Raman enhancement induced by exciton hybridization in molecules and 2D materials

The work addresses how electronic coupling between molecules and 2D materials governs Raman enhancement in semiconductor-based SERS. Unlike noble-metal SERS where electromagnetic plasmonic effects dominate, 2D materials enhance Raman signals mainly via chemical enhancement through electronic coupling (charge transfer, molecular absorption, or exciton resonance). The study investigates whether aligning molecular excitons (µmol) with 2D material excitons (µ2D mat) can produce co-resonance and exciton hybridization that boosts Raman intensities. SnS2 (Hexc ~2.3 eV) and Rhodamine 6G (Rh 6G; S0,0–S1,0 ~2.33 eV) provide a model system with nearly degenerate excitons, enabling a systematic test of exciton alignment effects on Raman enhancement and detection sensitivity.

Since the discovery of graphene-enhanced Raman scattering (GERS) in 2010, Raman enhancement has been reported on multiple 2D materials (graphene, h-BN, black phosphorus, TMDs, MXenes). The Lombardi–Birke theory predicts strong enhancement when excitation coincides with charge-transfer (µCT), molecular absorption (µmol), or exciton resonance (Hex), and potentially greater enhancement under multiple co-resonances. Most prior studies on 2D semiconductors emphasize charge-transfer resonances; fewer examine dual resonances or explicit alignment between molecular and substrate excitons. Some have hinted that aligning exciton resonances with charge transfer can help, and exciton hybridization has been proposed (e.g., PTCDA on WSe2). However, a systematic exploration of matching molecular and substrate excitons to induce hybridization and enhance Raman signals has been lacking, motivating this study.

- Materials and substrates: Bulk crystals of SnS2, MoS2, WSe2, and graphene (HQ Graphene) were mechanically exfoliated and transferred to cleaned 300 nm SiO2/Si substrates. Layer number of SnS2 was determined by AFM (Bruker Dimension 3000, tapping mode). - Probe molecules: Rhodamine 6G (Rh 6G), Rhodamine B (Rh B), and Rhodamine 123 (Rh 123) (Sigma-Aldrich) were dissolved in isopropyl alcohol (IPA) to form 10−4 M stock solutions. Exfoliated 2D crystals were submerged in solution for 2 h, then rinsed with IPA. For concentration-dependent studies, Rh 6G was serially diluted from 10−4 to 10−14 M. Samples were soaked 2 h in the lowest concentration, measured, then sequentially exposed to higher concentrations. - Optical spectroscopy: Micro-Raman (Horiba-JY T64000, triple-grating, 1800 g/mm) with ×100 objective; laser power <1 mW; 60 s acquisition. Excitation-dependent Raman used 11 laser lines from a Kr+/Ar+ ion laser (458–647 nm; 2.70–1.91 eV). Spectra were baseline-corrected where needed to remove photoluminescence. - Raman excitation profiles (REP): For Rh 6G/SnS2 and controls (Rh 6G on MoS2, WSe2, graphene; Rh B and Rh 123 on SnS2), REPs of selected vibrational modes were built from intensities versus excitation energy; calibrated to a quartz reference and normalized. - Absorption spectroscopy: UV–vis absorption of Rh 6G in IPA and in SnS2 colloid dispersions (Agilent CARY 5000) at varying concentrations (30–377 nM) to probe interactions (e.g., isosbestic points). Micro-absorption of dried SnS2 flakes before/after Rh 6G deposition (reflectance/transmittance) was also performed. - Band alignment considerations: Literature values for HOMO/LUMO of Rh 6G and band edges (CBM/VBM) of SnS2 were used to assess possible charge-transfer resonance energies relative to the excitation range. - Thickness dependence: Raman of Rh 6G on SnS2 with different layer counts (e.g., 3L, 11L, 18L) assessed enhancement versus thickness.

- Strong SERS with SnS2: Rh 6G on few-layer SnS2 shows prominent Raman bands, including an unusually intense 613 cm−1 C–C in-plane bending mode under resonant excitation (2.33 eV), consistent with strong vibronic (Herzberg–Teller) coupling. - Trace-level detection: Limit of detection (LOD) for Rh 6G on SnS2 reaches 10−13 M; at this concentration, the 613 cm−1 mode remains observable. This LOD is on par with state-of-the-art 2D SERS reports and comparable to plasmonic SERS. - Thickness dependence: The 613 cm−1 intensity on a 3-layer SnS2 sample is about 10× stronger than on an 18-layer sample, reflecting thickness-dependent enhancement due to optical absorption and band structure effects. - Absorption evidence of interaction: In Rh 6G/SnS2 mixtures, isosbestic points at 511 and 547 nm indicate interaction between Rh 6G and SnS2, though micro-absorption difference spectra of dried flakes appeared featureless. - REPs reveal dual resonances: For Rh 6G/SnS2, all monitored modes exhibit a primary resonance R1 near ~2.35 eV (matching Rh 6G S0,0–S1,0 at ~2.33 eV) and a secondary resonance R2 at 2.54 eV. Bending modes (613, 774 cm−1) peak at 2.33 eV; xanthene ring modes (>1350 cm−1) peak at ~2.38 eV, indicating vibronic character. - Substrate A1g REP change: The SnS2 A1g REP shows a single maximum at ~2.38 eV before Rh 6G deposition, but gains enhanced activity near ~2.54 eV after Rh 6G deposition, coincident with R2. - R2 origin analysis: • Scattered light resonance is unlikely: It would be mode-dependent (predicted at 2.41 eV for 613 cm−1; 2.50 eV for 1361 cm−1; 2.53 eV for 1650 cm−1), whereas R2 appears at the same 2.54 eV for all modes. • Photo-induced charge transfer is unlikely within the tested window: Expected CT transitions are from SnS2 VB → Rh 6G LUMO at ~3.8 eV and Rh 6G HOMO → SnS2 CB at ~0.7 eV, outside the 1.91–2.70 eV excitation range. • Electronic hybridization (Morton–Jensen) with MoS2 or WSe2 does not yield an R2: Rh 6G on MoS2 and WSe2 shows REPs peaking at the molecular S0,0–S1,0 but no 2.54 eV feature. - Control substrates: • Rh 6G/graphene: REPs show a main peak at ~2.38 eV and a shoulder at ~2.47 eV (matching Rh 6G S0,0–S1,1 at 2.44 eV), not at 2.54 eV. The ~0.1 eV blueshift of R2 (2.54 eV) in Rh 6G/SnS2 relative to this shoulder supports exciton hybridization. - Control molecules (excitonic misalignment): • Rh B/SnS2: Absorption S0,0–S1,0 at 2.24 eV and S0,0–S1,1 at 2.34 eV. REP follows absorption; no R2 at ~2.5 eV. • Rh 123/SnS2: Absorption S0,0–S1,0 at 2.41 eV and S0,0–S1,1 at 2.54 eV. REP maxima match these transitions; no additional shift or R2 beyond molecular features. - Conclusion from REPs: R2 at 2.54 eV in Rh 6G/SnS2 arises from hybridization of degenerate excitons (molecule and substrate), creating new excitonic states and enhancing Raman scattering, particularly enabling strong mode-selective enhancement (e.g., 613 cm−1) and ultralow LOD.

The central question was whether aligning molecular and substrate excitons in a molecule/2D-material complex can induce exciton hybridization that enhances Raman scattering beyond single-resonance conditions. The excitation-dependent Raman data and REPs demonstrate that when Rh 6G (µmol ~2.33 eV) is adsorbed on SnS2 (Hexc ~2.3 eV), a new resonance feature (R2 at 2.54 eV) emerges for both the molecule and the substrate (SnS2 A1g), consistent with hybridized exciton formation. Systematic controls rule out scattered light resonance and direct photo-induced charge transfer as the origin of R2. Additional controls with Rh 6G on MoS2, WSe2, and graphene, and with Rh B or Rh 123 on SnS2 (where excitons are not degenerate with SnS2) do not show the R2 feature. Thus, precise exciton energy alignment is pivotal for hybridization. This hybridization enhances vibronic coupling, explaining the unusually strong 613 cm−1 mode under resonance and underpinning the achieved 10−13 M LOD. The findings highlight a chemical enhancement route based on exciton co-resonance and hybridization, offering a rational design principle for semiconductor SERS substrates: engineer band structures and molecular choices to align excitonic energies and maximize coupling for sensitivity and selectivity.

This work provides a systematic demonstration that exciton degeneracy between molecules and 2D semiconductors can induce exciton hybridization, producing a distinct REP resonance (R2 ~2.54 eV) and enabling strong Raman enhancement. Using Rh 6G/SnS2 as a model system with nearly degenerate excitons, the study achieves a 10−13 M LOD and reveals mode-selective enhancement consistent with enhanced vibronic coupling. Control experiments (alternative 2D substrates and rhodamine analogs) confirm that proper exciton alignment is required for the hybridization effect. The results establish exciton hybridization as a powerful mechanism for chemical enhancement in 2D SERS and suggest practical strategies for substrate design: tune the 2D material electronic structure (defects, strain, heterostructures) and select molecules to align excitonic transitions, thereby boosting sensitivity to plasmon-free, trace-level detection.

- Excitation energy window limitation: The laser energies (1.91–2.70 eV) do not directly probe predicted charge-transfer resonances (~0.7 eV and ~3.8 eV for Rh 6G/SnS2), limiting conclusions about CT contributions under other excitations. - Generality across analytes: Ultralow LOD is demonstrated for Rh 6G; while controls with Rh B and Rh 123 support the alignment principle, broader validation across diverse molecular classes is not shown here. - Substrate scope: Exciton hybridization is evidenced specifically for SnS2 with Rh 6G under near-degenerate conditions; other material systems may require careful engineering (defects, strain, heterostructures) to achieve similar alignment. - Absorption signatures: Conventional absorption spectra of dried flakes did not reveal distinct hybrid features, necessitating reliance on REPs to detect coupling.