V. Conclusions and Unresolved Issues

There is considerable debate over the composition and even the existence of some homo- and heteroleptic aquo-, fluoro-, and hydroxo complexes of silicon-(IV), which makes it impossible to predict what species might be found in real potable water supplies that are fluoridated or those that naturally contain fluoride and silicates as background ions. The only agreement seems to be that hexafluorosilicate does not undergo cumulative, consecutive displacements of hydroxide for fluoride. Even this agreed-upon “fact” would seem to be drawn into question by some of the observations of partial hydrolysis products in moist air, suggesting instead that the analytical tools were (are?) incapable of detecting the very low concentrations that might exist. Given the disparity in the speciation models, there is hardly conclusive evidence that consecutive, cumulative substitutions of hydroxide for fluoride are impossible. This assertion is based on the inability to fit potentiometric and some spectrometric data to a suitable system of equations, although it appears to be supported by multiple investigations. The available evidence suggests that neither hexafluorosilicate nor its partial dissociation/hydrolysis products would complex with any transition- metal cations.

Many of the studies of the transformation of hexafluorosilicate have been geared toward ensuring a minimal concentration of free fluoride as a public health measure rather than knowing the equilibrium concentrations of all fluorosilicon(IV) complexes. Whether residual fluorosilicates or fluorosilicon(IV) complexes will be detectable with current instrumentation is debatable. Accounting for the effects of other chemicals on the fluoride speciation will make such analyses even more difficult. Fluoro complexes of some metals are well-studied, but only homoleptic complexes rather than various heteroleptic species that might be encountered in a typical aquatic system where a range of background mineral salts abound. Accordingly, there is a need for further study of heteroleptic fluoride complexes (especially with the common anions in drinking water) of aluminum(III) and possibly other metal cations. However, such research can reasonably be restricted to using fluoride salts (e.g., NaF or KF) rather than fluorosilicates.

Perturbation by dilution (as in a concentration jump type of experiment) will probably be necessary to understand whether industrial-grade hexafluorosilicic acid actually contains polymers such as H2Si2F10 which must also undergo reaction when used as drinking water additives. Such approaches may also dispel unsupported claims that heteroleptic species such as SiF4(H2O)2 can spontaneously decompose to give Si(OH)4 and SiF6 2-, akin to a disproportionation in redox chemistry.

The kinetics of the dissociation and hydrolysis of hexafluorosilicate are poorly understood from a mechanistic or fundamental perspective. Most of the studies have been rather crude, simply adding a certain amount of the material to water and waiting a set time. The analytical tools applied have not necessarily been chosen for their optimal performance on such a task. The stability of silicon tetrafluoride in water, the formation of aquo (or other) adducts, and the rate of SiF4 hydrolysis have been studied in a very cursory fashion and barely at all by adding tetrafluorosilane gas to water. Possible inhibitory effects from fluoride have not been investigated; only one qualitative report exists in the literature. Accelerative effects expected from various metal cations or hydrogen ion have not been fully probed.

On the other hand, all the rate data suggest that equilibrium should have been achieved by the time the water reaches the consumer’s tap if not by the time it leaves the waterworks plant. Thus, better knowledge of the conditions at equilibrium are critical for planning any pharmacokinetic, pharmacological, toxicokinetic, or toxicological experiments. The EPA is aware of papers positing links between fluoridation agents and lead in the bloodstream or challenging the accepted chemistry. [117,134,135] To truly investigate such hypotheses, better chemical knowledge of the speciation is required. For the time being, it is probably best to stop using qualified expressions such as “virtually complete” or “essentially complete” in favor of more rigorous and quantitative descriptions, even if that hinders communication with the lay public. Once the equilibrium speciation and the rate laws have been better elucidated, it may be possible to perform tests to define a level of completeness and the time required to attain that level based on health effects data, which would appear to be most suited to the public health objectives behind drinking water regulation. A purely chemical definition of complete would appear to be rather arbitrary in nature, albeit easier to specify.