🔬 optics

Spatiotemporal Electron Microscopy of Phonon Polaritons in MoO3

This study expands the operational bandwidth of photon-induced near-field electron microscopy (PINEM) to 12 μm, enabling the direct spatiotemporal imaging of phonon polariton dynamics and lifetime in anisotropic α-MoO₃.

Original authors: Harel Nahari, Yaniv Kurman, Raphael Dahan, Yuval Adiv, Michael Yannai, Hanan Herzig Shenfux, Frank H. L. Koppens, Ido Kaminer

Published 2026-05-29

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Harel Nahari, Yaniv Kurman, Raphael Dahan, Yuval Adiv, Michael Yannai, Hanan Herzig Shenfux, Frank H. L. Koppens, Ido Kaminer

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to watch a ghost dance. The ghost is invisible to the naked eye, moves incredibly fast, and exists in a world too small for normal cameras to see. This is the challenge scientists face when trying to observe phonon polaritons—tiny ripples of light and matter that travel inside special crystals.

In this study, a team of researchers built a super-powered "ghost camera" to watch these ripples dance inside a crystal called alpha-Molybdenum Trioxide (α-MoO₃). Here is how they did it, explained in simple terms:

1. The Special Crystal (The Stage)

Think of the α-MoO₃ crystal as a unique dance floor. Unlike a normal floor where you can walk in any direction, this floor has a special rule: you can only dance easily in one specific direction, and the floor behaves differently depending on which way you face. This is called "anisotropy." Because of this, the light ripples (phonon polaritons) inside it behave in strange, exotic ways, like bouncing back and forth in a specific pattern.

2. The "Ghost Camera" (PINEM)

To see these invisible ripples, the scientists used a technique called PINEM (Photon-Induced Near-field Electron Microscopy).

The Setup: Imagine a high-speed electron microscope (the camera) and a laser (the flashlight).
The Trick: They fired a burst of infrared light (the flashlight) at the crystal to make the ripples start dancing. Then, they fired a stream of electrons (the camera) right through the crystal.
The Interaction: As the electrons flew through, they bumped into the light ripples. Some electrons gained a tiny bit of speed (energy) from the ripples, while others lost a bit.
The Image: By filtering the electrons and only counting the ones that gained speed, the scientists could reconstruct a picture of the ripples. It's like seeing the ghost not by looking at it, but by seeing which people in a crowd got pushed by it.

3. Pushing the Limits (Going Deeper)

Previous versions of this "ghost camera" could only see ripples with wavelengths up to 7 micrometers (a very small distance). In this study, the researchers upgraded their equipment to see ripples up to 12 micrometers.

The Analogy: If previous cameras could only see a person from a few feet away, this new camera can see them from across the room. This allowed them to study a deeper, more infrared part of the light spectrum that was previously hard to reach.

4. What They Saw (The Dance)

With this new, deeper vision, they watched the ripples inside a tiny flake of the crystal (about the size of a speck of dust).

The Movement: They saw the ripples start in the middle of the flake and race toward the edges.
The Speed: These ripples are incredibly slow compared to normal light. They moved at about 2.4 micrometers per picosecond. To put that in perspective, they were moving 125 times slower than light in a vacuum. As they hit the edge of the crystal, they slowed down even more.
The Shape: The ripples formed a specific shape with a "nodal line" (a quiet zone) right down the center, which is forced by the crystal's unique rules.

5. Measuring the Lifespan (How Long the Dance Lasts)

The scientists also wanted to know how long these ripples could keep dancing before fading away.

The Method: They measured the energy of the electrons at different moments in time.
The Result: They found that the ripples don't last long. At a wavelength of 11 micrometers, the dance lasted about 0.45 picoseconds (that's less than one-trillionth of a second). At 12 micrometers, the dance ended even faster, almost instantly.
The Discovery: They noticed a pattern: the longer the wavelength (the "deeper" into the infrared), the shorter the dance lasted.

Why This Matters

This paper is a breakthrough because it proves that this "ghost camera" can now see much deeper into the infrared spectrum than before. It shows that we can watch these tiny, fast-moving energy waves in real-time, even inside materials that are hard to study.

The researchers note one big challenge: getting the light to actually enter the crystal is tricky. It's like trying to shout a secret into a locked room; the sound (light) doesn't always get in easily. However, now that they have the camera working at these new depths, they can start exploring the "exotic" physics of these materials, potentially leading to new ways to control light in the future.

In short: They built a better camera, looked deeper into the infrared, and filmed a super-fast, super-slow dance of light inside a crystal, measuring exactly how long the dance lasted.

Technical Summary: Spatiotemporal Electron Microscopy of Phonon Polaritons in $\alpha$ -MoO $_3$

Problem and Motivation
Photon-induced near-field electron microscopy (PINEM) has established itself as a premier technique for imaging optical excitations with nanometer spatial and sub-picosecond temporal resolution. While recent advancements have extended PINEM operations from the visible range into the mid-infrared (up to 7 $\mu$ m), enabling the study of phonon polaritons (PhPs) in materials like hexagonal boron nitride (hBN), a significant spectral gap remains. The discovery of novel two-dimensional (2D) materials with exotic photonic properties, such as the in-plane hyperbolic nature of $\alpha$ -phase molybdenum trioxide ( $\alpha$ -MoO $_3$ ), necessitates further expansion of PINEM's operational bandwidth. Existing techniques like scattering-type scanning near-field optical microscopy (s-SNOM) struggle to access buried near-fields or fields extending outside the sample confines, limitations that PINEM inherently overcomes. However, probing $\alpha$ -MoO $_3$ PhPs requires wavelengths deeper into the infrared (beyond 7 $\mu$ m) to match the material's Reststrahlen bands.

Methodology
The authors developed an experimental setup that nearly doubles the spectral bandwidth of PINEM, extending its operational range up to 12 $\mu$ m. The core of the experiment involves a time-resolved interaction between ultrafast electron pulses and mid-infrared (mid-IR) optical pulses within an $\alpha$ -MoO $_3$ flake.

Excitation: A femtosecond laser is split into two branches. The bottom branch undergoes difference frequency generation (DFG) to produce a tunable mid-IR pulse (10–12 $\mu$ m), which excites PhP modes in a 150 nm thick, cavity-like $\alpha$ -MoO $_3$ flake (2.5 $\mu$ m $\times$ 7 $\mu$ m).
Probing: The top branch is converted via fourth-harmonic generation into a UV pulse (~257 nm) to photoelectrically generate the electron probe pulses.
Imaging and Spectroscopy: The electron pulses interact with the near-field of the excited PhPs. By energy-filtering the electrons that have gained energy (PINEM signal), the authors visualize the spatial distribution of the PhP modes. By varying the time delay between the pump (IR) and probe (electron) pulses, they capture the spatiotemporal evolution of the wavepackets. Additionally, electron energy spectroscopy is performed to analyze the energy variance over time, allowing for the extraction of mode lifetimes.

Key Contributions and Results

Bandwidth Extension: The study successfully extends PINEM operation to wavelengths of 12 $\mu$ m, significantly broadening the accessible spectral range for polariton imaging.
Spatiotemporal Imaging of $\alpha$ -MoO $_3$ PhPs: The authors directly imaged the PhP mode in an $\alpha$ -MoO $_3$ cavity-like flake. The filtered images revealed the out-of-plane electric field distribution both inside and outside the sample, showing a nodal line along the long axis ( $y$ ) and a dipole-shaped field outside. This pattern is attributed to the material's hyperbolic nature ( $\varepsilon_x > 0, \varepsilon_y < 0$ ), which inhibits modes with zero $k_y$ , enforcing a protected nodal line.
Wavepacket Dynamics: Time-resolved measurements tracked the propagation of the PhP wavepacket. The wavepacket was observed to move from the center of the flake toward the edge with an initial group velocity of 2.4 $\mu$ m/ps (approx. 125 times slower than $c$ ). Upon reaching the flake's edge, the velocity decelerated to 0.55 $\mu$ m/ps (approx. 545 times slower than $c$ ).
Wavelength-Dependent Lifetime: Using ultrafast electron energy spectroscopy and fitting the electron energy distribution to a theoretical model involving the PhP-mode lifetime ( $\tau$ ), the authors extracted the lifetime of the cavity mode. The results demonstrated a strong wavelength dependence: the lifetime was measured at 0.45 ps for an 11 $\mu$ m excitation, while for 12 $\mu$ m, the lifetime was significantly shorter (0.04 ps), falling below the measurement resolution. This inverse relationship between lifetime and wavelength aligns qualitatively with previous experimental findings.

Significance and Outlook
The paper claims that this work pushes the frontiers of PINEM imaging, highlighting its unique potential to probe the hard-to-access polaritonic properties of novel van der Waals materials. By visualizing fields outside the sample and capturing the full spatiotemporal dynamics, PINEM offers capabilities distinct from tip-based techniques like SNOM.

The authors identify that while the technique is powerful, challenges remain, particularly in coupling free-space light into high- $k$ vector components of 2D materials, which limits the full utilization of PINEM's high spatial resolution. They suggest that overcoming this momentum mismatch via nanopatterning or tip insertion could be a future direction. Furthermore, the extension to longer wavelengths opens possibilities for controlling electron wavefunctions and exploring regimes where macroscopic material parameters may fail, such as in pico-cavities. The study positions PINEM as an increasingly vital tool for the exploration of 2D material physics, specifically for observing phenomena like polariton canalization and diffraction-less propagation.