1 Introduction

Nano-materials, due to their large effective surface area with a potent number of reactive sites, are extensively used in biological applications, which include antimicrobial activities, drug delivery and medical imaging [1, 2]. The study of antimicrobial activity of nano-materials is essential for biological applications, especially to develop antimicrobial medicines [3], and to determine the suitability of a material for antimicrobial surfaces [4]. This is important, as the increasing resistance of microorganisms to multiple antibiotics has fueled the demand for effective, resistance-free, cheaper and biocompatible antimicrobial agents.

Nano-materials provide a novel way to replace antibiotics. Nano-particles, for instance, have been used to reduce skin diseases [1, 5] and to prevent the microbial colonies that form on the surface of devices like endotracheal tubes, catheters and prostheses [6]. Although nano-particles of metal (silver, gold etc.) and some oxides (like TiO2, ZnO, etc.) show convincing antimicrobial activity, their poor dispersion stability in the organic medium makes it difficult to fabricate polymer composites [7]. The drawback with silver nano-particles lies in the rapid oxidation and ease of agglomeration that changes their reactivity with the environment [8]. Besides this, it is an expensive material and its environmental hazards are a matter of concern [9].

Nano-structures of silica, on the other hand, are preferable over other nano-materials due to its noncorrosive and biocompatible properties. Also, they are chemically stable and not as expensive as silver or gold. The surface of silica structures can be functionalized in many ways that make its integration with polymers easier. Silicon derivatives with polymers have been used as anti-corrosive and chemically resistant coatings [10]. Nano-structures of silica can thus provide very useful antimicrobial properties to polymers [11]. Song et al. [12] have reported antimicrobial studies of silica nano-particles modified by polymer whereas Hebalkar et al. [13] have reported the synthesis of silica nano-particles for antibacterial and self-cleaning surfaces. Nano-particles of silica are used in biological applications [14, 15] as scaffolds for drug delivery by functionalizing their surfaces and attaching biomolecules [16]. Likewise, nanoparticles of silica are used to form composites with metal nanoparticles like silver to avoid their aggregation and hence enhance antibacterial activity [17]. They are also important with regard to dental applications, mainly fillings [18].

Furthermore, nano-silica showing a hollow structure is considered superior to solid structures, as hollow nano-structures possess a greater effective surface area. They are expected to provide the increased effective surface reactivity needed for charge accumulation, which can contribute to antimicrobial activity. Several methods, like pulsed laser ablation [19], chemical vapor deposition [20], and the template method involving sol–gel techniques [21, 22], have been reported for the synthesis of silica nanotubes. Most of these methods involve the use of harmful chemicals like silane and silane derivatives and some are catalyst dependent. Moreover, silica nanotubes synthesized by these methods have a wall thickness of the order of several nanometers. We present our work on the synthesis of silica-coated silicon nanoparticles (SCSNPs), and silica-coated thin-walled silicon nanotubes using catalyst-free thermal plasma-assisted gas phase condensation technique. Thermal plasma synthesis has the advantage of being a clean process which ensures a high yield [23]. It produces highly crystalline products that are free of contaminants, as it is a high temperature process. In addition, synthesized nanotubes have a wall thickness of less than one nanometer, which is rarely reported. The morphology and structure of silicon nano-structures was analyzed using transmission electron microscopy (TEM), scanning tunneling microscopy (STM) and Raman spectroscopy.

The antimicrobial activity of as synthesized nano-structures was studied for selected strains of bacteria which are pathogenic and frequently colonize medical devices. The Gram-positive bacteria included were Bacillus subtilis and Staphylococcus aureus, whereas Gram-negative bacteria were Escherichia coli and Pseudomonas aeruginosa. S. aureus is one of the major resistant pathogens found on the mucous membranes and human skin of around one-third of the population and it is very adaptable to antibiotics [24]. The antibacterial activity was assayed using the optical densitometry technique and viable cell counting method, using plating technique. Here we report the synthesis, antimicrobial activity and the potential use of ultra thin walled silica-coated silicon nano-structures in the formulation of new types of bactericidal materials.

2 Experimental details

2.1 Synthesis of nano-particles and nano-tubes

The nano-structures of silicon consisted mainly of nanoparticles and nanotubes, and these were synthesized in a DC arc plasma reactor. The plasma reactor consisted of a graphite anode (in the form of a crucible) and a tungsten cathode (in the form of a 4 mm diameter rod) mounted in a vertical geometry. The electrode system was housed inside a double-walled, water-cooled stainless steel chamber. The silicon powder, from CDH, India Ltd., was placed in the graphite anode. The schematics of the plasma reactor and the detailed morphological analysis of silica-coated, silicon nanotubes (SCSNTs) have been included in our earlier publications [23, 25]. The chamber was pre-evacuated to a base pressure of about 10−3 mbar and was filled with the desired gas to maintain a pressure of 700 mbar during arcing. The arc voltage and arc current were maintained at about 12 ± 0.4 V and 80 ± 5 A respectively. In such a reactor arcing generates a plasma plume, which produces heat that causes melting and evaporation of the anode material. Homogeneous nucleation of the evaporated atoms occurs on account of the steep temperature gradient outside the plasma plume. However, growth is restricted due to a high collision rate resulting from a high partial pressure of the inert gas that prevails in the chamber. The resulting nanoparticles then condense and are collected from the walls of the chamber.

Two batches of samples consisting of nano-silicon were prepared in the presence of (a) 100 % Ar gas and (b) 95 % Ar + 5 % H2 gas. The main constituents of batch ‘a’ and batch ‘b’ were SCSNPs and SCSNTs respectively.

2.2 Characterization of as synthesized nano-structures

The samples were dispersed in methanol and were coated on a holey carbon-coated copper grid for TEM analysis using a Technai G2 20 microscope, facilitated by a LaB6 filament and a charge coupled device (CCD) camera. Nanotubes were further studied by OMICRON STM in an ultra high vacuum (UHV) at room temperature. The elemental purity was revealed by energy dispersive X-ray (EDX) analysis in a scanning electron microscope (SEM) (JEOL model JSM-6360A). Raman spectra were recorded in a backscattering geometry at room temperature, using a Jobin-Yvon Labram HR800 spectrometer with a He–Ne laser (λ = 632.81 nm) for excitation. The effective surface area measurements were carried out by the nitrogen adsorption process in a single-point Brunauer, Emmett and Teller (BET) instrument (Smart Sorb 91 model).

2.3 Microbial strains and culture media

The bacterial strains were procured from the National Collection of Industrial Microorganisms (NCIM), India. The Gram-positive bacterial strains tested were S. aureus (NCIM 2079) and B. subtilis (NCIM 2063). The Gram-negative strains used were E. coli (NCIM 2065) and P. aeruginosa (NCIM 2200). The above bacterial strains were cultured in Nutrient Broth (NB) media (10 g yeast extract, 5 g sodium chloride, 10 g tryptone and 20 g Agar in one litre media with the pH maintained between 7 and 7.5). B. subtilis were pre-inoculated in 2 ml of NB media and grown at 30 °C overnight while all other bacteria were grown at 37 °C overnight. Except for B. subtilis, the population of microbes was maintained between 1 × 108 and 5 × 109 CFU/ml.

2.4 Antibacterial test

2.4.1 Estimation of minimum inhibitory concentration (MIC) using the optical densitometric technique

The micro dilution method was used for estimating the MIC of the samples to determine the antibacterial activity. The colloidal solutions of SCSNPs and SCSNTs, with a concentration of 1 mg/ml were prepared by ultrasonicating them with distilled water for 30 min. The experimental miniprep consisted of 900 μl of NB media, 100 μl of respective bacterial inoculums and silicon nano-structures in various concentrations (0, 10, 50, 100, 150 and 200 μg/ml). Sterile, distilled water equal to the reaction volume was added. B. subtilis were allowed to grow overnight at 30 °C and the rest of the strains were incubated at 37 °C overnight. Each and every test sample along with the control, was diluted up to 109 dilutions using 0.9 % sodium chloride (saline).

To determine the MIC values 100 μl of all the tested concentrations of inoculums, and their dilutions up to 109 were added to 96 wellplates. Optical absorbance at 600 nm was then recorded in an enzyme-linked immune sorbant assay (ELISA) reader (Thermo Scientific Multiskan EX). NB media was used as a negative control. 100 μl of the experimental mixture without SCSNPs and SCSNTs, was used as a positive control. Optical absorbances for both SCSNPs and SCSNTs at different concentrations without bacterial inoculums, along with their dilutions up to 109 in saline, were also recorded.

2.4.2 Determination of colony forming units (CFU)

From the above experimental miniprep, 100 μl of the sample was used from each test concentration, with dilutions of 109, 108 and 107, to plate the bacterial inoculums, including the control. In B. subtilis the plating was carried out in dilutions of 105, 104 and 103 owing to the low optical density. The CFU were determined by counting the bacterial colonies and then multiplying this with the dilution factor. All assays were carried out in duplicate.

3 Results and discussion

3.1 Characterization of nano-structures

3.1.1 BET measurements

The prime contribution of these nano-structures seems to arise from the large specific surface area that was estimated by the nitrogen absorption method using BET. The measured values were found to be 300 and 100 m2/g for SCSNPs and SCSNTs respectively.

3.1.2 TEM analysis

The morphological information was obtained from the TEM analysis. Figure 1a shows the TEM image of the sample: SCSNP, while Fig. 1b shows the TEM micrographs of the sample: SCSNT. Figure 1a shows the presence of surface-oxidized spherical silicon nano-particles, with sizes varying from 5 to 60 nm. In addition, it also consists of some fiber-like structures in a small proportion. The TEM micrograph of SCSNT (Fig. 1b) reveals that the sample contains nanotubes as well as nanoparticles. The nanotubes appear in bundles wound around each other, along with some nanoparticles of silicon. The nanotubes and nanoparticles are seen to be in the ratio of ~70:30. The nanoparticles are spherical in shape, with sizes varying in the range 5–25 nm, while the diameters of the nanotubes range between 9 and 30 nm. A large number of tubes have a diameter of 14 ± 2 nm, while their lengths are in the order of several hundreds of nanometers. Tubular formation is apparent from the circular open tip of a single nanotube as shown in the inset of Fig. 1b. The inner part of the nanotube appears homogeneously bright. At its tip the hollow opening is clearly visible. The thickness of the annular dark wall seen at the tip happens to be less than 1 nm. The dark lines running along the lengths at the two edges of the nanotubes appear to have a similar thickness. The thickness of the nanotube is equivalent to two or three monolayers of buckled Si (111) planes. Such nanotubes must have been formed by the rolling of planar sheets of silicon in a buckled geometry. The accuracy of the measurement is limited, and so the exact confirmation cannot be made.

Fig. 1
figure 1

Transmission electron micrograph of silicon nano-structures: a TEM of SCSNPs and b TEM of SCSNTs. Inset shows TEM image of the tip of a nano-tube

Full size image

3.1.3 STM analysis

The adsorption geometry of these tubular structures on an atomically flat surface was analyzed by STM. Figure 2a shows the STM image of a single silicon nanotube adsorbed on a freshly cleaved, highly oriented, pyrolytic graphite (HOPG) surface. Almost every nanotube found during the experiment showed a fairly small topographic height compared to its diameter as shown in the line profile (Fig. 2b) recorded along the marked line in Fig. 2a. Such a discrepancy occurs mainly due to the geometrical convolution between the STM tip shape and the nanotubes [26, 27]. However, the dependence of the tunneling gap [28, 29] on local conductivity should also be considered. For example, fullerene C60 appears almost 40 % lower than its actual height in the STM topography, even with a single atom at the apex of the STM tip. Thus, it is not easy in general to obtain the real geometry of the tube adsorbed on the substrate. From the line profile, we estimated [30] the tube diameter, with some errors approximately 11 nm, while the lower limit of the estimated height is 1.4 nm. This indicates that a strong radial compression of the nanotube is induced by van der Waals interaction between the tube and the substrate. A similar kind of radial compression occurs for carbon nanotubes in the case of less number of shells or a single wall [31] contrary to multiwall tubes. So, the tubes studied are possibly few layered or single-walled structures.

Fig. 2
figure 2

a STM image (360 × 360 nm2) of single silicon nano-tube on HOPG; Vbias = 1 V, Itunn = 0.95 nA. b Line profile along the yellow line drawn in a (Color figure online)

Full size image

3.1.4 SEM–EDX analysis

Further studies using SEM–EDX provide evidence for the presence of oxygen in both samples. Figure 3a, b shows the SEM–EDX spectra for samples of SCSNPs and SCSNTs respectively. The spectrum shows peaks at 0.525, 0.129 and 1.739 keV, which correspond to O Kα, Si L2,3 and Si Kα edge respectively. It is observed that the ratio of silicon to oxygen for SCSNPs is 0.88 whereas it is 1.27 for SCSNTs.

Fig. 3
figure 3

SEM–EDX profiles of silicon nano-structures: a SEM–EDX profile of SCSNPs and b SEM–EDX profile of SCSNTs

Full size image

3.1.5 Raman spectroscopic analysis

The presence of crystalline silicon was studied, using Raman spectroscopy. Figure 4a–c shows the Raman spectra for crystalline silicon, SCSNT and SCSNP, respectively. The peak for crystalline silicon is found to be symmetric and appears at 520.5 cm−1, with the full width at half maximum (FWHM) of the peak being 2.8 cm−1. Figure 4b shows an asymmetric peak at 511.5 cm−1 with a FWHM of 15.8 cm−1 for SCSNT. An asymmetric peak is also observed in Fig. 4c at 504 cm−1 with a FWHM of 22 cm−1. These peaks are red-shifted as compared to bulk silicon and have an asymmetric nature, which can be attributed to the phonon confinement effect in the nano-structures. The peak for the nanotube-rich sample, i.e., SCSNT is red-shifted by 9 cm−1, while that for the nanoparticles-rich sample, i.e., SCSNP by 16.5 cm−1. This can be attributed to the two-directional phonon confinement in the case of nanotubes, and to the three-dimensional confinements in nanoparticles.

Fig. 4
figure 4

Raman spectra of silicon (a) Raman spectra of crystalline silicon, (b) Raman spectra of SCSNTs and (c) Raman spectra of SCSNPs

Full size image

The Raman spectroscopic analysis thus confirms that silicon nanotubes as well as silicon nanoparticles are crystalline in nature whereas EDX analysis has shown that their surface has oxide layers. Furthermore, it has been inferred that the walls of these hollow nano-structures, made of silicon and silicon oxide, are extremely thin. Oxygen is present on the surface of the nanotubes, and since it has a higher electro-negativity than silicon it attracts the bonded pair of electron towards itself. Thus, oxygen acquires partial negative charge. When many such oxygen atoms come close to each other, it gives rise to surface charging effects, which are usually observed in silica. In the present case, the oxide layer, being ultra-thin will have a tendency to accumulate a higher charge density on its surface, and this could possibly be the reason for antibacterial activity.

3.2 Antibacterial activity

3.2.1 Optical densitometric analysis

The values of MIC for SCSNPs and SCSNTs were obtained from measurements of optical density at 600 nm for the inoculums containing different bacteria. These are shown in Fig. 5a, b. There are certain limitations associated with using optical density techniques to determine bacterial viability in the presence of nano-materials since the latter themselves contribute to optical density at different concentrations [32]. To resolve this issue, optical densities of the nano-structures, SCSNPs and SCSNTs, are given along with the other data for reference.

Fig. 5
figure 5

Effect of different concentrations of silicon nano-structures on bacterial strains tested. Standard absorbance values of silicon nano-structures at various concentrations from 0 μg/ml (positive control) to 200 μg/ml are provided. Experimental mixture having NB media with respective bacterial inoculums, without nano-structures was used as positive control. NB media alone was used as negative control. a Effect of SCSNPs on bacterial strains tested and b effect of SCSNTs on bacterial strains tested

Full size image

We found that the concentration of nano-particles at which the growth was inhibited was different for different bacteria. In SCSNP treatment the growth of E. coli and B. subtilis was inhibited at 10 μg/ml, while for P. aeruginosa, inhibition was observed at 50 μg/ml. S. aureus did not show any inhibition with the use of SCSNPs. In contrast, SCSNTs showed an MIC of 10 μg/ml for S. aureus. The IC-50 value is found to be 100 μg/ml. Reports from “The Center for Disease Control and Prevention” indicate that the number of annual multidrug-resistant S. aureus (MRSA) infections increased from 127,000 to 278,000 between 1999 and 2005 [32]. In this scenario, SCSNTs would be a potential candidate to target MRSA infections. The inhibition of E. coli and P. aeruginosa by SCSNTs is comparable to SCSNPs. In contrast to SCSNPs, the growth of B. subtilis cultures were not inhibited by SCSNTs. Thus, the data demonstrates the difference in the mechanism of inhibition of bacterial strains by these two nano-structures.

The inhibition generally occurs due to inhibition of reactive oxygen species, membrane disruption, protein inactivation, flocculation or other unknown mechanisms [32]. Further studies need to be carried out to understand the mechanism of inhibition by these nano-structures (SCSNPs and SCSNTs). The MIC was found to be of the order of micrograms for these tested nano-materials, which is comparable to those reported in metal oxides for both Gram-positive as well as Gram-negative bacteria [32]. In many cases, the MIC values obtained for silica nano-structures are better than those reported in other oxide systems [32].

3.2.2 CFU analysis

In order to measure the viable cells, CFU were determined by using serial dilutions of suspensions, followed by spread-plate colony counting. When exposed to different concentrations of SCSNPs, B. subtilis showed a reduced viability at 100 μg/ml (Fig. 6a), whereas an increased viability is observed at 200 μg/ml. This may be attributed to the interaction of SCSNPs with bacteria at high concentrations. The B. subtilis cultures, exposed to SCSNTs, showed a definite pattern of reduced viability (Fig. 6b), with an increase in SCSNTs concentration although initial optical densitometry analysis could not clearly discern inhibition (Fig. 5b). The IC-50 value of SCSNTs was 200 μg/ml in B. subtilis cultures.

Fig. 6
figure 6

CFU counting in Gram-positive bacterial strains calculated for different concentrations of nano-structures (0–200 μg/ml). a CFU of B. subtilis cultures calculated at the dilutions of 103, 104 and 105 for SCSNPs. b CFU of B. subtilis cultures calculated at the dilutions of 103, 104 and 105 for SCSNTs. c CFU of S. aureus cultures calculated at the dilutions of 107, 108 and 109 for SCSNPs. d CFU of S. aureus cultures calculated at the dilutions of 107, 108 and 109 for SCSNTs

Full size image

In S. aureus (Fig. 6c), the IC-50 value was found to be 100 μg/ml for SCSNPs. However, the optical densitometric analysis could not reveal the inhibition (Fig. 5a), possibly due to interference by the optical activity of SCSNPs. The 10 μg/ml of SCSNTs proved to be effective in controlling the S. aureus even at very low concentrations with an IC-50 value of 100 μg/ml (Fig. 6d). Although nano-materials like ZnO [33], Fe3O4 [34] and Ag particles proved to be effective against MRSA infection, SCSNTs proved to be effective even at very low concentrations (10 μg/ml). Thus, the biocompatible nature and cost-effective, eco-friendly synthesis add to the efficacy of thin-walled, SCSNTs as an effective antibacterial agent.

With Gram-negative bacteria like E. coli, the MIC was found to be 10 μg/ml for both SCSNPs and SCSNTs, which concurs with densitometric analysis (Fig. 7a, b). A definite pattern of inhibition as the concentration increases was observed in both SCSNPs and SCSNTs. In P. aeruginosa, 50 μg/ml was found to be effective in reducing the viability in both SCSNPs and SCSNTs (Fig. 7c, d). Thus, SCSNPs and SCSNTs are found to be effective in controlling both the Gram-positive and Gram-negative bacterial strains that were tested.

Fig. 7
figure 7

CFU counting in Gram-negative bacterial strains calculated at the dilutions of 107, 108 and 109 for different concentrations of nano-structures (0–200 μg/ml). a CFUs of E. coli cultures for SCSNPs. b CFU of E. coli cultures for SCSNTs. c CFU of P. aeruginosa cultures for SCSNPs. d CFU of P. aeruginosa cultures for SCSNTs

Full size image

4 Conclusions

We have shown that the growth of thin walled nano-tubes (<1 nm) is enabled by thermal plasma-assisted, gas phase condensation, which is a clean method of synthesis. The possibility of synthesizing metastable states increases in this method due to the high processing temperature and steep temperature gradient. In addition, we have shown that silica-coated silicon nano-structures can be a good substitute as a cheaper and biocompatible antimicrobial agent. The low values of MIC are encouraging, and indicate the specific surface properties of the silicon nano-structures. It is proposed that due to the extremely thin oxide layers of silicon nano-structures, their interaction with the bacteria is strong and capable of inhibition. In this study we emphasize the important role of SCSNTs in controlling MRSA infections, thus making them an efficient antiMRSA agent.