*3.2. Characterization of SF@Cu-NFs Product*

### 3.2.1. Surface Morphology Measurement by LSCM and SEM Image

The morphologies of the synthesized SF@Cu-NFs were observed using and SEM images (shown in Figure 4a,b). As seen from SEM images, the prepared SF@Cu-NFs displayed highly peony flower-like morphology with diameters of about 50 μm.

**Figure 4.** (**<sup>a</sup>**,**b**) SEM images of SF@Cu-NFs product with low and high magnification; (**c**) EDS pattern of SF@Cu-NFs product.

Meanwhile, the EDS result was shown in Figure 4c and the related EDS data were listed in Table S1. The results identified the chemical species of the SF@Cu-NFs products and confirmed the presence of Cu, P, C, and O. The C and O can be attributed to SF protein and Cu, P and O can be attributed to Cu3(PO4)2. Meanwhile, the appearance of Cl and Na may be brought by the residual PBS. However, the N element was not detected in the prepared SF@Cu-NFs product by EDS analysis. Because SF protein was generally considered as nitrogen rich [58], the abnormal absence of N may be attributed to the abundant O and C in the SF@Cu-NFs product, which would cover up the N peak in EDS.

### 3.2.2. Chemical Structure Investigation by FTIR and XRD

The phase structures of the as-prepared Cu3(PO4)2 and SF@Cu-NFs were investigated by the XRD analysis and shown in Figure 5a. As observed, the diffraction peaks of SF@Cu-NFs and unmodified Cu3(PO4)2 were in good agreemen<sup>t</sup> with the Joint Committee on Powder Diffraction File data for Cu3(PO4)2 (File NO. 00-022-0548). As a result, it could be concluded that the petals in the hybrid nanoflowers were formed by regular arrangemen<sup>t</sup> of Cu3(PO4)2 crystals and the inorganic composition of SF@Cu-NFs was Cu3(PO4)2.

**Figure 5.** (**a**) XRD spectrum of (i) copper phosphate; (ii) SF@Cu-NFs nanoflower. (**b**) FTIR spectrum of (i) copper phosphate; (ii) SF@Cu-NFs nanoflower; (iii) SF protein.

By assigning peaks to various groups and bonds, the detail FTIR spectra of Cu3(PO4)2 (spectrum i), SF@Cu-NFs product (spectrum ii) and SF protein (spectrum iii) were shown in Figure 5b. As it can be seen, SF@Cu-NFs showed (1) weak peaks at 563 cm<sup>−</sup><sup>1</sup> and 603 cm<sup>−</sup><sup>1</sup> corresponding to flexural vibration of P-O, (2) strong peaks at1035 cm<sup>−</sup><sup>1</sup> corresponding to stretching vibration of P-O [59,60]. These signals corresponded to the asymmetric and symmetric stretching vibrations of PO4<sup>3</sup><sup>−</sup>, which were also present in Cu3(PO4)2 (spectrum i). Meanwhile, SF@Cu-NFs also showed (1) peak at 1638 cm<sup>−</sup><sup>1</sup> corresponding to stretching vibration of C-O originated from amide I, (2) peak at 1536 cm<sup>−</sup><sup>1</sup> corresponding to superposition of bending vibration of N-H and stretching vibration of C-N originated from amide II, (3) peak at 1421 cm<sup>−</sup><sup>1</sup> corresponding to bending vibration of N-H originated from amide III [61]. These signals corresponded to the major amide bands originating from SF protein, which were also present in SF protein (spectrum iii). Compared with Cu3(PO4)2 and SF, SF@Cu-NFs showed no significant changes before and after the formation. Without new absorption peaks or obvious peak shifts, the above results indicated that SF protein was fixed by self-assembly rather than by covalent bonds. Meanwhile, the structural integrity of the silk fibroin protein remained intact after product formation, which furthermore identified the successful preparation of SF@Cu-NFs.
