One-Pot, Surfactant-Free Synthesis of Gold Nanostars and Evaluation of Their Antibacterial Effects against Propionibacterium acnes

Anisotropic gold nanoparticles, especially gold nanostars, are used in many ﬁ elds of biomedical applications such as sensing, targeted drug delivery, and diagnostic and photothermal therapy. In this study, we introduced a novel application of gold nanostars as an antimicrobial agent. While spherical gold nanoparticles have an inappreciable e ﬀ ect, gold nanostars exhibit signi ﬁ cant antibacterial activity. Besides, the seed-mediated method, a conventional technique for preparing gold nanostars, is rather complex and toxic to human and environment due to unsafe synthesized materials such as surfactants and reducers. In recent years, green chemistry for nanoparticle synthesis is attractive because of its advantages. Instead of the seed-mediated procedure, we present a facile and green procedure to synthesize gold nanostars using ascorbic acid as a reductant and chitosan as a directing-growth agent. The in ﬂ uences of reacting parameters were evaluated to determine the optimal conditions. Star-shaped gold nanoparticles were successfully synthesized with average size tunning from 137 : 0 ± 20 : 7 nm to 281 : 9 ± 25 : 8 nm of the core and 14 : 0 ± 4 : 4 nm to 54 : 2 ± 11 : 9nm of branches. Antibacterial activity against Propionibacterium acnes of gold nanostars was also investigated. Propionibacterium acnes is one of the main reasons causing acne vulgaris. The antibacterial test was evaluated by the plate count and well di ﬀ usion method. The results showed a signi ﬁ cant e ﬀ ect that gold nanostars could be the prospective agent for replacing antibiotics in acne treatment.


Introduction
Acne vulgaris is one of the most popular chronic inflammatory dermatological problems affecting around 10% of the global population [1]. Moreover, it is also among top of three cutaneous diseases treated by dermatological doctors [2]. Therein, adolescence in the age 12-25 years assumes 85% of cases affected by acne vulgaris, according to Global Burden of Disease (GBD) [3]. There are 85% of females reporting having acne while just 15% of males were affected [4]. The treatment for chronic inflammatory acne vulgaris requires a long period and patience. Lesions usually appear on skin including the back, shoulders, chest, and especially face, causing to take scars or bruises. This not only evokes the physical lesions but also impacts the quality of life [5]. The patients can have problems with depression, anxiety, social isolation, and low self-confidence [6]. There are many factors related to acne, but Propionibacterium acnes (P. acnes) is believed to have an important role in the information or progression of acne vulgaris [7]. Propionibacterium acnes is a rod, anaerobic, Gram-positive bacterium [8]. Besides the association with the inflammatory skin, it is believed to play a role in other human infections and clinical conditions [9]. Treatment options for acne such as oral or topical use include benzoyl peroxide, salicylic acid, retinol, or antibiotics [10]. However, each option has disadvantages and side effects as follows: peeling, itching, erythema, and allergic reaction in the case of benzoyl peroxide [11]; retinoid irritation, dryness, erythema, purging, and stinging/burning in the case of using retinol [12]; and difficulty in breathing, dryness and peeling of skin, itching, welling of the eyes, face, lips, or tongue, and tightness in the throat in the case of using salicylic acid [13]. Antibiotics are common therapy for acne treatment; however, P. acnes become resistant to most antibiotics because of long-term use which led to failure in acne vulgaris treatment [14,15].
Nanotechnology and nanomaterials are prospective methods for replacing conventional acne treatment [16,17]. Antimicrobial metallic nanoparticles such as noble-(copper, silver, and gold) [18][19][20], oxide-(zinc oxide, magnesium oxide, and tin oxide) [21][22][23], and carbon-based nanoparticles (carbon nanotube, graphene) [23,24] were attractive and studied by researchers in several years. Among them, silver nanoparticles are one of the most common subjects due to great antibacterial activity [25]. However, there is dispute on the cytotoxicity of silver nanoparticles [26,27]. By contrast, gold nanoparticles are high biocompatible and low toxic materials [28,29]. However, spherical gold nanoparticles have low or not significant antibacterial activity [30] ( Table 1). Recently, some types of anisotropic gold nanoparticles, especially gold nanostars (GNS), were attractive because of their physical, chemical, and biological properties. The brief method for the preparation of gold nanostars is the seed-mediated method in the presence of a strong/mild reducing agent and anionic or cationic surfactants [31]. Sodium borohydride (NaBH 4 ) is used as a reducing agent while cetyltrimethylammonium bromide (CTAB), sodium dodecyl sulfate (SDS), and Triton X-100 are surfactants. Nevertheless, these materials are harmful to human and environment [31][32][33].
In recent years, green chemistry becomes a common trend in many fields of chemistry and materials [40] including nanomaterial synthesis [41]. This technology uses safe agents derived from nature for replacing toxic materials as well as applying low-cost and energy-efficient equipment [42]. Surfactant-free preparation of gold nanostars was first introduced by Yuan et al. [43]. In this report, we present a one-pot, surfactant-free synthesis of gold nanostars using chitosan. The role of chitosan (CS) is to direct growth, derived from crabs, shrimps, and other crustaceans. CS has many advantages of biological properties such as biocompatibility and antimicrobial and antifungal activity [44][45][46]. There are many biopolymers used to prepare metallic nanoparticles [47][48][49][50]; however, just spherical nanoparticles were formed due to these biocompatible polymers as the matrix in the nanocomposite. The study of Phan et al. [51] was the first to use CS as a directing-growth agent. However, this procedure synthesized gold nanourchins with a large core and short tip. Besides, ascorbic acid (AA), also known as vitamin C, occurs in many natural foods and has wide applications in foods, cosmetics, and pharmaceuticals [52,53], playing a role as a reducing agent. The evaluation of various conditions including pH, concentration of CS, and AA was performed to determine their influences on the morphology of GNS. Furthermore, the antibacterial activity of GNS against P. acnes was investigated by the well diffusion and plate count method. The antibacterial mechanism was observed using a scanning electron microscope (SEM). Furthermore, the mechanism that GNS killed bacteria was identified by SEM observation using simple preparation without any fixative and dehydrating step.

One-Pot Preparation of Gold Nanostars.
Firstly, CS 2% solution in acetic acid 1% was made from low molecular weight CS powder and acetic acid. Next, 1 mL of HAuCl 4 2 × 10 −2 M solution was added to 9 mL CS 2% solution under stirring. Acetic acid was used to adjust pH. Finally, ascorbic acid (AA) 0.1 M was dropped immediately into the mixture. The solution above was kept at least 30 min at room temperature. The color of the solution changed from colorless to cobalt blue indicating star-shaped particles formed. The influences of conditions including pH, concentration of CS, and AA on the morphology of GNS were studied.

Characterization of Gold Nanostars.
A Jasco V-670 spectrophotometer was used to characterize the surface plasmon resonance (SPR) of GNS in the wavelength range of 400-900 nm and scanning rate of 200 nm/min. Interaction between GNS and CS was determined by Bruker Tensor 27 Fourier Transfer Infrared Spectroscopy (FTIR). All FTIR results were obtained from powder samples and did not apply smoothed or correction baseline. The crystal structure of GNS was determined employing X-ray diffraction (XRD). The scanning was carried out in the 2 theta range of 20°-100°u sing X-ray diffractometer Bruker D5005. Transmission electron microscope (TEM) analysis was examined by using JEM1010-JEOL. The average sizes of the core and branches were calculated by using ImageJ software (NHI Image) based on thirty particles of each sample. The GNS solutions were sonicated before measuring and examining.
2.4. Evaluation of Antibacterial Effects against P. acnes 2.4.1. Well Diffusion Method. The antibacterial activity of CScapped GNS was investigated using the agar well diffusion assay according to Qais et al. [54]. The organisms were grown in thioglycollate broth for 48 hours to obtain the colonyforming unit around 10 8 CFU/mL. Next, the bacterial inocu-lum was uniformly spread using a sterile cotton swab on sterile TSA plates. Wells of 6 mm diameter were punched by using sterilized micropipette tips. Different concentrations of CS-capped GNS (50, 100, 150, and 200 μg/mL) were poured into each well. All plates were incubated for 48 h at 37°C, and diameters of the zone of inhibition were measured.

Plate Count Method.
Viable bacterial cell concentrations were estimated by counting colony-forming units (CFU) with and without exposure to the CS-capped GNS on an agar plate [55]. The P. acnes bacteria were exposed to CS-capped GNS with different concentrations (25,50,75, and 100 μg/mL). Ampicillin (Amp) 100 μg/mL was a positive control while DI water was a negative control. After 1 minute, move 100 μL of the solution above into a plate and pour Tryptic Soy Agar (TSA) to a plate. All plates were then incubated at 37°C for 72 hours, and CFU were counted.   For the first time, a facile method was developed using a scanning electron microscope to observe the mechanism of GNS in destroying P. acnes. The procedure for the preparation of observing specimen is presented in Figure 1. To evaluate how GNS damaged the bacterium, we observed the uptake of gold nanoparticles into the bacterial membrane. After exposing gold colloidal solutions to P. acnes cultures, this mixture was directly dropped on stubs and evaporated in vacuum at 40°C without any fixative and dehydrating process. The SEM observation was carried out on Hitachi S-4800 equipment at accelerating voltage 3 kV.

Results and Discussion
3.1. The Influences of Reacting Conditions on the Morphology of GNS 3.1.1. Influence of pH. Surface resonance plasmon (SPR) of GNS consists of two absorption bands ranging from 500 to 900 nm, including a weaker peak at 500-575 nm referring to the core and an intense band in the NIR region (SPR max ) ranging from 600 to 1200 nm referring to the branches of GNS [56]. Figure 2 shows the UV-Vis spectra of GNS prepared in various pH conditions. It was clear that the intensity of SPR max increased according to the decreasing pH from 3.73 while absorption locations were almost stable at 871 nm. However, both intensity and SPR max declined when keeping on adjusting pH to 2.10. TEM images of GNS prepared in various pH are shown in Figure 3. At pH 2.31, the average size of the core and branches obtained 137:0 ± 20:7 nm and 33:5 ± 9:7 nm, respectively. The average size of the core expanded to 179:2 ± 43:9 nm while branches decreased down to 14:0 ± 4:4 nm. It can be explained based on the effect of pH on oxidation/reduction of AA. Electrons produced by the oxidation process of AA reduced Au 3+ to Au 0 . The oxidation rate of AA depended on pH, faster toward basic pH but slower toward acidic pH [57]. Many electrons promoted due to the fast oxidation rate of AA at pH 3.73 led to uncontrolled reducing reaction of Au 3+ ions to Au 0 . By contrast, less electrons produced at pH 2.10 were not enough to reduce all Au 3+ ions to Au 0 . Moreover, pH affected the interaction of CS and gold nanoparticles. CS was a directing-growth agent to assist star-shaped particle reaction. Value pH was driven to acidic condition which led to better dispersion of CS and interaction of N-acetyl groups NH 3 + onto the surface of gold nanoparticles. But too many N-acetyl groups NH 3 + attached onto nanoparticles' surface which led to steric hindrance, resulting in the decrease in the star-shaped yield.

Influence of CS Concentration.
The absorption spectra of GNS synthesized in a variety of CS concentrations are shown in Figure 4. By increasing CS concentration from 0.0.125% to 0.050%, the SPR max shifted toward the NIR region, from 766 nm to 866 nm. However, the SPR max almost were steady in spite of increasing CS concentration to 0.2%. Moreover, the intensity of the SPR max peak declined corresponding to the CS increase in concentration. Figure 5 shows the morphology and size of synthesized gold nanostars in  Journal of Nanomaterials different mass concentrations. CS mass concentration ranges from 0.0125% to 0.025%, and there was a slight decline in the core from 281:9 ± 25:8 nm to 216:5 ± 33:2 nm whereas branches were elongated from 20:4 ± 5:4 nm to 27:8 ± 6:0 nm. The branches continuously prolonged to 54:2 ± 11:9 nm in spite of increasing CS mass concentration to 0.05%. As discussed above, the role of CS was to direct growth to prepare star-shaped gold nanoparticles.

Influence of AA Concentration.
The effect of AA on plasmon surface resonance and morphology of GNS was exhibited by absorption spectra ( Figure 6) and TEM analysis (Figure 7). Although the SPR max was steady at 875 nm when increasing AA concentration from 12:5 × 10 −4 M to 25 × 10 −4 M, there was a significant increase in intensity of SPR max . It was noticeable that SPR max shifted down to 741 nm and 658 nm despite increasing AA concentration to 50 × 10 −4 M and 20 × 10 −3 M. The prepared GNS at AA 25 × 10 −4 M obtained 157:8 ± 13:6 nm of the average core and 42:0 ± 9:2 nm of the branched length. Meanwhile, the average length of branches was shorter and the average core was expanded, which resulted in 14:7 ± 4:1 nm and 195:3 ± 28:9 nm, respectively. The reduction of Au 3+ ions to Au 0 due to electrons promoted the oxidation process of AA. The higher the AA concentration was, the more electrons were produced. At optimal AA concentration, the reduction to form GNS reacted gradually, which resulted in the high yield of star-shaped gold nanoparticles. By contrast, many electrons were promoted because of too high AA concentration which led to uncontrolled reduction. Figure 8 shows the FTIR spectra of pure CS and CS-capped GNS. In the spectrum of pure CS, it was clear that a band located at around 2886 cm -1 corresponds to C-H group stretching vibration [58]. Besides, there were two peaks at 1647 cm -1 and 1333 cm -1 relating to C=O stretching vibration (amide I) and C-N stretching (amide III), respectively [59]. Another band appeared at 1591 cm -1 corresponding to the bending vibration of N-H groups (amide II). These bands above confirmed the N-acetyl groups of pure CS powder [60]. Compared with pure CS, the shift of bands was observed in the FTIR spectrum of CS-capped GNS. The C-H stretching shifted down to 2883 cm -1 while the bending of N-H bonds disappeared. The C=O groups of amide I and C-N groups of amide III stretching vibration were located at 1642 cm -1 and 1328 cm -1 , respectively. These shifts indicated the interaction between functional groups of CS and GNS [61].

Evaluation of Antibacterial Effects against P. acnes
3.4.1. Well Diffusion Method. Star-shaped gold nanoparticles synthesized in optimal conditions (pH 2.31, CS 0.05% and AA 25 × 10 −4 M) were used in antibacterial tests. The antibacterial activity against P. acnes of GNS using the well diffusion method is shown in Table 2. The results ( Figure 10)  Journal of Nanomaterials showed that GNS began to inhibit P. acnes at 100 μg/mL. The zone of inhibitions was wider in spite of increasing GNS concentration. Therefore, the lowest concentration exhibiting antibacterial activity was chosen as an end point in the plate count method. Figure 11 shows the antibacterial effect of GNS against P. acnes in various concentrations ranging from 25 μg/mL to 100 μg/mL with Di water as a negative control and amp 100 μg/mL as a positive control. The antibacterial efficiency was calculated as follows:     Journal of Nanomaterials where N control was the colonies counted in the negative control plate and N sample was the colonies counted in the sample plate exposed to GNS. It was noticeable that the antibacterial efficiency of amp 100 μg/mL was over 10%. At 25 μg/mL, there was no antibacterial activity against P. acnes of GNS. However, the antibacterial efficiency increased to around 45% at GNS 50 μg/mL concentration. The efficiency obtained more than 90% despite increasing GNS concentration to 75 μg/mL, and 95% of P. acnes bacteria exposed to 100 μg/mL GNS were killed. It was noticeable that the antibacterial efficacy of amp 100 μg/mL was just over 10%. According to the result, GNS exhibited significant antibacterial activity against P. acnes compared to ampicillin at the same concentration.

Plate Count Method.
3.4.3. Scanning Electron Microscope (SEM) Analysis. Figure 12 shows SEM micrographs of P. acnes before and after exposure to GNS. It was clear that bacterial cells were damaged by GNS. The normal cell membranes of P. acnes bacteria before exposure to GNS were still stable and clear. However, after exposure to star-shaped gold particles, the cell membranes were deformed and appeared to have an amount of gold nanoparticles (AuNPs) inside bacterial cells. According to many studies before, there are three main mechanisms to kill bacteria of metallic nanoparticles including reactive oxidative species (ROS), releasing ion, and interaction of nanoparticles with the cell membrane [64]. Because gold nanoparticles (AuNPs) have no ROS and ion release [65], the interaction of AuNPs with the cell membrane is the only way to kill bacteria. The antibacterial activity of AuNPs depends on facets which are determined by the shape of nanoparticles. The spherical nanoparticles mainly have {100} facets while the essential facets of starshaped nanoparticles are {111}. The facets {111} have high antibacterial reactivity than {100} [66,67]. Therefore, spherical gold nanoparticles almost do not have an antibacterial effect. In contrast to this, GNS have good activity because of their primary facets {111} in the crystalline structure. This can be demonstrated in XRD patterns of GNS in Figure 9. The mechanism to damage bacterium could be explained according to many earlier reported studies [68][69][70]; gold nanoparticles can interact with bacterial membranes and destroy them, leading to deformation or breaking of the cell membranes, which resulted in the leakage of bacterial components outside [71]. After that, they penetrate into the cytoplasm where they can inhibit the protein process. All of the mechanisms lead to bacterial death.

Conclusions
In this study, we successfully synthesized GNS using a rapid and green method using safe materials such as ascorbic acid and chitosan. The prepared GNS were characterized using spectroscopic methods combining TEM and XRD to determine the optimal conditions. The average size of starbranched particles was ranging from 14:0 ± 4:4 nm to 54:2 ± 17:9 nm depending on conditions. Besides, GNS revealed the significant antibacterial effects against Propionibacterium acnes determined by the well diffusion and plate count method. Gold nanostars are a prospective agent for replacing antibiotics to solve antibacterial resistance in acne treatment.

Data Availability
The data used to support the findings of this study have been deposited in the Data Availability Statement_No 6650661 repository. All data include figures, raw spectroscopy data, and .txt files (https://doi.org/10.6084/m9.figshare.13222175.v2/).

Conflicts of Interest
The authors declare that they have no conflicts of interest.