In vitro antibacterial effect of forsterite nanopowder: synthesis and characterization

The aims of this study were the preparation, characterization, and in vitro antibacterial activity evaluation of forsterite (FS, Mg2SiO4) nanopowder obtained by two major methods, namely sol-gel (FSsg) and co-precipitation (FSpp). The main aim was to determine the influence of preparation methodologies on physical properties and in vitro antibacterial activity of obtained forsterite nanopowder. To assess the best working temperature for the preparation of FSsg and FSpp, the synthesis and thermal treatment conditions were optimized on the basis of thermal gravimetric (TG) and differential scanning calorimetric (DSC) analysis performed on the dried gel and dried co-precipitated solid, respectively. The FSsg and FSpp powders were characterized by X-ray powder diffraction (XRD), indicating a high purity for both FSsg and FSpp powders. The morphology of FSsg and FSpp nanopowders was explored by scanning electron microscopy (SEM) coupled with energy dispersive X-ray spectroscopy (EDX) and atomic force microscopy (AFM). In vitro antibacterial activity was investigated using a targeted pathogen, namely Staphylococcus aureus (S. aureus) ATCC 6538 P as tested strain by broth dilution technique and inoculations on nutrient agar to highlight the bactericidal inhibitory effect. FSsg nanopowder has no inhibitory capacity, while FSpp produced inhibition, the effect being bactericidal at a concentration of 10 mg/mL. The superior bactericidal activity of FSpp against FSsg is due to variation in the own surface properties, such as specific surface area (SSA) and nano-regime particle size. The FSpp nanoparticles, NPs, obtained by co-precipitation method are reported for the first time as a novel bactericidal nanomaterial against S. aureus.


Introduction
During the last few years, a new bioactive silicate, namely forsterite (FS, Mg 2 SiO 4 ), was revealed as a suitable bioceramic for bone tissue engineering, due to its good bioactivity and biocompatibility (Naghiu et al. 2013, Choudhary et al. 2017, Choudhary et al. 2020, Gorea et al. 2020, Venkatraman et al. 2022. For instance, forsterite generates fibrous nano hydroxyapatite (HAP) in simulated body fluid, particularly when it is distributed in a polymer matrix and stimulates adhesion and proliferation of osteoblasts in vitro (Furtos et al. 2016). These findings clearly indicate that forsterite can be an excellent candidate for hard tissue engineering applications due to its bioactivity and biocompatibility.
Studies also revealed that forsterite nanopowder is a good carrier for silver nanoparticles, AgNPs (Avram et al. 2020). These findings are comparable to those obtained by Responsible Editor: Mohamed M. Abdel-Daim using hydroxyapatite, HAP, as carrier of AgNPs (Mocanu et al. 2014) having controlled antibacterial properties.
Removing heavy metal ions by sustainable methods has also been the goal of many studies using mainly hydroxyapatite, HAP, (Ince et al. 2016, Costescu et al. 2014, Zamani et al. 2013, Corami et al. 2008, bentonite (Hieu et al. 2016, Nehru andSumathi 2013), and zeolites (Nehru and Sumathi 2013). From our preliminary results, FS nanopowders are also suitable for removal of different water pollutants, like heavy metals similarly to hydroxyapatite, HAP, employing physical interaction such as ion exchange. The removal of natural wastes and biological contaminants is also of great interest (Gupta et al. 2011, Fernane et al. 2013, Narwade et al., 2014, Elwakeel et al. 2018and Hammad et al. 2020. Staphylococcus aureus (S. aureus) is a pathogen related to infectious diseases of the eyes, skin, and ears (Batista et al. 2017). S. aureus is responsible for causing hip joint infections (Choudhary et al. 2018). It was also reported that S. aureus is the most predominant strain in samples collected from sea water, being very adaptable (Gabutti et al. 2000). The methicillin-resistant Staphylococcus strain (MRSA) is even more problematic as it has acquired resistance to all known types of penicillin as well as β-lactam-based antimicrobial drugs (David andDaum 2010, Liang et al. 2022). This puts extra strain on the medical system due to the explosion of these types of cases. Methicillin-resistant S. aureus genes have been detected in wastewater habitats (Varela andManaia 2013, Ramaite et al. 2021).
As an alternative, an expensive treatment can be used involving systematic administration of antibiotics, which can cause severe problems in liver and kidneys. Usually, this treatment can also lead to antibiotic-resistant bacterial strains and often can cause implant failure leading to multiple surgeries (Menz et al. 2021). A careless use of a large spectrum of antibiotics can lead to both environmental contamination and the spread of antibiotic-resistant pathogens (Zaha et al. 2020, Bungau et al. 2021. Undoubtedly, the best strategy to avoid bacterial contaminations especially in implants surgery is to develop new biomaterials with intrinsic antibacterial activity.
Over the last decade, studies to tailor the intrinsic antibacterial properties of inorganic materials, mainly by ionic substitution with certain ions or elements in their native structure are in progress. For example, strontium substituted hydroxyapatite, HAP-Sr, has been reported to have antibacterial effects on Lactobacillus, E. coli, and S. aureus (Anastasiou et al. 2019), although the Sr 2+ effect is disputed in literature. The Zn substituted hydroxyapatite, HAP-Zn, has also been proven to exhibit antibacterial effects on both Gram positive (B. subtilis, S. aureus, S. mutans) and Gram negative (E. aerogenes, E. coli) bacteria (Chung et al. 2006, Anwar et al. 2016). Current trend is related to multi-functional hydroxyapatites, as highly biocompatible materials, with wide applications mainly as bone substitutes (Anastasiou et al. 2019, Garbo et al. 2020 or as coatings on metallic implants to enhance the osseointegration and fracture healing (Oltean-Dan et al. 2021, Oltean-Dan et al. 2019. These nanostructured materials can be further used as carriers of AgNPs and antibiotics, to avoid infections in orthopedic surgical applications. Forsterite has been also studied regarding bone implants (Furtos et al. 2016). Some studies evidenced a potential intrinsic antimicrobial property of forsterite (Choudhary et al. 2017), while others are focused on forsterite-based composites (Saqaei et al. 2016). Another study (Zhu et al. 2020) revealed the photochemically induced antibacterial activity of forsterite scaffolds irradiated with NIR. The antibacterial effect might be explained by the ability of Mg 2 SiO 4 to increase the pH in the culture medium, which inhibits the development of pathogens. Magnesium ions have the property of decreasing the adhesion of bacteria on surfaces (Demishtein et al. 2019); their presence might lead to the prevention of the biofilm formation.
Due to the high demand of biomaterials for bone regeneration the awareness of discovering new materials enriched in essential physiologic elements, like hydroxyapatite substituted with Mg, Zn, Sr, and Si, using innovative environmentally friendly synthesis is increasing day by day (Oltean-Dan et al. 2019, Garbo et al. 2020, Oltean-Dan et al. 2021. As an alternative, forsterite ceramics (Mg 2 SiO 4 ) represent a new generation of nanomaterials naturally enriched in Mg and Si, recently discovered to have beneficial effects on osteoblast cells and thereby on bone development. To the best of our knowledge, our previous research (Naghiu et al. 2013) is among the first in this field of research for forsterite ceramics with great potential for biomedical applications. Among the established methods to synthesize powders of forsterite ceramics, the chemical co-precipitation, solid-state reaction, and sol-gel have been categorized as top-ranking methods for the synthesis of pure forsterite powder (Douy 2002, Sanosh et al. 2010, Mathur et al. 2018. Generally, the synthesized forsterite is associated with periclase (MgO), silicon dioxide (SiO 2 ), and enstatite (MgSiO 3 ) (Douy 2002, Choudhary et al. 2017). This disadvantage could limit the medical applications of pure forsterite. Up to now, the role of different thermal treatment conditions on the optimal formation of pure forsterite and its antimicrobial activity is not yet deeply evaluated. Therefore, the main aim of this study was to assess the influence of preparation methodologies on physical properties and in vitro antibacterial activity of obtained forsterite nanopowder.
Accordingly, we hypothesized that the experimental methodology to synthesize pure FS could have an impact on its structure, morphology and antibacterial activity. To respond to this hypothesized research question, the FS was obtained by two major methods, namely sol-gel (FSsg) and co-precipitation (FSpp). Moreover, an advanced thermal treatment was employed during the FS synthesis for the verification of our hypothesis and for the best quality of obtained pure FS NPs. Definitely this work is of great interest particularly in designing synthetic FS nanomaterial for bone substitutes with intrinsic antibacterial activity. Undoubtedly, this approach is crucial to understanding the relationship between advanced methodology, physical properties, and antibacterial activity of optimized FS powder.
Clearly, the diversification of synthesis conditions as well as preparation methods causes the structural properties of resulting forsterite to differ greatly. To the best of our knowledge, the synthesis conditions and the processing of raw materials used so far to synthesize forsterite have never been subjected to deep analysis of the correlation between resulting forsterite nanoparticle size and intrinsic antimicrobial activity.
Accordingly, in the present work, we focused on using the same starting materials, Mg(NO 3 ) 2 and tetraethylorthosilicate, TEOS, as in our previous work (Avram et al. 2020), but the synthesis conditions, preparation methods, and the thermal treatment conditions are established differently to facilitate a better control on particle size during the synthesis and calcination process. Our purpose was to examine the research question whether a synthesis procedure might determine the size of forsterite nanoparticles leading to different properties, like revealing the intrinsic antimicrobial activity of synthesized forsterite. Then, the obtained FS powders are characterized by XRD, SEM-EDX, and AFM. Furthermore, the effect of these FS powders on the Staphylococcus aureus 6538P strain (ATCC) is explored by broth dilution method and inoculations on nutrient agar.
This study provides a better methodology for antibacterial therapy via the development of FS nanoparticles for coatings on orthopedic metallic implants able to modulate the surrounding biological tissue and cell functions. These coatings are also expected to facilitate osseointegration and alleviate the implant infections in orthopedic surgery as well as in fracture and bone healing (Esteves et al. 2022, Perrault et al. 2022, Wu et al. 2022).
FSsg nanopowder was prepared by a sol-gel method using magnesium nitrate hexahydrate (Mg(NO 3 ) 2 • 6H 2 O, 99.5% purity, Merck, Darmstadt, Germany) and tetraethyl orthosilicate (TEOS, Si(OC 2 H 5 ) 4 , 99 % purity Merck, Darmstadt, Germany ) as precursors and following a Mg:Si = 2:1 molar ratio for the forsterite formula. TEOS was added to an aqueous solution of magnesium nitrate, followed by the addition of a sucrose solution (10 g/25 mL water for 1 g FS final after combustion). The mixture was homogenized and brought to a pH of 1 using 65% nitric acid (p.a. Merck, Darmstadt, Germany). Gel maturation was done at room temperature for 24 h, followed by a drying process at 110°C for 4 h. Finally, the dried gel was calcined at 900°C for 2 h, with a heating rate of 10°C/min.
FSpp was synthesized using the same precursors and molar ratio described above. TEOS was added to an aqueous solution of magnesium nitrate according to the stoichiometric ratio. The mixture was vigorously stirred for 2 h on a magnetic stirrer at 1500 rpm. Ammonium hydroxide (NH 4 OH, 25%, p.a. Merck, Darmstadt, Germany) was added to the stirring solution until a pH of 12 was reached to avoid MgO formation and precipitation has begun. The dispersion containing the co-precipitate was continuously stirred for 30 min. The white co-precipitate was separated from supernatant by filtration and washed with ultrapure water for many times to remove the remaining NH 4 OH. Then, the co-precipitated solid was dried at 110°C for 4 h in a laboratory oven. Finally, the co-precipitated solid was calcined at 900°C for 2 h, using a heating rate similar to that one used for FSsg.

Characterization methods
The synthesis and thermal treatment conditions for FSsg and FSpp were optimized on the basis of thermal analysis (TG and DSC) performed on the dried gel and dried coprecipitated solid, respectively, using a SDTQ600 TA Instruments New Castle, USA, with a thermal analyzer equipped with alumina crucibles. The analysis was carried out in air with a heating rate of 10°C/min for a temperature interval of 30-1000°C.
Pure stoichiometric forsterite, FSsg and FSpp, powders obtained after calcination at 900°oC for 2 h, were characterized by XRD, SEM-EDX, and AFM prior to antibacterial investigation.
X-ray diffraction (XRD) spectra of the synthesized powders were obtained by using a Bruker D8 Advance diffractometer (Cu Kα1 of 1.5406 Å, operated at 35 kV and 40 mA), Karlsruhe, Germany. The spectra were collected on a 2θ (10°-70°) with a 0.02°/sec step.
Field emission scanning, transmission electron microscope, STEM, Hitachi HD-2700 Tokyo, Japan, operated at a maximum acceleration voltage of 200 kV was used for morphology investigation of FS powders. SEM images coupled with energy dispersive X-ray spectroscopy, EDX spectra, were used for chemical investigation of FS powders.
Atomic force microscopy (AFM) images were obtained by using an AFM JEOL 4210 Tokyo, Japan, equipment operated in tapping mode, using silicon nitride tip cantilevers, having resonance frequency in the range of 200-300 kHz and a spring constant of 17.5 N/m (Horovitz et al. 2007, Zdrenghea et al. 2011, Garbo et al. 2020. Both types of forsterites, FSsg and FSpp, nanopowders were subjected to ultrasonication process using a ultrasonic processor Sonics Vibra-Cell, model VCX 750 (Sonics & Material Inc., Newtown, CT, USA) in deionized water at room temperature for 3 h, and afterwards deposited by adsorption on glass plates as a thin layer for AFM investigation.
In vitro antibacterial testing of the effects of FSsg and FSpp was performed using a Staphylococcus aureus strain (ATCC 6538P). The forsterite samples (in powder form) were named as follows: 1 for (FSsg) samples, forsterite obtained through the sol-gel method, calcined 900°C for 2 h (i.e., 4 samples of 10 mg FSs each); and 2 for (FSpp) samples, forsterite obtained through the co-precipitation method, calcined at 900°C for 2 h (4 samples of 10 mg FSpp each). Testing of antimicrobial activity was performed by the method of dilutions in sterile nutrient broth (Merck, Darmstadt, Germany); to assess the inhibitory effect of bacteria growth (presence/absence; bactericidal/ bacteriostatic), inoculations were made on Petri dishes with Mueller-Hinton sterile nutrient agar (Merck, Darmstadt, Germany). For the dilution of each forsterite, FSsg or FSpp, in nutrient broth, two series of 5 test tubes, per each FS, of 100/10 mm were prepared: one series for product 1 (FSsg) and the second one for product 2 (FSpp), of which 4 tubes are for diluting forsterite and the fifth tube is considered for the control sample. Both series were identically tested, using the same procedure. The dilution scheme was as follows: firstly, the nutrient broth was distributed in tubes in the following quantities: 1 mL in tube 1 and 0.5 mL in tube 5 (the control), and 0.5 mL in tubes 2, 3, and 4 (see, Scheme 1). Then, in tube 1, 10 mg of forsterite powder was added, resulting a mixture of 10 mg FS/mL of broth which was homogenized by stirring; from tube 1, 0.5 mL dispersion was distributed into test tube 2, obtaining a dispersion of 5 mg FS/mL, which was homogenized; after, 0.5 mL from tube 2 was passed into tube 3, obtaining 2.5 mg FS/mL, which was homogenized, and 0.5 mL was passed into test tube 4 obtaining a dispersion of 1.25 mg FS/mL; after homogenization from tube 4, about 0.5 ml of dispersion was removed. After completion of dilutions, all 5 test tubes of each series, corresponding to FSsg and separately to FSpp, were inoculated with 10 μL culture (0.5 McFarland density standard/1x10 8 CFU) of Staphylococcus aureus strain ATCC 6538P (see Scheme 1). Subsequently, the tubes were placed in an incubator at 37°C for Scheme 1 Forsterite dilutions made with nutrient broth. CFU stands for colony forming units, in S. aureus culture 1 3 24 h, after which the interpretation was made, assessing the turbidity and sediment in each tube.
Because the forsterite powder does not dissolve in the nutrient broth, the suspended FS particles or sediment, respectively, do not allow an appreciation of the inhibitory effect in the broth tubes. To verify that an inhibitory effect occurred, each tube with the corresponding dilution was inoculated on a Petri dish with Mueller-Hinton agar, divided into 5 sectors (4 sectors, each one for every tube with forsterite dilution and the 5th sector for the control tube). The plate was incubated at 37°C, for 24 h, after which it was interpreted. Each measurement was repeated for three times

Synthesis of FSsg and FSpp and thermal treatment
To prepare FSsg and FSpp, the precursors were wet mixed, processed and then dried, obtaining dried gel and dried precipitated powder, respectively. The minimum calcination temperature requested to induce the formation of crystalline forsterite itself, either FSsg or FSpp, was determined by TG and DSC analysis performed on dried concentrated mixtures, respectively (Douy 2002, Sanosh et al. 2010, Mathur et al. 2018. Figures 1 and 2 show the TG and DSC curves for the two samples, dried concentrated gel for FSsg and dried precipitated powder for FSpp, to evaluate the associated thermal treatment with the formation of forsterite crystals. The thermal treatment, revealed in Fig. 1, was performed on the dried gel obtained following the hydrolysis and condensation processes that take place in solution after homogenization and heating. The thermal diagram shows a broad endothermic peak of low intensity at relatively low temperatures, up to 233°C which is due to the removal of adsorbed water from dried concentrated gel. The weight loss is 22.69%. Up to 500°C, the oxidation of organic components occurs in the gel structure with the formation of volatile compounds that leave the sample. These processes are represented by two exothermic peaks, one with lower intensity, between 233 and 381°C and a second one between 381 and 465°C with high intensity. The mass loss that occurs is significant, 11.19% for the first process and 38.17% for the second one. The slow mass loss that is observed between 465 and 1000°C is due to the gradual elimination of the residue of volatile compounds formed by the oxidation of organic components (2.78%). The total weight loss is 74.83% and is specific to this sol-gel method, due to the use of a large amount of chelating agents. At around 833°C, an exothermic peak of low intensity is observed; see the inset in Fig. 1, without a significant mass loss that can be attributed to the formation of the new crystalline forsterite phase, namely FSsg. Therefore, the minimum calcination temperature of about 833°C is determined as the necessary temperature to induce the formation of crystalline FSsg powder. In consequence, the dried concentrated gel was further calcined at 900°C for 2 h, using a heating rate of 10°C/min. Figure 2 presents the TG curve and heat flow as DSC curve during heating of co-precipitated hydroxide powder. Two endothermic peaks can be observed at lower temperatures, up to 360°C that can be attributed to adsorbed water and residual ammonium nitrate removal from the coprecipitated powder. These successive endothermic effects are associated with the mass loss of 16.08% and 16.19%, respectively. Another two sequential endothermic processes take place at higher temperatures and can be attributed to the dehydroxylation of the intermediate precipitated products containing -MgOH groups and -SiOH groups, which can be further converted into oxides upon heating (Zampiva et al. 2017).  Dehydroxylation of -MgOH groups takes place between 360 and 582°C with a continuous mass loss of 10.11% and that of -SiOH groups takes place on a temperature interval that goes from 582 to 800°C and has a mass loss of 9.70%. The mass loss between 800 and 1000°C is due to the gradual elimination of the remaining of volatile compounds formed by the oxidation of organic components (3.71%). The overall mass loss for the co-precipitated powder is 55.79%. An exothermic effect without a large mass variation can be observed on the DSC curve at around 847°C (see inset in Fig. 2). This effect might be attributed to the formation of crystalline forsterite phase, noted FSpp. Consequently, the dried co-precipitated powder was calcined at 900°C for 2 h, using a heating rate of 10°C/min.

X-ray powder diffraction
The X-ray powder diffraction (XRD) patterns are presented in Fig. 3 for forsterite nanopowder obtained through both sol-gel and co-precipitation methods. The XRD patterns show the presence of a well-crystallized phase, characterized by well-defined narrow peaks, for the two FSsg and FSpp samples.
For FSsg, the synthetic forsterite (PDF 83-0542) was identified as the only detectable phase, suggesting that the heat treatment and general synthesis conditions are optimal. Well-defined peaks are also found for FSpp, forsterite (PDF 83-0542), indicating the absence of a second phase such as MgO or MgSiO 3 , which are usually present in forsterite conventional synthesis. Although FSpp is the main predominant phase (around 98%), some traces of quartz (PDF 83-0542) and stishovite (PDF 86-2333) can be found.
Certainly, both FSsg and FSpp are synthetic forsterite of high purity, with crystallinity of 67% and 52%, respectively. They crystallize in the orthorhombic system, space group Pbnm with cell parameters: a=0.4747 nm; b=1.020 nm, and c=0.5990 nm, for FSsg, and cell parameters: a=0.4754 nm; b=1.020; c=0.5980 nm, for FSpp. The crystallite size is about 41.8 nm for FSsg and around 20.3 nm for FSpp. Figure 4 shows the SEM image (Fig. 4a) coupled with EDX analysis (Fig 4b) for FSsg and Fig. 5, the SEM image (Fig. 5a) coupled with EDX analysis (Fig. 5b) for FSpp. A difference in morphology can be observed between the two forsterite samples indicating that the synthesis method influences the size and distribution of particles. The FSpp sample has a more compact appearance, the particles being smaller and having a more pronounced tendency to agglomerate. FSsg particles are better individualized, having larger dimensions when compared to FSpp. However, it can be observed that in the case of both samples the nanoparticles have a homogeneous granulometric (particle size) distribution in a narrow dimensional range, evenly distributed, a fact highlighted by the histograms in Fig. 6.

Chemical analysis of forsterite nanoparticles
SEM coupled with EDX is used for the chemical analysis of FS samples (FSsg and FSpp). EDX spectra confirm the presence of Mg, Si, and O, the constituent elements of forsterite (Mg 2 SiO 4 ). The compositions given by these spectra are reasonably close to the expected values for both FSsg and FSpp (Table 1): Si/Mg mole ratio is 0.5, or for weight ratio is 0.578 for theoretical values. For instance, from the spectrum in Fig. 5b, the Si/Mg ratio is found to be 0.52 for mole ratio, respectively 0.60 for weight ratio. The Si/Mg ratio is closer to the theoretical one in the case of FSpp which suggests a better homogeneity of the initial mixture.
The particles size distribution was obtained for each of the two kinds of forsterite samples, FSsg and FSpp, by measuring the diameters of over a hundred particles on SEM images, and it is presented in Fig. 6.
The mean diameters ( d ) and standard deviation (SD) values for the two samples are d=23.8 nm; SD = 5.6 nm, for FSpp, and d=40.1 nm; SD = 7.1 nm, for FSsg. From the one-way ANOVA test, at the 0.05 level, the means are significantly different. Forsterite obtained by the precipitation method contains smaller particles and their size distribution is rather narrower.

Atomic force microscopy (AFM)
The FSsg sample exhibits the formation of a smooth layer of very well adsorbed nanoparticles. Its characteristic 2D topography image is presented in Fig. 7a highlighting nanoparticles with well-individualized rounded shapes and without a tendency to coalesce. This is due in part to the 3-h ultrasonication time of the dispersion before deposition and the short adsorption time that prevented the formation of clusters on the glass surface. The phase image, Fig. 7b, shows very distinctly the forsterite nanoparticles which are brown in color. This is further evidence of the good  . 6 The particles versus size distribution for FSsg (a) and FSpp (b). On the y-axis the particle fractions (in %) are represented versus their diameters (nm) in certain size ranges (i.e., as shown on the x-axis) individualization of the forsterite nanoparticles synthesized by the sol-gel method. The amplitude image, Fig. 7c, shows that the scan took place in optimal conditions and there are no assembly defects on the surface of the FSsg layer. The smooth appearance with well-individualized FSsg nanoparticles is very well seen in the 3D image, Fig. 7d. The cross-section profile captures three successive FSsg nanoparticles on the deposited film on glass support. The rounded shape and the narrow diameter of around 37±4 nm can be observed.
The AFM images for the FSpp sample are presented in Fig. 8, having a slightly porous structure.
The formation of a rather porous layer of fine FSpp nanoparticles adsorbed from aqueous dispersion on glass plate. They are well individualized and appear very close to each other. In the case of Fig. 8, the film is less dense, and the nanoparticles seem to be better seen. This is more visible in the 3D image in Fig. 8d. It can be noticed that the deposited film is quite smooth. The nanoparticles appear well individualized in the phase image, Fig 8b, having a brown color while the free space between them has a yellow hue. The amplitude image, Fig. 8c, shows that the FSpp film was adsorbed uniformly and is free of defects. The profile in Fig. 8e shows a sequence of 3 adjacent nanoparticles of FSpp. The profile shows a diameter of 21±2 nm.
The size of FSpp nanoparticles is smaller than the size of FSsg nanoparticles in substantial agreement with the size of these NPs visualized in SEM images, and with the data obtained from XRD.
Definitely, the significant difference in size between the two types of forsterite particles has been revealed by XRD, SEM, and AFM. Usually, it is assumed that the particles are spherical without porosity, and in turn the specific surface area, SSA, can be estimated (Brantley and Mellott 2000) by the relation: where d is the average diameter of nanoparticles and ρ is the density of forsterite, namely about 3.3 g/cm 3 (Lloyd andBailey 1975, Zhang et al. 2010). With the measured (1) SSA = 6∕(dρ) image, c amplitude image, d 3D image, and e cross-section profile on the arrow in image (a); scanning area was of 1 μm × 1 μm, and surface roughness was around 2 nm diameters, a specific surface area of about 45 m 2 /g was found for FSsg and around 83 m 2 /g for FSpp.

Antibacterial testing
The efficacy of FSsg (noted 1) and FSpp (noted 2) against S. aureus was studied using broth dilution technique (Scheme 1) since the FS samples are insoluble in organic solvents and consequently their diffusion in agar is limited. Because the forsterite powder does not dissolve in the nutrient broth, the suspended FS particles do not allow an appreciation of the inhibitory effect in the broth tubes.
Firstly, to verify that an inhibitory effect occurred, about 10 μL dispersion from each tube (see Scheme 1) with the corresponding FS dilution was inoculated on a Petri dish with Mueller-Hinton agar, previously divided into 5 sectors (i.e., 4 sectors, each one for every tube with forsterite dilution and the 5 th sector for the control (culture) tube). The plate was incubated at 37°C, for 24 h, after which it was interpreted.
In sample 1 (FSsg), no inhibitory effect was found (Fig. 9a) at any dilution of forsterite, prepared according to Scheme 1. The development of S. aureus colonies was observed in all the corresponding triangles of the Petri dish in which the S. aureus was inoculated, namely in Fig. 9a, positions 1, 2, 3, and 4; this situation was also revealed in the triangle corresponding to the control tube, position 5 in Fig. 9a.
In sample 2 (FSpp), an inhibitory effect was found in Fig. 9b (position 1), only for the composition given (Scheme 1) in tube 1 (10 mg forsterite powder and 1 mL nutrient broth). This inhibitory effect is indicated by the absence of S. aureus colonies in the corresponding triangle (position 1, Fig. 9b). At the other dilutions, respectively 5 mg, 2.5 mg, and 1.25 mg, see Scheme 1, no inhibitory effect appears in Fig. 9b, as it is identified by the development of S. aureus colonies observed in all the corresponding triangles (positions 2, 3, and 4); the S. aureus colonies are also detected in the triangle of position 5 (Fig. 9b) which corresponds to the control tube (control culture, in Scheme 1).
The agar plates shown in Fig. 9 were kept under observation for another 5 days (in wet chambers), without any changes compared to the initial appearance.
Identical results were obtained after re-testing of samples 1 and 2, thus demonstrating that FSpp induces an inhibitory effect at the first dilution (10 mg FSpp in 1 mL nutrient broth) and has no inhibitory effect at lower tested dilutions.
Secondly, the primary tubes, with sample 1 (FSsg), 10 mg in 1 mL nutrient broth, and with sample 2 (FSpp), 10 mg in 1 mL nutrient broth, were kept for 48 h at room temperature. At that time, from these tubes, the inoculation was performed on Muller-Hinton agar, divided in two halves (Fig. 10), the left side for 1 (FSsg) and the right side for 2 (FSpp). Next, the Petri dish was incubated at 37°C for 24 h. Upon interpretation, there was a lack of inhibitory effect on sample 1 (FSsg), with S. aureus colonies development, Fig. 10 (part 1), and an inhibitory effect on sample 2 (FSpp), with lack of the S. aureus colonies development, Fig. 10 (part 2).
Sample 1 (FSsg): lack of inhibitory effect (see, S. aureus colonies, position 1); sample 2 (FSpp): induction of the inhibitory effect (lack of the S. aureus colonies, position 2) Thirdly, the two primary tubes, with sample 1 (FSsg) and respectively with sample 2 (FSpp), were also kept 72 h, from their initial preparation, at laboratory temperature. Then, from these dispersions the inoculations of 10 μL were made Fig. 9 Antibacterial testing for samples 1 (FSsg) given in panel (a) and 2 (FSpp) in (b) against Staphylococcus aureus: sectors: 1 for 10 mg/mL; 2 for 5 mg/ mL; 3 for 2.5 mg/mL; 4 for 1.25 mg/mL; 5 for control (S. aureus culture without forsterite) in two tubes, each having nutrient broth (5 mL). Subsequent, they were followed by incubation at 37°C for 24 h. At that time, they are shown in Fig. 11.
On the interpretation of this situation, it was found that in the tube containing sample 1 (FSsg), there was intense turbidity and easy to homogenize sediment, because of staphylococcal bacterial growth, which means that there was no inhibitory effect. At the control made by the Gram stain method, staphylococci were stained having their characteristic morphology. In the tube containing sample 2 (FSpp), it was found that the inoculated broth remained clear, being sterile, which meant that the sample had an inhibitory effect on bacteria growth (Fig. 11).
To check if the inhibition of the staphylococci was bactericidal, from the two primary tubes, with samples 1 (FSsg) and 2 (FSpp) at contact time 72 h, inoculations were performed on a Petri dish with Mueller-Hinton agar, divided in two halves: left side for sample 1 (FSsg) and right side for sample 2 (FSpp). Then, the plate was placed in the incubator at 37°C for 24 h, after which the result was evaluated (Fig. 12).
It was found that sample 1 (FSsg) has no inhibitory effect, with typical Staphylococcus colonies developing in the corresponding area (left side of the plate). Sample 2 (FSpp) was found that it caused inhibition of staphylococcal growth in the corresponding area (right side of the plate), without the development of colonies, so the nutrient medium remaining sterile. The plate was kept under observation for a further 3 days (wet chamber), without any changes to the original appearance. It can be deduced that the inhibitory effect produced by sample 2 (FSpp) is evidently bactericidal.
To see if the inhibitory effect is correlated with the contact time, both powder samples, 1 (FSsg) of 10 mg FSsg and 2 (FSpp) of 10 mg FSpp, were each hydrated as follows: 10 mg solid product and 1 mL saline solution (B Braun Ag Germany) and left in contact for 24 h, the tubes being shaken several times on a vortex mixer. After 24 h, 10 μL (calibrated loop) of liquid culture of Staphylococcus aureus (strain 6538P ATCC), at a dilution of 0.5 according to McFarland standard, was introduced into the tubes (for both samples). From these tubes, inoculations were made on a new Petri dish with Mueller-Hinton agar, divided into 5 triangular sectors, after the following contact times (minutes): 5′, 10′, 15′, 30′, and 60′. The plates were incubated and evaluated after 24 h (Fig. 13).
In the plate with sample 1 (FSsg), it was found that S. aureus colonies were developed in all the corresponding triangles, which meant that no inhibitory effect occurred at any of the contact times (Fig. 13a). The plate with sample 2 (FSpp) was also found that S. aureus colonies were developed in all the corresponding triangles (Fig. 13b), being more abundant at the contact times 5, 10, and 15 min, and sparse at contact time of 30 min and 60 min.
In order to verify whether the inhibition on staphylococci still occurs after a time exceeding 60 min, the tubes with saline dispersions were kept at laboratory temperature for another 24 h. From these tubes, inoculations were performed on a Petri dish with the Mueller-Hinton medium, divided into two halves, which were inoculated by streaking the left side for product 1 (FSsg) and the right side for product 2 (FSpp). Subsequently, the plate was incubated at 37°C for 24 h, after which the interpretation was performed on Fig. 14.
Sample 1 (FSsg) was found to have no inhibitory effect, and typical Staphylococcus colonies were developed in the corresponding area (Fig. 14, left side). In sample 2 (FSpp), the inhibition of Staphylococcus development in the corresponding area (Fig. 14, right side) was observed, without the appearance of colonies. The plate was kept under observation for a further 5 days, without any changes to the initial appearance. It has been confirmed that the sample 2 (FSpp) has bactericidal effect on staphylococci.
In this study, the FS samples were subjected to a complex antimicrobial testing process on S. aureus. Testing of the samples, FSsg and FSpp, by the method of dilutions in nutrient broth, showed that the inhibitory effect against the S. aureus strain was found only for the FSpp, at a dilution of 10 mg/mL. Inoculation on Petri dish with Mueller-Hinton agar confirmed that FSpp has bactericidal inhibitory effect against S. aureus. Inoculation in tubes with nutrient broth also confirmed that FSpp has inhibitory effect (broth remained clear, being sterile). In relation to the different contact times, 5, 10, 15, 30, and 60 min, FSpp samples do not have an inhibitory effect on S. aureus growth in the control Petri dish. Extending the contact time to 24 h, the FSpp powder produced inhibition, the effect being bactericidal on S. aureus.
In this work, the FSpp obtained by precipitation method is discovered for the first time as a novel bactericidal nanomaterial against S. aureus. This bactericidal activity of FSpp is rather superior to previous reports where FS was prepared through mechanical activation and it was unable to inhibit the bacterial growth in concentration range from 25 to 200 mg/mL (Saqaei et al. 2016).
Another study working with forsterite synthesized by a sol-gel combustion method found an inhibition effect on S. aureus at 2 mg/mL broth dilution for a percentage inhibition of about 76% (Choudhary et al. 2018). An earlier study, on the same forsterite but using agar diffusion, evidenced an effect starting at 100 mg/mL (Choudhary et al. 2017). The discrepancy in the antimicrobial activity data existing in literature could be explained by the influence of the synthesis method and conditions on the characteristics of the final FS product (Zhu et al. 2020).
Our results for bactericidal effect of FSpp nanopowder against S. aureus are found to be reliable at a nutrient broth dilution of 10 mg/mL. Furthermore, inoculation on Petri dishes with Mueller-Hinton agar, confirmed that FSpp has bactericidal inhibitory effect (lack of S. aureus colonies development on the culture plate). Moreover, inoculation in nutrient broth also confirmed that FSpp product has inhibitory effect.

Mechanism of FSpp antibacterial activity
The mechanism by which FSpp nanoparticles produce an inhibitory effect on the tested Staphylococcus aureus strain might be explained by assuming the adsorption of nanoparticles on the cell wall of bacteria. In this hypothesis, it is possible to consider FSpp particles adsorbed and agglomerated on the surface of bacteria, causing mechanical damage to the cell wall and cytoplasmic membrane, disrupting their barrier function, especially the transfer of nutrients through the plasma membrane. Such a mode of action was found on some bacteria (e.g., Staphylococcus aureus and Escherichia coli) by electron microscopy showing that the bacteria treated with various composites containing forsterite were covered with layers of composite particles, and thus, the inhibitory effect was more pronounced against S. aureus, depending on the composite that had a higher ratio of forsterite (Choudhary et al. 2020).
The agglomeration of FSpp nanopaticles around the surface of S. aureus may inhibit the bacterial membrane processes resulting in cell death by leakage of genetic materials, proteins and minerals (Li et al. 2008).
FSpp powder possesses the smallest nanoparticles and has a specific surface area of almost twice that corresponding to FSsg. Consequently, it is to be admitted that the released Mg 2+ amount from FSpp in the broth is higher than that from FSsg during the incubation period. In turn, the released Mg 2+ ions can increase significantly the broth pH by ionic exchange with the H + ions from broth to alkaline pH.
The Mg 2+ ions are found to have a binding capacity to S. aureus membrane (Xie and Yang 2016) which varied with pH of medium (Choudhary et al. 2018;Choudhary et al. 2020). Specifically, Mg 2+ ions interact with cardiolipin (Sennato et al. 2005, Som et al. 2009, Subramani et al. 2016), which is a major lipid component in S. aureus membrane (Short andWhite 1971, Tsai et al. 2011) and form complexes (Cutinelli andGaldiero 1967, Rand andSengupta 1972), and thus disrupt S. aureus membrane (Li et al. 2008) and destabilize membrane processes and finally kill bacteria (Emelyanenko et al. 2021, Wang et al. 2022. Clearly, FSpp has the potential to arrest the bacterial growth and the effect is bactericidal on S. aureus. It is also mentioned that changing the pH of the environment, in the sense of increasing its value, also plays an important role in antibacterial effect of FS. This change in pH is considered the most important activity for the bactericidal action of forsterite. There is research indicating an increase in pH in culture media, from 7.2 in the initial stage of the broth to 8.93 in the broth with nanopowders (Saqaei et al. 2016), respectively from 6.8 to 7.5, after 24 h of incubation (Choudhary et al. 2018;Choudhary et al. 2020). Subsequently, it is believed that an increase in pH determines the denaturation of proteins and several cytoplasmic components of cells, which can precipitate in the environment, resulting in a decrease in the bacteria population.
In relation to different contact times (5, 10, 15, 30, and 60 min), it was found that both samples (FSsg and FSpp) do not have inhibitory capacity at the mentioned times, but by prolonging the contact time to 24 h and by hydrating the sample, it was found that the FSpp induced inhibition (absence of colony development), the effect being bactericidal, and the FSsg product did not have inhibitory capacity (colony of bacteria was developed).
Furthermore, FSpp nanopowder having intrinsic antimicrobial property is appropriate for bacteria removal through biosorption, particularly for S. aureus, which might be present in the surface water. In addition, from our preliminary results, both forsterite FSsg and FSpp nanopowders are suitable for removal of different water pollutants, like heavy metal ions similarly to hydroxyapatite, employing physical interaction such as ion exchange. Moreover, the FSpp powder can better monitor the water pollutants due to predominantly smaller size of NPs, and consequently a higher specific surface area of its constitutive NPs compared to FSsg.
These findings suggest that suitable synthesis conditions can assist in fabrication of pure FSpp powder leading to specific surface properties of its nanoparticles of controlled size having a superior antimicrobial activity against S. aureus. Thus, FSpp powder can serve as coatings on metallic implants decreasing the risk of implant infections during bone surgery.
In summary, this study suggests that Mg 2+ ions may have implications in targeted eradication of S. aureus pathogen, being membrane active against S. aureus and destabilizing S. aureus membrane and killing bacteria.

Conclusion
Overall, the present work extends our earlier study (Avram et al. 2020) by developing a better methodology through precipitation route to obtain forsterite (FSpp) nanoparticles having inherent antibacterial activity.
This study provides two methods, sol-gel and co-precipitation, for the synthesis of FSsg and FSpp nanoparticles, respectively, of high purity. The FSpp obtained by co-precipitation method is reported for the first time as a novel bactericidal nanomaterial against Staphylococcus aureus, at a concentration of 10 mg/mL. This bactericidal activity of FSpp nanoparticles is rather superior to previous reports on FS prepared by different methods, like sol-gel method. Thus, FSpp exhibits a rather strong intrinsic antibacterial property against S. aureus. Furthermore, the nanoparticles of FSpp might enhance the effect of antibiotics against S. aureus. Accordingly, FSpp may serve as an effective antibacterial nanomaterial for its biomedical applications in pharmaceutical fields.
This study also demonstrates the feasibility of adding these FSpp nanoparticles as carriers for sustained delivery of antibiotics and jointly enhancing the effects of antibiotics against S. aureus, and so to support innovation, research, and development in antimicrobial medicine.
These findings revealed an advanced methodology to serve in future design of new carriers of antibiotics and help researchers to make experienced decisions as to which nanomaterials should enter in preclinical testing stage, and thus decreasing animal studies in antimicrobial medicine. Undoubtedly, both FSpp and FSsg nanopowders can be also used as carriers for silver nanoparticles and for antibiotics as coatings on metallic orthopedic implants against infections caused by S. aureus.
Despite current advances in the field of nanomaterials potentially used to treat bacterial infections, the combination of native morphology with physiologic function remains a challenge in antimicrobial medicine.
Further research is needed to develop inorganic nanoparticles to be used in antimicrobial coatings on orthopedic metallic implants able to modulate the surrounding biological environment and cell functions. These coatings are expected to facilitate osseointegration and alleviate possible adverse tissue responses including the implant infection, during orthopedic surgery, and fracture healing. Moreover, they could match physiological processes by their surface functionalization with various biocompounds, like collagen, to enhance cell interactions and their ability to be used in regenerative medicine.
Author contribution AA participated in FS synthesis and characterization, data analysis, literature research, manuscript writing. SR performed antimicrobial testing and data analysis. MG participated in FS synthesis and morphological characterization. GhT participated in research planning and antibacterial study. AM designed the experiment and participated in the FS physical characterization and data analysis. OH participated in design of this study, data analysis, and assisted the revision of the manuscript. GhR performed antimicrobial testing and data analysis, and participated in the manuscript writing and revision. MTC conceived and designed the study, aimed antibacterial mechanism, and assisted the changes to the revised manuscript based on the comments of the reviewers. MTC is corresponding author. The authors agree to be accountable for all aspects of this study. All authors have equal contributions. The authors have read and approved the final manuscript.
Funding This work was supported by grants of the Ministry of Research, Innovation and Digitization, CNCS/CCCDI-UEFISCDI, project number 186 and 481, within PNCDI III.

Competing interests
The authors declare no competing interests.