Probing Keratin Fiber Aging with Terahertz Time-Domain Spectroscopy: Linking Disulfide Cleavage and Crystallinity Loss.

Bio-based protein fibers such as cashmere are valued for low weight and thermal insulation but are vulnerable to environmental aging (UV, hygrothermal, alkaline), which undermines hierarchical structure and limits service life in high-demand applications. Rapid, non-destructive assays that bridge chemical fingerprints with crystalline and supramolecular readouts are therefore needed. This study integrates Fourier-transform infrared spectroscopy (FTIR), Raman spectroscopy, X-ray diffraction (XRD), and terahertz time-domain spectroscopy (THz-TDS) to systematically characterize the multiscale aging of cashmere fibers under ultraviolet (UV) and alkaline conditions. Alkaline media preferentially and rapidly induce disulfide (S-S) bond scission, with the Raman I511/I1451 ratio dropping from ∼0.65 to 0.15 (≈77% decrease) after 72 h in NaOH, whereas UV irradiation-by virtue of deeper penetration-predominantly compromises interior crystalline domains, reducing the XRD Segal crystallinity index from 0.74 to 0.09 (≈88% decrease) after 600 h. Both pathways drive a common structural evolution characterized by decreased α-helical content, increased amide III disorder (I1245/I1451↑), and a net loss of relative crystallinity (alkaline: 0.74→0.52, ≈30% decrease). THz-TDS, via the 1.5 THz absorption, records a concomitant attenuation (from 26.74 to 5.02 under UV, ≈81% decrease; to 8.32 under alkali, ≈69% decrease), indicating that low-frequency collective modes are jointly governed by crystallinity, intermolecular coupling, and chain-segment constraints. These results show that THz-TDS effectively complements FTIR/Raman/XRD, which are less responsive to ensemble interactions, and is therefore well suited for rapid, non-destructive aging assessment and pathway discrimination. In this multimodal framework, THz-TDS serves as an ensemble-level probe that captures low-frequency collective vibrations and intermolecular coupling, thereby complementing the local chemical fingerprints from FTIR/Raman and the crystallinity information from XRD. The proposed multimodal framework provides a methodological basis for durability evaluation and potential online monitoring of protein fibers, and it lays the groundwork for future multimodal quantitative modeling.