**3. Discussion**

In the present study, both of the synthesis methods proposed using sodium citrate or sodium borohydride as reducing agents, led to the anchorage between the silver nanoparticles and calcium glycerophosphate (Figure 2). Besides, in general, the nanocomposites (AgCaGP) were effective against reference strains of *Candida albicans* and *Streptococcus mutans*. Notably, CaGP substantively increased the antimicrobial effectiveness of silver in the AgCaGP, reducing up to a quarter their minimum inhibitory concentration when compared to the respective AgNP controls (Tables 1 and 2).

Although the CaGP has been previously nanoparticulated before the Ag-CaGP synthesis, in our study it was not characterized as being in nanoparticulated form when associated with silver. It might be happen due to the poor solubility of calcium at pH = 7 [29], even when using the same dispersant (NH-PM), as preconized by Miranda et al., whom synthesized AgNP with hydroxyapatite. A pastier bulk was particularly noted in micrographics of Ag-CaGP when water was used as solvent instead of isopropanol (Figure 2c–f), regardless of the reducing agen<sup>t</sup> that was used in the reaction. Although there has not been difference between micro and nanoparticulated-CaGP in the SEM images, its form influenced the amount of silver ions in the compounds (Table 1). In addition, our results showed the antimicrobial effectiveness against *C. albicans* and *S. mutans* for the samples of group C, and it could be explained by the highest amount of silver ions that are present in those compounds [4,30–35].

This expressive difference in the quantity of silver ions between groups B and C would be related to the characteristics of the reducing agents used, being sodium borohydride considered a stronger reducing agen<sup>t</sup> than sodium citrate [33]. Although silver ions are effective to kill several pathogenic microorganisms, they are easily dispersed, which quickly decreases its local concentration to levels of low effectivity. Moreover, ambient light reduces ionic silver forming typical black spots on skin or on any surface of contact [36]. This process causes aesthetic problems and it has potential to injure healthy living tissues. Silver nanoparticles, contrary to ionic silver, induce the production of reactive oxygen species (ROS), which is the primary antimicrobial mechanism [37]. However, AgNP tend to form aggregates in the absence of any support, reducing their efficacy. Therefore, substrates decorated with immobilized AgNP exhibits enhanced antimicrobial activity for longer periods, reducing the undesirable secondary effects that are associated to free ionic silver [38]. Although the difficult to separate the impact of free ionic silver from the AgNP antimicrobial action, differences that were observed in minimum inhibitory concentrations (MIC), as shown in Table 1, for *C. albicans* and *S. mutans* suggests the influence of their respective metabolism on the efficacy of silver against each microorganism.

Furthermore, other factors may influence the antimicrobial potential of AgNP [34]. For instance, how the compound containing silver interacts with the microorganisms would dependent on the characteristics of the AgNP formed, as well as the chemical and physical changes that may occur when they are added to the medium of interest [33]. In general, for the synthesis of AgNP, AgNO3 is used as source of silver, water or ethanol as solvent, and sodium borohydride or sodium citrate as reducing agen<sup>t</sup> [39]. Fabricated under similar conditions, the AgNP would have a negative surface charge [33,40], and this fact is noteworthy to elucidate the lower effectiveness of the compounds against *S. mutans*. Bacteria have a negative outer membrane charge [41] and the electrostatic attraction may have been hampered, and hence the action of the AgNP associated or not with CaGP on the *S. mutans*

was diminished. On the other hand, apart from fungi present a neutral surface charge [41] and might enhance the attraction of AgNP, the presence of phospholipid components, which contain phosphate groups, may have improved the antimicrobial activity of silver by targeting these sites [42,43]. Indeed, the control of AgNP reduced by sodium citrate showed a lower amount of ions (107.2 μg Ag+/mL) than the control that was produced using sodium borohydride (576.2 μg Ag+/mL), and it was more effective against *C. albicans*, suggesting an antifungal potential of AgNP by itself, which may have afforded the disruption of the *C. albicans* cell membrane by damaging the inner layers of the cell wall, increasing their permeabilization and then allowing for the passage of these particles to into the cell.

On the contrary, against *S. mutans* plaktonic cells, Ag+ may have played a preponderant role, particularly in view of the MIC values that are found for AgNO3 (21.2 μg/mL) when compared to those for AgNP, regardless of the reducing agen<sup>t</sup> that is used in the reaction (250 and 125 μg/mL, respectively, for AgNP (Na3C6H5O7) and AgNP (NaBH4)). Noteworthy was the effect that was produced against *S. mutans* when CaGP was associated with AgNP (Table 1). CaGP afforded an increment in the silver activity and it could be related to the acidogenic and acidic characteristic of *S. mutans*, acting CaGP probably as a buffer, and hence might have prevented the proliferation of the cells in the medium [44–46]. So that, the CaGP buffer activity and the highest amount of Ag+ ions could account for the better effectiveness of the samples of C group against that gram positive bacteria tested.

#### **4. Materials and Methods**

#### *4.1. Synthesis of Silver-Calcium Glycerophosphate (Ag/CaGP) Nanocomposites*

Ag/CaGP nanocomposites were synthesized at the Interdisciplinary Laboratory of Electrochemistry and Ceramics of the Chemistry Department in Federal University of São Carlos. Initially, the commercial form of calcium glycerophosphate (80% β-isomer and 20% rac-α-isomer, CAS 58409-70-4, Sigma-Aldrich Chemical Co., St. Louis, MO, USA) was acquired and was nanoparticulated using a ball mill for 24 h at 120 rpm, obtaining nanoparticles of approximately 10 nm. Then, two chemistry methods were employed for the synthesis. The first method was employed using sodium borohydride as reducing agen<sup>t</sup> (NaBH4, Sigma-Aldrich Chemical Co., St. Louis, MO, USA) and was based on the methodology that was proposed by Miranda et al. [29]. The synthesis was carried out in an alcoholic medium (isopropanol) or deionized water. For this, suspensions containing 5 g of CaGP and silver nitrate (AgNO3 Merck KGaA, Darmstadt, Hessen, Germany) at 0.85 or 0.085 g were prepared in the presence of 0.5 mL of a surfactant (ammonium salt of polymethacrylic acid (NH-PM), Polysciences Inc., Warrington, PA, USA) (Table 1). Then, NaBH4 (0.015 g) was added to each suspension, which caused the reduction of Ag+ to metallic silver nanoparticles in the presence of CaGP. The molar stoichiometric ratio between Ag+ and NaBH4 was 1:1.26, respectively. The second method was based on that proposed by Turkevich et al. [47] and Gorup et al. [48] (2011, p. 355). The reducing agen<sup>t</sup> of AgNO3 was sodium citrate (Na3C6H5O7, Merck KGaA) and the stoichiometric ratio of each compound was, respectively, 1:3. Thus, in a flask containing 100 mL of deionized H2O 5 g of CaGP was added following of 0.5 mL NH-PM and 1.4 g of Na3C6H5O7. This mixture was kept under magnetic stirring and heating. After reaching 95 ◦C temperature, AgNO3 was added and this suspension was maintained stirring for 30 min until occurring the color change, which qualitatively indicated the formation of AgNP. Controls containing only the reducing agents and surfactant, and AgNP produced by both reducing were also prepared.

#### *4.2. Characterization of Ag-CaGP Nanocomposites*

In order to demonstrate the presence of AgNP and CaGP in the compounds, the UV-Vis absorption spectroscopy was employed. The measure is based on the phenomenon of plasmon resonance band, as observed in metallic nanoparticles. Thus, UV-Vis spectra of Ag-CaGP nanocomposites were obtained from aqueous solutions poured out in a commercial quartz cuvette with 1 cm optical path using a spectrophotometer (Shimadzu MultSpec-1501 spectrophotometer; Shimadzu Corporation, Tokyo, Japan) at 300 to 800 nm. Water was used as blank.

After a drying step, the resulting powder, Ag-CaGP, was subjected to X-ray diffraction (XRD) phase characterization using Cu K α radiation (λ = 1.5406 Å), generated at a voltage of 30 kV and a current of 30 mA with continuous sweep in the range of 5◦ < 2θ < 80◦, at a scan rate of 2◦/min (Diffractometer Rigaku DMax-2000PC, Rigaku Corporation, Tokyo, Japan). The particles morphology was also characterized by scanning electron microscopy (SEM) on a Zeiss Supra 35VP microscope (S-360 Microscope, Leo, Cambridge, MA, USA), with field emission gun electron effect (FEG-SEM) operating at 10 kV. A drop of each sample were added with a micropipette and deposited on silicon metal plate (111) and dried at 40 ◦C for 2 h. With this technique, we can identify in the synthesized biomaterials the presence of silver, oxygen, silicon, phosphate, and calcium, which were artificially colored (Figures 3 and 4).

#### *4.3. Minimum Inhibitory Concentration (MIC)*

The MIC values for each sample were determined through the microdilution method and followed the Clinical Laboratory Standards Institute guidelines (CLSI, documents M27-A2 and M07-A9). *Candida albicans* (ATCC 10231) was cultivated on Sabouraud Dextrose Agar (SDA, Difco, Le Pont de Claix, France) and *S. mutans* (ATCC 25175) on Brain Heart Infusion Agar (BHI, Difco, Le Pont de Claix, France). Inocula from 24 h cultures on the respective media were adjusted to a turbidity equivalent to a 0.5 McFarland standard in saline solution (0.85% NaCl). This suspension was diluted (1:5) in saline solution, and afterwards diluted (1:20) in RPMI 1640 or BHI. Initially, the Ag-CaGP nanocomposite was diluted in deionized water in a geometric progression, from 2 to 1024 times. Afterwards, each Ag-CaGP nanocomposite concentration obtained previously was diluted (1:5) in RPMI 1640 medium (Sigma-Aldrich) for *C. albicans* and in BHI for *S. mutans*. The final concentrations of Ag-CaGP nanocomposite in the dispersion ranged from 5 to 0.01 mg/mL. Each inoculum (100 μL) was added to the respective well of microtiter plates containing 100 μL of each specific concentration of the Ag-CaGP nanocomposite solution. The microtiter plates were incubated at 37 ◦C, and the MIC values were determined visually as the lowest concentration of Ag-CaGP with no microorganism growth after 48 h for *C. albicans* and 24 h for *S. mutans*. All of the assays were repeated in triplicate on three different occasions.

#### *4.4. Determination of Ag+ Concentration*

The evaluation of Ag+ concentration in Ag-CaGP and AgNP, as obtained by both reducing agents, was determined by a specific electrode 9616 BNWP (Thermo Scientific, Beverly, MA, USA) coupled to an ion analyzer (Orion 720 A+, Thermo Scientific, Beverly, MA, USA). A 1000 μg/mL silver standard was prepared placing 1.57 g of dried AgNO3 into a 1 L volumetric flask containing deionized water. This solution was stored in an opaque bottle in a dark location and diluted in deionized water at the moment of dosage in order to achieve the standard concentrations used. Thus, the combined electrode was calibrated with standards containing 6.25 to 100 μg Ag/mL at the same conditions of samples. A silver ionic strength adjuster solution (ISA, Cat. No. 940011) that provides a constant background ionic strength was used (1 mL of each sample/standard: 0.02 mL ISA).
