Calcium and phosphate release from resin-based materials containing different calcium orthophosphate nanoparticles Marcela C. Rodrigues, Livia C. Natale, Victor E. Arana-Chaves, Roberto R. Braga ~ o Paulo, School of Dentistry, Sa ~o Paulo, Sa ~o Paulo, Brazil Department of Biomaterials and Oral Biology, University of Sa Received 22 July 2014; revised 6 October 2014; accepted 26 October 2014 Published online 21 January 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.b.33327 Abstract: The study compared ion release from resin-based materials containing calcium orthophosphates. Amorphous calcium phosphate (ACP), dicalcium phosphate anhydrous (DCPA), dicalcium phosphate dihydrate (DCPD), and tricalcium phosphate (b-TCP) nanoparticles were characterized by X-ray diffraction (XRD), transmission electron microscopy (TEM), dynamic light scattering (DLS), and surface area (nitrogen adsorption isotherms, BET method). Nanoparticles were added to a dimethacrylate-based resin and materials were tested for degree of conversion (DC) and calcium/phosphate release up to 28 days under pH 5.5 and 7.0. Data were analyzed by ANOVA/Tukey test (alpha: 0.05).The crystallinity of DCPA, DCPD, and b-TCP were confirmed, as well as the ACP amorphous nature. DCPD and bTCP presented larger agglomerates than DCPA and ACP.
The surface area of ACP was 5–11 times higher than those of the other nanoparticles. Materials showed similar DC. The material containing ACP released significantly more ions than the others, which released similar amounts of calcium and, in most cases, phosphate. Ion release was not affected by pH. Calcium release decreased between 7 and 21 days, while phosphate levels remained constant after 14 days. In conclusion, ACP higher ion release can be ascribed to its high surface area. DCPA, DCPD, and b-TCP C 2015 had similar performances as ion-releasing fillers. V Wiley Periodicals, Inc. J Biomed Mater Res Part B: Appl Biomater, 103B: 1670–1678, 2015.
Key Words: calcium phosphate, resin composite, characterization, ion release, degree of conversion
How to cite this article: Rodrigues MC, Natale LC, Arana-Chaves VE, Braga RR. 2015. Calcium and phosphate release from resin-based materials containing different calcium orthophosphate nanoparticles. J Biomed Mater Res Part B 2015:103B:1670–1678.
INTRODUCTION
Resin-based materials with ion-releasing capabilities have been investigated since the early years of resin composites use.1,2 Recently, luting and restorative composites, as well as liners and sealants containing calcium phosphates became available for clinical use.3,4 Manufacturers claim that these materials can provide protection against caries by offering an extra source of calcium and phosphate ions to create a supersaturated environment that would favor remineralization. In fact, in vitro studies using experimental materials reported their ability to remineralize enamel subsurface lesions. For example, caries-like lesions in contact with a composite containing 40% (by weight) of ACP recovered 38% of the mineral content after 2 weeks in a pHcycling regimen.5 A recent study observed 22% of enamel remineralization after 30 days of pH cycling with the use of a composite containing 40 wt % of ACP nanoparticles, in comparison to 6% with a commercial fluoride-releasing composite.6 Calcium and phosphate ions are also fundamental in the remineralization process involving fluorides.7 They are necessary to maintain fluoride concentration at cariostatic lev-
els after topic fluoride treatments, either in the form of “calcium fluoride-like” (phosphate-stabilized calcium fluoride) deposits on the tooth surface or as bacterial bound fluoride in the dental plaque.8 These biological reservoirs, particularly calcium fluoride-like deposits, need an additional source of calcium to be formed at effective concentrations.9 The ionic concentration released from bioactive composites and the corresponding effect on remineralization has also been studied. In demineralized dentin, resin cements containing 73–78 wt % of dicalcium phosphate anhydrous (DCPA) and tetracalcium phosphate (TTCP) particles released 0.3–0.5 mmol/L of calcium and 0.05–0.1 mmol/L of phosphate, and remineralized 38–47% of the lesion after 5 weeks.10 In enamel, a composite containing 40 wt % of ACP particles released 0.74 mmol/L of calcium and 0.54 mmol/L of phosphate, remineralizing 14% of the lesion after 30 days, in comparison to 4% achieved by a fluoride-containing commercial cement.11 Recently, an in situ study verified that under accumulated biofilm, a composite containing 40 wt % of ACP nanoparticles was able to reduce mineral loss at the enamel/restoration interface
Correspondence to: R. R. Braga; e-mail: [email protected] ~o Paulo Research Foundation); contract grant number: 2012/04532-4 Contract grant sponsor: FAPESP (Sa
1670
C 2015 WILEY PERIODICALS, INC. V
ORIGINAL RESEARCH REPORT
TABLE I. Calcium Phosphates Used in the Study Name (Manufacturer) Amorphous calcium phosphate (Sigma–Aldrich, Milwaukee, WI) Dicalcium phosphate anhydrous (Synth, Diadema, SP, Brazil) Dicalcium phosphate dihydrate (Synth, Diadema, SP, Brazil) b-Tricalcium phosphate (Sigma–Aldrich, Milwaukee, WI)
by 2.4 times compared to a composite containing only barium glass. The cumulative release of calcium and phosphate at pH 5.5 after 14 days (corresponding to the duration of the in situ study) was approximately 4 and 1 mmol/L, respectively.12 As demonstrated with the abovementioned studies, calcium phosphates with different particle sizes and calciumto-phosphate (Ca/P) ratios have been tested as ion-releasing fillers and, therefore, their results are not comparable. Reducing particle size and consequently, increasing its surface area has been shown to significantly increase ion release.13,14 The availability of ions is also influenced by the pH of the immersion medium. A fourfold increase in calcium and phosphate release was observed with a composite containing 75 wt % of TTCP between pH 7.4 and 4.0.15 Furthermore, calcium phosphates have different critical pHs (i.e., pH at which the solution is saturated with respect to a particular calcium phosphate and below which it starts to dissolve), which also may affect ion release. For example, while saliva becomes saturated with hydroxyapatite at pH 5.5, DCPD starts to dissolve at pHs below 6.3 and b-TCP at 6.1.16 The type of calcium phosphate may also affect the concentration of ions released in solution. In general, the solubility of calcium phosphates increases at low pH.17 However, they show very low solubility in aqueous media at neutral pH. Actually, some are virtually insoluble.18 In order of decreasing solubility, dicalcium phosphate dihydrate (DCPD) is more soluble than DCPA, followed by ACP and btricalcium phosphate (b-TCP).19 It is interesting to notice that the majority of studies focus on variables other than calcium phosphate type, for example, nanoparticle content,15 ratio between bioactive and reinforcing fillers,20 or material presentation10 and no direct comparison between resinbased materials containing different types of calcium phosphates is available. Based on the above, the purpose of the present study was to compare the concentration of calcium and phosphate released from resin-based materials containing nanoparticles of different calcium orthophosphates, at pH 5.5 and 7.0. To substantiate the discussion, the nanoparticles were characterized in terms of crystallinity, micromorphology, particle-size distribution, surface area, and true density. Additionally, the DC of the resin-based materials was determined. The null hypothesis was that the nanoparticle characteristics or the pH of the immersion medium do not influence calcium and phosphate release.
Abbreviation
Chemical Formula
Ca/P
ACP
CaxHy(PO4)znH2O
1.15–1.38
DCPA DCPD b-TCP
CaHPO4 CaHPO4 2H2O Ca3(PO4)2
1.1 1.1 1.5
METHODS
Characterization of calcium phosphates Four commercially available calcium phosphate nanoparticles were chosen for this study (Table I). The type of structure (crystalline or amorphous) was identified by X-ray diffraction (XRD), using nickel filtered CuKa radiation at 40 kV and 30 mA (MultiFlex, Rigaku, Tokyo, Japan). The equipment geometry was h/2h. Readings were continuous at angles from 10 to 60 at 0.05 intervals, 10 s per interval. Nanoparticle morphology was observed under transmission electron microscopy (TEM; JEOL, model 1010, Tokyo, Japan), using an accelerating voltage of 80 kV. Powders were dispersed in isopropyl alcohol and, after 24 h, a few drops of the supernatant were placed on nickel grids (200 mesh) covered with a poly(vinyl formal) pellicle (Formvar). Nanoparticle dimensions were measured using the ImageJ software (National Institute of Health, Bethesda, MD). Twenty nanoparticles were measured for each material. Additionally, particle-size distribution was estimated by dynamic light scattering (DLS, Nanotrac 252, Microtrac, Montgomeryville, PA). The powders were dispersed in isopropyl alcohol and sonicated for 10 min. Then, after 20 min, the supernatant was collected for analysis. Nitrogen adsorption isotherms were obtained (Quantacrome, model Nova 1200e, Boynton Beach, FL) and the surface area of the nanoparticles was estimated by the BET (Brunauer–Emmett–Teller) method. Prior to the analysis, the powders were kept for 16 h at 150 C under vacuum to remove adsorbed contaminants. Material formulation and DC measurements A photopolymerizable resin mixture composed by equal parts in mols of BisGMA (bisphenol A glycidyl methacrylate) and TEGDMA (triethylene-glycol dimethacrylate) was prepared, with the addition of 0.5 wt % of camphorquinone and 0.5 wt % EDMAB (ethyl-4-dimethylamino benzoate) as photoinitiators (all chemical from Sigma–Aldrich, Milwaukee, WI). Density values of the nanoparticles were determined in a Helium picnometer (Ultrapyc 1200e, Quantachrome Instruments, Boynton Beach, FL) and used to formulate the resin-based materials with the same calcium phosphate content by volume (Table II). Four different materials were prepared by incorporating 20 % in volume of one of the nanoparticles described in Table I, with the use of a mechanical mixer under vacuum (Speedmixer DAC 150.1 FVZ-K, FlackTek, Landrum, SC). DC (n 5 3) was obtained through Fourier-transformed near-infrared (NIR) spectroscopy (Vertex 70; Bruker Optik
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | NOV 2015 VOL 103B, ISSUE 8
1671
of emissions with wavelengths at 184 and 177 nm, respectively.
TABLE II. Median Particle Size, Surface Area, and True Density of the Calcium Phosphates
Particle size (D50, nm) Surface area (m2/g) True density (g/cm3)
ACP
DCPA
DCPD
b-TCP
255 30.9 3.57
207 5.5 2.92
637 2.7 2.39
496 6.1 3.09
GmbH, Ettlingen, Germany). Specimens (0.8 mm thick and 8 mm in diameter) were built using poly(dimethylsiloxane) molds (Rapid System Pack; Colte`ne/Whaledent, Altst€atten, Switzerland). The mold was placed on a glass slide, filled with the composite and covered with a second glass slide. The spectrum of the nonpolymerized composite was obtained by the coaddition of 32 scans at 4 cm21 resolution in absorbance mode. Then, the composite was light-cured through the glass slide for 40 s, using an LED unit with 1200 mW/cm2 irradiance (Radii Cal, SDI, Bayswater, Australia). The specimens were kept in the mold and between the glass slides during storage for 24 h at 37 C. After that, a new spectrum was obtained and DC was calculated as the ratio between the areas under the peak located at 6165 cm21, corresponding to the 5C-H absorption, according to the formula: DC51003
polimerized non-polymerized
Calcium and phosphate release Calcium and phosphate release was determined by inductively coupled plasma - optical emission spectrometry (ICPOES, Agilent Technologies, Santa Clara, CA). Briefly, this analytical technique uses radiofrequency to produce a magnetic field responsible for ionizing argon gas, creating a hightemperature plasma. Atoms or ions in the sample collide with electrons and ions in the plasma, reaching an excited state. On returning to a relaxed state, they emit photons at characteristic wavelengths, and their intensity is related to the concentration of the element in the analyzed solution. Photons with a particular wavelength (i.e., characteristic of the element of interest) are captured by a photodetector and converted to an electrical signal. Disk-shaped specimens (5 3 1 mm, n 5 6) were drystored at 37 C for 24 h. Then, they were individually immersed in 5 mL of NaCl solution (133 mmol/L) buffered to pH 5.5 with 50 mmol/L acetic acid or to pH 7.0 with 50 mmol/L HEPES.20 The specimens were transferred to new vials with fresh solution after 7, 14, 21, and 28 days. Prior to the ICP-OES analysis, the samples were filtered (pore size: 3 mm) and acidified with 5 mL of 10% nitric acid (pH 5 0.2). Calibration curves were built using solutions containing 0, 0.6, 1.2, 2, 4, 5, 10, 15, 20, 25, and 30 ppm of calcium and 0, 2, 4, 6, 10, 20, and 30 ppm of phosphorous, both prepared by dilution of 1000 ppm standards (SpecSol Quimlab, Jacareı, SP, Brazil). Calcium and phosphorous readings were obtained simultaneously, based on the detection
1672
RODRIGUES ET AL.
Statistical analysis DC data was subjected to one-way ANOVA, while calcium and phosphate release data was subjected to three-way ANOVA (type of nanoparticle, pH and immersion period as main factors). In both cases, Tukey test was used for multiple comparisons and the global significance level was set at 0.05. RESULTS
Characterization of calcium phosphates Difractograms of the four nanoparticles are shown in Figure 1. The amorphous structure of ACP is evidenced by the absence of defined peaks, while the other materials revealed well-defined crystalline planes. For the crystalline materials, compound identification was confirmed using the ICDD/ JCPDS (International Center for Diffraction Data/Joint Committee on Powder Diffraction Standards—PDF 09–0080: DCPA, 09–0077: DCPD, and 09–0359: b-TCP). The images of the nanoparticles under TEM revealed distinct micromorphologies (Figure 2). ACP presented several interconnected round structures, approximately 80 nm in diameter, with globular empty spaces in their interior. DCPA presented irregularly rounded particles with approximately 120 nm in their long axis, also with empty round structures in the interior. DCPD presented highly agglomerated rod-like particles with approximately 60 nm. b-TCP displayed plate-like crystals, with approximately 400 nm in the longest dimension. Particle size distribution is shown in Figure 3, and the median (D50) particle size is shown in Table II. ACP and DCPA presented similar D50 values, but different distributions, while DCPD and b-TCP presented larger agglomerates, with similar D50 and overall distribution. Surface area results are shown in Table II. ACP presented surface area 5 to 11 times higher than the other nanoparticles. DCPA and TCP presented similar surface areas, about twice the surface area of DCPD. DC and calcium/phosphate release of experimental resin-based materials The four materials reached statistically similar degrees of conversion, with average values between 80 and 82% (p 5 0.577). Calcium release as a function of type of calcium phosphate, immersion time and pH of the immersion medium is shown in Table III. None of the interactions were statistically significant, neither was the effect of pH (p > 0.05). Calcium release varied only among materials and immersion periods (p < 0.001). Figure 4 shows the pooled averages, standard deviations and results of statistical analysis for the type of calcium phosphate only (regardless of immersion period and pH). The ACP-containing material released approximately twice the amount of calcium than the other materials. The pooled results for immersion periods only are shown in Figure 5. Successive reductions in calcium release were observed between 7 and 14 days, and subsequently after 21 days. After 28 days, the concentration
ION RELEASE FROM RESIN-BASED MATERIALS WITH CALCIUM PHOSPHATES
ORIGINAL RESEARCH REPORT
FIGURE 1. XRD of the calcium phosphate nanoparticles tested in the study. A: ACP, B: DCPA, C: DCPD, D: b-TCP
FIGURE 2. TEM images of the nanoparticles. A: ACP, B: DCPA, C: DCPD, D: b-TCP
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | NOV 2015 VOL 103B, ISSUE 8
1673
FIGURE 3. Particle size distribution of the nanoparticles, determined by DLS. A: ACP, B: DCPA, C: DCPD, D: b-TCP. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
of calcium in the solution increased to a level similar to the amount measured at 14 days. For phosphate release, the second-order interaction was statistically significant (p < 0.05, Table IV). With one exception (b-TCP after 28 days), the pH of the immersion medium did not affect phosphate release. Materials containing ACP and DCPD (both pHs) and DCPA (pH 5.5) presented a decrease in phosphate release between 7 and 14 days and sustained statistically similar levels after that. Materials containing b-TCP (both pHs) and DCPA (pH 7.0) released statistically similar amounts of phosphate during the entire
experiment. The ACP-containing material released more phosphate than the other materials, regardless of the pH or the immersion period. With a few exceptions, the other materials released similar amounts of phosphate at any given pH and immersion period. DISCUSSION
The commercial nanoparticles used to formulate the resinbased materials had their crystalline structures confirmed by the XRD, except ACP, as expected, whose diffractogram is characterized by a broad peak located between 20 and
TABLE III. Averages and Standard Deviations for Calcium Release (in mmol/L) from Resin-Based Materials Containing 20 vol % of Calcium Phosphate Nanoparticles, as a Function of Type of Nanoparticle, pH of the Immersion Media and Immersion Period ACP pH 5.5 7 days 14 days 21 days 28 days
0.63 0.32 0.18 0.35
(0.23)A (0.21)AB (0.02)AB (0.23)AB
DCPA pH 7.0
0.60 0.33 0.19 0.35
(0.21)A (0.17)AB (0.02)AB (0.21)AB
pH 5.5 0.38 0.18 0.04 0.19
(0.22)AB (0.18)AB (0.01)B (0.19)AB
DCPD pH 7.0
0.30 0.20 0.04 0.15
(0.17)AB (0.14)AB (0.02)B (0.16)AB
pH 5.5 0.36 0.17 0.03 0.17
(0.20)AB (0.17)AB (0.02)B (0.15)AB
b-TCP pH 7.0
0.25 0.16 0.03 0.21
(0.15)AB (0.15)AB (0.00)B (0.19)AB
pH 5.5 0.25 0.17 0.03 0.27
(0.25)AB (0.16)AB (0.01)B (0.34)AB
pH 7.0 0.25 0.13 0.02 0.15
(0.18)AB (0.10)AB (0.01)B (0.14)AB
Three-way ANOVA detected a significant effect of type of calcium phosphate and immersion period only. Reader must refer to text and Figures 4 and 5 for detailed description of the statistical analysis. Similar superscripts indicate lack of statistically significant differences among individual groups (three-way ANOVA/Tukey test, p > 0.05)
1674
RODRIGUES ET AL.
ION RELEASE FROM RESIN-BASED MATERIALS WITH CALCIUM PHOSPHATES
ORIGINAL RESEARCH REPORT
FIGURE 4. Pooled averages and standard-deviations for calcium release from resin-based specimens containing 20 vol % of calcium phosphate nanoparticles, as a function of nanoparticle type. The horizontal line connects calcium phosphates that showed statistically similar calcium release (ANOVA/Tukey test, p > 0.05).
40 .21 Their morphologies, primary particle sizes, and particle size distributions were also different. One limitation of using proprietary calcium phosphates is that manufacturers may add modifiers during the synthesis to tailor crystalline habit, particle size, reduce agglomeration, and modify their properties.21 Therefore, the results of the present study must not be generalized for all representatives of a specific type of calcium phosphate. The tested materials reached statistically similar degrees of conversion, meaning that the addition of 20 vol % of different calcium phosphates to the resin matrix did not affect polymer network formation. This is particularly important because crosslinking density affects water sorption by the polymer,22 which could be a confounding factor in the study. Calcium and phosphate release is usually determined by spectrophotometry UV-Vis10,12,23 or atomic emission spectroscopy.11 In both methods, the sample is acidified to make sure it is fully ionized. However, it is possible that some of the calcium and phosphate released from the material was not in their ionic forms, as nanoparticles or small agglomerates detached from the surface of the specimen remained in suspension until the sample was acidified. For instance, available data on the critical pH of DCPD and b-TCP indicate that the immersion medium should be undersaturated for both salts at pH below 6.0.16 In other words, both forms of calcium phosphate should be dissociated at pH 5.5, but not at pH 7.0. The fact that at neutral pH some of the released calcium phosphates remain in their nondissociated forms is clinically relevant, as the ionic species must be available for remineralization to occur.16,24 Notwithstanding, these nondissociated calcium phosphates can still be useful, for example, if they ionize during a pH challenge. Previous studies reported significantly higher calcium and phosphate cumulative release at pH 5.5 or 6.0, in comparison to neutral conditions.15,20,25 In this study, however, the difference in calcium and phosphate release between both pH conditions was only numerical, except for phosphate release from b-TCP after 28 days. Data in the present
work are not cumulative and, for this reason, cannot be directly compared to the abovementioned studies. Nevertheless, if the partial weekly values are added, cumulative calcium and phosphate release at pH 5.5 was, in average, 14 and 25% higher, respectively, than at pH 7.0. The behavior of dimethacrylate-based polymers in terms of water sorption and solubility is not significantly affected by pH conditions.26 Likewise, mechanical properties are not affected by prolonged storage at low pH,25 suggesting that the higher release under more acidic conditions is not a consequence of resin matrix degradation. On the other hand, ceramic materials are susceptible to water erosion, which is accelerated in acidic environments.27 That may be valid to calcium phosphate agglomerates at surface of the specimen, explaining the trend for higher release at pH 5.5. Among the tested materials, the one containing ACP released significantly higher amounts of calcium and phosphate. DCPA, DCPD, and b-TCP released statistically similar amounts of calcium and, with a few exceptions, also phosphate. This finding suggests that calcium and phosphate release was not related to the solubility of the calcium phosphates. If that was the case, the material containing ACP would release the lowest amount of ions, while a higher release could be expected from the material containing DCPD.19 Instead, calcium and phosphate release seems to be determined by the surface area of the nanoparticles. The surface area of ACP nanoparticles was 5–11 times higher than the other calcium phosphates, which provides a larger interface with the immersion medium and facilitates both ion release and the detachment of the nanoparticle agglomerates from the specimen surface. Previous studies also verified that an increase in surface area of DCPA particles, ranging from 0.17 to 18.6 m2/g significantly increased the cumulative ion release after 56 days.13,28 Particle size measurements partially agree with BET measurements, except for the fact that DCPA showed similar median particle size
FIGURE 5. Pooled averages and standard-deviations for calcium release from resin-based specimens containing 20 vol% of calcium phosphate nanoparticles, as a function of immersion time in saline solution. Similar letters indicate absence of statistically significant differences among groups (ANOVA/Tukey test, p > 0.05).
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | NOV 2015 VOL 103B, ISSUE 8
1675
A3 A2 A2 A2a (0.03) (0.00) (0.00) (0.00) 0.08 0.05 0.04 0.04
pH 7.0
A Bb Bb Bbc 0.25 0.08 0.06 0.07
(0.00) (0.01) (0.01) (0.01)
b
pH 5.5
0.13 0.05 0.04 0.04
(0.02) (0.01) (0.00) (0.00)
A A2 A2 A2
23
pH 7.0
0.18 0.06 0.05 0.05
(0.04) (0.02) (0.01) (0.01)
A Bb Bb Bc
b
pH 5.5
0.51 0.28 0.31 0.31
(0.02) (0.02) (0.02) (0.04)
A B1 B1 B1
1
pH 7.0
RODRIGUES ET AL.
7 days 14 days 21 days 28 days
0.60 0.29 0.30 0.31
(0.03) (0.01) (0.03) (0.04)
A Ba Ba Ba
a
pH 5.5
In the same column, similar upper-case letters indicate lack of statistically significant difference between averages (three-way ANOVA/Tukey test, p > 0.05). In the same line, similar superscript letters indicate lack of statistically significant differences between materials for pH 5.5 In the same line, similar superscript numbers indicate lack of statistically significant differences or pH 7.0 a Indicates statistically significant difference between pHs.
(0.00) (0.00) (0.00) (0.15) (0.04) (0.00) (0.00) (0.00) 0.19 0.08 0.06 0.08
pH 7.0
A B2 B2 B2
2
0.07 0.05 0.04 0.16
pH 5.5
AB Bb Bb Aba
c
b-TCP DCPD DCPA ACP
22 TABLE IV. Averages and Standard Deviations for Phosphate (PO32 4 or HPO4 ) Release (in mmol/L) from Resin-Based Materials Containing 20 vol % of Calcium Phosphate Nanoparticles as a Function of Nanoparticle Type, pH of the Immersion Media and Immersion Period
1676
to ACP, but much smaller surface area. The ACP microstructure shows a network of interconnected globules, which increases the porosity within the agglomerate and, consequently, its surface area. DCPA agglomerates probably have lower internal porosity, which would explain their lower surface area. Another interesting aspect is that the Ca/P ratio detected in the solution did not match the ratio of the calcium phosphate salts added to the resin matrix. Except for the material formulated with ACP, the Ca/P ratios found in the solutions were up to 3.5 times higher than those of the initial salts (Figure 6). This is a common finding in similar studies.10,11,20,23 Also, Ca/P ratio fluctuated during the study, most noticeably for the crystalline calcium phopshates. At 21 days, the decrease in Ca/P ratio found for all materials occurred due to the overall reduction in calcium release. Calcium and phosphate concentrations in the immersion media can change due to precipitation of hydroxyapatite.10,28 However, hydroxyapatite precipitation would consume more calcium than phosphate, causing the Ca/P in the solution to drop below the ratio of the initial compounds, as its Ca/P is 1.67.21 Another explanation for the higher calcium release is that in the crystalline structures of calcium orthophosphates, calcium ions occupy the interstices of the PO4 tetrahedra and, due to the ionic nature of the bond between calcium and phosphate, the former is more readily removed from the crystal lattice. In this study, calcium and phosphate were reported in a weekly release basis, instead of using cumulative values. Cumulative values must be analyzed carefully, as in the oral environment ions are constantly being rinsed away by the salivary flow. Although weekly release values also do not accurately represent the concentrations of calcium and phosphate available for remineralization at a pH challenge event, it indicates more objectively the levels released during that particular time interval. Converting the data of this study into cumulative values, the results for calcium and
FIGURE 6. Calcium : phosphate ratio calculated from the average ionic concentrations released by each material containing different calcium orthophosphates during the 28-day period. Data obtained at pH 5.5 and 7.0 were averaged. The Ca/P ratios of the initial salts are: 1.15–1.38 for ACP, 1.1 for DCPA and DCPD and 1.5 for b-TCP.
ION RELEASE FROM RESIN-BASED MATERIALS WITH CALCIUM PHOSPHATES
ORIGINAL RESEARCH REPORT
phosphate release (in mmol/L) after 28 days at pH 5.5 are, respectively, ACP: 1.48 and 1.50, DCPD: 0.73 and 0.46, DCPA: 0.79 and 0.34, and b-TCP: 0.72 and 0.32. These values are in the same range of those reported in other studies for similar periods and pH conditions,20,25 and are compatible to values reported in remineralization studies.10,11 However, it is important to emphasize that between seven and fourteen days overall calcium release dropped by approximately 50% and at 21 days it averaged less than 0.1 mmol/ L. Likewise, except for the material with ACP, the average phosphate release between 14 and 28 days was below 0.1 mmol/L. As a reference, the concentrations of ionized calcium and phosphate found in the human stimulated whole saliva were 0.59 and 2.53 mmol/L, respectively,29 suggesting that the contribution of phosphate release from resinbased materials for creating supersaturated levels in the oral environment is questionable. Other variables may influence calcium and phosphate release from resin-based materials, for example, the hydrophilicity of the organic matrix30 and the percentage of calcium phosphate nanoparticles present in the material.25 Laboratorial studies on calcium and phosphate release and remineralization of artificial caries lesions are very limited in face of the complexities of the oral environment. Clinically, remineralization is an incredibly complex process involving ever-changing pH and temperature conditions, interactions between calcium and phosphate ions and salivary proteins, presence of fluoride, biofilm, and salivary flux. In conclusion, the null hypothesis of the present study can be partially rejected, as calcium and phosphate release were significantly influenced by the type of calcium phosphate, but not by the pH of the immersion media. The surface area of the nanoparticles was the most determinant factor. When added to a resin matrix, DCPD, DCPA, and b-TCP were equally efficient as ion sources, regardless of their different solubilities. The main advantage of ACP (a calcium phosphate phase favored in several commercially available products) is its high surface area, which seemed to be the primary factor responsible for its superior ion release. Further studies are necessary to find out if the apparently low levels of calcium and phosphate released after a few weeks effectively contribute to the remineralization of caries lesions. ACKNOWLEDGMENTS
The authors would like to thank Douglas Nesadal de Souza and Dr. Alyne Sim~ oes (University of S~ao Paulo, Department of Biomaterials and Oral Biology) for the technical support with the ion release measurements, Giancarlo Brito (University of S~ao Paulo Institute of Physics) and Thiago Hewer (University of S~ao Paulo, Polytechnic School) for their help with X-ray diffraction and surface area analyses, and Sergio Hiroshi Toma (Department of Fundamental Chemistry, University of S~ao Paulo) for helping with the DLS analysis. REFERENCES 1. Legeros RZ, Retief H, Sundstrom B, Harris BE, Penugonda B. Physicochemical properties of fluoridized human-enamel. J Dent Res 1984;63:185–185.
2. Antonucci JM, Misra DN, Peckoo RJ. The accelerative and adhesive bonding capabilities of surface-active accelerators. J Dent Res 1981;60:1332–1342. 3. Silva KG, Pedrini D, Delbem AC, Ferreira L, Cannon M. In situ evaluation of the remineralizing capacity of pit and fissure sealants containing amorphous calcium phosphate and/or fluoride. Acta Odontol Scand 2010;68:11–18. 4. Uysal T, Amasyali M, Koyuturk AE, Ozcan S, Sagdic D. Amorphous calcium phosphate-containing orthodontic composites. Do they prevent demineralisation around orthodontic brackets? Aust Orthod J 2010;26:10–15. 5. Skrtic D, Hailer AW, Takagi S, Antonucci JM, Eanes ED. Quantitative assessment of the efficacy of amorphous calcium phosphate/ methacrylate composites in remineralizing caries-like lesions artificially produced in bovine enamel. J Dent Res 1996;75:1679–1686. 6. Weir MD, Chow LC, Xu HH. Remineralization of demineralized enamel via calcium phosphate nanocomposite. J Dent Res 2012; 91:979–984. 7. Reynolds EC, Cai F, Cochrane NJ, Shen P, Walker GD, Morgan MV, Reynolds C. Fluoride and casein phosphopeptide-amorphous calcium phosphate. J Dent Res 2008;87:344–348. 8. Vogel GL. Oral fluoride reservoirs and the prevention of dental caries. Monogr Oral Sci 2011;22:146–157. 9. Vogel GL, Tenuta LM, Schumacher GE, Chow LC. No calciumfluoride-like deposits detected in plaque shortly after a sodium fluoride mouthrinse. Caries Res 2010;44:108–115. 10. Dickens SH, Flaim GM, Takagi S. Mechanical properties and biochemical activity of remineralizing resin-based Ca-PO4 cements. Dent Mater 2003;19:558–566. 11. Langhorst SE, O’Donnell JN, Skrtic D. In vitro remineralization of enamel by polymeric amorphous calcium phosphate composite: Quantitative microradiographic study. Dent Mater 2009;25:884–891. 12. Melo MA, Weir MD, Rodrigues LK, Xu HH. Novel calcium phosphate nanocomposite with caries-inhibition in a human in situ model. Dent Mater 2013;29:231–240. 13. Xu HH, Weir MD, Sun L. Nanocomposites with Ca and PO4 release: Effects of reinforcement, dicalcium phosphate particle size and silanization. Dent Mater 2007;23:1482–1491. 14. Xu HH, Carey LE, Simon CG, Jr. Premixed macroporous calcium phosphate cement scaffold. J Mater Sci Mater Med 2007;18:1345– 1353. 15. Xu HH, Weir MD, Sun L. Calcium and phosphate ion releasing composite: Effect of pH on release and mechanical properties. Dent Mater 2009;25:535–542. 16. Larsen MJ, Pearce EI. Saturation of human saliva with respect to calcium salts. Arch Oral Biol 2003;48:317–322. 17. van Kemenade MJJM, de Bruyn PL. A kinetic study of precipitation from supersaturated calcium phosphate solutions. J Colloid Interface Sci 1987;118:564–585. 18. Elliott JC. Structure and Chemistry of the Apatites and Other Calcium Orthophosphates. Amsterdam, The Netherlands: Elsevier; 1994. 19. Dorozhkin SV. Calcium orthophosphates—Occurrence, properties, biomineralization, pathological calcification and biomimetic applications. Biomaterr 2011;1:121–164. 20. Xu HH, Moreau JL. Dental glass-reinforced composite for caries inhibition: Calcium phosphate ion release and mechanical properties. J Biomed Mater Res B Appl Biomater 2010;92:332–340. 21. LeGeros RZ. Calcium Phosphates in Oral Biology and Medicine. Karger; 1991. 22. Schneider LFJ, Consani S, Sakaguchi RL, Ferracane JL. Alternative photoinitiator system reduces the rate of stress development without compromising the final properties of the dental composite. Dental Mater 2009;25:566–572. 23. Skrtic D, Antonucci JM, Eanes ED. Improved properties of amorphous calcium phosphate fillers in remineralizing resin composites. Dent Mater 1996;12:295–301. 24. Cochrane NJ, Cai F, Huq NL, Burrow MF, Reynolds EC. New approaches to enhanced remineralization of tooth enamel. J Dent Res 2010;89:1187–1197. 25. Xu HH, Moreau JL, Sun L, Chow LC. Nanocomposite containing amorphous calcium phosphate nanoparticles for caries inhibition. Dent Mater 2011;27:762–769.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | NOV 2015 VOL 103B, ISSUE 8
1677
26. Ortengren U, Andersson F, Elgh U, Terselius B, Karlsson S Influence of pH and storage time on the sorption and solubility behaviour of three composite resin materials. J Dent 2001;29:35– 41. 27. Milleding P, Karlsson S, Nyborg L. On the surface elemental composition of non-corroded and corroded dental ceramic materials in vitro. J Mater Sci Mater Med 2003;14:557–566. 28. Xu HH, Weir MD, Sun L, Ngai S, Takagi S, Chow LC. Effect of filler level and particle size on dental caries-
1678
RODRIGUES ET AL.
inhibiting Ca-PO(4) composite. J Mater Sci Mater Med 2009;20: 1771–1779. 29. Gron P. State of calcium and inorganic orthophosphate in human saliva. Arch Oral Biol 1973;18:1365–1378. 30. Antonucci JM, Skrtic D. Fine-tuning of polymeric resins and their interfaces with amorphous calcium phosphate. A strategy for designing effective remineralizing dental composites. Polymers (Basel). 2010;2:378–392.
ION RELEASE FROM RESIN-BASED MATERIALS WITH CALCIUM PHOSPHATES
The surface area of ACP was 5–11 times higher than those of the other nanoparticles. Materials showed similar DC. The material containing ACP released significantly more ions than the others, which released similar amounts of calcium and, in most cases, phosphate. Ion release was not affected by pH. Calcium release decreased between 7 and 21 days, while phosphate levels remained constant after 14 days. In conclusion, ACP higher ion release can be ascribed to its high surface area. DCPA, DCPD, and b-TCP C 2015 had similar performances as ion-releasing fillers. V Wiley Periodicals, Inc. J Biomed Mater Res Part B: Appl Biomater, 103B: 1670–1678, 2015.
Key Words: calcium phosphate, resin composite, characterization, ion release, degree of conversion
How to cite this article: Rodrigues MC, Natale LC, Arana-Chaves VE, Braga RR. 2015. Calcium and phosphate release from resin-based materials containing different calcium orthophosphate nanoparticles. J Biomed Mater Res Part B 2015:103B:1670–1678.
INTRODUCTION
Resin-based materials with ion-releasing capabilities have been investigated since the early years of resin composites use.1,2 Recently, luting and restorative composites, as well as liners and sealants containing calcium phosphates became available for clinical use.3,4 Manufacturers claim that these materials can provide protection against caries by offering an extra source of calcium and phosphate ions to create a supersaturated environment that would favor remineralization. In fact, in vitro studies using experimental materials reported their ability to remineralize enamel subsurface lesions. For example, caries-like lesions in contact with a composite containing 40% (by weight) of ACP recovered 38% of the mineral content after 2 weeks in a pHcycling regimen.5 A recent study observed 22% of enamel remineralization after 30 days of pH cycling with the use of a composite containing 40 wt % of ACP nanoparticles, in comparison to 6% with a commercial fluoride-releasing composite.6 Calcium and phosphate ions are also fundamental in the remineralization process involving fluorides.7 They are necessary to maintain fluoride concentration at cariostatic lev-
els after topic fluoride treatments, either in the form of “calcium fluoride-like” (phosphate-stabilized calcium fluoride) deposits on the tooth surface or as bacterial bound fluoride in the dental plaque.8 These biological reservoirs, particularly calcium fluoride-like deposits, need an additional source of calcium to be formed at effective concentrations.9 The ionic concentration released from bioactive composites and the corresponding effect on remineralization has also been studied. In demineralized dentin, resin cements containing 73–78 wt % of dicalcium phosphate anhydrous (DCPA) and tetracalcium phosphate (TTCP) particles released 0.3–0.5 mmol/L of calcium and 0.05–0.1 mmol/L of phosphate, and remineralized 38–47% of the lesion after 5 weeks.10 In enamel, a composite containing 40 wt % of ACP particles released 0.74 mmol/L of calcium and 0.54 mmol/L of phosphate, remineralizing 14% of the lesion after 30 days, in comparison to 4% achieved by a fluoride-containing commercial cement.11 Recently, an in situ study verified that under accumulated biofilm, a composite containing 40 wt % of ACP nanoparticles was able to reduce mineral loss at the enamel/restoration interface
Correspondence to: R. R. Braga; e-mail: [email protected] ~o Paulo Research Foundation); contract grant number: 2012/04532-4 Contract grant sponsor: FAPESP (Sa
1670
C 2015 WILEY PERIODICALS, INC. V
ORIGINAL RESEARCH REPORT
TABLE I. Calcium Phosphates Used in the Study Name (Manufacturer) Amorphous calcium phosphate (Sigma–Aldrich, Milwaukee, WI) Dicalcium phosphate anhydrous (Synth, Diadema, SP, Brazil) Dicalcium phosphate dihydrate (Synth, Diadema, SP, Brazil) b-Tricalcium phosphate (Sigma–Aldrich, Milwaukee, WI)
by 2.4 times compared to a composite containing only barium glass. The cumulative release of calcium and phosphate at pH 5.5 after 14 days (corresponding to the duration of the in situ study) was approximately 4 and 1 mmol/L, respectively.12 As demonstrated with the abovementioned studies, calcium phosphates with different particle sizes and calciumto-phosphate (Ca/P) ratios have been tested as ion-releasing fillers and, therefore, their results are not comparable. Reducing particle size and consequently, increasing its surface area has been shown to significantly increase ion release.13,14 The availability of ions is also influenced by the pH of the immersion medium. A fourfold increase in calcium and phosphate release was observed with a composite containing 75 wt % of TTCP between pH 7.4 and 4.0.15 Furthermore, calcium phosphates have different critical pHs (i.e., pH at which the solution is saturated with respect to a particular calcium phosphate and below which it starts to dissolve), which also may affect ion release. For example, while saliva becomes saturated with hydroxyapatite at pH 5.5, DCPD starts to dissolve at pHs below 6.3 and b-TCP at 6.1.16 The type of calcium phosphate may also affect the concentration of ions released in solution. In general, the solubility of calcium phosphates increases at low pH.17 However, they show very low solubility in aqueous media at neutral pH. Actually, some are virtually insoluble.18 In order of decreasing solubility, dicalcium phosphate dihydrate (DCPD) is more soluble than DCPA, followed by ACP and btricalcium phosphate (b-TCP).19 It is interesting to notice that the majority of studies focus on variables other than calcium phosphate type, for example, nanoparticle content,15 ratio between bioactive and reinforcing fillers,20 or material presentation10 and no direct comparison between resinbased materials containing different types of calcium phosphates is available. Based on the above, the purpose of the present study was to compare the concentration of calcium and phosphate released from resin-based materials containing nanoparticles of different calcium orthophosphates, at pH 5.5 and 7.0. To substantiate the discussion, the nanoparticles were characterized in terms of crystallinity, micromorphology, particle-size distribution, surface area, and true density. Additionally, the DC of the resin-based materials was determined. The null hypothesis was that the nanoparticle characteristics or the pH of the immersion medium do not influence calcium and phosphate release.
Abbreviation
Chemical Formula
Ca/P
ACP
CaxHy(PO4)znH2O
1.15–1.38
DCPA DCPD b-TCP
CaHPO4 CaHPO4 2H2O Ca3(PO4)2
1.1 1.1 1.5
METHODS
Characterization of calcium phosphates Four commercially available calcium phosphate nanoparticles were chosen for this study (Table I). The type of structure (crystalline or amorphous) was identified by X-ray diffraction (XRD), using nickel filtered CuKa radiation at 40 kV and 30 mA (MultiFlex, Rigaku, Tokyo, Japan). The equipment geometry was h/2h. Readings were continuous at angles from 10 to 60 at 0.05 intervals, 10 s per interval. Nanoparticle morphology was observed under transmission electron microscopy (TEM; JEOL, model 1010, Tokyo, Japan), using an accelerating voltage of 80 kV. Powders were dispersed in isopropyl alcohol and, after 24 h, a few drops of the supernatant were placed on nickel grids (200 mesh) covered with a poly(vinyl formal) pellicle (Formvar). Nanoparticle dimensions were measured using the ImageJ software (National Institute of Health, Bethesda, MD). Twenty nanoparticles were measured for each material. Additionally, particle-size distribution was estimated by dynamic light scattering (DLS, Nanotrac 252, Microtrac, Montgomeryville, PA). The powders were dispersed in isopropyl alcohol and sonicated for 10 min. Then, after 20 min, the supernatant was collected for analysis. Nitrogen adsorption isotherms were obtained (Quantacrome, model Nova 1200e, Boynton Beach, FL) and the surface area of the nanoparticles was estimated by the BET (Brunauer–Emmett–Teller) method. Prior to the analysis, the powders were kept for 16 h at 150 C under vacuum to remove adsorbed contaminants. Material formulation and DC measurements A photopolymerizable resin mixture composed by equal parts in mols of BisGMA (bisphenol A glycidyl methacrylate) and TEGDMA (triethylene-glycol dimethacrylate) was prepared, with the addition of 0.5 wt % of camphorquinone and 0.5 wt % EDMAB (ethyl-4-dimethylamino benzoate) as photoinitiators (all chemical from Sigma–Aldrich, Milwaukee, WI). Density values of the nanoparticles were determined in a Helium picnometer (Ultrapyc 1200e, Quantachrome Instruments, Boynton Beach, FL) and used to formulate the resin-based materials with the same calcium phosphate content by volume (Table II). Four different materials were prepared by incorporating 20 % in volume of one of the nanoparticles described in Table I, with the use of a mechanical mixer under vacuum (Speedmixer DAC 150.1 FVZ-K, FlackTek, Landrum, SC). DC (n 5 3) was obtained through Fourier-transformed near-infrared (NIR) spectroscopy (Vertex 70; Bruker Optik
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | NOV 2015 VOL 103B, ISSUE 8
1671
of emissions with wavelengths at 184 and 177 nm, respectively.
TABLE II. Median Particle Size, Surface Area, and True Density of the Calcium Phosphates
Particle size (D50, nm) Surface area (m2/g) True density (g/cm3)
ACP
DCPA
DCPD
b-TCP
255 30.9 3.57
207 5.5 2.92
637 2.7 2.39
496 6.1 3.09
GmbH, Ettlingen, Germany). Specimens (0.8 mm thick and 8 mm in diameter) were built using poly(dimethylsiloxane) molds (Rapid System Pack; Colte`ne/Whaledent, Altst€atten, Switzerland). The mold was placed on a glass slide, filled with the composite and covered with a second glass slide. The spectrum of the nonpolymerized composite was obtained by the coaddition of 32 scans at 4 cm21 resolution in absorbance mode. Then, the composite was light-cured through the glass slide for 40 s, using an LED unit with 1200 mW/cm2 irradiance (Radii Cal, SDI, Bayswater, Australia). The specimens were kept in the mold and between the glass slides during storage for 24 h at 37 C. After that, a new spectrum was obtained and DC was calculated as the ratio between the areas under the peak located at 6165 cm21, corresponding to the 5C-H absorption, according to the formula: DC51003
polimerized non-polymerized
Calcium and phosphate release Calcium and phosphate release was determined by inductively coupled plasma - optical emission spectrometry (ICPOES, Agilent Technologies, Santa Clara, CA). Briefly, this analytical technique uses radiofrequency to produce a magnetic field responsible for ionizing argon gas, creating a hightemperature plasma. Atoms or ions in the sample collide with electrons and ions in the plasma, reaching an excited state. On returning to a relaxed state, they emit photons at characteristic wavelengths, and their intensity is related to the concentration of the element in the analyzed solution. Photons with a particular wavelength (i.e., characteristic of the element of interest) are captured by a photodetector and converted to an electrical signal. Disk-shaped specimens (5 3 1 mm, n 5 6) were drystored at 37 C for 24 h. Then, they were individually immersed in 5 mL of NaCl solution (133 mmol/L) buffered to pH 5.5 with 50 mmol/L acetic acid or to pH 7.0 with 50 mmol/L HEPES.20 The specimens were transferred to new vials with fresh solution after 7, 14, 21, and 28 days. Prior to the ICP-OES analysis, the samples were filtered (pore size: 3 mm) and acidified with 5 mL of 10% nitric acid (pH 5 0.2). Calibration curves were built using solutions containing 0, 0.6, 1.2, 2, 4, 5, 10, 15, 20, 25, and 30 ppm of calcium and 0, 2, 4, 6, 10, 20, and 30 ppm of phosphorous, both prepared by dilution of 1000 ppm standards (SpecSol Quimlab, Jacareı, SP, Brazil). Calcium and phosphorous readings were obtained simultaneously, based on the detection
1672
RODRIGUES ET AL.
Statistical analysis DC data was subjected to one-way ANOVA, while calcium and phosphate release data was subjected to three-way ANOVA (type of nanoparticle, pH and immersion period as main factors). In both cases, Tukey test was used for multiple comparisons and the global significance level was set at 0.05. RESULTS
Characterization of calcium phosphates Difractograms of the four nanoparticles are shown in Figure 1. The amorphous structure of ACP is evidenced by the absence of defined peaks, while the other materials revealed well-defined crystalline planes. For the crystalline materials, compound identification was confirmed using the ICDD/ JCPDS (International Center for Diffraction Data/Joint Committee on Powder Diffraction Standards—PDF 09–0080: DCPA, 09–0077: DCPD, and 09–0359: b-TCP). The images of the nanoparticles under TEM revealed distinct micromorphologies (Figure 2). ACP presented several interconnected round structures, approximately 80 nm in diameter, with globular empty spaces in their interior. DCPA presented irregularly rounded particles with approximately 120 nm in their long axis, also with empty round structures in the interior. DCPD presented highly agglomerated rod-like particles with approximately 60 nm. b-TCP displayed plate-like crystals, with approximately 400 nm in the longest dimension. Particle size distribution is shown in Figure 3, and the median (D50) particle size is shown in Table II. ACP and DCPA presented similar D50 values, but different distributions, while DCPD and b-TCP presented larger agglomerates, with similar D50 and overall distribution. Surface area results are shown in Table II. ACP presented surface area 5 to 11 times higher than the other nanoparticles. DCPA and TCP presented similar surface areas, about twice the surface area of DCPD. DC and calcium/phosphate release of experimental resin-based materials The four materials reached statistically similar degrees of conversion, with average values between 80 and 82% (p 5 0.577). Calcium release as a function of type of calcium phosphate, immersion time and pH of the immersion medium is shown in Table III. None of the interactions were statistically significant, neither was the effect of pH (p > 0.05). Calcium release varied only among materials and immersion periods (p < 0.001). Figure 4 shows the pooled averages, standard deviations and results of statistical analysis for the type of calcium phosphate only (regardless of immersion period and pH). The ACP-containing material released approximately twice the amount of calcium than the other materials. The pooled results for immersion periods only are shown in Figure 5. Successive reductions in calcium release were observed between 7 and 14 days, and subsequently after 21 days. After 28 days, the concentration
ION RELEASE FROM RESIN-BASED MATERIALS WITH CALCIUM PHOSPHATES
ORIGINAL RESEARCH REPORT
FIGURE 1. XRD of the calcium phosphate nanoparticles tested in the study. A: ACP, B: DCPA, C: DCPD, D: b-TCP
FIGURE 2. TEM images of the nanoparticles. A: ACP, B: DCPA, C: DCPD, D: b-TCP
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | NOV 2015 VOL 103B, ISSUE 8
1673
FIGURE 3. Particle size distribution of the nanoparticles, determined by DLS. A: ACP, B: DCPA, C: DCPD, D: b-TCP. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
of calcium in the solution increased to a level similar to the amount measured at 14 days. For phosphate release, the second-order interaction was statistically significant (p < 0.05, Table IV). With one exception (b-TCP after 28 days), the pH of the immersion medium did not affect phosphate release. Materials containing ACP and DCPD (both pHs) and DCPA (pH 5.5) presented a decrease in phosphate release between 7 and 14 days and sustained statistically similar levels after that. Materials containing b-TCP (both pHs) and DCPA (pH 7.0) released statistically similar amounts of phosphate during the entire
experiment. The ACP-containing material released more phosphate than the other materials, regardless of the pH or the immersion period. With a few exceptions, the other materials released similar amounts of phosphate at any given pH and immersion period. DISCUSSION
The commercial nanoparticles used to formulate the resinbased materials had their crystalline structures confirmed by the XRD, except ACP, as expected, whose diffractogram is characterized by a broad peak located between 20 and
TABLE III. Averages and Standard Deviations for Calcium Release (in mmol/L) from Resin-Based Materials Containing 20 vol % of Calcium Phosphate Nanoparticles, as a Function of Type of Nanoparticle, pH of the Immersion Media and Immersion Period ACP pH 5.5 7 days 14 days 21 days 28 days
0.63 0.32 0.18 0.35
(0.23)A (0.21)AB (0.02)AB (0.23)AB
DCPA pH 7.0
0.60 0.33 0.19 0.35
(0.21)A (0.17)AB (0.02)AB (0.21)AB
pH 5.5 0.38 0.18 0.04 0.19
(0.22)AB (0.18)AB (0.01)B (0.19)AB
DCPD pH 7.0
0.30 0.20 0.04 0.15
(0.17)AB (0.14)AB (0.02)B (0.16)AB
pH 5.5 0.36 0.17 0.03 0.17
(0.20)AB (0.17)AB (0.02)B (0.15)AB
b-TCP pH 7.0
0.25 0.16 0.03 0.21
(0.15)AB (0.15)AB (0.00)B (0.19)AB
pH 5.5 0.25 0.17 0.03 0.27
(0.25)AB (0.16)AB (0.01)B (0.34)AB
pH 7.0 0.25 0.13 0.02 0.15
(0.18)AB (0.10)AB (0.01)B (0.14)AB
Three-way ANOVA detected a significant effect of type of calcium phosphate and immersion period only. Reader must refer to text and Figures 4 and 5 for detailed description of the statistical analysis. Similar superscripts indicate lack of statistically significant differences among individual groups (three-way ANOVA/Tukey test, p > 0.05)
1674
RODRIGUES ET AL.
ION RELEASE FROM RESIN-BASED MATERIALS WITH CALCIUM PHOSPHATES
ORIGINAL RESEARCH REPORT
FIGURE 4. Pooled averages and standard-deviations for calcium release from resin-based specimens containing 20 vol % of calcium phosphate nanoparticles, as a function of nanoparticle type. The horizontal line connects calcium phosphates that showed statistically similar calcium release (ANOVA/Tukey test, p > 0.05).
40 .21 Their morphologies, primary particle sizes, and particle size distributions were also different. One limitation of using proprietary calcium phosphates is that manufacturers may add modifiers during the synthesis to tailor crystalline habit, particle size, reduce agglomeration, and modify their properties.21 Therefore, the results of the present study must not be generalized for all representatives of a specific type of calcium phosphate. The tested materials reached statistically similar degrees of conversion, meaning that the addition of 20 vol % of different calcium phosphates to the resin matrix did not affect polymer network formation. This is particularly important because crosslinking density affects water sorption by the polymer,22 which could be a confounding factor in the study. Calcium and phosphate release is usually determined by spectrophotometry UV-Vis10,12,23 or atomic emission spectroscopy.11 In both methods, the sample is acidified to make sure it is fully ionized. However, it is possible that some of the calcium and phosphate released from the material was not in their ionic forms, as nanoparticles or small agglomerates detached from the surface of the specimen remained in suspension until the sample was acidified. For instance, available data on the critical pH of DCPD and b-TCP indicate that the immersion medium should be undersaturated for both salts at pH below 6.0.16 In other words, both forms of calcium phosphate should be dissociated at pH 5.5, but not at pH 7.0. The fact that at neutral pH some of the released calcium phosphates remain in their nondissociated forms is clinically relevant, as the ionic species must be available for remineralization to occur.16,24 Notwithstanding, these nondissociated calcium phosphates can still be useful, for example, if they ionize during a pH challenge. Previous studies reported significantly higher calcium and phosphate cumulative release at pH 5.5 or 6.0, in comparison to neutral conditions.15,20,25 In this study, however, the difference in calcium and phosphate release between both pH conditions was only numerical, except for phosphate release from b-TCP after 28 days. Data in the present
work are not cumulative and, for this reason, cannot be directly compared to the abovementioned studies. Nevertheless, if the partial weekly values are added, cumulative calcium and phosphate release at pH 5.5 was, in average, 14 and 25% higher, respectively, than at pH 7.0. The behavior of dimethacrylate-based polymers in terms of water sorption and solubility is not significantly affected by pH conditions.26 Likewise, mechanical properties are not affected by prolonged storage at low pH,25 suggesting that the higher release under more acidic conditions is not a consequence of resin matrix degradation. On the other hand, ceramic materials are susceptible to water erosion, which is accelerated in acidic environments.27 That may be valid to calcium phosphate agglomerates at surface of the specimen, explaining the trend for higher release at pH 5.5. Among the tested materials, the one containing ACP released significantly higher amounts of calcium and phosphate. DCPA, DCPD, and b-TCP released statistically similar amounts of calcium and, with a few exceptions, also phosphate. This finding suggests that calcium and phosphate release was not related to the solubility of the calcium phosphates. If that was the case, the material containing ACP would release the lowest amount of ions, while a higher release could be expected from the material containing DCPD.19 Instead, calcium and phosphate release seems to be determined by the surface area of the nanoparticles. The surface area of ACP nanoparticles was 5–11 times higher than the other calcium phosphates, which provides a larger interface with the immersion medium and facilitates both ion release and the detachment of the nanoparticle agglomerates from the specimen surface. Previous studies also verified that an increase in surface area of DCPA particles, ranging from 0.17 to 18.6 m2/g significantly increased the cumulative ion release after 56 days.13,28 Particle size measurements partially agree with BET measurements, except for the fact that DCPA showed similar median particle size
FIGURE 5. Pooled averages and standard-deviations for calcium release from resin-based specimens containing 20 vol% of calcium phosphate nanoparticles, as a function of immersion time in saline solution. Similar letters indicate absence of statistically significant differences among groups (ANOVA/Tukey test, p > 0.05).
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | NOV 2015 VOL 103B, ISSUE 8
1675
A3 A2 A2 A2a (0.03) (0.00) (0.00) (0.00) 0.08 0.05 0.04 0.04
pH 7.0
A Bb Bb Bbc 0.25 0.08 0.06 0.07
(0.00) (0.01) (0.01) (0.01)
b
pH 5.5
0.13 0.05 0.04 0.04
(0.02) (0.01) (0.00) (0.00)
A A2 A2 A2
23
pH 7.0
0.18 0.06 0.05 0.05
(0.04) (0.02) (0.01) (0.01)
A Bb Bb Bc
b
pH 5.5
0.51 0.28 0.31 0.31
(0.02) (0.02) (0.02) (0.04)
A B1 B1 B1
1
pH 7.0
RODRIGUES ET AL.
7 days 14 days 21 days 28 days
0.60 0.29 0.30 0.31
(0.03) (0.01) (0.03) (0.04)
A Ba Ba Ba
a
pH 5.5
In the same column, similar upper-case letters indicate lack of statistically significant difference between averages (three-way ANOVA/Tukey test, p > 0.05). In the same line, similar superscript letters indicate lack of statistically significant differences between materials for pH 5.5 In the same line, similar superscript numbers indicate lack of statistically significant differences or pH 7.0 a Indicates statistically significant difference between pHs.
(0.00) (0.00) (0.00) (0.15) (0.04) (0.00) (0.00) (0.00) 0.19 0.08 0.06 0.08
pH 7.0
A B2 B2 B2
2
0.07 0.05 0.04 0.16
pH 5.5
AB Bb Bb Aba
c
b-TCP DCPD DCPA ACP
22 TABLE IV. Averages and Standard Deviations for Phosphate (PO32 4 or HPO4 ) Release (in mmol/L) from Resin-Based Materials Containing 20 vol % of Calcium Phosphate Nanoparticles as a Function of Nanoparticle Type, pH of the Immersion Media and Immersion Period
1676
to ACP, but much smaller surface area. The ACP microstructure shows a network of interconnected globules, which increases the porosity within the agglomerate and, consequently, its surface area. DCPA agglomerates probably have lower internal porosity, which would explain their lower surface area. Another interesting aspect is that the Ca/P ratio detected in the solution did not match the ratio of the calcium phosphate salts added to the resin matrix. Except for the material formulated with ACP, the Ca/P ratios found in the solutions were up to 3.5 times higher than those of the initial salts (Figure 6). This is a common finding in similar studies.10,11,20,23 Also, Ca/P ratio fluctuated during the study, most noticeably for the crystalline calcium phopshates. At 21 days, the decrease in Ca/P ratio found for all materials occurred due to the overall reduction in calcium release. Calcium and phosphate concentrations in the immersion media can change due to precipitation of hydroxyapatite.10,28 However, hydroxyapatite precipitation would consume more calcium than phosphate, causing the Ca/P in the solution to drop below the ratio of the initial compounds, as its Ca/P is 1.67.21 Another explanation for the higher calcium release is that in the crystalline structures of calcium orthophosphates, calcium ions occupy the interstices of the PO4 tetrahedra and, due to the ionic nature of the bond between calcium and phosphate, the former is more readily removed from the crystal lattice. In this study, calcium and phosphate were reported in a weekly release basis, instead of using cumulative values. Cumulative values must be analyzed carefully, as in the oral environment ions are constantly being rinsed away by the salivary flow. Although weekly release values also do not accurately represent the concentrations of calcium and phosphate available for remineralization at a pH challenge event, it indicates more objectively the levels released during that particular time interval. Converting the data of this study into cumulative values, the results for calcium and
FIGURE 6. Calcium : phosphate ratio calculated from the average ionic concentrations released by each material containing different calcium orthophosphates during the 28-day period. Data obtained at pH 5.5 and 7.0 were averaged. The Ca/P ratios of the initial salts are: 1.15–1.38 for ACP, 1.1 for DCPA and DCPD and 1.5 for b-TCP.
ION RELEASE FROM RESIN-BASED MATERIALS WITH CALCIUM PHOSPHATES
ORIGINAL RESEARCH REPORT
phosphate release (in mmol/L) after 28 days at pH 5.5 are, respectively, ACP: 1.48 and 1.50, DCPD: 0.73 and 0.46, DCPA: 0.79 and 0.34, and b-TCP: 0.72 and 0.32. These values are in the same range of those reported in other studies for similar periods and pH conditions,20,25 and are compatible to values reported in remineralization studies.10,11 However, it is important to emphasize that between seven and fourteen days overall calcium release dropped by approximately 50% and at 21 days it averaged less than 0.1 mmol/ L. Likewise, except for the material with ACP, the average phosphate release between 14 and 28 days was below 0.1 mmol/L. As a reference, the concentrations of ionized calcium and phosphate found in the human stimulated whole saliva were 0.59 and 2.53 mmol/L, respectively,29 suggesting that the contribution of phosphate release from resinbased materials for creating supersaturated levels in the oral environment is questionable. Other variables may influence calcium and phosphate release from resin-based materials, for example, the hydrophilicity of the organic matrix30 and the percentage of calcium phosphate nanoparticles present in the material.25 Laboratorial studies on calcium and phosphate release and remineralization of artificial caries lesions are very limited in face of the complexities of the oral environment. Clinically, remineralization is an incredibly complex process involving ever-changing pH and temperature conditions, interactions between calcium and phosphate ions and salivary proteins, presence of fluoride, biofilm, and salivary flux. In conclusion, the null hypothesis of the present study can be partially rejected, as calcium and phosphate release were significantly influenced by the type of calcium phosphate, but not by the pH of the immersion media. The surface area of the nanoparticles was the most determinant factor. When added to a resin matrix, DCPD, DCPA, and b-TCP were equally efficient as ion sources, regardless of their different solubilities. The main advantage of ACP (a calcium phosphate phase favored in several commercially available products) is its high surface area, which seemed to be the primary factor responsible for its superior ion release. Further studies are necessary to find out if the apparently low levels of calcium and phosphate released after a few weeks effectively contribute to the remineralization of caries lesions. ACKNOWLEDGMENTS
The authors would like to thank Douglas Nesadal de Souza and Dr. Alyne Sim~ oes (University of S~ao Paulo, Department of Biomaterials and Oral Biology) for the technical support with the ion release measurements, Giancarlo Brito (University of S~ao Paulo Institute of Physics) and Thiago Hewer (University of S~ao Paulo, Polytechnic School) for their help with X-ray diffraction and surface area analyses, and Sergio Hiroshi Toma (Department of Fundamental Chemistry, University of S~ao Paulo) for helping with the DLS analysis. REFERENCES 1. Legeros RZ, Retief H, Sundstrom B, Harris BE, Penugonda B. Physicochemical properties of fluoridized human-enamel. J Dent Res 1984;63:185–185.
2. Antonucci JM, Misra DN, Peckoo RJ. The accelerative and adhesive bonding capabilities of surface-active accelerators. J Dent Res 1981;60:1332–1342. 3. Silva KG, Pedrini D, Delbem AC, Ferreira L, Cannon M. In situ evaluation of the remineralizing capacity of pit and fissure sealants containing amorphous calcium phosphate and/or fluoride. Acta Odontol Scand 2010;68:11–18. 4. Uysal T, Amasyali M, Koyuturk AE, Ozcan S, Sagdic D. Amorphous calcium phosphate-containing orthodontic composites. Do they prevent demineralisation around orthodontic brackets? Aust Orthod J 2010;26:10–15. 5. Skrtic D, Hailer AW, Takagi S, Antonucci JM, Eanes ED. Quantitative assessment of the efficacy of amorphous calcium phosphate/ methacrylate composites in remineralizing caries-like lesions artificially produced in bovine enamel. J Dent Res 1996;75:1679–1686. 6. Weir MD, Chow LC, Xu HH. Remineralization of demineralized enamel via calcium phosphate nanocomposite. J Dent Res 2012; 91:979–984. 7. Reynolds EC, Cai F, Cochrane NJ, Shen P, Walker GD, Morgan MV, Reynolds C. Fluoride and casein phosphopeptide-amorphous calcium phosphate. J Dent Res 2008;87:344–348. 8. Vogel GL. Oral fluoride reservoirs and the prevention of dental caries. Monogr Oral Sci 2011;22:146–157. 9. Vogel GL, Tenuta LM, Schumacher GE, Chow LC. No calciumfluoride-like deposits detected in plaque shortly after a sodium fluoride mouthrinse. Caries Res 2010;44:108–115. 10. Dickens SH, Flaim GM, Takagi S. Mechanical properties and biochemical activity of remineralizing resin-based Ca-PO4 cements. Dent Mater 2003;19:558–566. 11. Langhorst SE, O’Donnell JN, Skrtic D. In vitro remineralization of enamel by polymeric amorphous calcium phosphate composite: Quantitative microradiographic study. Dent Mater 2009;25:884–891. 12. Melo MA, Weir MD, Rodrigues LK, Xu HH. Novel calcium phosphate nanocomposite with caries-inhibition in a human in situ model. Dent Mater 2013;29:231–240. 13. Xu HH, Weir MD, Sun L. Nanocomposites with Ca and PO4 release: Effects of reinforcement, dicalcium phosphate particle size and silanization. Dent Mater 2007;23:1482–1491. 14. Xu HH, Carey LE, Simon CG, Jr. Premixed macroporous calcium phosphate cement scaffold. J Mater Sci Mater Med 2007;18:1345– 1353. 15. Xu HH, Weir MD, Sun L. Calcium and phosphate ion releasing composite: Effect of pH on release and mechanical properties. Dent Mater 2009;25:535–542. 16. Larsen MJ, Pearce EI. Saturation of human saliva with respect to calcium salts. Arch Oral Biol 2003;48:317–322. 17. van Kemenade MJJM, de Bruyn PL. A kinetic study of precipitation from supersaturated calcium phosphate solutions. J Colloid Interface Sci 1987;118:564–585. 18. Elliott JC. Structure and Chemistry of the Apatites and Other Calcium Orthophosphates. Amsterdam, The Netherlands: Elsevier; 1994. 19. Dorozhkin SV. Calcium orthophosphates—Occurrence, properties, biomineralization, pathological calcification and biomimetic applications. Biomaterr 2011;1:121–164. 20. Xu HH, Moreau JL. Dental glass-reinforced composite for caries inhibition: Calcium phosphate ion release and mechanical properties. J Biomed Mater Res B Appl Biomater 2010;92:332–340. 21. LeGeros RZ. Calcium Phosphates in Oral Biology and Medicine. Karger; 1991. 22. Schneider LFJ, Consani S, Sakaguchi RL, Ferracane JL. Alternative photoinitiator system reduces the rate of stress development without compromising the final properties of the dental composite. Dental Mater 2009;25:566–572. 23. Skrtic D, Antonucci JM, Eanes ED. Improved properties of amorphous calcium phosphate fillers in remineralizing resin composites. Dent Mater 1996;12:295–301. 24. Cochrane NJ, Cai F, Huq NL, Burrow MF, Reynolds EC. New approaches to enhanced remineralization of tooth enamel. J Dent Res 2010;89:1187–1197. 25. Xu HH, Moreau JL, Sun L, Chow LC. Nanocomposite containing amorphous calcium phosphate nanoparticles for caries inhibition. Dent Mater 2011;27:762–769.
JOURNAL OF BIOMEDICAL MATERIALS RESEARCH B: APPLIED BIOMATERIALS | NOV 2015 VOL 103B, ISSUE 8
1677
26. Ortengren U, Andersson F, Elgh U, Terselius B, Karlsson S Influence of pH and storage time on the sorption and solubility behaviour of three composite resin materials. J Dent 2001;29:35– 41. 27. Milleding P, Karlsson S, Nyborg L. On the surface elemental composition of non-corroded and corroded dental ceramic materials in vitro. J Mater Sci Mater Med 2003;14:557–566. 28. Xu HH, Weir MD, Sun L, Ngai S, Takagi S, Chow LC. Effect of filler level and particle size on dental caries-
1678
RODRIGUES ET AL.
inhibiting Ca-PO(4) composite. J Mater Sci Mater Med 2009;20: 1771–1779. 29. Gron P. State of calcium and inorganic orthophosphate in human saliva. Arch Oral Biol 1973;18:1365–1378. 30. Antonucci JM, Skrtic D. Fine-tuning of polymeric resins and their interfaces with amorphous calcium phosphate. A strategy for designing effective remineralizing dental composites. Polymers (Basel). 2010;2:378–392.
ION RELEASE FROM RESIN-BASED MATERIALS WITH CALCIUM PHOSPHATES
