Introduction

Harmful microorganisms may cause some serious problems to human health such as watery diarrhea to severe dysentery and fever, if they are not controlled properly (Chen et al. 2010; Liga et al. 2013; Mitoraj et al. 2007; Seo et al. 2007). Antibacterial using different methods or materials have been widely studied (Burt 2004; Feng et al. 2000; Marambio-Jones and Hoek 2010; Molan 2006; Pal et al. 2007). Of the achieved so far, titanium dioxide (TiO2) and silver (Ag) nanoparticles (NPs) have attracted more attention because of cost-saving, stability, and high efficiency, although they have different antibacterial mechanisms. For the TiO2 particles, the bactericidal property was firstly reported by Matsunaga et al. who found that the Coenzyme A of microbial cells can be oxidized photo electrochemically with platinum-loaded titanium oxide, leading to the inhibition of respiratory activity and to the death of bacteria (Matsunaga et al. 1985). Specifically, the infiltration and photocatalytic oxidation of TiO2 NPs can lead to the damage of the plasma membrane and loss of enzymatic activity which result in the metabolic inactivation of the microorganisms (Thabet et al. 2013). The antimicrobial activity of TiO2 NPs may be strongly dependent on the high surface areas and the enhanced surface electronic effect (Thiel et al. 2007). However, the use of TiO2 is limited in only absorbing UV light (wavelength < 380 nm) due to its wide bandgap (~3.2 eV for bulk anatase TiO2).

Different from the TiO2 NPs, Ag NPs can incorporate bacteria surface and intracellular uptake of the bacteria (e.g., Escherichia coli, noted as E. coli), leading silver to binding with sulfur-containing membranes and cytoplasmic proteins to inactivate bacteria (McDonnell and Russell 1999; Pal et al. 2007). In addition, silver ions (Ag+) can inhibit phosphate uptake and exchange in E. coli and cause efflux of accumulated phosphate, which finally results in cell collapse (Schreurs and Rosenberg 1982). Silver ions may also lead to disorder of DNA replication and detachment of the cytoplasm membrane from the cell wall (Feng et al. 2000; Yamanaka et al. 2005). Silver atoms and ions (Ag0 and Ag+) have been found effective antimicrobial performance in wound dressing (Ip et al. 2006), textiles (YeonáLee et al. 2007), bone implants (Schneider et al. 2008), and self-sterilizing (Loher et al. 2008).

In addition, Ag nanoparticles can be found in diverse commercial products, especially food storage materials. However, the cytotoxicity of Ag NPs has caused wide concerns by scientists and engineers in the past (Liu et al. 2011 and relevant references in this paper). In vitro cell line studies have shown decreased mitochondrial function after exposure to Ag NPs in human skin carcinoma cells, hepatoma cells, alveolar epithelial and macrophage cell lines, human epidermal keratinocytes and fibroblasts, human lung fibroblast cells (IMR-90), and glioblastoma cells (U251). Ingestion of Ag can cause agyria, the benign condition characterized by the bluish-graying of the skin that occurs through the preferential deposition of Ag in the basal lamina of soft tissues such as the liver, and spleen and blood vessels, gastrointestinal tract, and kidney. Liu et al. demonstrated that oxidative stress might contribute to Ag NPs cytotoxicity, though the exact mechanism behind Ag NPs toxicity is suggested oxidative stress and lipid peroxidation playing an important role in Ag NPs elicited cell membrane disruption, DNA damage, protein damage, and subsequent cell death; however, to reveal whether apoptosis involved in Ag NPs toxicity, further studies are underway.

Furthermore, although the TiO2 nanoscale colloids have been widely used in sunscreen for anti-UV irradiation, the potential toxicity of such photo-activated TiO2 particles on exposure of humans and the environment remains unknown. A good example in this area reported by Petkovic (Petković et al. 2011 and relevant reference in this paper), who studied whether pre-irradiation of TiO2 particles with UV influences their cytotoxic and genotoxic potential in human hepatoma HepG2 cells. They found that the non-irradiated TiO2 at 1.0–250 μg/ml did not reduce viability of HepG2 cells, nor induce significant increases in DNA strand breaks; only TiO2(<25 nm in diameter) induced significant increases in oxidative DNA damage. After UV pre-irradiation, both TiO2 (<25 nm) and TiO2 (>100 nm) reduced cell viability and induced significant increases in DNA strand breaks and oxidative DNA damage. That is, UV pre-irradiation of anatase TiO2 particles results in increased cytotoxic and genotoxic potential. In our study, no human cells but E. coli bacteria was tested, and there may be some common toxic mechanisms from Ag or TiO2 nanoparticles in acting on both human cells and E. coli bacteria. However, more work needs to be performed for detailed investigations regarding the biosafety of Ag and TiO2 nanoparticles.

What will happen for the hybrid TiO2 and Ag nanostructures, either Ag@TiO2 core–shell structure or Ag-decorated TiO2 structure in antibacterial? Our recent studies revealed that for the Ag@TiO2 core–shell nanocomposites, the photogenerated electron-hole (e−-h+) pairs could be efficiently separated due to Ag as an electron trap to facilitate charge separation and hence enhance photocatalytic efficiency, with nearly double efficiency compared to the pure TiO2 NPs and P25 TiO2 (Wang et al. 2012; Wu and Long 2011). The Ag@TiO2 core–shell particles may also improve the Ag stability and longevity in antibacterial, however, little reported in the literature. It is also expected that superior antibacterial performance will be delivered by the Ag@TiO2 core–shell nanostructures to individual TiO2 or Ag nanoparticles. A number of methods have been proposed for fabricating Ag-TiO2 nanocomposites, for example, Ag@TiO2 core–shell nanoparticles can be prepared by a so-called one-step route using N,N,-Dimethylformamide (DMF) to reduce silver nitrate, and TiO2 was anchored to the metallic core by hydrolysis of titanium isopropoxide (Zhang et al. 2006). The metal@TiO2 core–shell nanocomposites can also be obtained by a hydrothermal treatment of TiF4 and colloidal noble metal particles (Zhang et al. 2011). Isabel et al. reported a nanostructure composed of silver cores and 1–2 nm layer of TiO2 produced by a simultaneous reduction of silver and condensation of titanium butoxide (Pastoriza-Santos et al. 2000).

In this study, a facile but effective synthesis method will be developed to prepare hybrid Ag-TiO2 nanostructures under mild conditions. Different from other synthesis methods, this approach can be operated under mild condition (≤100 °C, 2 h). The as-prepared core–shell structures show well coating of TiO2 on Ag particles, high surface area of TiO2 layer, mixed anatase, and rutile phases, and have been proved with significantly high photocatalytic property (Yang et al. 2013). The microstructure, composition, and properties will be characterized by various advanced techniques, such as transmission electron microscope (TEM), energy dispersive X-ray spectroscopy, X-ray diffraction (XRD), Brunauer Emmett Teller (BET), and ultraviolet-visible (UV-Vis) spectroscopy. In particular, the novelty of this work is to test and compare the antibacterial ability of Ag@TiO2 core–shell nanoparticles and Ag-decorated TiO2 nanoparticles in both bacterial growth inhibition and killing grown bacterial (e.g., E. coli). Finally, the antibacterial mechanisms of different Ag-TiO2 nanostructures and the relationship between structure and antibacterial behavior will be discussed. The significance of these findings is to help developing highly efficient photocatalysts for environmental remediation in terms of antibacterial, deodorisation, and water purification.

Experimental section

Preparation of Ag@TiO2 core–shell structure

Synthesis of Ag@TiO2 core–shell structure was conducted, by reference to our recent studies (Yang et al. 2013), but some modifications have been made. In a typical protocol of generating Ag cores, a few steps were involved. First, 10 mL of EG (ethylene glycol) solution containing a small amount of polyvinylpyrrolidone (PVP, 0.050 g) was stirred vigorously until all the PVP dissolved completely and the solution became transparent. 0.02548 g of AgNO3 (0.15 mmol) was dissolved in 5 mL of EG solution, followed by vigorous stirring until the solution became homogeneous. Second, two solutions were put into a 25-mL vial and mixed homogeneously by stirring, followed by placing the vial into a Teflon-lined stainless steel autoclave (50 mL capacity). The autoclave was heated in an oven and maintained at a temperature range of 80–180 °C for different times (30 min to a few hours). The color of the reaction system changed from light yellow to brown. Finally, Ag nanoparticles were collected by centrifuge and washed several times with ethanol and water for further use.

The preparation of Ag@TiO2 core–shell structures could be demonstrated as below: First, 0.05 mL titanium (IV) butoxide 97% (TBT) was added to 10 mL ethylene glycol (EG). The mixture was magnetically stirred for ~8 h at room temperature, labeled as solution A. Second, 3 mL of 0.01 M Ag colloids suspension was poured into 10 mL acetone under stirring for 5 min, labeled as solution B. Third, 0.5 mL solution A was added into solution B, leaving it for 1 h without disturbance. Some yellow brown precipitates were collected by centrifugation and then refluxed in boiling water for ~2 h. Finally, to get Ag@TiO2 core–shell powders, the solution was centrifuged at a rate of 3000 rps for 10 min, and the precipitates were washed with alcohol and water several times, followed by drying in an oven at 60 °C for 24 h for further characterizations.

Preparation of TiO2@Ag surface-doped nanoparticles

TiO2 colloids were synthesized by following the approach described in our recent study but with some modifications (Yang et al. 2012). To generate Ag-decorated TiO2 nanostructures with well control, 0.01 mM of TiO2 colloids was dispersed in 15 mL EG solution and stirred to make sure homogeneously. Then, 0.01 mL of 0.1 M AgNO3 and 60 µL hexylamine were added to the above solution simultaneously, and then left without disturbance at room temperature for 2 h to allow the Ag particles to deposit onto the surface of TiO2 nanospheres. Finally, the precipitate was collected by centrifugation, washed with ethanol and water for several times, following by oven drying at 60 °C for 6 h.

Characterizations

Various advanced techniques were employed to characterize the as-prepared nanoparticles, including:

  1. (i)

    Irradiation of the TiO2/Ag hybrid nanostructure was carried out in a quartz cuvette under UV lamp (NEC, 15 W, 300–400 nm, maximum emission 350 nm, Germany);

  2. (ii)

    Transmission electron microscopy (TEM) images were conducted with a JEOL1400 operated at an acceleration voltage of 100 kV, in which the specimen was prepared by dropping the particle suspension onto copper grids covered with amorphous carbon and air drying;

  3. (iii)

    Energy dispersive spectroscopy (EDS) analysis was conducted on an energy dispersive X-ray spectrometer attached to a Philips CM200 emission scanning electron microscope;

  4. (iv)

    The UV-Vis absorption spectrum was obtained on a CARY 5G UV-visible spectrophotometer (Varian) with a 1 cm quartz cell;

  5. (v)

    Powder X-ray diffraction (XRD) pattern of the optimized product was recorded on a Philip MPD diffractometer with Cu Kα radiation, the analysis was done by analysis software XRDUSEM;

  6. (vi)

    X-ray photoemission spectroscopic (XPS) analysis was conducted with a Physical Electronics PHI 5000 Versa probe spectrometer with Al Kα radiation (1486 eV); Analysis of the spectra was conducted using the Physical Electronics Multipak software package; and

  7. (vii)

    Specific surface area of the samples was measured by N2 adsorption at 77 K using the BET method (Tristar, Micromeritics); Prior to the measurement, the samples were pre-treated at 150 °C under vacuum (VacPrep, Micromeritics) for at least an hour to remove surface-adsorbed moisture and other volatile organic impurities.

Bacterial strains cultivation

Inactivation experiments were performed in pure culture of E. coli O157:H7 strains and E. coli BJ5183 strains. Both were cultivated in LB medium (10 g L−1 tryptone, 5 g L−1 yeast extract, and 5 g L−1 NaCl) at 37 °C for 8 h. Optical densities and viable cell numbers were periodically measured to observe the growth behaviors of the strains. Viable cell counts were determined by tenfold serial dilution of 1 mL broth culture in phosphate buffer solution (PBS), followed by inoculation of 0.1 mL aliquots on nutrient agar, by incubation of the plates at 37 °C for 24 h prior to colony plate counting.

Growth inhibition

Escherichia coli inoculums were prepared by growing culture in an incubator at 37 °C overnight and 280 rpm in 50 mL LB medium from a single agar plate colony. A measured volume of 2 mL of the culture was transferred into 100 mL fresh LB and further conditioned for 0.5 h at 37 °C and ~280 rpm to prepare the inoculum. For each growth inhibition experiment, 5 mL E. coli inoculums was added into 50 mL LB followed by different loadings of nanoparticles. The experiments were performed in triplicate with hourly sampling measured by optical density at 600 nm (OD600) using a UV-Vis spectrophotometer (Hitachi). Two types of control were employed: a growth control without nanoparticles and the nanoparticles control without E. coli, of which the latter was served as a reference to obtain the OD600.

Bactericidal test

The bactericidal activities of Ag@TiO2 or TiO2@Ag nanocomposites were determined on each strain, respectively. Disinfection efficiencies of the as-prepared samples were determined under various periods of UV light pre-irradiation. After UV pre-irradiation, inactivation of 7Log10 CFU/mL inoculum concentrations was performed by adding 1 mL of 7Log10 CFU/mL bacterial suspensions into 9 mL suspension containing 80 mg/L Ag@TiO2 core–shell or TiO2@Ag nanoparticles at 25°C. After 5-min exposure, 1 mL aliquot of treated sample was neutralized in PBS for 3 min. The surviving bacteria population was determined through viable cell counting. All results were presented as an average value of triplicates. The bacterial population survival ratio can be defined as follows:

$${\text{Population survival ratio}}\; = \;1 - \frac{{\log \left( {\text{A}} \right) - \log \left( {\text{B}} \right)}}{{\log \left( {\text{A}} \right)}} \times 100\,\%,$$

where A is the number of original bacterial population, and B is the number of bacterial population after inactivation process.

Statistical analysis

The quantitative data of antibacterial tests were obtained from triplicate experiments on three samples, shown as mean value ± standard deviation (SD).The difference across groups was analyzed by one-way analysis of variance (ANOVA), followed by a Post-Hoc Bonferroni test. Among the groups, the results were further analyzed by a non-parametric Kruskal–Wallis test, followed by a non-parametric Post-Hoc test. P (two-tailed) < 0.05 was considered as significant. Statistical analyses were performed using IBM SPSS Statistics. 21 (IBM Chicago, IL).

Results and discussion

Particle structure and composition

The structure of nanoparticles was characterized. Figure 1 shows TEM images that the Ag@TiO2 core–shell nanoparticles have Ag cores with diameters of 50–100 nm generated by a polyol-thermal approach (Fig. 1a) (Jiang et al. 2011b). The TiO2 colloids were synthesized with mild condition by water boiling titanium glycolate for 1–2 h (Fig. 1b), according to our recent report (Yang et al. 2012). Figure 1c shows the TiO2-coated Ag nanostructures, generated by a simple TiO2 coating method (Yang et al. 2013). TiO2 shell thickness could be tunable from a few to tens nanometers, depending on the ratio of titanium to silver (Jiang et al. 2003). The brownish Ag@TiO2 powder containing Ag, Ti, and O elements was identified by EDS technique (Fig. 1e). The peaks at 3.2 and 0.7 keV can be assigned to Ag and O element, the peaks at 0.3 and 4.6 keV are assigned to Ti element, while the observed peaks of Cu may be originated from the copper grid used for the TEM observation.

Fig. 1
figure 1

TEM images of a Ag nanoparticles, b TiO2 nanoparticles, c Ag@TiO2 core–shell nanocomposites, d TiO2 surface decorated by Ag nanocomposites, e and f EDS spectra for c and d nanocomposites

Full size image

Meanwhile, the nanostructures of TiO2 surface decorated by Ag colloids (Fig. 1d) were prepared via a developed hexylamine reduction approach, in which the silver nitrate was dispersed in ethylene glycol containing evenly dispersed TiO2 nanospheres, then reduced by hexylamine and finally the Ag colloids deposited on the surface of TiO2 nanoparticles. Figure 1d shows that the small silver NPs with size of 3–10 nm (dark dot) deposited on the surface of TiO2 nanospheres (~100 nm in diameter). From the EDS spectrum shown in Fig. 1f, an intensive Ag peak compared to that from the Ag@TiO2 core–shell particles was observed and centered at around 3.2 keV, probably caused by the exposed Ag particles on TiO2 surface. The elemental Ti and O peaks can also be clearly identified in the EDS spectrum.

XRD analysis

To identify the composition, the XRD technique was applied in this work (Fig. 2). The diffraction peaks centered at 38.12°, 44.3°, 64.4°, and 77.4° can be assigned to Ag (111), (200), (220), and (330) crystal planes of the face-centered-cubic (fcc) silver, respectively. While for those centered at 25.84°, 38.48°, 48.44°, 54.98°, and 63.62° could be assigned to the anatase TiO2(101), (004), (200), (211), and (204) crystal planes, respectively (Jiang et al. 2003). For the Ag@TiO2 core–shell nanostructures, the XRD pattern shows combined peaks originated from both Ag and anatase TiO2 (Fig. 2a). The Ag content in such nanostructures is estimated to 58.4 wt% (Table 1). With careful inspection, the presence of Ag2O and rutile TiO2 phase in the Ag@TiO2 core–shell structures could also be proved by the XRD technique. It was noted that the Ag2O has been demonstrated to impart additional antimicrobial activity in previous reports (Bellantone et al. 2002; Gunawan et al. 2009; Wang et al. 2009).

Fig. 2
figure 2

XRD patterns of the as-prepared Ag@TiO2 core–shell nanocomposites A and TiO2 surface decorated by Ag nanocomposites B

Full size image
Table 1 Comparison of Ag@TiO2 core–shell and TiO2@Ag surface-doped structures
Full size table

Figure 2b shows the XRD pattern of TiO2@Ag nanoparticles, in which the diffraction peaks positioned at 2θ = 37.71 , 43.83 , 63.71 , and 76.47° could be assigned to metal silver, whereas the peaks centered at 25.27°, 38.51°, 47.98°, 54.99°, 62.57°, and 74.9° could be assigned to the anatase TiO2 phase. The Ag content in such nanostructures is estimated to 17.4 wt% (Table 1). Surprisingly, no rutile TiO2 and Ag2O were detected and observed in such Ag-decorated TiO2 nanoparticles. This means that the Ag@TiO2 core–shell nanostructures are composed of more compositions such as rutile TiO2 and Ag2O, which may play a role in the photocatalysis, as discussed in our recent work (Yang et al. 2013).

XPS elemental analysis

To make clear, the Ag valence states (Ag+ or Ag0) are very helpful for understanding the antimicrobial function. The composition in the Ag@TiO2 core–shell structure was also investigated by XPS analysis. The characteristic energy spectra consist of five kinds of atoms: titanium (Ti), oxygen (O), silver, a trace amount of carbon (C), and nitrogen (N) in the survey spectrum (Fig. 3a). The existence of C may be ascribed to the adventitious hydrocarbon from the XPS instrument itself. The photoelectron peaks for Ti2p, O1s, and Ag3d were detected and analyzed with binding energies (BE) around 459, 5301, and 3672 eV, respectively.

Fig. 3
figure 3

XPS spectra of Ag@TiO2 core–shell nanocomposites a full spectrum and b–d spectrum for elemental Ti, O, and Ag, respectively

Full size image

To further confirm the elemental presence status, the high energy resolution XPS spectra of Ti2p region and O1s region are displayed in Fig. 3b, c, respectively. The peaks for Ti2p appear at 459.6 eV (Ti2p3/2) and 465.4 eV (Ti2p1/2), with slight shifting toward higher binding energies compared to those of the pure anatase TiO2 (Ti2p3/2= 458.8 eV and Ti2p1/2= 464.8 eV) (Jiang et al. 2011a). This is probably caused by the electrons in Ti–O bond transfer partially to the grain interface between Ag and TiO2. The shifting of peaks in the high resolution O1s spectra (Fig. 3c) also indicates electrons immigration between Ti–O bonds (Orts-Gil et al. 2013). In addition, the binding energies of Ag3d5 peak, 368.3 eV and 367.5 eV were calculated which could be assigned to Ag0 and Ag2O (Fig. 3d), respectively. This indicates that the Ag species in the core–shell composites are in the form of Ag and Ag2O, consistent with the XRD characterizations (Fig. 2a).

Optical property

UV-Vis spectrometer was used to measure the optical properties of nanoparticles. Figure 4 shows the UV-Vis spectra for the as-prepared Ag@TiO2 core–shell structures and TiO2 surface decorated by Ag nanostructures and Ag nanoparticles (diameter of ~60 nm) which were obtained by a polyol-thermal approach (Jiang et al. 2011a, b), and pure anatase TiO2 nanoparticles which were prepared by a water boiling method (Yang et al. 2012). The surface plasmon resonance of the Ag nanoparticles centered at ~445 nm (Fig. 4a) weakened when coated with amorphous or anatase TiO2 nanoparticles, and also shifted in the absorption wavelength region.

Fig. 4
figure 4

UV-Vis absorption spectra of Ag nanoparticles A, pure anatase TiO2 nanoparticles B, Ag@TiO2 core–shell nanostructures C, and TiO2 surface decorated by Ag nanocomposites D, corresponding to those particles shown in Fig. 1 a–d

Full size image

For the Ag@TiO2 core–shell particles, the UV-Vis absorption peak is centered at ~320 nm (Fig. 4c), similar to that of pure anatase TiO2 particles (Fig. 4b). That is, the surface plasmon resonance of Ag nanoparticles is heavily reduced by the surface coated TiO2. While for those TiO2@Ag nanostructures, the surface plasmon adsorption intensity of Ag (~445 nm) dramatically decreased, despite the position is nearly kept, which may be caused by the small size (3–10 nm colloids) or a small ratio in the whole structure. Moreover, a significant blue shift from ~320 to 260 nm occurred for the TiO2 absorption, when Ag deposited on the surface. However, the real reason is not clear yet at this moment.

BET surface analysis

Surface area or surface-to-volume ratio is critical for the functional properties of nanoparticles, and for their applications in catalysis, gas sensing and those related to contact surfaces. The surface area of the Ag@TiO2 core–shell or Ag-decorated TiO2 nanostructures may highly affect the antibacterial performance. In this study, the BET surface analysis was conducted. The N2 adsorption–desorption isotherms and the pore size distribution for Ag@TiO2 core–shell structure are shown in Fig. 5. The isotherms can be ascribed to type IV (BDDT classification), indicating the presence of mesoporous pore structures. The BET surface area of the as-prepared Ag@TiO2 core–shell structure was estimated to ~170 m2/g, and the average pore size was 4.9 nm. Similar procedures to the Ag-decorated TiO2 nanostructures were carried out, and the BET surface area was estimated to ~151 m2/g, with average pore size of 5.4 nm. All the obtained data are listed in Table 1.

Fig. 5
figure 5

BET surface area test for Ag@TiO2 core–shell structures showing SAg@TiO2= ~170.6 m2/g, and average pore size is 4.9 nm

Full size image

Bacterial growth inhibition

The bacterial growth inhibition capability of hybrid TiO2 and Ag nanostructures was assessed using E. coli O157:H7, as a target microorganism. A calibration curve was made at the beginning of growth inhibition test to determine the conversion factor between OD reading at 600 nm and the number of colony forming units growing (per mL) (Fig. 6). Under standard operation conditions and in the absence of nanoparticles, E. coli growth in Luria Bertani (LB) nutrient media was characterized by a relatively short lag phase of 1 h followed by 3–4 h active log phase before entering stationary phase after 6 h (See Fig. 7 control curve). As expected, the trend of bacterial growth inhibition increases at higher particle loadings. Figure 7 shows the kinetics of E. coli growth in the presence of different loadings of the as-prepared Ag@TiO2 core–shell particles. The control trial without any particles was conducted as a reference. The presence of 5 mg/L Ag@TiO2 has a little effect on inhibiting bacterial growth; while at 50 mg/L, a prolonged lag phase was observed. The presence of 100 mg/L of Ag@TiO2 core–shell particles could completely suppress bacterial growth under 6 h. At this stage, the presence of Ag0 atoms or Ag+ ions may play an essential role in inhibiting bacterial growth (Gunawan et al. 2009).

Fig. 6
figure 6

Calibration curve to calculate the conversion factor between OD reading at 600 nm and the number of colony forming units growing per mL (CFU/mL)

Full size image
Fig. 7
figure 7

Kinetic study of E. coli growth in Luria Bertani nutrient media at 37 °C in the presence of different loadings of the as-prepared Ag@TiO2 core–shell particles as well as without any particles (control). There were significant differences across the five groups, (P < 0.001). 100 mg/L group had a significant difference as compared with any of other four groups (P ≤ 0.015). There was no significant difference between the other four groups

Full size image

In comparison, Ag-decorated TiO2 nanostructures were tested in the bacterial growth inhibition. Under the same concentration of 80 mg/L, three types of nanoparticles Ag@TiO2, TiO2@Ag, and pure TiO2 particles were homogenously dispersed into the LB media, respectively. Then the E. coli inoculum was transferred into the media suspension for observation. Figure 8 shows the E. coli growth trends in the presence of these three kinds of nanoparticles. It was found that the Ag decorated TiO2 nanostructures show the most effective performance in bacterial growth inhibition, where no growth indication was observed up to 8 h. The small Ag particles depositing on the surface of TiO2 support makes it easier to release Ag+ ions in the aqueous culture media, which consequently contributes to the observed bacterial growth inhibition.

Fig. 8
figure 8

Kinetic study of E. coli growth in Luria Bertani nutrient media at 37 °C in the presence of 80 mg/L of the as-prepared particles as well as a blank. There were significant differences across the five groups, (P ≤ 0.002). Significant differences were found between TiO2@Ag and Ag@TiO2 groups (P = 0.010), TiO2@Ag and control (P = 0.015), and TiO2@Ag and TiO2(P  = 0.009). There was no significant difference between other groups

Full size image

For the Ag@TiO2 core–shell nanoparticles, the bacteria may grow in the first 2 h, but no active lag phase growth afterward. On the contrary, there was insignificant effect on bacteria growth suppression for the pure TiO2 nanoparticles, similar to the blank with control. However, the reason why pure TiO2 has no contribution but Ag-decorated TiO2 nanostructures can inhibit E. coli growth needs to be addressed, as below. Without UV light irradiation, TiO2 has less photo-activity and generates less electron-hole pairs, and thus weaken performance in E. coli growth inhibition, as shown in Fig. 8. While for the Ag@TiO2 core–shell ones with TiO2 shells, better inhibition control of E. coli growth suggests that it is the silver core nanoparticles that may play a key role in suppressing bacterial growth. The principles underlying antibacterial by silver colloids have been extensively discussed in the literature (Gunawan et al. 2009; Sotiriou and Pratsinis 2010; Xiu et al. 2012). Some researchers demonstrated that the Ag particle size effects on antibacterial activity (Sotiriou and Pratsinis 2010), e.g., for smaller ones (<10 nm), Ag+ ions easily released, and the antibacterial activity was mainly dominated by Ag+ ions rather than Ag0 atoms (Xiu et al. 2012); while for larger Ag ones (>20 nm), the particle itself influenced the antimicrobial activity, despite of less Ag+ ions released. That is, Ag+ ions are the definitive molecular toxicant, and the antimicrobial activity could be controlled by modulating Ag+ ions release. One possible reason for the Ag@TiO2 core–shell structures to show good performance in growth inhibition is probably that the Ag2O (~14%) maybe easy to produce Ag+ ions. Another reason is the Ag can be used as an electron trap for electrons generated by TiO2 shell, which may further reduce the possibility of electron-hole recombination, and benefit for generating more ·OH radicals on the core–shell particle surface to inhibit bacterial growth. It needs to point out that the detailed information regarding the relationship between Ag+ ions release and antibacterial requires more experimental and theoretical work to do in the near future.

Bactericidal assessment

In addition to suppressing bacteria growth from inoculum, antibacterial activities of the as-prepared nanocomposites were also assessed in killing the grown bacteria, when they are in stationary phase. Under dark conditions, the hybrid Ag-decorated TiO2 nanostructures show poor capability to kill the grown bacteria (Fig. 9). When the concentration of nanocomposites increased to 200 mg/L in aqueous solution, the percentage of survival bacterial population could still be over ~90%.

Fig. 9
figure 9

Dose-dependent growth inhibitory effect of Ag@TiO2 core–shell particles and the nanostructures of TiO2 surface decorated by Ag particles on E. coli (O157:H7) in dark (UV-free) condition. There were significant differences across two groups (P ≤ 0.012). Error bars correspond to the standard deviation (SD) between three independent measurements

Full size image

On the contrary, after UV irradiation, both types of nanocomposites Ag@TiO2 core–shell and Ag-decorated TiO2 nanostructures exhibited highly enhanced bactericidal activity. Figure 10a shows the survival rate of the grown bacteria population after mixed with nanoparticles at a concentration of 80 mg/L, with different UV irradiation times prior to testing. The cultured E. coli O157:H7 was initially diluted with LB media to an optical density of 0.05 at 600 nm, about 107 colony forming units (CFU)/mL. The survival of E. coli was estimated by a plate count method. It was found, for the UV irradiated pure TiO2 particles for 240 min, a minimal effect was observed on bacteria inactivation, where >80% bacteria still survived.

Fig. 10
figure 10

a E. coli O157:H7 and b E. coli BJ5183 inactivation by pure TiO2 particles, Ag@TiO2 core–shell particles and the nanostructures of TiO2 surface decorated by Ag particles under different UV irradiation times. There were significant differences across three groups (P ≤ 0.036). No significant differences were found between Ag@TiO2 and TiO2 (P = 0.042). There was no significant difference between other groups. b There were significant differences across three groups (P ≤ 0.050), while insignificant difference between Ag@TiO2 and TiO2 (P = 0.018) was found. There were no significant differences between other groups

Full size image

Moreover, the UV irradiated Ag@TiO2 core–shell and Ag-decorated TiO2 nanostructures exhibited stronger killing bacteria capacities, related closely to irradiation time. For those Ag-decorated TiO2 surface nanostructures, the bacteria inactivation performance gradually increased with increasing the UV irradiation time, e.g., bacterial killing 20 % and 50 % after 120-min and 240-min irradiation, respectively. The bactericidal activities of the Ag@TiO2 core–shell particles were remarkably enhanced under the same conditions. After 240-min UV irradiation, they can kill more than 90 % E. coli O157:H7. As a further confirmation, E. coli BJ5183, another strain of E. coli, was employed to repeat the assessment (Fig. 10b), and similar results in killing bacterial were confirmed. That is, the bacterial killing capabilities are in the following order: Ag@TiO2 (core–shell) > TiO2@Ag (surface decorated with Ag) > pure TiO2.

To well understand the bactericidal performance, Scheme 1 was used to illustrate the working principles of hybrid silver-titania nanostructures toward bacteria inactivation. For the Ag@TiO2 core–shell nanostructures, the ground-state electrons at the valence band (VB) are excited to the conduction band (CB) under UV irradiation. Then the CB electrons subsequently migrate to the Ag in a typical Schottky-type mechanism, and the Ag cores attract more electrons from TiO2 (Chuang and Chen 2009). The holes generated at the valence band react with water molecules to produce hydroxyl radicals (OH·), and partial electrons may also transfer to the adsorbed oxygen on particle surface to produce superoxide anions (∙O2). In addition, Ag cores may release some silver ions in this structure due to the formation of Ag2O. All these factors will benefit for the bacterial inactivation. On the contrary, for UV irradiated TiO2 surface decorated by Ag nanostructures, the circumstance is a little complicated, because the surface plasmon resonance of silver leads hot electrons to transferring to the TiO2 (Cushing et al. 2012), while the newly formed Ag+ ions on TiO2 surface may accept photogenerated electrons from TiO2 to be reduced to Ag0.

Scheme 1
scheme 1

Illustrating the working species toward microbe inactivation of a Ag@TiO2 core–shell nanoparticles, and b Ag-decorated TiO2 nanoparticles under UV light irradiation

Full size image

Based on the above analysis, it can be assumed that the hydroxyl radicals (OH·) and superoxide anion (·O2) play a crucial role in the antibacterial activity than the Ag0 and Ag+ ions under UV irradiation. Under the same conditions, more OH· and ·O2 radicals were produced in the Ag@TiO2 core–shell structures than in the Ag-decorated TiO2 surface nanostructures, leading to an enhanced performance in killing bacteria.

Conclusions

In this study, we have demonstrated facile but effective methods to prepare hybrid Ag-TiO2 nanocomposites: Ag@TiO2 core–shell and Ag-decorated TiO2 surface nanostructures. They exhibit different capabilities in antimicrobial (i) without UV irradiation, the Ag-decorated TiO2 surface nanoparticles show superior capability in bacterial growth inhibition to Ag@TiO2 core–shell ones or pure anatase TiO2 particles, in which Ag0 and Ag+ ions generated on TiO2 surface play a key role in antimicrobial; (ii) under UV light irradiation, the Ag@TiO2 core–shell nanocomposites exhibit superior bactericidal performance to others, because Ag cores tend to facilitate charge separation for the photogenerated TiO2, and hence produce more hydroxyl radicals (OH·) and superoxide anion (·O2) that play a critical role in antibacterial activity; and (iii) two types of bacteria, E. coli O157:H7 and E. coli BJ5183 were tested and assessed in antibacterial behavior, and the similar trend was found for both Ag@TiO2 core–shell and Ag-decorated TiO2 nanostructures. However, we need to point out that the quantitative information regarding the relationship between the produced hydroxyl radicals (OH·) and survival bacterial, and the released Ag+ ions from Ag cores or Ag2O and their impact on the antibacterial activities need to be further investigated, experimentally and theoretically. These findings would be useful for design and construction of antibacterial nanomaterials with a high efficiency. This study may open a new path to generate hybrid metal and metal oxide nanostructures with superior functionalities for energy and environmental applications.