Optical sensing systems often need to distinguish particular wavelengths or respond preferentially to light arriving from a chosen direction. A nanoantenna photodetector offers one way to build these functions into the detector structure itself. In a reported device, an array of 400 Yagi-Uda nanoantennas was integrated with a metal-semiconductor-metal (MSM) rectifying photodetector on silicon. Electrical measurements showed wavelength-dependent photocurrent peaks and a response consistent with the antenna array’s directionality.
The results demonstrate how concepts from radio-frequency antenna design can inform optical devices, while also showing why nanoscale designs must account for plasmonic effects, materials, and the detector’s electrical structure.
How a Yagi-Uda nanoantenna photodetector works
A conventional Yagi-Uda antenna uses elements with different roles: a feed element, a reflector, and directors. In the receiving device described here, the feed element is also the detector element. The nanoscale rods are arranged to couple incident infrared light into the antenna structure, concentrating electromagnetic energy around the metal elements.
At optical frequencies, this coupling involves surface plasmons: collective oscillations of electrons at a metal surface. When the antenna is illuminated at a suitable wavelength, the excited plasmons can lose energy through several pathways, including reradiation, transfer to phonons, or the generation of energetic, or “hot,” electrons. Some hot electrons can cross the metal-silicon Schottky barrier and enter the semiconductor’s conduction band. This charge transfer contributes to the measured photocurrent.
The detector combines the nanoantenna metal with silicon and two larger gold counter contacts to form an MSM structure. Its two junctions rectify the optical-frequency response into a photocurrent that can be measured at low frequency. This makes the electrical output a direct way to assess the integrated antenna-detector system, rather than relying only on optical transmission, reflection, or absorption measurements.
Electrical measurements reveal wavelength selectivity
The device was illuminated with monochromatic light across a free-space wavelength range of 1,000 to 2,000 nm. Measurements compared the antenna array with a reference device that lacked the array. At near-glancing incidence, the array device showed photocurrent resonances at 1,110 nm and 1,690 nm. The blank reference did not show corresponding resonances.
The angular measurements also indicated directional behavior. At near-glancing incidence, the incoming light’s direction was expected to align more closely with the Yagi-Uda antenna’s main lobe. At 45-degree incidence, clear resonance peaks were not observed. Average photocurrent was higher at 45 degrees, however, showing that the total photocurrent level and the presence of distinct antenna resonances are different aspects of the response.
Estimated quantum efficiencies for the array device were 5.1% at 1,110 nm and 3.1% at 1,690 nm. For the reference device, the estimates were 1.2% and 0.64%, respectively. These results represent an approximately fourfold increase over the reference at the reported wavelengths. They apply to the tested prototype and conditions; they should not be treated as a general performance guarantee for other detector designs.
Why nanoscale antenna dimensions do not scale directly
At radio frequencies, antenna dimensions can often be scaled in proportion to the free-space wavelength. That simple relationship becomes less reliable as dimensions shrink to the nanoscale. The metal’s optical behavior and plasmonic effects alter the wavelength experienced by the antenna, so the effective wavelength differs from the wavelength of the incoming light in free space.
For the reported device, the two free-space resonance wavelengths corresponded to effective wavelengths of 388 nm and 776 nm. These values matched full-wavelength and half-wavelength operation, respectively, of the detector element. The comparison supports a scaling model that accounts for the dielectric environment and capacitive effects associated with the nanorod ends.
This has a practical design implication: selecting rod lengths from free-space wavelength alone may not predict the resonant response. The substrate, surrounding medium, and element geometry all influence the effective wavelength. In the device, the gold rods sat on silicon and were exposed to air, creating a nonuniform dielectric environment. The rod-end geometry also affected capacitive coupling.
Design considerations for integrated optical sensing
Embedding an antenna array in a rectifying detector can combine optical selection with electrical signal extraction. The direct electrical measurement approach is useful because it evaluates photocurrent generated by the complete structure. It also makes clear that detector behavior depends on more than antenna resonance alone: charge injection across the Schottky barrier and transport through the biased MSM device are part of the signal-generation process.
The reported prototype also points to factors that can affect efficiency and response across an array:
- Array layout: Antennas can shadow elements behind them, so their positions may influence how much each contributes to the photocurrent.
- Electrical field distribution: Band bending is not uniform across the biased device, which can change where generated carriers contribute most effectively.
- Bias and carrier transport: The likelihood that hot electrons reach a contact before losing energy is relevant to signal collection.
- Illumination conditions: Yagi-Uda response depends on direction and polarization, so source orientation and alignment matter.
These are design considerations rather than demonstrated improvements in the tested device. The work identifies them as potential routes for further optimization, alongside investigation of polarization sensitivity and operation across a wider range of infrared and visible wavelengths.
Conclusion: integrating optical function with detection
The Yagi-Uda array shows that nanoscale antenna geometry can impart wavelength selectivity and directional response to an MSM photodetector, while producing a measurable electrical output. Its resonances depend on plasmonic scaling and the device’s dielectric and geometric details, not simply on conventional antenna dimensions. For engineers assessing optical sensing architectures, the central lesson is to design the antenna, semiconductor interface, and carrier-collection path as one system. The reported results establish a prototype approach; further optimization would be needed to determine its suitability for a particular sensing application.