Brown alpaca fibres were collected from Australia pasture and were thoroughly cleaned by acetone/ethanol mixture prior to use. Hydrochloric acid (HCl, 32%), tetraethyl orthosilicate (TEOS, 98%), ammonium hydroxide (30%), ethanol (100%), acetone (100%), methyl methacrylate (MMA, 99%), and 1,1’-Azobis (cyclohexanecarbonitrile) (AICN, 98%) were purchased from Sigma-Aldrich.

The extraction method was conducted using the reported acid hydrolysis method [2]. Typically, 2 g of the brown fibres were mixed with 20 mL of HCl (1%) in a round-bottom flask and stirred for 5 min to remove the water-soluble components. After centrifugation, the precipitates were immersed in 80 mL of 32% HCl and heated at 100˚C in an oil bath with mechanical stirring for 3 h. The precipitates were then washed with deionized water and centrifuged at 4000 rpm for 5 min repeatedly until the neutral pH of the suspension was approached. The supernatant was removed and the samples were vacuum dried at room temperature in the dark.

The silica coating (SC) method was employed as described in the previous work [20]. Specifically, 10 mg of the melanin was dispersed uniformly in 1.5 mL of distilled water and 1.5 mL of ethanol. 0.4 mL of ammonia was mixed in 6 mL ethanol and poured into the melanin solution. After mechanical stirring for 10 min, a mixture of 0.05 mL of TEOS and 2 mL of ethanol was added dropwise. The final products were obtained after different stirring duration (5 h and 20 h) and were then washed with deionized water for three times, followed by drying at room temperature in the dark.

The core-shell structures of MSCP before and after SC were examined on a JEOL 2100 LaB6 TEM. The size and distribution of MSNP were measured by Dynamic light scattering (DLS) on Malvern Zetasizer. Samples were dispersed in deionized water, placed in low volume disposable cuvettes and measured under size mode. The absorption properties of the melanin solutions before and after SC were scanned with a Cary 3 UV-Vis spectrometer from 200 to 800 nm. The Fourier transform infrared spectroscopy (FTIR) spectra were recorded on a Bruker FTIR Spectrophotometer with ATR accessories from 4000 to 600 cm⁻¹ under the accumulation of 32 scans at 4 cm⁻¹ resolution. The simulated solar irradiation test was conducted with an Atlas Suntester XLS instrument that equipped with a 1500 W xenon arc lamp (light range: 300 - 800 nm). The irradiation dose was 350 W/m² and the chamber temperature was 35˚C. The solutions were under mechanical stirring during irradiation and the concentrations were kept same. 1 mL of each solution was taken out every 10 h and scanned with UV-Vis spectroscopy.

The morphologies of the original melanin and MSNP are shown in Figure 1.

The original melanin particles exhibit round or elliptical shapes with distinct sizes, which is inconsistent with the extracted melanin from other mammal’s hair samples in previous research [21] [22]. It can be seen that the particle surface is cratered instead of smooth which is due to the special bio-synthetic pathway and inhomogeneous nature of melanin that polymerizing the oligomers with a variety of units through different of bonds, such as π-stacking interactions [23]. Figure 1(b) of the MSNP-5 h barely changes compare with the original melanin in addition to the appearance of the exotic substances, meaning the formation of the silica shell outside melanin particles. The particles also seem to be separated apart from each other at this stage. When the coating duration reached 20 h, a conspicuous melanin core and silica shell structure was obtained (as shown in Figure 1(c)). The formation of the silica shell was slowly, continuously and evenly owing to the well-studied Stöber method. Specifically, the addition of ammonia assists the synthesis of silica from TEOS, and the NH₄⁺that adsorbs onto melanin helps to get a positively charged surface so that the hydrogen bond could be formed between NH 4 + and the OH⁻ of silica. Then, the continuous hydrolysis and condensation reactions took place in this alkali system, thereby achieving a uniform silica layer outside the melanin particles [24].

DLS measurement exhibits the distribution of the three particles in aqueous solutions (Figure 2). Apparently, the DLS result shows that original melanin is very easy to get serious particle aggregation, as described in previous study [25]. However, a 5h coating can largely decrease the aggregation of melanin, showing a particle size distribution from 200 to 1000 nm and an average size of 480 nm which is consistent with the TEM result (Figure 1(b)). The average particle size of MSNP-20 h becomes 530 nm with a 20 h coating. The increase contributes to the growing of the silica shell thickness and may imply the average shell thickness is around 25 nm (Figure 1(c)). The photo of solutions of MSNP-5 h and MSNP-20 h after standing overnight demonstrate again the welldispersity of melanin solution (inset of Figure 2).

The typical UV-Vis spectra of melanin can be obtained from the MSNP solution (Figure 3(a)) [26] and the UV-Vis spectrum of MSNP-20 h shows an increasing UV absorbing ability compared with the original melanin and MSNP-5 h at the same concentration, which probably owing to the reduced particle aggregating phenomenon that enhanced the overall particle surface area and solution dispersity [27].

The representative functional groups in the FTIR spectrum of the original melanin coincides well with the reported work (Figure 3(b)) [28] [29], including

the −OH, −NH₂ groups at around 3400 cm⁻¹, and aromatic C=C, COO−, C=O groups at 1600 cm⁻¹. The spectra of the MSNP samples exhibit significant difference below 1300 cm⁻¹. Especially, a strong peak at around 1070 cm⁻¹ refers linearsilica structure (Si-O-Si), a weak peak at 796 cm⁻¹ represents the network of silica [30], and a gentle slope at around 1220 cm⁻¹ indicates the unreacted TEOS residues [31].

It has been proved that coated with silica shell could inhibit the photocatalytic activity of the core material from irradiation to some extent [20] [32]. Therefore, the photoprotective effect of the SC on melanin particles is studied. The degradation trends of each melanin sample with increasing dose of simulated solar irradiation are shown in Figure 4. The A/A₀ refers to the ratio of the absorption after specific irradiation duration to the initial absorption at the wavelength of 500 nm. It can be seen that all of the three samples showed an overall downtrend after irradiation.

Specifically, the original melanin has barely degraded for the first 10 h irradiation. This probably caused by the severe aggregation of the melanin particles in the solution that slowed down the photo-induced impact on thesingle particle. On the contrary, a serious photodegradation (30%) of the MSNP-5 h took place at the first 10 h. In terms of the DLS tests, MSNP-5 h has already resulted in a great improvement of particle aggregation compared with the original melanin, but the TEM images and UV-Vis spectra indicated that the silica shell could not be fully formed for a 5 h coating. Therefore, the overall degradation trend of the aggregated original melanin is slower than the well separated but uncompleted MSNP-5 h. Notably, the MSNP-20 h showed a linearly degrading rate, and is much slower than the other solutions after irradiation for 30 h. This may owe to the uniformly dispersed MSNP that could absorb the light evenly, causing the

constant degradation speed. After exposure for 80 h, the melanin in MSNP-5 h has almost been exhausted, while less than 20% of the original melanin remained. Meanwhile, around 50% of the melanin within MSNP-20h has been protected, suggesting the efficiency of using the silica coating method on suppressing melanin from photodegradation.

Figure 5 displays the TEM images and optical photos of the 80 h irradiated samples. It is found that the original melanin particle has been largely shrunken (Figure 5(a)), and the boundary of the particle surface is getting vague and out of shape, which can reflect the irreversible degradation effect of the irradiation on melanin. Similarly, the MSNP-5 h in Figure 5(b) presents a small dark core inside the silica shell. The shape of the shell still maintained as an original melanin particle but become transparent due to the shrinkage of the melanin core. By comparison, the MSNP-20 h (Figure 5(c)) shows more melanin contained within the shell, indicating a better protection provided by this structure.

The colors of the irradiated solutions, on the other side, intuitionally supported the conclusions, which are darkened from MSNP-5 h (white), original melanin (pale yellow) to MSNP-20 h (yellow). The fading color coincident well with the UV-Vis results.