How strong can light get in a gap one atom-layer wide?
Light enhancement in gold nanogaps rises as the gap shrinks, but at a single molybdenum-disulfide layer it falls well short of classical predictions, pointing to quantum electron tunneling.
Source
Probing the limits of plasmonic enhancement using a two-dimensional atomic crystal probe
Study at a glance
- Design
- Other — Lab experiment on gold nanoparticle-on-mirror antennas with 1-3 layer MoS2 gaps; plasmon resonance tuned by successive Al2O3 coatings while measuring dark-field scattering and SERS, compared with classical electromagnetic models.
- N
- Individual antennas: 17 monolayer, six bilayer and four trilayer MoS2 nanoparticle-on-mirror structures.
- Population
- 50 nm gold nanoparticles on ultrasmooth gold film separated by one to three layers of MoS2
- Outcome
- Surface-averaged vertical and horizontal SERS enhancement factors and derived local field enhancement versus gap distance
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The team placed gold nanoparticles on a flat gold film with one, two or three layers of MoS2 in between, giving gaps of 0.62, 1.24 and 1.86 nm. The MoS2 itself served as the Raman probe: its out-of-plane vibration reports the vertical field and its in-plane vibration the horizontal field. They added aluminium-oxide layers step by step to shift the plasmon resonance across the 785 nm laser line, measuring dark-field scattering and Raman enhancement after each step, and compared the maxima with classical electromagnetic simulations.
What they found
Vertical enhancement dominated: in monolayer devices the averaged enhancement factor peaked around 1.5 × 10^8, roughly 2400 times the horizontal one, when the resonance sat between the laser and Raman wavelengths. Maximum enhancement grew as the gap narrowed, and classical models matched the measurements at 1.24 and 1.86 nm. At 0.62 nm the measured field enhancement of about 114 was 38.4% below the classical prediction of about 185, consistent with electron tunneling. Signals were stable, fluctuating 4.3% over 150 minutes.
The limits
What it doesn't show
Gap size can only change in steps of one MoS2 layer, so the exact onset of quantum effects between 0.62 and 1.24 nm is not resolved. The tunneling explanation is inferred from disagreement with classical models; a quantum-corrected model was not fitted because the gold-doped MoS2's tunneling conductivity is unknown. Sample numbers per gap are small (as few as four antennas), and horizontal enhancements exceeded calculations, possibly due to ripples in the MoS2.
Key terms
- Nanoparticle-on-mirror (NPOM)
- A metal nanoparticle sitting a tiny distance above a flat metal film, which acts like a dimer with its mirror image and traps light in the gap.
- SERS enhancement factor
- How much stronger Raman scattering is near the nanostructure than in a reference, roughly the fourth power of the local field enhancement.
- Plasmon-scanned SERS
- Tuning the antenna's resonance (here by dielectric coatings) across a fixed laser wavelength instead of tuning the laser.
- Electron tunneling
- Quantum leakage of electrons across a very thin barrier, which short-circuits the gap and limits field enhancement.
- A1g and E2g phonons
- Out-of-plane and in-plane lattice vibrations of MoS2 that respond to vertical and horizontal fields respectively.
Flashcards
0 of 10 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What sets the gap size in these antennas?
Common questions
Why use MoS2 instead of dye molecules as the probe?
MoS2 fills the whole gap uniformly with a known thickness and orientation, so the probe area and alignment are defined, which molecules cannot guarantee.
Why doesn't a smaller gap always give stronger fields?
At sub-nanometre gaps electrons can tunnel across and screening becomes nonlocal, which reduces the charge build-up that creates the strong field.
Why add aluminium oxide layers?
The high-index coating redshifts the plasmon so it can be matched to the fixed 785 nm laser and the outgoing Raman light.
More on Plasmonics