The effect of mesoporous silica doped with silver nanoparticles on glass ionomer cements; physiochemical, mechanical and ion release analysis

Abstract

Background

The purpose of the study was to evaluate the effect of adding mesoporous silica with silver nanoparticles to conventional glass ionomer cements (GIC) on its, physical, chemical, mechanical properties and ion release analysis.

Methods

Synthesized mesoporous silica with silver nanoparticles were added in 1, 3 and 5% by weight to the liquid component of GIC forming three experimental groups and compared with plain GIC as control group. Physical and chemical characterization were conducted using nano-computerized tomography (NanoCT) and Fourier transform infrared spectroscopy. Surface microhardness, water sorption and solubility were analyzed. Ion release was investigated using Inductive Coupled Plasma-Optical Emission Spectroscopy and High Performance Liquid Chromatography. Statistical analysis between different groups for the set parameters using parametric and non-parametric tests. The results were analysed using one way analysis of variance (p < 0.05).

Results

Synthesized mesoporous silica with silver nanoparticles were of 7.28 ± 5.0 nm in diameter with a spherical morphology. NanoCT revealed less porosities for 3 wt %. Microhardness showed a statistically significant difference for 5 wt% at day 1 and 21 ( p < 0.0001). Water sorption values decreased significantly on day 14 compared to day 7 for control, 1, 3 and 5 wt%. Control specimens showed highest concentration of fluoride release followed by 5, 3 and 1 wt%.

Conclusion

Mesoporous silica with silver nanoparticles modified glass ionomer cements showed comparable microhardness to conventional GIC. Ion release was evident from the modified specimens. Silver remained within the GIC for atleast four weeks following incorporation.

Introduction

Glass ionomer cements (GIC) are widely recognised as a pivotal restorative biomaterial specially in the field of paediatric dentistry because of their tendency to chemically bond with dental hard tissue [1] and release fluoride to trigger remineralization of dental tissues [2]. The World Health Organization has enlisted GIC’s in the Model list of Essential Medicine, which are medicine that fulfil the populations priority healthcare needs [3]. They exhibit an acid base reaction in between the ion leachable glass powder and aqueous solution of polyacrylic acid. GIC forms a rich bioactive water based restorative biomaterials with an ability for therapeutic ion supply [4]. Although their use has been widely accepted in the clinical field for decades there are still some limitations that researchers are trying to overcome. Amongst these are its poor mechanical properties, sensitivity to moisture [1], limited antibacterial effect after setting [3] and secondary caries [5]. Some studies suggest that the antibacterial activity of GIC is due to the fluoride release, whilst other argue that low pH of GIC during setting contributes more to the bactericidal effect and nearly no antibacterial activity exists after setting [5]. In addition, biofilms can conveniently grow on dental and GIC surfaces due to the broad variety of bacterial species that are harbouring the oral cavity [6].

In order to overcome these investigators have incorporated antibacterial agents into GIC for preventing secondary caries. These modifications of GIC should be achieved without affecting their physical properties [3]. Despite all the advantages of GIC recurrent decay is the most common cause of failure which indicates that the released fluoride is not as strong enough to inhibit bacterial growth and protect tooth structures. Therefore, alternative approaches to enhance the properties of the existing formulation are being explored. Nanotechnology is being harnessed to synthesize, characterize and discover their unique properties to make these materials desirable in biomedical science. Due to the acceptable antimicrobial potential and biocompatibility silver nanoparticles (AgNP) have been adapted in clinical care for different specialities [7]. They have been widely implicated due to the broad range of micro-organisms it can affect within a wide therapeutic window. The concentration, time dependant toxicity and the particle size of AgNP have been discussed in the literature [4, 8, 9]. Nanocomposites release silver ions due to the high surface area of the nano sized silver exposed in the biomaterial, so particles being slowly oxidised into Ag₂O in aerobic conditions [4].

Similar to AgNP, mesoporous silica nanoparticle (MSnP) have also been introduced as promising nanocarriers to entrap bioactive agents due to their large surface area, tuneable pore size and ease of surface functionalization [10]. MSnP are high stability, resistant to microbial attach, show higher mechanical strength, durability and biocompatibility due to their unique helical highly porous ultrastructure morphology [11]. Their addition of dental resin composites have also been reported earlier to improve mechanical and antimicrobial properties [11–13].

Therefore, this study aimed to synthesize mesoporous silica particles (MSnP) with silver nanoparticles (AgNP) and to incorporate them into commercially available glass ionomer cements in different percentages. Additionally, the second purpose was to examine the effect of addition of these nanoparticles in different ratios on mechanical, physical properties, ion release analysis. It was hypothesized that by using such nanoparticles the physical, chemical, and mechanical properties would be enhanced.

Materials and methods

Synthesis of mesoporous silica (MS) and doping silver nanoparticles

Mesoporous silica (MS) was prepared by adopting the sol-gel technique in the presence of a biopolymer. Chitosan solution was prepared by dissolving in acetic acid, followed by the addition of the required quantity of tetraethyl orthosilicate, (TEOS), precursor (to give 5 wt% of silica) whilst continuous stirring. Then required amount of water and ethanol was added to solution and stirred at room temperature (RT). Water to precursor (TEOS) ratio was 2:1. The prepared solutions were casted in petri dishes and left at 50 °C for 24 h (hr). The obtained films are then further dried for 48 h at 70 °C under vacuum to obtain chitosan − 5wt% silica hybrid films. To obtain the mesoporous silica nanoparticles powder, the hybrid films were pyrolyzed in air at temperature in the range of 450–800 °C for 6–8 h in a tubular furnace. Ag nanoparticles (AgNP) were incorporated onto the mesoporous silica (MS) by adding the silver precursor at the required weight% during the sol-gel reaction or prepared separately using a silver precursors as reducing agent (e.g. trisodium citrate (TSC), sodium borohydride (NaBH₄), ascorbic acid, sodium citrate. etc.) and then incorporated at the required wt% into the MS to acquire mesoporous silica with silver nanoparticles (MSAgNP).

Characterization of prepared Mesoporous silica with silver nanoparticles (MSAgNP)

X-ray Photoelectron microscopy (XPS) and transmission electron microscopy of prepared mesoporous silica with silver nanoparticles

ESCA lab250xi instrument (Thermo Scientific, UK) equipped with a monochromator Al Kα X-ray irradiation source of power 1486.5 eV and charge compensation flood gun had been used for analyzing the MSAgNP. After the vacuum reached x10^-7 the particles were transferred to the analysis chamber for scanning. The scanning and peak fit process was conducted with Thermo Advantage software (v5.956, Thermo Fisher Scientific). The full survey spectra were collected on a range of 0-1300 eV at pass energy, dwell time and step size of 150 eV, 50 ms and 1 eV, respectively. The data was transferred to OriginPro 2021 software version 9.8 (OriginLab Corporation, Northampton, MA, USA) to process the high and low resolution spectra of Si and Ag in the prepared MSAgNP. In order to analyze the size and morphology of the particles Transmission Electron Microscopy (TEM) was conducted at a magnification of X60000 and X120000. The TEM samples were prepared by dropping 10 µl of the solution onto the copper grids using a micropipette. The solution was allowed to air dry in a fume hood prior to analyzing. Once the grid was completely dry, it was handled by tweezers and placed onto the specimen holder for examination under the TEM. The particle sizes were measured by Fiji / ImageJ Software (National institute of health, MD, USA).

Preparation of GIC specimens

Riva Selfcure (SDI^® Limited, Bayswater, VIC, Australia) ( two component, Polyacrylic acid 20–30 wt%, tarataric acid 10–15 wt% and Fluoroaluminosilicate glass 90–95 wt% and polyacrylic acid 5–10 wt%) was used as the base GIC. Experimental MSAgNP were mixed in the liquid phase of GIC with the help of a vortex mixed for 2 min (min) at 1, 3 and 5 wt% of MSAgNP. Then the MSAgNP modified liquid phase was hand mixed with the GIC powder according to the manufacturer’s instructions ( spoonful of level powdered surface was mixed with a single drop of liquid). Perspex moulds of different sizes were used to shape GIC specimens according to various characterizations conducted. They were filled with the material and then covered with a glass slide which was flattened and gently pressed by hand in order to remove air bubbles. The materials was allowed to set for 10 min and then the glass slides were gently retrieved and specimens were removed from their respective moulds. These were then hand polished and then sonicated. The experimental flow chart is shown in Fig. 1. All together 4 groups were prepared. Group 1, control ( without MSAgNP), Group 2, 1 wt% MSAgNP, Group 3, 3 wt% MSAgNP and Group 4, 5 wt% MSAgNP.

Fig. 1.

Experimental flow chart for synthesis and characterization used

Fourier transform infrared spectroscopy

Fourier transform infrared spectroscopy (FTIR) was performed on Bruker Tensor 27 spectrometer (Bruker Optics, Inc Billerica, MA, USA) in attenuated total reflectance mode (ATR). Blank backgrounds were acquired before measuring each specimen. The prepared MSAgNP were placed in contact with the ATR crystal and spectra were acquired in the range of 500 to 4000 cm^− 1 at an optical resolution of 16 cm^− 1. Similarly the liquid component of control and experimental specimens with 1, 3 and 5 wt % MSAgNP were also analysed by placing 10 ul on the ATR crystal. Alternatively, spectra were also acquired from the set specimens as well and finger print regions were studies. 32 scans were conducted for each specimen using the OPUS 7.5 Software ( Bruker Optics Inc, Billerica, MA, USA). The data was exported as data point table and processed for peak analysis and peak shift using OMNIC Software (Thermo Scientific, USA).

Nano-computerized tomography

Nano computerized tomography (NanoCT) of the control and experimental specimens 1, 3 and 5 wt % were conducted using the Phoenix nanotom^® ( General Electric Sensing and Inspection Technologies, Wunstorf, GmbH, Germany). Images were acquired at a final isotropic resolution of 5 micron meters per voxel at an x-ray tube voltage of 100 kV and current of 100uA with a 0.5 aluminum filter. Each specimen took 1.5 h for single scan that resulted in 2000 images. Samples were scanned over 360° with a rotation step. The reconstruction process were conducted on VG Studio Max 3.4 Software ( Volume Graphics, GmbH Heidlberg, Germany). Before starting the porosity inclusion analysis, data segmentation or surface determination was conducted. A threshold grey value was assigned to edge voxels. Based on the results of the surface reconstruction, the Porosity/Inclusion analysis module within the VG studio max software was used and it was able to run automatic and fast detection of material discontinuities, such as pores and inclusions. Data reported in the current study measured the materials volume (mm³), defect volume (mm³) and defect volume percentage (%).

Surface microhardness

A polymethylmethacrylate mould was used to prepare disc shaped specimens with a 10 mm diameter and 2 mm thickness in accordance with ISO 9917-1:2007 specifications. All specimens were prepared at RT in 70% relative humidity and stored in distilled water at 37 °C for 24 h before testing. Five-disc shaped specimens from each groups were collected for surface microhardness evaluation. The test was conducted using a digital microhardness tester with a load of 50 g for 20 s on the surface of the specimen with a Vickers diamond indentor (CV Instruments 400DAT/3, Sheffield, UK) at RT. The average value of 10 points randomly selected on each sample were taken into account for further analysis. The diagonal lengths of the indentations were measured by an objective lens at 40X. Microhardness in g µm² was calculated from the equation:

HV = 1854.4P/ d².

Where HV is the Vickers Hardness, P is the load set in grams (g) and d is the diagonal’s length in µm.

Water sorption and solubility measurements

After preparation, each specimen was weighed (M0) using an analytical balance instrument ( Shimadzu AUW 220D, Shimadzu Corporation, Kyoto, Japan) with an accuracy of 0.001 mg. The diameter and thickness were measured using a Mitutoyo digital calliper with an accuracy of up to 0.01 mm ( Digimatic, Mitutoyo Corporation, Tokyo, Japan). The diameter was measured at two points at right angles to each other in order to calculate the mean diameter. Consequently the volume (V) of each specimen was calculated using the formula V = 𝜋 x r² x h in cubic millimetres (mm³). Where r is the mean sample radius and h is the mean thickness. The specimens were then immersed in two well plates containing 5 mL of distilled water for each time of 7 and 14 days separately. The well plates were tightly sealed with parafilm and placed in an incubator at 37 °C for 24 h ( Shel Lab 2406, Sheldon Mfg, Inc, Cornelius, USA). After each time point ( 7 and 14 days) the surface water of the specimens was removed and mass of each specimen was recorded again. Water pH was measured at each time point and after each period (pH.5.5 ).The water sorption (W_so) and solubility (W_sl) were calculated using the equations mentioned below (ISO 4049:2000).

W_so = M1 – M2 / V.

W_sl= M0 – M2 / V.

Where by M0 was the specimen mass before immersion, M1 was the specimen mass after immersion and M2 was the specimen mass after desiccation. V is the volume of the specimen before immersion (mm³).

Inductive coupled plasma- optical emission spectroscopy (ICP-OES) and high performance liquid chromatography (HPLC)

Freshly prepared specimens were used for ion release studies. A split mould of 6 mm height and 4 mm diameter was used for ion release studies. After preparation the specimens were stored individually in 5 cm³ of distilled water. Ion release was analyzed on after week 1, 2, 3 and 4. At the designated time point specimens were taken out and placed in a fresh vial. The solutions were retained and their Al, Ca, Na, P, and Ag ion concentrations were determined by using Inductive coupled plasma-optical emission spectroscopy (ICP-OES). Analysis was performed on Varian ICP-OES (Model Varian 710-ES, South Carolina, USA). Data was represented as parts per million (ppm) release and subjected to statistical analysis by one way ANOVA.

The F ion release over the experimental period was also analyzed from the solutions retrieved for ICP-OES analysis. These solutions were analysed by high performance liquid chromatography (Prominence Shimadzu, Europe GmbH, Germany) with a UV–vis spectrometric detector to analyze and quantify the release of F^- from the control and experimental groups. An IC-A₃ column was used with the mobile phase containing 4-hydroxy benzoic acid, Bis, Tris, Boric acid, phosphate (PO₄^3-), Fluoride (F^-), Chlorie (Cl^-), Bromide (Br^-), Nitrate (NO₃^-) and sulphate (SO₄^2-). The data was acquired as mg / ml and subject to statistical analysis as well.

Statistical analysis

The normality of distribution was defined by the Shapiro–Wilk and Kolmogrovs-Smirnov test. Microhardness, water solubility and sorption data were analyzed with one-way analysis of variance (ANOVA) and Tukey’s post-hoc tests. Data from the ion release was analyzed considering two variation factors: materials and time: The data showed homogenous distribution and were then subjected to a two-way ANOVA followed by Tukey’s post hoc test. The outcome of each statistical test was considered to be significant if p ≤ 0.05 and highly significant with p < 0.0001. The analyses were performed using GraphPad Prism^® software (Version 8.0.1; GraphPad Software Inc., San Diego, CA, USA).

Results

The oxidation state and the presence of silver in MSAgNP were confirmed with XPS measurements (Fig. 2a, b and c ). The survey scan is shown in Fig. 2a which exhibited a peaks at around 532.16, 102.97, 368.08 and 374.04 eV due to O 1 s, Si2p, Ag3d, 5/2 and Ag3d, 3/2 and the atomic percentages acquired from the scan are also reported (Table 1). The high resolution XPS spectra of Si showed one peak whilst that of Ag showed two binding beaks ( splitting of the spin orbital energy level). This confirms the presence of metallic Ag(0) state. The nano scale morphology of the MSAgNP revealed rounded and spherical particles ( approximate diameter of 7.28 ± 5.0 nm). They are slightly aggregated and few are observed to be dispersed as well. The particle size range of the MSAgNP was from 2 to 15 nm.

Fig. 2.

Representing the survey spectra of silica doped silver nanoparticles showing the oxygen, silver, sodium, carbon and silver peaks representing the binding energy. Binding energy graphs of (b) Silica and (c) Silver and their molecular arrangement along with Silver and its different molecular configurations. (d) Transmission electron microscopy images of mesoporous silica nanoparticles doped with silver nanoparticles. Image is scaled at 50 nm. Inset image of the frequency distribution of AgNP

Table 1.

The atomic percentage of the elements acquired from XPS survey scan of MSAgNP

Elements	Peak BE	Atomic %
C1s	284.66	5.29
C1s A	285.92	4.24
C1s B	287.96	1.79
O1s	532.16	53.25
Si2p	102.97	10.63
Ag3d, 5/2	368.08	4.53
Ag3d, 3/2	374.04	2.96

Spectroscopic analysis of the control and experimental liquid components of examined GIC and MSAgNP is shown in Fig. 3. Mesoporous silica gives an absorption band at 3400, 1639, 1057 cm^-1, these are assigned to the silanol stretching vibrations (Si-OH), symmetric stretching, bending vibrations, ( Si-O-Si) bending and trisiloxane bonds. Significant shifts in wavenumbers and intensity were observed in the control and experimental specimens (1, 3 and 5 wt%). Slight band shifts were noted to the lower wavenumbers in experimental specimens. The hydroxyl groups (OH) from the water component of GIC from the liquid is observed at 3400 cm^-1 and another at 1639 cm^-1. The GIC liquid is a solution of polymeric acid with carboxylic acid functional groups (-COOH), which can be noted by the intensity of the band at 1709 cm^-1. The setting reactions causes a reduction in the COOH groups which can be noted in the set cement spectra (Fig. 3c) as it sets it reorganizes as -COO groups ( 1583 and 1622 cm^-1). The carboxylate salts ( Fig. 3c and d) can be observed in the set cement since these are released from the glass and bind to the carboxylate groups on the polymer. The bands are due to aluminum carboxylate, calcium carboxylate and sodium carboxylate.

Fig. 3.

FTIR-ATR spectral data of (a) Mesoporous silica with silver nanoparticles (b) the liquid component of control and experimental groups ( 1, 3 and 5 wt%) (c) Set GIC cement spectra collected after 24 h., (d) Finger print region from 500 to 1800 cm^-1 of the GIC liquid and set cement

The cross-sectional representative nano-CT images of control and experimental specimens are presented in Fig. 4a to d. The specimens displayed cracks and porosities distributed within the matrix. Cracks, connecting pores or starting from them are observed in the control and 1 wt % specimens. The white radiopaque particles are representative of the unreacted ultrafine highly reactive glass particles. These porosities are homogenously distributed throughout the specimens. The ultrastructural morphology of the pores are circular. The matrix in 3 and 5 wt% is more compact. The 3 and 5 wt % have lesser pores comparatively to control and 1 wt% specimens. Results of the porosity inclusion analysis are shown in Fig. 5. No significant difference were noted with respect to material volume. However a statistically significant difference was observed in the defect volume (mm³) and defect volume (%). A significant difference was noted when control (13.61 ± 0.35%) was compared with 3wt% (10.61 ± 0.32%) ( p < 0.0001), significant difference was also observed for 3 wt% and 5 wt % (= p < 0.01).

Fig. 4.

Nano computerized tomography of (a) control, (b) 1 wt%, (c) 3 wt % and (d) 5 wt %. Images shown are cross-sectional and longitudinal sections. Porosities can be visualized along with their distribution

Fig. 5.

Porosity inclusion analysis of control and experimental specimens (a) Materials Volume (mm³). (b) Defect volume ( mm³), (C) Defect volume (%).). Statistically significant is denoted by, * = p < 0.01, ** = p < 0.001, *** = p < 0.0001

Results from the Vickers microhardness (HV) measurements conducted at different time points are shown in Figure: 6. All groups showed a gradual increment in their surface microhardness with time from day 1 to 21. A statistically significant difference is noted in between 5 wt% at day 1 and day 21 ( p < 0.0001). 3 wt% also showed a significant increment from day 7 to day 12 (p < 0.001). On day 21 no significant differences were observed in between the control and experimental groups.

Fig. 6.

Microhardness (B) Compressive strength of control, 1, 3 and 5 wt% specimens over a period of 21 days. Values shown are mean ± SD ( n = 3). Statistically significant is denoted by, ** = p < 0.001, **** = p < 0.0001

The mean and standard deviation for water sorption and solubility on day 7 and 14 are shown in Table 2. Significant interactions was perceived for water sorption in between day 7 and 14 for the tested groups ( p < 0.0001). Water sorption values decreased significantly on day 14 compared to day 7 for control, 1, 3 and 5 wt%. Although 1 wt% showed higher water sorption values (149 ± 8.94) this was not significant when compared to experimental and control specimens on day 7 (131.5 ± 0.25). With respect to water solubility no significant differences were observed in between the groups and time points Statistically significant differences in water solubility between the groups and at different time points were observed.

Table 2.

Water sorption and solubility analysis conducted on day 7 and 14. Values shown are mean ± SD ( n = 3) in µg / mm³

Water Sorption
	Control	1wt%	3wt%	5wt%
D 7	131.5 ± 0.25	149.8 ± 8.94	142.6 ± 8.82	138.4 ± 11.85
D 14	10.80 ± 11.77	5.84 ± 2.42	5.12 ± 0.41	5.15 ± 0.41
Water Solubility
	Control	1wt%	3wt%	5wt%
D 7	47.4 ± 3.29	65.7 ± 11.55	59.7 ± 4.12	52.3 ± 8.20
D 14	40.2 ± 8.13	58.4 ± 24.22	51.3 ± 4.04	51.5 ± 4.12

Results from the ion release analysis for Al⁺, Ca⁺, Na⁺, P, Ag⁺ and F ¯ release analysis conducted at different time points ( week 1, 2, 3 and 4) are represented in Fig. 7a to f. Results revealed that a there was a significant decline in the ion concentration over the period of 4 weeks. Highly significant difference was observed in the control specimen for Al⁺ release from week 1 to week 2 ( p < 0.0001). Ca⁺ ion also showed a significant difference from week 1 to 4 within the control group. Significantly higher values were noted for 5 wt% between week 1 and 3 ( p > 0.01) as well. Na ions showed a significant decline with time from week 1 to 4. A statistically significant difference was noted for 3 wt% from week 3 ( 4.573 ppm ) to week 4 ( 2.663 ppm) ( p > 0.001). With respect to Ag⁺ release, at week 2 significant difference was noted between 1 and 5 wt% (0.0037ppm) ( p < 0.001). A significant difference was noted for F release in the control and 5 wt% specimens from Week 1 to 4 ( p < 0.001). No significant difference was observed within the groups.

Fig. 7.

Ion release analysis conducted by ICP-OES for control, 1, 3 and 5 wt% specimens to quantify the release of (a) Al, (b) Ca, (c) Na, (d) P, (e) Ag ions in parts per million (ppm) and (f) Fluoride release analysed by HPLC in mg/ml over a period of 4 weeks. Values shown are mean ± SD. Statistically significant is denoted by, ** = p < 0.001, **** = p < 0.0001

Discussion

This study assessed the effect of addition of mesoporous silica with silver nanoparticles (MSAgNP) to Glass ionomer cement ( Riva Selfcure, SDI Limited, Bayswater, Australia) on the physical, mechanical and ion release analysis. The incorporation of AgNP and MSnP separately within dental biomaterials is not a new concept [5, 10, 14]. However, studies investigating synergistic role of AgNP and MSnP and their effect on ions released with respect to time, microhardness and anti-biofilm properties are still limited.

Based on the results of the current study, the null hypothesis had to be rejected since there were no statistically significant difference noted in the microhardness, and anti-biofilm potential of the control specimens vs. experimental specimens with MSAgNP. Findings from the studies conducted in the past on similar nanoparticles indicated that such combinations remarkably enhanced the materials toughness, surface hardness and antimicrobial properties [11, 15–17]. Insights into the effect of not just mechanical, physical and antimicrobial properties of such additives can aid in developing the next generation of cements, but also the effect of such additions on ion-release can aid in developing better formulation that can reinforce and overcome the existing drawbacks with conventional GIC. In this study MSnP with AgNP were successfully synthesized via sol-gel technique. This technique was adapted due to its ability to tailor make ultrastructural features ( size, porosity and shape ) during synthesis. The MSnP contain uniformly sized pores and are characterized by having a large surface area. They have been regarded as promising nano carrier platforms to develop intelligent delivery systems with high drug loading capacity, targeted delivery and stimuli responsive release capabilities [10]. AgNP being very small have the tendency to fill up the porous network of MSnP, and also distribute themselves over the silica surface [18]. FTIR results revealed shifts in the peak height and peak numbers. Which pointed towards weak physical interaction rather than chemical crosslinking within the matrix. These findings from the spectroscopic investigations were also in agreement with the study conducted by Tsuzuki et al. [19].It has also been reported that the phenomenon accountable for AgNP stabilization is called polyelectrolyte bridging interaction. It is reliant on a charged polyacrylic acid polymer chain and adsorption is mediated by columbic interactions. Therefore, an electrostatic stabilization occurs triggered by steric and electrostatic repulsions as polyelectrolyte adsorbs on a colloidal positively charges particle surface [4].

Studies in the past have also reported tomography to visualize and quantify porosity percentage of GIC [20, 21]. The presence of voids or porosities are usually attributed to hand mixing of cements or due to air entrapment when the mold is being filled. Therefore they can be attributed to either mixing methodology adapted and or operator dependant parameters [22]. However porosity seems to be lower in hand mixed GIC than those manipulated using automated machines [23]. Previously it has been reported that the addition of nanoparticles of zirconium oxide, alumina and titanium dioxide resulted in less microscopic voids in high viscosity conventional GIC. However, these were visualized using electron microscopy [24]. Whereas, the current study employed nano-CT for bulk analysis and 3 dimensional rendering aided in visualization and distribution of the pores. Over all 3 and 5 wt% showed less pores compared to control and 1wt % specimens. The results are consistent with the findings whereby large pores more readily occur in lower viscosity materials, reducing the strength [22]. Addition of nanoparticles effects the viscosity and pore generation consequently. The ultrastructural morphology of GIC observed in the current study demonstrates many artifacts in the form of inclusions of reduces and increased density. This has observed by Tian et al. and they mentioned that such artifacts are often responsible for initiating cracks in GIC restorations [25].

With respect to the microhardness results addition of up to 5wt % MSAgNP had no significant effect on the hardness. Similar results have also been reported in the past whereby the addition of 2 and 4 wt% nano silica did not affect the microhardness of Ketac™ Molar (3 M ESPE, MN, USA) [26]. Contrary to this they also observed that the addition of similar weight percentages in Fuji IX GP( GC Corporation, Japan) showed statistically significant results. These results suggest that certain glass ionomer cement formulations are more sensitive to the incorporation of silica as compared to others [26]. Elshenawy et al. also modified GIC with quaternized chitosan coated MSnP in similar concentrations of 1, 3 and 5 wt%. Their observations revealed that 1 to 3% addition significantly improved the mechanical properties ( microhardness and flexural strength). However, 5 wt% group showed no significant difference when compared to control [6]. In another study, a higher concentration of MSnP beyond 5 wt% revealed decline in the mechanical properties. They attributed this to the large agglomerates of nanoparticles that could serve as weak zones in the cement matrix [27, 28]. The results observed in this study may be ascribed to either the insufficient dispersion of the MSAgNP or potentially inadequate loading. Another study indicated that in cases where AgNP were not firmly bonded to the matrix, significant enhancement in mechanical properties was not observed. This phenomenon could be attributed to the nano size of the particles, facilitating dispersion between and around polymer chains [29, 30]. Alternatively the MSAgNP may not have interacted with the GIC matrix or may be chemically incompatible with the commercially available formulation. The proximity of the values provided by a commercial GIC indicates that the experimental specimens were relevant for dental applications. A strict comparison of the results observed in the current study with existing literature is difficult at this point due to the limited number of studies conducted on incorporation of MSAgNP [31].

Solubility is one of the pivotal factors used for evaluating the quality of biomaterials used for liners, bases, luting agents and restorations [32]. GIC’s are inherently sensitive to water and exhibit high solubility which can trigger delayed setting and decreased mechanical properties. Moreover, at the early stage of setting, water can both decrease mechanical properties and degrade bonds between tooth and restorative interface [33]. This can occur through two effects: lamination and degradation [34]. Based on the ISO standard 4049, the sorption limit is 40 ug / mm³ after 7 days of water storage. However the current values were higher than these for the control and experimental values. Similar higher water sorption values of GIC ( Vitro Molar, Maxxion R 158, Vitremer, 114) have also been reported earlier by Lima et al. [34]. These values could be due to the cement composition, pore structure of the cements may be poorly controlled, hand mixing can also effect this phenomenon. Furthermore the presence of hygroscopic fillers could also contribute to higher water sorption values [34]. Riva Selfcure contains ionglass™ filler, which are made up of a unique blend of difference sizes of ultra fine highly reactive glass particles. The sorption and solubility behaviour can be partly ascribed to the ion glass filler or the presence of tartaric acid. However, the information and research pertaining to these ion glass fillers are currently scarce [35].

Although the ion release of GIC’s is influenced by the pH of the surrounding medium. In an acidic environment a faster dissolution is triggered consequently higher ion release is observed, conversely a neutral or alkaline pH conditions may reduce ion release [36]. Nevertheless, the pattern observed in the current study can be attributed to what has been observed previously [37]. The first step of ion release is referred to as an early wash out and is the first order dissolution stage that stops after a little while. The second stage is a longer term slower process that is diffusion based [37]. Billington et al. proposed another mechanism called the rapid burst of ion release non-linear with respect to time [38]. Which we can correlate with the findings from the ion release analysis in the current investigation. When water is present apart from H⁺ and OH^- all ions must be derived from glass. These ions may be monovalent such as Na⁺ and F¯, divalent ( Ca²⁺, Zn²+, Sr²⁺ ) or trivalent (Al³⁺). Al³⁺ is a critical ion that interacts with a number of ion species and functional groups within GIC. The carboxylate groups of GIC liquid are the primary functional groups that interact with Al³⁺. A relatively higher Al³⁺ ions was released from the control compared to the experimental groups, this suggests that the ion is bound into the matrix of GIC. This early release of Al³⁺ from the glass appears necessary for faster setting time, this phenomenon has also been observed previously [19, 39]. With respect to calcium, it has been reported to be involved in the setting process of cements and therefore no release has been observed in the current study which is also consistent with previous observations [36]. Interestingly the 1wt% was noted to release a significantly higher Ca²⁺ as compared to other. In regard to monovalent cations (Na²⁺), it has been reported not to exhibit a time-dependent relationship [38]. Phosphorus release was very less. A finding which has been consistent with the observations of Nicolson et al. [37]. They attributed such low values due to the long term storage and regular changes of storage medium. It also suggests that this ion has the smallest driving force for release amongst the ions studied in the current investigation [37]. Ag⁺ ions are usually added in the powder component of GIC formulation, however few studies added them in the liquid component as well [4, 9]. Silver ions could potentially be liberated from the nanoparticles over time as a result of dissolution or surface oxidation mechanisms. Experimental specimens displayed minimal release of silver ions at low concentrations, a result consistent with findings reported in the study conducted by Porter et al. [9]. This release can also be attributed to just surface dissolution of the cement whilst rest of the nanoparticles remain embedded in the cement matrix. and Furthermore, no burst release was noted which is unlike the pattern observed for other ions. Interestingly, the 5 wt% on week two showed a higher released followed by 3 wt% specimens. The primary explanation for this observation could be attributed to the elevated specific surface area of AgNP, rendering them more prone to releasing ions owing to their enhanced surface reactivity [40]. Although The F release observed in the current study was consistent with the findings in the literature whereby there is a burst release and then a plateau [41]. This release has been attributed to the surface wash out effect controlled by the higher diffusion of loosely bound F in the cement matrix. The F content in these biomaterials has a correlation to the increment in the amount of F¯ release by driving the concentration gradiant in between the material and solution. Whilst, the nature of the matrix formation of GIC based biomaterials is a process of maturation impeded F¯ release by decreasing F¯ diffusion. The increase in hardness reflects the maturation of GIC [42]. This can be observed in the current investigation where by the increment in hardness values after 21 days can be correlated to decrease in F¯ release at week 2,3 and 4. Studies have also shown that the F¯ release is influenced by the frequency with which the storage medium is changed [37].

When adding MSAgNP into the liquid component of commercially available GIC we suspected that the interactions in between the two components would primarily be physical in nature. There could be some adsorption or surface interactions between the nanoparticles and polyacrylic acid. The nanoparticles would remain suspended due to steric hinderance and electrostatic repulsion between particles. This suspension would permit a homogenous distribution within the GIC matrix. It is also possible that the particle become partially entangled and entrapped within the polymeric matrix. Whilst the addition of MSAgNP did not significantly alter the immediate physical and mechanical properties of GIC. The addition of MSAgNP could influence the release of therapeutic ions which is critical for remineralization. Improved ion release can also enhance the bioactivity. Under the specific clinical conditions or over extended periods these modified GIC could potentially influence the biomaterials performance. Furthermore, whilst the current study used silver nanoparticles, mesoporous silica particles can be tailored to incorporate other bioactive agents within its ultrastructure.

Conclusion

The addition of mesoporous silica with silver nanoparticles in GIC can improve the mechanical properties, reduce porosity and enhance ion release and potentially extend the clinical lifespan and effectiveness of dental restorations, making them a promising biomaterials for clinical use. Therefore, although MSAgNP can be adapted to carry different novel nanoparticles alone and in composite formulation their incorporation into GIC requires further in-depth investigations.

Author contributions

SBQ acquired funding, experimental work, wrote the manuscript, prepared the figures,, AB synthesized, characterized, reviewed, and edited the manuscript. Both authors reviewed the manuscript.

Funding

This research grant from research sector of Kuwait University DB 02/20.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethical approval and consent to participate

All methods were performed in line with the relevant guidelines and regulations. This study did not involve human or animal experiments and the cell lines used were artificially cultured and did not require ethical approval or moral explanations.

Consent for publication

Not applicable.

Conflict of interest

The authors declare no conflict of interest.