Kinetics of canine dental calculus crystallization: An in vitro study on the influence of inorganic components of canine saliva

Journal of Colloid and Interface Science 425 (2014) 20–26

Contents lists available at ScienceDirect

Journal of Colloid and Interface Science www.elsevier.com/locate/jcis

Kinetics of canine dental calculus crystallization: An in vitro study on the influence of inorganic components of canine saliva Ballav M. Borah a,1, Timothy J. Halter a,1, Baoquan Xie a, Zachary J. Henneman a, Thomas R. Siudzinski a, Stephen Harris b, Matthew Elliott c, George H. Nancollas a,⇑ a b c

Department of Chemistry, University at Buffalo, The State University of New York, Buffalo, NY 14260, USA Waltham Centre for Pet Nutrition, Leicestershire LE14 4RT, UK Mars Care & Treats Europe, Oakwell Way, Birstall WF17 9LU, UK

a r t i c l e

i n f o

Article history: Received 6 December 2013 Accepted 13 March 2014 Available online 21 March 2014 Keywords: Carbonated hydroxyapatite Calcium phosphate Calcium carbonate Dental calculus Saliva Biomaterial Nucleation Supersaturated solution

a b s t r a c t This work identifies carbonated hydroxyapatite (CAP) as the primary component of canine dental calculus, and corrects the long held belief that canine dental calculus is primarily CaCO3 (calcite). CAP is known to be the principal crystalline component of human dental calculus, suggesting that there are previously unknown similarities in the calcification that occurs in these two unique oral environments. In vitro kinetic experiments mimicking the inorganic components of canine saliva have examined the mechanisms of dental calculus formation. The solutions were prepared so as to mimic the inorganic components of canine saliva; phosphate, carbonate, and magnesium ion concentrations were varied individually to investigate the roll of these ions in controlling the nature of the phases that is nucleated. To date, the inorganic components of the canine oral systems have not been investigated at concentrations that mimic those in vivo. The mineral composition of the synthetic calculi grown under these conditions closely resembled samples excised from canines. This finding adds new information about calculus formation in humans and canines, and their sensitivity to chemicals used to treat these conditions. Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction Attempts have been made to study mineralization events that take place in canine saliva. The pH and inorganic ionic concentrations differ greatly, and are known to markedly influence reactions at tooth surfaces [1,2]. Dental calculus is composed primarily of mineral and organic components and is usually classified by location as supra- or sub-gingival calculus [3]. Supra-gingival and sub-gingival calculi differ from each other with respect to the principal crystal constituents and inorganic elemental components [4]. As soon as enamel is exposed to saliva, dental plaque is formed as the result of bacterial colonization, followed by subsequent mineralization of phases involving calcium, phosphate, carbonate, and other ions in the contacting saliva [5]. Canine calculi appear as granular, yellow–brown masses on the buckle surfaces of teeth and are largest on molar teeth of the upper jaw near salivary duct orifices. They are associated with gingivitis, ⇑ Corresponding author. Fax: +1 716 645 6947. 1

E-mail addresses: [email protected], [email protected] (G.H. Nancollas). These authors made equal contributions.

http://dx.doi.org/10.1016/j.jcis.2014.03.029 0021-9797/Ó 2014 Elsevier Inc. All rights reserved.

periodontitis, and erosion of tooth surfaces. Four types of calcium phosphates, varying in relative abundance from sample to sample, are reported present in human dental calculi [6,7]. These calcium phosphates are: brushite (dicalcium phosphate dihydrate, DCPD, CaHPO42H2O), octacalcium phosphate (Ca8H2(PO4)65H2O), magnesium-containing whitlockite, b-TCP [(Ca,Mg)3(PO4)2], and carbonate-containing hydroxyapatite [(Ca,M)10(CO3,HPO4,PO4)6 (OH,X)2], where M represents other cations capable of substituting for the Ca2+, e.g., Sr2+, Pb2+, K+, Na+, etc.; X = Cl or F. Calculus formation can be divided into two parts; (1), adhesion between calcified plaque and the tooth surface and, (2), the mineralization of this phase involving calcium, phosphate, carbonate, and other ions in the contacting saliva [8]. Calcification is also a kinetic process controlled by nucleation and growth of calcium-phosphate-rich phases on the tooth surface [9]. The rate of calculus formation can be elevated by salivary pH, salivary calcium concentration, bacterial protein and lipid concentration, and concentration of protein in the sub-mandibular salivary gland secretions as well as low inhibitory factors and higher total salivary lipid levels. The most widely cited work on canine dental calculi identifies the main component the calculi as calcium carbonate, with minor amounts of calcium phosphate [1]. The carbonate concentration in

21

B.M. Borah et al. / Journal of Colloid and Interface Science 425 (2014) 20–26

canine saliva much is greater than that in humans and the lack of sufficient analysis of whole calculus samples led previous investigations likely led to conclusions. Preliminary speciation calculations showing relative supersaturation (r) values for hydroxyapatite (rHAP) that are more than 10 times those for calcite (rcalcite), suggested that canine calculus is predominantly composed of calcium phosphate minerals. Carbonate is known to markedly influence the mineralization of calcium phosphates; under pseudo canine conditions of relatively high carbonate and low phosphate concentrations, with high pH (7.7–8.5), the mineralization events were never studied quantitatively, nor the roles of specific inorganic components. An in-depth investigation of excised samples has revealed that the previously held views on canine dental calculus were incorrect and apatitic calcium phosphate minerals with minor inclusions of carbonate are principal components of canine dental calculus. 2. Materials and methods 2.1. Canine saliva collection Canine saliva and calculus samples were collected from dogs housed at The WALTHAM Center (UK). Dogs were trained using a clicker for positive reinforcement to provide saliva samples by chewing or licking on large cotton wool pads. The pads were centrifuged at 500g for 5 min in a Salivette (Sarstedt), accumulated saliva was removed and the pads were spun at 1000g for 5 min to remove the remaining saliva fluid. The saliva samples were analyzed by Mars Care & Treats Europe, Oakwell Way, Birstall, WF17 9LU, UK. 2.2. Driving force and solution speciation Solution speciation calculations were made to determine the relative supersaturations, r, for mineral phases under canine salivary conditions (Table S3). These calculations were made using the extended Debye–Hückel equation proposed by Davies, with mass balance expressions for total calcium and phosphate incorporating appropriate ion pair equilibrium constants by successive approximation for the ionic strengths, I (Eq. (1)).

r ¼ S  1 ¼ ½IAP=K SP 1=m  1

ð1Þ

where IAP is the ionic activity product of free lattice ions in solution, KSP the activity solubility product (HAP 2.88  10112 mol18 L18) [10] and t the number of ions in a formula unit (HAP v = 18). Conditions used for the calculations are shown in Table 1. The aim of these calculations was to use the analytically determined total concentrations of the dissolved inorganic calculus components and estimate the extent of ion-pairing, as well was the mineral phases which form under canine oral conditions to form dental calculus. 2.3. FE-SEM and TEM Crystallites were collected from solution by filtration using polycarbonate membrane filters (200 nm pore size). The samples

were then dried, sputter coated with graphite under vacuum, and examined by field-emission scanning electron microscopy (FE-SEM, Hitachi SU-70) at 20 kV. Transmission electron microscopy (TEM) investigations of nucleated particles removed at various time intervals in Ultrathin Carbon type A (400 mesh copper grids) were made using a JEOL-2010 TEM at an accelerating voltage of 200 kV. 2.3.1. Experimental method Reactions were made in TeflonÒ covered, 150 mL double-walled PyrexÒ jacketed cells, thermostated at 38.0 ± 0.1 °C by a circulating water bath. Supersaturated solutions were prepared by first adding triple distilled water (TDW) to the cell, followed by slowly mixing the appropriate amounts of CaCl2, MgCl2, KH2PO4, and NaHCO3with NaCl to established the ionic strength, I. The pH was adjusted to the desired value (pH = 8.50) using 0.05 M KOH solutions. Titrant addition was potentiometrically controlled by a glass electrode (Orion 91-01) and reference electrode (Orion 900100). During growth, the electrode potential was continuously compared with the pre-set voltage bias and a difference in signal (D P +0.10 mV) activated the motor-driven titrant burettes, which maintained a constant thermodynamic driving force by the addition of titrant solutions. The final reaction solution concentrations approximated those of canine saliva. Hydrogen ion activity was monitored by a pH electrode coupled to a Ag/AgCl single junction reference electrode and a pH meter. Data were analyzed using the Nernst equation and calibration of the electrodes was performed using the following buffers: (1) pH 9.088 (9.96 mM Na2B4O7) and (2) pH 7.385 (8.70 mM KH2PO4 and 30.43 mM Na2HPO4, at 37 °C). After calibration, the electrodes were washed extensively with triple distilled water (TDW) and allowed to stabilize in the reaction cell. When a stable emf value had been reached, the automatic titrator potentiostat was set to this value (i.e. the ‘‘voltage bias’’). As nucleation and growth took place, the removal of ions from the bulk solution caused a deviation in the cell solution concentration and emf from the set value. If greater than ±0.050 mV, the addition of titrant solutions from mechanically coupled burets was triggered to restore the emf to the ‘‘voltage bias’’. A computer program recorded the titrant addition volume as a function of time (Fig. S4). Prior to the start and at the end of each experiment, a weighed aliquot of solution was removed for calcium and phosphate analysis. At the end of the experiment, cell solutions were vacuum filtered and the remaining precipitates were saved for further solid phase analysis. Phosphate and calcium analyses were made using UV–Vis (as the vanadomolybdate complex, Perkin Elmer Lambda 25) and atomic absorption spectroscopy (AAS) (Perkin Elmer 3100), respectively. 2.3.2. Titrant solutions Addition of the titrant solution containing KOH and NaCl was triggered to maintain constant pH and ionic strength. The reaction solution concentration (Wi), of the free ion, i, following titrant addition was given by:

Wi ¼

W i V T þ T i dV t  xdn V T þ nb dV t

ð2Þ

Table 1 Solution concentration for synthetic canine dental calculus growth experiments (I = 0.10 M, pH = 8.50 at 38.0 °C). Component

Calcium (as CaCl2) Phosphate (as KH2PO4) Magnesium (as MgCl2) Bicarbonate (as NaHCO3) Sodium chloride

Concentration (mmol L1)

2.57 0.56 2.29 10.10 84.23

Mole ratio

[Ca]/[HCO3] = 0.25 [Ca]/[PO4] = 4.59 [Ca]/[Mg] = 1.12

Calculated relative supersaturation Phase (xx)

r(xx)

Calcite (CaCO3) Hydroxyapatite (HAP) Tricalcium phosphate (TCP) MgCO3

3.27 47.70 9.06 0.66

22

B.M. Borah et al. / Journal of Colloid and Interface Science 425 (2014) 20–26

where VT is the volume of reaction solution, Ti the titrant concentration, dVt the volume of titrant added, x the stoichiometric coefficient of the species, dn the number of moles grown, and nb the number of titrant tubes. Ti was determined by Eq. (3),

T i ¼ W i N þ C eff

ð3Þ

in the presence of magnesium (Fig. 1C) under canine oral conditions, and this is in contrast to the inhibition of calcium phosphate mineral formation by magnesium under previously investigated human oral conditions [11,12]. 3.2. Influence of titrant components

4

Ceff (Ceff = 2.0  10 M) was the number of moles of mineral grown per liter of added titrant, Ceff is defined by:

C eff ¼ dn=dV t

ð4Þ

and for total chloride:

2W CaCl2 þ W NaCl ¼

ð2W CaCl2 þ W NaCl ÞV T þ 2T CaCl2 dV t þ T NaCl dV t V T þ nb dV t ð5Þ

Total sodium:

W KH2PO4 þ W NaCl þ W KOH ¼

W NaHCO3 þ W NaCl þ W KOH ¼

The formation of a magnesium-containing carbonated apatite (Mg-CAP) phase indicated that titrant equations, initially designed for calcite growth, required modification to account for apatite precipitation. In the subsequent sets of experiments, the original calcite titrant equations (Eqs. (7)–(11)) were modified to better-fit apatite conditions (Eqs. (13)–(16)). Total sodium:

ðW NaHCO3 þ W NaCl þ W KOH ÞV T þ T NaHCO3 dV t þ T NaCl dV t þ T KOH dV t ð6Þ V T þ nb dV t

When combined, Eqs. (3)–(6) enabled calculation of the titrant concentrations for calcium chloride (x = 1) and sodium bicarbonate (x = 1) for calcite stoichiometry and potassium phosphate (x = 3) for HAP stoichiometry.

ðW KH2PO4 þ W NaCl þ W KOH ÞV T þ T KH2PO4 dV t þ T NaCl dV t þ T KOH dV t ð12Þ V T þ nb dV t

When combined, Eqs. (5), (11), and (12) enable calculation of the titrant concentrations for calcium chloride (x = 5) and potassium phosphate (x = 3) for HAP stoichiometry.

T CaCl2 ¼ nb W CaCl2 þ 5C eff

ð13Þ

T KH2PO4 ¼ nb W KH2PO4 þ 3C eff

ð14Þ

T CaCl2 ¼ nb W CaCl2 þ C eff

ð7Þ

T NaCl ¼ nb W NaCl 10C eff

ð15Þ

T NaHCO3 ¼ nb W NaHCO3 þ C eff

ð8Þ

T KOH ¼ nb W KOH þ 7C eff

ð16Þ

T NaCl ¼ nb W NaCl 2C eff

ð9Þ

T KOH ¼ nb W KOH þ C eff

ð10Þ

T KH2PO4 ¼ nb W KH2PO4 þ 3C eff

ð11Þ

The influence of titrant solution containing phosphate and bicarbonate was investigated. Under conditions similar to those in experiment 1C, but phosphate was added to the titrant solution. This was done to compensate for the formation of calcium phosphate minerals (Fig. 2), thus accelerating the reaction by maintaining the driving force. Elemental analyses of reaction solutions without phosphate in the titrant were compared to those containing phosphate (Table 2). At the conclusion of the reactions the solutions were filtered, the suspended crystallites collected, dried at 60 °C, and analyzed. Nano-structured spherical particles with an average diameter of 100 nm were observed using SEM (Fig. 3a and b). Samples from four replicate experiments were pooled to acquire adequate material for examination by powder X-ray diffraction (XRD). The broad

3. Results and discussion This work designed to study mineralization that occurs in canine saliva has shown the dramatic influence of Mg2+, HPO2 4 and HCO 3 on the kinetics of nucleation of minerals, and the nature of the phases precipitated. 3.1. Growth at pH 8.50 – influence of solution components Crystal growth experiments were made to study the influence of magnesium and phosphate on the kinetics of mineral formation in solutions containing ionic calcium, carbonate, sodium, and chloride, at a pH of 8.50. Titrant solutions containing crystal lattice ions and electrolytes were added throughout the experiments; these additions were potentiometrically controlled using glass electrodes. The precipitates were examined using FE-SEM and elemental analyses were done by Energy-dispersive X-ray Spectroscopy (EDS). Crystal growth experiments, with phosphate and magnesium present in solution resulted in the precipitation of amorphous magnesium-containing apatitic calcium phosphate (Supplemental S5-B, Ca/P = 1.82). This magnesium-containing apatitic calcium phosphate was also observed in natural canine samples which had a Ca/P = 1.98 (Table 4). The absence of phosphate resulted in the rapid precipitation of calcite (Fig. 1A) without detection of MgCO3 (Supplemental S5-A). In experiments with phosphate but in the absence of magnesium, an amorphous calcium phosphate phase (Ca/P = 1.35) was precipitated (Supplemental S5-C), similar to that detected in the initial stages of growth in the presence of magnesium. These results along with titrant volume consumed, suggest that calcium phosphate mineralization was accelerated

Fig. 1. Titrant addition curves for experiments with (A) 2.29 mM magnesium without phosphate, (B) 0.56 mM phosphate without magnesium, and (C) 2.29 mM magnesium and 0.56 mM phosphate. All experiments made at I = 0.10 M and pH 8.50 at 38.0 °C.

23

B.M. Borah et al. / Journal of Colloid and Interface Science 425 (2014) 20–26

Fig. 2. Comparison of growth (A) with, and (B) without phosphate added to the titrant solution (rcalcite = 3.27, rHAP = 47.71). All experiments made at I = 0.10 M and pH 8.50 at 38.0 °C.

non-ordered peaks at 2h = 31°and 47 ° are characteristic peaks of carbonated apatite, the peaks are broadened by the less crystalline nature of carbonated apatite compared to hydroxyapatite. These results combined with SEM, EDS, and SAED were used to identified as nano-structured amorphous carbonated apatite (Fig. 3d) [13,14]. Earlier studies on synthetic calcite systems demonstrated that the addition of phosphate results in the formation of carbonate-containing apatites, similar in crystallographic and spectroscopic characteristics to bone and dentin [15,16]. Replicate experiments were done to obtain sufficient filtered material for elemental analysis of the dissolved crystallites by AAS and UV–Vis. 10 mg of precipitate from each set of pooled samples were dissolved in a 2.0% by vol. conc HCl solution, and analyzed for calcium, magnesium (Table 3). A similar analysis was also made for natural canine calculus (Table 4). Micro-Raman analyses of known synthetic calcium phosphate and calcium carbonate phases were compared to the precipitated crystallites. The most appropriate phase match for the precipitated crystallites was CAP (Supplemental S8). The empirical formulae calculated from the compositions of the precipitates obtained from the experiments, and natural canine calculus samples can be expressed as Ca1.9 Mg0.6 (PO4) (CO3) and Ca2.0 (PO4) (CO3)0.5 respectively. The greater incorporation of magnesium into synthetic samples is believed to be a result of the biological control of the process in vivo. This incorporation may be the result of the formation of kinetically favorable ion pairs of [MgCO3]0.6, which would act as nucleation sites for the precipitation of kinetically less favorable but thermodynamically more

stable calcium phosphate amorphous phases. It was also found the precipitation of calcium phosphate is accelerated by the addition of magnesium when the carbonate concentration of the solution is similar to that which is typically found in canine saliva (Fig. 1), a phenomenon not observed under human salary conditions. This finding would indicate magnesium is very important during the kinetically controlled initial stages of nucleation and growth. Based on earlier studies on synthetic Mg-substituted apatites it was suggested that biologic apatites should be represented by a formula such as (Ca,Na,Mg)10(PO4,HPO4,CO3)6(OH,Cl,F)2 [17,18], however it is still not clear whether magnesium can be incorporated into the crystal lattices of Ca–P biominerals. Elemental analysis of excised canine samples found <0.5% Mg, supporting previous findings that Mg ions are stored in the extracellular matrices or adsorbed on crystal surfaces as a contaminant [19,20]. The formation of a magnesium-containing carbonated apatite (Mg-CAP) phase suggested the titrant equations, initially designed for calcite growth, and required modification to account for apatite precipitation. The original calcite titrant equations (Eqs. (7)–(11)) were modified to account for the formation of apatite (Eqs. (13)–(16)). 3.3. Influence of pH in growth experiments The in vivo pH of canine saliva lies between pH 7.20 and 8.50 (mean value 7.70) [21]. A constant Ca2+, PO3 4 concentration could not be maintained in experiments done at >pH 8.30, the pH at which instantaneous precipitation occurred. The most appropriate pH for kinetic studies of mineral formation at these ionic concentrations determined to be pH 8.25. Under these conditions precipitation could not be detected prior to the activation of the automated titrators, and the Ca2+, PO3 concentration could be 4 maintained within 5% throughout the experiments. At lower pH values of 7.70–8.20 no mineralization could be detected by pH electrode (Fig. 4a and b). During these experiments, the pH drifted towards the alkaline region (Fig. 4b). We speculate that bicarbonate (pKa = 10.33) buffering was the reason for this result. To investigate the influence of bicarbonate buffering on crystallite growth, TRIS buffer (pKa = 8.06) was introduced. In experiments done using TRIS buffer at pH 6 8.25 an opaque viscous layer formed on the walls of the reaction cell 1 h after the final pH adjustment, which coalesced into a white gel-like deposit 6 h into the reaction, after 24 h the gel layer solidified (Supplemental S10). This solid was identified as amorphous carbonated apatite, and it was speculated to have formed via a liquid–liquid (spinodal) phase separation mechanism [22]. Samples removed throughout the experiments were analyzed using TEM and STEM conversionEDS (Fig. 5b, inset); analysis confirmed that the initial phases were amorphous. This was especially interesting since it clearly showed that crystallization of the mineral phase(s) was pH driven. Under

Table 2 Analysis of growth reaction solutions (I = 0.10 M, pH 8.50 at 38.0 °C). Analyzed using UV–Vis (PO4) and AAS (Ca). Set

Components

Sample

mM Ca

mM PO4

DCaexpa (%)

DPO4,expa (%)

DCacanineb (%)

DPO4,canineb (%)

1A

w/Mg, wo/PO4

16.70

n.a.

51.00

n.a.

w/PO4, wo/Mg

38.50

24.90

53.60

w/Mg and PO4

0.00

30.00

29.20

64.30

2

w/Mg and PO4

0.26 0.16 0.20 0.26 0.21 0.23

2.10

1C

1.26 1.47 1.93 1.89 1.82 1.82 1.94 1.87

n.a.

1B

a b a b a b a b

24.5

62.3

3.6

7.1

Set 1 – without phosphate in the titrant solution; Set 2 – with phosphate in the titrant solution. Sample (a) and (b) – solutions collected at the start and the end of experiment monitored by pH electrode. a Percent change between initial and final concentration of components in experimental solution. b Percent difference in initial solution component concentration versus canine saliva conditions ([Ca] = 2.57 mM, [PO4] = 0.56 mM as shown in Table 1).

24

B.M. Borah et al. / Journal of Colloid and Interface Science 425 (2014) 20–26

Fig. 3. (a) and (b) SEM micrographs of crystallites withdrawn from growth experiments with phosphate in the titrant and a mixture of magnesium (2.29 mM) and phosphate (0.56 mM) in the reaction solution, (c) energy dispersive X-ray spectroscopy and (d) powder XRD of precipitates obtained from growth experiments.

Table 3 Elemental analysis of precipitate from growth experiments (I = 0.10 M, pH 8.50 at 38.0 °C). Samples dissolved in 2% HCl and analyzed using UV–Vis (PO4) and AAS (Ca, Mg). Sample

% Ca

% Mg

% PO4

Ca/Pa

Mg/Pa

Ca/Mga

A B

25.20 24.90

4.50 4.30

31.00 31.90

1.92 1.85

0.57 0.53

3.40 3.49

A – without phosphate in the titrant solution; B – with phosphate in the titrant solution. a Units of mole ratio.

Table 4 Elemental analysis of natural canine calculus. Samples dissolved in 10% HCl and analyzed using UV–Vis (PO4) and AAS (Ca, Mg, and Na).

a

Sample

% PO4

% Ca

% Mg

% Na

Ca/Pa

Ca/Mga

1 2 3 4 5

27.53 24.50 28.95 28.46 27.55

22.44 24.32 22.05 21.43 23.36

0.16 0.38 0.45 0.39 0.39

0.73 0.69 0.85 0.87 0.70

1.93 2.35 1.80 1.78 2.01

84.09 39.13 29.98 33.24 36.08

Mean Std. dev.

27.40 1.73

22.72 1.14

0.35 0.11

0.77 0.08

1.98 0.23

48.97 25.93

Units of mole ratio.

the previous pH conditions (pH P 8.30; no TRIS), bicarbonate itself was acting as a weak pH buffer, and the rapid drop in pH upon precipitate formation thermodynamically drove the crystallization reaction to form Mg-CAP. TRIS buffer maintained the pH, preventing the pH-driven rapid crystallization of apatite, and favoring the formation of an amorphous calcium phosphate (ACP). STEM

conversion micrographs (Fig. 5c and d) and the EDS spectra (Supplemental 12) confirmed the presence of an ACP phase having Ca/P = 1.03. Free drift experiments to study the influence of dissolved carbon dioxide were carried out in the absence of both phosphate and bicarbonate at ionic strength I = 0.10 M, pH = 8.20 ± 0.01. In a first set of experiments, nitrogen saturated with water vapor at 38.0 °C was purged through the reaction solution to exclude carbon dioxide. In a second experimental set, the solution was not bubbled with water-saturated N2 gas. The experiments were monitored by glass and calcium ion selective electrodes (Orion 93-20). It was found that in the presence of dissolved carbon dioxide, possibly due to the formation of soluble calcium-carbonate complexes, both pH as well as free Ca2+ ion concentration decreased with time (Fig. 6I-A and II-A). In the absence of dissolved carbon dioxide, both pH and free calcium ion concentration remained constant (Fig. 6I-B and II-B). These results indicate that calcium can form Ca–CO3 ionic clusters with the dissolved CO2 without precipitation. The presence of stable pre nucleation ion clusters in solutions of dissolved calcium carbonate have been reported by Gebauer et al., and the study also identifies amorphous calcium carbonate (ACC) as a post nucleation-stage precursor to calcium carbonate mineralization. It was suggested that a substantial gain in entropy is provided by the release of water from hydrated layers of ions, favoring pre-nucleation cluster formation [22]. This is against classical concept for nucleation. In classical concept, nucleation occurs in a solution of ions that has become supersaturated, leading to the formation of the solid phase by stochastic solute clustering (metastable ion clusters) where the earliest crystal precursor is considered to be a cluster of critical size. Tribello et al. examined the precipitation of calcium carbonate in water using a combination of molecular dynamics and umbrella

B.M. Borah et al. / Journal of Colloid and Interface Science 425 (2014) 20–26

25

Fig. 4. (a) Volume (mL) of titrant added plotted against time and (b) pH against time at (A) pH 7.70, rHAP = 21.65, (B) pH 7.90, rHAP = 26.83, (C) pH 8.10, rHAP = 32.89, (D) pH 8.20, rHAP = 36.26, (E) pH 8.30, rHAP = 39.87, (F) pH 8.50, rHAP = 47.71 and (G) pH 8.25, rHAP = 38.04 using solution concentrations.

Fig. 5. (a) and (b) TEM micrographs and selected area electron diffractograms, (c) and (d) Transmission-SEM micrographs of precipitate collected in the first 5 min of experiments made with TRIS buffer, pH 8.25 at 38.0 °C.

sampling [23]. They found that, although calcite nano-crystals are stable in solution, at high supersaturations kinetically favored particles of strongly hydrated amorphous phases are formed. The formation of these amorphous nano-spheres has been explained using a liquid–liquid (spinodal) phase separation mechanism. Meldrum et al. suggested that the pre-critical clusters observed by Gebauer et al. might be the result of stabilization of calcium carbonate clusters by another species present as an impurity [24].

Based on our experimental data and studies [22–24], the formation of pre-nucleation ionic clusters of calcium carbonate is proposed. The negligible activation barrier for the formation of pre-nucleation ionic clusters of calcium carbonate results in its initial formation. These species subsequently act as nuclei for the heterogeneous nucleation of calcium phosphate into ionic clusters. The pre-nucleated calcium carbonate ionic clusters may behave as precursors, leading to nucleation of calcium phosphate and by coalescence amorphous carbonated apatite is formed.

26

B.M. Borah et al. / Journal of Colloid and Interface Science 425 (2014) 20–26

Fig. 6. (I) pH (glass electrode) plotted against time and (II) Ca2+ ion selective electrode potential (mV) plotted against time (A) without bubbling N2 (B) With N2 bubbling. I = 0.10 M and pH 8.20 at 38.0 °C.

4. Conclusions

Appendix A. Supplementary material

This work identifies carbonated hydroxyapatite as the primary component of canine dental calculus, and corrects the long-held belief that this phase is primarily calcite. Our in-depth investigations of dental calculus excised from live canines revealed that in every sample tested, calcium phosphate minerals were the principal components of canine dental calculus. This claim is supported by analytical data obtained using FE-SEM, TEM, STEM conversion, EDS, powder-XRD, and elemental analysis of the dissolved solids by AAS and UV–Vis. Experiments done to study the kinetics of mineral precipitation under canine salivary conditions yielded a carbonated apatite like product, further supporting this claim. The data on the mineral formation presented in this work will alter current techniques for canine dental calculus treatment, and should improve proactive measures currently taken to prevent calculus formation in domesticated dogs around world. The calcium phosphate mineralization studies in this work provided unique data about mineralization under conditions never before investigated. The acceleration of calcium phosphate precipitation by magnesium at high carbonate concentration is fascinating, and the phenomenon warrants further study. The understanding of the influence of various inorganic components in canine saliva on the mechanisms of dental calculus formation has been advanced. The work presented here is a stepping stone in the development of physio-chemical experimental methods for studying the kinetics of biomineralization, and further develops experimental methods for testing potential commercial calculus inhibitors and modifiers.

Speciation calculations, CC technique and solution analysis, titrant equations, EDX and micro Raman are available online only. Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jcis.2014.03.029.

Acknowledgments We thank Waltham Centre for Pet Nutrition, Waltham on the Wolds, Melton Mowbray, Leicestershire, LE14 4RT, UK and Mars Care & Treats Europe, Oakwell Way, Birstall, WF17 9LU, UK and National Institute of Dental and Craniofacial Research (NIDCR) Grant DE003223 to G.H.N. for support of this work. We are also grateful to Dr. Sarbajit Banerjee for the micro-Raman studies.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

R.Z. Legeros, I.L. Shannon, J. Dent. Res. 58 (1979) 2371–2377. M.J. Larsen, E.I. Pearce, Arch. Oral Biol. 48 (2003) 317–322. Y. Jin, H.K. Yip, Crit. Rev. Oral Biol. Med. 13 (2002) 426–441. J. Friskopp, G. Isacsson, Scand. J. Dent. Res. 92 (1984) 25–32. H.M. Uyen, L.J. van Dijk, H.J. Busscher, J. Clin. Periodontol. 16 (1989) 393–397. A.T. Jensen, M. Danø, J. Dent. Res. 33 (1954) 741–750. R.Z. LeGeros, J. Dent. Res. 53 (1974) 45–50. H.M. Uyen, L.J. van Dijk, H.J. Busscher, J. Clin. Periodontol. 16 (1989) 393–397. J. Christoffersen, M.R. Christoffersen, N. Kjaergaard, J. Cryst. Growth 43 (1978) 501–511. L.J. Shyu, L. Perez, S.J. Zawacki, J.C. Heughebaert, G.H. Nancollas, J. Dent. Res. 62 (1983) 398–400. H. Luoma, Magnes. Res. 1 (1988) 223–230. X. Yang, B. Xie, L. Wang, Y. Qin, Z.J. Henneman, G.H. Nancollas, CrystEngComm 13 (2011) 1153–1158. E.H. Chowdhur, A. Maruyama, A. Kano, M. Nagaoka, M. Kotaka, S. Hirose, M. Kunou, A. Akaike, Gene 376 (2006) 87–94. E.H. Chowdhury, Adv. Biomater. Sci. Biomed. Appl. (2013) 265–292. Ch10. D.I. Hay, E.C. Moreno, D.H. Schlesinger, Inorg. Perspect. Biol. Med. 2 (1979) 271–285. E. Saitoh, S. Isemura, K. Sanada, Arch. Oral Biol. 30 (1985) 641–643. S. Raz, S. Weiner, L. Addadi, Adv. Mater. (Weinheim Germany) 12 (2000) 38– 42. J.H. Schroeder, E.J. Dwornik, J.J. Papike, Geol. Soc. Am. Bull. 80 (1969) 1613– 1616. L. Bertinetti, C. Drouet, C. Combes, C. Rey, A. Tampieri, S. Coluccia, G. Martra, Langmuir 25 (2009) 5647–5654. C. Robinson, J. Kirkham, R. Shore, Editors, Dental Enamel: Formation to Destruction, 1995. M. Larmas, A. Scheinin, Acta Odontol. Scand. 29 (1971) 205–214. D. Gebauer, A. Volkel, H. Colfen, Science 322 (2008) 1819–1822. A. Tribello Gareth, F. Bruneval, C. Liew, M. Parrinello, J. Phys. Chem. B 113 (2009) 11680–11687. C. Meldrum Fiona, P. Sear Richard, Science 322 (2008) 1802–1803.