AuNP/Chitosan Nanocomposites Synthesized through Plasma Induced Liquid Chemistry and Their Applications in Photothermal Induced Bacteria Eradication

In this work, a facile direct current atmospheric pressure micro-plasma (APM) technology was deployed for the synthesis of functional gold nanoparticle/chitosan (AuNP/CS) nanocomposites for the first time. Different experimental parameters, such as metal salt precursor concentration and chitosan viscosity, have been investigated to understand their effects on the resulting nanocomposite structures and properties. The nanocomposites were fully characterized using a wide range of material characterization techniques such as UV–vis, transmission electron microscope (TEM), Fourier transform infrared (FTIR) spectra and X-ray photoelectron spectroscopy (XPS) analyses. Potential reaction pathways have been proposed for the nanocomposite synthesis process. Finally, potential of the synthesized nanocomposites towards photothermal conversion and bacteria eradiation applications has been demonstrated. The results show that APM is a facile, rapid and versatile technique for the synthesis of AuNP/CS functional nanocomposites. Through this work, a more in-depth understanding of the multi-phase system (consisting of gas, plasma, liquid and solid) has been established and such understanding could shine a light on the future design and fabrication of new functional nanocomposites deploying the APM technique.


Introduction
Gold nanoparticles (AuNPs), amongst various functional nanomaterials, have attracted great attention in the biomedical field due to their promising physicochemical properties [1]. For instance, when surface modified (e.g., by thiol or amine groups), AuNPs can bind with therapeutic agents to achieve targeted drug delivery [2]. Additionally, AuNPs exhibit localized Surface Plasmon Resonance (LSPR) effect and can convert Near-Infrared (NIR) laser light energy into heat [3]. Such unique features have enabled their applications in biosensing [4], cancer hyperthermia [5] and eradication of drug resistant bacteria [6]. Despite their promising applications, the health risk associated with free AuNPs often presents a concern on the final fate of AuNPs in vivo [7]. To minimize the uncontrolled release of AuNPs, polymer-based nanocomposites incorporating AuNPs have been developed as an alternative solution [8,9]. Chitosan (CS), a biocompatible and biodegradable natural polymer, is commonly used in a wide range of biomedical applications such as prevention of biofilm formation [10], tissue engineering [11] and wound dressings [12]. Combining CS with AuNPs is shown to improve the chemical, physical and biological properties of AuNPs [10], as CS and their derivates can enhance the stability/dispersion  After 30 min of stirring, the HAuCl 4 /CS mixtures were subjected to APM treatment. Scheme 1 shows the APM set-up used in the present work. The anode (platinum rod) was immersed in the HAuCl 4 /CS aqueous mixtures, while the cathode (stainless steel capillary with an inner diameter of 250 µm) was positioned approximately 1 mm above the mixture surface. Helium (He) gas was supplied through the capillary and the flow rate was regulated to 50 sccm using a mass flow controller (CS200-A, Sevenstar, Beijing, China) and a read-out unit (D08-2F, Sevenstar, China). The plasma can be ignited at the gas/liquid interface at a voltage of~4 kV (DC power supply: LAS-10KV-12Ma-P, Boer Co., Jiangsu, China). Afterwards, the APM processing current was adjusted to 5 mA to sustain the plasma and the HAuCl 4 /CS solutions were gently stirred using a magnetic stirrer throughout the treatment (10 min). For comparison, pure CS solutions with different viscosity were also treated under the same APM condition. out the treatment (10 min). For comparison, pure CS solutions with different viscosity were also treated under the same APM condition.
AuNP/CS composite films were prepared by casting 9 mL of each as-synthesized composite liquid sample onto a 5 cm × 9 cm Teflon slide. The samples were then ovendried at 60 °C overnight (until no further weight loss). Control film sample (pure CS and APM treated CS) were also prepared for comparison. Scheme 1. Diagram of the APM set-up used in this work.

Characterization
UV−Vis spectroscopy (UV-2600, Shimadzu, Kyoto, Japan) was used to analyze the optical absorption of the resulting AuNP/CS nanocomposite solution. The sample was diluted 10 times using DI water, drop-casted on to TEM grid and dried completely before transmission electron microscope (TEM) analysis (Tecnai G2 F20/30, FEI, Hillsboro, OR, USA). The TEM images were analyzed by "FUJI" software based on >100 particles for each sample. The viscosity of samples was tested using a sensing viscometer (DV-2 pro, Brookfield, Middleboro, MA, USA). Zetasizer nanoparticle analyzer (Zetasizer Nano ZSE, Malvern, UK) was used for particle size and surface charge when appropriate. All data were averaged over three measurements to ensure repeatability. X-ray photoelectron spectroscopy (XPS) of AuNP/CS samples was conducted using a XPS system (ESCALAB 250Xi, Thermo Fisher Scientific, Waltham, MA, USA). The survey spectra were obtained with a pass energy of 150 eV, while the core level spectra were obtained using a pass energy of 20 eV. All spectra were calibrated with the C1s core peak at the binding energy of 284.8 eV and the results were analyzed using open-source software CasaXPS. Fourier transform infrared (FTIR) spectra analysis was carried out using a FT-IR Spectrometer (Spectrum 3, Perkin-Elmer, Akron, OH, USA). The thermal properties of composite film samples were investigated by differential scanning calorimetry (DSC) analysis in an N2 atmosphere at a heating rate of 10 °C min −1 using Netzsch DSC 404 F3 (Netzsch, Weimar, Germany). The photothermal conversion effect of each composite film sample (5 mm diameter discs) was tested using an NIR CO2 laser (wavelength = 808 nm, power density = 0.8 W cm −2 ). The temperature of the samples was monitored in situ using an IR thermal camera (FOTRIC 340) during laser irradiation.
The swelling ratio of the film samples was tested in phosphate buffer solution (PBS) at 37 °C following gravimetric method [36]: AuNP/CS composite films were prepared by casting 9 mL of each as-synthesized composite liquid sample onto a 5 cm × 9 cm Teflon slide. The samples were then oven-dried at 60 • C overnight (until no further weight loss). Control film sample (pure CS and APM treated CS) were also prepared for comparison.

Characterization
UV−Vis spectroscopy (UV-2600, Shimadzu, Kyoto, Japan) was used to analyze the optical absorption of the resulting AuNP/CS nanocomposite solution. The sample was diluted 10 times using DI water, drop-casted on to TEM grid and dried completely before transmission electron microscope (TEM) analysis (Tecnai G2 F20/30, FEI, Hillsboro, OR, USA). The TEM images were analyzed by "FUJI" software based on >100 particles for each sample. The viscosity of samples was tested using a sensing viscometer (DV-2 pro, Brookfield, Middleboro, MA, USA). Zetasizer nanoparticle analyzer (Zetasizer Nano ZSE, Malvern, UK) was used for particle size and surface charge when appropriate. All data were averaged over three measurements to ensure repeatability. X-ray photoelectron spectroscopy (XPS) of AuNP/CS samples was conducted using a XPS system (ESCALAB 250Xi, Thermo Fisher Scientific, Waltham, MA, USA). The survey spectra were obtained with a pass energy of 150 eV, while the core level spectra were obtained using a pass energy of 20 eV. All spectra were calibrated with the C1s core peak at the binding energy of 284.8 eV and the results were analyzed using open-source software CasaXPS. Fourier transform infrared (FTIR) spectra analysis was carried out using a FT-IR Spectrometer (Spectrum 3, Perkin-Elmer, Akron, OH, USA). The thermal properties of composite film samples were investigated by differential scanning calorimetry (DSC) analysis in an N 2 atmosphere at a heating rate of 10 • C min −1 using Netzsch DSC 404 F3 (Netzsch, Weimar, Germany). The photothermal conversion effect of each composite film sample (5 mm diameter discs) was tested using an NIR CO 2 laser (wavelength = 808 nm, power density = 0.8 W cm −2 ). The temperature of the samples was monitored in situ using an IR thermal camera (FOTRIC 340) during laser irradiation.
The swelling ratio of the film samples was tested in phosphate buffer solution (PBS) at 37 • C following gravimetric method [36]: where the W s and W d represent the weight of the swollen film and the dried film, respectively.

Antibacterial Testing
Gram-negative Escherichia coli (E. coli ATCC 11303) and gram-positive Staphylococcus aureus (S. aureus ATCC 6538) were used to evaluate the antibacterial properties of the AuNP/CS nanocomposites. Composite film discs (5 mm) were placed in a 96-well cell culture plate and incubated with 100 µL (1 × 10 5 CFU/mL) bacterial solution under NIR laser irradiation (0.8 W cm −2 , 808 nm) for 0, 30, and 180 s, respectively. Blank well was used as a control. After NIR laser irradiation, the culture was further incubated at 37 • C for 4 h, the co-cultured bacterial suspensions were diluted with 15 mL of PBS, and 50 µL of the bacterial suspensions was removed and spread onto Mueller-Hinton agar plates and incubated at 37 • C overnight. The bacterial colonies were counted. Tests were repeated five times for repeatability.

Statistical Analysis
Data were expressed as mean ± standard deviation and analyzed using the SPSS 16.0 software (SPSS Inc., Chicago, IL, USA). Significant differences among groups were assessed using the one-way analysis of variance (ANOVA), followed by the least significant difference multiple comparison test. Statistical high significance and very high significance were defined as ** p < 0.01 and *** p < 0.001, respectively.

Physical Properties of AuNP/CS Nanocomposites
During the APM treatment, all sample solutions showed immediate color change (Figure 1a), compared to the pure m-CS solution (Figure 1a). The color of the plasma treated samples ranged from pale pink to mahogany with increasing HAuCl 4 concentration. The optical property of each solution sample was analyzed using UV-vis spectroscopy ( Figure 1b). In the absence of HAuCl 4 , the APM treatment had no significant influence on the absorption spectrum of CS. For AuNP/m-CS nanocomposite solutions, the concentration of HAuCl 4 caused a red shift of the absorption resonant peak from 511 nm (0.1 mM) to 529 nm (2.0 mM). The characteristic peak wavelength is in agreement with the LSPR band of AuNPs [37] and the red shift of LSPR is indicative of the growing AuNP size and the particle anisotropy [37,38]. The intensity of LSPR, on the other hand, is related to the concentration of AuNP in the solution, which gradually increased with the increasing HAuCl 4 concentration. Although the light scattering caused by the nanoparticles could to certain extent affect the overall detected optical signal, this is considered insignificant and is not the main focus of this study, hence it will not be further investigated here. The morphology and size of AuNPs formed within the APM treated AuNP/m-CS were further investigated by TEM. Representative TEM images in Figure 1c illustrate the typical AuNP morphologies resulted from different HAuCl 4 concentration. At lower HAuCl 4 concentrations, the AuNPs (dark particles) are well-dispersed in the CS matrix and are mostly in spherical shapes. The particle size distribution analysis (Figure 1d) suggests that the average AuNP size increases with increasing HAuCl 4 concentration, consistent with the red shift of the LSPR peak found in the UV-vis analysis. The result is similar to previous studies where the size of APM synthesized metal nanoparticles increases with metal salt concentration [34,36]. On the other hand, with increasing HAuCl 4 concentration, AuNPs also start to show signs of anisotropy (e.g., triangular, hexagonal, rods, etc.). This is in line with the literature where the metal salt precursor concentration is reported to have an impact on the AuNP growth kinetics [39]. The UV-vis and TEM results suggest that the HAuCl 4 concentration plays a significant role on the AuNP size, morphology and the size distribution within the AuNP/CS nanocomposites. To investigate the effect of CS viscosity on the morphology and size of the synthesized AuNPs, CS with different viscosities, namely l-CS (50-100 mPa·s), m-CS (200-400 mPa·s) and h-CS (>400 mPa·s) were deployed. All APM-treated samples exhibit consistent color (ruby red) ( Figure 2a). Figure 2b shows that the LSPR for APM-treated HAuCl4/(l-CS, m-CS, h-CS) was independent of the CS viscosity. This result suggests that the AuNP size and concentration were not influenced by the initial CS viscosity. This can be further supported by the TEM images in Figure 2c, where all APM-treated samples contain welldispersed spherical AuNPs within the CS with similar particle size distribution (11.1 ± 3.1, 12.4 ± 3.5, and 12.6 ± 2.8 nm, respectively) ( Figure 2d).
The fact that the AuNP size is independent on the initial CS viscosity is in contrast to the results reported in [33] where AuNP synthesized by in situ thermal method showed a decreasing trend with increasing CS viscosity. This is because the APM treatment has significantly altered the CS viscosity during the treatment ( Figure S1). It is reported that the plasma-induced liquid chemistry can produce abundant energetic/reactive species (such as solvated electrons, radicals, OH•, H•, and O•, etc.) [33,35]. It is widely accepted that OH• can induce the cleavage of β-1-4 glycosidic linkages and the oxygenation of d-glucopyranose rings within the CS chains, leading to effective CS chain scission [40,41], hence reducing the polymer viscosity. All APM treated samples showed similar viscosity regardless of the initial CS concentration. As a result, their impact on the AuNP growth kinetics is similar. To investigate the effect of CS viscosity on the morphology and size of the synthesized AuNPs, CS with different viscosities, namely l-CS (50-100 mPa·s), m-CS (200-400 mPa·s) and h-CS (>400 mPa·s) were deployed. All APM-treated samples exhibit consistent color (ruby red) ( Figure 2a). Figure 2b shows that the LSPR for APM-treated HAuCl 4 /(l-CS, m-CS, h-CS) was independent of the CS viscosity. This result suggests that the AuNP size and concentration were not influenced by the initial CS viscosity. This can be further supported by the TEM images in Figure 2c, where all APM-treated samples contain well-dispersed spherical AuNPs within the CS with similar particle size distribution (11.1 ± 3.1, 12.4 ± 3.5, and 12.6 ± 2.8 nm, respectively) ( Figure 2d). To investigate the effect of CS viscosity on the morphology and size of the synthesized AuNPs, CS with different viscosities, namely l-CS (50-100 mPa·s), m-CS (200-400 mPa·s) and h-CS (>400 mPa·s) were deployed. All APM-treated samples exhibit consistent color (ruby red) (Figure 2a). Figure 2b shows that the LSPR for APM-treated HAuCl4/(l-CS, m-CS, h-CS) was independent of the CS viscosity. This result suggests that the AuNP size and concentration were not influenced by the initial CS viscosity. This can be further supported by the TEM images in Figure 2c, where all APM-treated samples contain welldispersed spherical AuNPs within the CS with similar particle size distribution (11.1 ± 3.1, 12.4 ± 3.5, and 12.6 ± 2.8 nm, respectively) ( Figure 2d).
The fact that the AuNP size is independent on the initial CS viscosity is in contrast to the results reported in [33] where AuNP synthesized by in situ thermal method showed a decreasing trend with increasing CS viscosity. This is because the APM treatment has significantly altered the CS viscosity during the treatment ( Figure S1). It is reported that the plasma-induced liquid chemistry can produce abundant energetic/reactive species (such as solvated electrons, radicals, OH•, H•, and O•, etc.) [33,35]. It is widely accepted that OH• can induce the cleavage of β-1-4 glycosidic linkages and the oxygenation of d-glucopyranose rings within the CS chains, leading to effective CS chain scission [40,41], hence reducing the polymer viscosity. All APM treated samples showed similar viscosity regardless of the initial CS concentration. As a result, their impact on the AuNP growth kinetics is similar. The fact that the AuNP size is independent on the initial CS viscosity is in contrast to the results reported in [33] where AuNP synthesized by in situ thermal method showed a decreasing trend with increasing CS viscosity. This is because the APM treatment has significantly altered the CS viscosity during the treatment ( Figure S1). It is reported that the plasma-induced liquid chemistry can produce abundant energetic/reactive species (such as solvated electrons, radicals, OH•, H•, and O•, etc.) [33,35]. It is widely accepted that OH• can induce the cleavage of β-1-4 glycosidic linkages and the oxygenation of d-glucopyranose rings within the CS chains, leading to effective CS chain scission [40,41], hence reducing the polymer viscosity. All APM treated samples showed similar viscosity regardless of the initial CS concentration. As a result, their impact on the AuNP growth kinetics is similar.
The chemical structural properties of the treated samples were studied by FTIR ( Figure S3). The broadband ranging from 3200 to 3500 cm −1 can be attributed to the -OH and -NH 2 group stretching vibrations within the CS molecular chains. The peaks around 2927, 2848, 1334, and 1250 cm −1 can be correlated to the CH 2 bending within pyranose rings [42]. The peaks at around 1639 cm −1 are due to the vibrations of C=O (amide I) of O-C-NHR, while the peak at around 1540 cm −1 is the NH bending (amide II) (NH 2 ) peak. The band at 1152 and 1064 cm −1 were attributed to the C-O-C bend of saccharide structures. The band at 1408 cm −1 is due to CH 3 wagging [43] and the peak at 1338 cm −1 can be attributed to the C-C stretching of the glucosamine group of CS [44]. It is noticed that all peak intensities of APM-treated AuNP/CS were weakened as compared to pure CS, indicating the coverage of CS signal by AuNPs.
XPS analysis of APM-treated AuNP/CS nanocomposites containing (l, m, h)-CS was carried out to confirm the atomic state of Au element. The survey scans of the three samples all present distinguishing Au peaks (Figure 3a). The core Au 4f of each sample (Figure 3b-d) can be well fitted into two doublet centering at binding energy (BE) of 84.0 and 87.7 eV, respectively. The two element states of Au 4f are consistent with the profiles of reduced metallic Au 0 [45], suggesting Au 3+ ions have been reduced to form Au 0 NPs by APM treatment. The detailed BE and atomic fraction of all fitted elements are listed in Table S1. No absorbance at 313 nm (characteristic LSPR of Au 3+ ) has been detected in APM-treated samples ( Figure S2), indicating the depletion of Au 3+ (i.e., 100% reduction) during the APM process [46]. The results suggest our APM technique is an efficient technique in producing well dispersed AuNPs in comparison to the traditional thermal synthesis approach. As with the latter method, 100% Au 3+ → Au 0 reduction could not be achieved and the assynthesized AuNPs were more aggregated and anisotropic in morphology [32].

AuNP/CS Nanocomposite Formation Mechanisms
The formation of AuNPs within the CS network is influenced by both the plasma induced liquid chemistry and the presence of CS. For HAuCl4 aqueous solution (without CS), two absorption bands are present at 220 and 280 nm ( Figure S2-green curve), which

AuNP/CS Nanocomposite Formation Mechanisms
The formation of AuNPs within the CS network is influenced by both the plasma induced liquid chemistry and the presence of CS. For HAuCl 4 aqueous solution (without CS), two absorption bands are present at 220 and 280 nm ( Figure S2-green curve), which correspond to the p π 5d x 2 −y 2 and p σ 5d x 2 −y 2 ligand−metal transition modes in the [AuCl 4 ] − complex structures, respectively [47]. For HAuCl 4 /CS mixture solution, the corresponding absorption peaks showed decreased intensity and were red shifted, indicating part of [AuCl 4 ] − were anchored onto the CS ( Figure S2-red curve) [48]. This transition can be attributed to the interaction between [AuCl 4 ] − complex and the CS -NH 2 group through ion exchanging reaction as -NH 2 becomes ionized into -NH − and H + in water [49]. In contrast, no ion exchange should in principle take place between -OH and [AuCl 4 ] − because −OH cannot be ionized in water [34] (Scheme 2). This interaction is also evidenced by the UV-vis spectra shown in Figure S2 The formation mechanism of AuNP/CS nanocomposites as a result of APM treatment has been illustrated in Scheme 3. Solvated electrons (RI) and H2O2 (RII) are believed to be the two main species that trigger the initial reduction of [AuCl4] − [50][51][52]. When the sample was subject to APM treatment, the [AuCl4] − anchored on the CS chain and the unbound free [AuCl4] − ions in the bulk solution were reduced into free Au 0 atoms through both RI and RII reaction pathways, see Equations (2) and (3).
In addition, the APM process is also accompanied by UV radiation [53,54]. The UV light generated in the plasma could also partly contribute to the AuNP formation (Scheme 3 RIII) following Equation (4) [55]. It is widely accepted that the OH• and UV produced by the APM-liquid interaction can lead to the scission of CS chains (Scheme 3 RIV) [40,41].
It is worth pointing out that the reducing power of short-lived solvated electrons is usually constrained at the plasma-liquid interfaces, while H2O2 molecules can diffuse into the bulk solution and enable the reduction of the remaining [AuCl4] − via Equation (3) [40]. Although the dominant reduction mechanism is not easy to quantify, we hypothesize the reaction pathway RI is mainly responsible for the initial AuNP nucleation and the subsequent AuNP growth is via aggregation and coalescence near the plasma-liquid interface [36]. This is because the solvated electrons travel much faster than H2O2 [52]. AuNP reduced by H2O2 would be mainly driven by the coalescence and surface growth in the bulk solution [56]. In the presence of CS, the formation of the first Au 0 clusters can take place on the Au-N-C bonding sites on the CS surface (Step 1). Depending on the HAuCl4 concentration/CS ratio, this process is also possible for unbound free [AuCl4] − in the bulk solution (Step 2). The as-formed Au 0 clusters further coalesce with other clusters to form stable Au seed (Step 3). In the event where residual Au ions are still present in the solution after step 3, the ions will be attracted towards the seed AuNPs due to their surface electric double layer. The H2O2 within the bulk liquid will further reduce these ions into Au 0 as part of the continuous surface growth process (Step 4) [36,57,58]. The AuNPs formed in the bulk solution were negatively charged (zeta potential: ~30 mV, see Figure 4) and they could interact electrostatically with the positively charged CS (zeta potential: ~120 mV, see Figure 4) to form AuNP/CS complex (Step 5). Within the solution, the Au seed growth will continue until the HAuCl4 supply is depleted. For instance, in the 2.5 µM AuNP/m- The formation mechanism of AuNP/CS nanocomposites as a result of APM treatment has been illustrated in Scheme 3. Solvated electrons (RI) and H 2 O 2 (RII) are believed to be the two main species that trigger the initial reduction of [AuCl 4 ] − [50][51][52]. When the sample was subject to APM treatment, the [AuCl 4 ] − anchored on the CS chain and the unbound free [AuCl 4 ] − ions in the bulk solution were reduced into free Au 0 atoms through both RI and RII reaction pathways, see Equations (2) and (3).

Thermal and Swelling Properties of AuNP/CS Nanocomposite Film
The thermal properties of typical composite film samples were analyzed by DSC and the results are shown in Figure 5. The DSC thermograms of all samples exhibit a broad In addition, the APM process is also accompanied by UV radiation [53,54]. The UV light generated in the plasma could also partly contribute to the AuNP formation (Scheme 3 RIII) following Equation (4) [55]. It is widely accepted that the OH• and UV produced by the APM-liquid interaction can lead to the scission of CS chains (Scheme 3 RIV) [40,41].
It is worth pointing out that the reducing power of short-lived solvated electrons is usually constrained at the plasma-liquid interfaces, while H 2 O 2 molecules can diffuse into the bulk solution and enable the reduction of the remaining [AuCl 4 ] − via Equation (3) [40]. Although the dominant reduction mechanism is not easy to quantify, we hypothesize the reaction pathway RI is mainly responsible for the initial AuNP nucleation and the subsequent AuNP growth is via aggregation and coalescence near the plasma-liquid interface [36]. This is because the solvated electrons travel much faster than H 2 O 2 [52]. AuNP reduced by H 2 O 2 would be mainly driven by the coalescence and surface growth in the bulk solution [56]. In the presence of CS, the formation of the first Au 0 clusters can take place on the Au-N-C bonding sites on the CS surface (Step 1). Depending on the HAuCl 4 concentration/CS ratio, this process is also possible for unbound free [AuCl 4 ] − in the bulk solution ( Step 2). The as-formed Au 0 clusters further coalesce with other clusters to form stable Au seed (Step 3). In the event where residual Au ions are still present in the solution after step 3, the ions will be attracted towards the seed AuNPs due to their surface electric double layer. The H 2 O 2 within the bulk liquid will further reduce these ions into Au 0 as part of the continuous surface growth process (Step 4) [36,57,58]. The AuNPs formed in the bulk solution were negatively charged (zeta potential:~30 mV, see Figure 4) and they could interact electrostatically with the positively charged CS (zeta potential:~120 mV, see Figure 4) to form AuNP/CS complex (Step 5). Within the solution, the Au seed growth will continue until the HAuCl 4 supply is depleted. For instance, in the 2.

Thermal and Swelling Properties of AuNP/CS Nanocomposite Film
The thermal properties of typical composite film samples were analyzed by DSC and the results are shown in Figure 5. The DSC thermograms of all samples exhibit a broad endothermic peak (first stage) ranging from 150 to 200 • C, which is due to the dehydration process of water constrained in the sample via hydrogen bonds [33,59]. The exothermic stage (second stage) starting from 275 • C corresponds to the decomposition of the CS polymer chains. In Figure 5a, the lower thermal decomposition temperature of AuNP/m-CS nanocomposites (275 • C) comparing to pure m-CS (290 • C) maybe due to the presence of AuNP within CS network. It is reported that AuNP has a catalytic competency over the polymer decomposition [60]. For AuNP/CS synthesized from CS of different viscosities, the endothermic peak showed more intense peak at 171, 179 and 184 • C for l-CS, m-CS and h-CS of CS. This may be due to the greater water content in the higher viscosity CS requires greater thermal absorption. Figure 5b shows that there is no variation in the exothermic peak among the three composite film samples, suggesting initial CS viscosity has no influence over the decomposition behavior of the APM synthesized composites. The water uptake ability of the nanocomposites is key to their applications particu larly in the biomedical field [61]. The swelling behaviors of APM-treated CS and AuNP/CS films were evaluated and their swelling ratios are shown in Figure 6. Pure CS films dis solved completely in PBS after 24 h, whereas APM-treated l-CS, m-CS and h-CS remained stable in PBS ( Figure S4). This is because the APM process has altered the molecular struc tures of CS, leading to a more robust cross-linked network consisting of shorter CS mo lecular chains [5]. This has in turn enhanced the stability of CS and prevented its dissolu tion in PBS. Figure S5 shows that there is no statistical difference in the viscosity of APM treated l-CS, m-CS and h-CS, suggesting the APM treatment has led to CS solutions with similar polymer molecular structures. Compared to APM-treated pure CS, the swelling ratio of AuNP/CS nanocomposites (with HAuCl4 concentration > 0.5 mM) decreases with increasing initial HAuCl4 concentration (Figure 6a). This may be due to the greater AuNP contents within the polymer network allows less room for holding water [62]. The swell ing ratio of AuNP/CS nanocomposites synthesized from different initial CS viscosities wa shown in Figure 6b and no significant difference was found between the three groups confirming the independence of the composite swelling ratio on the initial CS viscosity. The water uptake ability of the nanocomposites is key to their applications particularly in the biomedical field [61]. The swelling behaviors of APM-treated CS and AuNP/CS films were evaluated and their swelling ratios are shown in Figure 6. Pure CS films dissolved completely in PBS after 24 h, whereas APM-treated l-CS, m-CS and h-CS remained stable in PBS ( Figure S4). This is because the APM process has altered the molecular structures of CS, leading to a more robust cross-linked network consisting of shorter CS molecular chains [5]. This has in turn enhanced the stability of CS and prevented its dissolution in PBS. Figure S5 shows that there is no statistical difference in the viscosity of APMtreated l-CS, m-CS and h-CS, suggesting the APM treatment has led to CS solutions with similar polymer molecular structures. Compared to APM-treated pure CS, the swelling ratio of AuNP/CS nanocomposites (with HAuCl 4 concentration > 0.5 mM) decreases with increasing initial HAuCl 4 concentration (Figure 6a). This may be due to the greater AuNP contents within the polymer network allows less room for holding water [62]. The swelling ratio of AuNP/CS nanocomposites synthesized from different initial CS viscosities was shown in Figure 6b and no significant difference was found between the three groups, confirming the independence of the composite swelling ratio on the initial CS viscosity. similar polymer molecular structures. Compared to APM-treated pure CS, the swelling ratio of AuNP/CS nanocomposites (with HAuCl4 concentration > 0.5 mM) decreases with increasing initial HAuCl4 concentration (Figure 6a). This may be due to the greater AuNP contents within the polymer network allows less room for holding water [62]. The swelling ratio of AuNP/CS nanocomposites synthesized from different initial CS viscosities was shown in Figure 6b and no significant difference was found between the three groups, confirming the independence of the composite swelling ratio on the initial CS viscosity.

Photothermal Conversion of Anti-Bacterial Properties of AuNP/CS Nanocomposite Film
The photothermal conversion efficiency of our nanocomposite film was assessed under irradiation of 808 nm CO 2 laser (Figure 7). All samples showed drastic temperature rise in the first 30 s, after which the temperatures curves plateaued. Results show that the sample synthesized from a greater HAuCl 4 concentration exhibits a higher photothermal conversion efficiency (Figure 7a,b). The trend of temperature rise follows 2.0 mM AuNP/m-CS > 1.0 mM AuNP/m-CS> 0.5 mM AuNP/m-CS > 0.1 mM AuNP/m-CS > 2.5 µM AuNP/m-CS. The trend is in line with the increasing AuNP size as indicated by the UV-vis data, suggesting the particle size plays an important role in the photothermal effect. The sample of 2.0 mM AuNP/m-CS (laser power density = 0.8 W cm −2 ) exceeded 45 • C in 30 s, reaching the therapeutic window for photothermal bacteria killing [63] and cancer hyperthermia [11]. In the literature, Jo et al. created an AuNSw core-shell cancer therapy system consisting of a glycol CS core and small AuNP shell [64]. However, the photothermal conversion efficiency of such system is relatively low-it took 7 min for their samples to reach 55.5 • C under 1W/cm 2 NIR laser irradiation. Salem et al. also developed nanogold-loaded CS nanocomposites for drug release and chemo-photothermal cancer therapy applications [65]. Yet their best sample only reached the therapeutic temperature window after~10 min of irradiation using a 532 nm laser (200 mW).
Literature suggests that the photothermal conversion efficiencies of AuNPs can be influenced by the particle size, shape and concentration [66]. Potential formation of assemblies or aggregates during the sample preparation could also lead to red shift of the surface plasmon absorption maximum to the NIR region [67]. On the other hand, past studies report anisotropic AuNPs with multiple sharp spikes have increased number of localized plasmonic modes and could result in stronger light-heat conversion efficacy [68]. As such, the faster temperature rise profiles seen in our 2.0 mM AuNP/m-CS could be the results of growing AuNP size/concentration as well as the increasing particle anisotropy within the sample. Figure 7c,d confirms that the initial CS viscosity does not have significant impact on the resulting composite photothermal conversion efficiency. This is because the AMPinduced liquid chemistry has broken down the CS molecular chains in all samples, leading to CS solutions with the same viscosity. As such, the presence of CS can no longer affect the AuNP particle formation kinetics, and hence the resulting AuNP size/concentration and morphology are independent of the initial CS viscosity.  The strong photothermal effect of our AuNP/CS nanocomposites have been demonstrated for bacterial eradication applications. As shown in Figure 8, the bacteriostatic rate (BR) of as-prepared CS is ~65% against E. coli and ~60% against S. aureus. This is due to the inherit anti-microbial properties of CS [69]. The antimicrobial property of CS is proposed as a result of the entrance of CS molecules into bacterial cell membrane, which leads to their interactions with DNA and thus the damage of bacteria [70]. Upon 30 s of NIR laser irradiation (power density = 0.8 W cm −2 ), the BR of AuNP/CS nanocomposites dramatically increased with increasing initial HAuCl4 concentration, and the BR can reach as high as ~ 99% (Figure 8a,b). The trend of BR (i.e., increasing with initial HAuCL4 concentration) is consistent with that of the temperature rise profile shown by Figure 7a. For samples synthesized from greater HAuCl4 concentration, the resulting AuNP particle size, concentration and morphology favor the light-to-heat conversion. Samples made from higher HAuCl4 concentration will reach the temperature threshold sooner and photothermal damage of the bacteria cell membrane occurs as a result of thermal disintegration [71].
Comparing Figure 8a,b, our samples demonstrated slighter higher antibacterial efficiency towards E. coli as compared to S. aureus This is because S. aureus is surrounded by a thicker cell wall, whereas the E. coli cell wall comprises of a thin layer of negatively charged lipopolysaccharide [24], making it more susceptible to thermal damage. Results show that our sample demonstrate much greater bacteria killing efficiency when compared to the thiol chitosan-wrapped gold nano-shell samples developed by Manivasagan The strong photothermal effect of our AuNP/CS nanocomposites have been demonstrated for bacterial eradication applications. As shown in Figure 8, the bacteriostatic rate (BR) of as-prepared CS is~65% against E. coli and~60% against S. aureus. This is due to the inherit anti-microbial properties of CS [69]. The antimicrobial property of CS is proposed as a result of the entrance of CS molecules into bacterial cell membrane, which leads to their interactions with DNA and thus the damage of bacteria [70]. Upon 30 s of NIR laser irradiation (power density = 0.8 W cm −2 ), the BR of AuNP/CS nanocomposites dramatically increased with increasing initial HAuCl 4 concentration, and the BR can reach as high as 99% (Figure 8a,b). The trend of BR (i.e., increasing with initial HAuCL 4 concentration) is consistent with that of the temperature rise profile shown by Figure 7a. For samples synthesized from greater HAuCl 4 concentration, the resulting AuNP particle size, concentration and morphology favor the light-to-heat conversion. Samples made from higher HAuCl 4 concentration will reach the temperature threshold sooner and photothermal damage of the bacteria cell membrane occurs as a result of thermal disintegration [71].
Comparing Figure 8a,b, our samples demonstrated slighter higher antibacterial efficiency towards E. coli as compared to S. aureus This is because S. aureus is surrounded by a thicker cell wall, whereas the E. coli cell wall comprises of a thin layer of negatively charged lipopolysaccharide [24], making it more susceptible to thermal damage. Results show that our sample demonstrate much greater bacteria killing efficiency when compared to the thiol chitosan-wrapped gold nano-shell samples developed by Manivasagan et al. In their study, the total eradication of E. coli was only achieved after 5 min of 808 nm laser irradiation (at a power density of 0.95 W cm −2 ) [72].
We further investigate the effect of initial CS viscosity on the BR efficiency of our composite film samples. Without NIR laser irradiation, 0.5 mM AuNP/l-CS, 0.5 mM AuNP/m-CS, and 0.5 mM AuNP/h-CS exhibited a similar BR towards E. coli (~65%) and S. aureus (~60%) (Figure 8c,d). With the application of NIR laser, near total eradication against E. coli can be achieved within 30 s, and that for S. aureus can be achieved within 180 s, regardless of the initial CS viscosity. Results confirm that APM-synthesized AuNP/CS nanocomposites can be successfully deployed for killing pathogenic bacteria, and the killing efficiency is directly influenced by the AuNP characteristics (size, concentration and morphology) within the composite sample. This is directed related to the initial HAuCl4 concentration, but is independent of the initial CS viscosity.

Conclusions
In this study, we have successfully demonstrated a facile approach for the synthesis of AuNP loaded CS nanocomposites deploying atmospheric pressure micro-plasma. The complicated interaction within the multiphase phase reaction system and the associated reaction pathways have been elucidated in detail. An in-depth understanding on how synthetic parameters such as HAuCl4 concentration and the initial polymer viscosity can affect the resulting nanocomposite structures and properties has been established for the first time. Our APM-synthesized AuNP/CS nanocomposites demonstrated excellent photothermal conversion efficiencies and anti-microbial properties under NIR laser irradiation. Such properties are highly desirable for potential applications such as antibacterial coatings and cancer hyperthermia. Although our APM approach is currently limited to bench scale synthesis, in future, the technology has the potential to be scaled up for rapid We further investigate the effect of initial CS viscosity on the BR efficiency of our composite film samples. Without NIR laser irradiation, 0.5 mM AuNP/l-CS, 0.5 mM AuNP/m-CS, and 0.5 mM AuNP/h-CS exhibited a similar BR towards E. coli (~65%) and S. aureus (~60%) (Figure 8c,d). With the application of NIR laser, near total eradication against E. coli can be achieved within 30 s, and that for S. aureus can be achieved within 180 s, regardless of the initial CS viscosity. Results confirm that APM-synthesized AuNP/CS nanocomposites can be successfully deployed for killing pathogenic bacteria, and the killing efficiency is directly influenced by the AuNP characteristics (size, concentration and morphology) within the composite sample. This is directed related to the initial HAuCl4 concentration, but is independent of the initial CS viscosity.

Conclusions
In this study, we have successfully demonstrated a facile approach for the synthesis of AuNP loaded CS nanocomposites deploying atmospheric pressure micro-plasma. The complicated interaction within the multiphase phase reaction system and the associated reaction pathways have been elucidated in detail. An in-depth understanding on how synthetic parameters such as HAuCl 4 concentration and the initial polymer viscosity can affect the resulting nanocomposite structures and properties has been established for the first time. Our APM-synthesized AuNP/CS nanocomposites demonstrated excellent photothermal conversion efficiencies and anti-microbial properties under NIR laser irradiation. Such properties are highly desirable for potential applications such as antibacterial coatings and cancer hyperthermia. Although our APM approach is currently limited to bench scale synthesis, in future, the technology has the potential to be scaled up for rapid synthesis of wide range of functional nanocomposites. In future work, our study would also hugely ben-efit from in vitro and in vivo evaluation (such as animal studies), to warrant the potential of our composites in practical biomedical applications.