# Physical chemical effects of zinc on in vitro enamel demineralization

## Physical chemical effects of zinc on in vitro enamel demineralization

Journal of Dentistry, 2014-09-01, Volume 42, Issue 9, Pages 1096-1104, Copyright © 2014 The Authors

## Objectives

Zinc salts are formulated into oral health products as antibacterial agents, yet their interaction with enamel is not clearly understood. The aim was to investigate the effect of zinc concentration [Zn 2+ ] on the in vitro demineralization of enamel during exposure to caries-simulating conditions. Furthermore, the possible mechanism of zinc’s action for reducing demineralization was determined.

## Methods

Enamel blocks and synthetic hydroxyapatite (HAp) were demineralized in a range of zinc-containing acidic solutions (0–3565 ppm [Zn 2+ ]) at pH 4.0 and 37 °C. Inductively coupled-plasma optical emission spectroscopy (ICP-OES) was used to measure ion release into solution. Enamel blocks were analysed by Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy (ATR-FTIR), and HAp by X-ray diffraction (XRD) and neutron diffraction (ND).

## Results

ICP-OES analysis of the acidic solutions showed a decrease in [Ca 2+ ] and [PO 4 3− ] release with increasing [Zn 2+ ]. FTIR revealed a α-hopeite (α-Zn 3 (PO 4 ) 2 .4H 2 O)-like phase on the enamel surfaces at >107 ppm [Zn 2+ ]. XRD and ND analysis confirmed a zinc-phosphate phase present alongside the HAp.

## Conclusions

This study confirms that zinc reduces enamel demineralization. Under the conditions studied, zinc acts predominantly on enamel surfaces at PO 4 3− sites in the HAp lattice to possibly form an α-hopeite-like phase.

## Clinical significance

These results have a significant implication on the understanding of the fundamental chemistry of zinc in toothpastes and demonstrate its therapeutic potential in preventing tooth mineral loss.

## Introduction

Despite the considerable progress in preventive dentistry, caries still remains a major problem in both adults and children in Westernized countries. As a result, enamel has been studied extensively because of the economic and social costs of restorative dental treatment attributable to the disease. Caries involves the demineralization of a calcium-deficient carbonated hydroxyapatite, the main component of tooth enamel, due to acids produced over time by the interactions between microorganisms in plaque and dietary fermentable carbohydrates. The disease develops beneath the bacteria-rich plaque that can build up at sites where the saliva turnover is low or where oral surfaces cannot be adequately cleaned.

Zinc salts are formulated into oral health products to reduce malodor, and control plaque and calculus formation. The antibacterial actions of zinc salts are due to their ability to inhibit bacterial adhesion, metabolic activity and growth. Additionally, zinc is known to inhibit the crystal-growth of hydroxyapatite (HAp) and its reported precursors’ dicalcium phosphate dihydrate (DCPD) and octacalcium phosphate (OCP). This ability of zinc to modify the crystal-growth pathways of calcium phosphates has been exploited to control calculus formation. A mechanism of adsorption or incorporation of zinc ions into the apatite lattice structure has been proposed. While the effects of zinc as an anti-plaque and calculus agent have been investigated extensively, its interaction with enamel and its putative role in demineralization and remineralization is considerably less well understood. In the early sixties, Brudevold et al., reported that zinc reduces the solubility of enamel, but were not clear on the fundamental chemistry of this mechanism.

The aim of this in vitro study was to investigate the effect of zinc ion concentration [Zn 2+ ] on the demineralization of dental enamel and synthetic hydroxyapatite (Ca 10 (PO 4 ) 6 (OH) 2 , HAp), during exposure to acidic solutions simulating caries (pH 4.0). Furthermore, the possible mechanistic action of zinc with enamel and HAp was determined under the conditions used. Synthetically pure HAp was also investigated due to the significant variability that exists naturally in the composition of human enamel. Inductively coupled plasma-optical emission spectroscopy (ICP-OES) was used to quantitatively measure the concentration of calcium and phosphorus ions released into solution. The treated HAp powders were characterized by X-ray diffraction (XRD) and neutron diffraction (ND). Furthermore, the surfaces of the enamel blocks were analysed using Attenuated Total Reflectance-Fourier Transform Infrared Spectroscopy (ATR-FTIR) to obtain structural information on the mineral phases formed on enamel surfaces as a function of [Zn 2+ ] in the acidic conditions.

## Preparation of demineralizing solutions

A series of acidic solutions were made from analytical grade reagents. A 10.0 L stock of 0.1 mol/L acetic acid (AnalaR NORMAPUR, VWR International, France) was prepared with deionized water and buffered to pH 4.0. Subsequently, this was divided into ten 1.0 L flasks. Zinc acetate (Zn(O 2 CCH 3 ) 2 , Sigma–Aldrich, Dorset) was added to the acid reservoirs to prepare solutions containing: 0, 9, 36, 107, 356, 1782, 3565 mg/L [Zn 2+ ] respectively. This concentration range was used to mimic those found in dentifrices, plaque fluid and saliva. The lower concentration range bracketed those typical of saliva at some considerable time after application, whereas the upper concentration range might be typical of immediate post-application.

## Preparation of enamel blocks

Anonymized caries-free permanent molars extracted for orthodontic purposes were randomly selected. Ethical approval was obtained from Queen Mary Research Ethics Committee (QMREC 2011/99). Enamel blocks (∼5 mm × 5 mm) with a thickness of ∼2 mm were cut from each tooth (3 teeth in total, n = 3) using an annular diamond blade (Microslice 2, Malvern Instruments, UK) and dentine was polished off using a P600 grit silicon carbide paper.

## Enamel blocks

6× enamel blocks were placed in a container with 50 ml of demineralizing solution (0 mg/L [Zn 2+ ]) at 37 °C in a shaking incubator (KS 4000i control, IKA, UK) at 60 rpm for 24 h. Subsequently, each sample was immersed in 50 ml of a [Zn 2+ ]-containing demineralizing solution (0, 36, 107, 356, 1782 and 3565 mg/L) at 37 °C for 96 h.

## HAp powder

1 g of HAp powder (Biotal-R HAP, surface area 20 m 2 /g) was immersed directly in 50 ml of a [Zn 2+ ]-containing demineralizing solution (9, 36, 107, 356, 1782 and 3565 mg/L) at 37 °C for 9 h. The solution was centrifuged to remove any HAp powder. The resulting suspensions were dried at 37 °C for ∼48 h.

After reaction with enamel and HAp powders, the [Zn 2+ ]-containing demineralizing solutions were prepared for ICP-OES analysis by 1/4, 1/10, 1/20 and 1/200 dilutions with the original demineralizing acid stock solution. These dilutions were used to analyse the Ca, P and Zn levels at a detectable range. Each measurement was replicated three times (n = 3). Calibration solutions were also prepared with the same original demineralizing acid stock solution to give the same background Na levels and similar ionic strengths to the test samples. All test samples and calibration solutions were prepared with an acidification of 1% (0.1 ml of 69% nitric acid in 10 ml) for analysis using ICP-OES (ICP; Varian Vista-PRO, Varian Ltd., Oxford, UK). Calcium and phosphorus standards were used in the concentration range 1–50 mg/L, whereas zinc standards were in the 0.1–20 mg/L range.

The Wilcoxon Signed-Rank Test was carried out, in turn for Ca and P release for both HAp and enamel individually, with the zinc concentrations [Zn 2+ ] as the independent variables. The null hypothesis was that there is no difference in the Ca and/or P release in HAp and/or enamel for all combinations of [Zn 2+ ].

## X-ray diffraction

The zinc treated HAp powders (0–3565 mg/L) were analysed by XRD (X’Pert PRO MPD, PANalytical, Cambridge, UK; 40 kV/40 mA, Cu Kα). Data was collected at room temperature with a 0.033° 2 θ step size and a count rate of 99.6 s step −1 , from 2 θ values of 10° to 60°.

## Neutron diffraction

For the control (0 mg/L [Zn 2+ ]) and 3565 mg/L [Zn 2+ ] treated HAp, neutron powder diffraction data were collected on the Near and InterMediate Range Order Diffractometer (NIMROD), at the ISIS neutron Facility (Rutherford Appleton Laboratory, Didcot, Oxon, United Kingdom). Approximately 1 g of each sample was loaded into a thin-walled titanium-zirconium alloy container (2 mm thickness) and data were collected at room temperature for ∼2 h. The time-of-flight neutron diffraction data were collected with a wavelength range of 0.5–14 Å, which gave a continual working Q range of 0.02–50 Å −1 . The beam size used was 30 mm × 30 mm.

## ATR-FTIR analysis of enamel blocks

The enamel blocks were analysed by ATR-FTIR (Spectrum GX, Perkin-Elmer, Waltham, MA, USA) without any further sample preparation. This enabled direct analysis of zinc treated surfaces without the bulk of the samples contributing to the spectra.

## Results

Fig. 1 a and b shows the [Ca 2+ ] and [PO 4 3− ] release into the acidic solution following demineralization of enamel and HAp samples respectively. For enamel, the Wilcoxon Signed-Rank Test results showed no statistical significance in the reduction of calcium release at p ≤ 0.05 when comparing the 0 ppm [Zn 2+ ] with the 36 ppm [Zn 2+ ]. Thus the null hypothesis was not rejected. However, there was a statistical significance between the 0 ppm [Zn 2+ ] and all the remaining [Zn 2+ ], as well as every other combination of two [Zn 2+ ] at p ≤ 0.05. This indicates that there is a statistically significant reduction in the calcium release from enamel on exposure to [Zn 2+ ] >36 ppm, and therefore the null hypothesis was rejected.

(a) ICP-OES measurements showing the Ca and P detected for enamel blocks immersed in a range of [Zn 2+ ]-containing acidic solutions (96 h). The dashed lines show the Ca and P release for the control enamel (0 mg/L [Zn 2+ ]). The error bars show the Standard Error of the Mean (SEM) ( n = 3). (b) ICP-OES measurements showing the Ca and P detected for HAp immersed in a range of [Zn 2+ ]-containing acidic solutions (96 h). The dashed lines show the Ca and P release for the control HAp sample (0 mg/L [Zn 2+ ]). The error bars show the Standard Error of the Mean (SEM) ( n = 3). (c) Ca/P ratio in solution as a function of [Zn 2+ ] in the acidic solutions shown for the enamel blocks. The dashed lines show the Ca/P ratio for the control enamel (0 mg/L [Zn 2+ ]). (d) Ca/P ratio in solution as a function of [Zn 2+ ] in the acidic solutions shown for HAp samples. The dashed lines show the Ca/P ratio for the control HAp (0 mg/L [Zn 2+ ]). (e) Compares the [Zn 2+ ] uptake from solution with the [PO 4 3− ] released into solution for HAp samples. The error bars show the Standard Error of the Mean (SEM) ( n = 3).

For the Ca release in HAp, and P release in both enamel and HAp, there was a statistically significant difference for all combinations of [Zn 2+ ] at p ≤ 0.05. Therefore, there is a statistically significant reduction in Ca release from HAp, and P release from both HAp and enamel for [Zn 2+ ] ≥36 ppm. Thus the null hypothesis was rejected.

The [PO 4 3− ] release was considerably lower than the [Ca 2+ ] release. The [Zn 2+ ] is plotted on a log-scale and the concentration range of [Ca 2+ ] and [PO 4 3− ] was plotted between 10 and 1000 mg/L for HAp, whereas for enamel it was plotted between 0.01 and 10 mg/L/mm 2 . This difference is predominantly due to the difference in surface area between enamel and HAp used.

Fig. 1 c and d shows the Ca/P ratio in the acidic solution as a function of log [Zn 2+ ] for enamel and HAp samples respectively. The Ca/P ratio rises with increases in [Zn 2+ ] in the acidic solution. This indicates that the solution is calcium-rich with respect to phosphorus, and therefore less PO 4 3− groups are released from the apatite structure with increasing [Zn 2+ ].

Fig. 1 e compares the [Zn 2+ ] uptake from the solution by the HAp powder, and the [PO 4 3− ] released into solution as a function of increasing [Zn 2+ ] in the acidic solutions. The graph illustrates that as there is an increase in zinc removed from the aqueous solution by the apatite, there is a bigger reduction in [PO 4 3− ] detected in the solution.

Fig. 2 a displays the XRD patterns for HAp powders treated with acidic solutions containing a range of [Zn 2+ ]. All the samples show well-resolved diffraction patterns for HAp (typical apatite peaks appear at ∼2.8–2.6, 3.4 and 8 Å). Above 356 mg/L [Zn 2+ ], new diffraction peaks appear at ∼9.1 and 4.5 Å ( Fig. 2 b) which corresponds with the two strongest peaks in the α-hopeite reference. The third strongest peak for α-hopeite is at 2.86 Å; however, this peak was not visible in the [Zn 2+ ] treated HAp due to overlap with the apatite peaks. Fig. 3 shows the neutron diffraction data for the HAp samples treated with acidic solutions containing no [Zn 2+ ] (0 mg/L) and a high [Zn 2+ ] (3565 mg/L). The zinc-treated sample (3565 mg/L [Zn 2+ ]) possesses an additional peak at ∼9 Å which corresponds to the (0 2 0) reflection in the XRD pattern of the α-hopeite reference.

(a) The XRD patterns for HAp immersed in a range of [Zn 2+ ]-containing acidic solutions (0–3565 mg/L). A XRD pattern for α-hopeite is shown as a reference. The * show new diffraction peaks that are absent in the control sample. (b) The XRD patterns for HAp immersed in a range of [Zn 2+ ]-containing acidic solutions (0, 1782 and 3565 mg/L) in the region 4–10 Å. A XRD pattern for α-hopeite is shown as a reference. The * show new diffraction peaks that are absent in the control sample.
The neutron diffraction analysis for HAp powder immersed in acidic solutions containing [Zn 2+ ] (0 and 3565 mg/L). The * represents an additional peak in 3565 mg/L [Zn 2+ ] sample that is not present in the control sample (0 mg/L [Zn 2+ ]).

Fig. 4 a–c shows the ATR-FTIR spectra of enamel blocks treated in a range of [Zn 2+ ]-containing acidic solutions. In Fig. 4 a, the apatite bands (963, 1026 and 1087 cm −1 ) characteristic for PO 4 3− groups are progressively distorted with increasing [Zn 2+ ] in the acidic solutions, indicating that there is a new mineral phase forming on enamel surfaces that is less apatite-like ( Fig. 4 a). The bands observed at ∼876, 1413, 1455 and 1549 cm −1 correspond to the CO 3 2− groups in enamel ( Fig. 4 a and b). At 356 mg/L [Zn 2+ ] and above, the absorption bands show characteristic features for a mineral phase that resembles α-hopeite (α-Zn 3 (PO 4 ) 2 ·4H 2 O). Fig. 4 b shows the broad band at 1640 cm −1 grows in sharpness and intensity whereas Fig. 4 c shows the band at ∼3400 cm −1 becomes shifted to a lower wavenumber centred around 3300 cm −1 .

(a) The ATR-FTIR spectra of enamel blocks immersed in a range of [Zn 2+ ]-containing acidic solutions (0, 107, 356 and 3565 mg/L) in the region 500–1300 cm −1 . A spectra for α-hopeite is shown as a reference. . (b) The ATR-FTIR spectra of enamel blocks immersed in a range of [Zn 2+ ]-containing acidic solutions (0, 356 and 3565 mg/L) in the region 1200–2000 cm −1 . A spectra for α-hopeite is shown as a reference. . (c) The ATR-FTIR spectra of enamel blocks immersed in a range of [Zn 2+ ]-containing acidic solutions (0, 356, 1782 and 3565 mg/L) in the region 2000–4000 cm −1 . A spectra for α-hopeite is shown as a reference. The dashed line highlights the shift in band from ∼3500 cm −1 (in control sample) to ∼3300 cm −1 in the treated samples.

## Discussion

The present study investigates the physical chemical effects of zinc ion concentrations on the in vitro demineralization of enamel, during exposure to caries-simulating conditions, where the system was buffered at pH 4.0 and the temperature was maintained at 37 °C. At a fundamental level, the dissolution mechanism of biological apatite mineral is highly complex, not only due to gradients existing with depth in solubility and porosity, but also due to the available surface area for reaction. Therefore, the demineralization of enamel is not a one-dimensional process, but one that involves simultaneous processes or steps that is not limited to the diffusion of H + from bulk solution to the solid/liquid interface, adsorption of ions or chemical species onto the apatite surface, desorption of products (ions of calcium and orthophosphate) from the crystal surface and their diffusion into the bulk solution.

Although the in vitro model described in the present study does not fully replicate the intricacies that exist in the oral environment, it does provide a strategic approach for the strict chemical control of individual variables of a multi-factorial disease. Quantitative analysis of ions released into solution following demineralization of samples confirmed that zinc reduces the rate of demineralization as a function of concentration ( Fig. 1 a and b). An overall decrease in the [Ca 2+ ] and [PO 4 3− ] in solution was observed with increasing [Zn 2+ ], whereby a greater reduction was measured at higher [Zn 2+ ] (≥1782 mg/L). Further, the [PO 4 3− ] release was considerably lower than the [Ca 2+ ] release, particularly at 1782 mg/L [Zn 2+ ] and above, and this was more substantial in the HAp powder. Additionally the Ca/P ratio in solution increased as a function of [Zn 2+ ] i.e. the solution was more phosphorus-deficient with respect to the apatite stoichiometry (1.67) ( Fig. 1 c and d). These findings are indicative that zinc interacts predominantly at the PO 4 3− sites on the apatite structure during its mechanism in reducing demineralization, which further suppresses the [PO 4 3− ] ion release. This is also supported by the relationship observed between the [PO 4 3− ] ion released from the HAp structure and the amount of [Zn 2+ ] taken up from solution ( Fig. 1 e). With increasing [Zn 2+ ] in solution, there is a rise in the amount of zinc taken up by the apatite, which corresponded with the substantial decrease in the [PO 4 3− ] released. Thus, it seems evident that the mechanistic behaviour of zinc is dependent on its concentration and the availability of PO 4 3− binding sites.

The XRD analysis of the zinc treated HAp confirmed the formation of a zinc phosphate species above 356 mg/L [Zn 2+ ] ( Fig. 2 a and b). Previous studies found the concentration of the lattice PO 4 3− ions available on HAp surfaces as 3.02 ions/nm 2 , and show that most of the surface is covered by oxygen ions of the orthophosphate groups. In this study, the HAp material has a larger specific surface area (20 m 2 /g) than the enamel blocks and it is therefore less limited by binding sites available to the zinc ions on the apatite surface. Earlier studies have suggested that the amount of zinc adsorbed is directly proportional to the HAp surface area, whereby a Langmuir-type adsorption of zinc was observed exhibiting monolayer coverage. Furthermore, data from mechanistic studies show that the surface adsorption of zinc is also concentration-dependent.

The neutron diffraction patterns from the HAp sample treated with 3565 mg/L [Zn 2+ ] also showed the formation of zinc phosphate, which maybe present within the apatite, or in amongst the apatite crystals ( Fig. 3 ). Neutron diffraction data provides similar crystallographic information as X-ray diffraction. It is useful to compare diffraction patterns from both sources since neutrons can probe all atoms in a large sample due to their weak interaction with matter, and therefore unlike X-rays can penetrate the entire bulk of the sample eliminating surface effects. Thus the neutron diffraction results suggest that zinc is not only forming a zinc phosphate species at the HAp surface, but could also potentially diffuse into the bulk apatite lattice structure.

ATR-FTIR, a vibrational spectroscopic technique, was used with a view to obtain structural information about the physical and chemical changes taking place on the enamel surface (∼5–10 μm) during demineralization in the presence of [Zn 2+ ]. Above 107 mg/L [Zn 2+ ], the apatite bands characteristic for PO 4 3− groups (963, 1026 and 1087 cm −1 ) become progressively distorted with increasing [Zn 2+ ] in the acidic solutions, indicating the formation of a new mineral phase on the enamel surfaces which is less apatite-like ( Fig. 4 a). In the control sample, the broad band at 1640 cm −1 ( Fig. 4 b) is characteristic for absorbed water and the broad band at ∼3400 cm −1 ( Fig. 4 c) corresponds to OH stretch in the apatite structure. With increasing [Zn 2+ ] above 107 mg/L, the broad band at 1640 cm −1 grows in sharpness and intensity whereas the band at ∼3400 cm −1 becomes shifted to a lower wavenumber centred around 3300 cm −1 . These bands have been previously reported for α-hopeite (α-Zn 3 (PO 4 ) 2 ·4H 2 O), in which the largest peak around 1640 cm −1 corresponds to the internal bending vibration of water molecules, while the broad, very strong band centred around 3300 cm −1 represents OH stretching. Also above 356 mg/L [Zn 2+ ], a shoulder at ∼1068 cm −1 and 1148 cm −1 appears that are characteristic absorptions due to the PO 4 3− group in α-hopeite ( Fig. 4 a).

The structure of α-hopeite consists of ZnO 2 (H 2 O) 4 octahedra, ZnO 4 tetrahedra, and PO 4 tetrahedra, none of which are regular; these polyhedra share corners and edges. Previous studies have reported the growth of crystals of α-hopeite on the outermost surfaces of zinc-containing dental cements, which is in direct contact to enamel and dentine.

Many studies have demonstrated that HAp can bind to a variety of divalent metal cations and has a high capacity for the removal of Zn 2+ and other metal cations in aqueous solutions. One could also perceive it as metal cations bind to surface sites on HAp. Whichever way round it may be, the uptake of zinc from aqueous solution to coexist on the apatite surface may occur non-exclusively by three main processes: (1) surface adsorption or complexation, which is limited to the accumulation of sorbate on the external surface of the apatite; (2) absorption, ion exchange or diffusion into the solid; and (3) dissolution of HAp and precipitation of metal phosphates or coprecipitation (substitution of Ca in HAp by metals during recrystallization). The interaction of zinc with HAp/enamel is very likely to occur by more than one mechanism, and would most certainly vary depending on the pH conditions of reaction and saturation levels with respect to HAp of the system used.

3Zn2++2H2PO4+4H2OZn3(PO4)24H2O+4H+ $3{\text{Zn}}^{2+}+2{\text{H}}_{2}\text{P}{\text{O}}_{4}^{-}+4{\text{H}}_{2}\text{O}\to {\text{Zn}}_{3}{\left({\text{PO}}_{4}\right)}_{2}\cdot 4{\text{H}}_{2}\text{O}+4{\text{H}}^{+}$

In the context of the present study, the low pH condition releases dissolved calcium and phosphate ions into the aqueous solution, however in the presence of zinc ions, the precipitation of a hopeite-like phase is promoted on enamel surfaces, which in turn suppresses the release of further phosphate ions from the apatite structure (Eq. (1) ). This dissolution/re-precipitation mechanism would also explain the reduction in free [PO 4 3− ] measured in solution with increasing zinc ions. Nevertheless, it does appear that with increasing [Zn 2+ ] in the acid solutions, the driving force for the formation of a zinc phosphate species is much greater than the dissolution of apatite, and reduction in demineralization is observed as a function of [Zn 2+ ]. The results indicate that the formation of a hopeite-like phase is very dependent on pH and a high zinc concentration.

The adsorption of zinc and formation of a complex at the apatite surfaces is also a possibility in zinc’s mechanistic interaction with enamel. Fig. 5 is a schematic model illustrating the possible coordination geometry of Zn 2+ at the exposed PO 4 3− sites on enamel surfaces, in which it is more energetically favourable for zinc to adopt a tetrahedral configuration. The formation of these zinc adsorption complexes would be dependent on the zinc concentration and the surface area for reaction i.e. the availability of PO 4 3− binding sites on the apatite surfaces. As the PO 4 3− binding sites become saturated with zinc ions, there is a significant reduction in tooth mineral loss. In principle, the phosphate ion is probably the most important component of HAp both inside and outside the lattice. Its state of protonation significantly affects the stability and/or dissolution of the crystal. With a decrease in pH, some of the PO 4 3− ions, most likely those situated at the accessible surfaces of the crystals, becomes protonated to form H 2 PO 4 , thus lowering the solubility of the mineral.

The possible coordination of Zn 2+ at the PO 4 3− sites on the enamel surface.

These findings are clinically significant, particularly when considering the oral disposition of zinc following application from, for example, mouth-rinses and toothpastes. Some zinc is present naturally in both plaque and saliva, at ‘background’ concentrations. Following application, zinc is cleared from saliva bi-modally, with relatively high post-application concentrations falling rapidly over the first hour. Thereafter, lower concentrations, still significantly elevated when compared to baseline, may persist for many hours. A similar trend is seen in plaque. Here, elevated concentrations can persist for at least 12 h after application. However, in order to influence enamel demineralization under cariogenic conditions, zinc must be available in the plaque-fluid at a concentration sufficient to reduce or inhibit tooth mineral loss. Pharmacokinetic data relating to zinc concentrations in plaque-fluid following application are apparently almost absent from the literature, with only one reported value (15.2 ppm) for plaque-fluid measured 1 h after use of a mouth-rinse containing zinc citrate at 0.5%. This was substantially higher than the background value (0.2 ppm), and the data from the present study suggest that the increased zinc concentration over background would effect a reduction in demineralization.

However, it is also likely that higher plaque-fluid zinc concentrations than this 1 h value might occur. While the mouthwash contained 0.5% zinc citrate, toothpastes containing 2.0% are widely available. Assuming that the efficiency of zinc delivery is at least broadly similar for mouth-rinses and toothpastes, it seems probable that a substantial increase in plaque-fluid zinc citrate concentration would result following application of zinc at this higher concentration. As stated above, zinc concentrations in whole plaque fall rapidly in the hour following application, so it is likely that zinc concentration in plaque-fluid follow a similar trend i.e. concentrations are considerably higher in the hour following application. Further, zinc binds to lipoteichoic acid (LTA) associated with plaque bacteria in model biofilms in the same way as calcium, and it has been proposed that half of the bound zinc would be released under cariogenic conditions, largely through protonation of carboxylate and phosphate groups in LTA. Therefore more zinc may be available during an acidic challenge.

## Conclusion

This study demonstrates the ability of zinc to reduce the demineralization of enamel during in vitro caries-simulating conditions. Under the conditions used, zinc acts at the PO 4 3− sites on enamel surfaces to form an α-hopeite-like phase. Zinc phosphates have a low solubility product constant, and therefore these phases formed on enamel surfaces are unlikely to be very soluble under acidic conditions. In light of the current findings, it would appear there is scope for exploring and optimizing the therapeutic potential of zinc, not only as an antibacterial agent but also as a possible preventive treatment for caries. Thus, these results have a significant implication on the understanding of the fundamental chemistry aspects of zinc in toothpastes and demonstrate its therapeutic potential in reducing tooth mineral loss.

## Conflict of interest

The authors declare no potential conflicts of interest with respect to the authorship and/or publication of this article.

## Acknowledgements

This study was funded by GlaxoSmithKline Grant No. ( STU10018277 ) via an Case PhD Studentship. The authors would like to thank Rory Wilson from the School of Engineering and Materials Science for technical support with XRD and Tristan Youngs, the NIMROD beam-line scientist at the ISIS neutron facility for technical support with the Neutron Diffraction experiment 1310347.

## References

• 1. Peterson P.E., Bourgeois D., Ogawa H., Estupinan-Day S., Ndiaye C.: The global burden of oral diseases and risks to oral health. Bulletin of the World Health Organization 2005; 83: pp. 661-669.
• 2. Robinson C., Shore R.C., Brookes S.J., Strafford S., Wood S.R., Kirkham J.: The chemistry of enamel caries. Critical Reviews in Oral Biology and Medicine 2000; 11: pp. 481-495.
• 3. Fedorowicz Z., Aljufairi H., Nasser M., Outhouse T.L., Pedrazzi V.: Mouthrinses for the treatment of halitosis. Cochrane Database of Systematic Reviews 2008; Art. No.: CD006701
• 4. Young A., Jonski G., Rolla G.: Inhibition of orally produced volatile sulfur compounds by zinc, chlorhexidine or cetylpyridinium chloride – effect of concentration. European Journal of Oral Sciences 2003; 111: pp. 400-404.
• 5. Saxton C.A., Harrap G.J., Lloyd A.M.: The effect of dentifrices containing zinc citrate on plaque growth and oral zinc levels. Journal of Clinical Periodontology 1986; 13: pp. 301-306.
• 6. Chow L.C., Eanes E.D.: Octacalcium Phosphate.2001.KargerSwitzerland
• 7. LeGeros R.Z., Bleiwas C.B., Retino M., Rohanizadeh R., LeGeros J.P.: Zinc effect on the in vitro formation of calcium phosphates: relevance to clinical inhibition of calculus formation. American Journal of Dentistry 1999; 12: pp. 65-71.
• 8. Shepherd D., Best S.M.: Production of zinc substituted hydroxyapatite using various precipitation routes. Biomedical Materials 2013; 8:
• 9. Brudevold F., Steadman L.T., Spinelli M.A., Amdur B.H., Gron P.: A study of zinc in human teeth. Archives of Oral Biology 1963; 8: pp. 135-144.
• 10. Pawlig O., Trettin R.: Synthesis and characterization of alpha-hopeite, Zn-3(Po-4)(2)center dot 4H(2)O. Materials Research Bulletin 1999; 34: pp. 1959-1966.
• 11. Elliott J.C.: Calcium phosphate biominerals. Phosphates: Geochemical, Geobiological, and Materials Importance 2002; 48: pp. 427-453.
• 12. Elliott J.C.: Structure and Chemistry of the Apatites and Other Calcium Orthophosphates.1994.ElsevierMilton Keynes
• 13. Kukura M.B.L., Posner A.M., Quirk J.P.: Radioisotope determination of the surface concentrations of calcium and phosphorus on hydroxyapatite in aqueous solution. Journal of Physical Chemistry 1972; 76: pp. 900-904.
• 14. Okazaki M., Sato M., Takahashi J.: Space-cutting model of hydroxyapatite. Biomaterials 1995; 16: pp. 45-49.
• 15. Ingram G.S., Horay C.P., Stead W.J.: Interaction of zinc with dental mineral. Caries Research 1992; 26: pp. 248-253.
• 16. Fuierer T.A., Lore M., Puckett S.A., Nancollas G.H.: A mineralization adsorption and mobility study of hydroxyapatite surfaces in the presence of zinc and magnesium ions. Langmuir 1994; 10: pp. 4721-4725.
• 17. Frost R.L.: An infrared and Raman spectroscopic study of natural zinc phosphates. Spectrochimica Acta. Part A-Molecular and Biomolecular Spectroscopy 2004; 60: pp. 1439-1445.
• 18. Crisp S., O‘Neill I.K., Prosser H.J., Stuart B., Wilson A.D.: Infrared spectroscopic studies on the development of crystallinity in dental zinc phosphate cements. Journal of Dental Research 1978; 57: pp. 245-254.
• 19. Xu Y., Schwartz F.W., Traina S.J.: . Environmental Science & Technology 1994; 28: pp. 1472-1480.
• 20. Cheung C.W., Porter J.F., McKay G.: Removal of Cu(II) and Zn(II) ions by sorption onto bone char using batch agitation. Langmuir 2002; 18: pp. 650-656.
• 21. Lee Y.J., Elzinga E.J., Reeder R.J.: Sorption mechanisms of zinc on hydroxyapatite: systematic uptake studies and EXAFS spectroscopy analysis. Environmental Science & Technology 2005; 39: pp. 4042-4048.
• 22. Corami A., Mignardi S., Ferrini V.: Cadmium removal from single- and multi-metal (Cd plus Pb plus Zn plus Cu) solutions by sorption on hydroxyapatite. Journal of Colloid and Interface Science 2008; 317: pp. 402-408.
• 23. Robinson C., Kirkham J., Shore R.: Dental enamel: formation to destruction.1995.CRCBoca Raton, FL/London
• 24. Cummins D., Creeth J.: Delivery of anti-plaque agents from dentifrices, gels and mouthwashes. Journal of Dental Research 1992; 71: pp. 143-149.
• 25. Gilbert R.J., Ingram G.S.: The oral disposition of zinc following the use of an anti-calculus toothpaste containing 0.5% zinc citrate. Journal of Pharmacy and Pharmacology 1988; 40: pp. 399-402.
• 26. Afseth J., Oppermann R.V., Rolla G.: Accumulation of Cu and Zn in human dental plaque in vivo. Caries Research 1983; 17: pp. 310-314.
• 27. Adams S.E., Theobald A.J., Jones N.M., Brading M.G., Cox T.F., Mendez A., et. al.: The effect of a toothpaste containing 2% zinc citrate and 0.3% Triclosan on bacterial viability and plaque growth in vivo compared to a toothpaste containing 0.3% Triclosan and 2% copolymer. International Dental Journal 2003; 53: pp. 398-403.
• 28. Rose R.K.: . Caries Research 1996; 30: pp. 71-75.