Super-resolution Optical Near-field EM for bio- and materials science

TL;DR

The authors present Optical Near-field Electron Microscopy (ONEM), a non-invasive, label-free approach that converts optical near-field intensities into photoelectrons for LEEM-based imaging. Achieving sub-diffraction spatial resolution ($>31\mathrm{nm}$, i.e., $\lambda/13$ at $405\mathrm{nm}$) and sub-second temporal resolution, ONEM images interfaces without exposing samples to high-energy electrons. They demonstrate three proof-of-principle domains—polarization-resolved plasmon modes in metal nanostructures, orientation-aware 3D imaging of live E. coli in liquid, and real-time copper electrodeposition—highlighting the method’s potential across biology, electrochemistry, and nanophotonics. The work outlines clear paths for improvements (thinner supports, higher quantum efficiency photocathodes, better detectors) and discusses broad implications, including applications in protein dynamics, lipid membranes, and connectomics. Overall, ONEM offers a versatile, damage-free platform for fast, nanoscale imaging of dynamic interfacial processes in both vacuum and liquid environments.

Abstract

Microscopy has been key to tremendous advances in science, technology, and medicine, revealing structure and dynamics across time and length scales. However, combining high spatial and temporal resolution in a non-invasive, label-free imaging technique remains a central challenge in microscopy. Here, we introduce Optical Near-field Electron Microscopy (ONEM), a method that converts optical near-field intensity patterns into photoelectron emission, enabling nanometer-scale imaging using low-energy electron microscopy. ONEM achieves 31 nm spatial and sub-second temporal resolution without exposing the sample to electrons, preserving structural and functional integrity. We demonstrate ONEM across three distinct domains: imaging polarization-dependent plasmon modes in metal nanostructures; visualizing live Escherichia coli in liquid with orientation-resolved contrast in 3D; and capturing real-time electrodeposition of copper nanoclusters from solution. These results establish ONEM as a versatile platform for damage-free super-resolution imaging of interface dynamics in both vacuum and liquid, with broad implications for biology, electrochemistry, and nanophotonics.

Figures (13)

• Figure 1: Optical Near-field Electron Microscopy (ONEM) setup and resolution.a, Sketch: ONEM combines an optical illumination module (green) with electron detection in a low-energy electron microscope (LEEM, red). In ONEM, the electron gun of the LEEM system is off. Instead, the sample is illuminated with light from a single-mode fiber. The near-field intensities close to the sample are converted into a photoelectron flux using a photocathode. Closed-loop polarization control allows for setting any desired polarization of the illumination. Electrons emitted by the photocathode are collected by the electron objective lens and imaged by the LEEM. b, ONEM on lithographically fabricated test sample: Photons pass through a fused silica substrate and interact with a geometrical pattern of chromium and fused silica. The resulting near-field intensities are converted to electrons in the cesiated graphene photocathode. c, ONEM image of a single $250\,\mathrm{nm}$ wide line of the geometrical pattern at high magnification. d, An error function fit (red) to a (10 pixel wide) line cut (blue) of c yields an upper bound of the spatial resolution of $2\sigma=31\,\mathrm{nm}$.
• Figure 1: Demonstration of ONEM capability to distinguish the orientation of plasmons within 1 second acquisition time. ONEM image of a $100\,\mathrm{nm}$ Ag nanocube illuminated with two perpendicular linear polarizations. Exposure times of 1 second are sufficient to determine the orientation of the plasmon. I.N. - intensity normalized to the background.
• Figure 2: Surface plasmon response imaged with ONEM.a, Sample geometry. $100\,\mathrm{nm}$ Ag nanocubes on top of $20\,\mathrm{nm}$ Si$_3$N$_4$ are illuminated with $405\,\mathrm{nm}$ collimated light. A $3\,\mathrm{nm}$ layer of caesiated chromium acts as a photocathode. b, Sample area imaged by both SEM (left), and ONEM using circularly polarized illumination (right). c, Zoom: nanocube in SEM. d, Magnified ONEM images of nanocube using circular (left) and linear (right) polarization. e, Polarization response of nanocube. Top row: (linear) polarization angles chosen, except: LCP - left circular polarized. Second row: ONEM images, showing the intensities normalized to the background (I.N). Third row: simulated intensities in the photocathode. Fourth row: third row convoluted with a Gaussian with $2\sigma$ = $35\,\mathrm{nm}$. f, Line profile of images in second column of e, taken between arrows, comparing ONEM and (broadened) simulation. Note the hot spot in the center. g, ONEM image and (broadened) simulation of a nanosphere illuminated with linearly polarized light. Note the minimum in the center.
• Figure 2: ONEM liquid cell.a, Cross-cut of the ONEM liquid cell, showcasing both the outer shell (in red and gray) and the inner shell (from bottom to top Si$_3$N$_4$ chip, O-ring, and glass window). The inset shows the Si$_3$N$_4$ window, with slanted Si walls next to it. Light reflections from these walls lead to interference patterns in our ONEM images, which are removed using Fourier filtering. b, The assembled liquid cell is displayed next to a 1 Euro coin. c, The liquid cell is placed at the tip of the LEEM sample holder. The magnified image shows a pogo pin pressing against the coated Si$_3$N$_4$ chip. It ensures a secure electrical contact required for electrochemical experiments.
• Figure 3: Live E. coli bacteria imaged with ONEM.a, Schematic of the ONEM liquid cell. E. coli bacteria (blue capsules, not to scale) are confined between a glass window and a Si$_3$N$_4$ chip, separated by an O-ring. The closest distance of the bacterium to the photocathode, consisting of Cr (dark gray) and Cs (yellow), is indicated by $d$, while its orientation relative to the photocathode plane is given by the angle $\theta$. b, the bacterial trajectory (blue) over a 22-second recording, starting at $t=0.3\,\mathrm{s}$ when the bacterium enters the field-of-view. c, d, e, selected frames. f, g, h, simulations for a $20\,\mathrm{\text{\textmu m}}$ long and $1\,\mathrm{\text{\textmu m}}$ wide bacterium f with $d= 1\,\mathrm{\text{\textmu m}}, \theta = 30^\circ$, g with $d= 3.15\,\mathrm{\text{\textmu m}}, \theta = 90^\circ$, and h with $d= 0.5\,\mathrm{\text{\textmu m}}, \theta = 60^\circ$.