High precision micro-optical elements on fiber facets via focused-ion beam machining

arXiv:2604.18426 · physics.optics, physics.app-ph, physics.atom-ph, quant-ph · Submitted 2026-04-20 · Read on arXiv

Listen

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: I'm Kai, and with me are Mira and Lev, guest researcher.

Mira: Today's paper: "High precision micro-optical elements on fiber facets via focused-ion beam machining".

Kai: Fiber-integrated micro-optical elements fabricated on fiber facets promise a scalable approach to photon collection and beam shaping for quantum information processing.

Mira: First, who's behind it and why it matters.

Title and authors: Kai: So, we're diving into this paper now titled "High precision micro-optical elements on fiber facets via focused-ion beam machining," and it looks like they’ve really tackled getting these tiny optical structures directly onto fiber cores.

Mira: That's right, Kai; the authors are focusing on using focused-ion beam machining to create micro-spherical, microspiral, and microaxicon structures right on a single-mode fiber facet, aiming for extremely precise shapes.

Lev: From my side as someone who deals with actual hardware setups, I’m curious if these fabrication steps are truly repeatable enough for us to integrate this into a functioning quantum processor without constant recalibration of the ion beam alignment.

Kai: That's a valid concern, Lev; it seems they laid out a pretty detailed process starting with core identification using chemical etching in twenty:one buffered oxide etch for fifteen minutes on the fiber facet <ref:2604.18426#pg0>.

Mira: And that etching step is crucial because it allows them to deterministically locate the central core, an intermediate annular region, and the outer cladding, which sets up the deterministic alignment needed for subsequent focused-ion beam processing.

Lev: If they can achieve that deterministic alignment between the ion column axis and the fiber facet with a tilt of fifty-two degrees as mentioned in their method section, that simplifies things immensely for scaling up experimental setups <ref:2604.18426#pg0>.

Kai: They did mention using a Thermo Fisher Helios G5 dual-beam instrument with a liquid gallium metal ion source, and they are using bitmap pattern files to encode the local milling depth for these structures.

Mira: The specific parameters they used included a beam voltage of thirty kV and a beam current of two point six nA for creating micro-concave or micro-convex features, which gives us a concrete idea of the precision they are aiming for in their fabrication process.

Lev: That current level suggests that if we were to implement this in an error correction context, the resulting surface topography would have a certain level of fidelity that we could potentially model as noise or as a controllable feature.

Kai: The results they reported are pretty impressive when you look at the atomic force microscopy data; they found micro-concave and micro-convex surfaces achieving shape accuracies of lambda/eighty and lambda/fifty respectively, when the wavelength is seven hundred eighty nm <ref:2604.18426#pg0,micro-concave and micro-convex>.

Mira: That level of precision means that by using optical characterization with a He-Ne laser at six hundred thirty-three nm, they verified expected far-field donut beam patterns for the microspiral and microaxicon fibers as well <ref:2604.18426#pg0,expected far-field donut beam patterns>.

Lev: If we're looking at real hardware performance, achieving those shape accuracies over a forty-five percent or fifty percent area, as shown in their curve-fitting analysis of residuals, would give us a very specific metric to test against our theoretical models for light-matter interaction <ref:2604.18426#pg2>.

Kai: They also showed the phase structure verification using Mach-Zehnder interferometry, confirming that the microspiral fiber displays a spiral profile with exactly one fringe per azimuthal rotation of 2π <ref:2604.18426#pg0>.

Title and authors: Mira: That confirms the radial and azimuthal phase structure they were aiming for by showing it matches their analytical height profile derived from a sphere with a desired radius of curvature, like the one hundred six point one three µm for micro-concave surfaces <ref:2604.18426#pg2>.

Lev: Knowing that they've verified this phase structure using interferometry means we have a solid baseline to start checking if our quantum states are actually coherent when interacting with these fabricated elements.

Kai: They also dedicated a section to surface quality analysis, showing that while the areal RMS roughness increased from zero point seven six nm to zero point nine five nm after FIB processing, this added roughness is sub-nanometer and negligible compared to the optical wavelengths involved.

Mira: That's important because they confirmed using two-dimensional surface power spectral density analysis that FIB doesn't introduce new roughness components at any lateral scale relevant to optical wavelengths, meaning the optical quality is maintained.

Lev: Preserving nanoscale surface quality while introducing these specific geometric features is a key finding; it tells us that the fabrication method respects the underlying physics of light propagation, which is vital for minimizing scattering losses in our quantum hardware.

Kai: Looking ahead at their conclusions, they highlight that this single-step FIB approach gives them flexibility to define more general surface profiles compared to older laser ablation or etching techniques.

Mira: That flexibility is what makes it so promising; it moves beyond simple geometry into defining more complex functional surfaces directly on the fiber core for photon collection and beam shaping.

Lev: For error correction, that generalized control over the surface profile could potentially be used to tailor the coupling environment precisely, which is something we've been struggling to achieve with simpler geometries.

Kai: So, to wrap up this discussion on "High precision micro-optical elements on fiber facets via focused-ion beam machining," we see a clear demonstration of how FIB can reproducibly create these high-precision micro-optical elements directly integrated onto single-mode fibers.

Mira: The main implication is that we have a scalable route for creating photon collection and beam shaping components that meet quantum standards, which opens up possibilities for fiber-integrated microcavities and structured light generation.

Lev: For practical quantum hardware, the ability to define these complex surface profiles with this level of accuracy could lead to more controlled interaction zones within a chip or cavity.

Kai: I think what this paper really establishes is that FIB is a viable platform for creating precise optical elements that are compatible with quantum information science applications.

Mira: It sets a new benchmark for achieving high-precision micro-optics in fiber systems by confirming shape accuracies of lambda/eighty and lambda/fifty which is a very specific metric we need to keep in mind <ref:2604.18426#pg0>.

Lev: I just want to say that if the fabrication process proves robust enough for real hardware operation, it gives us a concrete path toward realizing more sophisticated quantum light sources and beam shapers.

Kai: We're going to pause here for now, but next up we'll discuss how these micro-optical elements can actually be used in practice to build those quantum devices we were talking about earlier.

The paper's summary: Kai: So, to quickly summarize what we just went over, this paper is about using focused ion beam machining to precisely build micro-optical elements directly onto fiber facets for quantum applications.

Mira: Exactly; the core idea is that FIB milling can achieve a very high degree of shape accuracy—better than lambda/fifty—which establishes a reliable pathway for creating fiber-integrated micro-optics suitable for quantum information processing.

Lev: From my viewpoint, if they can reliably produce these structures with that level of precision, it means we have a fabrication method that respects the optical constraints needed for high-fidelity operations on real hardware.

Kai: Right, and the results are pretty compelling because they didn't just build some random shapes; they verified specific optical functions like donut beams for microspirals and Bessel beams for axicons using far-field imaging.

Mira: That’s where the theoretical underpinning gets interesting; confirming those specific phase structures through interferometry shows that their geometric design actually translates into the intended quantum properties, which validates their modeling assumptions about light propagation in these confined geometries.

Lev: If we think about running this on a real chip, that level of precision means we aren't fighting significant shape errors when trying to control atom trapping or beam shaping; it moves the challenge from fabrication noise to fundamental physics control.

Kai: And they also addressed surface quality, showing that even with the milling, the added roughness is so small at relevant optical wavelengths that it doesn't significantly degrade performance.

Mira: That’s a crucial detail for me because it tells us that our assumptions about scattering losses in these fiber systems hold up even when introducing these complex three dee geometries; the process preserves the necessary optical surface quality.

Lev: So, if we take this precision and combine it with our work on Lindbladians or error correction, we could start designing quantum hardware where the optical elements themselves are tailored to optimize those specific quantum dynamics.

Kai: That’s a big leap from just building a component; it opens the door to designing entire integrated systems where the light is engineered for specific quantum interactions.

Mira: Indeed, this work suggests that creating fiber-integrated micro-optics isn't just about getting a shape right; it’s about using deterministic fabrication to realize specific optical functions required for advanced quantum protocols.

Lev: It really pushes us toward a future where the hardware design is intrinsically linked to the light field manipulation, which is exactly what we need for robust quantum computation.

Kai: We've seen how this technique can be used for things like neutral atom trapping and cavity QED; it's showing real promise for making these complex optical setups practical.

Mira: The implication here is that FIB offers a flexible tool beyond simple etching, allowing us to define more generalized surface profiles needed for complex quantum light-matter interfaces.

Lev: And I think the real impact will be in how we build fault-tolerant systems, because if the component fabrication is this precise, we can focus our error correction efforts on the quantum states themselves rather than fighting structural imperfections.

Kai: So, to wrap up this summary of their findings, they’ve shown a scalable way to manufacture these high-precision optical elements directly on fiber facets that meet quantum standards.

Mira: It solidifies the idea that deterministic fabrication is a viable route toward building sophisticated quantum photonic devices integrated with fiber systems.

Lev: This provides a concrete, precise platform for implementing the complex light-matter interactions we’ve been modeling, giving our error correction research something tangible to test against in terms of physical layout challenges.

Kai: It certainly gives us more to work with when planning the next generation of quantum hardware designs that need highly controlled optical environments.

The paper's improvements: Kai: So, to quickly summarize our previous discussion, the paper demonstrated how focused ion beam machining can precisely create micro-optical elements right on fiber facets for quantum applications, achieving accuracy levels better than lambda/fifty and verifying specific optical functions like donut beams.

Mira: Exactly; they proved that FIB is a robust way to move beyond simple etching and actually define complex shapes with high fidelity, which is key for the underlying physics of light confinement we need to model accurately.

Lev: From my perspective, this precision is exactly what we’d need if we were trying to build real hardware where every component has a known, predictable optical environment for error correction.

Kai: And they also looked at how to make the process more flexible; they suggested that the bitmap pattern files can be adapted to define a wider variety of surface profiles than before, which is really useful for prototyping different interaction geometries.

Mira: That flexibility is important because it means we aren't locked into one specific shape; we can tailor the micro-optics to match the exact phase requirements of our Lindbladian models or our desired topological charge states.

Lev: If they can adapt the pattern generation, it suggests a pathway toward designing elements that are optimized not just for a single use case, but for a whole family of quantum experiments.

Kai: I’m also interested in their suggestion to integrate surface metrology feedback directly into the design loop; having an AFM or SEM measurement feed back into the next milling iteration sounds like a way to push accuracy even further than what's currently achievable.

Mira: That concept is interesting because it moves us from a static design to an iterative optimization process, allowing us to use experimental data immediately to refine the theoretical predictions about how those geometries will behave optically.

Lev: If we can couple that metrology feedback with our quantum simulation algorithms, we could create a closed-loop system where the hardware actively learns and optimizes its own optical structure for better coherence.

Kai: That kind of active feedback mechanism would be fantastic for dynamic adjustments in cold atom experiments or when dealing with fluctuating noise in free-space links.

Mira: It points toward a more holistic approach where we aren't just fabricating a component, but designing an entire optical system that is self-optimizing based on real-world measurement constraints.

Lev: That level of integrated optimization could significantly reduce the experimental overhead required to tune complex quantum systems to work properly.

Kai: It really shows how the fabrication process itself can be a part of the quantum control strategy, not just a prerequisite step before we start our main experiments.

Mira: This paper suggests that for advanced fiber-integrated quantum hardware, we should prioritize methods that allow for this kind of closed-loop optimization between fabrication and optical characterization.

Lev: It gives us a clearer roadmap on how to translate the theoretical requirements of fault tolerance into physical device specifications with high certainty.

Kai: It’s exciting because it bridges the gap between our theoretical designs and what we can actually build and cool in a lab setting.

Conclusion: Kai: So, to close out this session on "High precision micro-optical elements on fiber facets via focused-ion beam machining," we've established that FIB milling is a reliable way to create these high-accuracy micro-optical structures directly onto fiber cores for quantum applications.

Mira: That’s right; the main takeaway is that this technique provides a scalable route toward creating fiber-integrated micro-optics with shape accuracies we can actually measure and control, which supports our theoretical models about light confinement.

Lev: For error correction purposes, this fabrication method offers a level of structural precision that makes it much easier to design quantum hardware where the optical environment is highly controlled, which is essential for running any meaningful experiments.

Kai: I think the implication here is that we can start designing experimental setups with much more predictable light fields when we integrate these precisely fabricated elements into our quantum chips.

Mira: Indeed, this opens up exciting new avenues for cavity QED and structured light generation because the geometry isn't just arbitrary; it's defined by a deterministic fabrication process.

Lev: If we can reliably build these components, we can start testing whether our quantum error correction codes perform better when they are interacting with these specific, precisely shaped optical modes.

Kai: It’s really encouraging to see this work move from pure concept into a demonstrable fabrication method for complex quantum optics components.

Mira: We've shown that the surface quality is maintained, which means we aren't sacrificing optical performance for geometric complexity, validating the assumptions about scattering losses in these systems.

Lev: I think that preservation of optical grade surface quality is vital because any added noise from fabrication would immediately introduce errors into our quantum state preparation protocols.

Kai: So, to wrap up, this paper on "High precision micro-optical elements on fiber facets via focused-ion beam machining" gives us a solid foundation for building more sophisticated quantum devices with highly controlled optical features integrated directly onto fibers.

Mira: It's a significant step in realizing the potential of deterministic fabrication for complex quantum systems.

Lev: We have a clearer path now to design hardware where the light field itself is an engineered, controllable feature rather than just an incidental byproduct of the setup.

Kai: I think this sets a high bar for what we can expect from future experimental setups in this area.

Mira: Absolutely; it’s about using advanced manufacturing to realize the intricate optical requirements we theorize are necessary for strong quantum interactions.

Lev: It gives us a tangible piece of hardware that meets stringent precision requirements, which is exactly what we need to push the boundaries of what's possible in fault-tolerant quantum computing.

Kai: Alright team, I think this is a great paper to keep on our radar as we look toward designing those next generation quantum architectures.

Raman Kumar, Sebastian Will

Quantum Information Science and Technology, Brookhaven National Laboratory · Department of Physics, Columbia University

physics.optics, physics.app-ph, physics.atom-ph, quant-ph

Submitted: 2026-04-20

Updated: 2026-10-03

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 79/100

The gist: Fiber-integrated micro-optical elements fabricated on fiber facets promise a scalable approach to photon collection and beam shaping for quantum information processing.

Key concepts

FIB Milling: Focused-Ion Beam milling uses a dual-beam instrument with a liquid gallium metal ion source to precisely remove material from the fiber facet. By carefully controlling the ion beam's direction and current, researchers can create complex micro-optical shapes by defining local milling depths based on digital patterns.
Shape Accuracy Verification: The precision of the created elements is measured using Atomic Force Microscopy (AFM) to map surface topography and calculate residuals against a perfect sphere. This method confirms that the resulting surfaces meet stringent requirements, achieving shape accuracies of lambda/80 over key areas.
Optical Performance Metrics: Optical performance is confirmed by testing the structures with specific light sources. For example, far-field imaging verifies expected patterns like donut beams for micro-spiral elements, and Mach-Zehnder interferometry confirms the precise phase structure of the created optical components.

Terminology

Core — 30× in this paper · explained in 2 episodes
Accuracy — 11× in this paper · explained in 1 episode
Alignment — 11× in this paper · explained in 1 episode
Entanglement — 7× in this paper · explained in 3 episodes
Fidelity — 6× in this paper · explained in 2 episodes
Ablation — 3× in this paper · explained in 1 episode
Aperture — 3× in this paper · explained in 1 episode
Verification — 3× in this paper · explained in 2 episodes
Brookhaven National Laboratory — 2× in this paper
Benchmark — 1× in this paper · explained in 2 episodes
Closed-Loop Optimization — 1× in this paper · explained in 2 episodes
Closed-Loop System — 1× in this paper · explained in 4 episodes
Department of Energy — 1× in this paper
Multipartite Entanglement — 1× in this paper · explained in 2 episodes
Orbital Angular Momentum (OAM) — 1× in this paper · explained in 3 episodes

Summary

Fiber-integrated micro-optical elements fabricated on fiber facets promise a scalable approach to photon collection and beam shaping for quantum information processing. The gist: FIB milling can reproducibly achieve shape accuracies better than λ/50, establishing this technique as a route toward quantum-grade, fiber-integrated micro-optics.

Fabrication Process and Core Identification

The fabrication begins with preparing single-mode fibers, which are cleaved and cleaned to reveal distinct regions. A critical step involves precisely locating the fiber core prior to machining; this is achieved by etching the fiber tip with a hydrofluoric acid-based solution and subsequently immersing the freshly cleaved fiber in a 20:1 buffered oxide etch (BOE) solution for 15 minutes. This etching reveals three distinguishable regions rather than a simple core-cladding structure: a central core, an annular intermediate region, and the outer cladding. Following this, the fibers are coated with a thin (∼10 nm) layer of gold using an ion sputter coater to make the facet electrically conductive.

Focused-Ion Beam Processing

FIB milling is performed using a Thermo Fisher Helios G5 dual-beam instrument equipped with a liquid gallium metal ion source. To ensure accurate fabrication results, it is critical to ensure normal incidence of the ion beam on the fiber facet for accurate fabrication results. The sample stage is adjusted iteratively to bring the fiber facet into perpendicular alignment with the ion column axis by tilting it to 52 degrees. Micro-optical elements are defined using bitmap pattern files (512 × 512 pixels) that encode local milling depth. For spherical elements, the pattern is generated from an analytical height profile of a sphere with the desired radius of curvature. Typical parameters include a beam voltage of 30 kV and various beam currents optimized for different features, such as a beam current of 2.6 nA for micro-concave/convex.

Structural Characterization and Precision

The resulting structures are characterized using multiple metrology techniques to verify shape accuracy. Atomic force microscopy (AFM) topography maps are used to characterize the surface profile, yielding radii of curvature (ROC) such as 106.13 µm for micro-concave surfaces and 28.13 µm for micro-convex surfaces. The shape accuracy is assessed by computing residuals between the measured height profile and the best-fit spherical surface; specifically, the residuals are λ/80 (at λ = 780 nm) over the central 45% of the total feature area for micro-concave elements. This level of precision is quantified via equivalent Strehl ratio calculations, demonstrating that surfaces achieve a Strehl ratio significantly exceeding the diffraction limit criterion of S ≥ 0.80.

Optical Performance Verification

The optical performance is verified through various imaging and interferometric methods to confirm the intended function of the structures. For micro-spiral elements, far-field imaging using a He-Ne laser (633 nm) confirms the expected far-field donut beam patterns, which is consistent with a topological charge of l = 1. Similarly, Mach-Zehnder interferometry verifies the phase structure: When the micro-spiral fiber is placed in one arm [Fig. 5(c)], the coaxial interference pattern displays a spiral profile with exactly one fringe per azimuthal rotation of 2π. For micro-axicon structures, far-field images exhibit the concentric ring pattern characteristic of a Bessel beam, consistent with the radial phase imparted by the element.

Surface Quality Analysis

Surface quality is rigorously analyzed to ensure optical grade performance. A comparison between unprocessed and FIB-processed areas shows that while the areal RMS roughness increases from Sq = 0.76nm to Sq = 0.95nm, this added sub-nanometer roughness which is negligible in comparison to optical wavelengths. Furthermore, the two-dimensional surface power spectral density (PSD) analysis confirms that FIB does not introduce new roughness components at any characteristic lateral scale relevant to optical wavelengths, indicating the preservation of optical surface quality. This demonstrates that the FIB process preserves nanoscale surface quality at wavelengths relevant to quantum optics, minimizing scattering losses.

Applications and Conclusion

The demonstrated capabilities enable broad use in quantum technology, including fiber micro-cavities for cavity quantum electrodynamics, beam shaping for neutral atom trapping, and the generation of structured light for free-space quantum network links. The single-step FIB fabrication approach provides flexibility in defining more general surface profiles than prior laser ablation or etching techniques, establishing it as a promising route toward quantum-grade, fiber-integrated micro-optics. This work establishes FIB-fabricated fiber micro-optics as a precise and flexible platform for quantum information science applications.

Improvements for AI systems

Based on the scientific paper, here are several specific improvements to AI systems that could be enabled by these micro-optical elements:

To enhance quantum information processing platforms based on neutral atom arrays or trapped ions, the improved AI system could perform real-time optimization of light-matter coupling strengths.
The improved system could precisely control the generation of structured light (like Bessel beams or vortex beams) to optimize beam shaping for neutral atom trapping, leading to more efficient cooling and state preparation in quantum computing modules.
The AI could be trained on the precise phase profiles imparted by micro-spiral and micro-axicon structures to develop algorithms for generating topological charge states in photons with deterministic precision, crucial for high-fidelity entanglement distribution in free-space quantum networks.
An improved system could utilize the high focal accuracy of the micro-concave elements to perform real-time, adaptive focusing for cold atom experiments, allowing for dynamic adjustment of interaction zones within a quantum processor without manual repositioning.
The AI could be used to design optimal optical cavity configurations (using the demonstrated small mode volumes) that maximize the cooperativity between atoms and photons in fiber microcavities, enabling higher fidelity two-qubit quantum gate operations.
The system could incorporate surface metrology feedback (AFM/SEM data) into a machine learning loop to predict the optical performance (Strehl ratio) of a fabricated micro-optical element before physical implementation, significantly accelerating the design cycle for quantum photonic components.
For quantum key distribution over long distances, the AI could dynamically adjust beam shaping parameters based on real-time turbulence models to maintain high fidelity in free-space links using the resilient Bessel beam generated by the micro-axicon.