably bad value
4 = bad value
5 = changed value
6 = value below detection
7 = value in excess
8 = interpolated value
9 = missing value