Most toxicological studies have been conducted with silver in the free, elemental or 0 oxidation state, and with the +1 monovalent silver ion. The rarer +2 and +3 oxidation states have not been studied adequately (Anon. 1990, Taylor 1964, MRI 1975, Thompson 1973, Boyle 1968, Mulvey 1978, Bertine et al. 1971, DeMayo et al. 1979, Klein 1978 and Presant et al. 1965).
The majority of silver resulting from photo-processing occurs in an insoluble form. Theoretical calculations of organic and inorganic silver complexes indicate that, due to the low solubility of silver sulphide and the high affinity of silver for sulphide, little free silver, usually <10 to 12 µg/L, would occur at equilibrium, in effluents or surface waters that contained any sulphide (IJC 1982). Bioassays have demonstrated that 'free' silver caused the death of fathead minnows at 5 to 16 µg/L, but the salts, silver thiosulphate and silver sulphide, had no effect at 11 000 to 21 000 µg/L (LeBlanc et al. 1984).
A method for measuring 'log silver ion activity' (pAg) with a specific ion electrode has been developed to measure 'free' silver in surface waters. Applying this method to Lake Ontario surface waters yielded pAg values in the 8.5 to 11.1 range, mostly 9.2 to 9.4. These values were claimed to be equivalent to 'free' silver ion levels of 0.340 to 0.00079 µg/L, primarily 0.054 to 0.63 µg/L (Lockhart 1980, Bionomics 1980a, Bionomics 1980b and Chudd 1979).
The authors indicate that the information in the prior paragraph supports an aquatic life objective of about 0.1 µg/L 'free' silver in order to limit that fraction of the total silver that is biologically active or toxic. The Aquatic Ecosystem Objectives Committee of IJC agrees that 'free' silver is a better measure of toxicity than total silver but does not recommend the adoption of an objective based on 'free' silver for the following reasons:
-near an effluent, silver may not yet be in equilibrium with all the available complexing agents.
-sulphide, which is the only reactant likely to reduce silver adequately, is readily oxidized to sulphate when oxygen is available (Chen et al. 1972). The half-life of sulphide is about 50 hours but oxidation rates may be increased fivefold by metals such as calcium. Sulphide has not been found, even at detection limits of 0.0001 µg/L, in the Great Lakes.
-weaker organic complexing agents for silver still permit 'free' silver to exist at concentrations near the objective level of 0.1 µg/L.
-the method presented by Kodak for measuring silver ion activity (Chudd 1979) is not adequate for low environmental concentrations; there are difficulties with reproducibility, dependability and comparability. The method is reliable at levels of 'free' silver higher than 11 µg/L.
Until a reliable method is developed to measure 'free' silver at concentrations below 0.1 µg/L, silver objectives should be expressed as total silver (IJC 1982). This is still the routine total or dissolved silver detection limit for water samples analyzed in labs used by the British Columbia Ministry of Environment for analysis of ambient water samples.
In this report the statement `water of hardness 230' indicates water with a hardness equivalent to that caused by 230 mg/L of CaCO3.