Can Fluoridation Affect Water Lead(II) Levels and Lead(II) Neurotoxicity

[Guest guest]

Hi People

I found this article, and as you have been discussing Fluoridation in water,

I thought it might be of some interest.

Regards

Jeff

 

Can Fluoridation Affect Water Lead(II) Levels and Lead(II) Neurotoxicity?a,b

 

Edward T. Urbanskyc and Michael R. Schockd

United States Environmental Protection Agency (EPA), Office of Research and

Development,

National Risk Management Research Laboratory, Water Supply and Water

Resources Division,

Cincinnati, Ohio 45268 USA

Abstract

Recent reports have attempted to show that certain approaches to

fluoridating potable water is

linked to increased levels of lead(II) in the blood. We examine these claims

in light of the established

science and critically evaluate their significance. The completeness of

hexafluorosilicate hydrolysis

is of paramount importance in ensuring that total water quality is

maintained. The possible impacts

of such complexes as PbII–F–SiF5 or PbFx

(2–x) are discussed as are the contributions of fluoridation

byproducts to total acid content. We calculate the fractional distribution

of aqueous species based

on known chemical equilibria and show the species concentrations for a

selected model tap water

with a composition that would favor lead fluoride and silicofluoride

complexation. We discuss and

quantitatively show the effects of other complexing anions, such as

carbonate or hydroxide. Overall,

we conclude that no credible evidence exists to show that water fluoridation

has any quantifiable effects on the solubility, bioavailability,

bioaccumulation, or reactivity of lead(0) or lead(II) compounds. The

governing factors are the concentrations of a number of other species, such

as (bi)carbonate, hydroxide, or chloride, whose effects far exceed those of

fluoride or fluorosilicates under drinking water conditions. Lastly, we

consider some previous epidemiological studies of lead(II) exposure and how

recent papers fare methodologically.

Background

Controversy over water fluoridation varies in nature and intensity. Recent

papers have implications

for water fluoridation since they suggest that certain adverse health or

social conditions may

stem from interactions between lead(II) and inorganic fluoro-compounds,

specifically, fluorosilicates

and fluoride.1-3 In order to assess the validity of these assertions, it is

necessary to have a firm

foundation of the aqueous chemistry of H2SiF6 and HF, as well as of Pb.

There is a considerable body

of fundamental chemical literature on these species. Nonetheless, some gaps

do remain,4 and little

effort has been expended in combining the known chemistry into one

comprehensive and

authoritative volume. Accordingly, we believe it to be worthwhile to revisit

the chemical concepts

and relationships involved in water fluoridation at a fundamental level in

light of well-established

science, and thus to examine the potential validity of some of the

hypotheses of adverse chemical

interactions suggested by the recent papers.1-3 We have previously reviewed

the state of knowledge

in considerable detail,4 but here emphasis is placed on the chemistry and

conditions of most relevance to the public drinking water treatment and

health communities.

The sheer number of people consuming fluoridated potable water makes

fluoridation issues

relevant. In 1992, the Centers for Disease Control Fluoridation Census found

that 62.1% of the U.S.

population served by public suppliers drank fluoridated water.5 The CDC also

surveyed utilities

regarding fluoridating agents (see Table 1). Most commonly used are

hexafluorosilicic acid (H2SiF6)

or its sodium salt (Na2SiF6), which hydrolyze to produce fluoride ion upon

dilution (eqs 1-2).

However, sodium fluoride (NaF) is sometimes used as a direct fluoride source

(eq 3).

H2SiF6(aq) + 4 H2O 6 6 HF(aq) + Si(OH)4(aq) (1)

Na2SiF6(aq) + 4 H2O 6 4 HF(aq) + 2 NaF(aq) + Si(OH)4(aq) (2)

NaF 6 Na+(aq) + F–(aq) (3)

Table 1. Water fluoridation chemicals used by U.S. public water suppliers in

1992

hexafluorosilicic acid sodium hexafluorosilicate sodium fluoride

formula H2SiF6 Na2SiF6 NaF

common

synonyms

fluosilicic acid

fluorosilicic acid

hydrofluosilicic acid

sodium silicofluoride

sodium fluorosilicate

—

population

served

80,019,175

62.6%†

36,084,896

28.2%†

11,701,979

9.2%†

utilities

using

5876 1635 2491

*Total US population: 258,544,000. Population and utility data were taken

from reference 4.

†Percentages are based on total population of 127.8 million persons drinking

fluoride-fortified public

water and does not include those drinking water naturally high in fluoride.

Hexafluorosilicic acid is a cheap and readily available source of fluoride.

However, it is difficult to

handle and the handling costs can only be offset by the volume discount in

large water treatment

plants. Although more systems rely on sodium fluoride than sodium

hexafluorosilicate, these serve

only 9.2% of the U.S. population. Because sodium fluoride is the easiest of

the three to handle and

dispense, small systems are the primary users of NaF. Although the EPA

regulates drinking water,

the US Public Health Service has been involved in water fluoridation for

historical reasons (primarily

because the practice of fluoridation pre-dates EPA). The purpose of

fluoridating water is the

prevention of dental caries; therefore, the publication of water

fluoridation how-to manuals falls

under the purview of the CDC. These manuals discuss dosing and other

practical matters of concern

to the treatment plant operator.6

Drinking water contains a large number of chemical species, including

disinfection byproducts,

residual oxidants, dissolved organic matter, trace metals, minerals, and

additives (such as fluoride).

As a result, drinking water science is a complicated interplay among the

chemical constituents as

well as the physical conditions. We think it is useful to frame the issues

recently raised about

adverse interactions between aqueous lead and fluoridation species,1-3 in a

question-answer format,

to help water managers, scientists, and engineers understand and respond to

them.

1. What is the residual concentration of hexafluorosilicate ion (SiF6

2–) after hydrolysis

and how fast does hydrolysis take place?

Hexafluorosilicate ion reacts with water to produce fluoride ion and an

assortment of silicon

oxyanions,7-8 e.g., SiO3

2–, SiO4

4–, Si(OH)O3

3–. We represent the oxyanions as SiIV(aq) without further speciation at

this time.

SiF6

2– + n H2O º SiIV(aq) + 6 HF(aq) (4)

The actual speciation of silicon oxyanions is a function of acidity, i.e.,

[H+]. Busey et al.9

showed that virtually 100% of the hexafluorosilicate is hydrolyzed to

silicon oxyanions at pH 6, even

when there is a free fluoride concentration of 0.01 M. Meanwhile,

fluoridated drinking water

contains only -1 mg/L of fluoride, which equates to 5 × 10–5 M. Previous

investigations10-11 found

a non-negligible concentration of residual SiF4 when this gas was passed

through water. Ciavatta10

et al. investigated fluorosilicate equilibria with 0.3 # [H+] # 3 molal [m,

mol F– (kg water)–1] and

ionic strength fixed at 3 M, adjusted with LiClO4. They concluded that the

mixed ligand species

SiF(OH)3 and SiF(OH)2(H2O)+ are significant contributors to total

silicon(IV) in addition to SiF4,

SiF6

2–, and HSiF6

· under these conditions. Nonetheless, their results showed that

fluoro-complexes

comprised less than 5 mol% of the total silicon(IV) in 0.01 m H+ and 10–4 m

F–. Korobitsyn et al.11

examined the hydrolysis of sodium hexafluorosilicate in sodium carbonate

solution. Their work was

geared towards an industrial process for producing sodium fluoride and is

not directly applicable here.

The use of chemical shift information derived from 19F NMR spectrometry in

understanding the

formation of fluoro-ligated species is well-established.12-17 Fluoride

ligand exchange occurs rapidly

between HF and SiF6

2– at temperatures above –10 °C,14 and the identification of aqueous

fluorosilicate species and the measurement of the concomitant equilibrium

constants has been done

almost entirely by 19F NMR spectroscopy and spectrophotometry.15-17 The

Gmelin Handbook of

Inorganic Chemistry tabulates values for the equilibrium constants

expression (6) of the hydrolysis

reaction (5) at temperatures from 0 to 60 °C.18

SiF6

2– + 4 H2O º Si(OH)4 + 4 H+ + 6 F– (5)

K ‘

[si(OH)4 ] [H% ] 4 [F & ] 6

[siF 2 &

6 ]

(6)

The smallest value at ambient temperature reported for K is 10–31.6. Using

this value at [H+] = 10–6

M and [F–] = 5 × 10–5 M, the ratio [si(OH)4]/[siF6

2–] = 1.6 × 1018. Note that less than 1% of fluoride

exists as HF at drinking water acid levels (i.e., pH > 5.2) since pKa

HF = 3.17.19 Even if the hydrolysis

constant were off by a factor of 1000, it would not matter. There would

still be essentially no

hexafluorosilicate ion. A fractional distribution plot in Gmelin18 shows

that other fluorosilicates (i.e.,

SiF4 and SiF5

–) also drop off dramatically as free fluoride concentration, and not [F–]T,

decreases towards 10–4 M, even in silica-saturated 4 M perchloric acid. For

this solution, total fluoride concentration is expressible as (7),

neglecting any mixed fluorohydroxo-ligated species:

[F–]T = [HF] + [F–] + 4 [siF4] + 5 [siF5

–] + 6 [siF6

2–] (7)

Crosby studied the dissociation of sodium hexafluorosilicate and

hexafluorosilicic acid in deionized water.20 He found that about 99 mol% of

the hexafluorosilicate had hydrolyzed when added to water to produce a 1

mg/L fluoride solution; however, the pH of this solution was 4.20,

considerably below a potable water pH. An important factor must be

considered in potable water fluoridation as Crosby explained:

It should be remembered that the actual ionic population of most public

drinking-water supplies is somewhat

different from the experimental conditions used in the present and previous

studies. Thus, the pH is normally

adjusted to about 7 to 8, and the presence of additional salts may further

influence the equilibrium owing to the

formation of complexes with calcium and other metals.

If the pH of a treated drinking water is too low, it is adjusted to comply

with regulations (or

consumer complaints) and minimize corrosion. Crosby’s results were obtained

in a water that was

demineralized and completely devoid of buffering agents. Consequently, the

dissociation of

hexafluorosilicate was hindered by the drop in pH. Thus, Crosby’s fractional

dissociation data

cannot be applied directly to a potable water supply without correcting them

for pH. Of course, that

correction is the effect we have computed above, namely, the complete

hydrolysis of fluorosilicates.This is precisely what Crosby was emphasizing.

This observation hints at the effect on pH, which we shall come back to

shortly.

Interestingly enough, a number of species actually promote the dissociation

of hexafluorosilicate,

including ferric ion.21 While the compound PbSiF6•2H2O can be synthesized,

it decomposes

quickly in moist air and slowly when dry.22 Perhaps then lead(II) itself

promotes hexafluorosilicate

decomposition, such as through the formation of plumbous fluoride. Because

moist air promotes this

compound’s destruction, we can infer that it would not be stable in aqueous

solution at all. There is

essentially no hexafluorosilicate remaining in drinking water at

equilibrium.

Now we must consider how fast hydrolysis takes place. In the 1970s,

Plakhotnik conducted

studies into the effects of lithium and calcium cations on the rate of

hexafluorosilicate (and

tetrafluoroborate) hydrolysis.23-24 Based on Plakhotnik’s results, we

calculated4 that the hydrolysis

would be 99 mol% complete in 12 minutes if carried entirely by the

uncatalyzed pathway. That

notwithstanding, natural water supplies do contain calcium and other

divalent metals as well as

trivalent metal cations (e.g., Al3+, Fe3+); hence, the actual hydrolysis

rate would be even faster so that

equilibrium is reached long before water reaches the consumer’s tap.

Based on the above information on both the thermodynamics of the hydrolysis

reaction and its

kinetics, we can safely conclude that there is essentially no (« 1 part per

trillion) hexafluorosilicate

remaining in drinking water at equilibrium and that equilibrium is rapidly

reached from the combined uncatalyzed and metal-catalyzed reactions.

2. Can F– or residual SiF6

2– complex with Pb2+ and make it more bioavailable?

Another way to ask this is: Do fluoro-species complex with Pb(II), promoting

permeation of the

gastric mucosa and absorption into the bloodstream? Even though we have

demonstrated that there

is no hexafluorosilicate remaining by the time water reaches the consumer’s

tap, the following scenarios nicely illustrate the magnitude of the effects

on lead(II).

To produce 1.0 mg/L fluoride requires an initial hexafluorosilicate

concentration of 8.8 µM. The

hydrolysis reaction (eqn 2) is a reversible equilibrium, and in the most

acidic gastric conditions, the

pH could be as low as 1.5 so that [H+] = 10–1.5 M (assuming unit

activities). Using this hydrogen ion

concentration, we calculate the ratio [si(OH)4]/[siF6

2–] = 4.5 × 105. This means that only 0.00022%

of the total silicon(IV) is present as the hexafluorosilicate ion so that

[siF6 2–] = 1.9 × 10–11 M = 19 picomolar (pM).

Haque and Cyr25 showed that hexafluorosilicate anion forms complexes with

several metal

cations: CuII, NiII, CoII, and FeIII. The largest stability constant they

obtained was for the reaction

with ferrous ion, with K = 1.2 = 100.08. Let us assume that the lead(II) ion

forms a stabler complex

and set the stability constant for eq 18 to an arbitrarily high value of

100. In addition, we shall pretend that the hydrolysis computed above has

not occurred, that all 8.8 µM of the silicon(IV) remains in the form of

hexafluorosilicate ion. In this worst case, only 0.088 mol% of the total

lead(II)would be in the form of a hexafluorosilicate complex. The µ-fluoro

ligand would serve as a link between the silicon(IV) and lead(II).

Pb2+ + SiF6

2– º PbSiF6(aq), K = 100 (hypothetical) (8)

Now consider the actual case where [siF6

2–] = 19 pM. If the equilibrium constant for eqn (8)

were 10,000 times larger, say 106, so that the reaction could be treated as

going to near completion,

there would still be less than 19 pM Pb–µ-F–SiF5. Because of the magnitude

of the equilibrium

constant for eqn (6), the equilibrium constant for eqn 8 would have to

exceed 1025 in order to have

a quantitatable effect by preventing hexafluorosilicate hydrolysis. There is

no basis in fact for such

an assertion. As a final point, we note that the national primary drinking

water standards are intentionally

predicated on the assumption that all lead is bioavailable, and the water

utilities should be complying with these standards.

What about the effect of the fluoride itself? Can it promote lead(II)

bioabsorption? Is there an association between lead(II) and fluoride?

Therefore, dosing the chemical NaF does affect pH indirectly via eqn (9)

because HF is a weak acid with a pKa of 3.17.19 However, its effect can

generally be neglected since the pH of drinking water is controlled by many

different buffering species,26 which will be discussed later.

F– + H2O º HF + OH–, Kb . 10–10.5 (9)

There is only a very small association27-28 between Na+ and F– (Table 2, eqn

F1). The magnitude

of this stability constant is so small as to be negligible; however it can

still be calculated. On the other hand, other cations present in reasonably

high concentrations, most notably aluminum, bind to fluoride much more

strongly. Table 2 summarizes these chemical equilibria and their stability

constants.

Table 2. Cumulative stability constants for formation of fluoro-complexes*

Fluoro-complexation Eqn log $

Na+ + F– º NaF(aq) (F1) –0.24†

Al3+ + F– º AlF2+ (F2) 7.0†

Al3+ + 2 F– º AlF2

+ (F3) 12.7†

Al3+ + 3 F– º AlF3(aq) (F4) 16.8†

Al3+ + 4 F– º AlF4

– (F5) 19.4†

Al3+ + 5 F– º AlF5

2– (F6) 20.6†

Al3+ + 6 F– º AlF6

3– (F7) 20.6†

Al3+ + H2O + F– º AlOHF+ (F16) 0.0§

Al3+ +H2O + 2F- º AlOHF2° (F17) 20.6§

Fe3+ + F– º FeF2+ (F8) 5.2‡

Fe3+ + 2 F– º FeF2

+ (F9) 9.1‡

Fe3+ + 3 F– º FeF3(aq) (F10) 11.9‡

Ca2+ + F– º CaF+ (F11) 0.94†

Mg2+ + F– º MgF+ (F12) 1.82†

Cu2+ + F– º CuF+ (F13) 1.2‡

H+ + F– º HF(aq) (F14) 3.18†

H+ + 2 F– º HF2

– (F15) 3.76†

2H+ + 2 F– º H2F2° (F18) 6.77§

*These stability constants are used for the construction of Figures 1-4 with

the exception of Equations (F8)-(F10) and

(F13). †Values taken from reference 28. ‡Values taken from reference 19.

§Values taken from reference 32.

There are many metal cations competing for the fluoride; therefore, the free

fluoride available

to complex with the lead(II) ion is very small. In addition, most, if not

all, of the competing metal

cations are in greater abundance than lead(II) by orders of magnitude.

Further reducing the lead(II)

are such ligands as hydroxide, chloride, carbonate, bicarbonate, and

sulfate, all of which compete

with fluoride for the lead(II) and are present in far greater

concentrations. Table 3 summarizes these

equilibria and their stability constants. For pH > 6, the free lead(II)

concentration drops off

dramatically from hydroxo- and (bi)carbonato-complexation. That drinking

water contains a

substantial fraction of fluoroaluminum complexes rather than free fluoride

was highlighted by Pitter

as a concern because free fluoride is more effective in protecting against

tooth decay.29 We shall take

these and other factors into account in speciating the lead(II).

Table 3. Lead(II) equilibria and constants*

Equilibrium Eqn log $

Pb2+ + H2O º PbOH+ (L1) –7.22

Pb2+ + 2 H2O º Pb(OH)2(aq) + 2 H+ (L2) –16.91

Pb2+ + 3 H2O º Pb(OH)3

– + 3 H+ (L3) –28.08

Pb2+ + 4 H2O º Pb(OH)4

2– + 4 H+ (L4) –39.72

2 Pb2+ + H2O º Pb2OH3+ + H+ (L5) –6.36

3 Pb2+ + 4 H2O º Pb3(OH)4

2+ + 4 H+ (L6) –23.86

Pb2+ + CO3

2– º PbCO3(aq) (L7) 7.10

Pb2+ + 2 CO3

2– º Pb(CO3)2

2– (L8) 10.33

Pb2+ + H+ + CO3

2– º PbHCO3

+ (L9) 12.59

Pb2+ + SO4

2– º PbSO4(aq) (L10) 2.73

Pb2+ + 2 SO4

2– º Pb(SO4)2

2– (L11) 3.50

Pb2+ + Cl– º PbCl+ (L12) 1.6

Pb2+ + 2 Cl– º PbCl2(aq) (L13) 1.8

Pb2+ + 3 Cl– º PbCl3

– (L14) 1.7

Pb2+ + 4 Cl– º PbCl4

2– (L15) 1.4

Pb2+ + F– º PbF+ (L16) 2.06

Pb2+ + 2 F– º PbF2(aq) (L17) 3.42

Pb2++ H4SiO4(aq) + 4 H+ + 6 F– º (L18)

Pb-F-SiF5(aq) + 4 H2O

35.18†

*Values derived from Table 4-16 in reference 30 at 25 EC and zero ionic

strength. These equilibria are used in the construction

of Figures 1-4.

†Computed from combining the dissociation constant for the reaction Si(OH)4

+ 4 H+ + 6 F º SiF6

2– + 4 H2O, log K =

30.18 (from reference 28) with Equation (8), but using an extremely

exaggerated hypothetical value for K in equation

(8). We believe this value to be an intentional overestimate by a factor of

at least 1000-2000 over the likely value of the

true stability constant, which has not been measured.

One might logically inquire whether PbF2 can precipitate under drinking

water or physiological conditions. The solubility product expression for

plumbous fluoride is:

PbF2(s) º Pb2+ + 2 F–, Ksp = 10–7.44 (reference 18) (10)

One way to estimate the minimum amount of lead in solution needed to

precipitate lead fluoride

would be to back-calculate from the concentration of the solubility of the

aqueous uncharged

difluorolead(II) coordination complex, PbF2(aq), the concentration of which

can be obtained by

combining equations (L17) and (10). This yields: [PbF2(aq)]max = 9.5 × 10–5

M. Pretending that

there are no competing metal cations and no competing coordinating ligands,

the total Pb(II) concentration is given by eqn (11):

[PbII]T = [Pb2+] + [PbF+] + [PbF2] = [Pb2+] (1 + $1[F–] + $2[F–]2) (11)

where $1 and $2 come from eqns L16 and L17, respectively. In 1.0 mg L–1 free

fluoride (5.3 × 10–5

M) solution, the approximate fractional speciation is as follows:4 fPb2+ =

99.904%, fPbF+ = 0.096%,

and fPbF2(aq) = 0.000099%. We draw attention to the fact that, in

fluoridated water, the number 5.3

× 10–5 M really refers to the total fluoride, which is expressible as (12):

[F–]T = [F–] + [PbF+] + 2 [PbF2] = 5.3 × 10–5 M (12)

Nevertheless, because [F–]T . [F–] (less than 0.1% difference), there is no

point in distinguishing

between these two concentrations. However, in a real water, there are other

metals competing for

fluoride and other ligands competing for lead(II). The competition of other

metal cations for fluoride

as a ligand substantially reduces the free fluoride concentration. Thus, the

required concentration of

lead(II) in solution would be impossibly high:

..9.6 M (13) [Pb II]T

‘

[PbF2(aq)]

fPbF2(aq)

‘ 9.5 × 10–5 M

0.000 000 99

To make the situation even more extreme, aluminum, iron(III), calcium,

magnesium, and copper(II)

all compete with lead(II) for fluoride. Meanwhile hydroxide, carbonate,

phosphate, and sulfate

compete with fluoride for lead(II). The net result of these simultaneous

competitions is that PbF2

cannot precipitate as a solid. Even with an extremely high 90th percentile

lead(II) level of -210 µg

L–1 (- 1 µM), plumbous fluoride would be orders of magnitude from

precipitating.

The formation of soluble fluoro-complexes of Pb(II) is governed solely by

the stability

equilibria, and no simple stoichiometric ratio exists among the

concentrations of lead(II), fluoride,

and the fluor-complexes. If ligand availability alone were the determining

factor, chloride itself

would usually be far more important than fluoride. Considering the relative

stability constants for

the complexes given in Table 3, a chloride concentration of 50 mg L–1 (= 1.4

mM = 1400 µM) is

about 26 times the normal fluoride concentration. In the vast majority of

all cases in drinking water,

concentrations of lead(II) complexes with chloride (and even sulfate)

considerably exceed those of

fluoride.

3. Do fluoridation additives significantly affect the acidity and pH of

consumed drinks constituted with tap water?

How does the acid contribution from hexafluorosilicate hydrolysis compare

with that from

other sources of acidity? Deionized water treated to contain 1 mg L–1

fluoride would contain 53 µM

HF(aq). If one were to drink this solution of 53 µM HF(aq), which is 93 mol%

dissociated to

hydrogen and fluoride ions, it would contain 49 µM H+ and its solution would

have a pH of 4.3. We

will show later that drinking water contains buffering components that

essentially neutralize even

this effect. Meanwhile, the high extreme for stomach pH (lowest acidity) is

about 3 (1000 µM H+);

the lowest stomach pH is about 1.5 (for optimal pepsin enzymatic activity in

the digestion of protein). At pH 3, roughly half of the HF will not ionize

since it is a weak acid. Meanwhile, some foods are equally or more acidic,

for example, apple juice (pH 2.9).

If this reasoning were correct, consuming soft drinks made with unbuffered

tap water should

be high risk, given the high concentrations of complexing organic acids

(e.g., citric and tartaric acids

in powdered fruit drink mixes) or inorganic acids (e.g., phosphoric and

carbonic acids in colas). In

fact, Coleman et al. showed that chelating organic bases (e.g., citrate,

ascorbate, EDTA) promote

the transport of lead(II) in the small intestine.31 The acidic components of

these beverages completely

overwhelm the contribution from HF in the water used to prepare them.

Whether any of these other

species is present in sufficient concentration to influence bioavailability

is unknown. Regardless, acid from hexafluorosilicate-based fluoridation is

negligible compared to other dietary sources.

Consequently, one cannot demonstrate that an increase in blood lead(II) ion

levels can be linked to

acidity from SiF6

2– hydrolysis any more than one can demonstrate it results from consuming

soft drinks.

In the small intestine, bile (produced by the gall bladder) and bicarbonate

(secreted by the

pancreas) raise the pH and effectively buffer against pH change. Partly

digested food in the chyme

also acts as a buffer. Moreover, normal gastric biophysiology resists

changes in acidity by a

mechanism involving gastrin secretion and activity for which a detailed

description is beyond the

scope of this work. In conclusion, the production of acid from fluoridation

of water is insignificant

when compared to other acids and bases supplied by a normal diet or

physiological mechanisms.

4. What are the concentrations of the lead(II) and fluoride species in a

typical drinking water?

To test different hypotheses about the impacts of fluoride ligands on lead

solubility, several

aqueous solutions were modeled with MINEQL+.32 The effect of background ions

such as CO3

2-,

HCO3

-, and PO4

3- and water quality parameters such as pH have been extensively

investigated and

reported in the water treatment literature.26,30, 33,35-43 Free lead(II)

ion, Pb2+, is a very small fraction of

the soluble lead in most drinking water systems.

We have taken into account equilibria of lead(II), aluminum, calcium, and

other metals for such ligands as carbonate, chloride, hydroxide, sulfate,

and, of course, fluoride. These were given earlier in Tables 2 and 3. Other

necessary equilibria and their constants that we have used for this modeling

exercise are shown in Table 6.

Table 6. Other equilibria used to calculate the fractional distribution of

aqueous species*

Equilibrium Eqn log $

CO2(g) º CO2(aq) (E1) –1.468

CO2(aq) + H2O º HCO3

– + H+ (E2) –6.352

HCO3

– º CO3

2– + H+ (E3) –10.329

Na+ + HCO3

– º NaHCO3(aq) (E4) –0.25

Na+ + CO3

2– º NaCO3

– (E5) 1.27

Ca2+ + HCO3

– º CaHCO3

+ (E6) 1.106

Ca2+ + CO3

2– º CaCO3(aq) (E7) 3.224

Mg2+ + HCO3

– º MgHCO3

+ (E8) 1.07

Mg2+ + CO3

2– º MgCO3(aq) (E9) 2.98

Si(OH)4(aq) º SiO(OH)3

– + H+ (E10) –9.83

Si(OH)4(aq) º SiO2(OH)2

2– + 2 H+ (E11) –23.0

HSO4

– º SO4

2– + H+ (E12) –1.988

Ca2+ + SO4

2– º CaSO4(aq) (E13) 2.30

Mg2+ + SO4

2– º MgSO4(aq) (E14) 2.37

Al3+ + SO4

2– º AlSO4

+ (E15) 3.02

Al3+ + 2 SO4

2– º Al(SO4)2

– (E16) 4.92

Al3+ + HSO4

– º AlHSO4

2+ (E17) 0.46

H2O º H+ + OH– (E18) –14.00

Na+ + H2O º NaOH(aq) + H+ (E19) –14.18

Ca2+ + H2O º CaOH+ + H+ (E20) –12.78

Mg2+ + H2O º MgOH+ + H+ (E21) –11.44

Al3+ + H2O º AlOH2+ + H+ (E22) –5.00

Al3+ + 2 H2O º Al(OH)2

+ + 2 H+ (E23) –10.1

Al3+ + 3 H2O º Al(OH)3(aq) + 3 H+ (E24) –16.9

Al3+ + 4 H2O º Al(OH)4

– + 4 H+ (E25) –22.7

Al3+ + SiO(OH)3

– º AlH3SiO4

2+ (E26) –0.785

*Values taken from reference 29, except (E26) from reference 32. These

equilibria are used in the construction of Figures 1–4. Equations (59)-(60),

(67)-(68), and (75) are used for Figure 5.

Calculations were performed for the following hypothetical water solutions,

as a means to

test some plausible limits on when fluoride or fluorosilicate complexes

might be of consequence

with respect to solubility. Conditions are summarized in Table 7. The

lead(II) concentration used

represents the regulatory 90th percentile action level (AL) for public water

supplies under the Lead

and Copper Rule.44-48 Because the general distribution of lead levels in

residences and buildings is

widely accepted to follow a lognormal pattern according to virtually all

published research, this

assumption will provide an example that is very biased towards the highest

lead occurrence.49-50 For

all modeling, temperature was set to 25 °C, and an ionic strength of 0.005 M

was assumed. Stability

constants for important chemical species of lead are not available for other

temperatures, but

temperature effects on the speciation of other metals have not been reported

to be so dramatic as to

affect the speciation by the orders of magnitude that would be necessary to

change the conclusions

of this investigation.

This background water is characteristic of many areas of the United States

where geological

and hydrological conditions create soft waters of low carbonate content

(dissolved inorganic carbon,

DIC) and ionic strength. These are the most susceptible waters for any

effect of fluoridation on lead

speciation, because strong competitive complexation by carbonate in higher

DIC waters reduces the

amount of lead complexed by fluoride, sulfate and chloride.4 Small

differences in hardness, silica concentration, or other major ions will have

mimimal impact on the aqueous lead speciation.

The concentration of aluminum in Table 7 represents a moderate to somewhat

high residual

carried over from coagulation with aluminum sulfate (alum), as commonly

occurs with surface water

treatment plants. The point of addition of fluoride varies widely, from

before coagulation to the clearwell or entry point to the distribution

system. The final effluent pH is often adjusted for corrosion control and

the Lead and Copper Rule, ofsetting acidity created by fluoride chemical

addition.

Table 7. Water quality parameters for speciation modeling

Species Concentration, mg L–1 Concentration, mol L–1

[siO2]T

† 5.0 8.3 × 10-5

[Pb2+]T

‡ 0.015 7.2 × 10-8

[F–]T

· 1.0 5.3 × 10-5

[CO2]T° as C 5.0 4.2 × 10-4

Ca2+ 5.0 1.2 × 10–4

Mg2+ 2.0 8.2 × 10–5

Na+ 10.0 4.4 × 10–4

Al3+ 0.20 7.4 × 10–6

Cl– 10.0 2.8 × 10–4

SO4

2– 5.0 5.2 × 10–5

†[siO2]T = total silicon(IV) concentration, expressed as silicon dioxide.

‡[Pb2+]T = total lead(II) concentration, all species.

+[F–]T

+ = total fluoride concentration = [F–] + [HF] + E n[MFn (q–n)].

°[CO2]T = DIC = [CO2(aq)] + [H2CO3] + [HCO3

–] + [CO3

2–] (dissolved inorganic carbon).

Mass-based concentration is expressed as C not CO2).

As noted previously, for these calculations we intentionally exceedingly

overestimated the

highest conceivable value for the equilibrium constant for equation (8),

using a value of 10000. The

summary plots (Figures 1-4) clearly show that hexafluorosilicate and

fluoride complexes are

minuscule contributors to lead(II) in a drinking water matrix. Figure 1

shows the computed relative

eBuffer intensity is usually represented by the symbol $, which we find to

be an unfortunate coincidence as it leads to confusion between this quantity

and cumulative stability constants, for which $ is often used

simultaneously. As a result, here we have broken with convention and use the

symbol B to stand for buffer intensity.

amounts of the associations of fluoride with hydrogen ion, dissolved silica,

aluminum, lead and the

group calcium+magnesium+sodium. Figures 2 and 3 show that Pb2+(aq) is the

dominant species

only at low pH, with the bicarbonate and carbonatolead(II) complexs already

beginning to dominate

soluble lead speciation by pH 7. By pH . 7.2, the hydroxolead(II) ion also

exceeds the free lead(II).As pH increases to -8.4, only -1% of the total

lead is the free aquated ion. In Figures 2 and 3, we see that the mono- and

difluorolead(II) complexes always account for less than 0.3% of the total

lead(II). Note that the species PbSiF6 0 is present at such low

concentrations that we would expect to find only one molecule of this

complex in more than 1000 liters of tap water at pH 6, which of course, far

exceeds the volume possible for water consumption and the human stomach.

Note the broken ordinate.

In Figure 4, we show the minor species, including the sulfato-, fluoro-, and

chloro-complexes

of lead for this hypothetical but realistic water. The carbonato-complexes

of lead(II) are much

stronger than the halo-complexes—as reflected by their stability constants,

which are 5-8 orders of

magnitude higher than those of the comparable halide complexes, combined

with the higher molar

concentration of total DIC ([CO2]T). Since the increase in lead complexation

by carbonate has been

shown elsewhere,26,30 we did not repeat the calculations here. Clearly, at

the higher [CO2]T

concentrations, the fluoro-complexes become even less significant. We note

that there is less than

one molecule of PbSiF6 per liter of water even with the extreme exaggeration

of the possible value

for the stability constant with the fluorosilicate anion.

The insignificance of any SiF6

2– can be logically determined another way. Even if the

formation constant for a hypothetical PbSiF6(aq) complex were ten times

higher than the strongest

complex found by Haque and Cyr,25 it would have a similar stability to PbF+.

Assuming all of the

fluoride present in drinking water were SiF6

2-, Figures 1-4 show that it would still be approximately

3 (pH 6) to more than 6 (pH 10) orders of magnitude lower than the soluble

lead level, which is

governed by the concentrations of other Lewis bases. Because complexation

with carbonate and

bicarbonate dominates aqueous lead speciation at drinking water pH, any

increased [CO2]T level

makes contribution of the fluoro-complexes to [PbII]T even less significant.

The bar graphs in Figures

3 and 4 clearly illustrates how free lead(II), hydroxo-, and

(bi)carbonato-complexes dominate the speciation of lead(II) at all drinking

water pH values while fluoro-complexes are always in the minority.

5. Can fluorosilicates or fluoride affect the pH of a finished water?

We have previously stated that naturally occurring buffers have a

significant impact on

drinking water chemistry. At this point, we will quantitatively illustrate

the magnitude of this impact.

Figure 5 shows the buffer intensity (capacity) B as a function of pH.e The

buffer intensity is a

quantitative description of a solution’s resistance to changes in pH upon

addition of acid or base and

is defined as (14):

B = –dCa/d(pH) = dCb/d(pH) (14)

where Ca is the formal concentration (formality) of added acid and Cb is the

formality of added base.

Because we are discussing infinitesimal (differential) quantities of added

acid or base, the effects are

equal and opposite for added acid versus added base. Note that the

derivative described in eqn 14 is the slope of a curve of Cb versus pH,

which is the inverse of a “titration curve” where base molecules are added

directly to a solution of acid so that the titer is zero and there is no

change in volume, only changes in concentration.

Virtually all potable waters contain some dissolved inorganic carbon,

represented here as

[CO2]T; therefore, the buffer intensity will be controlled by the

simultaneous conjugate acid-base

equilibria of the carbon dioxide-carbonic acid-bicarbonate-carbonate system,

unless it is very low

in concentration relative to orthophosphate or other weak acid/base systems.

Although not

conceptually difficult to understand, the derivation of a quantitative

definition of the buffer intensity

B for a given system can be cumbersome. We have previously written about the

importance of

buffer capacity and we and others have derived the formulae for a diprotic

species, such as carbonic

acid.26,51-53

B = (15) (ln 10)

Kw

[H% ]

% [H% ] % [CO2]T

$1[H% ] % 4$2[H% ]2 % $1$2[H% ]3

(1 % $1[H% ] % $2[H% ]2)2

where B has units of M (pH unit)–1 when all concentrations are expressed in

molarities. Upon inspection of (15), it can be seen that the buffer

intensity can readily be divided into contributions from [OH–], [H+], and

[CO2]T.

Figure 5 shows an illustration of a water with a [CO2]T of 5 mg L–1 as C and

3 mg L–1 as PO4

of an orthophosphate corrosion inhibitor at 25°C. It can clearly be seen

that much of the buffer

intensity is derived from the carbon dioxide-bicarbonate-carbonate system,

and that contribution gets

more and more significant as [CO2]T increases. For the example water used in

these calculations, the

minimum buffer intensity contributed by the [CO2]T and water is

approximately B = 0.25 mM (pH

unit)–1. The contribution of acid from undissociated SiF6

2– can again be proved negligible from the

following extreme example. Even if 10% of the total fluoride input were as

SiF6

2– at pH 7, the acid

input would be )Ca = 5 × 10–6 M. I . 0.005 M, T = 25 °C. Thus, we can

compute the change in pH

directly using the buffer intensity calculated above: )(pH) . –)Ca/B

= –0.0050 mM/[0.25 mM (pH

unit)–1] = –0.020 pH unit. This value is within the limits of a linear

approximation of buffer capacity.

Such a small effect on pH is analytically undetectable and inconsequential

with respect to other

sources of variability in factors affecting lead release from plumbing

materials. Masters and Coplan

assert without any field data or simple calculation, that slow dissociation

of silicofluoride in

distribution systems can increase acidity and increase lead release.2

However, to the contrary, the

concepts of chemical equilibria are well-established and measured

equilibrium constants are

sufficiently accurate and precise to show that fluoride and fluorosilicate

essentially do not affect the

solubility distribution of lead(II) species under potable water conditions.

Many water systems are

also compelled under the Lead and Copper Rule to conduct distribution system

pH monitoring, and

dangerous pH decreases from any cause would usually be uncovered and

treatment adjustments made.

6. Can additives be responsible for contaminants in the water supply?

Water treatment chemicals are subject to National Sanitation Foundation

specifications,

which require that additives contain a maximum allowable level (MAL) less

than or equal to 10%

of the maximum contaminant level (MCL) for any regulated contaminant in the

national primary

drinking water standards.54 It is worth reviewing how fluoridation chemicals

are made to see whether

there are any steps where contamination could occur. Most hexafluorosilicate

and fluorosilicic acid

are derived from the processing of phosphate rock by the fertilizer

industry.55 In this process, apatite

and fluoroapatite (which can be thought of as a blend of fluorite and

apatite for this purpose) are

decomposed with sulfuric acid. HF and SiF4 are removed as gases so that

there is little chance of lead

contamination from the crushed rock. The resulting 23% w/w hexafluorosilicic

acid is a strong acid

and quite corrosive, but no evidence has been put forth to suggest that this

additive has become contaminated prior to use. Moreover, testing either the

water at the plant or the stock fluoridating agent itself would also be

sufficient to rule this out.

7. What is required to have a valid sampling scheme for measuring lead(II)

intake from tap water? How do regulation-required samplings relate to

exposure?

The total lead in a first draw sample mostly reflects the nature of the

building plumbing

system. A one liter sample volume dominantly represents the metals picked up

in contact with the

last approximately 17 to 26 feet of plumbing material before the consumers’

tap, presuming “ ½-

inch” pipe of commonly-used materials. Comprehensive water sampling for

epidemiological and

other health effects studies for lead(II) is logistically complicated and

expensive; therefore, it is very

tempting to try to use available regulatory tap water monitoring data for

this purpose. The

temptation must be resisted, however as the monitoring program specified in

the United States

drinking water regulations is both statistically and physically invalid for

this purpose, and was never

intended to be an exposure assessment sampling program.44,48-50 The

regulatory targeting scheme is

intentionally biased towards reducing the highest lead exposures through

central water treatment. It does not capture the highest copper exposures.

It does not give any information on the levels of metals to which the

general population is exposed from old non-lead plumbing materials, or many

other corrosivity-related characteristics too numerous to list.

As previously noted, the vast preponderance of the lead(II) in nearly all

tap waters originates

from the plumbing materials located between the water distribution mains and

the end of the faucet

used by the consumer. Individuals consume water under innumerable

combinations of volumes of

water, interior plumbing system configurations and ages, and lengths of

stagnation of the water in

the plumbing between uses. Data reported from many tap water sampling

experiences throughout

the US and Europe indicate tap water lead levels tended to follow a

log-normal distribution, and

both within-site and between-site variability tended to be large relative to

the lead(II)

concentrations.49,50 Keeping this in mind, the American standard for lead in

drinking water was

crafted to focus on the lowering of lead(II) levels by central water

treatment for the plumbing

configurations most likely to represent nearly the worst cases for the most

vulnerable humans, i.e.,

infants, children, pregnant women (see Q6). Some attempts have been made to

define reasonable

statistical bases for comparing soluble metal release from parallel pipe

loops used for corrosion

control testing, and the required number and frequency of samples directly

relates to the intrinsic

variability of the metal release and the confidence levels one wishes to

place in the characterization

of the metal levels.56,57 After a cursory examination of the requirements

for a statistically valid

sampling program accounting for needed levels of predictive confidence

across all sources of

variability observed, one realizes that it would take literally hundreds or

thousands of samples at

great frequency from cities of all sizes to try to adequately characterize

tap water lead levels for even

a single uniformly applied national sampling protocol.

Obviously, the water chemistry at the point the distributed finished water

enters the domestic

or commercial building plumbing system plays a very significant role in

affecting lead release into

the water, but many other physical factors also

operate.26,30,33-36,49,50,56,58,59 The water at this point may

have undergone chemical changes during its passage through the distribution

system from the

treatment plant or well, and changes in treatment or changes in water

sources may also cause the

chemical characteristics of the water to change periodically, especially in

such important aspects as

pH, and concentrations of alkalinity, natural organic matter, oxidant

levels, and a variety of

potentially aggressive anions. Even the season may influence lead levels in

complicated ways, by

changes in ground temperature, or temperatures in buildings where pipes run

through basements,

unheated crawl spaces, concrete slabs, or nearby heating or air conditioning

ducts. A single snapshot

sampling event cannot capture this.

The drinking water literature is full of papers that show how difficult it

is to correlate lead

levels with any one or even a mix of several water quality parameters (even

when frequently sampled

and sophisticated statistics are applied60 ), and a complete discussion of

the matter is beyond the scope of this article. There may be countless other

physical or chemical quantities that may be statistically correlated with

lead(II) levels but nonetheless be totally unrelated mechanistically.

Clearly, aggregate measures such as a small number of first-draw or

fully-flushed water samples

taken infrequently from an intentionally biased relatively small pool of

sampling sites throughout

a water system cannot quantitatively and precisely predict the exposure of

any individuals to lead

from drinking water. To accurately determine lead(II) intake, sampling

schemes using diverters or

proportional sampling devices that capture a representative fraction of the

water actually drawn at

the faucet by the consumer seem to be the only feasible approach.30

Interestingly, the bibliographies

of the Masters and Coplan study most strongly asserting the adverse effects

of silicofluoride shows

only a single reference related to sampling of drinking water or the control

of lead or other metals

by water treatment, so the level of awareness in the design of the studies

and interpretation of the

data is highly questionable. By not measuring or statistically testing

numerous other water and

plumbing characteristics that could correlate with lead(II) levels with

equal to or greater statistical

significance than those relationships that were put forth, the studies of

Reference 2 are intentionally

biased towards what appears to be a preconceived conclusion. Even simple

analytes that are known

fIt is unclear from where these numbers originated. Reference 2 mentions

averaged 90th percentile values. We take

this to mean that 90th percentile values from two or more rounds of

regulatory testing were averaged. Given that the

normal distribution of monitoring data contains many non-detects, and the

number of samples varies somewhat with

system size so that the number and extent of values above the reported 90th

percentile is unpredictable, the meaning

of this table is very difficult to determine.

to affect lead mobility, such as pH or alkalinity, or analytes known to play

important dietary roles

in health, such as calcium, sodium or magnesium, were not reported to be

measured in their study,

so possible confounding variables are conspicuously excluded from

evaluation. Needless to say, tap water intake is highly variable with

beverage preferences of individuals, and that factor needs to also be taken

into account in any assessment of exposure and behavioral implications.

8. How is lead(II) concentration measured in blood?

The best methods that analytical chemistry has to offer are spectroscopic in

nature: AAS,

ICP-MS, or ICPS. Quick screening tests, as we have pointed out previously,4

have several

weaknesses in terms of precision and accuracy. What is appropriate to screen

children for exposure

prior to a more expensive and more elaborate test is not necessarily

appropriate for investigating overall lead exposure from drinking water.

Acceptable uncertainty in a yes-no screening test, for instance, would not

be appropriate when seeking a quantitative relationship.

Masters and Coplan did not give the total lead concentrations in the first

draw water samples,

so we cannot directly compare blood lead levels with water lead levels.2

They did give blood lead

levels divided up by those water systems where first draw samples were

divided by a cut-off of 15

µg L–1 of lead(II); see Table 5.f

Table 5. [PbII]blood (µg L–1) for fluoridation processes*

[PbII]water none NaF Na2SiF6 H2SiF6

<15 µg L–1 19.7 21.1 23.7 23.1

n = 86 31 6 26

>15 µg L–1 21.8 19 43.8 32.7

n = 29 8 1 25

*Taken from reference 2.

Because the sodium hexafluorosilicate data are based on one system with

[PbII]water > 15 µg L–1, it

is impossible to treat that value as significantly supporting any

hypothesis. Without some estimation

of the uncertainties of [PbII]blood, we also cannot be assured that 23 µg

L–1 is distinct from 33 µg L–1.

Reporting 3 significant digits in the blood lead(II) concentrations seems

suspect. We expect that the

numbers are probably good to about 10–15%. Masters and Coplan also failed to

include the

possibility of naturally occurring fluoride and silicates in the

unfluoridated water systems, which

gThe references cited here include a representative sampling over the last

two decades of the kinds of work that have

been done. These references are not intended to comprise a complete listing

or review of the studies in this area.would be necessary to substantiate

their thesis, as naturally occurring silica and fluoride should chemically

react to produce the same effects.

In the Masters and Coplan studies that most strongly assert to implicate

drinking water

fluoridation in lead neurotoxicity,1,2 there is no report of efforts to

obtain appropriate exposure data

and then attempt to correlate the consumed water quantity and quality from

an individual building

or house with the blood lead levels of individuals residing or spending

significant time there. There

is no indication that there is any connection between sampled taps and

sampled persons. In other

words, nothing indicates that a person living in the sampled house had his

blood drawn. Instead, the

authors rely on quartile divisions of both water lead levels and blood lead

levels. For there to be a

correlation between the lead(II) levels in blood and water, there must be a

link between the samples.

It is possible that the highest blood levels of lead(II) are closely linked

to other exposures, such as

paint, soil, or mine run-off. The authors appeared to notice and somewhat

acknowledge the

[unsurprising]lack of correlation of the water lead levels with blood lead

levels in one part of their

paper,2 even after postulating earlier in the paper that slow dissociation

of the fluoridation chemical

would cause more acidity and more lead release. To overcome this, they then

invoke a kind of

black-box “...biochemical effects...” to justify their conclusion. The

postulation of some

“...chemical effects that maintain lead in suspension...” is contradictory,

because that would be

reflected in higher lead levels at the tap using regulatory analytical

procedures. Another interesting

apples-versus-oranges comparison is made where they attempt to see if

silicofluoride could enhance

lead uptake for exposure to lead paint and dust by looking at old housing

(pre-1940 and post-1940)

combined with other data that includes 90th percentile lead levels. Of

course, the targeting scheme

for the sampling sites under the Lead and Copper Rule have only the most

indirect of relationships

to housing age, and houses can easily be remodeled, repainted and

re-landscaped making those differentiations very problematic.

Thus, when all of these methodological problems are coupled with the failure

to account

forthe quantitative level of fundamental chemical interactions, the

relationships posed between any

lead(II) speciation and water fluoridation become unjustifiable.

9. What are the routes of lead exposure besides drinking water?

Exposure routes have been the subject of multiple studies. The problem is

further

complicated by incomplete understanding of subacute toxicity and

dose-response.61 Much of the

exposure to lead occurs through dust, air-borne particulates, soil, paint,

ceramic glazes, and sundry

other sources, including drinking water.62-65 One of the special concerns

for drinking water is that the

lead(II) appears to be far more bioavailable.62 This is probably because

aqueous lead(II) is far more

likely to pass through mucous membranes than insoluble plumbous minerals.

However, there is some

evidence to suggest that even insoluble minerals can release lead(II) when

ingested under the right

conditions.66 A number of studiesg have concentrated on other factors

affecting bioavailability and

bioabsorption, including other nutrients, alcohol, cigarettes, water

hardness, plumbing, and

lifestyle.37,67-72 At least one study has also shown some data indicating

that lead associated with

orthophosphate (either ingested as particles or simultaneously ingested in

solution may be less

bioavailable through the intestinal system, because of the higher levels of

phosphate present in that

organ causing the formation of insoluble lead phosphate particles that would

not be readily absorbed.

This is very interesting from a drinking water perspective, because of the

widespread use of

orthophosphoric acid or orthophosphate-containing corrosion inhibitor

chemicals. The main

conclusion that can be drawn from these studies is that the biological

availability, absorption, and

accumulation of lead and its compounds depend on a wide variety of factors,

making this a very

complicated puzzle to solve. 39

Conclusions

Recent reports1,2 that purport to link certain water fluoridating agents,

such as

hexafluorosilicic acid and sodium hexafluorosilicate to human lead uptake

are inconsistent with

accepted scientific knowledge. The authors of those reports fail to identify

or account for these

inconsistencies, and mainly argue on the basis of speculation stated without

proof as fact. The

sampling scheme employed in the studies is entirely unrelated to any

credible statistically-based

study design to identify drinking water lead and fluoride exposure as a

significant source of blood

lead in the individuals. The authors use aggregated data unrelated in space

and time and then attempt

to selectively apply gross statistical techniques that do not include any of

thousands of other possible

water quality or exposure variables which could show similar levels of

correlation utterly by

accident. Many of the chemical assumptions are scientifically unjustified,

are contradicted by known

chemistry data and principles, and alternate explanations (such as multiple

routes of PbII exposure)

have not been satisfactorily addressed. The choice in water fluoridation

approach is often made for

economic, commercial or engineering reasons that may have a regional

component that could also be related to various community socio-economic

measures, and so should not be considered to be a purely independent

variable without investigation.

At present, the highly-promoted studies asserting enhanced lead uptake from

drinking water

and increased neruotoxicity still provide no credible evidence to suggest

that the common practice

of fluoridating drinking water has any untoward health impacts via effects

on lead(II) when done

properly under established guidelines so as to maintain total water quality.

Our conclusion supports

current EPA and PHS/CDC policies on water fluoridation.

References

1. R.D. Masters and M. Coplan. “Brain Biochemistry and the Violence

Epidemic: Toward a ‘Win-

Win’ Strategy for Reducing Crime” Super-Optimizing Examples: Across Public

Policy Problems, Stuart S. Nagel, ed., Nova Science Publishers, Inc, New

York (1999).

2. R.D. Masters and M. Coplan. “Water Treatment with Silicofluorides and

Lead Toxicity,”

Intern. J. Environmental Studies 56, 435-449 (1999).

3. Crime Times: Research Reviews and Information on Biological Causes of

Criminal, Violent, and Psychopathic Behavior 4(4), 1-2 (1998).

4. E.T. Urbansky and M.R. Schock, “Can Fluoridaton Affect Lead(II) in

Potable Water? Hexafluorosilicate and Fluoride Equilibria in Aqueous

Solution,” Intern. J. Environmental Studies, in press (2000).

5. US Department of Health and Human Services, US Public Health Service,

Centers for Disease Control and Prevention. Water Fluoridation Census 1992.

(July 1993). 6. US Department of Health and Human Services, Public Health

Service, Centers for Disease Control and Prevention. (a) Water Fluoridation:

A Manual for Engineers and Technicians. (September 1986); (b) Water

Fluoridation: A Manual for Water Plant Operators. (August 1993).

7. C.F. Baes Jr. and R.E. Mesmer. The Hydrolysis of Cations. Krieger

Publishing Co.: Malabar, Fla., 1976; Ch. 15.

8. R.H. Busey and R.E. Mesmer. “Ionization Equilibria of Silicic Acid and

Polysilicate Formation in Aqueous Sodium Chloride Solutions.” Inorg. Chem.

16(10), 2444-2450 (1977). 9. R.H. Busey, E. Schwartz, and R.E. Mesmer.

“Fluorosilicate Equilibria in Sodium Chloride Solutions from 0 to 60 °C.”

Inorg. Chem. 19(3), 758-761 (1980). 10. L. Ciavatta, M. Iulianno, and R.

Porto. “Fluorosilicate Equilibria in Acid Solutions.”

Polyhedron 7(18), 1773-1779 (1988).

11. A.S. Korobitsyn, G.N. Bogachov, and D.F. Rize. “Influence of Temperature

on the Rate of Decomposition of Crystalline Sodium Fluosilicate by Sodium

Carbonate Solutions.” J. Appl.

Chem. (USSR) 45(4), 736-740 (1972). Translated from Zh. Prikl. Khim. 45(4),

722-726 (1972).

12. R.E. Connick and R.E. Poulson. “19F Nuclear Magnetic Resonance of

Various Metal-Fluoride

Complexes in Aqueous Solution.” J. Phys. Chem. 63, 568-569 (1959). 13. E.L.

Mutterties and W.D. Phillips. “Structure and Exchange Processes in Some

Inorganic Fluorides by Nuclear Magnetic Resonance.” J. Am. Chem. Soc. 81,

1084-1088 (1959).

14. A.G. Kucheryaev, V.A. Lebedev, and I.M. Ovchinnikov. “Stability of the

Silicon

Hexafluoroanion in Hydrofluoric Acid, From 19F NMR Data.” J. Struct. Chem.

(USSR) 11(5),

858-859 (1970). Translated from Zh. Strukt. Khim. 11(5), 925-926 (1970).

15. P.M. Borodin and N.K. Zao. “Equilibria in the Li2SiF6–HClO4–H2O System

by the 19F Nuclear

Magnetic Resonance Method.” Russ. J. Inorg. Chem. 16(12), 1720-1722 (1971).

16. P.M. Borodin and N.K. Zao. “Study of Mobile Equilibria in the

HF–SiO2–H2O System by the

19F NMR Method.” Russ. J. Inorg. Chem. 17(7), 959-962 (1972).

17. P.M. Borodin, N.K. Zao, and N.S. Petrov. “Stability of the

Hexafluorosilicate Ion in Aqueous

Solutions of Fluorosilicic Acid System, from 19F NMR Data.” J. Struct. Chem.

(USSR) 14(4),

564-568 (1973). Translated from Zh. Strukt. Khim. 14(4), 605-610 (1973).

18. Gmelin Handbook of Inorganic Chemistry. 8th ed. Silicon.

Springer-Verlag: New York, 1995;

“H2SiF6.” Supplement vol. B8.

19. R.M. Smith and A.E. Martell. Critical Stability Constants: Inorganic

Complexes. Plenum Press: New York, 1976; vol. 4, pp. 96-103.

20. N.T. Crosby. “Equilibria of Fluorosilicate Solutions with Special

Reference to the Fluoridation

of Public Water Supplies.” J.Appl. Chem. 19, 100-103 (1969). See also

references therein. 21. V.N. Krylov, E.V. Komarov, and M.F. Pushlenkov.

“Stability of the SiF6 2– Ion in Aqueous Solutions in the Presence of Ions

of Various Valence.” Sov. Radiochem. 13(3), 445-449 (1971). Translated from

Radiokhim. 13(3), 430-434 (1971).

22. K.K. Worthington and M.M. Haring. “Methods for Determining the

Solubilities of Some

Fluosilicates.” Ind. & Eng. Chem. 3(1), 7-9 (1931).

23. V.N. Plakhotnik. “Kinetics and Mechanism of the Hydrolysis of Some

Acidofluoro-Complexes

in the Presence of Polarizing Cations.” Kinet. Catal. 16(6), 1211-1214

(1975). Translated from

Kinet. Katal. 16(6), 1395-1399 (1975).

24. V.N. Plakhotnik. “The Catalytic Influence of Li+ on the Alkaline

Hydrolysis of the Hexafluorosilicate

Anion.” Kinet. Catal. 14, 1417-1418 (1973).

25. R. Haque and N. Cyr. “Fluorine Magnetic Resonance Studies of

Metal-Hexafluorosilicate Ion-

Pairs in Solution.” Faraday Soc. Trans. 66, 1848-1853 (1970). 26. M.R.

Schock. “Internal Corrosion and Deposition Control.” Ch. 17. In Water

Quality and Treatment: A Handbook of Community Water Supplies, 5th ed,

American Water Works Association, McGraw-Hill, Inc., New York (1999).

27. A.E. Martell and Robert M. Smith Critical Stability Constants, First

Supplement. Plenum Press:

New York, 1982; vol. 5, pp. 415-417.

28. D.K. Nordstrom, L.N. Plummer, D. Langmuir, E. Busenberg, H.M. May, B.F.

Jones, and D.L. Parkhurst. “Revised Chemical Equilibrium Data for Major

Water-Mineral Reactions and Their Limitations.” In Chemical Modeling of

Aqueous Systems II. ACS Symposium Series 416, R. L. Bassett and D. C.

Melchior, eds., American Chemical Society, Washington, DC, 1990; Ch. 31.

29. P. Pitter. “Forms of Occurrence of Fluorine in Drinking Water.” Water

Res. 19(3) 281-284 (1985).

30. M.R. Schock, I. Wagner, and R.J. Oliphant. “Corrosion and Solubility of

Lead in Drinking Water.” In Internal Corrosion of Water Distribution

Systems, 2nd ed. American Water Works Association Research Foundation:

Denver, Colo., 1996; Ch. 4. 31. I.P.L. Coleman, J.A. Blair, and M.E.

Hilburn. “Effects of Dietary and Synthetic Chelating Agents on the

Intestinal Absorption of Lead.” Intern. J. Environmental Studies 18, 187-191

(1982).

32. W. D. Schecher and D.C. McAvoy. MINEQL+: A Chemical Equilibrium Modeling

System, Version 4.0 for Windows. Environmental Research Software: Hallowell,

Maine, 1998.

33. M.R. Schock. “Understanding Corrosion Control Strategies for Lead.” J.

Am. Water Works

Assn. vol., 88-100 (1989).

34. N.-H. Lin, A. Torrents, A.P. Davis, M. Zeinali, and F.A. Taylor. “Lead

Corrosion Control from

Lead, Copper-Lead Solder, and Brass Coupons in Drinking Water Employing Free

and

Combined Chlorine.” J. Environ. Sci. Health A32(4), 865-884 (1997).

35. M.R. Schock and M.C. Gardels. “Plumbosolvency Reduction by High pH and

Low

Carbonate—Solubility Relationships.” J. Am. Water Works Assn. 75(2), 87-91

(1983).

36. J.H. Colling, P.A.E. Whincup, and C.R. Hayes. “The Measurement of

Plumbosolvency

Propensity to Guide the Control of Lead in Tapwaters.” J. Inst. Water &

Environ. Management

(J. IWEM) 1(3), 263-269 (1987).

37. J. Sherlock, G. Smart, G.I. Forbes, M.R. Moore, W.J. Patterson, W.N.

Richards, and T.S. Wilson. “Assessment of Lead Intakes and Dose-Response

for a Population in Ayr Exposed to a Plumbosolvent Water Supply.” Human

Toxicol. 1, 115-122 (1982). 38. H.F. Thomas, P.C. Elwood, C. Toothill, and

M. Morton. “Blood and Water Lead in a Hard Water Area.” Lancet 1047-1048

(May 9, 1981).

39. W.L. Marcus. “Lead Health Effects in Drinking Water.” Toxicol. Ind.

Health 2(4), 363-407 (1986).

40. M.R. Schock. “Response of Lead Solubility to Dissolved Carbonate in

Drinking Water.” J. Am.

Water Works Assn. 72(12), 695-704 (1980).

41. M.R. Schock. “Response of Lead Solubility to Dissolved Carbonate in

Drinking Water.”

Erratum. J. Am. Water Works Assn. 73(3), 36 (news) (1981).

42. M.R. Schock and J.A. Clement. “Control of Lead and Copper with Non-Zinc

Orthophosphate.”

J. New Engl. Water Works Assn. 112(1) 20-242 (1998).

43. R. Gregory and P.J. Jackson. “Central Water Treatment to Reduce Lead

Solubility.” Proc.

AWWA Annual Conference (Dallas, TX, 1984).

44. Federal Register, “Maximum Contaminant Level Goals and National Primary

Drinking Water Regulations for Lead and Copper.” Final Rule, U. S.

Environmental Protection Agency, 40 CFR Parts 141 and 142, 56:110:26460

(June 7, 1991).

45. Federal Register, “Drinking Water Regulations; Maximum Contaminant Level

Goals and National Primary Drinking Water Regulations for Lead and Copper.”

Final Rule; correction, U. S. Environmental Protection Agency, 40 CFR Parts

141 and 142, 56:135:32112 (July 15, 1991).

46. Federal Register, “Drinking Water Regulations: Maximum Contaminant Level

Goals and National Primary Drinking Water Regulations for Lead and Copper.”

Final Rule; correcting amendments, U. S. Environmental Protection Agency, 40

CFR Parts 141 and 142, 57:125:28785 (June 29, 1992).

47. Federal Register, “Drinking Water Regulations; Maximum Contaminant Level

Goals and National Primary Drinking Water Regulations for Lead and Copper.”

Final Rule; technical corrections, U. S. Environmental Protection Agency, 40

CFR Parts 141 and 142, 59:125:33860 (June 30, 1994).

48. Federal Register, “National Primary Drinking Water Regulations for Lead

and Copper.” Final Rule, U. S. Environmental Protection Agency, 40 CFR Parts

141 and 142, 65:8:1949 (January 12, 2000).

49. R. Levin, M.R. Schock, and A. Marcus. “Exposure to Lead in U.S. Drinking

Water.” Twentythird

Annual Conference on Trace Substances in Environmental Health, Cincinnati,

Ohio, May 30–June 1 (1989).

50. M. R. Schock. “Causes of Temporal Variability of Lead in Domestic

Plumbing Systems.”

Environ. Monitoring Assessment 15, 59-82 (1990).

51. E.T. Urbansky and M.R. Schock. “Understanding, Deriving, and Computing

Buffer Capacity.”

Journal of Chemical Education, in press (2000).

52. M.R. Schock,”Reasons for Corrosion Control other than the Lead and

Copper Rule.” Journal

of the New England Water Works Association, 113(2), 128-150 (1999).

53. J.N. Butler. Ionic Equilibrium: Solubility and pH Calculations. Wiley

Interscience, New York (1998).

54. N.I. McClelland, D.A. Gregorka, and B.D. Carlton. “The Drinking-Water

Additives Program.”

Environ. Sci. Technol. 23(1), 14-18 (1989).

55. P.J. Chenier. Survey of Industrial Chemistry, 2nd rev. ed. Wiley-VCH:

New York, 1992; pp.

50-51.

56. AWWARF, “Lead Control Strategies,” AWWA Research Foundation and AWWA,

Denver, CO, (1990).

57. AWWARF, “Development of a Pipe Loop Protocol for Lead Control,” AWWA

Research Foundation, Denver, CO (1994).

58. A.D. Hulsman. “Particulate Lead in Water Supplies.” J. IWEM 4(1), 19-25

(1990).

59. J.H. Colling, B.T. Croll, and P.A.E. Whincup. “Plumbosolvency Effects

and Control in Hard

Waters.” J. IWEM 6(3) 259-268 (1992).

60. C.H. Neff, M.R. Schock and J.I. Marden. Relationships between Water

Quality and Corrosion

of Plumbing Materials in Buildings, U. S. Environmental Protection Agency,

EPA/600/S2-87/036, Risk Reduction Engineering Laboratory, Cincinnati, OH

(1987). 61. A.C. J.G. Wetmur, J.M. Moline, J.H. Godbold, S.M. Levin,

and P.J. Landrigan.

“Unraveling the Chronic Toxicity of Lead: An Essential Priority for

Environmental Health.”

Environ. Health Perspect. 104(1), 141-146 (1996).

62. G.M. Raab, D.P.H. Laxen, and M. Fulton. “Lead from Dust and Water as

Exposure Sources for

Children.” Environ. Geochem. & Health 9(3-4), 80-85 (1987).

63. C.D. Strehlow and D. Barltrop. “Temporal Trends in Urban and Rural Blood

Lead

Concentrations.” Environ. Geochem. & Health 9(3-4), 74-79 (1987).

64. J.L. Pirkle, R.B. Kaufmann, D.J. Brody, T. Hickman, E.W. Gunter, and

D.C. Paschal.

“Exposure of the U.S. Population to Lead, 1991–1994.” Environ. Health.

Perspect. 106(11)

(1998).

65. M.B. Rabinowitz and D.C. Bellinger. “Soil Lead–Blood Lead Relationship

Among Boston

Children.” Bull. Environ. Contam. Toxicol. 41, 791-797 (1988).

66. U.G. Gasser, W.J. Walker, R.A. Dahlgren, R.S. Borch, and R.G. Burau.

“Lead Release from

Smelter and Mine Waste Impacted Materials under Simulated Gastric Conditions

and Relation

to Speciation.” Environ. Sci. Technol. 30(3), 761-769 (1996).

67. C.R. Angle and K.L. Stelmak. “Lead and Iron Deficiency.” Trace

Substances in Environmental

Health – IX. Proceedings of the University of Missouri’s 9th Annual

Conference on Trace Substances in Environmental Health. D.D. Hemphill, ed.

(Columbia, MO, June 10-12, 1975); pp. 377-386.

68. D. Barltrop and H.E. Khoo. “Nutritional Determinants ofLead Absorption.”

Trace Substances in Environmental Health – IX. Proceedings of the University

of Missouri’s 9th Annual Conference on Trace Substances in Environmental

Health. D.D. Hemphill, ed. (Columbia, Mo., June 10-12, 1975); pp. 369-376.

69. M. Rabinowitz, G. Wetherill, and J. Kopple. “Absorption, Storage, and

Excretion of Lead by Normal Humans.” Trace Substances in Environmental

Health – IX. Proceedings of the University of Missouri’s 9th Annual

Conference on Trace Substances in Environmental Health. D.D. Hemphill, ed.

(Columbia, Mo., June 10-12, 1975); pp. 361-368.

70. S.J. Pocock, A.G. Shaper, M. Walker, C.J. Wale, B. Clayton, T. Delves,

R.F. Lacey, R.F. Packham, and P. Powell. “Effects of Tap Water Lead, Water

Hardness, Alcohol, and Cigarettes on Blood Lead Concentrations.” J.

Epidemiol. & Community Health 37, 1-7 (1983). 71. H.M. James, M.E. Hilburn,

and J.A. Blair. “Effects of Meals and Meal Times on Uptake of Lead from the

Gastrointestinal Tract in Humans.” Human Toxicol. 4, 401-407 (1985).

72. M. Berode, V. Wietlisbach, M. Rickenbach, and M.P. Guillemin. “Lifestyle

and Environmental

Factors as Determinants of Blood Lead Levels in a Swiss Population.”

Environ. Res. 55, 1-17 (1991).

73. P. Zhang, J.A. Ryan and J. Yang, “In Vitro Soil Pb Solubility in the

Presence of Hydroxyapatite.” Environmental Science and Technology, 32(18),

2763-2768 (1998).

pH

6 7 8 9 10

log [mol/L]

· 40.0

· 35.0

· 30.0

· 25.0

· 20.0

· 15.0

· 10.0

· 5.0

free F-ion

F- associated with H+

F- complexed with SiO2

F- complexed with Al

F- complexed with Na, Ca & Mg

F- complexed with Pb

Figure 1. Illustration of concentration of fluoride species bound to

different metal groups for hypothetical low-DIC water, assuming 15 µg L-1 Pb

and background ion concentrations given in Table 7 of the text.

pH

6 7 8 9 10

log [mol/L]

· 13.0

· 12.0

· 11.0

· 10.0

· 9.0

· 8.0

· 7.0

· 6.0

15 µg/L as Pb

free Pb2+ ion

in hydroxides

in carbonates

in fluorides

in sulfates

in chlorides

Figure 2. Illustration of concentration of soluble lead bound to different

ligand groups for hypothetical low-DIC water, assuming 15 µg L-1 Pb and

background ion concentrations given in Table 7 of the text.

pH

6 7 8 9 10

% in Form

0.1

1

10

100

free Pb2+ ion

in hydroxides

in carbonates

in fluorides

in sulfates

in chlorides

Figure 3. Illustration of fractions of soluble lead bound to different

ligand groups for hypothetical low-DIC water, assuming 15 µg L-1 Pb and

background ion concentrations given in Table 7 of the text. Note logarithmic

scale for “% in Form..”

pH

6 7 8 9 10

log [mol/L]

· 50.0

· 48.0

· 46.0

· 44.0

· 30.0

· 28.0

· 26.0

· 24.0

· 22.0

· 20.0

· 18.0

· 16.0

· 14.0

· 12.0

· 10.0

· 8.0

· 6.0

· 4.0

PbF+

PbF2°

PbSiF6°

PbSO4°

Pb(SO4)2

2-

PbCl+

PbCl2°

PbCl3

-

PbCl4

2-

Approximately 1 PbSiF6 molecule in 1 L

total ingested volume

PbSiF6°

Figure 4. Minor lead species distribution in hypothetical water described in

Table 7. Computations were done for 25°C, I=0.001. PbSiF6° complex was

included in the model, assuming of log b = 5.

pH

4 5 6 7 8 9 10

Buffer Intensity, (mol/L)/pH unit

10-8

10-7

10-6

10-5

10-4

10-3

Buffer Intensity: H2O

Buffer Intensity: 3 mg/L PO4

Buffer Intensity: 5 mg/L DIC as C

Total Buffer Intensity

Figure 5. Components of buffer intensity for a hypothetical water with DIC =

5 mg L-1 as C (4.16 x 10-4 M), and orthophosphate at 3 mg L-1 as PO4 (3.16 x

10-5 M) at 25°C.

 

 

 

 

 

--

 

 

Version: 7.0.308 / Virus Database: 266.9.7 - Release 12/04/2005