Microemulsion synthesis of zinc-containing mesoporous bioactive silicate glass nanoparticles: In vitro bioactivity and drug release studies

2.1 Materials

Tetraethyl orthosilicate (TEOS, Si(OCH₂CH₃)₄), CTAB (≥99% as surfactant), cyclohexane (≥99%), and bovine serum albumin (BSA) were obtained from Aldrich. Triethyl phosphate (TEP, Merck), polyvinyl alcohol (PVA), calcium nitrate tetrahydrate, and zinc nitrate hexahydrate were purchased from Fluka (Buchs, Switzerland). Absolute ethanol (EtOH) was purchased from EL Nasr Pharmaceutical Chemicals Co (Egypt). Deionized water (Millipore-Q system, resistivity value: 18.2 MΩ. cm, Darmstadt, Germany).

2.2 Preparation of MSN and MBG particles

MSN was synthesized using a microemulsion technique, as reported by Ashour et al. [19]. Two separate solutions, (A) and (B), were initially prepared. Solution (A) was prepared by adding 4 g of PVA to 150 mL of ethanol and 250 mL of deionized water. At 70°C, the solution was stirred, and 0.8 g of CTAB was gradually added, and the entire mixture was vigorously stirred until it turned into a clear liquid. Solution (B) was obtained by mixing 25 mL of TEOS with 20 mL of cyclohexane for 30 min at room temperature, and the pH value was adjusted to 9 with ammonium hydroxide. The reason for adding cyclohexane to the reaction medium is to form a layer to control the rate of TEOS addition during the nucleation and growth of silica particles. Moreover, solution (B) was added slowly to solution (A) and stirred continuously for 3 h, forming a microemulsion state. The resultant mixture was centrifuged and then washed with absolute ethanol and distilled water. The precipitate was left to dry for 10 h at 80°C, followed by calcination for 3 h at 700°C to eliminate the surfactant. The sample is denoted as S1.

The MBG system containing 75% SiO₂–5% P₂O₅–20% CaO (in wt%) was prepared by the microemulsion technique. ZnO was added to MBG at the expense of SiO₂. MBG compositions were described after the addition by the following formula: [(75 − X)% SiO₂–5% P₂O₅–20% CaO–X% ZnO, X = 0, 2.5, and 5 wt%]. The samples are denoted as S2, S3, and S4, respectively. Solution (A) was prepared by adding 4 g of PVA to 150 mL of ethanol and 250 mL of deionized water. At 70°C, the solution was stirred, and 0.8 g of CTAB was gradually added, and the entire mixture was vigorously stirred until it turned into a clear liquid. Solution (B) was obtained by mixing 20 mL of TEOS with 18 mL of cyclohexane and stirring for 30 min at room temperature. The reagents TEP, Ca(NO)₃·4H₂O, and Zn(NO₃)₂·6H₂O were sequentially added to the TEOS solution, and the pH value was adjusted to 9 with ammonium hydroxide, allowing 45 min for a complete reaction. Afterward, solution (B) was slowly mixed with solution (A), stirring the mixture continuously for 3 h to form a microemulsion state. After vigorously stirring for 4 h, the resultant mixture was centrifuged and then washed with absolute ethanol and distilled water. The obtained mixture was then oven-dried at 80°C for 10 h, followed by calcination for 3 h at 700°C to remove the organic template.

2.3 Characterization techniques

2.4 N₂ adsorption–desorption isotherms and zeta potentials

The textural parameters of the prepared samples were measured using adsorption–desorption nitrogen gas at 77 K using a fully automated system (Quanta Chrome Nova Gas Sorption 1.12). MSN and MBG samples were degassed for 24 h at 453.15 K (180℃) before the testing. The specific surface area was determined using the Brunauer–Emmet–Teller (BET) technique. The Barrett–Joyner–Halenda model was used to calculate the pore volumes, pore diameters, and pore size distributions in the adsorption branch of the isotherm [22]. The surface charges of the prepared samples were evaluated using a Zetasizer with a 633 nm laser (Nano ZS, Malvern Instruments Ltd, UK). Before being analyzed, 50 mg of MSN and MBG samples were individually mixed with 10 mL of deionized water at pH 7.4, sonicated for 45 min at 25°C, and then filtered using a micro-filter with a pore size of 0.22 μm. The samples were measured in suspension form by placing them in a measuring cell (DTS1060, Malvern). Data analysis was performed using Malvern technology software (Version 4.0). The data collected by electrophoretic mobility provided the basis for computing the zeta potential values.

2.5 In vitro bioactivity in SBF

To assess the bioactivity of MBG and MSN samples in vitro, the ability to form apatite on their surfaces was assessed after their immersion in SBF solution. The Kokubo and Takadama procedure was applied to immerse the prepared samples in SBF solution [23]. Ion concentrations in SBF were approximately the same as those found in human plasma. The prepared samples were immersed in a static SBF solution (pH = 7.4) with a fixed solid-to-liquid ratio of 10 mg·mL⁻¹ and stored at 37°C for 2 weeks in a thermodynamic incubator. After the immersion period ended, samples were withdrawn from the SBF, filtered, gently washed three times with deionized water, and dried completely at room temperature.

2.6 Protein adsorption

A common model protein, BSA, was used in this study. To prepare the soaking protein solution, 2 mg·mL⁻¹ BSA was added to phosphate-buffered saline (PBS). In order to determine the amount of protein absorbed by MSN and MBGs, 50 mg of these powders were placed in 20 mL of protein solution at a fixed ratio (1 mg·mL⁻¹). The samples were incubated for various predetermined times (1, 2, 4, 8, and 16 h) in a shaking incubator at 37°C. Once the given periods for adsorption had elapsed, the samples were instantly removed and stored at 4°C in a refrigerator for later use. The non-adsorbed protein was removed from the collected samples by washing with deionized water and drying overnight at 60°C. Particle-free protein solution was chosen as a reference sample. A UV spectrophotometer (UV Jenway 4600, England) was used to measure the amount of adsorbed BSA at a specific wavelength of 595 nm. The experiments were performed in triplicate for each sample. The amount of total adsorbed protein (Q = mg·g⁻¹) in each sample was estimated using the calibration curve of pure albumin using the following equation [24]:

where C ₀ and C ₁ (mg·mL⁻¹) are the initial and residual concentrations of BSA, respectively, in the PBS solution, V (mL) is the total volume of the starting solution, and M (mg) is the weight of the glass sample immersed in the solution.

2.7 DEX release profile

A suspension containing 100 mg of MSN and MBG nanoparticles was prepared by mixing them with 100 mL of deionized water and stirring the mixture. A total of 100 mg of dexamethasone (DEX) was dissolved in this suspension. Then, 150 mg of DEX-loaded MSN and MBG nanoparticles were immersed in 100 mL of PBS buffer solution at 37°C to evaluate their drug release behavior. To measure the amount of DEX release, 5 mL of solution was withdrawn at various periods, including 2, 4, 8, 12, 24, 48, 96, 168, 240, 312, 384, and 480 h. PBS solution was refreshed with 5 mL of fresh PBS each time when 5 mL was removed for analysis to maintain the sink conditions. Subsequently, a UV Jenway 4600 spectrophotometer at a characteristic wavelength of 273 nm was used to quantify the concentration of released DEX. Each sample was tested at least three times to determine the mean and standard deviations. A calibration curve was plotted before the determination. To obtain information about both the release kinetics and drug release mechanism of the prepared samples, three different mathematical models were applied, such as zero-order, Korsmeyer–Peppas, and Higuchi models. The add-in program DDSolver was used to calculate the dissolution data models and determine the kinetic model of best fit [25].

3.1 XRD before soaking in SBF

Figure 1 illustrates the XRD patterns of MSN and MBG in the wide 2θ region (10°–60°). All samples show amorphous structural support, clearly revealing the structural disorder of the silica network synthesized using the microemulsion technique, as shown by a halo peak at 2θ = 20°–25° (JCPDS card no. 01-086-1561). As reported by Zhang et al. [26], XRD characterizations of silica particles show that they are always formed with an amorphous structure. This result is in agreement with the fact that the silicates are characterized by their small nanosized particles [27]. Also, there are no peaks of metal oxide addition in the XRD pattern, thus giving it a non-crystalline structure. This indicates that the added Ca and Zn particles exist as small-sized nanoparticles in an amorphous phase. This means that the addition of these ions to the glass network was successful without affecting its structure [28]. Therefore, the characteristics of MBG may be tailored by integrating divalent cationic ions such as Ca and Zn.

Figure 1

XRD patterns of MSN and MBG with and without Zn addition before soaking in SBF.

3.2 FTIR spectra before soaking in SBF

The FTIR absorption spectra of MSN and MBG samples are illustrated in Figure 2. As seen in this figure, the absorption spectra were split into two distinct areas: 400–1,720 and 2,950–4,000 cm⁻¹. Three main characteristic bands are observed in all samples corresponding to different modes of silicate groups within the Si–O network. The first band, at 474 cm⁻¹, is ascribed to the bending mode of the bridging oxygen atom in the Si–O–Si group. At 802 cm⁻¹, the second band is ascribed to the silicate binding mode of the bridging oxygen atoms Si–O (b) BO that form the SiO₄ tetrahedra [29].

Figure 2

FTIR spectra of MSN and MBG with and without Zn addition before soaking in SBF.

At 960 cm⁻¹, the third band is associated with the stretching modes Si–O, where each SiO₄ tetrahedron has one non-bridging oxygen (Si–O–NBO) [30]. This band was observed in S2, S3, and S4, but not in S1. This is due to the presence of Ca and Zn ions in MBG, but not in MSN. In addition, the asymmetric stretching mode of the Si–O–Si group is associated with significant characteristic bands covering the range from 1,100 to 1,240 cm⁻¹ [29,31]. However, when P₂O₅ is added to the silicate network, the band at 597 cm⁻¹ appears in S2, S3, and S4. This band is associated with amorphous symmetric bending modes of the P–O bonds. Figure 2 also shows a band located at 1,637 cm⁻¹, attributed to the bending absorption vibrations of the H–O–H group. The presence of this band is corroborated by another broad band attributed to O–H stretching at 3,442 cm⁻¹. Also, the broad absorption band located at 1,421 cm⁻¹, which corresponds to the carbonate groups CO 3 2 − , was detected.

As shown in Figure 2, there are differences between S1 and the other samples. In S1, the Si–O–Si band at 1,100 cm⁻¹ splits into two bands in S2, S3, and S4 at 1,105 and 1,068 cm⁻¹. This splitting is due to the addition of network modifiers such as CaO and ZnO to the silicon network. It is also the reason for the presence of a new band at 966 cm⁻¹ for the non-bridged oxygen (Si–O–NBO) groups. Furthermore, increasing the addition of ZnO at the expense of silicon reduces the intensity of the Si–O–Si stretching band while increasing the intensity of the Si–O–NBO band. The incorporation of cations into the SiO₄ networks is responsible for the appearance of these non-bridging oxygen groups. This leads to a reduction in the degree of connectivity of the [SiO₄] tetrahedron groups and induces the formation of SiO^– groups [32].

3.3 Thermal analysis

A thermal analysis was performed to identify the organic content and thermal stability of mesoporous silica. Figure 3 shows DSC/TGA curves of MSN (S1) and MBG with Zn 5% (S4) from 31 to 1,000°C. As seen in the TGA curve (Figure 3(a)), the total weight loss of S1 is about 35.5%, which occurred between 31°C and 660°C. This total weight loss is divided into three main stages in the temperature ranges of 30–214, 215–360, and 360–660°C. The first stage (6.5%) resulted from the evaporation of physically adsorbed water (drying) and residual solvent, which is confirmed by the appearance of an endothermic peak in DSC at 132°C [33]. The second stage (15%) is a major weight loss due to the degradation of surfactant (CTAB) from the pores of silica, which is confirmed by the appearance of a small exothermic peak in DSC at 330°C [34]. The third stage (14.1%) is caused by the evaporation of trapped water and alcoholic groups formed by the gradual condensation of Si–OH and Si(OC₂H₅) groups, which is confirmed by the appearance of a broad exothermic peak in DSC at 470°C. Above 660°C, no additional weight loss was detected, showing that the extracted mesoporous silica is thermally stable.

Figure 3

DTA/TGA curves of (a) MSN (S1) and (b) MBG with 5% Zn (S4).

As shown in the TGA curve (Figure 3(b)), the total weight loss of S4 is about 56.6%, which occurred between 29 and 690°C. This total weight loss is divided into three main stages in the temperature ranges of 29–210, 230–420, and 420–680°C. The first stage (21.3%) resulted from the evaporation of physically adsorbed water (drying) and residual solvent, which is confirmed by the appearance of an endothermic peak in DSC at 124°C. The second stage (12.3%) is due to the release of chemically adsorbed water and the degradation of surfactant (CTAB) from the pore of silica, which is confirmed by the appearance of a small exothermic peak in DSC at 275°C [35]. The third stage (24.5%) resulted from the decomposition of nitrate ( NO 3 − ), which is present in large amounts due to the use of calcium nitrate and zinc nitrate as raw materials. This weight loss stage may also suggest further condensation of hydroxyl groups caused by the evaporation of alcoholic groups on the glass surface, which is confirmed by the appearance of a broad endothermic peak in DSC at 494°C. The observed higher weight loss in S4 may be attributed to the increased microemulsion residues in this sample, such as alcohols and decomposition of the surfactant (CTAB), as well as to the removal of adsorbed water and nitrate from the surface of the glass sample. The exothermic peak observed at 850°C is ascribed to the crystallization process of MBG. The weight loss stabilized after 680°C, and the thermal stability of MBG (S4) was assessed.

3.4 TEM and ICP

TEM was used to investigate the size, shape, and dispersion value of the synthesized nanoparticles. Figure 4 shows the TEM images, EDS, and particle size distribution of MSN (S1) and MBG (S2, S4). It is seen that the particles of MSN are uniform, spherical, and discrete, without any agglomerations, with average sizes 8.7 nm (Figure 4a, d and g). In contrast, the particles of MBG S2 and S4 have a nearly spherical shape and are agglomerated in a random pattern, with average sizes of 31.6 and 54 nm, respectively. This is due to the addition of Zn, which increased the concentration of metal cations in the SiO₂ network. This, in turn, reduced the negative charge on the surface of MBG and increased their ability to agglomerate [36], as shown by the zeta potential results. The EDS analysis was done to determine the elemental composition of S1, S2, and S4, as shown in Figure 4a, c, and d. The EDS spectra exhibited distinct peaks corresponding to Si, O, and C elements, providing strong evidence that the MSN consists of these elements (Figure 4a). On the other hand, new peaks were detected in the elemental composition of MBG (S2 and S4). These peaks were specific to Si, Ca, P, and Zn elements, indicating successful incorporation (Figure 4b and c).

Figure 4

TEM image, EDS, and particle size distribution: (a), (d), (g) S1, (b), (e), (h) S2, and (c), (f), (i) S4.

The results of the ICP-OES elemental analysis in Table 1 show the actual chemical composition of the MBG samples, and they are compared with the nominal composition. It was observed that the nominal compositional values differ from the actual compositional values obtained from MBG. For example, the initial values of SiO₂ content for all samples were 75, 72.5, and 70 wt%, which then increased significantly to 92.3, 89.7, and 86.9 wt% for S2, S3, and S4, respectively. This difference is due to the use of the microemulsion method to prepare MBG, which lacks proper composition control. During the preparation, metal ions loosely adsorb onto the SiO₂ surface and subsequently integrate into the SiO₂ network after the calcination process. Therefore, excessive rinsing, which is required to remove the organic phase and prevent agglomeration, easily washes away the ions. As also observed, there is a significant decrease in the CaO content compared to its ratio in the nominal composition, while the ZnO content does not significantly decrease. This decrease may be due to the easier adsorption of Zn²⁺ ions, which have a higher electronegativity compared to Ca²⁺ ions [37].

3.5 N₂ adsorption–desorption analyses

The textural features of the prepared samples are described using nitrogen adsorption measurements. Several theoretical models specified by the International Union of Pure and Applied Chemistry were compared with the resulting isotherms. The observed isotherm is IV-type, indicating that all prepared samples have a mesoporous structure, as shown in Figure 5(a). Moreover, all MBG samples show hysteresis loops of shape type H1, confirming that the mesopores are cylindrical in shape with open ends [18]. Furthermore, all samples exhibit pore size distribution curves in a single-modal form, with relatively narrow distribution peaks centered at 4.7 nm for S1 and around 3.8 nm for the remaining samples, as shown in Figure 5(b). This narrow distribution is indicative of the structural uniformity of the mesopore size distribution [38]. Textural parameters such as S _BET, V _P, and D _P are summarized in Table 1. According to the linear section of the BET curves, all samples have a specific surface area ranging from 280 to 381 m²·g⁻¹. Furthermore, the total volumes of adsorption at a relative pressure of 0.9 ranged from 0.3 to 0.4 cm³·g⁻¹ for all prepared samples. Consequently, these samples exhibit decreased surface area and average pore volume. Table 2 indicates that the pore diameter of S1 (4.7 nm) is greater than that of S2, S3, and S4 (4.1, 3.6, and 3.8 nm), respectively. This decrease appears to be related to the competition between Zn²⁺ and Ca²⁺, where Zn²⁺ has a smaller ionic radius and higher electronegativity compared to Ca²⁺, making it easier to adsorb and incorporate into the SiO₂ network. This incorporation led to less cross-linking, disturbed the usual arrangement of the silica network during the self-assembly process, and decreased the gaps between their particles. Such changes could result in structural defects in the atomic array, which could further modify the mesopore structure and shape. These results are in agreement with previous studies [39,40]. Nevertheless, despite the partial disruption of the ordered mesoporous structure, the pore diameters still ranged from 3 to 5 nm and a surface area ranging from 280 to 381 m²·g⁻¹, even after adding 2.5 and 5% Zn to the MBG samples. These optimum values are necessary to provide a suitable environment that facilitates the absorption of many biomolecules, such as enzymes, growth factors, drugs, bioactive proteins, and antibiotics [41,42].

Figure 5

(a) Nitrogen adsorption–desorption isotherm; and (b) pore size distribution of MSN and MBG with and without Zn addition.

Additionally, the surface charge and dispersity of the prepared samples in the physiological solution PBS were evaluated. Surface charge assessment is necessary for the preparation of stable and well-dispersed suspensions of mesoporous particles in water. Furthermore, this property is essential when considering their use as fillers in composites or as drug/ion carriers. The zeta potential represents the surface charge of particles. At physiological pH, all samples exhibited negatively charged surfaces based on their zeta potential values, as shown in Table 2. Moreover, these results showed that the zeta potential decreased to the negative value and shifted from −26.5 to −15.3 mV, leading to a reduction in surface negativity and a tendency to approach zero. This reduction might be owing to the addition of positively charged Zn ions with electronegativity values (1.6) lower than those of Si (1.8) into the MBG network, leading to a decrease in the zeta potential with Zn addition, which is consistent with earlier studies [43,44].

3.6 XRD after soaking in SBF

Figure 6 shows the XRD patterns of MSN and MBG after 15 days of SBF soaking. It was evident from the XRD patterns that the soaked samples showed sharp diffraction peaks with varying intensities. In sample S1, three small sharp peaks appeared at 2θ = 25.78°, 31.72°, and 32.94°, which can be ascribed to (0 0 2), (2 1 1), and (300), respectively. The standard JCPDS card No. (76-0694) indicates that these peaks correspond to the reflection planes of one crystalline phase, the hydroxyapatite (HAp) phase, as shown in Figure 6.

Figure 6

XRD patterns of MSN and MBG with and without Zn addition after soaking in SBF.

In contrast, four additional reflection peaks were detected next to the reflection peaks of the hydroxyapatite crystal in the other samples. These four peaks appear at 2θ = 29.36ᵒ, 39.42ᵒ, 47.56ᵒ, and 57.28ᵒ, which are assigned to the calcite phase (CaCO₃) with the coordinates (1 0 4), (1 1 3), (0 1 8), and (1 1 2), respectively, as described in the standard JCPDS card no. (81-2027). By comparing the samples after immersion in the solution, it was noted that the calcite phase did not appear in the first sample, but it appeared in the remaining three samples. The reason for the formation of the calcite phase is due to the addition of calcium oxide in all samples, except in S1. As is well known, the reaction between Ca²⁺ and HCO 3 2 − creates calcite, which depends on how much Ca²⁺ is left after the MBG dissolves in the SBF solution, as shown in the following reaction [45]: Ca²⁺ + 2HCO₃ → CaCO₃ + CO₂ + H₂O. MBG dissolves in the solution, which determines the concentration of Ca²⁺ ions in the solution. This percentage depends on the ratio of the amount of glass to the SBF solution and the length of time the MBG is immersed. The formation of calcite on the surface of MBG is due to two factors, as reported by Lukito [46]: first, calcite is formed on the MBG surface if the soaking period exceeds 6 h (for one day or more). Second, when the ratio of BG to SBF solution is greater than 2 mg·mL⁻¹ (such as 10 mg·mL⁻¹), the MBG surface becomes susceptible to calcite formation. HA formation dominates calcite formation in the SBF solution during the first 6 h of soaking, owing to the high concentration of PO 4 3 − in the SBF solution and the ion exchange between Ca²⁺ and H₃O⁺. This is because there is competition between the PO 4 3 − groups and the HCO 3 2 − groups for binding to Ca²⁺ ions. However, after soaking for 6 h, the SBF solution showed a strong ability to form calcite because more Ca²⁺ ions remained from the dissolving MBG that were not attached to the phosphate groups and were instead attached to the HCO 3 2 − group in the solution.

It is also noted that the number and intensity of HA crystal peaks of MBG increase with an increase in ZnO content. This indicates that the HA deposited on the surface of the samples developed significantly as a result of zinc addition. This is because bioactive behavior, specifically apatite formation, pertains to an interface-driven phenomenon. The parameters, including surface charge, chemical compositions, and morphology, play an important role in apatite layer formation and how the material interacts with the surrounding environment [47]. Measuring the changes in the surface zeta potential of MBG in an electrolyte solution could help understand the crucial reactions involved in ion exchange, degradation, and precipitation that influence the bioactive response. The average electronegativity of the cations in MBG was reduced when silicon was substituted with zinc. Thus, these lower electronegativity cations form more ionic bonds with oxygen. As a result, more ionic bonds are weaker and easier to break in aqueous environments, enhancing the dissolution rate of MBG [48]. This rapid dissolution releases more Ca²⁺ and PO 4 3 − ions into the surrounding environment, promoting HA precipitation.

3.7 FTIR after soaking in SBF

The FTIR spectra of MSN and MBG with ZnO content after 15 days in SBF are shown in Figure 7. There are noticeable changes in the absorption bands of silicate groups after soaking in SBF, as well as the formation of new bands. The intensity of the absorption bands at 474 and 804 cm⁻¹, which are associated with silicates, decreased, particularly in S2, S3, and S4. Additionally, the non-bridging oxygen (NBO) bands at 960 cm⁻¹ disappeared. Meanwhile, all IR spectra show a small shoulder at 950 cm⁻¹, corresponding to the Si-OH stretching vibration. Moreover, four new absorption bands appear at 565, 603, 677, and 1,510 cm⁻¹. Two distinct new bands identified at 562 and 603 cm⁻¹ are ascribed to P–O bending (crystal) in all samples. The intensities of these two bands appear to be higher in S2, S3, and S4 compared to S1, and also increase with higher zinc addition. The absorption bands at and 1,510 cm⁻¹ are assigned to the carbonate group, suggesting that the apatite structure contains CO 3 2 − ions [49,50]. These results indicate that the HA crystals formed on the MBG were of hydroxycarbonate apatite (HCA). Furthermore, the absorption bands at 1,645 and 3,450 cm⁻¹ are ascribed to H–O–H stretching vibrations. These bands reveal significant water content due to the hygroscopic nature of the precipitated apatite, which appears due to the nucleophilic P–OH or Ca–OH groups, thereby promoting humidity absorption [51].

Figure 7

FTIR spectra of MSN and MBG with and without Zn addition after soaking in SBF.

By comparing these results with those obtained before immersion, it was revealed that the apatite phase increased as calcium and zinc contents increased. This can be attributed to several factors that play an important role in influencing bioactivity (biomaterialization). In the present study, some of these factors are involved in increasing or decreasing biomineralization (HA formation), such as chemical composition, specific surface area, etc. As it is known, the specific surface area plays a major role in enhancing bioactivity by promoting the formation of the apatite layer on the surface of the material; the higher the surface area, the higher the apatite formation and bioactivity [52]. However, in this study, two impacts of adding zinc at the expense of silica were observed: first, an increase in non-bridging oxygens (NBO) at the level of the chemical composition of the silica network. The second is the reduction of the surface area of the material at the level of the surface texture of MBG. Both have opposing effects on the bioactivity of the material. Thus, the reason for the enhanced bioactivities of S2, S3, and S4 (MBG) is ascribed to the increase in NBO. Therefore, a low surface area combined with a suitable chemical composition, like adding zinc to a silica network, may increase HA layer formation but at a slower rate with more crystallization than if the surface area were high. On the other hand, comparing S1(MSN) with other MBG samples added to calcium and zinc, it was observed that the amount of HA deposited on the surface of S1(MSN) was less, although it had a large surface area, as reported in previous studies [53]. Consequently, it can be deduced that enhancing the bioactivity of the material does not only depend on increasing the surface area but may also be attributed to its chemical composition.

The structural changes in the silica network of MBG resulting from the incorporation of zinc at the expense of silicate can be explained as follows: When Zn ions are added to MBG, the continuous and interconnected tetrahedral [SiO₄] network will be disrupted [54]. During disruption, the additional modifier ZnO changes the silicate glass network by breaking the network chains and creating more NBO. This is an important step in the bioactivity process because the concentration of NBO groups affects the rate of silica dissolution and solubility by forming Si–OH groups at the MBG surface [55]. Thus, these silanol groups promote the growth of heterogeneous HA nuclei. Once nucleation sites form on the surface, they start to grow spontaneously because the SBF contains Ca²⁺ and P⁵⁺ ions released from the glass, helping the formation of the HCA layer. IR spectral analysis confirmed that the addition of ZnO in small proportions improves the ability to form HCA and enhances the bioactivity and biomineralization of MBG.

3.8 Scanning electron microscopy (SEM-EDS)

Figure 8 shows the SEM micrographs of MSN and MBG with the addition of different contents of Zn before and after soaking in SBF. The surfaces of S1 and S3 exhibited a uniform microstructure and smooth surface before soaking in SBF solution, as shown in Figure 8(a and c). After soaking for 15 days, numerous small, spherical particles with a diameter range of 6–7 µm appeared and covered the surface of all samples. These spherical particles, which formed in varying quantities on the surface of the samples, represent the apatite layer [24]. For S1 (MSN), it was observed that the apatite layer particles were not well developed, displaying poor crystallization, and did not cover the whole surface (Figure 8b). This is attributed to the insufficient formation of nucleation sites on its surface [56]. On the other hand, the amount of the apatite layer deposited on the surfaces of MBG samples (S2, S3, and S4) varied depending on the zinc content (Figure 8d–f). This is because different elements in the silica network, such as Ca, P, and Zn, which participate in the formation of the apatite phase, provide sufficient nucleation sites on their surface. EDS analysis was conducted to confirm the formation of an apatite layer on the sample surface by determining the Ca/P ratio after soaking in the SBF solution, as illustrated in Figure (8g–i). The EDS patterns of S1, S2, and S4 reveal the presence of Ca and P elements, with corresponding Ca/P ratios of 1.53, 1.59, and 1.64, respectively. These ratios suggest that the apatite layer covering S1 and S2 may be non-stoichiometric HA, as it is lower than the standard Ca/P ratio (1.67) and falls within the previously published range of Ca/P ratios between 1.3 and 2.0 for hydroxyl apatite [57]. However, the Ca/P ratio for S4 is closer to the standard value, indicating that the resulting apatite layer approaches stoichiometric HA. Furthermore, these results confirm the findings from XRD and FTIR studies.

Figure 8

SEM images of S1 (a) and S3 (c) before soaking in SBF, S1 (b), S2 (d), S3 (e), and S4 (f) after soaking in SBF; EDS patterns of the layer formed on S1 (g), S2 (h), and S4 (i) after soaking in SBF.

3.9 Protein adsorption

Protein adsorption on solid surfaces plays a vital role in many biomedical applications, such as drug delivery systems, biosensors, or artificial tissues, organs, etc. BSA is an important protein because it can bind and deliver drugs and nanoparticles. The amount of BSA adsorption on MSN and MBG up to 16 h of incubation in a BPS solution is shown in Figure 9. All samples exhibit the ability to adsorb protein, but the amount of protein adsorbed varies from one sample to another. A greater amount of protein was absorbed on the MBG than on MSN. Moreover, an MBG sample with 5 mol% of Zn ions (S4) shows significant protein adsorption as compared to that of S2 and S3. These results showed that the incorporation of Zn ions could increase protein adsorption, as S3 and S4 adsorbed a larger amount of BSA than S1 and S2. This is attributed to the variation in the surface charge value between the prepared samples and the adsorbed protein. The surface with a lower negative charge might increase the adsorption of BSA, as evidenced by the zeta potential values listed in Table 2. As is known, BSA has a negatively charged surface and therefore prefers to bind to surfaces with a positive or less negative charge [58]. In addition, it was observed that the amount of absorbed BSA increased with time, especially at the initial periods of incubation (1, 2, and 4 h), but saturation occurred after 16 h. This saturation might be due to exposure to more hydrophilic groups on the surface of the MBG sample and thus not finding free surface sites that can adsorb the BSA [59]. MBG after doping with Zn showed significant protein adsorption properties and thus could be chosen as the suitable composition used in bone tissue engineering.

Figure 9

Protein adsorption profiles of MSN and MBG with and without Zn addition in PBS at 37°C for different incubation times.

3.10 Analysis of drug release kinetics

3.10.1 Drug release behavior

Several textural nanoparticle properties influence selective drug loading and release, such as the composition, pore volume, pore size, specific surface area, and surface charge [60]. Figure 10(a) shows release profiles of DEX from MSN and MBG with and without Zn content after 20 days. It is evident that MSN showed a rapid DEX release rate, while the MBG samples exhibited a slow DEX release rate. It was also observed that adding zinc to MBG reduced the rate of DEX release kinetics. These findings demonstrate that adding Zn ions to MBG reduced DEX’s release rate due to the decreased degradation rate of the silica network. Thus, this decrease in the degradation rate might be the result of strong interactions that developed between MBG and the drug [15].

Figure 10

Cumulative percentage of drug (DEX) release from MSN and MBG with and without Zn addition at pH 7.4 (a); fitting the dissolution data with the drug release kinetic models: zero-order model (b), Korsmeyer–Peppas model (c), Higuchi model (d); cumulative percentage of drug (DEX) release at first 24 h (e); fitting the dissolution data with Higuchi model at first 24 h (f). Each experimental data point is represented as mean ± SD (n = 3).

For instance, after the release for 20 days, the cumulative release of DEX from S1, S2, S3, and S4 was assessed to be at 80.7, 74, 69.2, and 60.7%, respectively. As shown in this figure, the release of DEX from all samples took place in two stages [61]: the drug was rapidly released in the initial 24 h of incubation in the form of a burst (Stage I). This is due to the rapid release of physically bound drug molecules on the outer surface of MBG because of their interaction with water. After that, the drug was slowly released until the incubation period ended (20 days), a process known as the slow-release stage (Stage II). It is caused by releasing drug molecules that have entered the glass pores.

Generally, drug release depends on the type of carrier–drug interaction. If there is a chemical reaction between the prepared substance (carrier) and the drug, the degradation rate of the drug-loaded samples controls and regulates the drug release rate. Conversely, if there is a physical adsorption interaction between the drug and the carrier’s surface, diffusion becomes the primary mechanism [14]. Consequently, the experimental data were analyzed using three mathematical models to determine if drug-loaded samples followed any of these mechanisms.