Infrared absorption spectroscopy, while powerful for label-free molecular identification, suffers from inherently weak molecular vibrational absorption. Nanophotonic structures offer a solution by amplifying absorption via enhanced near-field intensity in localized "hot spots." However, conventional analyte delivery methods (e.g., spin coating, droplets) distribute molecules unevenly, limiting the fraction interacting with hot spots and thus reducing sensitivity. Existing approaches, including microfluidics and active trapping mechanisms (dielectrophoresis, optical trapping), have limitations, particularly for small molecules. This research introduces a novel nanophotonic sensor design that passively concentrates and traps analyte molecules in hot spots during solvent evaporation, addressing this critical limitation.

Various nanophotonic structures (nanorods, split-ring resonators, colloidal nanoparticles, metal-insulator-metal structures) have been explored for surface-enhanced infrared absorption (SEIRA) sensing. High field confinement is achieved using nanometric gaps, enabling detection of a few hundred molecules. However, the challenge remains in delivering a significant portion of the analyte molecules to these small hot spots. Techniques like self-assembly, spin coating, and biomolecule immobilization distribute molecules across the sensor surface inefficiently. Micro- and nanofluidic approaches can guide solutions, but lack concentration mechanisms. Active trapping methods require external energy and are unsuitable for small molecules. Super-hydrophobic surfaces have shown promise for confining large biomolecules but haven't achieved targeted delivery to hot spots.

The proposed sensor employs a metal-insulator-metal (MIM) structure with periodic arrays of aluminum (Al) ribbons on top of germanium (Ge) ribbons, backed by a gold (Au) reflector. Nano-trenches flanking each Ge ribbon create hot spots where the electric field is highly enhanced (simulated enhancement > 40). Finite-difference time-domain (FDTD) simulations were used to optimize the Al ribbon width (w), nano-trench width (L), and Ge ribbon height (d) for targeted spectral responses to specific molecular vibrational modes. The fabrication process involves a single lithography and dry etching step, simpler than methods using nanometric gaps or super-hydrophobic surfaces. The passive trapping mechanism relies on the evaporation of a solution droplet placed on the sensor surface. Surface tension and the nano-trench geometry retain the solution, leading to analyte molecule precipitation within the hot spots. The sensor's performance was evaluated experimentally using L-proline, D-glucose, and sodium chloride solutions. Optical microscopy and SEM imaging were used to verify the device structure and analyte distribution. FDTD simulations explored the effects of varying w and L on reflection spectra and sensitivity, showing stronger responses with wider nano-trenches and analyte located on the outer edges of the trenches.

FDTD simulations demonstrated a wide spectral tuning range by adjusting geometric parameters (w, L, d). The simulations indicated that placing analyte molecules on the outer side of the nano-trenches yielded stronger spectral responses compared to the inner side. Experimental results confirmed the passive trapping functionality. Optical microscopy images showed analyte precipitation predominantly concentrated near the Al ribbon edges, consistent with the nano-trench trapping. The sensor achieved a reflection change of a few percentage points in response to picogram-level masses of proline and glucose, significantly less than the mass of a molecular monolayer covering the same area. The fabrication process was shown to be robust and reproducible, using standard nanofabrication techniques.

The results demonstrate that the proposed nanophotonic sensor design significantly improves sensitivity by actively concentrating analyte molecules in hot spots. The passive trapping mechanism eliminates the need for external energy sources and complex microfluidic systems, offering a simple, robust, and highly sensitive platform. The picogram-level detection achieved significantly surpasses the sensitivity of conventional SEIRA sensors. The observed enhanced sensitivity is attributed to the efficient delivery and concentration of analyte molecules within the high-field regions of the nano-trenches. This approach holds potential for various sensing applications involving small molecules, nanoparticles, and biomolecules.

This work presents a novel nanophotonic sensor design incorporating passive analyte molecule trapping in hot spots. The design's simplicity, enhanced sensitivity (picogram-level detection), and robustness highlight its potential for diverse sensing applications. Future work could explore the application of this design to a broader range of analytes and investigate the integration of this sensor with microfluidic devices for automated sample handling and analysis.

The current study primarily focuses on small molecule detection. Further investigation is needed to optimize the design for larger biomolecules and nanoparticles. While the fabrication process is relatively simple, scalability and high-throughput manufacturing need further consideration for widespread implementation. The sensitivity might be affected by variations in the evaporation process, potentially leading to inconsistencies in analyte distribution.