Characterization of the nonlinear salinity dependence of glass pH electrodes: A simplified spectrophotometric calibration procedure for potentiometric seawater pH measurements at 25 °C in marine and brackish waters: 0.5 ≤ S ≤ 36

Abstract

Glass electrodes are commonly used to measure the pH of natural waters over various, sometimes wide, ranges of salinity (S). For such applications, the electrodes must be calibrated against solutions of known pH and salinity identical to those of the sample solutions. Well-characterized buffer solutions may be used for these calibrations, but if a wide range of salinity is to be encountered in the samples (e.g., as in estuarine transects), this approach is quite laborious. Previous work has demonstrated that for 28.5 < S < 36.1, pH electrodes can be efficiently calibrated spectrophotometrically in seawater because electrode intercept potential E0 (a key calibration parameter) varies linearly with salinity over that range. The present work (a) characterizes pH electrode calibration parameters in seawater over a wider range of salinity (0.5 < S < 36) and (b) provides a simple and efficient method for creating and maintaining “river-to-sea” electrode calibrations over periods of months. Electrode calibration slope (g') was found to be insensitive to salinity, as expected. The value of this parameter, measured at S > 5, was reliably consistent with theoretical expectations, such that repeat verification needs to be conducted only occasionally. Electrode intercept potential (E0), in contrast, was found to depend substantially on salinity: approximately linearly for 5 ≤ S ≤ 36 and substantially nonlinearly for 0.5 ≤ S < 5. Ignoring this dependence of E0 on S can lead to pH misestimates as large as 0.24, with the problem being most severe at lower salinities. Based on these observations, a method was developed by which the dependence of E0 on S can be rapidly ascertained by simultaneously measuring pH (spectrophotometrically) and electromotive force (potentiometrically) in seawater that is serially diluted to produce the full range of salinities to be encountered in sampling. Because no acid titrations are required, a full river-to-sea calibration can be acquired in <3 h. With occasional (daily to weekly) one-point checks/corrections for electrode drift, this calibration is stable for weeks to months.