Alzheimer's disease (AD) is a neurodegenerative disease associated with aging, and the number of people affected is rapidly increasing. Abnormally hyperphosphorylated tau filaments and extracellular deposits of amyloid β-peptides (Aβ) fibrils are two important pathological hallmarks of AD. Currently, stopping the production of Aβ and blocking its aggregation is the main strategy for the treatment of AD. Turmeric is effective in treating neurodegenerative diseases, but there is no effective way to identify active compounds from their complicated chemical compositions. Instead of using conventional extraction and separation methods with low efficiency and time-consuming, our group tried to use atomic materials in high-throughput chemical screening due to their structural characteristics and the unique advantages of surface atomic. Herein, a novel atomic zinc sites with hierarchical porous carbon (Zn-HPC) was synthesized to quickly screen potential inhibitors of Aβ aggregation in turmeric. As-combined Aβ@Zn-HPC demonstrates superior storage stability and high selectivity, outperforming the most reported supporters for ligand fishing. Five compounds with strong affinity on Aβ@Zn-HPC were selected by high-performance liquid chromatography-hybrid linear ion trap/orbitrap mass spectrometer after incubation with turmeric extract. Finally, it was shown that curcumin and bisdemethoxycurcumin can inhibit Aβ aggregation by using thioflavin-T fluorescence assay and biolayer interferometry. A new application for the accurate identification of Aβ aggregation inhibitors from turmeric were developed based on the active compounds possessing binding affinity to Aβ to inhibit its aggregation. The developed method could provide a promising tool for efficient drug discovery from natural product resources.