High-Precision Strontium Isotope Analysis and Origin Tracing of Trace-Level Natural Silk Samples

BRIEF REPORTSignificance: Provenance tracing of silk artifacts is crucial for studying the Silk Road, yet current methods lack scientific tools. Here, we develop a highly sensitive technique for Sr isotope analysis of trace silk samples, overcoming low Sr content and organic complexity. Our results reveal that cocoon 87Sr/86Sr ratios are uniform within sources but vary primarily by geography, thereby establishing Sr isotopes as a robust tool for silk provenance and advancing historical trade research.Introduction: Silk, a key commodity on the ancient Silk Road, played a pivotal role in the development of human civilization[1]. Tracing the provenance of silk artifacts is essential for reconstructing ancient trade networks. However, current provenance studies mostly rely on art historical evidence such as patterns and craftsmanship, lacking precise scientific methods[2-3]. Radiogenic strontium isotopes (87Sr/86Sr), with strong regional signatures, offer a powerful alternative[4-5]. Bioavailable Sr from weathered bedrock enters organisms through soil–water–plant systems, preserving a distinct “isotopic fingerprint”[6]. This technique has proven effective in archaeology, including tracing the origins of shipwreck timbers[7], Roman glass[8], and ancient textiles[10-11]. Expanding this approach to silk artifacts could offer new insights into historical Silk Road trade.　　However, the inherently low Sr content (~1 μg/g) and complex organic matrix of silk present significant analytical challenges[13]. Although previous studies have demonstrated regional differences in 87Sr/86Sr ratios in cocoons, they relied on high-consumption methods unsuitable for precious artifacts[12-13]. Therefore, the development of high-sensitivity techniques for trace silk samples is crucial to enable minimally destructive provenance analysis.　　In this study, refined purification techniques were coupled to eliminate organic matrix interferences with an optimized TIMS loading protocol to enhance ionization efficiency. These advancements enabled high-precision 87Sr/86Sr analysis of silk cocoons while significantly reducing sample consumption. This approach was subsequently applied to samples across five major Chinese sericultural regions to evaluate the efficacy of Sr isotopes as a robust tool for geographic provenance.Methods: Silk cocoons were collected from five major production regions in China: Hubei, Guangxi, Yunnan, Zhejiang, and Shandong Provinces. To remove sericin, the cocoons were boiled in deionized water at 100℃ for 2 h, then dried and ground into powder. Approximately 10 mg of each powdered sample was digested sequentially: first with 1 mL HNO3 and 0.5 mL H2O2 for 2 h; then with 2 mL HNO3 under sealed conditions at 120℃ for 24 h; and finally with a mixture of 1 mL HNO3, 1 mL HCl, and 1 mL deionized water at 120℃ for 12 h. The final residue was dissolved in 3 mol/L HNO3, centrifuged, and stored for chemical purification.　　Sr isotope purification was performed using Sr-specific resin (50–100 μm, Eichrom) packed into a ~0.5 mL custom-made Teflon column, which was pre-cleaned and conditioned with 3 mol/L HNO3. After sample loading, matrix elements were removed by sequential rinses with 3 mol/L HNO3, and Sr was eluted with deionized water (details in Table 1). To ensure complete organic matter removal, a second purification step was applied. The purified Sr fractions were evaporated to dryness and re-dissolved for filament loading.　　Isotopic measurements were performed at the China University of Geosciences (Wuhan) using a Triton TI thermal ionization mass spectrometer (TIMS, Thermo Finnigan), equipped with nine 1011 Ω Faraday cups. Samples were loaded onto rhenium filaments using a modified “sandwich” method with purified tungstosilicic acid as the ionization activator[14]. Ionization efficiency of the sample loading protocol was evaluated using the total evaporation method with 1 ng of NBS987. Static multi-collector mode was employed to simultaneously detect 84Sr, 85Rb, 86Sr, 87Sr, and 88Sr. Mass fractionation was corrected using the exponential law and a normalization ratio of 88Sr/86Sr = 8.375209. Isobaric interference from 87Rb was corrected using the natural isotopic ratio 87Rb/85Rb = 0.3856. To monitor analytical stability, the NBS987 Sr reference materials (0.5–1 ng each) were loaded three times along with samples in every batch of measurements. The Sr loading blank and total procedural blank, as determined by an 84Sr spike, were 2–3 pg and ≤35 pg, respectively.Data and Results: (1) Residual organic matter remained in silk samples after treatment with nitric acid and hydrogen peroxide but was effectively removed by the following two-stage Sr Spec resin column separation. The presence of organic components did not influence the adsorption behavior of Sr Spec resin (Fig. 1). The total Sr recovery of the two-stage columns was as high as 99%, thus preventing any potential Sr isotope fractionation resulting from Sr losses during the column separation process. Furthermore, the total blank throughout the entire procedure was no more than 35 pg for Sr, which is negligible compared to the sample size. (2) Compared to the traditional TaF5-based method (4%–5% ionization efficiency), a modified “sandwich” loading technique using 4 μL of tungstosilicic acid and phosphoric acid achieved up to ~10% ionization efficiency—nearly double that of TaF5. What’s more, ionization efficiency was highly sensitive to tungstosilicic acid volume but not to TaF5 quantity (Fig. 2a). A negative correlation between ion beam intensity and efficiency was observed (Fig. 2b), likely due to premature Sr evaporation at high filament temperatures. This loading protocol significantly enhanced efficiency while maintaining low blanks (2–3 pg), providing a robust approach for high-precision Sr isotopic analysis of trace level silk samples. (3) The novel loading protocol enabled TIMS to accurately measure NBS987 standards down to 0.5 ng (Fig. 3). This method achieved 0.000033 (2SD) precision for as low as 10 mg cocoon samples, which reduced sample consumption by 1–2 orders of magnitude compared to conventional approaches. (4) High-precision Sr isotope analysis revealed remarkable isotopic homogeneity within individual cocoons (Table 2). Furthermore, cocoons from the same geographical origins and genetic varieties displayed highly consistent Sr isotope compositions. However, significant differences were observed between different genetic varieties and regions (Fig. 4), with geographical factors playing a dominant role in Sr isotope variation.