The selenium powder (99.5%, 325 mesh, Alfa Aesar), sodium ascorbate (99%, Adamas-Beta), PVP (MW = 40,000, K30, Sigma), SnCl₂·2H₂O (99.99%), NaBH₄ (98%), and HAuCl₄·3H₂O (99.9%) were purchased from Aladdin Biochemical Technology Co., Ltd. (Shanghai, China). Methanol (GR) and K₂CO₃ (AR) were purchased from Sinopharm Chemical Reagent Co., Ltd. Luciferase isothiocyanate (FITC, ≥90%), 3, 3′, 5, 5′-tetramethylbenzidine (TMB), and methylene blue (MB, 95%) were purchased from Sigma-Aldrich Co., Ltd. DCFH-DA and Hoechst 33342 staining solution were purchased from Beyotime Co., Ltd. All chemical agents in this work were utilized without further purification.

First, an A stock Se precursor solution was prepared. In brief, 0.104 g Se powder was added into a 100-mL beaker filled with 40 mL deionized (DI) water, heated in advance at 60°C, and then stirred vigorously for 15 min. When the Se powder floated on the upper surface of the water, 0.20 g NaBH₄ was added and stirred for another 15 min until the solution became transparent. Then, the B stock Sn precursor solution was prepared by adding 0.20 g SnCl₂·2H₂O into 24 mL of 24 mg/mL sodium ascorbate solution with stirring until the solution turned milky white. Finally, A and B were mixed and transferred to a 100-mL autoclave vessel and then reacted at 200°C for 20 h. The products were washed with DI water three times and then dried at 60°C to obtain the SnSe nanorod powder.

For the synthesis of Au–SnSe, 10 mL of 20 mM HAuCl₄ solution and 10 mL of 20% methanol were added to the 10 mL of 1 mg/mL SnSe aqueous dispersion system. The pH of the solution was adjusted to 9.7 with K₂CO₃ and then sonicated (40 kHz, 80 W) for 1 h. Finally, the products were washed with DI water three times to obtain Au–SnSe.

For PVP modification, 100 mg Au–SnSe and 0.5 g PVP were dispersed in 100 mL DI water and sonicated at 60°C for 3 h. After being centrifugated with DI water three times, ASNPs were obtained.

Transmission electron microscopy (TEM) images were collected using JEOL 200CX at 300 kV. X-ray diffractometry (XRD) measurement was performed on a Rigaku D/MAX-2250 V X-ray diffractometer at Cu Kα (λ = 0.154056 nm) within the 2θ range of 20–90° at a scanning rate of 5°/min. Fourier transform infrared (FT-IR) spectra were obtained using a Bruker Tensor II FTIR spectrometer. X-ray photoelectron spectroscopy (XPS) plots were detected by Shimadzu, AXIS SUPRA. Zeta-potential data were collected with Nicomp Z3000 SOP. Ultraviolet-Visible (UV-Vis) absorption spectra were recorded on Shimadzu UV-3600 Plus. An infrared thermal imager (FLIR A325SC camera) was used to record temperature changes and thermal image information. MTT assays were conducted with a Spark™ multimode microplate reader. Confocal laser scanning microscopy (CLSM) images were obtained using a Nikon A1+R-980 confocal microscope.

The photothermal performances of SnSe–PVP and ASNPs were measured using the infrared thermal imager under 808-nm laser irradiation. The solution temperature was recorded under SnSe–PVP and ASNPs aqueous (1 mL, 50 μg/mL) with different 808-nm laser power densities (0.5, 1, and 1.5 W/cm²) and 808-nm laser (1 W/cm²) with different SnSe–PVP and ASNPs concentrations (100, 50, 25, and 12.5 μg/mL). Then, the ASNPs aqueous dispersion was subjected to multiple laser irradiation heating and natural cooling cycles, and the temperature curve was recorded to measure the photothermal stability and calculate the photothermal conversion efficiency.

A total of two typical ∙OH indicators, MB, and TMB were used to detect ∙OH. First, SnSe–PVP and ASNPs were dispersed in MB solution at a concentration of 50 μg/mL. The as-achieved solution was then irradiated with an 808-nm laser (1 W/cm²) for 5 min, then cooled naturally, and cycled different times (1, 2, and 4 times). Subsequently, SnSe–PVP and ASNPs were dispersed in H₂O and H₂O₂ at a concentration of 50 μg/mL, then irradiated with an 808-nm laser (1 W/cm²) for 5 min, and cooled naturally. The UV-Vis spectra of MB were measured using a UV-Vis spectrophotometer. Similarly, TMB was also used to further detect the ∙OH productivity of ASNPs with different laser power densities and material concentrations.

The 4T1 murine breast adenocarcinoma cell line (obtained from the Shanghai Institute of Cells, Chinese Academy of Sciences) was cultured in RPMI 1640 (Gibco, United States) supplemented with 10% heat-inactivated fetal bovine serum, streptomycin (100 mg/mL), and penicillin (100 units/mL) and cultured at 37°C in a humidified atmosphere along with 5% CO₂.

Cytotoxicity and phototoxicity were tested on 4T1 cells. The 4T1 cells were seeded into 96-well culturing plates in a density of 5×10³ cells/well and incubated overnight. Next, ASNPs dissolved in the fresh culture medium with the concentrations of 0, 12.5, 25, 50, and 100 ppm were added to exchange the culture medium in the 96-well plates. After a 6 h co-incubation, the culture medium was refreshed with the fresh culture medium, then irradiated with an 808-nm laser (1 W/cm²) for 5 min, and incubated for another 24 h. Another 96-well plate inoculated with 4T1 cells was irradiated with an 808-nm laser at different power densities (0.5, 1, 1.25, and 1.5 W/cm²) for 5 min without ASNPs and then cultured for another 24 h. The cell viability was examined via a standard MTT assay on a microplate reader (Bio-TekELx800, United States) by testing the absorbance at 490 nm.

The 4T1 cells were seeded with 1×10⁵/mL cells in RPMI 1640 in confocal laser scanning microscopy (CLSM) culture vessels and permitted to adhere overnight.

The 4T1 cells were cocultured with FITC-labeled ASNPs for 4 or 6 h and then washed with PBS. The endocytosis was observed under confocal mode after Hoechst 33342 staining.

The 4T1 cells were first cocultured with ASNPs (50 μg/mL) for 6 h. After refreshing the culture medium, the cells were irradiated with an 808-nm laser (1 W/cm²) for 5 min and then incubated for another 12 h. The cells were stained with Hoechst 33342 as nuclear staining solution and DCFH-DA as the ROS fluorescent probe before CLSM observation.

All animal experiments were conducted according to the guidelines of the Institutional Animal Care and Use Committee and the Experimental Animal Ethics Committee of East China Normal University, and the accreditation number was m⁺R20190701.

ICR mice (female, 7 weeks old) and BALB/c mice (female, 6 weeks old) were obtained from the Laboratory Animal Center of East China Normal University.

A total of fifteen ICR mice were randomly divided into three groups. The control group was intravenously (i.v.) injected with 100-μL saline solution, 3 and 30-days groups were i.v. injected with 100-μL ASNPs solution (30 mg/kg). The mice in the control and 30-days groups were weighed every 3 days. Three and 30 days after injection, one mouse from each group was randomly selected and killed to take major organs (heart, liver, spleen, lung, and kidney) for further histopathology analysis using a typical hematoxylin and eosin (H&E) staining assay. Meanwhile, the blood samples of the other mice in each group were collected for routine blood and biochemical analysis.

The BALB/c mice were subcutaneously inoculated with 4T1 cells (1×10⁶, suspended in 100 μL PBS) on the right posterior axilla. When the tumor size reached 200 mm³, the mice were randomly divided into four groups: control group, ASNPs group (i.v., 15 mg/kg), near-infrared (NIR) group (808 nm laser, 1 W/cm², 5 min), and NIR + ASNPs group (i.v., 15 mg/kg; 808 nm laser, 1 W/cm², 5 min), with six mice in each group. One day after treatment, one mouse in each group was randomly selected and killed, and the tumors were taken for H&E and TUNEL immunofluorescence staining. The body weight and tumor volume [V= (length × width × width)/2] of the remaining mice were recorded every other day. The mice were euthanized when the tumor volume reached the ethical value.

Figure 1A is a schematic of ASNPs generating ∙OH in the PEDT process. Upon the irradiation of the 808-nm laser, heat was generated along the nanorod, and the temperature difference was formed in the heat conduction process. Under a temperature difference, the carriers (e^– and h⁺) in SnSe could be activated and driven to the hotter and colder sides, respectively. Owing to the differences of potential between electron acceptor Au-NPs and SnSe, the e^– accumulated on Au-NPs, hence promoting the carriers’ mobility and e^– and h⁺ separation, facilitating the gathering of h⁺ at the low-temperature end of ASNPs. This resulted in a significant ∙OH productivity increase due to e^–-rich Au causing H₂O₂ splitting (Lv et al., 2020) and h⁺–H₂O reactions. The TEM images (Figures 1B,D) show that the average particle size of SnSe and ASNPs is approximately 200 and 250 nm, respectively. The high-resolution TEM (HRTEM) images (Figures 1C,E) indicate considerable changes in interplanar spacing of SnSe and ASNPs, suggesting a distinct lattice distortion in ASNPs induced by the modification of Au-NPs. XRD patterns show that Au–SnSe composite structures were successfully constructed (Figure 1F). The FT-IR spectra (Figure 1G) and zeta-potential (Supplementary Figure S1) results show that PVP was successfully modified onto the surface of Au–SnSe. In Figure 2, XPS was used to characterize the electron transfer of Au–SnSe compared with SnSe (Figures 2A–F). The binding energies of the Sn 3d and Au 4f peaks in Au–SnSe shifted to the high and low fields, respectively, after the construction of Au-NPs, indicating that electrons transferred from SnSe to Au, thereby proving the enhanced separation of e^– and h⁺ property of Au–SnSe nanocomposites.

Figures 3A–F show the temperature change of SnSe–PVP and ASNPs under different material concentrations and 808-nm laser intensities. The temperature change is proportional to the material concentration and laser intensity, and the photothermal conversion ability of ASNPs is slightly better than that of SnSe–PVP under the same conditions. The temperature variation curves of ASNPs were recorded under irradiation of an 808-nm laser at 1 W/cm² and in the process of natural cooling (Figures 3G, Supplementary Figure S2). These results indicated that ASNPs had stable photothermal conversion ability. By measuring the UV-Vis-NIR absorption spectra of ASNPs aqueous dispersion, the extinction coefficient of ASNPs at 808 nm was calculated to be 12.41 L∙g⁻¹∙cm⁻¹ (Supplementary Figures S3, S4). According to the aforementioned experimental results and calculation methods in previous literature reports (Lyu et al., 2016), the photothermal conversion efficiency of ASNPs was estimated to be 32.62%.

Compared with SnSe–PVP, ASNPs exhibited superior MB degradation ability, indicating higher ∙OH productivity of ASNPs upon 808-nm laser irradiation (Figures 4A,B). Furthermore, no obvious ∙OH production differences were observed in the H₂O or H₂O₂-dispersed SnSe–PVP group, whereas the ASNPs group presented a higher ∙OH productivity in H₂O₂ media (Figures 4C,D). The aforementioned results further confirmed that Au-NPs could conduct e^– out of ASNPs and promoted the separation of e^– and h⁺ in the thermoelectric process, and the e^–-rich Au-NPs could catalyze H₂O₂ into ∙OH. Thus, the reaction efficiency of h⁺–H₂O half-reaction was improved, and the e^–-rich Au-NPs-induced H₂O₂ homolysis process was constructed, thereby significantly improving PEDT performance. In addition, TMB was used to test the influence of ASNPs concentrations and laser power densities (resulting in different temperature differences) on the ∙OH yield, and the results showed that both of them are positively correlated with ∙OH yield (Figures 4E,F).

First, the biocompatibility of ASNPs was evaluated. From Figure 5A, ASNPs showed good biocompatibility in the absence of NIR irradiation, and the viability of 4T1 cells remained above 95% when the concentration of ASNPs reached 100 ppm. Under NIR irradiation, ASNPs exhibited a good cell-killing effect with IC₅₀ of approximately 50 ppm. Next, the cytotoxicity of NIR was evaluated, and almost no cytotoxicity was observed with NIR irradiation alone (Figure 5B). Figure 5C shows the endocytosis ability of 4T1 cells of FITC-labeled ASNPs. ASNPs were endocytosed into cells after cocultured for 4 h, and a high concentration level was still maintained in cells at 6 h. DCFH-DA was used as the ROS fluorescent probe to detect the PEDT process at the cell level. Compared with the other three groups, the intensity of ROS signaled by green fluorescence (DCFH-DA) was significantly increased in 4T1 cells treated with ASNPs + NIR (Figure 5D). The quantitative changes of ROS intensity were further statistically analyzed (Supplementary Figure S5). It can be seen that the fluorescence intensity of the ASNPs + NIR group was 3.84, 5.65, and 4.41 times stronger than that of the control, NIR, and ASNPs groups, respectively, showing the same trend as CLSM images of ROS.

The biosafety of ASNPs was evaluated before its in vivo treatment applications. There were no abnormalities in body weight, organ sections, and blood indexes of mice in each group, indicating the good biosafety of ASNPs (Supplementary Figures S6–S8).

During treatment, no obvious body weight changes were observed among the control, ASNPs (i.v., 15 mg/kg), NIR (808-nm laser, 1 W/cm², 5 min), and NIR + ASNPs (i.v., 15 mg/kg; 808-nm laser, 1 W/cm², 5 min) groups (Figure 6A). From the statistical results of relative tumor volume changes in each group and images of each group of mice 15 days after treatment (Figures 6B,C), the ASNPs + NIR group exhibited the best tumor growth inhibition performance, whereas the other three groups showed almost no tumor inhibition effect and with little differences. H&E and TUNEL-stained tumor tissue (Figure 6D) showed that cells in the control, ASNPs, and NIR groups hardly suffered any damage, whereas cells in the ASNPs + NIR group showed obvious typical apoptosis, indicating the tremendous tumor cell-killing effect of PEDT–PTT synergetic therapy.