Self-assembled metal cluster-carbon quantum dot heterostructures with photothermal antibacterial properties

Abstract

Noble metal (Au, Ag, Cu) cluster is an emerging category of promising interest functions in form of designed constructions. Among various candidates, carbon dots could be treated as one interesting component for synthesis functional candidates while heterogeneous contents are orderly integrated together. In this work, we successfully fabricate heterostructural nanoparticles (HNPs) based on noble metal clusters (Au, Ag and Cu) integrated with carbon quantum dots (CQDs) through self-assembling approach. These HNPs demonstrate remarkable photothermal efficiency, high stability, low hemolysis ratio, excellent biocompatibility and significant bactericidal effects, making them promising candidates for photothermal applications. Notably, the Au–C achieved remarkable photothermal conversion efficiency (PTE) of 54.16% and antibacterial rate over 99%, which also significantly accelerated the healing process in methicillin-resistant Staphylococcus aureus (MRSA)-infected subcutaneous abscess model mice. Our findings highlight the potential of these self-assembled heterostructures, especially Au–C, as effective and promising photothermal agents with antibacterial functionality.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s11671-025-04243-5.

Introduction

Orderly stacking of functional nanomaterials has played a significant role in expanding the real applications and properties. Through various self-assembly mechanisms, liquid phase in the solution system enables the creation of different dimensions of functional constructions, including complex semiconductors, electrical materials, and biological materials [1–7]. To enhance the stability of these assembled constructions, the peptides or amino molecules are often utilized due to their different functional groups. As assembled structures become more widely used, they developed potential utilizations toward metal clusters in forms of fundamentals and opportunities [8–13]. Owing to the demands for exploring functional fabrications, carbon quantum dots (CQDs) have emerged various of agent materials in form of fluorescent biosensors and photoelectrochemical agents [14–18]. This has prompted the exploration of noble metal clusters (Au, Ag, Cu) assembled with CQDs as heterostructure nanoparticles (HNPs) with exceptional physical properties, which can be effectively controlled through assembled approaches and demonstrate therapeutic potential [19, 20].

Photothermal Therapy (PTT) is a non-invasive therapeutic technique that uses light, typically in the near-infrared (NIR) region, to generate heat in targeted tissues, which then induces localized damage to pathogenic bacteria by denaturing proteins and disrupting cell membranes. Traditional clinical antibacterial treatments primarily rely on the use of antibiotics, associated with side effects such as bacterial resistance. As a physical antibacterial approach, PTT has minimal side effects. Existing PTT materials mainly include a solitary PTT active monomer, such as porphyrins, heptamethine, BODIPY-family dyes and black phosphorus, and nanoparticle construct including multi photoactive materials [21]. However, there are still certain limitations in terms of stability, photothermal efficiency, biocompatibility, and synthesis complexity. Carbon-based materials like CQDs have been reported for PTT exactly [22–24]. HNPs, formed by assembling atomic-level noble metal clusters with CQDs through self-assembly [25–27], creating a “metal” and “semiconductor” complex that enhances the functionality of the two original components. Metal clusters, whose properties are determined by internal metal atoms and the surface molecules, can construct or integrate other material unit for fabricating heterostructures [28–32]. In order to avoid strong steric hindrance and anisotropy, these atomic-level noble metal clusters can assemble with CQDs via surface functional group bonding functions [33–35]. Based on original characterizations, noble metal clusters and CQDs exhibit weak photothermal properties, unless photothermally sensitive modifications are induced by cyanine, porphyrin or polydopamine [36–40]. Nevertheless, their integration has the potential to create a stable, biocompatible structure with enhanced photothermal activity.

In this study, we fabricate HNPs from atomic-level noble metal clusters and CQDs for PTT. During this self-assembly procedure, these materials demonstrate higher stability in alkalinity mediums than in acidic atmospheres. The cross-linkage was achieved through bonding interaction via amidation. The assumed structure (Au–C, Ag–C, Cu–C) information is simulated as one heterogeneous porous construction, as illustrated in Fig. 1. The enhanced PTT activity is attributed to the interactions between the metal and semiconductor components, driven by the “hot-electron” effect [41]. These synthesized HNPs exhibit good biocompatibility and bactericidal properties, providing a novel approach for the development of effective and low-side-effect clinical antibacterial treatments.

Fig. 1.

Schematic diagram of the formation and synthesis processes of metal clusters (Au, Ag, Cu clusters) self-assembled with CQDs into HNPs. Representative TEM images of assembled HNPs. a Au–C, b Ag–C and c Cu–C HNPs. The relative image is STEM image of each HNP, and the EDS-Mapping showed the C, Au, Ag and Cu elements, respectively. Insert scale bars in a, b, c are 50 nm, 50 nm, and 500 nm, respectively

Materials and methods

Chemical and reagents

Chloroauric acid (HAuCl₄, 99%), urea (95%), sodium citrate (95%), sodium hydroxide (NaOH, 95%), isopropanol (95%), CuCl₂ (99%), AgAc (99%), L-cysteine (Cys, 98%), L-glutathione (GSH, 99%), sodium borohydride (NaBH₄, 95%), and the liquid substrate system for ELISA were purchased from Sigma-Aldrich (Milwaukee, WI, USA). All chemicals were used without further purification.

Preparation of nitrogen-doped carbon quantum dots (CQDs) and photoluminescent metal clusters such as gold, silver and copper clusters

Urea (1.01 g, 20 mmol) and sodium citrate (0.81 g, 2.7 mmol) were mixed well with an agate mortar for approximately 40 min to obtain a uniform powder. Then, the mixture was transferred into a polytetrafluoroethylene flask, and the temperature was kept at 180 °C for 2 h. Finally, the resultant material was collected and cooled to room temperature (RT) for additional purification. During the purification process, a small amount of ultrapure water was added to the reaction flask, and the product was dissolved and transferred to a round-bottom flask. Ethanol was added, the mixture was washed 3 times by centrifugation, and the precipitate was recovered and dried at 30 °C under vacuum for 24 h (this batch of samples contained 4 POTS, 3.06 g in total). Dialysis was carried out for 24 h, during which the water was changed several times. In the process of dialysis, to prevent the dialysis bag from bursting, the external water gradually penetrated the dialysis bag.

Au clusters were synthesized by combining 5.16 mL of a 13.67 mmol HAuCl₄ aqueous solution and L-glutathione (L-GSH, 30 mmol) in water medium, and the whole mixture was heated at 80 °C for approximately 5 h under ambient atmosphere. The color of the solution changed from clear to light-yellow. This color change indicated that gold nanoclusters were synthesized, and their optical properties could be measured using different instruments. Ag clusters were synthesized previously [42]. By using a microinjection system, silver metal precursors (AgAc, 0.25 mmol) and L-GSH (1 mmol) molecules were added to aqueous solution (50 mL) in an ice bath. After a simple stirring reaction of approximately 15 min, a cloudy white liquid was obtained, indicating the formation of a suspension containing silver and thiolate. A solution of NaBH4 (12.5 mL, 2.5 mmol) was added dropwise to the ice-cold mixture, the color of the solution changed from white to slightly yellow immediately, and then to deep brownish-black after approximately 3 h. Cu clusters were synthesized as in follow approach, 2 mL of a copper precursor (CuCl₂, 20 mM) and 10 mL of BSA (15 mg mL^–1) were mixed in aqueous solution and stirred for 0.5 h. Then, a sodium hydroxide solution with a pH of approximately 12 was added. The whole mixture was kept at 55 °C for 6 h with continuous stirring.

Preparation of photoluminescent metal clusters self-assembled with CQD into heterostructural nanoparticles (HNPs)

HNPs were carried out in mixed aqueous medium by adjusting the pH under different conditions. After this reaction, the mixtures were exposed to an ambient atmosphere and stirred for approximately 10 h to ensure the metal clusters were completely mixed with the CQDs in aqueous medium and self-assembled HNPs were obtained. In order to remove excess metal ions and the L-glutathione complex, the HNPs were precipitated and purified three times by adding isopropanol to the system and centrifuging at 10,000 rpm min⁻¹ for 15 min. The precipitate was completely dispersed in deionized water or a buffer solution. For further utilization, these samples were placed in the dark for at least one week. Over time, a white precipitate was generated, and the solution gradually became nontransparent. This precipitate, obtained through the self-assembly process, was the final product.

Electron microscopy and optical characterization

UV–Vis absorption spectra were recorded in the range of 200–800 nm by using Varian Cary 50 UV–Vis Spectrophotometer in absorbance mode. The photoluminescence (PL) spectra were obtained by a FLUOROMAX-4 spectrofluorometer equipped with a Xenon lamp. A transmission electron microscope (TEM, JEM-2100F, JEOL Ltd., Japan) with an energy dispersive spectrophotometer (TECNAI G2 F20) was employed for TEM characterization with an accelerating voltage of 200 kV and a Gatan SC200 CCD camera. FT-IR spectra data were recorded with a Bruker Vertex 70 spectrometer in the range from 4000 cm⁻¹ to 400 cm⁻¹.

In vitro photothermal performance

All the HNPs and single components solutions (4.4 mg mL⁻¹, in PBS) were added to EP tubes under the 808 nm irradiation (0.9 W cm⁻², 600 s). The HNP solution with various concentrations (1.76, 2.64, 3.52 mg mL⁻¹) was tested at the same irradiation condition. Additionally, various irradiance (0.7 W cm⁻², 0.9 W cm⁻², 1.1 W cm⁻², 1.3 W cm⁻²) were used to irradiate the Au–C solution (2.64 mg mL⁻¹) for 10 min. To assess the photothermal stability of the Au–C solution (4.4 mg mL⁻¹), 5 photothermal cycles were carried out under irradiance of 0.9 W cm⁻². For each cycle, 6 min of irradiation was used to increase the temperature to its maximum, followed by cooling to ambient temperature. A thermal imaging system (288, Fotric, China) was used to monitor the temperature.

In vitro antibacterial efficiency

A spread plate method was used to evaluate the antibacterial efficiency in vitro with Escherichia coli (E.coli, ATCC25922) and Staphylococcus aureus (S. aureus, ATCC25923) chosen as gram-negative and gram-positive model bacteria. First, 500 μL of the bacterial suspension (OD600 = 0.6) was prepared in each EP tube. The bacteria were collected by centrifugation (10,000 rpf, 3 min) and rinsed with PBS three times. Next, 200 μL of Au–C, Ag–C, and Cu–C solutions (4.4 mg mL⁻¹) and PBS were gently mixed with the bacteria in the EP tubes. These four groups were placed at 37 °C for 10 min. Another four groups were irradiated with an 808 nm laser (0.9 W cm⁻²) for 10 min. Finally, the bacterial suspensions were diluted 1 × 1010-fold, and then, 25 μL of the suspensions was spread evenly on an agar plate. All plates were then cultured in an incubator (37 °C) for 1 day. After the colonies formed, they were counted to determine the bacterial survival rate using Eq. 1:

$$\text{Survival}\text{Rate}(\%)=\frac{{\text{CFU}}_{\text{eg}}}{{\text{CFU}}_{\text{cg}}}\times 10$$ 1

where CFU_eg is the quantity of colony-forming units of bacteria in the experimental groups and CFU_cg is the quantity in the control groups. The bacteria were collected via centrifugation, and a live/dead bacteria-staining experiment was performed according to the protocols of the LIVE/DEAD BacLight Bacterial Viability Kit (Molecular Probes, the U.S.). The live and dead stain, containing 50% SYTO9 dye and 50% propidium iodide, was mixed with the bacterial suspension and cultured in an incubator (37 °C) for 20 min. Images of a sample of each group were captured with a fluorescence microscope (EVOS FL Auto, Life Technologies, America).

Morphological studies of bacteria

The morphological changes of the bacteria after the different treatments in vitro antibacterial experiment were observed under a scanning electron microscope. The bacteria in each group were fixed with a 2.5% glutaraldehyde solution for 24 h after washing with PBS 3 times. Then, all the bacteria were dehydrated with alcohol at various concentrations (40, 60, 80, 90, 95 and 100%). The bacteria were then dried with a critical point dryer (EMCPD300, Leica, Germany) and coated with platinum to observe the bacterial morphology with a scanning electron microscope (SU8010, Hitachi, Japan).

Hemolytic activity assay in vitro

Fresh mouse blood was centrifuged at 3000 r min⁻¹ for 5 min, the supernatant was discarded, and phosphate-buffered saline (PBS) was added to wash the RBC sediment. Then, the solution was centrifuged again at 3000 r min⁻¹ for 3 min, and the supernatant was discarded; this procedure was repeated 3 times to obtain a clean RBC sediment. Then, 1 mL RBC sediment was added to 9 mL PBS to obtain a 10% erythrocyte suspension. Then, 0.5 mL of the erythrocyte suspension was mixed with 0.5 mL of the Au–C, Ag–C and Cu–C suspensions of different concentrations (0.88, 1.76, 2.64, 3.52 mg mL⁻¹ in PBS). All the mixtures were placed in an incubator (37 °C) for 1 h, followed by centrifugation at 3000 r min⁻¹ for 10 min. The supernatant of each group was transferred to a new 96-well plate, and the absorbance was measured at 540 nm wavelength with a multimode plate reader (ELX800, BioTec, America). The hemolysis ratio was calculated using Eq. 2:

$$\text{Hemolysis}\text{Ratio}(\%)=\frac{{\text{A}}_{\text{S}}-{\text{A}}_{\text{N}}}{{\text{A}}_{\text{P}}-{\text{A}}_{\text{N}}}$$ 2

where A_S represents the absorbance of Au–C–, Ag–C– or Cu–C–treated groups, A_N represents the PBS-treated group and A_P represents the Triton-treated group.

Cytotoxicity assay in vitro

To explore the cytotoxicity of the photothermal materials at different concentrations, L929 cells were cultured in 96-well plates at a density of 5.0 × 10³ cells per well for 24 h first. After that, culture medium containing different concentrations of Au–C, Ag–C, and Cu–C (0.88 mg mL⁻¹, 1.76 mg mL⁻¹, 2.64 mg mL⁻¹ and 3.52 mg mL⁻¹ diluted with the culture medium) was added to replace the original medium. These cells were cultured continuously for an additional 1, 3 and 5 days. Then, the cell viability was evaluated via a Cell Counting Kit 8 (CCK8) following the manufacturer’s instructions. Absorbance (A450) was measured with a multimode plate reader (ELX800, BioTec, America). The cell viability was calculated according to Eq. 3:

$$\text{Cell}\text{Viability}(\%)=\frac{{\text{A}}_{450}\text{of}\text{treated}\text{cells}}{{\text{A}}_{450}\text{of}\text{control}\text{cells}}\times 100\%$$ 3

To further quantify cytotoxicity, a live/dead cell staining experiment was performed. L929 cells were cultured with Cu–C, Ag–C and Au–C (2.64 mg mL⁻¹ in culture medium) for 3 days. The staining procedure and subsequent visualization were similar to the live and dead bacteria staining experiment, which was outlined before.

Antimicrobial effect of Au–C in vivo

The animal experiments were permitted by the Ethics Committee in Stomatological Hospital, Chongqing Medical University, Chongqing, China (Approval number: CQHS-REC-2020 LSNo.52).

Male BALB/c mice (6 weeks old) were used for the subcutaneous abscess mouse model. Methicillin-resistant Staphylococcus aureus (MRSA) (ATCC43300) was used in this study. Briefly, after overnight culture, 50 μL of MRSA (10⁸ CFU mL⁻¹) was subcutaneously injected into the hips of the mice. Twenty-four hours later, abscesses could be observed, and the mice were randomly divided into 8 groups (n = 7): Au–C + NIR, Ag–C + NIR, Cu–C + NIR, PBS + NIR, Au–C, Ag–C, Cu–C, and PBS. Fifty microliters of the Au–C, Ag–C, and Cu–C solutions (2.64 mg mL⁻¹ in PBS) and PBS were injected into the abscess sites. Then, the abscesses of the mice in the four NIR groups were treated for 10 min with NIR laser (808 nm, 0.9 W cm⁻²) irradiation, and the other four groups were treated equally but without NIR laser irradiation. After one day later, two mice from each group were sacrificed, and the abscess site of the tissue was dissected for antibacterial analysis and H&E staining. The abscesses of the remaining mice were photographed, and an electronic balance was used to record their weight every other day. After 17 days, all the mice were sacrificed, and the skin tissues and major organs (heart, liver, spleen, lung and kidney) were collected for H&E staining.

Statistical analysis

SPSS software (Version 19.0, America) was used in this study to analyze all the data, which were evaluated as the mean ± standard deviation and contrasted via Kruskal–Wallis one-way analysis of variance. The data were considered significantly different for P < 0.05.

Results and discussion

Synthesis and characterization of the obtained HNPs.

Representative TEM images of assembled HNPs as showed in Fig. 1, the mean sizes of Au–C, Ag–C and Cu–C HNPs reached to 50 nm, 25 nm, and 1.2 μm, respectively. According to elemental mapping, it obviously showed each constitute such as C, Au,Ag and Cu exactly. It fully proved that these self-assembling processes were well constructed. For instance, it conveyed relative growth process of Au cluster assembled with CQDs form initial stage to resulted Au–C HNPs. Moreover, the structural information illustrated the morphological evolution and implied the surface characterization with pore particularity (Fig. S1). The outer morphology was similar to spherical shape, and porous characterization by HR-TEM (Fig. S2). As apparent from the formation analysis of the HNPs, their homogeneous characteristic distributed separately in EDS-mapping, it included conductor (Au, Ag, Cu clusters) and semiconductor components (CQDs), respectively. According to the size of individual C QDs (Fig. S3), it proved that HNPs were successfully fabricated via self-assembly procedures in alkaline atmosphere. It indicated one possibility in forming assembled heterogeneous structure with achieved hot-electron effect, and it also improved the photothermal conversion efficiency when exposed to photoirradiation [43]. The internal structure of HNPs completely changed since metal and carbon atoms formed cross-links during self-assembly [44–49]. Ascribed to the distinction of and protected ligands and their spatial resistance, the resulted morphology of HNPs showed the less difference in preparation.

The optical property confirmed the colloidal solution could disperse well in aqueous mediums (Fig. 2a), and photoluminescence (PL) showed the emission spectra of HNPs (Fig. 2b) [50]. Compared with the individual constitutes, it obviously induced a Stokes shift after self-assembled fabrication upon excitation (λ_ex = 365 nm). Due to existed PL property without quenching phenomenon, it conveyed that linkage interaction between metal cluster and carbon, which could keep functionalization balance. It proved the assembling process might be induced through chemical bonding and electrostatic adsorption between two components. By using glutathione (GSH) and bovine serum albumin (BSA) molecules as the protected ligands for forming metal clusters, it could provide numerous sophisticated, ingenious and accurate structures in self-assemble behavior. Figure 2c showed zeta potential while HNPs (Cu–C, Ag–C and Au–C) incubating with DDW and PBS (pH = 7.4), respectively. The zeta value conveyed HNPs still maintained stability and can be preparing for further applications. It still maintained the stable status as CQDs solution negative charge (Fig. S4). The main mechanisms in whole assembling process can be ascribed to the formation of electrostatic interactions and hydrogen bonds. For this investigation, the existed functional groups (–SH, –COOH, –NH₂) provided higher possibility in spontaneous assembling HNPs. As showed in FT-IR characterization (Fig. S5), the absence of distinct infrared peaks corresponding to assemble information. The thiol group (–SH, 2520 cm⁻¹), it indicated that surface status of the pure clusters but also conveyed assembled HNPs structure process after clusters integrated with CQDs. There has no sharp peaks indicated that amino group (–NH₂, 3317 cm⁻¹) were functionalized with carboxyl (–COOH, C = O: 1760 cm⁻¹, –OH: 2500 ~ 3300 cm⁻¹) under acylamino response. The weak response (vibration absorption) toward hydroxide (–OH) and carbonyl group (C = O) indeed implied the assembled HNPs were achieved exactly [51]. Due to the cooperative effect of electrostatic interactions and hydrogen bonds, HNPs were constructed finally. After integration, its interface and the surface functionalities were interacted through amide reaction between the carboxylic and amino groups of metal clusters. Hydrogen bonds might cause the aggregation of particles forming a bigger nanosphere.

Fig. 2.

Characterization of the obtained HNPs. a The magnified UV–vis absorption at the range of 250–700 nm. b Normalized emission spectra of water-dispersed samples. c Zeta potential curves of HNPs d Single-wavelength dynamics of the HNPs probed at 365 nm. The curves are fitting plots taking into account of electron–phonon thermalization in form of regular absorption phenomenon. e The correlation of each HNP with different regular absorption phenomenon. f Photothermal stability assessment of Au–C. g Photothermal performance of HNPs and the single components at the same concentration. h Photothermal conversion efficiency of HNPs. i Heating rates of HNPs and the single components. j Photothermal performance of Au–C under different NIR irradiances. * P < 0.05, ** P < 0.01, *** P < 0.001

The curves are fitting plots taking into account of electron–phonon thermalization in form of regular absorption phenomenon through the single-wavelength dynamics of each HNPs (Fig. 2d). The regular absorption spectroscopy was about the dynamics of hot electrons might be attributed to the electron–phonon relaxation of hot electrons [52, 53], and tracked by the regular absorption which could lead to cooling into phonon modes. The Au–C exerted the highest hot electron generation efficiency (Fig. 2e). It conveyed same tendency while HNPs (Au–C, Ag–C, Cu–C) were characterized in photothermal characterizations (808 nm laser irradiation). The temperature of all the HNPS gradually increased to a plateau status with continuous irradiation (Fig. 2g). Under same conditions (concentration: 4.4 mg mL⁻¹; irradiance: 0.9 W cm⁻²; time: 600 s), Au–C, Ag–C and Cu–C exhibited significantly higher heating effect than the individual components (CQDs, AuNCs, AgNCs, CuNCs) (Fig. 2g, 2i), which clearly proved HNPs offered higher photothermal effect compared with single components. All the HNPs displayed a concentration-dependent temperature increase under same irradiation (Fig. S6). The photothermal conversion efficiency (PTE) of all the HNPs were calculated according to Jin et.al [54] (Fig. 2h). Among all the HNPs, Au–C exhibited the best photothermal effect, with the highest heating rate of 4.37 °C/min and PTE of 54.16%. Therefore, we further investigated Au–C and found that, at the same concentration, it exhibited an irradiance-dependent temperature increase (Fig. 2j). Additionally, in the photothermal stability assessment, Au–C maintained a stable maximum temperature over five consecutive cycles (Fig. 2f), indicating its long-term stability and efficient photothermal capability. Au–C is a promising material for photothermal antibacterial applications. It has been reported that temperatures above 50 °C can potentially induce cell death through overheating [55]. This result can be attributed to an induced hot-electron transfer efficiency between Au (conductor) and CQDs (semiconductor) under 808 nm laser irradiation. Furthermore, the 808 nm laser promoted an electron transition in molecular orbitals between electrons and the electron levels of electrons generated by valence electrons [56].

In vitro antibacterial effect

The photothermal antibacterial effects of Cu–C, Ag–C and Au–C were evaluated in vitro against both gram-negative bacteria (E. coli) and gram-positive bacteria (S. aureus). Live/dead bacterial staining images showed that green fluorescence (live bacteria stain) was observed in the Cu–C, Ag–C, Au–C, PBS and PBS + NIR groups. A large area of green fluorescence with nearly no red fluorescence was observed in no NIR groups and PBS groups. In the Cu–C + NIR group, nearly little red-stained bacteria were observed while several bacteria were stained red in Ag–C + NIR group, which reflected that the antibacterial efficacy of Cu–C and Ag–C was limited. Interestingly, red fluorescence predominated in the Au–C + NIR group, indicating the apparent bactericidal effect of Au–C after NIR irradiation (Fig. 3a). The photothermal antibacterial effect of HNPs was further verified quantitatively by bacterial culture. As showed in Fig. 3b, there are significantly lower survival rate of E. coli and S. aureus in HNPs + NIR groups compared with no NIR groups and PBS groups. HNPs under NIR also exhibited excellent antibacterial rate, in which Au–C performed the highest antibacterial rate of 99.38% to E.coli and 99.06% to S. aureus. Furthermore, changes in the bacteria morphology were observed under a scanning electron microscope (Fig. 3c). Compared with PBS and all the groups without NIR, the bacterial surfaces in the HNPs + NIR treatment groups exhibited varying degrees of morphological destruction. Small wrinkles were observed on the bacterial surface in the Cu–C + NIR groups, whereas the bacterial surfaces in the Ag–C + NIR groups were rougher and more wrinkled. The Au–C + NIR groups showed the highest bacterial destruction effect, and total destruction of the basic cell morphology with exposure of the cellular contents occurred. These results above indicated that Au–C had the best antibacterial ability among the HNPs, which was consistent with its excellent photothermal performance.

Fig. 3.

Antibacterial effect of HNPs with/without NIR in vitro. a The live and dead bacteria staining assay. (Scale bar = 25 μm). b Bacteria survival rate and HNPs antibacterial rate analysis. c SEM images of the morphologies of bacteria treated with PBS, Cu–C, Ag–C, and Au–C with or without NIR irradiation. (Scale bar = 250 nm) * P < 0.05, ** P < 0.01, *** P < 0.001

In vitro biocompatibility assays

Biocompatibility is a crucial prerequisite for the in vivo application of biomaterials. Since HNPs exhibited a concentration-dependent temperature increase under the same irradiation (Fig. 2g), it is essential to conduct biocompatibility assays to identify the optimal concentration of HNPs that provides effective photothermal performance while minimizing potential side effects. In the hemolysis assay, obvious red colour was observed in the supernatants of the positive control group (2% Triton) while the supernatants of PBS and HNPs were almost colourless and transparent (Fig. 4a). Nearly all the HNPs (at 0.88–2.64 mg/mL) had hemolysis ratios lower than the permissible limit (5%) [57, 58]. When at 3.52 mg mL⁻¹, Ag–C had a hemolysis ratio of 7.49%, while Au–C and Cu–C still possessed low hemolysis ratios. The CCK8 assay was used to further explore the cytotoxicity of the HNPs at different concentrations (Fig. 4b). When the concentration at 0.88–2.64 mg/mL, Ag–C and Au–C exhibited excellent biocompatibility as the relative cell activity reached to over 80% at day 5. As for Cu–C, the cytotoxicity was acceptable only at a low concentration (0.88 mg mL⁻¹). At a high concentration of 3.52 mg. mL⁻¹, the relative cell activity of all the groups was lower than 80%. The live and dead cells staining assay of HNPs at a concentration of 2.64 mg mL⁻¹ was consistent with the above results (Fig. 4c). Because Au–C exhibited the highest PTE under the same concentration and irradiance among the HNPs. Based on the results above, we conclude that a concentration of 2.64 mg mL⁻¹ is likely the optimal concentration of Au–C, as it provides effective photothermal performance while maintaining a negligible hemolysis rate and relatively high cell viability.

Fig. 4.

Biocompatibility test of HNPs. a The relative hemolysis ratios of various concentrations of HNPs and the corresponding photographs. b The relative cell activity after treatment with HNPs at different concentrations for 5 days. c The live/dead cell staining result of HNPs at a concentration of 2.64 mg mL.⁻¹ at day 3. (Scale bar = 50 μm) * P < 0.05, ** P < 0.01, *** P < 0.001

In vivo antibacterial effect

Inspired by the biocompatibility and the outstanding bactericidal efficacy of Au–C in vitro, the in vivo therapeutic effect was further investigated. BALB/c mice with MRSA-infected subcutaneous abscesses were used as animal models [59]. After subcutaneous injection of MRSA, weight of all the mice decreased obviously in 3 days (Fig. S7). After NIR irradiation (10 min), the temperature in the abscess region of Cu–C + NIR-, Ag–C + NIR- and Au–C + NIR-treated mice increased from approximately 37 °C to 45 °C, 48 °C and 52 °C, respectively, whereas the temperature only reached 38 °C in abscess regions of mice treated with PBS + NIR (Fig. 5d–f). This result indicated that all the materials exhibited photothermal effects to different extents and that the Au–C + NIR group showed the best photothermal effect in vivo. Heating above 50 °C in vivo results in damage to bacterial enzymes, lipids and proteins, while tissue shows visible denaturation when heated above 55 °C [60, 61]. The temperature induced by the Au–C with NIR group was approximately 52 °C, which can cause bactericidal effects but little damage to tissue. One day after treatment, the bactericidal effect was analyzed and observed the infected tissue, the skin tissue surrounding the abscess site was collected from each group of mice. Severe inflammation was observed in the skin of each group compared with normal tissue [62]. Abundant neutrophils emerged, and the connective tissue and collagen fibrils were affected one day after MRSA infection (Fig S7c–d). Fewer colonies were observed on the agar plates of both the Ag–C + NIR and Au–C + NIR groups than of the PBS group, while the Au–C + NIR group showed the best antibacterial effect (Fig. S7a). As time passed, mice gained weight and the lesion size decreased gradually (Fig. 5a-b, Fig S8). The lesion size in the Au–C + NIR group decreased faster than other groups, indicating that mice treated with Au–C + NIR healed faster than the other groups (Fig. 5b). At day 17 after treatment, the wounds in Au–C + NIR groups were basically healed, while in mice from other groups remained, demonstrating the considerable bactericidal effect of Au–C in vivo (Fig. 5c). H&E staining was performed on all groups to assess lesion healing. On Day 17, the Au–C + NIR group exhibited the shortest scar length compared to the other groups (Fig. S7b, 7d). Skin infections usually trigger the immune system to generate abundant inflammatory cells, such as neutrophils [63]. Compared to untreated normal skin, numbers of polymorphonuclear leukocytes were observed in the infected skin with tissue necrosis (Fig. S7c). Compared with other groups, vessels containing red blood cells formed and there were fewer neutrophils in the Au–C + NIR group (Fig. S7d). It has been confirmed that when tissue heals, blood vessels are generated in the tissue microenvironment to supply blood containing nutrients and oxygen [64]. As a necessary element in the evaluation of the biocompatibility of HNPs, H&E staining of the major organs (heart, liver, spleen, lungs, and kidneys) was carried out to evaluate the systemic toxicity in vivo, and no evident tissue destruction was observed according to Fig. S9.

Fig. 5.

a Photographs of the abscesses in mice in the different groups. (Scale bar = 1 cm) b Lesion size change curves in different groups. c Lesion size of mice in day 17. d Thermal images in NIR groups during irradiation. e Temperature curves in NIR groups during irradiation. f Temperature in NIR groups at 10 min of irradiation. * P < 0.05, ** P < 0.01, *** P < 0.001

Conclusions

In summary, the HNPs (Au–C, Ag–C and Cu–C) were synthesized through self-assembly and exhibited better photothermal effect compared with noble metal (Au, Ag and Cu) clusters and CQDs. Au–C was the most effective photothermal material with PTE of 54.16% and negligible biotoxicity even at relatively high concentrations. Both in vitro and in vivo antibacterial assays consistently identified Au–C as the most promising candidate for bacterial eradication and abscess healing. Furthermore, all of the photothermal materials exhibited minimal long-term toxicity to critical organs. Given the above advantages, our work constructed a promising photothermal material for effective bacterial elimination with the possibility of further enhancing its applications. Expanding the composition scope and designing new components may enable the application of heterogeneous HNPs for bioimaging, biosensing, and biocatalysis.

Author contributions

Conceptualization: Xulei Yuan, Shaojun Liu, Jun Wang; Methodology: Xulei Yuan, Jun Wang; Formal analysis and investigation: Shaojun Liu, Jiaqi Liu, Fang Qin; Writing—original draft preparation :Xulei Yuan, Jun Wang, Xiang Mao; Writing—review and editing: Jinling Song, Xiang Mao; Funding acquisition: Xiang Mao,Jinling Song; Resources: Jun Wang, Min Zhang; Supervision: Xiang Mao, Jiunling Song. All the authors read and approved the manuscript.

Funding

This projected was funded “Chongqing Municipal Natural Science Foundation (CSTB2022NSCQ-MSX0113); Scientific Foundation Project of Chongqing Education Commission (KJQN201900436); National Natural Science Foundation of China (32070826)”.

Data availability

Data is provided within the manuscript or supplementary information files.

Code availability

Not applicable.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent to publication

Not applicable.

Competing interest

The authors declare no competing interests.