A comparison of hydrocarbon and alkali metal response in the flame ionization detector used in subcritical water chromatography

Abstract

The flame ionization detector (FID) response toward alkali metals and hydrocarbons was compared. Optimal hydrogen flame gas flow rates were found near 40 mL/min for hydrocarbon response and 80 mL/min for alkali response. While each displayed a linear FID response, alkali metals produced several orders of magnitude greater detector sensitivity than hydrocarbons. Of note, KCl, NaCl, LiCl, and ethanol yielded respective FID sensitivity of about 7500, 980, 130, and 1 mV/μg analyte. This was subsequently demonstrated to greatly alter the FID response of organic salts. For example, while formic acid is normally unresponsive in an FID, its potassium salt could be readily detected here at picogram levels. Conversely, this phenomenon also rendered the FID unsuitable for use with buffered mobile phases containing such salts. In particular, FID background and baseline noise levels for formic acid – sodium formate buffers were about 10 times larger than equivalent experiments with methanol–water and up to two orders of magnitude larger than pure water. Overall, the results show that alkali metals respond much stronger in the FID than do hydrocarbons. Accordingly, their presence in organic analytes or mobile phases must therefore be accounted for when using this detector, particularly in areas such as subcritical water chromatography where it is commonly employed.