Rationally designed dual-plasmonic gold nanorod@cuprous selenide hybrid heterostructures by regioselective overgrowth for in vivo photothermal tumor ablation in the second near-infrared biowindow

NIR-II plasmonic materials offer multiple functionalities for in vivo biomedical applications, such as photothermal tumor ablation, surface-enhanced Raman scattering biosensing, photoacoustic imaging, and drug carriers. However, integration of noble metals and plasmonic semiconductors is greatly challenging because of the large lattice-mismatch. This study reports the regioselective overgrowth of Cu2-xSe on gold nanorods (GNRs) for preparation of dual-plasmonic GNR@Cu2-xSe hybrid heterostructures with tunable NIR-II plasmon resonance absorption for in vivo photothermal tumor ablation. Methods: The regioselective deposition of amorphous Se and its subsequent conversion into Cu2-xSe on the GNRs are performed by altering capping agents to produce the GNR@Cu2-xSe heterostructures of various morphologies. Their photothermal performances for NIR-II photothermal tumor ablation are evaluated both in vitro and in vivo. Results: We find that the lateral one- and two-side deposition, conformal core-shell coating and island growth of Cu2-xSe on the GNRs can be achieved using different capping agents. The Cu2-xSe domain size in these hybrids can be effectively adjusted by the SeO2 concentration, thereby tuning the NIR-II plasmon bands. A photothermal conversion efficiency up to 58-85% and superior photostability of these dual-plasmonic hybrids can be achieved under the NIR-II laser. Results also show that the photothermal conversion efficiency is dependent on the proportion of optical absorption converted into heat; however, the temperature rise is tightly related to the concentration of their constituents. The excellent NIR-II photothermal effect is further verified in the following in vitro and in vivo experiments. Conclusions: This study achieves one-side or two-side deposition, conformal core-shell coating, and island deposition of Cu2-xSe on GNRs for GNR@Cu2-xSe heterostructures with NIR-II plasmonic absorption, and further demonstrates their excellent NIR-II photothermal tumor ablation in vivo. This study provides a promising strategy for the rational design of NIR-II dual-plasmonic heterostructures and highlights their therapeutic in vivo potential.

The lattice mismatch between the FCC gold domain and the cubic berzelianite Cu2-xSe domain in the GNR@Cu2-xSe heterostructures can be calculated on the basis of the conventional definition using the following formula:  Figure S1. FDTD-calculated optical extinction spectra of GNRs, core-shell GNR@Se structures and coreshell GNR@Cu2-xSe structures. The models for FDTD simulations are shown as well. The dimensional parameters (length × width) of GNRs are adopted from the statistical results shown in Figure S1. The thickness of shell of core-shell GNR@Se structures and GNR@Cu2-xSe heterostructures is taken from GNR@Cu2-xSe structures prepared with PVP and 1.12 mM SeO2. S5 Figure S3. Large scale SEM image of as-prepared GNRs by the seed-mediated growth method. Figure S4. Optical absorption spectra of aqueous suspensions of GNRs modified with CTAB, CTAC, PVP, PSS and PDDA capping agents, respectively. Negligible spectral change in all these capping agentmodified GNRs was observed, indicating no significant aggregation.

S2. In vitro hemolysis assay
Hemolysis test was performed to evaluate the hemocompatibility of PEG-GNR@Cu2-xSe hybrids.
Typically, 1 mL of mice red blood cell (RBC) solution was obtained after anticoagulation and washing three times with PBS solution, and then dispersed in 4 mL of PBS. After that, 0.2 mL of RBCs obtained above was mixed with 0.8 mL of 50,75,100,200, 400 µg/mL). Ultrapure water and PBS were used as the positive control and negative control, respectively. Then the samples were shaken and kept stable at 37 ℃ for 4 h, the mixtures were centrifuged at 10000 rpm for 10 min.
Optical photographs were taken, and hemoglobin in the supernatant was estimated on the basis of the absorbance at 545 nm. The hemolysis ratio was calculated using the following equation: where Asample, Anegative and Apositive are the absorbance of the samples, negative control and positive control, respectively.

S3. Test of blood biochemistry
The analysis of blood biochemistry was performed with three groups of healthy mice (n=3). Blood was collected directly from either healthy mice injected with 50 μL PBS or at 24 h-and 48 h-postinjection from heathy mice intravenously administrated with PEGylated GNR@Cu2-xSe solution (50 μL, 200 μg/mL for gold). Levels of different makers including alanine transaminase (ALT), aspartate transaminase (AST), alkaline phosphatase (ALP), blood urea nitrogen (BUN), globulin (GLB) and creatinine (CREA) were tested to evaluate the hepatic and kidney functions. Figure S41. Analysis of blood biochemistry of heathy mice intravenously administrated with GNR@Cu2-xSe hybrids. ALP, AST, ALT, BUN, GLB and CREA levels from heathy mice after administration with GNR@Cu2-xSe hybrids at 24 h-and 48 h-post-injections are compared with the control mice. No significant difference for all these levels was observed, indicating no hepatic and kidney toxicity by GNR@Cu2-xSe hybrids prepared with PVP and 1.12 mM SeO2. Figure S42. Individual body weight change curves of breast tumor-bearing mice shown in Figure 7C, with various treatments. Figure S43. Photographs of breast tumor-bearing mice following various treatments including PBS + NIR-II laser, GNR@Cu2-xSe heterostructures only, and GNR@Cu2-xSe heterostructures prepared with various capping agents + NIR-II laser, respectively. Figure S44. Representative H&E histological images of tumor tissues of tumor-bearing mice following various treatments with PBS + NIR-II laser, GNR@Cu2-xSe heterostructures prepared with PVP, and GNR@Cu2-xSe heterostructures (prepared with various capping agents) + NIR-II laser. Scale bar: 50 μm.