Deep-ultraviolet plasmonic molybdenum nanoholes for adenine detection via deep-learning-enhanced Raman scattering spectroscopy

Deep ultraviolet surface-enhanced Raman scattering (DUV-SERS) technology demonstrates its potential for biomolecule detection. Materials exhibiting localized surface plasmon resonance (LSPR) effects generate intense electromagnetic field enhancements, which are critical for improving DUV-SERS detection sensitivity. However, the chemical stability of SERS substrates in complex biological environments remains a substantial challenge in practical applications. In this work, we have developed an LSPR-enhanced DUV-SERS substrate material utilizing molybdenum (Mo) random nanoholes fabricated on monocrystalline films, providing an efficient and anti-corrosion solution for ultrasensitive molecular detection in complex biological environments. Under 325 nm DUV laser excitation, the Mo random nanoholes substrate achieves a Raman enhancement factor (EF) of approximately 10 7 for adenine detection and a detection limit as low as 10 −10 mol/L. Finite-difference time-domain (FDTD) simulations further confirm the Mo’s significant localized field enhancement effect, providing theoretical support for the high EF value. Additionally, a deep-learning-based spectral analysis approach was implemented to achieve a classification accuracy of 92.89% for adenine as a representative of biological molecules. The sensitive and robust Mo nanoholes material offers a cost-effective, high-reliability platform for detection of biomolecules, promoting the future development of portable biosensing and rapid molecular recognition devices.