Defect engineered 2D MoS2 materials for ML-enabled SERS detection of neurotransmitters.

Mechanisms of neurotransmitter detection using a two-dimensional MoS2 platform are discussed. Machine learning methods are applied to distinguish the spectra of structurally similar analytes. The attachment of catechol-containing neurotransmitter molecules to a defect-engineered MoS2 is demonstrated, leading to activation of SERS through molecular charge transfer. The SERS effect and selective docking of biomolecules are boosted through introducing sulfur vacancies in MoS2 monolayer films via soft plasma etching. The quality of the defect-engineered material was controlled by Raman, photoluminescence, AFM, and XPS characterization, thus allowing for optimization of the process of defect formation and achieving sensing selectivity. Defect-engineered MoS2 has enabled SERS detection of dopamine and epinephrine down to the sub-nanomolar range (5 × 10-10 M), with strong calibration reliability (R2 = 0.95 and 0.99 for pure samples). The sensor material showed no response to serotonin, confirming the specificity of attachment due to S vacancies, which enables catechol-specific molecular adsorption. PCA-LDA achieved 100% accuracy in distinguishing dopamine and epinephrine, establishing defect-engineered MoS2 as a tunable, low-cost SERS platform for future sensing applications.