🔬 optics

Transverse quantum-state characterization of programmable electron optics

This paper demonstrates the first direct measurement of the transverse quantum-state purity of a programmable electrostatic spiral phase plate using mixed-state ptychography, revealing that the device produces a substantially mixed beam rather than a pure coherent wave, a finding that enables in situ calibration and suggests significant potential improvements in dose efficiency for advanced electron imaging techniques.

Original authors: Shengbo You, Paolo Rosi, Enzo Rotunno, Alberto Roncaglia, Luca Belsito, Amir H. Tavabi, Rafal E. Dunin-Borkowski, Vincenzo Grillo, Philipp M. Pelz

Published 2026-08-07

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Shengbo You, Paolo Rosi, Enzo Rotunno, Alberto Roncaglia, Luca Belsito, Amir H. Tavabi, Rafal E. Dunin-Borkowski, Vincenzo Grillo, Philipp M. Pelz

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 trying to paint a masterpiece with a brush that you can't see. In the world of electron microscopy, scientists use beams of electrons instead of light to see things smaller than atoms. For years, researchers have been trying to twist these electron beams into tiny, swirling tornadoes called "vortex beams." Think of a regular electron beam like a straight arrow flying through the air, but a vortex beam is like a corkscrew or a spiral staircase made of invisible energy. These spirals carry a special kind of spin called "orbital angular momentum," which makes them incredibly powerful tools for studying tiny magnets, chiral molecules, and even for creating new types of X-rays.

The big dream is to build "programmable" electron optics—devices that act like a digital dial, allowing scientists to twist the electron beam into any shape or spin they want just by turning a voltage knob. It's like having a magic remote control for the shape of an electron. However, there's a catch: everyone has been assuming these beams come out perfectly pure and smooth, like a laser. But in the real world, nothing is perfect. Just like a cheap projector might throw a blurry, flickering image instead of a crisp one, these electron devices might be delivering a messy, "mixed-up" beam without anyone knowing. If the beam isn't as pure as we think, all the amazing predictions about what these beams can do might be overestimating their power. So, the big question is: Are these programmable electron beams actually perfect, or are they secretly messy?

This paper is like a high-tech quality control check for those programmable electron beams. The researchers, led by a team from Germany and Italy, decided to stop guessing and actually measure the "purity" of the electron beam coming out of a tiny, programmable device called a Micro-Electro-Mechanical System (MEMS) spiral phase plate. Instead of just looking at the shadow the beam casts, they used a clever technique called "mixed-state ptychography." You can think of this like taking a single, super-detailed 4D scan of the beam and then using a computer to reverse-engineer exactly what the beam looks like inside, revealing not just its shape, but how "messy" or "mixed" it is.

What they found was a bit of a surprise. The device does work, and it can twist the beam into different spins as the scientists programmed it to do. However, the beam is far from the perfect, pure vortex everyone hoped for. As the scientists turned up the voltage to create stronger twists (higher "charge"), the beam actually got messier. It's as if the more they tried to twist the beam, the more it started to wobble and lose its perfect spiral shape. The "purity" of the beam dropped significantly, falling from about 0.47 to 0.24 as the twist increased. This means the beam is a "mixed state"—a jumbled combination of different spins rather than a single, clean one.

The team also discovered that this messiness isn't just a simple blur caused by the electron source being slightly out of focus. If it were just a blur, the beam would actually look cleaner as it got bigger, but the opposite happened. Instead, the messiness seems to come from the device itself, possibly due to tiny electrical fluctuations or the way the device's electrodes interact with the beam. Because the beam is so mixed, the researchers calculated that the efficiency of these beams for taking pictures is much lower than previously thought—roughly three times worse than if the beam were perfect.

In short, this paper doesn't say the technology is broken, but it does say we need to stop assuming it's perfect. By using a single scan to measure the "quantum state" of the beam, the authors have provided a new way to test these devices. They show that while we can program the twist, the beam we get out is a partially decohered, mixed-up version of what we asked for. This finding is crucial because it means that future experiments relying on these beams need to account for this "messiness" to get accurate results, and it suggests that if we can clean up the device design, we could make these electron microscopes much more powerful and efficient.

Technical Summary: Transverse Quantum-State Characterization of Programmable Electron Optics

Problem Statement
Programmable electron optics, particularly microelectromechanical system (MEMS) electrostatic spiral phase plates (SPPs), are central to emerging proposals for dose-efficient phase imaging, shaped-electron X-ray sources, and quantum electron optics. These applications fundamentally assume that the delivered electron beam is a pure, fully coherent wavefunction. However, the purity and coherence of these beams have never been directly measured; verification has historically relied on intensity-based signatures (e.g., doughnut diffraction patterns) or partial projections that fail to capture the off-diagonal coherences of a mixed state. As noted in existing reviews, current diagnostics work well for single pure orbital angular momentum (OAM) states but are insufficient for the mixed states actually produced by real devices. This gap leaves critical application proposals—such as magnetic contrast scaling with $\ell$ or dose-efficient phase contrast—unquantified regarding their true signal-to-noise potential.

Methodology
The authors address this gap by reconstructing the transverse density matrix of a programmable MEMS electrostatic SPP using mixed-state ptychography on standard 4D-STEM data. The approach requires no additional hardware or dedicated column optics. The methodology employs a two-stage reconstruction scheme:

Direct Probe Estimation: A non-iterative stage estimates the complex aperture function and effective phase winding directly from the shifts of the bright-field disk. This utilizes a loop integral to separate the vortex circulation from scan-distortion contributions, providing an initial probe estimate.
Mixed-State Ptychographic Refinement: Initialized with the first-stage probe, a mixed-state ptychographic algorithm refines the complex probe modes. This stage reconstructs an effective, finite-rank density matrix of the illuminating probe from the same scanning-diffraction data used for imaging.

The study applies this pipeline to a charge series of $\ell = -17$ to $+17$ on a MEMS SPP. The reconstructed modes are used to calculate mode powers, spatial coherence (in real and Fourier space), and the spectral signal-to-noise ratio (SSNR) and detective quantum efficiency (DQE) for each state. A "virtual OAM sorter" is implemented computationally by projecting the reconstructed mutual-coherence operator onto an OAM basis, allowing for the extraction of OAM populations and inter-channel coherences without hardware sorting.

Key Results

Substantial Beam Mixing: The delivered beam is found to be substantially mixed, not pure. The purity ( $\text{Tr } \rho^2$ ) decreases monotonically from approximately 0.47 at $\ell=0$ to 0.24 at the highest charges ( $|\ell| \approx 17$ ). Correspondingly, the von Neumann entropy rises from $\approx 1.1$ nats to $\approx 1.6$ nats, indicating an increase in effective mode number from 3 to 5.
Inconsistency with Source Blur: The observed decay in purity with increasing charge is inconsistent with a fixed lateral source-blur model, which would predict improving relative coherence as the vortex ring radius grows. This suggests the mixing arises from device-specific mechanisms (e.g., bias noise, electrode charging) rather than simple source limitations.
Real-Space Coherence Stability: Despite the mixing, the real-space coherence width remains remarkably constant at approximately 1 nm across all programmed charges.
Virtual OAM Sorting and Coherence: The virtual sorting reveals that the beam is neither a pure OAM eigenstate nor a fully incoherent mixture. It is a partially coherent, mixed-OAM vortex with appreciable off-diagonal coherence between adjacent OAM channels. The degree of coherence ( $g$ ) between the dominant channel and its neighbor drops from $\approx 0.78$ (for the round beam) to a plateau of $\approx 0.30$ for higher charges.
Dose Efficiency Impact: The measured mode mixing significantly reduces dose efficiency. While a coherent vortex mode offers a marginal advantage over an ideal round aperture at low spatial frequencies, the measured mixed state achieves only 24–50% of the ideal round aperture's dose efficiency. Purifying the output could theoretically improve dose efficiency roughly threefold.
In-Situ Calibration: The scans provide an in-situ calibration of the device, showing a linear relationship between applied bias and the delivered mean OAM ( $\approx 16.8 \hbar/\text{V}$ ), with a slope of 1.68 relative to the programmed integer charge.

Significance and Claims
The paper claims to provide the first direct, quantitative characterization of the transverse quantum state of a programmable electron optic. By converting assumed beam properties into measured, per-charge-state specifications, the work establishes a "quantum-state acceptance test" for such devices.

The authors emphasize that the reconstruction is entirely computational, meaning the same 4D-STEM acquisition used for imaging can simultaneously validate the probe's quantum state. This capability is presented as a prerequisite for the field: application proposals that rely on perfect purity or specific coherence properties must now be evaluated against measured reality rather than idealized assumptions. The results suggest that current device designs (specifically the electrode geometry relative to the illuminated aperture) introduce significant mixing, and that optimizing these parameters could yield purer states. The paper concludes that this method enables closed-loop, self-calibrating electron optics where the delivered quantum state is programmed, measured, and corrected in situ.

More like this