The determination of fluoride in water in the presence of organic acids and iron using a fluoride-selective electrode

Environ Geochem Health. 1993 Sep;15(2-3):185-6. doi: 10.1007/BF02627837.

Abstract

Surface waters in Scotland, notably from upland catchment areas, are commonly enriched in iron and organic acids. This study investigated the impact of these species on the direct potentiometric determination of fluoride using a fluoride-selective electrode. As the electrode technique is commonly used to monitor the fluoride content of potable waters, it is important that these effects are evaluated if such waters are to be fluoridated.The determination method used was that defined by Nicholson (1983) and Nicholson and Duff (1981) to minimise errors. This employs the TISAB III-TAC buffer system with the following composition (in 1L): 58.0 g sodium chloride, 57.0 mL glacial acetic acid, 4.0 g CDTA, 243.0 g tri-ammonium citrate adjusted to pH 5.4 with 10 M sodium hydroxide. Experimental solutions of fluoride with organic acid or iron were prepared, and the effect on fluoride concentration determined for each combination. Concentrations used: fluoride: 0.1, 1.0 mg L(-1); humic acid: 5, 10, 100, 1000, 10,000 mg L(-1); oxalic acid: 5, 10, 100, 1000, 10,000 mg L(-1); iron(III): 1, 10, 100, 500 mg L(-1);Humic acid (HA) concentrations had no impact on the determination of fluoride at the 1.0 mg L(-1) level. However, with 0.1 mg L(-1); F(-), an increase in the apparent fluoride concentration was observed when HA > 1,000 mg L(-1); this increased with increasing HA content to a maximum of ∼600%. Oxalic acid (OA) generally had no impact on the fluoride determinations at the 1.0 mg L(-1); level, but at an OA concentration of 10,000 mg L(-1); fluoride concentrations were reduced by ∼50%. At 0.1 mg L(-1) F(-), increasing OA concentrations produce a steady increase in the fluoride concentration of up to 200% with 1,000 mg L(-1); OA, greater OA contents produce a fall in the fluoride content. Iron had no effect on the fluoride determinations at both the 0.1 and 1.0 mg L(-1) levels.The causes of the apparent increases in fluoride concentration have not been determined, although fluoride contamination by the reagents has been ruled-out. However, the results demonstrate that the defined method and buffer system is suitable for the determination of fluoride in the presence of iron and organic acids at naturally occurring levels, and that fluoride will not be masked from detection.