POSSIBLE ERRORS CAUSED PRIOR TO MEASUREMENT OF MERCURY IN NATURAL-WATERS WITH SPECIAL REFERENCE TO SEAWATER

Taking seawater as an example, analytical problems in the mercury determination of natural waters are presented, with a special reference to errors caused prior to measurement of mercury. A mercury concentration of 0.5 ppb decreases rapidly even in an acidified solution, but the presence of sodium chloride prevents the adsorption loss onto the bottle wall. Therefore, if a seawater sample is acidified to 0.2 M with sulfuric acid at sampling, the mercury is stable for at least 60 days. Polyethylene bottles are unreliable because of mercury contamination. Glass bottles contaminated with mercury can be cleaned by heating at 500 °C or rinsing with diluted hydrofluoric acid. When there is no contamination and no mercury loss, the apparent mercury concentration of an acidified seawater sample increases for a couple of weeks due to a change in form of the mercury species in the seawater, and then reaches a constant value which is consistent with the value determined after digestion with a mixture of nitric acid and sulfuric acid, or a mixture of nitric acid, potassium permanganate, and persulfate. The level of 5-6 ng of Hg/L determined with the considerations described above seems to be the base-line concentration of mercury in the oceans. © 1979, American Chemical Society. All rights reserved.