Bactericidal Silver Nanoparticles by Atmospheric Pressure Solution Plasma Processing

Silver nanoparticles have applications in plasmonics, medicine, catalysis and electronics. We report a simple, cost-effective, facile and reproducible technique to synthesise silver nanoparticles via plasma-induced non-equilibrium liquid chemistry with the absence of a chemical reducing agent. Silver nanoparticles with tuneable sizes from 5.4 to 17.8 nm are synthesised and characterised using Transmission Electron Microscopy (TEM) and other analytic techniques. A mechanism for silver nanoparticle formation is also proposed. The antibacterial activity of the silver nanoparticles was investigated with gram-positive and gram-negative bacteria. The inhibition of both bacteria types was observed. This is a promising alternative method for the instant synthesis of silver nanoparticles, instead of the conventional chemical reduction route, for numerous applications.


Introduction
Silver (Ag) nanoparticles have been widely studied due to their applications in a broad area of industries, such as biomedical engineering, textiles, healthcare, electronics, optics, and chemical engineering [1][2][3][4]. A number of synthesis techniques have been developed for the generation of silver nanoparticles [5]. Chemical reduction is one of the most popular techniques that employs reducing agents, such as ascorbic acid, to reduce the silver from its precursor solution [6]. However, thorough water treatment processes are required to remove the unused reducing agents from the final product since some reducing agents, such as hydrazine and sodium borohydride, are toxic [2]. Nanoparticle agglomeration occurs due to the high active surface area and capping agents, such as Poly (Viny Pyrrolidone) (PVP) or Poly (Vinyl Alcohol) (PVA), which are used to prevent the agglomeration of the formed nanoparticles [4]. A number of other methods are also used for the silver nanoparticle synthesis such as biosynthesis, photochemical and sonochemical pathways [3,4]. These techniques still require either several processing steps or time to obtain pure yield of the synthesised product. Here, we address these issues and demonstrate a versatile synthesis of functional silver nanoparticles by atmospheric pressure solution plasma processing. Figure 1 shows a schematic diagram of the DBD plasma reactor used for this work. The reactor is a cylindrical structure made of quartz. The silver precursor solution was prepared by mixing 1.2 mM of aqueous Silver Nitrate (AgNO 3 ) solution with 34 mM of Sodium Citrate solution in the ratio of 1:3 v/v. The hollow centre part of the plasma chamber could be filled with the precursor solution and could be covered with a quartz lid. The centre part was connected to the gas feeding line and the outlet. The total volume of the chamber is 25 mL and during the experiment 4 mL of the precursor solution was used. The lid and the bottom of the quartz chamber act as the dielectric barriers, with their thickness being 2 mm and the distance between the dielectric barriers being 8 mm. A high voltage was applied to the stainless-steel electrodes by CTP 2000K power supply and the amplitude and frequency were 30 kV and 9.1 kHz, respectively. Voltage and current measurements were made using Rigol DS6104 oscilloscope. Nanomaterials 2020, 10, x FOR PEER REVIEW 3 of 10 and frequency were 30 kV and 9.1 kHz, respectively. Voltage and current measurements were made using Rigol DS6104 oscilloscope. Optical emission spectroscopy (OES) were conducted using an Andor SR-500 spectrometer, Newton CCD camera and fibre optic cable (Oxford Instruments plc, Oxon, UK). The absorption spectra of the resulting solutions were recorded on a UV-Vis spectrometer (Cary 60 UV-Vis Spectrophotometer, Agilent Technologies Inc, Santa Clara, CA, USA). Transmission electron microscopy (TEM, JEOL 2100, JEOL Ltd, Tokyo, Japan) was used to analyse the geometry and size distribution of the nanoparticles. During the sample preparation, an aqueous solution of suspended nanoparticles was added dropwise on carbon-coated copper TEM grids and dried under ambient conditions overnight. A selected area electron diffraction (SAED) ring pattern was used to calculate the crystal spacings using the reported procedure [15]. Lattice fringe spacing was determined by the profile plot of Image J software. Malvern Zetasizer Nano ZS (Malvern Panalytical Ltd, Malvern, UK) used for the Dynamic Light Scattering (DLS) measurements to determine the size distribution. Glass cuvettes were used as sample holders and the samples were filtered through "Hydraflon" hydrophilised PTFE filters with pore size of 0.22 μm before the data acquisition. The measurement parameters are: material refractive index 1.40, Dispersant Refractive index 1.330, and measurement temperature 25 °C.

Experimental Methods
For the antibacterial studies, a modified version of the Mueller-Hinton method was used [16,17]. 6-mm Whatman ® filter papers and Luria-Bertani (LB) broth and agar were supplied by Sigma-Aldrich Pty Ltd, Castle Hill, NSW, Australia. The E. coli DH5α strain was purchased from New England Biolabs Ltd, Ipswich, MA, USA. Staphylococcus aureus ATCC ® 25923™ was purchased from Thermo Fisher Scientific Pty Ltd, Waltham, MA, USA. E. coli and S. aureus were cultured on LB agar plates and single colonies were used to inoculate 10 mL of LB broth and grown at 37 °C until the optical density (OD) at 600 nm reached 0.6. A total of 200 μL of the liquid culture was evenly spread on nutrient agar plates. A total of 10 μL of each silver nano particle sample exposed to different plasma exposure times was added to the centre of a 6-mm filter paper disk prior to placement on the inoculated LB agar plates and incubated overnight at 37 °C to observe zones of bacterial growth inhibition. Optical emission spectroscopy (OES) were conducted using an Andor SR-500 spectrometer, Newton CCD camera and fibre optic cable (Oxford Instruments plc, Oxon, UK). The absorption spectra of the resulting solutions were recorded on a UV-Vis spectrometer (Cary 60 UV-Vis Spectrophotometer, Agilent Technologies Inc., Santa Clara, CA, USA). Transmission electron microscopy (TEM, JEOL 2100, JEOL Ltd., Tokyo, Japan) was used to analyse the geometry and size distribution of the nanoparticles. During the sample preparation, an aqueous solution of suspended nanoparticles was added dropwise on carbon-coated copper TEM grids and dried under ambient conditions overnight. A selected area electron diffraction (SAED) ring pattern was used to calculate the crystal spacings using the reported procedure [15]. Lattice fringe spacing was determined by the profile plot of Image J software. Malvern Zetasizer Nano ZS (Malvern Panalytical Ltd., Malvern, UK) used for the Dynamic Light Scattering (DLS) measurements to determine the size distribution. Glass cuvettes were used as sample holders and the samples were filtered through "Hydraflon" hydrophilised PTFE filters with pore size of 0.22 µm before the data acquisition. The measurement parameters are: material refractive index 1.40, Dispersant Refractive index 1.330, and measurement temperature 25 • C.

Particle Formation and Characterisation
For the antibacterial studies, a modified version of the Mueller-Hinton method was used [16,17]. 6-mm Whatman ® filter papers and Luria-Bertani (LB) broth and agar were supplied by Sigma-Aldrich Pty Ltd., Castle Hill, NSW, Australia. The E. coli DH5α strain was purchased from New England Biolabs Ltd., Ipswich, MA, USA. Staphylococcus aureus ATCC ® 25923™ was purchased from Thermo Fisher Scientific Pty Ltd., Waltham, MA, USA. E. coli and S. aureus were cultured on LB agar plates and single colonies were used to inoculate 10 mL of LB broth and grown at 37 • C until the optical density (OD) at 600 nm reached 0.6. A total of 200 µL of the liquid culture was evenly spread on nutrient agar plates. A total of 10 µL of each silver nano particle sample exposed to different plasma exposure times was added to the centre of a 6-mm filter paper disk prior to placement on the inoculated LB agar plates and incubated overnight at 37 • C to observe zones of bacterial growth inhibition.

Particle Formation and Characterisation
The colour of the precursor solution changed to yellow after a few minutes of the plasma treatment, indicating the formation of the nanoparticles. All the samples present the characteristic surface plasmon resonance peak of silver nanoparticles ( Figure 2). A slight but significant peak was observable for 1 min of plasma exposure, but this later became more prominent. The width of the each plasmon is related to the size distribution of the nanoparticles. The 3-min, 5-min and 7-min samples produced peaks at 369 nm, 394 nm and 396 nm, respectively. The 1-min and 10-min samples have their surface plasmon peaks at 404 nm and 416 nm, respectively. Since equal volumes of samples reacted using the same concentration of precursor, the peak height of the absorbance can be directly proportional to the concentration of the synthesised nanoparticles. The nanoparticle concentration increases until it reaches 5 min and then there is a slight decrease at 7 min. At 10-min plasma exposure, the decrease is more prominent and the plasmon resonance peak is red shifted, which can be attributed to polydisperse distribution of silver nanoparticles [18].
The colour of the precursor solution changed to yellow after a few minutes of the plasma treatment, indicating the formation of the nanoparticles. All the samples present the characteristic surface plasmon resonance peak of silver nanoparticles ( Figure 2). A slight but significant peak was observable for 1 min of plasma exposure, but this later became more prominent. The width of the each plasmon is related to the size distribution of the nanoparticles. The 3-min, 5-min and 7-min samples produced peaks at 369 nm, 394 nm and 396 nm, respectively. The 1-min and 10-min samples have their surface plasmon peaks at 404 nm and 416 nm, respectively. Since equal volumes of samples reacted using the same concentration of precursor, the peak height of the absorbance can be directly proportional to the concentration of the synthesised nanoparticles. The nanoparticle concentration increases until it reaches 5 min and then there is a slight decrease at 7 min. At 10-min plasma exposure, the decrease is more prominent and the plasmon resonance peak is red shifted, which can be attributed to polydisperse distribution of silver nanoparticles [18]. Unlike microscopic techniques, DLS measures the hydrodynamic diameter of the theoretical sphere (rather than the physical size) that diffuses with the same speed as the measured nanoparticle. This is determined by the stabilisers adsorbed on to the nanoparticle, and the solvation shell also moves along with the particle [19]. Therefore, the size measured using the DLS technique is slightly larger than the results obtained by the microscopic techniques. Moreover, DLS measurements are hardly possible on smaller particles in a polydisperse solution [20]. According to the DLS data obtained from the samples (Table 1 and Figure 2 right), there is a decrease in nanoparticle size with an increase in plasma exposure time. The largest average diameter of 20.06 nm is observed at 3 min of plasma treatment and it was then observed that the diameter decreases gradually with an increase in treatment time. The lowest average particle diameter of 9.99 nm was observed at 10 min of plasma exposure. This is an important finding since sub 10-nm silver nanoparticles has been proven to be extremely effective on antibacterial activity. Since DLS measures the hydrodynamic diameter of the nanoparticles, the actual size of the resulting particles can be much smaller than 9.99 nm, which will be far more effective on disinfection processes. Additionally, more than 90% of the volume percentage can be found within the peak area, indicating a narrow distribution of nanoparticles formed during the experiment.  Unlike microscopic techniques, DLS measures the hydrodynamic diameter of the theoretical sphere (rather than the physical size) that diffuses with the same speed as the measured nanoparticle. This is determined by the stabilisers adsorbed on to the nanoparticle, and the solvation shell also moves along with the particle [19]. Therefore, the size measured using the DLS technique is slightly larger than the results obtained by the microscopic techniques. Moreover, DLS measurements are hardly possible on smaller particles in a polydisperse solution [20]. According to the DLS data obtained from the samples (Table 1 and Figure 2 right), there is a decrease in nanoparticle size with an increase in plasma exposure time. The largest average diameter of 20.06 nm is observed at 3 min of plasma treatment and it was then observed that the diameter decreases gradually with an increase in treatment time. The lowest average particle diameter of 9.99 nm was observed at 10 min of plasma exposure. This is an important finding since sub 10-nm silver nanoparticles has been proven to be extremely effective on antibacterial activity. Since DLS measures the hydrodynamic diameter of the nanoparticles, the actual size of the resulting particles can be much smaller than 9.99 nm, which will be far more effective on disinfection processes. Additionally, more than 90% of the volume percentage can be found within the peak area, indicating a narrow distribution of nanoparticles formed during the experiment. The morphology of the silver nanoparticles has been analysed using TEM. The morphology of the silver nanoparticles has been analysed using TEM. Figure 3 (a-d) show the images of the particles synthesised under 3, 5, 7 and 10 min of cold plasma exposure. The shape is near spherical and the average diameters of the nanoparticles were 17.9 ± 8.9, 14.3 ± 1.6, 10.6 ± 2.3, and 5.4 ± 1.4 nm, respectively. The High-resolution transmission electron microscopy (HRTEM) images of the silver nanoparticles synthesised under different plasma exposure times are shown in the Figure 3 (e-g). The fringe spacing is measured to be 0.235, 0.14 and 0.202 nm, which corresponds to the (111), (220) and (200) planes of silver. The d spacings are 0.236, 0.220, 0.141 and 0.121, indicating the facecentred cubic (fcc) crystal structure. The results are consistent with the previous reports in literature [5,21,22]. The sizes obtained on these samples match with the data obtained in the above DLS measurements.

Antibacterial Properties
Equal volumes of nanoparticle suspensions were infused onto the filter paper disks that were added to bacterial plates and the growth inhibition zones were observed after incubation for 24 h. The variation in the size of the inhibition zones was shown to depend on the plasma exposure (treatment) time, as shown in Figure 4. The diameter of the growth inhibition zone was observed to increase when the plasma exposure (treatment) time was increased. S. aureus was seen to be more sensitive to the silver nano particles compared to E. coli. Based on the UV-Vis data, there was a slight difference in the silver nanoparticle concentrations in the 5-min, 7-min and 10-min plasma-treated The fringe spacing is measured to be 0.235, 0.14 and 0.202 nm, which corresponds to the (111), (220) and (200) planes of silver. The d spacings are 0.236, 0.220, 0.141 and 0.121, indicating the face-centred cubic (fcc) crystal structure. The results are consistent with the previous reports in literature [5,21,22]. The sizes obtained on these samples match with the data obtained in the above DLS measurements.

Antibacterial Properties
Equal volumes of nanoparticle suspensions were infused onto the filter paper disks that were added to bacterial plates and the growth inhibition zones were observed after incubation for 24 h. The variation in the size of the inhibition zones was shown to depend on the plasma exposure (treatment) time, as shown in Figure 4. The diameter of the growth inhibition zone was observed to increase when the plasma exposure (treatment) time was increased. S. aureus was seen to be more sensitive to the silver nano particles compared to E. coli. Based on the UV-Vis data, there was a slight difference in the silver nanoparticle concentrations in the 5-min, 7-min and 10-min plasma-treated samples. However, this antibacterial study showed an inverse relationship between the silver nanoparticle size and the antibacterial activity (Figure 4). These observations align with previous studies where decreased particle size increased the antibacterial effect of silver nanoparticles [23][24][25][26].
Nanomaterials 2020, 10, x FOR PEER REVIEW 6 of 10 samples. However, this antibacterial study showed an inverse relationship between the silver nanoparticle size and the antibacterial activity ( Figure 4). These observations align with previous studies where decreased particle size increased the antibacterial effect of silver nanoparticles [23][24][25][26]. Silver nanoparticles are capable of increasing concentrations of reactive oxygen species and decreasing reactive nitrogen species, which cause oxidative stress in bacteria [27,28]. Anti-fungal and anti-viral activities are also reported for silver nanoparticles [29]. The attachment of nanoparticles to the cell wall and cell membranes damages the cell wall and internal structures and causes changes to biochemical pathways. Properties of the nanoparticles, such as size, shape, surface coating and surface charge, also affect the antibacterial activity [30]. S. aureus inhibition by silver nanoparticles has been reported to occur through cell wall damage [31]. For E. coli, a variety of toxic effects have been reported upon exposure to silver nanoparticles, including respiratory inhibition, plasma membrane depolarisation, the leakage of intercellular potassium ions, and metabolic activity inhibition [32].

Mechanism of Nanoparticle Formation
The mechanism of nanoparticle formation is related to the species generated by the plasma and is illustrated in Figure 5 (right). The chemical reactions taking place during the experiment occur in three different regions. These are the plasma region, plasma-liquid interface and liquid region. Species present in the plasma are excited atoms, ions, electrons and UV photons. In this experiment, Ar DBD has been used and, hence, the main energy carriers for the generation of reactive species are excited Ar atoms and few nitrogen species. The origin of nitrogen species may be due to the dissolved nitrogen in the solution or impurities of the gas. Species such as OH, O3, and H2O2 can also be generated due to the presence of water vapour from the aqueous precursor solution. Optical emission spectroscopy (OES) measurements were performed to identify the species in the plasma discharge. This analyses the light emissions from the constituents of the plasma, such as neutral or ionised gas atoms, free radicals and other molecules. Figure 5 (left) shows the emission spectra of the plasma, which is dominated by argon bands. The hydroxyl band (A 2 Σ to X 2 Π, 0 -0, 1 -1, 2 -2), at 309 nm [33], Silver nanoparticles are capable of increasing concentrations of reactive oxygen species and decreasing reactive nitrogen species, which cause oxidative stress in bacteria [27,28]. Anti-fungal and anti-viral activities are also reported for silver nanoparticles [29]. The attachment of nanoparticles to the cell wall and cell membranes damages the cell wall and internal structures and causes changes to biochemical pathways. Properties of the nanoparticles, such as size, shape, surface coating and surface charge, also affect the antibacterial activity [30]. S. aureus inhibition by silver nanoparticles has been reported to occur through cell wall damage [31]. For E. coli, a variety of toxic effects have been reported upon exposure to silver nanoparticles, including respiratory inhibition, plasma membrane depolarisation, the leakage of intercellular potassium ions, and metabolic activity inhibition [32].

Mechanism of Nanoparticle Formation
The mechanism of nanoparticle formation is related to the species generated by the plasma and is illustrated in Figure 5 (right). The chemical reactions taking place during the experiment occur in three different regions. These are the plasma region, plasma-liquid interface and liquid region. Species present in the plasma are excited atoms, ions, electrons and UV photons. In this experiment, Ar DBD has been used and, hence, the main energy carriers for the generation of reactive species are excited Ar atoms and few nitrogen species. The origin of nitrogen species may be due to the dissolved nitrogen in the solution or impurities of the gas. Species such as OH, O 3 , and H 2 O 2 can also be generated due to the presence of water vapour from the aqueous precursor solution. Optical emission spectroscopy (OES) measurements were performed to identify the species in the plasma discharge. This analyses the light emissions from the constituents of the plasma, such as neutral or ionised gas atoms, free radicals and other molecules. Figure 5 (left) shows the emission spectra of the plasma, which is dominated by argon bands. The hydroxyl band (A 2 Σ to X 2 Π, 0 -0, 1 -1, 2 -2), at 309 nm [33], can be identified and would originate due to the dissociation of water. The second positive system emission of N 2 (C 3 Π µ to B 3 Π g ) can be seen in low intensity at 337, 357 and 380 nm. The nitrogen lines may occur due to the dissolved nitrogen in the aqueous precursor solution. Moreover, oxygen emission lines can be found Nanomaterials 2020, 10, 874 7 of 10 in 777 nm and 844 nm. Identified spectral lines for argon are included in Table 2 including transitions, relative intensities, and energy levels. Table 2. Observed optical emission spectroscopy (OES) data of the argon emission spectral lines. Transition, initial energy levels (Ei) and final energy levels (Ef) taken from Ref [10]. The transfer of these species occurs in the plasma-liquid interface and a fraction of these species is transported into the liquid interface. Within the gas-liquid interface, the possible reactions to take part are the recombination, absorption or desorption of reactive species and the solvation of the electrons [34]. Due to the complex mechanisms of these chemical reactions, a number of unexplained questions remain regarding the plasma-induced liquid chemistry. Nanomaterials 2020, 10, x FOR PEER REVIEW 7 of 10 can be identified and would originate due to the dissociation of water. The second positive system emission of N2 (C 3 Πμ to B 3 Πg) can be seen in low intensity at 337, 357 and 380 nm. The nitrogen lines may occur due to the dissolved nitrogen in the aqueous precursor solution. Moreover, oxygen emission lines can be found in 777 nm and 844 nm. Identified spectral lines for argon are included in Table 2 including transitions, relative intensities, and energy levels. The transfer of these species occurs in the plasma-liquid interface and a fraction of these species is transported into the liquid interface. Within the gas-liquid interface, the possible reactions to take part are the recombination, absorption or desorption of reactive species and the solvation of the electrons [34]. Due to the complex mechanisms of these chemical reactions, a number of unexplained questions remain regarding the plasma-induced liquid chemistry. The solvated electrons absorbed into the liquid penetrate a few nanometers and react with solvent molecules to induce a number of chemical reactions [12] inclusive of the reduction of silver ions Equation (1). The redox potential of Ag + /Ag is (+0.799V), which makes the instant capture of the The solvated electrons absorbed into the liquid penetrate a few nanometers and react with solvent molecules to induce a number of chemical reactions [12] inclusive of the reduction of silver ions Equation (1). The redox potential of Ag + /Ag is (+0.799 V), which makes the instant capture of the solvated electrons and synthesis of the reduced silver atoms. The atmospheric pressure plasma may have electrons with an energy level up to 20.2 eV and it is possible for the positive ions to also have energy levels above 10 eV [3]. The following reaction is possible within the few nanometres of the liquid surface. e aq + Ag + → Ag (1) Furthermore, solvated electrons react with water and H + to form more H, which reacts with silver ions to reduce them to Ag Equations (2) and (3). These reactions form H, which also contributes to the further reduction of silver ions.

Wavelength
The synthesis of silver nanoparticles in the liquid occurs by the nucleation of the reduced metal atoms due to the liquid chemistry initiated by the DBD plasma. According to the OES spectra, excited Ar, OH and N 2 species were identified. The reduction could occur due to a number of species, such as plasma electrons, hydrogen radicals, solvated electrons and hydrogen peroxide [14]. Thai et al. reported that fragments of water continuously ionised to produce OH + , O + , and H + because the ionisation energies are located below the energy threshold. Additionally, the reaction between H and silver cations also forms silver metal atoms by the following reaction in Equations (2) and (3), which also facilitates the formation of Ag seeds.
Nanoparticle growth occurs eventually by the combination of metallic atom clusters which are surrounded by the citrate capping agents. According to Equations (1)-(4), the amount of reducing agents required to form silver ions depends on the exposure time, and the reaction kinetics can be controlled by varying it. According to the TEM images, the size and the size distribution of the synthesised nanoparticles depend on the plasma exposure. This phenomenon can be explained using the La Mer's model [2,35]. On low plasma exposure duration, the reducing species in the solution are scarce and few nuclei are formed for the nucleation step. As such, only particle growth can be observed with a broader size range. When the exposure time increases, the enhanced reduction rate generates more nuclei, which leads to the formation of silver nanoparticles of a smaller size. The amount of the reactants available for the particle growth step decreases due to the large number of nuclei, and a relatively smaller number of nanoparticles can be observed in a narrow size distribution.

Conclusions
Silver nanoparticles were successfully synthesised by DBD solution plasma processing. Our results suggest that the plasma exposure time is an important factor for the resulting silver nanoparticle concentration and morphology. The morphology was characterised by Transmission Electron Microscopy (TEM) and the optical properties were determined by ultraviolet and visible spectroscopy. The DLS technique was used to determine the hydrodynamic diameter of the particles. OES measurements were made to identify the species present in the cold plasma. The antibacterial properties of the synthesised silver nanoparticles were studied using S. aureus as a model of gram-positive bacteria and E. coli as a gram-negative bacterium. The results indicate that the synthesised nanoparticles demonstrated increased anti-microbial activity with decreasing nanoparticle size. This work demonstrates the possibility of producing silver nanoparticles without the need for toxic reducing agents. The resulting nanomaterials can be successfully employed in electronics, plasmonic and biological applications. The same technique can be used for the synthesis of other various metal nanoparticles, removing the need for chemical reducing agents, leading to much more economically and environmentally favozurable processes.