PhD Dissertation: Sümeyra Vural Kaymaz, FIELD DEPLOYABLE RAMAN SPECTROSCOPY-BASED PESTICIDE DETECTION FROM AGRICULTURAL PRODUCE, Date & Time: July 09, 2026 – 5:00 PM, Place: FENS G035
FIELD DEPLOYABLE RAMAN SPECTROSCOPY-BASED
PESTICIDE DETECTION FROM AGRICULTURAL PRODUCE
Sümeyra Vural Kaymaz
Molecular Biology, Genetics and Bioengineering, PhD Dissertation, 2026
Thesis Jury
Assoc. Prof. Meral Yüce (Thesis Supervisor)
Assoc. Prof. Serkan Ateş
Asst. Prof. Stuart James Lucas
Prof. Sedat Nizamoğlu
Prof. Alpan Bek
Date & Time: 9th July, 2026 – 5:00 PM
Place: FENS G035
Keywords : PERS, MIM Metasurfaces, Pesticide Detection, Machine Learning,
Food Safety Monitoring
Abstract
Agricultural pesticide residues pose significant challenges to food safety and environmental monitoring, creating a growing demand for analytical technologies capable of rapid, sensitive, and on-site detection. Raman spectroscopy provides molecularly specific fingerprint information without labeling or extensive sample preparation; however, its practical application is limited by inherently weak scattering efficiency. To overcome this limitation, this thesis investigates metal-insulator-metal (MIM) plasmonic metasurfaces as engineered platforms for plasmon-enhanced Raman spectroscopy (PERS). Through electromagnetic simulations, nanofabrication, and experimental characterization, MIM metasurfaces were designed to support localized plasmonic resonances and enhanced light-matter interactions. The optimized architectures generated confined electromagnetic fields, enabling molecular detection with enhancement factors approaching 107 and detection limits down to the femtomolar level. Furthermore, the lithographically fabricated metasurfaces demonstrated structural reproducibility and optical consistency, addressing key challenges associated with practical plasmonic sensing applications. The developed PERS platform was applied to the detection of pesticide residues in complex food matrices. Characteristic molecular fingerprints of various fungicides, insecticides, and plant growth regulators were successfully identified and quantified at concentration levels approaching or below current regulatory limits. To facilitate spectral interpretation, preprocessing methods were integrated with machine learning algorithms for classification, clustering, and concentration-level discrimination, enabling the analysis of both laboratory-prepared and authentic food samples. Overall, this work demonstrates the potential of combining engineered MIM plasmonic metasurfaces with data-driven spectral analysis for sensitive molecular detection in complex samples, contributing to the development of portable Raman-based sensing technologies for food safety and environmental monitoring.