1：Experimental Design and Method Selection

The synthesis of C-QDs was achieved via a single-step thermal decomposition process using fennel seeds. Machine learning techniques such as PCA, MCR-ALS, and NMF-ARD-SO were employed to analyze the PL data-set.

2：Sample Selection and Data Sources

Fennel seeds (Foeniculum vulgare) were used as the natural carbon source. PL data was acquired at various pH levels and excitation wavelengths.

3：List of Experimental Equipment and Materials

Mixer grinder (Tiger mixer grinder, Japan),Crucible cup (AS ONE, Japan),Heat plate (AS ONE, Japan),Dialysis kit (Float-A-Lyzer G2 Dialysis, Japan),UV-Vis absorption spectroscopy (V-530, JASCO, Japan),Photoluminescence spectroscopy (FP-6500, JASCO, Japan),Transmission electron microscopy (TEM, JEM2002-FS, JEOL),Zeta-sizer (Nano-ZS 90 Zetasizer, Malvern Instruments Ltd),Energy-dispersive X-ray spectroscopy (EDS, JEM-2300F, JEOL),Laser Raman spectroscopy (Lab RAM, HR-800, Horiba JOBIN YVON S.A.S.),Fourier transform infrared spectra (FTIR-4100, JASCO, Japan),X-ray photoelectron spectroscopy (PHI Quantes, ULVAC-PHI. Inc.),Quantum efficiency measurement system (QE-2000, Otsuka Electronics, Japan)

4：Experimental Procedures and Operational Workflow

Fennel seeds were crushed, heated at 500 °C for 3 hours, dissolved in deionized water, sonicated, centrifuged, filtered, and dialyzed to purify the C-QDs. PL measurements were taken at various pH levels and excitation wavelengths.

5：Data Analysis Methods

PCA, MCR-ALS, and NMF-ARD-SO techniques were used to analyze the PL data-set.