Modified Sol–Gel Synthesis of Mesoporous Borate Bioactive Glasses for Potential Use in Wound Healing

In this study, we successfully utilized nitrate precursors for the synthesis of silver (Ag)-doped borate-based mesoporous bioactive glass (MBGs) based on the 1393B3 glass formulation in the presence of a polymeric substrate (polyvinyl alcohol (PVA)) as a stabilizer of boric acid. The X-ray diffraction (XRD) analysis confirmed the glassy state of all the MBGs. The incorporation of 7.5 mol% Ag into the glass composition led to a decrease in the glass transition temperature (Tg). Improvements in the particle size, zeta potential, surface roughness, and surface area values were observed in the Ag-doped MBGs. The MBGs (1 mg/mL) had no adverse effect on the viability of fibroblasts. In addition, Ag-doped MBGs exhibited potent antibacterial activity against gram-positive and gram-negative species. In summary, a modified sol–gel method was confirmed for producing the Ag-doped 1393B3 glasses, and the primary in vitro outcomes hold promise for conducting in vivo studies for managing burns.


Introduction
Chronic non-healing wounds (CNHW) constitute a significant health concern throughout the world and affect the life quality of nearly 2.5% of United States people [1]. The financial burden is higher than the $25 billion/year spent on CNHW treatment in just the United States [1]. Therefore, numerous attempts have been made to manage CNHW, including the use of traditional wound dressings (e.g., gauze and ointments) and advanced dressings (e.g., biodegradable and biocompatible materials) [2,3]. However, bacterial infections are still among the most challenging issues ahead of treating CNHW in the clinic and may impede wound healing. In this subject, eradicating multidrug-resistant (MDR) bacteria (e.g., Staphylococcus aureus and Pseudomonas aeruginosa) is an unsolved problem and needs innovative and novel therapeutic approaches. Accordingly, the use of antibacterial materials is of utmost importance for preventing bacteria contamination and colonization in the wound bed. Up to now, plentiful natural and synthetic antibacterial substances have

Glass Synthesis
The Ag-doped 1393B3 borate MBGs were synthesized in a 54.6B 2 O 3 -(22.1-X) CaO-XAg 2 O-7.6MgO-7.9K 2 O-6.1Na 2 O-1.7P 2 O 5 (X = 0, 1, 2.5, 5, 7.5) multi-component system. For this aim, appropriate amounts of reagents, including B(OH) 3 , Ca(NO 3 ) 2 .4H 2 O, AgNO 3 , Mg(NO3) 2 .6H 2 O, KNO 3 , NaNO 3 , and (C 2 H 5 ) 3 PO 4 (Triethyl phosphate (TEP)) were defined using HSC chemistry ® software (HSC chemistry ® 9.4, Outotec, Espo, Finland) to obtain 10 g of the desired BGs (Table 1). In the experimental section, 10 g of B(OH) 3 was dissolved in 200 mL absolute ethanol (Merck, Darmstadt, Germany) at 70 • C for 45 min by using a magnetic stirrer at a speed of 1000 rpm. The nitrate reagents were then dissolved in deionized water and added to the B(OH) 3 -containing solution in 45 min intervals. TEP was hydrolyzed in the presence of deionized water and then it was added to the obtained solutions (batch 1). In another batch, 10 g of PVA (M W = 8800 g/mol) was dissolved in 300 mL of deionized water for 60 min at a temperature of 80 • C for 2 h (batch 2). The pH of batch 2 was increased up to 14 by using ammonium hydroxide solution (25% NH 3 in H 2 O). After that, batch 1 was added to batch 2 with a drop rate of 30 drops/min under constant stirring, and the gel-like samples were immediately obtained. The aging process was carried out by storage of the prepared samples in sealed bottles for 7 days. In order to dry the samples, they were freeze-dried for 48 h. To initially remove the polymeric substrate, the samples were then pre-treated at a temperature of 250 • C for 24 h. Finally, the samples were heat-treated at 600 • C at a rate of 1 • C/min in the air. Figure 1 shows the schematic presentation of the process designed for synthesizing the sol-gel borate BGs.

Thermal Behavior
In order to determine the thermal behavior of the glasses, the freeze-dried samples were analyzed using thermogravimetric analysis (TGA) and differential thermal analysis (DTA) analyses (STA 503, BAHR, Hullhorst, Germany) with a heating rate of 10 °C/min in the air. To discuss the glass transition temperature (Tg) data, the Ag-doped borate-based BG compositions were adapted from SciGlass database version 7.12. The obtained data from the database were processed using SPSS Modeler (version 18.22, IBM, Armonk, NY, USA). It should be noted that we removed the duplicated data and replaced them with their average.

Elemental Composition Analysis
The compositions of the calcined MBG particles were analyzed by inductively coupled plasma atomic emission spectroscopy (ICP-AES, Spectro Arcos, Kleve, Germany) after being digested in the aqua regia.

XRD Analysis
The phase composition of the synthesized Ag-doped 1393B3 MBGs was investigated using X-ray diffraction (XRD) analysis (D8-Advance, Bruker, Karlsruhe, Germany) before and after immersion in SBF. The instrument conditions were set to do scanning at a 2θ range of 20-80°. The Cu-Kα radiation was employed with a step size of 0.05° and a time

Thermal Behavior
In order to determine the thermal behavior of the glasses, the freeze-dried samples were analyzed using thermogravimetric analysis (TGA) and differential thermal analysis (DTA) analyses (STA 503, BAHR, Hullhorst, Germany) with a heating rate of 10 • C/min in the air. To discuss the glass transition temperature (T g ) data, the Ag-doped borate-based BG compositions were adapted from SciGlass database version 7.12. The obtained data from the database were processed using SPSS Modeler (version 18.22, IBM, Armonk, NY, USA). It should be noted that we removed the duplicated data and replaced them with their average.

Elemental Composition Analysis
The compositions of the calcined MBG particles were analyzed by inductively coupled plasma atomic emission spectroscopy (ICP-AES, Spectro Arcos, Kleve, Germany) after being digested in the aqua regia.

XRD Analysis
The phase composition of the synthesized Ag-doped 1393B3 MBGs was investigated using X-ray diffraction (XRD) analysis (D8-Advance, Bruker, Karlsruhe, Germany) before and after immersion in SBF. The instrument conditions were set to do scanning at a 2θ range of 20-80 • . The Cu-Kα radiation was employed with a step size of 0.05 • and a time per step of 2 s. The characteristics of the crystalline HAp were studied using the Rietveld refinement technique (Profex 4.2.2 package, Solothurn, Switzerland) after the glass phase transformation in SBF. The rate constant (K) of the phase transformation of the MBGs to crystalline ceramic phase was calculated as suggested in Equation (1) [21].
where C0 and Ca represent the initial amount of the crystalline ceramic after the first incubation time in SBF (i.e., 24 h (86,400 s)) and the amount of HAp after a time of T (s), respectively.

FTIR Study
The primary bands of the synthesized MBGs, including B-O bands, were investigated using Fourier-transform infrared spectroscopy (FTIR) analysis on pellets consisting of a fixed amount of sample and KBr (Thermo Nicolet AVATAR 370, USA) over the range of 400-4000 cm −1 . Furthermore, the main bands of the hydroxy carbonated layer (HCA), i.e., P-O, O-H, and C-O bands, were investigated in the SBF-incubated glasses.

DLS and Surface Charge
The particle size values of the synthesized MBGs were measured by dynamic light scattering (DLS) (Vasco3, Cordouan, France) analysis. The zeta (ζ) potential and mobility values of the samples were measured using Zeta potential analyzer (NANO-flex ® II, Thermo Fisher Scientific, Waltham, MA, USA). For this purpose, the MBG powders (0.01 g) were first dispersed in absolute ethanol (10 mL) by applying ultrasonic waves (FR USC 22 LQ, 400 w, 20%, Taiwan) for 5 min and then introduced to the instrument.

Electron Microscopy Observations
In order to observe the surface morphology, the samples were first gold sputter-coated and then introduced to field emission scanning electron microscopy (FESEM) (MIRA3, TESCAN, CZ) before and after incubation in SBF. The effect of the Ag-doped MBGs on the particle size was investigated by using transmission electron microscopy (TEM) (EM 208S, Philips, Amsterdam, The Netherlands). In addition, the impact of the dopant concentrations on the surface topography and roughness of the glasses was evaluated by AFM analysis (Nano Wizard II; JPK Instruments, Berlin, Germany). It is mentioned that TEM and AFM analyses were carried out through dispersion of 0.01 g of the MBGs powders in 30 mL of absolute ethanol with the assistance of ultrasonic waves (FR USC 22 LQ, 400 w, 20%, Taiwan) for 10 min. Then a drop of the suspensions was picked up with a lam and standard TEM grade for AFM and TEM analysis.

N 2 Adsorption-Desorption Analysis
The mesoporous characteristics of the glass particles were determined using Brunauer-Emmett-Teller (BET) and Barrett-Joyner-Halenda (BJH) analyses (Quantachrome Instrument, Japan). Before the test, the glass powders were degassed at 250 • C for 6 h in a vacuum process.

In Vitro Bioactivity Assessment
According to Kokubo's method [22], we prepared SBF to evaluate the bioactivity of the synthesized MBG particles. For this purpose, 0.15 g of the BG particles were incubated in 100 mL of SBF, as reported in [23]. Then the samples were shaken at a speed of 20 rpm at the temperature of 37 • C for 1, 3, and 7 days. Meanwhile, the pH changes of the samples were measured by using a digital pH meter (AZ pH Meter 86552, Taiwan). The phase, morphology, and the released ions concentration were detected using XRD, FTIR, FESEM, and ICP-AES analyses.

Cell Viability
MTT ([3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl tetrazolium bromide]) assay was performed to reveal the impact of the borate-based MBGs on the viability of mouse fibroblasts (NIH-3T3 cell line; National Cell Bank, Pasteur Institute of Iran). For this purpose, 1 × 10 4 cells were plated in 96-well cell culture treated plates and cultured in RPMI-1640 supplemented by 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin solution (Gibco, Waltham, MA, USA). The 1, 3, and 5 mg/mL of each sample were incubated with RPMI for 24 h in order to prepare a conditioned media. After 24 h, the cell culture media were replaced with the conditioned counterparts, and the cells were further cultured for an additional 24 h. Then the media were pulled out and replaced with the MTT solution (5 mg/mL). After 4 h, the MTT medium was aspirated and replaced with dimethyl sulfoxide (DMSO) (Sigma-Aldrich, Burlington, MA, USA) and shaken for 10 min. Finally, the optical density of the cell culture wells was read using a microplate reader (Synergy HT, BioTek, Winooski, VT, USA) at 570 nm.

Antibacterial Activity
Gram-positive bacteria Staphylococcus aureus (S. aureus) (PTCC: 1112) and Bacillus cereus (B. cereus) (PTCC: 1247), as well as gram-negative bacteria Escherichia coli (E. coli) (PTCC: 1330) and Pseudomonas aeruginosa (P. aeruginosa) (PTCC: 1074), were cultured in the nutrient broth (NB) in an incubator at 37 • C for 24 h. The MBG particles were dissolved in a solvent containing deionized water and dimethyl sulfoxide (5 v/v DMSO) to obtain extracts of the BGs. The prepared samples were then incubated with bacteria for 24 h. Finally, the number of viable bacteria was counted, and the antibacterial activity of the glasses was calculated by using Equation (2), as follows: where R represents the antibacterial activity of the samples, A is the average number of viable bacteria before the test, B and C represent the average number of bacteria in the control group (DMSO) and experiment group (MBGs) after 24 h of incubation.

ICP Results
The ICP results of the MBGs digested in the aqua regia solution are represented in Table 2. The data confirmed that all of the compositions were almost matched with the designed compositions.

DTA/TGA Analysis
The DTA and TGA graphs of all samples after performing the freeze-drying are illustrated in Figure 2A-E. In general, the endothermic peaks at temperatures lower than 100 • C are related to physically adsorbed water in PVA and other compound structures [24]. The endothermic peaks around 230-250 • C are associated with the elimination of water from the PVA structure and the polyene formation [24]. The peaks around 400-500 • C can be attributed to intramolecular condensation of decomposition of polyene to acetaldehyde, benzaldehyde, acrolein, cis, and trans derivatives. The peaks around 500-600 • C belong to the oxidation of the remained structural groups of PVA [24]. Moreover, the endothermic peaks around between 500 and 700 • C could be related to the decomposition of the remaining nitrate groups [23]. According to the DTA graphs, the glass transition of the Ag-free MBGs, 1, 2.5, 5, and 7.5 Ag-doped BGs were about 573, 540, 532, 529, and 526 • C. As can be seen in the TGA graphs, the major weight loss of the samples is related to the thermal decomposition of the PVA structure. Increasing the weight in the TGA graphs at a temperature higher than 550 • C can be associated with the crystallization of oxide phases [25]. The decision tree was considered based on the composition of this study, with a focus on borate-based bioactive compounds containing Ag or Ag 2 O ( Figure 3A). The frequency distribution histogram of B 2 O 3 , Ag 2 O, and T g in the extracted data is shown in Figure 3B. According to the data, doping of Ag or Ag 2 O to borate-based compounds reduces their T g . Based on the data in Figure 3C, the amount of T g (y) as the function of the concentration of Ag 2 O (x) is y = −34.48 ln(x) + 463.54.

XRD Patterns
The results of XRD, along with the Rietveld analyses before and after the MBGs immersion in SBF, are shown in Figure 4A-E. Besides, the results of the calculation of crystallinity, crystallite size, lattice constant, and rate constant of the BGs phase transformation to HAp are reported in Table 4. The XRD patterns of the BGs before immersion in SBF confirmed a major amorphous state of the synthesized particles with traces of crystalline calcium borate (CaB 2 O 4 , ICDD ref. cod. 01-076-0747, orthorhombic) ( Figure 4A-D). Substitution of Ca with 1, 2.5, 5, and 7.5% Ag decreases the amounts of crystallized calcium borate from 11% to 7, 4, 4, and lower than 1 wt.%, respectively. The XRD graphs clearly indicate the presence of crystallized HAp phase (ICCDD ref. cod. 9-0432) in all the SBF-immersed glasses with a crystallinity degree lower than 40% and a crystallite size lower than 43 nm after 7-days of immersing.  Table 4, the lattice constants (a = b, c) of the formed HAp showed an increase in the MBGs immersed in SBF for 7 days compared to the immersed samples for 3 days. In addition, the situation of the (211) peak of HAp (peak with the highest (100%) intensity) was shifted to the lower angles. The HAp phase was crystallized through a phase transformation of the glass phase to ceramics; the rate constant of this transformation for un-doped, 1, 2.5, 5, and 7.5 mol% Ag-doped MBGs were 7.4 × 10 −7 , 14 × 10 −7 , 13 × 10 −7 , 11 × 10 −7 , and 10 × 10 −7 s −1 .  [26]. A slight shift to a higher wavenumber was observed for the Ag-doped MBGs compared to their Ag-free counterparts. The FTIR spectra of the SBFimmersed MBGs after 7 days are shown in Figure 5B. The marked bands (553-663) cm −1 , and (1000-1100) cm −1 are related to P-O bands [27]. The observed broad bands in the range of (1460-1560) cm −1 are connected with carbonated groups of the hydroxycarbonate apatite (HCA) layer [27].  Table 4, the lattice constants (a = b, c) of the formed HAp showed an increase in the MBGs immersed in SBF for 7 days compared to the immersed samples for 3 days. In addition, the situation of the (211) peak of HAp (peak with the highest (100%) intensity) was shifted to the lower angles. The HAp phase was crystallized through a phase transformation of the glass phase to ceramics; the rate constant of this transformation for un-doped, 1, 2.5, 5, and 7.5 mol% Ag-doped MBGs were 7.4 × 10 −7 , 14 × 10 −7 , 13 × 10 −7 , 11 × 10 −7 , and 10 × 10 −7 s −1 .

FESEM Observations
The surface morphology of the Ag-free and Ag-doped borate MBGs is presented in Figure 6. As clearly observed, a HAp-like layer was formed on the MBGs during 7 days of incubation in SBF. However, higher percentages (between 5 and 7.5 mol%) of the dopant (Ag) in the glasses can interfere with their bioactivity. planar [BO3] 3− [26]. A slight shift to a higher wavenumber was observed for the Ag-doped MBGs compared to their Ag-free counterparts. The FTIR spectra of the SBF-immersed MBGs after 7 days are shown in Figure 5B. The marked bands (553-663) cm −1 , and (1000-1100) cm −1 are related to P-O bands [27]. The observed broad bands in the range of (1460-1560) cm −1 are connected with carbonated groups of the hydroxycarbonate apatite (HCA) layer [27].

FESEM Observations
The surface morphology of the Ag-free and Ag-doped borate MBGs is presented in Figure 6. As clearly observed, a HAp-like layer was formed on the MBGs during 7 days of incubation in SBF. However, higher percentages (between 5 and 7.5 mol%) of the dopant (Ag) in the glasses can interfere with their bioactivity.

TEM Images
TEM micrographs of the Ag-free and Ag-doped borate-based MBGs are observed in Figure 7. Regarding the glassy nature of samples, the particles do not show specified or oriented morphologies. The porous particles with a 50-100 nm size were observed in the un-doped glass. As observed in the TEM images, doping Ag into the MBGs reduced their particle size to 15-60 nm. It should be pointed out that the well-ordered porosity can be seen in the Ag-doped samples with 7.5 mol%.

TEM Images
TEM micrographs of the Ag-free and Ag-doped borate-based MBGs are observed in Figure 7. Regarding the glassy nature of samples, the particles do not show specified or oriented morphologies. The porous particles with a 50-100 nm size were observed in the un-doped glass. As observed in the TEM images, doping Ag into the MBGs reduced their particle size to 15-60 nm. It should be pointed out that the well-ordered porosity can be seen in the Ag-doped samples with 7.5 mol%.

AFM Micrographs
The AFM images of MBGs are displayed in Figure 8. As shown, the surface roughness of the MBGs significantly increased (p < 0.05) after doping Ag into the samples' network. The average values of the surface roughness were 35 ± 12, 15 ± 8, 61 ± 6, 6 ± 0.4, and 2 ± 0.2 nm for the un-doped, 1, 2.5, 5, and 7.5 mol% Ag-doped BGs, respectively. It should be noted that the well-dispersed nature of the particles is important for the evaluation of surface roughness.

AFM Micrographs
The AFM images of MBGs are displayed in Figure 8. As shown, the surface roughness of the MBGs significantly increased (p < 0.05) after doping Ag into the samples' network.
The average values of the surface roughness were 35 ± 12, 15 ± 8, 61 ± 6, 6 ± 0.4, and 2 ± 0.2 nm for the un-doped, 1, 2.5, 5, and 7.5 mol% Ag-doped BGs, respectively. It should be noted that the well-dispersed nature of the particles is important for the evaluation of surface roughness.  Figure 9 displays the BET/BJH outcomes of the borate MBGs. Given the results and based on Brunauer-Deming-Deming-Teller theory (BDDT) classification, the glass samples belong to category IV and are closely attributed to the H4 class of mesoporous particles according to [27]. This class is related to irregular and broad-sized and/or ordered mesoporous particles.  Figure 9 displays the BET/BJH outcomes of the borate MBGs. Given the results and based on Brunauer-Deming-Deming-Teller theory (BDDT) classification, the glass samples belong to category IV and are closely attributed to the H4 class of mesoporous particles according to [27]. This class is related to irregular and broad-sized and/or ordered mesoporous particles. Table 5 represents the mesoporous characteristics of the MBG particles in detail. The S BET of the un-doped, 1, 2.5, 5, and 7.5 mol% Ag-doped MBGs was 47, 87, 104, 145, and 167 m 2 /g, respectively. The corresponding pore radius range of the MBGs was 27-33, 17-25, 29-33, 23-31, and 6-12 nm for the un-doped, 1, 2.5, 5, and 7.5 mol% Ag-doped samples, respectively. The intersection of two adsorption and desorption curves as the representative of adsorption energy for the mentioned MBGs were 0.79, 0.80, 0.74, 0.82, and 0.4, respectively. Figure 9 displays the BET/BJH outcomes of the borate MBGs. Given the results and based on Brunauer-Deming-Deming-Teller theory (BDDT) classification, the glass samples belong to category IV and are closely attributed to the H4 class of mesoporous particles according to [27]. This class is related to irregular and broad-sized and/or ordered mesoporous particles.  Table 5 represents the mesoporous characteristics of the MBG particles in detail. The SBET of the un-doped, 1, 2.5, 5, and 7.5 mol% Ag-doped MBGs was 47, 87, 104, 145, and 167 m 2 /g, respectively. The corresponding pore radius range of the MBGs was 27-33, 17-25, 29-33, 23-31, and 6-12 nm for the un-doped, 1, 2.5, 5, and 7.5 mol% Ag-doped samples,

Ions Release Profile
The results of the ions released from the borate glasses into SBF are shown in Figure 10A-G. The calculated kinetics of ions released into the SBF are represented in Tables S1-S7. According to the data, B ions were first released from the MBGs into SBF during the first 72 h of incubation, and then they were desorbed to the samples from SBF after 168 h. In the case of P, a constant decrease in its release into SBF was observed over the incubation time (up to 168 h). In contrast, the release of Ca 2+ , Na + , and Mg 2+ ions showed a constant increase during the test. More importantly, the release profile of Ag + ions into SBF indicated a continuous increase during the test period; a burst release can be found in the first 24 h, followed by a slow upward slope in the release by 168 h post-incubation.

pH Variations
The pH changes in the glass containing SBF are shown in Figure 10H. Given the data, the pH was increased from 7.42 to 8.35, 8.42, 8.2, 8.2, and 8.25 for the un-doped, 1, 2.5, 5, and 7.5 mol% Ag-doped glasses.

Cell Viability
The effects of the prepared MBGs on cell viability are shown in Figure 11. According to the data, the viability of cells cultured with the conditioned media containing 5 mg/mL and 3 mg/mL of the MBG particles showed a reduction of up to 20 and 10% compared to control (cells culture without the MBGs), respectively. Incubation of the cells with the conditioned media containing 1 mg/mL of the MBGs had no significant negative effect (p > 0.05) on the viability of NIH-3T3 cells during 24 h of incubation.
S7. According to the data, B ions were first released from the MBGs into SBF during the first 72 h of incubation, and then they were desorbed to the samples from SBF after 168 h. In the case of P, a constant decrease in its release into SBF was observed over the incubation time (up to 168 h). In contrast, the release of Ca 2+ , Na + , and Mg 2+ ions showed a constant increase during the test. More importantly, the release profile of Ag + ions into SBF indicated a continuous increase during the test period; a burst release can be found in the first 24 h, followed by a slow upward slope in the release by 168 h post-incubation.

pH Variations
The pH changes in the glass containing SBF are shown in Figure 10H. Given the data, the pH was increased from 7.42 to 8.35, 8.42, 8.2, 8.2, and 8.25 for the un-doped, 1, 2.5, 5, and 7.5 mol% Ag-doped glasses.

Cell Viability
The effects of the prepared MBGs on cell viability are shown in Figure 11. According to the data, the viability of cells cultured with the conditioned media containing 5 mg/mL and 3 mg/mL of the MBG particles showed a reduction of up to 20 and 10% compared to control (cells culture without the MBGs), respectively. Incubation of the cells with the conditioned media containing 1 mg/mL of the MBGs had no significant negative effect (p > 0.05) on the viability of NIH-3T3 cells during 24 h of incubation.

Antibacterial Activity
The results of the antibacterial activity of BGs against P. aeruginosa and E. coli (gramnegative species) as well as S. aureus and B. cereus (gram-positive species) are shown in Figure 12A,B. According to the data, the un-doped BGs showed 50% antibacterial activity against both gram-positive and negative bacteria. Doping of Ag into BGs significantly (p < 0.05) increased the antibacterial activity of BGs by up to 98% against gram-negative bacteria ( Figure 12A). Besides, Ag-doping increased the antibacterial activity of BGs by up to Figure 11. MTT assay results of the Ag-free and Ag-doped MBGs. (* p ≤ 0.05, ** p ≤ 0.01, ns means not significant).

Antibacterial Activity
The results of the antibacterial activity of BGs against P. aeruginosa and E. coli (gramnegative species) as well as S. aureus and B. cereus (gram-positive species) are shown in Figure 12A,B. According to the data, the un-doped BGs showed 50% antibacterial activity against both gram-positive and negative bacteria. Doping of Ag into BGs significantly (p < 0.05) increased the antibacterial activity of BGs by up to 98% against gram-negative bacteria ( Figure 12A). Besides, Ag-doping increased the antibacterial activity of BGs by up to 80% against gram-positive bacteria ( Figure 12B).

Discussion
A series of Ag-doped borate 1393B3 MBGs were synthesized through a modified solgel process. In this study B(OH)3 was used as the precursors of boron. The decomposition of B(OH)3 in the water occurred as follows (R1): However, the decomposed ions are not stable, and the reaction could easily happen in the opposite direction of the sol-gel process steps (e.g., polycondensation); therefore, a sharp decrease occurs in the homogeneity of the products. Besides, the B(OH)3 and another borate source (e.g., tributyl borate) could deposit after evaporation of water from the solutions by heating the sols, as reported elsewhere [28]. To address this problem, along with reducing the costs needed for methoxyethoxide precursors, we used PVA substrate and proceeded with the synthesis at a high pH (pH = 14). In an aqueous media, the B(OH)3 and PVA react as follows (R2 and Supplementary Figure S1): where C2H4O is a single unit of PVA structure, and (C2H4O)2*B(OH)4 − is a single unit of polyvinyl borate ion's structure. This reaction impedes the B(OH)3 depositions in all the sol-gel processes. Moreover, increasing the pH of PVA-containing solutions could lead to in-situ depositions of all the used nitrate compounds. The results of the XRD pattern of the formed gels after this insitu deposition indicated the presence of PVA in the structures (Supplementary Figure  S2). In the heat-treatment process, the oxides or multi-component BGs could be formed. According to the following reaction (R3), the nucleation of the oxides instead of the BGs in low temperatures is very insignificant.
The glassy state of the particles was increased by increasing the amounts of dopants

Discussion
A series of Ag-doped borate 1393B3 MBGs were synthesized through a modified solgel process. In this study B(OH) 3 was used as the precursors of boron. The decomposition of B(OH) 3 in the water occurred as follows (R1): However, the decomposed ions are not stable, and the reaction could easily happen in the opposite direction of the sol-gel process steps (e.g., polycondensation); therefore, a sharp decrease occurs in the homogeneity of the products. Besides, the B(OH) 3 and another borate source (e.g., tributyl borate) could deposit after evaporation of water from the solutions by heating the sols, as reported elsewhere [28]. To address this problem, along with reducing the costs needed for methoxyethoxide precursors, we used PVA substrate and proceeded with the synthesis at a high pH (pH = 14). In an aqueous media, the B(OH) 3 and PVA react as follows (R2 and Supplementary Figure S1): The calculated equilibrium constant (K) of this reaction is about 10 308 , which indicates the formation of multi-component oxides instead of single ones. The minor amounts of the unwanted calcium borate phase were nucleated in our experimental section. Although this phase is water-soluble and could easily be removed from the BGs, the optimization of the process needs more investigation in future studies to avoid the formation of any unwanted phase.
The glassy state of the particles was increased by increasing the amounts of dopants (Table 4). From a thermodynamic point of view, the stability of multi-component BGs is higher (the higher value of entropy or higher structural disorder) [29]. Our calculations and estimations with HSC chemistry ® software showed that the entropy of the un-doped MBGs is 5818.138 J/K, and doping Ag up to 7.5 mol% increased the value to 6506.882 J/K. Besides, the calculated values of entropy for the un-doped and the 7.5 mol% Ag-doped samples are 38.048 R and 39.75 R, respectively (R is the gas constant and equals 8.314 J/mol × K). According to these calculations, the disordering of BGs in multi-components or the doped BGs is increased and prevents their nucleation in the sol-gel process.
Basically, Ag-doping of the glass network could break the network's bonds and generate non-bridging oxygen (NBO) [29]. As reported, the Urbach energy of the glass network is enhanced by increasing the amount of the Ag 2 O modifier [30]. This increment in energy could raise the disordering of the glass network [30] and thereby lead to changes in the thermal behavior of the glass. According to the DTA results (Figure 2), T g decreased in the Ag-doped MBGs from 573 to 526 • C. The results confirmed the effect of Ag-doping on preventing particle growth (the particle size was reduced from 105 to 25 nm) ( Table 3). As reported [31,32], the creation of NBO in the glass network could increase the mobility of ions; thus, particle growth is impeded. The effect of Ag 2 O doping on the borate glass network is presented as follows (R4): The zeta potential and mobility measurements showed an increase in the surface charge and mobility of the Ag-doped samples. The zeta potential and mobility of the Ag-free MBGs were −12 mV and −0.89 ± 0.2, which increased to −22 mV and −1.58 ± 0.3 µ/s/V/cm in the case of the MBG-Ag7.5), respectively.
A broad hump was observed in the XRD patterns of the MBGs (Figure 4), indicating the amorphous nature of the synthesized glasses. The crystalline phase appeared in the XRD in the case of SBF-doped glasses. The Ag-doping could increase the rate of glass to HAp transformation in all the samples and result in an increase in the rate constant of transformation from 7.4 × 10 −7 up to 14 × 10 −7 s −1 . As our calculation shows, the peak position and the lattice constant of HAp were changed while soaking the MBGs in SBF. This may be associated with the doping of the released ions (such as B and Ag + ) from SBF to the HAp network. The FTIR study ( Figure 5A) confirmed the existence of borate groups in all the samples. The obtained data is clear evidence for the structural modifications and the creation of NBOs in the structure of Ag-doped MBGs. The increased concentration of NBOs could lead to an enhancement in the ionic conductivity of the MBGs. Furthermore, the structural bands of HAp were detected after immersion of all the MBGs in SBF ( Figure 5B).
The surface morphology of the MBGs before and after immersion in SBF ( Figure 6) displayed the formation of the HCA layer onto the MBGs, which is in line with the XRD and FTIR results. The TEM micrographs (Figure 7) of the prepared glasses revealed that the presence of the dopant (Ag) up to 7.5 mol% decreased the particle size from 50 to 100 nm to below 15-60 nm. Moreover, the surface roughness of the MBGs was investigated using AFM analysis (Figure 8), and the obtained data uncovered the significant impact of the dopant on the MBGs surface. Recently, the gradient distribution of the dopants from the bulk to the surface of BGs was reported [33]. According to the AFM data, the surface roughness of the BGs was changed from 35 (Ag0) to 2 nm (Ag7.5). Furthermore, the dopants could act as the functionalization agent and generate surface defects (e.g., oxygen vacancies). Increasing the zeta potential values of the doped samples (Table 3) was in line with this hypothesis.
A summary of the mesoporous characteristics is given in Table 5. According to the data, all the glass samples have a mesoporous nature (pore size between 2 and 50 nm). The mesoporous structure of the samples is related to the nature of the sol-gel process as well as the burning out of the PVA template during the calcination process [34]. Increasing the amounts of S BET of the Ag-doped (from 47 (Ag0) to 167 m 2 /g (Ag7.5) samples (Figure 9) is correlated to changing the morphology and integrity of the particles as well as changing the ionic conductivity, as reported elsewhere [35,36]. The ion's release profile indicated a burst release for all the ions (except P 5+ ) during the first hours of incubation in SBF ( Figure 10). The study of the kinetics of ions released in SBF (Tables S1-S7) demonstrated the independence of the ions released from the amount of the dopant (Ag).
The MTT assay ( Figure 11) showed the glasses have no negative effects on the viability of NIH3T3 fibroblast cells at a concentration of 1 mg/mL. However, a significant inhibitory effect was detected in concentrations higher than 3 mg/mL of the MBGs. BGs were previously reported to have inhibitory impacts on cells' proliferation due to the burst release of ions from their network to the surrounding biological environment and also a sudden increase in pH. All these events can be moderated in vivo conditions due to the large volume of body fluids [37]. It has been reported that 1393B3 glasses have an intrinsic activity against bacterial growth and proliferation [38], due to the high release of ions from their structure into the environment. Our study demonstrated the antibacterial activity of the Ag-free MBGs (50% against both gram-positive and gram-negative bacteria), which was significantly increased after doping with Ag + ions (up to 80 and 98% against gram-positive and gram-negative species, respectively) ( Figure 12). Previously, it has been clarified that Ag + ions can (I) disrupt the bacterial cell wall and cytoplasmic membrane; (II) facilitate the denaturation of ribosomes; (III) interrupt the production of adenosine triphosphate (ATP); (IV) elevate ROS production; (V) interfere with the DNA replication [39]. Interestingly, silver was reported to have more lethal effects on gram-negative than gram-positive bacteria due to their different metabolism profile [20,40].

Conclusions
The Ag-doped 1393B3 borate MBGs were successfully synthesized through a modified sol-gel process using nitrate precursors. We used PVA substrate and implemented pH = 14 for stabilizing boron in the solution and deposition of all the nitrate precursors, respectively. The XRD results revealed the amorphous state of the Ag-free and Ag-doped MBGs. The particle size, zeta potential, surface roughness, and S BET values were 105 nm, −12 mV, 35 nm, and 47 m 2 /g for the Ag-free borate MBGs and changed to 25 nm, −22 mV, 2 nm, and 167 m 2 /g for MBG-Ag7.5. The incorporation of Ag at concentrations of 1, 2.5, and 5 (mol%) into the glass composition had no adverse effect on the bioactivity; however, 7.5 mol% of this dopant hinders the surface reactivity of the glasses. The FTIR spectroscopic study confirmed that doping Ag to the MBGs may increase the NBO concentrations. From a biological point of view, Ag-doped borate MBGs had no inhibitory effects on the growth and proliferation of fibroblast cells at a concentration of 1 mg/mL. More importantly, adding Ag to the borate BGs improved their antibacterial activity from 50% to about 80% and 98% for gram-positive and gram-negative bacteria. In summary, our study showed that Ag-doped 1393B3 MBGs are suitable antibacterial substances and can be utilized for a wide range of wound healing applications, such as the management of burns.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/bioengineering9090442/s1, Figure S1: The structure of poly vinyl borate which forms during the synthesis process; Figure S2: The XRD pattern of the formed gels after this in-situ deposition of compounds;