Dynamic Electrochemical Reorientation at SERS Hotspots Enables Composition, Length, and Sequence Reading of DNA Oligonucleotides

Extending surface-enhanced Raman spectroscopy (SERS) from small molecules to macromolecules is important for various applications, yet it remains challenging due to the "surface selection rule". Molecular segments that bind closer to the substrate surface experience stronger signal enhancement and thus dominate the spectrum, resulting in an incomplete or biased SERS response that does not represent the overall molecular structure. DNA oligonucleotides exemplify this challenge, where SERS struggles to extract sequence-level information such as length and primary sequence due to relevant vibrational features being overshadowed by dominant peaks and noise, restricting complete molecular characterization. In this work, we introduce an electrochemical modulation strategy that dynamically reorients oligonucleotides near plasmonic hotspots to enrich information for SERS profiling, achieving comprehensive decoding of oligonucleotide's composition, length, and sequence. By tuning the surface potential of the substrate, we force adsorbed DNA oligonucleotides through distinct configurations and sequentially bring different types and numbers of bases into hotspots. Combining these configuration-specific spectra realizes SERS superprofiles that enrich vibrational information, showcasing a good resolution of 98.4% classification accuracy for 44 oligonucleotides. Importantly, these superprofiles exhibit chemistry-grounded relationships with key sequence characteristics, where integration with a stepwise machine learning (ML) framework enables accurate prediction of unseen oligonucleotides with an average 3.4% difference for base composition, 0.9 bases difference for length, and 100% accuracy for primary sequence. Overall, our electrochemical reorientation strategy overcomes the orientation- and surface-selection biases in static SERS for reliable structural prediction of unknown DNA oligonucleotides. We envisage expanding these dynamic SERS profiles to a vast range of analytes and substrates with diverse chemical functionalities and surface potentials, realizing an extensive space of analyte-surface configurations for future molecular characterization across various fields.