Modern nano-optics and nanophotonics depend heavily on precise nanostructured light field formation and characterization. While significant progress has been made in high-precision light scattering detection (reaching 1 nm resolution), direct subwavelength light mapping without relying on scattering, fluorescence, or non-linear conversion remains limited. This research addresses this gap by proposing a novel technique for direct, non-destructive conversion of light to electrical signals with precise nanoscale imaging. The method is crucial for both device operation and understanding underlying physical mechanisms in nanophotonics. Existing techniques, like near-field optical microscopy, suffer from limitations such as the six-power scaling of scattered power with probe size, limiting practical resolution to around 50 nm at optical wavelengths. Furthermore, these methods are often invasive and introduce image distortions. Electron microscopy techniques such as EELS and CL microscopy offer sub-nanometer resolution but provide indirect access to optical information requiring complex data processing. Indirect methods based on scattered fields can also achieve high resolution but are often invasive and rely on assumptions to translate measurements into light field maps. This paper introduces a fundamentally different approach, aiming for direct and non-invasive nanoscale light-field imaging with significantly improved resolution.

The paper reviews existing methods for nanoscale light field imaging, highlighting their limitations. Near-field optical microscopy, while offering improved resolution compared to diffraction-limited techniques, suffers from low signal strength due to the six-power scaling of scattered power with probe size, resulting in practical resolution limits around 50 nm at optical wavelengths. The technique is also invasive, as the probe interacts with the light field being measured. Electron microscopy techniques such as electron energy-loss spectroscopy (EELS) and cathodoluminescence (CL) microscopy offer extremely high spatial resolution, down to the sub-nanometer scale, but provide indirect information about the optical field, requiring elaborate data processing. Indirect methods such as mapping Brownian motion of dye particles, surface-enhanced Raman scattering, and backscattered light with scattering scanning near-field optical microscopy (s-SNOM) have also shown high spatial resolution, but they are often invasive and rely on assumptions for accurate light field reconstruction. The authors contrast these methods with their proposed technique, which is designed to be direct, non-invasive, and high-resolution.

The proposed method utilizes the photoelectric effect in a p-n junction induced in graphene. A graphene sheet is positioned in the region of interest, connected to source and drain contacts. A gate dielectric is deposited on top, with two parallel metallic gates creating a controllable p-n junction. The p-n junction position is precisely controlled by an external gate voltage, allowing for scanning across the area of interest. The generated photocurrent is proportional to the light intensity at the p-n junction position. The authors carefully consider the contribution of thermoelectric effects and show that by keeping the source and drain at the same temperature and using the same material, the thermoelectric contribution can be minimized. The technique's spatial resolution is determined by the p-n junction width which is about ~20nm. This resolution can be further improved by reducing the thickness of the dielectric separator or lowering the temperature. For experimental demonstration, the researchers mapped the electric field profile of a gap surface plasmon (GSP) mode in a plasmonic slot waveguide operating at telecom wavelengths. The device was fabricated using electron beam lithography, and a combination of optical and electrical measurements is done to characterize the GSP mode. The graphene's gating characteristics were determined by applying the same gate voltage to both waveguide sides. The interaction of graphene with the GSP mode was verified by observing the modulation of the output light intensity upon gating. The main measurements involved applying a gate voltage to one contact while grounding the other, inducing a p-n junction. The photovoltage was then measured as a function of the gate voltage to reconstruct the electric field profile. The authors conducted Finite-Difference Time-Domain (FDTD) and Finite-Element Method (FEM) simulations to model light propagation and field distribution in the device for comparison with experimental results. The fabrication process involves wet transfer of graphene flakes onto a silicon substrate, etching to define graphene stripes, deposition of a hafnium oxide dielectric layer, and finally, fabrication of the silver waveguide structure and contacts using electron beam lithography and lift-off.

The researchers successfully demonstrated nanoscale light field mapping with a spatial resolution of approximately 20 nm, a significant improvement over existing methods. The technique directly measures the electric field of a confined optical mode without relying on light scattering or other indirect effects. The measured electric field profile of a plasmonic slot waveguide mode at telecom wavelengths shows excellent agreement with numerical simulations using both FDTD and FEM. The p-n junction in graphene effectively converts the light field into an electrical signal, and its position can be precisely controlled via gate voltage. The method proved to be non-invasive; the graphene's absorption is negligible, and it does not significantly perturb the electromagnetic field distribution. The experimental setup included a focused 1550-nm laser beam, an InGaAs short-wave infrared (SWIR) camera to monitor input and output light, and lock-in detection to measure the photocurrent induced in the graphene. Measurements of photovoltage and photocurrent as functions of gate voltage and light power were conducted, showing good linearity, confirming the photoelectric effect as the dominant mechanism. The spatial resolution, extracted from comparing the measured field profile to the calculated distribution, was approximately 22 nm (FWHM). The device also exhibited good electro-optical modulation characteristics, with a modulation depth of 0.12 dB/µm at a gate voltage amplitude of 6V. The maximum photocurrent produced in the graphene p-n junction was estimated to be 3 nA, which is close to the experimentally observed values.

The excellent agreement between the experimentally measured and simulated electric field profiles validates the proposed method as a highly accurate and effective technique for nanoscale light field imaging. The non-invasive nature of the method, along with its high spatial resolution, makes it particularly suitable for characterizing strongly confined optical modes in nanophotonic circuits. The success of the method stems from the precise control of the p-n junction position within the graphene sheet via gate voltage and the linear relationship between photocurrent and light intensity. The negligible absorption of light by graphene ensures that the measurement does not significantly affect the electromagnetic field being measured. The authors discuss the potential contribution of thermoelectric effects and conclude that they are minimal under their experimental conditions. The comparison between the experimental results and the FDTD/FEM simulations strengthens the confidence in the accuracy of the technique. The high spatial resolution demonstrated in this work surpasses the capabilities of current near-field microscopy techniques and opens up new possibilities for advanced nanophotonic device characterization and development.

This research successfully demonstrates a novel, non-invasive method for nanoscale light field imaging using a gate-voltage-controlled p-n junction in graphene. Achieving a spatial resolution of ~20 nm, the technique offers high precision and accuracy, as confirmed by the excellent agreement between experimental results and numerical simulations. The method's non-invasive nature and compatibility with existing nanophotonic fabrication processes make it a promising tool for characterizing a wide range of nanoscale optical devices. Future research could explore improvements in spatial resolution by further optimizing the device geometry and material properties, as well as extending the technique to two-dimensional imaging and different types of optical modes.

The current implementation of the technique is limited to one-dimensional imaging. The complex fabrication process might pose challenges for widespread adoption. The maximum gate voltage applied was restricted by the dielectric's breakdown voltage, potentially limiting the range of accessible field distributions. Further optimization of the device geometry and material choices could potentially further enhance the spatial resolution and sensitivity.