2.3. Coupling of IC to ICP-MS
The IC was connected to the ICP-MS with black PEEK tubing. This was spun between the outlet of the conductivity detector on the IC and the inlet of the nebulizer on the ICP-MS. When the instruments were connected in this way, the suppressor on the IC was regenerated with an external water supply at a flow rate of ca. 0.5-1 ml/min, delivered from an N2-pressurized container. (When the IC is not connected to the ICP-MS, the suppressor is regenerated with the waste solution from the detector). The ICP-MS method for this analysis was set up with 750 runs and 1 pass, equivalent to 3 minutes per analysis. In the sequence file, the chromatographic output mode was selected for each sample. While it is possible to run the ICP-MS in automatic mode where a new analysis is externally triggered via a relay cable between the IC and the ICP-MS, we decided against this option, because it would imply that the slit plate used for medium resolution in the ICP-MS would be corroded quickly over the course of several analyses that last up to 55 minutes each. Instead, we manually started the analysis on the ICP-MS ca. 1 minute before the expected arrival of the phosphite peak at the conductivity detector on the IC. The peak itself lasted for ca. 40 seconds. The ICP-MS method was allowed to run for 3 minutes, providing sufficient data before and after the phosphite peak to determine analytical background levels.
The final data were accessed via the Show program under Chromatogram > From Info File > Display Chromatogram, followed by Display > Data View and File > Export (select ASCII format). This dataset could be opened in Excel or Origin (Origin Lab) for further processing. Here, we used Origin for smoothing the data (using the fast furrier transform filter with a points of window value of 15), subtract background levels, and calculate the area under the peak as a metric for signal intensity. We also calculated the peak height, but the calibration curve was found to be less scattered when peak area was used instead. The method was tested with a series of standards containing 0.003 to 1.614 µmol/L phosphite (corresponding to 0.1 to 50 ppb P).