Spectrophotometric Winkler determination of dissolved oxygen: re-examination of critical factors and reliability

The spectrophotometric iodine measurement for oxygen determination by the Winkler method is re-examined for theoretical and operational aspects. It is shown that the selection of an isobestic point for measuring the mixture of iodine and tri-iodide in the solution enhances reliability. The wavelength value of 466 nm was selected after a spectrum study. Then, performances are assessed by mean of statistical approach based on repeated standardization curves at the isobestic point compared with other wavelengths. At 466 nm, the method is linear up to about 1000 μmol kg−1 of O2. Optimization and validation of the spectrophotometric method is realized through a robustness study based on factors that may alter the results (temperature, reagent volumes, storage time). Temperature is the only factor that exerts a significant influence on the absorbance (0.5% per °C), hence samples and standards should be kept at the same temperature within ±1 °C. Precision was estimated over a 10-day period. The standard deviation was 0.45% for inter-sample repeatability and 0.73% for reproducibility near 250 μmol kg−1. The repeatability of 12 samples taken at the same immersion was around 0.12%. The detection limit, determined from standardization curves, is <2 μmol kg−1. Accuracy was verified with O2 saturated seawater and found to be within the precision confidence interval. The method can be applied to the determination of dissolved oxygen in fresh or seawater samples, suboxic, or waters supersaturated in dissolved oxygen. It is suggested as being a suitable alternative to titration in most applications of the Winkler determination of dissolved oxygen.