1 Introduction

Implantation often leads to infection risks, which can result in complications and implantation failure, therefore, implant materials incorporated with antibacterial agents have been extensively investigated in recent decades. Silver containing materials have a very broad spectrum of antibacterial properties; silver has an ability to interact with proteins and enzymes of bacteria. These materials are used for different applications and in various forms due to the antibacterial activity of silver cation (Ag+) and other silver species like Ag0 [1–3]. In both forms it can be used for medical applications, such as antimicrobial bandages and silver coated catheters or implant materials. Such materials can avoid biofilm growth and prevent bacterial colonization on various surfaces [4, 5].

Calcium phosphate biomaterials have been widely used for bone repair and regeneration due to their biocompatibility [6, 7]. The structure and chemical composition of synthetic hydroxyapatite (HAp) is similar to the mineral phase of bone and teeth. The hard tissue mineral phase is HAp, but it contains ionic substituens in the crystalline lattice, like F−, Mg2+, Na+, CO3 2−, etc. HAp is a non-stoichiometric compound with the ability to accept compositional variations, such as substitutional inclusion of several ions [8–11]. Silver ions or silver particles incorporated in the HAp structure or scaffolds could act as antibacterial agents. The combination of silver antibacterial properties and HAp bioactivity, biocompatibility and osteoconductivity leads to a material, which is biocompatible and bactericidal [12–14].

In several studies, many processes are used to synthesize HAp/Ag. The wet precipitation method is among the most described ones [15–22]. It is a simple method and can be easily modified. Usually HAp/Ag is prepared from Ca(NO3)2·4H2O, AgNO3, NH4OH and (NH4)2HPO4 [15–18] or Ca(OH)2, AgNO3 and H3PO4 [19–22]. Sol–gel methods are developed for acquiring HAp/Ag [23, 24] and various approaches to produce HAp/Ag nanoparticles are carried out [25–27]. It has been determined that silver incorporation within the HAp structure is limited for wet precipitation methods [16, 19]. Badrour et al. [16] have estimated that, by using Ca(NO3)2·4H2O, AgNO3, NH4OH and (NH4)2HPO4 as raw materials, it is possible to incorporate silver below 5.5 wt% within the HAp structure; however, Rameshbabu et al. [19] Have specified that the silver amount within structure is below 4 wt%, if Ca(OH)2; (NH4)2HPO4 and AgNO3 are used as raw materials. Most of the prepared materials are used as nanoparticles or as-synthesized powders [16, 18, 20, 21, 23, 25–27]. HAp/Ag as-synthesized material contains a pure HAp phase without the presence of other phases [16, 18–22, 27]. Silver and silver oxide phases are observed for sintered HAp/Ag pallets and scaffolds [19, 24]. The preparation of sintered samples is relevant to gain the best mechanical properties as the as-synthesized HAp has lower mechanical strength than the sintered one. Due to this fact all samples in the current research were sintered prior to the in vitro antibacterial evaluation.

According to the literature, HAp/Ag shows good antibacterial activity towards both gram positive and gram negative bacteria, and the effectiveness is strongly dependent on the incorporated silver concentration [15, 19–24]. It has been testified that the silver release has a great effect on antibacterial efficiency [17]; therefore, it is important to evaluate all factors that define antibacterial effectiveness to gain implant materials with the best properties.

Various studies on HAp/Ag synthesis and evaluation can be found; however, none of them compares physical and antibacterial properties depending on the preparation method. There is a study comparing different precipitation routs of zinc substituted HAp to investigate the effect of the preparation route on obtaining phase pure stoichiometric zinc substituted HAp after heat treatment up to a temperature of 1,100 °C [28], but nothing can be found on the preparation of HAp/Ag with different routes and comparison of physical and antibacterial properties depending on the preparation method. In previous studies silver doped HAp has been prepared by a modified wet precipitation method from calcium oxide, phosphoric acid and silver nitrate as raw materials [29, 30], and the effect of sintering temperature on phase composition and antibacterial properties has been investigated [31]. Therefore, the following aims of the present study can be singled out: to obtain HAp/Ag mixed crystals by two different methods, to determine the phase composition of as-synthesized and sintered powders, to evaluate silver incorporation within HAp structure, to compare and estimate physical properties and to elucidate the effect of the preparation method on the silver release and antibacterial activity of the samples.

2 Materials and methods

2.1 Synthesis of HAp/Ag

Reference HAp and HAp/AgA samples were synthesized using a modified wet chemical method A [19–22] and HAp/AgB samples were synthesized using a modified wet chemical method B [15–18].

In the case of method A: The precursors included calcium oxide (≥97 %, FLUKA), phosphoric acid (85.5 %, Sigma Aldrich) and silver nitrate (≥99 %, Sigma Aldrich).

In the case of method B: The precursors included calcium nitrate tetrahydrate (≥99 %, Sigma Aldrich), ammonium phosphate dibasic (≥98 %, Sigma Aldrich), silver nitrate (≥99 %, Sigma Aldrich) and ammonium hydroxide solution (ACS reagent 28.0–30.0 % NH3 basis, Sigma-Aldrich).

Deionized water was used for all experiments. All synthesis was made with the amount [Ca + Ag] = 0.25 mol and the atomic ratio (Ca + Ag)/P fixed at 1.67.

In the case of method A: An appropriate amount of silver nitrate was added, with atomic ratios Ag/(Ag + Ca) = 0; 0.001; 0.003; 0.005; 0.007; 0.010; 0.012; 0.020; 0.050. Ca(OH)2 (0.25 mol/L) suspension was prepared by milling the required amount of CaO in 200 mL of water in a ball mill. The required amount of AgNO3 was dissolved in water (0.01 mol/L). Silver nitrate solution was added to the calcium hydroxide suspension in correct proportions and heated up to 90 °C with continuous stirring. Phosphoric acid (2 M) was added dropwise to the suspension and pH was controlled during the synthesis process.

In the case of method B: An appropriate amount of silver nitrate was added, with atomic ratios Ag/(Ag + Ca) = 0.003; 0.007; 0.020. The required amount of Ca(NO3)2·4H2O (0.25 mol/L) and AgNO3 (0.01 mol/L) was dissolved in water. Silver nitrate solution was added to calcium nitrate solution in correct proportions and heated up to 90 °C with continuous stirring. Ammonium phosphate dibasic (0.24 mol/L) was added dropwise to the solution and pH was adjusted to 8.0 with ammonium hydroxide solution.

The obtained suspensions were aged for 15 h, filtered and dried at 100 °C. Samples were sintered as scaffolds to determine the phase composition changes after sintering. Before sintering, powders were uniaxially pressed in 2 mm thick scaffolds (with a diameter of 10 mm). The sintering of samples was carried out at 1,000 °C for 2 h. All synthesis was repeated three to five times.

2.2 Characterization of the materials

All as-synthesized powders and sintered scaffolds were analyzed by X-ray powder diffraction (XRD, PANalytical zX’Pert Pro diffractometer, The Netherlands) using Cu Kα radiation at 40 kV and 30 mA, 2θ range of 10°–55°. The obtained patterns were compared to the pattern from the literature [32] and phase identification was made by XRD phase identification software HighScore. The porosity was determined using Archimede’s method. To determine the silver content in powder samples, X-ray fluorescence spectrometry (XRF, BRUKER, Pioneer S4) was used. Scanning electron microscopy (SEM) was used to evaluate the distribution of bacteria and silver on the scaffold surfaces and within the structure. Samples were sputter coated with gold using a fine coat ion sputter device (Emitech K550X, QUORUM TECHNOLOGIES Company) and observed using a Tescan Mira/LMU scanning electron microscope, equipped with a Mira\\LMU Schottky-Emission electron gun with a ZrO/W emitter at an accelerating voltage of 15 kV and magnification of 3,000×, and an energy dispersive X-ray spectroscope (Oxford Instrument EDX/7378). N2 sorptiometry was performed (at −195 °C) with a KELVIN 1042 (COSTECH) automatic sorptiometer powered with the standard BET method for determination of specific surface areas (adsorptive gas N2, carrier gas He, temperature 120 °C). Thermal stability and thermodynamical properties were determined using optical dilatometry (EM201, HT163).

2.3 Determination of in vitro silver release

Sintered HAp/Ag scaffolds were used for the determination of in vitro silver release kinetics. The silver release was measured by placing three replicate samples from each batch in 15 mL of simulated body fluid (SBF, prepared by a method described elsewhere [33, 34]) and incubated at 37 °C at a rotation speed 50 rpm. The rate of silver release was determined by taking 2 mL of SBF solution after 1, 24 and 50 h. 2 mL of fresh SBF solution was added to samples to keep a constant incubation volume.

Inductively coupled plasma (ICP, ELAN DRC-e, PerkinElmer) spectrometry was used to determine silver ion release kinetics.

2.4 Determination of in vitro antimicrobial properties

Reference cultures of Staphylococcus epidermidis ATCC 12228 and Pseudomonas aeruginosa ATCC 27853 were used in the study. S. epidermidis is a gram- positive facultative anaerobic coccus, while P. aeruginosa is a gram- negative facultative anaerobic rod. Both organisms are capable of biofilm formation on implant surfaces [35–37]. Bacteria suspensions were prepared from the microbiological cultures in 1 mL of triptycase soy broth (TSB, Oxoid, UK) with concentrations of 10, 102, 103 CFU/mL (colony forming units). Samples were cultivated at 37 °C for 2 h for the determination of bacteria adhesion. HAp/Ag and HAp samples were incubated in 1 mL of TSB for 24 h, to evaluate the level of the microorganism colonization. For the determination of both bacteria adhesion and colonization, the sonication-plate count method [38, 39] was applied. The unattached microorganisms were washed off after incubation. To separate the bacteria attached on the sample surface, samples were treated in an ultrasonic bath for 1 min (with a frequency of 45 kHz) and for 1 min in a centrifuge vortex at the maximum rotational speed. To determine the total count of microorganisms, several subcultures were prepared on a TSA growth medium (Triptycase soy agar, Oxoid, UK) and cultivated for 24 h at a temperature of 37 °C. For CFU calculations, 1 mm2 of the biomaterial sample surface was used. Samples for SEM analysis were fixed in a mixture of ether–ethanol (1:1).

2.5 Statistics

Results are presented as the mean value ± SD (standard deviation) of three experiments. Statistical analysis was performed on adhesion and colonization intensity data with unpaired Student’s t test, and P < 0.05 was used as a limit to indicate statistical significance.

3 Results and discussion

3.1 Characterization of HAp/Ag

In this study, silver doped HAp was prepared, using two modified wet chemical methods with different raw materials: (A) from calcium oxide and phosphoric acid and (B) from calcium nitrate, diammonium hydrogen phosphate and ammonia solution. For both methods silver nitrate was used as a source of silver. All samples were used as as-synthesized powders for the determination of physical properties and sintered as scaffolds at a temperature of 1,000 °C to determine phase composition and physical properties. S. epidermidis and P. aeruginosa bacterial strains were chosen to evaluate the in vitro antibacterial activity of the sintered scaffolds at a temperature of 1,000 °C.

For the determination of the incorporated silver amount and calcium/phosphorus ratio within HAp/Ag powder samples, XRF analysis was used. XRF results (Table 1) showed that silver retrievability decreased alongside the increase of the added silver amount. Significant differences in the retrievable silver amount were observed between samples prepared by different methods, which can be explained by the possible formation of a stable ion complex [Ag(NH3)2]+ applying method B [40]. Silver nitrate and ammonium solution was used as raw materials for the synthesis B; therefore, the formation of a stable ion complex [Ag(NH3)2]+ was limited by the concentration of free Ag+ ions. As a result, less Ag+ ions could be incorporated within the HAp structure. It has been reported that such a complex is forming in the case of using ammonia at zinc substituted HAp synthesis [28].

Table 1 Silver retrievability in HAp/Ag samples
Full size table

The specific surface area results obtained by the BET method indicated that the increase of the silver amount within the structure increased the specific surface area of HAp/Ag particles, which means an increased grain size (Table 1). Specific surface area for pure as-synthesized HAp powder was 99 ± 2 m2/g, while for as-synthesized HAp/Ag powders, specific surface area decreased from 80 ± 2 to 35 ± 2 m2/g depending on the incorporated silver amount. The least specific surface area was obtained for samples prepared by method B with the silver amount of 0.70 ± 0.05 wt%. By BET results the authors of this study conclude that an equivalent diameter for all prepared particles is in nanosize from 19 ± 3 nm for pure HAp up to 55 ± 2 nm for silver doped samples.

The porosity of sintered HAp/Ag scaffolds and pure HAp scaffolds treated in the same conditions was evaluated. Results of the BET surface area showed that higher surface area was detected for samples prepared by method A; hence, that open porosity was expected to be higher for those samples. The results showed that the open porosity of HAp/Ag scaffolds prepared by method A was in the range of 32–35 %, but for the samples prepared by method B—in the range of 23–28 %. The authors of the present study assume that the 10 % difference in the porosity could change mechanical and antibacterial properties for the implant materials prepared under the same conditions.

The HAp/Ag phase composition was identified for as-synthesized powders and sintered scaffolds. Sintered scaffolds were milled in a mortar and used for further XRD analysis. The XRD pattern for HAp/Ag as-synthesized powders corresponded to the hexagonal symmetry of HAp (ICDD 00-009-0432) (Fig. 1). Amorphous phase was observed for samples prepared by method A.

Fig. 1
figure 1

XRD patterns of HAp and HAp/Ag as-synthesized powders

Full size image

Sintered and milled HAp/Ag scaffolds were found to contain additional phases to the pure hexagonal symmetry of HAp. The results demonstrated that the phase composition depended on the preparation method—in the case of method A silver and silver oxide AgO phases were present or in the case of method B an additional silver oxide phase was observed (Table 2; Fig. 2). As mentioned above, silver and silver oxide phases were observed for the sintered HAp/Ag samples [19, 23] and they corresponded to ICDD reference samples—namely, ICDD 01-076-1489 and ICDD 00-004-0783 in the cases of silver oxide silver respectively.

Table 2 Phase composition for HAp/Ag samples
Full size table
Fig. 2
figure 2

XRD patterns of sintered and milled scaffolds

Full size image

Apart from HAp characteristic peaks for all sintered HAp/Ag samples, the characteristic peak of silver oxide was observed at 2θ 37.5° and the peaks of silver were observed at 2θ 38.2° and 44.4° for HAp/Ag prepared by method A. The characteristic peak of silver at 2θ 44.4° was not observed for HAp/Ag samples with a small amount of silver. The phase composition of HAp/Ag prepared by method B corresponded to the HAp phase and the additional silver oxide phase with a characteristic peak at 2θ 37.5° . Since silver can oxidize to AgO at 1,000 °C, according to the reaction: \(2{\text{Ag}}\; + \;{\text{O}}_{2\;} \;\underset {} \leftrightarrows \;2{\text{AgO}}\) [41], scaffolds containing higher amounts of silver, distributed on the scaffold surface, contained higher amount of the AgO phase.

After the analysis of optical dilatometry results, it was found that sintering temperature was higher for silver doped samples. Sintering temperature for HAp/Ag samples prepared by method A was determined to be higher by 150 ± 10 °C than that for pure HAp. It was also observed that the silver content within the scaffold did not affect the sintering temperature (Fig. 3a). For all silver doped samples obtained by method A the sintering temperature remained in the range of 850 ± 10 °C. These results can be explained by the similarities in BET results. All of the samples prepared by method A showed a similar specific surface area. Sintering temperature for HAp/Ag samples prepared by method B was increasing alongside the increase of the incorporated silver amount within the structure. The incorporation of 0.1 % of silver resulted in a sintering temperature increase by 50 ± 5 °C, while the incorporation of 0.7 % of silver increased the sintering temperature difference between HAp/Ag and HAp by more than 200 ± 10 °C (Fig. 3b). The diffusion mechanism is the most important one in the sintering process and the densification rate is dependent on the grain size. Densification occurs by the flux of matter from the grain boundaries to the pores. Faster densification will occur for the samples containing smaller grain sizes [42]. The size of the HAp/Ag particles prepared by method B increased with the increase of the incorporated silver amount. Therefore, sintering temperature was also increasing with the increase of the incorporated silver amount. The results suggest that the HAp/Ag structure stability depends on the preparation method, because the samples prepared by method B are more stable at higher temperature.

Fig. 3
figure 3

Optical dilatometry data. a HAp and HAp/Ag prepared by method A. b HAp and HAp/Ag prepared by method B

Full size image

3.2 In vitro silver release

The silver release from sintered samples was determined. Samples for the release tests were chosen with a silver amount of 0.35 ± 0.04 wt%. Sintered scaffolds were placed in 15 mL of SBF and incubated at 37 °C and 50 rpm. The rate of silver release was determined by taking 2 mL of SBF solution after 1, 24 and 5 0 h. 2 mL of fresh SBF solution was added to samples for a constant incubation volume. The released silver amount in SBF from samples prepared by method B was determined to be more than twice higher than that for samples prepared by method A (Fig. 4). After 50 h incubation, 0.66 ± 0.11 % of incorporated silver was released from the scaffolds prepared by method A. At the same time, already 1.61 ± 0.22 % of silver was passed in SBF from the scaffolds prepared by method B. The difference in the released amount could be explained by the differences of the silver phases in the samples, as the sintered HAp/Ag scaffolds prepared by method B contain the silver oxide phase, but the sintered HAp/Ag scaffolds prepared by method A contain silver and silver oxide phases. Solubility in water solutions for silver oxide is higher than that for pure silver. There are other effects that can influence the silver release rate, for instance, scaffold solubility. The scaffolds prepared by both methods have been reported to be more soluble in acetate or phosphate buffer solutions compared to pure HAp [15, 20]. Another factor that can influence the silver release rate is the formation of a pure HAp layer on the surface of samples from the SBF containing ions. It can be concluded, that the silver release rate is dependent on several factors, which interact within and with scaffolds.

Fig. 4
figure 4

Released silver amount in SBF

Full size image

3.3 Antibacterial activity of sintered scaffolds

The antibacterial activity of HAp and HAp/Ag scaffolds prepared by both methods and sintered at 1,000 °C was evaluated against S. epidermidis and P. aeruginosa bacterial strains. The results showed that the adhesion of S. epidermidis and P. aeruginosa on the surface of scaffolds began when the bacteria were incubated for 2 h at the concentration of 102 CFU/mL. Adhesion of P. aeruginosa was determined only for pure HAp and HAp/Ag containing, 0.14 wt% silver and prepared by method B (Table 3). The adhesion of S. epidermidis and P. aeruginosa on HAp/Ag samples, containing 0.35 wt% Ag was not observed. The authors of this study suggest that this effect could be associated with the silver release rate, because the samples prepared by method B in the first 2 h show a higher silver release than those prepared by method B. Obtained data indicated that antibacterial activity increased with the increase of the incorporated silver amount. The lowest bacteria adhesion was observed for HAp/Ag scaffolds prepared by method B, which correlated with the silver release data. Released silver concentration for HAp/Ag scaffolds prepared by method B was higher in the first hours; consequently, the bacteria could not adhere to the scaffold surface.

Table 3 Intensity of adhesion
Full size table

The samples containing pure HAp did not demonstrate any antibacterial activity. The colonization intensity of P. aeruginosa on the HAp samples reached 139,303 CFU/mm2 (at a concentration of 103 CFU/mL), and 123,232 CFU/mm2 (at a concentration of 103 CFU/mL) in the case of S. epidermidis. The colonization activity of both bacteria strains was inhibited on the Ag-containing samples, especially on the samples prepared by method A with a silver amount of 1.86 wt% (Fig. 5). Only rare and separate S. epidermidis cells with the absence of glycocalyx were found on the surface of Ag containing biomaterials (Fig. 6). For samples prepared by method A silver oxide particles were observed on the sample surfaces (Fig. 7; Table 4). EDX analysis of particles approved the silver oxide phase. These results correspond to XRD data.

Fig. 5
figure 5

Colonization intensity of a S. epidermidis, b P. aeruginosa

Full size image
Fig. 6
figure 6

SEM images of HAp/Ag scaffolds—colonization of S. epidermidis

Full size image
Fig. 7
figure 7

SEM–EDX analysis of HAp/Ag scaffold prepared by method A (1.86 wt% Ag)

Full size image
Table 4 EDX analysis of HAp/Ag surface
Full size table

The colonization intensity of P. aeruginosa (at a concentration of 103 CFU/mL) for samples prepared by method B significantly differed. Colonization intensity was 140 ± 6 CFU/mm2 for samples containing 0.35 wt% silver, while for samples containing 0.14 wt% silver it was 507 ± 14 CFU/mm2 (Fig. 5). The pronounced inhibition of P. aeruginosa colonization was observed for HAp/Ag scaffolds prepared by method A. Physical properties for both materials were found to differ; therefore, the intensity of colonization for both bacteria on HAp/Ag prepared samples can be affected by other factors, such as surface morphology, silver distribution, silver release etc.

The colonization test results showed that the bacterial growth was limited on all samples containing Ag and depended on the preparation method. A pronounced inhibition of Staphylococcus and Pseudomonas colonization was observed on HAp/Ag samples prepared by method A and containing 1.86 wt% silver.

4 Conclusions

The phase composition and thermal stability of the HAp/Ag mixed crystals depends on the silver amount incorporated in the samples and the preparation method. Sintered HAp/Ag samples prepared by method A contain three phases, i.e. HAp, silver and silver oxide, while sintered HAp/Ag samples prepared by method B contain two phases, i.e. HAp and silver oxide. The silver release rate and, consequently, the antibacterial activity of sintered scaffold can be controlled by choosing an appropriate preparation method. HAp/Ag samples prepared by method B presented lower bacteria adhesion but higher colonization intensity and therefore better antibacterial properties can be gained for HAp/Ag samples prepared by method A. The presence of silver oxide particles on the surface of scaffolds can interact even more with the bacterial strains, suggesting that biomaterials containing Ag (especially, prepared by method A) could be a better choice for the practical use to prevent device associated infections.