3.1. Removal of chloride in ion chromatograph
The chromatograms of the IC in chloride-removal mode (Fig. 1c, d) revealed complete removal of chloride while phosphite and other ions contained in the solution were retained (Fig. 2b). The comparison to the normal mode (Fig. 2a) shows that also nitrite was removed by this method as it elutes close to chloride. However, nitrate, phosphite, sulfate and phosphate were retained. Chloride removal was incomplete when the delay on Valve 2 (i.e., the time before Valve 2 switches form ‘load’ to ‘inject’ mode) was less than 2 minutes. It may be possible to also cut off anions that elute later than phosphite (such as sulfate or phosphate) by switching Valve 2 back to ‘load’ as soon as phosphite has been pushed from the clean-up column to the guard column. This was not explored in this study.
The chromatograms with the Cl-removal method display a larger carbonate peak than those generated in normal mode (Fig. 2) despite the use of carbonate-free NaOH. We suspect that some carbonate was already dissolved in the samples and standards, derived from atmospheric CO2. Additional CO2 may have dissolved in the NaOH solution during preparation. Care was taken to minimize ingrowth of atmospheric CO2 (Section 2.1), but we could not guarantee complete avoidance. The retention time of carbonate is only slightly longer than that of phosphite, and therefore the carbonate peak may reduce confidence in the detection and quantification of phosphite with the conductivity detector on the IC. However, this problem is resolved by IC-ICPMS coupling, as the ICP-MS measures phosphite by mass and is therefore not impacted by the presence of carbonate ions.