Three varieties of silk cocoons, white jade (WJ, known as Baegokjam in Korea), golden silk (GS), and PS (known as Yeonnokjam in Korea), were provided by the National Academy of Agricultural Sciences in Korea. Figures 1a-1c and Fig. 1a1, b1, and c1 show the images of the WJ, GS, and PS cocoons and threads, respectively. Under daylight, the PS, WJ, and GS cocoons exhibit light green, white, and yellow colors, respectively. Because of the nutrients absorbed by the mulberry leaves, the silk fibers of all the varieties display fluorescence under UV excitation (Fig. 1a2, b2, and c2)43. The WJ silk thread shows a weak blue fluorescence, whereas the fluorescence intensities of the colored silk cocoons increase; notably, the PS cocoons exhibit a strong green fluorescence (upper right of Fig. 1c). A microscopy image of a single PS-cocoon thread, captured under UV light, confirms the emission of a strong and uniform green fluorescence from the fiber surface, indicating the existence of fluorescent materials in sericin (Fig. 1c3). Scanning electron microscopy (SEM) images show that the surface of a PS cocoon is smooth and contains microfibers (diameter: ~16 µm), resulting in a large surface-to-volume ratio, which increases the sensitivity of the PS cocoon (Fig. 1d).
To investigate the fluorescence quantitatively, we constructed a spectrometer set-up as shown in Fig. 1e. The UV light emitted by a 365-nm light-emitting diode (LED), mounted on the spectrometer set-up, was collimated, and focused on the silk cocoon samples. The emitted fluorescent light was collected by an optical fiber and sent to the spectrometer to record fluorescent spectra. Figure 1f shows the fluorescent spectra of the WJ, GS, and PS cocoon mats. Interestingly, the PS mat exhibits a strong green fluorescence with a central wavelength of 552 nm and full width at half maximum of 65 nm, whereas the fluorescence intensities of the WJ and GS cocoon mats are negligible (magnified plot in the right panel of Fig. 1f). Because silk fiber is composed of two proteins (fibroin in the core and sericin in the sheath), identifying the location of the fluorescent pigment is essential. The PS fiber mat was fully degummed to remove the outer sericin layer and excited by the UV LED. As shown in Fig. 1g, the fluorescence totally disappears after the degumming, indicating that the pigment molecules are mainly present inside the sericin layer44. The reflectance spectra of the GS and PS cocoons indicate the existence of color pigments in both samples (Fig. S1). However, the colors expressed by the GS and PS cocoons can be attributed to absorption and fluorescence, respectively. In addition, the PS fluorescence was compared to those of the commercial fluorescent dyes, i.e., riboflavin and sodium fluorescein, which are widely used in biological and medical applications, and their gains are sufficient to induce lasing. We prepared separate silk fibroin films containing 1% riboflavin and sodium fluorescein and demonstrated the strong optical gain of the dye-containing silk fibroin film for lasing. The fluorescence emitted from the PS cocoon mat was relatively weaker than that emitted by sodium fluorescein, but comparable to that of riboflavin (Fig. 1h). Notably, qualitative analysis revealed that the quantum yield of the fluorescent pigment in the PS sericin was comparable to that of a commercial dye.
The PS is the progeny of Japanese-originated green silk (Jam 315) and Chinese-originated white silk (Jam 316) cross-species and can therefore simultaneously carry two genes Ga and Gb, which complement each other to produce a light green color. Another independently inherited Gc gene, which is dominant in white cocoons, produces green cocoons with a stronger color. Silkworms produce two main pigments, viz. flavonoids, which impart green color to the cocoons, and carotenoids, which control the yellow and pinkish colors of cocoons (such as the GS)45,46. A green silk species (Antheraea yamamai, AY) was prepared and its fluorescent spectra were measured for comparison (Fig. S2). The green silk cocoon exhibited a sky-blue fluorescence due to the existence of biliverdin47, but the intensity was much weaker than that emitted by the PS cocoon (Fig. S3). The pigments in the PS, AY, and GS co-coons were extracted using EtOH, and the resulting pigment solutions were absorbed on filter papers. The capillary force drove the upward movement of the pigment molecules and EtOH (Fig. S4). Under UV light, the GS pigment showed a sky-blue fluorescence due to the presence of carotenoids46. The fluorescence intensity emitted by the AY pigment was extremely weak and could not be observed with our naked eyes47. The PS pigment exhibited a strong green fluorescence, confirming the existence of quercetin 5-O-glucosides, which are the major constituents of cocoon flavonoids45.
PS cocoons can be utilized as large fluorescent gas sensors because of their large surface-to-volume ratios. HCl vapor was adopted as an analyte because it is widely used to produce organic compounds but is corrosive and harmful to humans. The lowest concentrations (LCLo) of HCl vapors, lethal for an individual human, are 1300 and 3000 ppm for 30-min and 5-min exposures, respectively48. The RD50 (i.e., exposure concentration that decreases the respiratory rate by 50%) test is another standard method used as an indicator of the hazards of vapors. The 10-min RD50 value for mice has been reported to be 309 ppm49. Moreover, the permissible HCl-vapor exposure limit for an 8-h time-weighted average (TWA-PEL) in working spaces is 5 ppm50. Safety limits are very low compared to lethal concentrations because they are set by considering a long-time exposure over many hours. To examine the gas-sensing response of the fluorescent PS, the PS cocoons were placed in a beaker containing HCl vapors (Fig. 2a), whose concentration was controlled by modulating the evaporated content of an HCl droplet. Figure 2b shows the absorption and emission spectra of the PS cocoons before (solid) and after (dotted) exposure to HCl vapors. Due to the exposure to a strong acid (300 ppm for 2 min), the absorbance and fluorescence intensities simultaneously deteriorate, and this feature can be used for fluorescent chemosensing. The effect of the HCl vapor exposure could be recognized with the naked eye. As shown in Fig. 2c, the PS mats that were cut to express letters revealed a bright green fluorescence under UV light. After the HCl vapor exposure, the daylight color and fluorescence gradually faded as evident from Fig. S5. The GS mats also showed a similar fading upon HCl-vapor exposure (Fig. S6); however, the rates at which the fluorescence intensity and absorbance decreased were relatively lower than those of the PS mats, and a weak fluorescence was detected.
For quantitative analysis, the PS cocoon mats were exposed to different concentrations of HCl vapors (from 5 to 3000 ppm). A distinct fluorescence decay (fading) is evident even at a concentration of 5 ppm (TWA-PEL limit) as shown in Fig. 3a. The fluorescence intensity decreases by 25% for a 5-min exposure, and this intensity decay can be easily detected by a simple photodetector. These results indicate the feasibility of using PS cocoon mats as cost-effective and compact sensing systems. Further, the fluorescence decay rate increases with the increasing HCl vapor concentration. For a 5-min exposure, the fluorescence ratios decrease to 65% and 79% at 100 and 300 ppm, respectively (Fig. 3b and Fig. S7c), and at a concentration of > 1000 ppm, the fluorescence completely vanishes/fades after 5 min (Fig. 3c and Fig. S7d). Figure 3d shows that the fading rate, determined by the ratio of the relative peak intensity change and exposure time, is high in the as-exposed stage and becomes constant. At 5 ppm, the fading rate is 15.0%p/min for 30 s and 3.6%p/min afterward, whereas the rates are 40.8%p/min and 10.0%p/min at 100 ppm. Notably, the fluorescence decays linearly and vanishes completely at all the HCl vapor concentrations, although the fading rate depends on the HCl concentrations.
To investigate the versatility of fluorescence sensing, the PS cocoon mats were exposed to different environments, and temperature and EtOH content, which play a crucial role in biological and environmental research, were analyzed as the sensing parameters. The sensing performance of a material is determined by its ability to detect the decay in fluorescence under external stimuli (Fig. 4a). Especially, the fluorescent pigment found in PS cocoons is a kind of flavonoid, and thus, these cocoons can efficiently respond to temperature changes and EtOH in water. Similar to the response of the cocoons under HCl exposure, the fluorescence of the cocoons diminished with a linear decay rate with the increasing heating time and/or immersion time in the EtOH solution at different concentrations of EtOH in water. Figure 4b shows the thermally induced changes in fluorescence exhibited by the PS cocoons upon heating where the temperature was increased slowly from 20 to 40 ℃. We observed that the transition to the fluorescence de-cay started at ~ 35 ℃, and above 40 ℃, the fluorescence decay followed a linear pattern up to 200 ℃; the total fluorescence decay was ~ 75% at 200 ℃ (Fig. 4b and Fig. S8). To investigate the EtOH sensing response, the PS cocoon mats were immersed in an EtOH-water solution with different EtOH concentrations. The fluorescence intensity rapidly faded during the first 60 s, and the highest decay rate was observed at an EtOH concentration of 100%. However, as the cocoon immersion time in EtOH increased, the fluorescence decayed linearly; a decay of ~ 80% was observed in the pure EtOH case after 30 min of immersion (Fig. 4c, Fig. S9, and Fig. S10). These results collectively suggest that because of the presence of the fluorescent pigment in PS, PS cocoons can be used as sensors under various biological and environmental stimuli. These sensing abilities of PS can be attributed to the changes in the flavonoid activity induced by various agents.