Dissolved organic carbon (DOC) analyses: vials, sample filtration and acidification, matrix effects and stability

<p>Despite the presence of an ISO protocol for the determination of dissolved organic carbon (DOC) since 2018, a variety of protocols is used in the literature. The way of sampling and storage is crucial to get reliable results, especially when DOC concentrations are low. This technical note describes experiments first carried out on DOC contribution from several materials: (i) opaque glass vials versus polypropylene (PP) vials, (ii) filter membranes and (iii) acids. The effect of glass vial decontamination, as well as the temperature of storage (4&#176; C versus -18&#176;C) with time were evaluated. The possible matrix effects due to the presence of sulfides (SH₂S), sodium (Na) or calcium (Ca) in the samples were tested.</p> <p>Opaque glass vial decontamination during 3 h at 450 &#176;C and filtering ultra-pure water through 0.45 &#181;m hydrophilic polytetrafluoroethylene (PTFE) filters previously rinsed with 20 mL resulted in the lowest DOC deviation from the baseline with a 2.6-factor and the lowest relative standard deviation (RSD) at 5% on nine replicates. Compared to the background signal, the lowest DOC concentration was obtained when the acidification was realized with puriss analytical grade hydrochloric acid (HCl) (4.8-factor, RSD = 5%, N= 5).</p> <p>Storage at 4&#176;C ensured minor DOC changes within one month for a 1 mg L^-1 DOC solution (factor of increase less than 1.5) whereas for lower concentrations close to the quantification limit (~ 0.5 mg L^-1), DOC concentrations in samples filtrated through 0.45 &#181;m PTFE filters varied up to 29% after one-week storage. Even if freezing might intuitively seem to be a reliable way to fix the chemistry of a sample with time, frozen samples showed drastic increases in DOC concentration after one month of storage, which went up to factor of increase from 10 for a 1 mg L^-1 DOC acidified solution to 142 for ultra-pure water only.</p> <p>The presence of sulfides (SH₂S) did not induce a significant change in DOC concentration (< 10%) whereas sodium (Na) or calcium (Ca) impacted DOC analyses with underestimations from 53% to 75%.</p>