Determinations of the Solubility of Oxygen in l?ure Water1

A recently published revision of the absorption cocfficienLs for oxygen in water has prompted a reinvestigation of the subject by a new microgasometric technique. A known volume of gas-free water (approximately 1 ml) was brought into contact with oxygen at constant pressure and temperature. The oxygen uptake and the amount of water were measured with a micrometer burette. The results obtained at 6”, 1‘2”, 18”, and 24” C confirm the standard handbook values of Winkler and Fox to within =tl% of Q. The absorption coefficients for oxygen in pure water as given in the standard handbooks and tables (1952, 1955-1956) are based upon data obtained by Winkler (1889, 1891) and Fox (1909). Winkler made two series of determinations, one with a titrimetric method, the other with a gasometric method. In the latter he measured the amount of oxygen taken up by a known volume of gas-free water. The two sets of figures checked each other within a fraction of one per cent. Fox employed a gasometric method of the same principle as Winkler’s. In the temperature range between 0” and 25°C #‘ox’s data are l-2 % higher than those of Winkler. The nitrogen solubility in water was determined by the same gasometric method as used for oxygen, and again Fox’s data were some 2 % higher than Winkler’s. Winkler’s results, both for oxygen and nitrogen have been checked at 20°C within 1% by Scholander et al. (1955), employing a micro volumetric gas extraction method. Recently Truesdale et al. (1.955), using the Winkler titration, published a revision of the solubility coefficients for oxygen in pure water and in sea water of varying salinity. The values, which are some 3 % lower than those of Winkler, were confirmed by Wheatland and Smith (1955), using a volumetric gas-extraction method. Their water samples were prepared by the technique of 1 I am greatly indebted to Professor 1’. I?. Scholander for encouraging discussion and advice and Mag. Scient. Helge Leivestad for aid in preparing the manuscript. Truesdale et al., involving transfer of the equilibrated water samples under exposure to air. The revised solubility values have already won entrance in oceanographic literature (Richards and Corwin 1956). Inasmuch as Winkler’s and Fox’s large series of data were cxccuted with a simple, strictly anaerobic, and apparently impeccable technique, it seemed desirable to check the proposed revision. The method described below can bc used to determine the solubility of any ideal gas in pure water with an accuracy of &l % or better. It can be used for sea water or other saline solutions only provided these are made gas fret by a technique which does not product any change in water content or salt composition.