🔬 applied physics

A complete fs-laser-ablation route to miniaturized single-crystal PMN-PT piezoelectric actuators

This paper presents a novel, fully femtosecond-laser-based fabrication route for miniaturized single-crystal PMN-PT piezoelectric actuators that utilizes a local thinning strategy and third-harmonic UV laser cutting to achieve lower operating voltages, improved edge quality, and enhanced integration capabilities for advanced quantum photonics applications.

Original authors: Menotti Markovic, Lucia Oberndorfer, Tobias M. Krieger, Ievgen Brytavskyi, Barbara Lehner, Julia Freund, Vishnu Prakash Karunakaran, Matthias Domke, Dorian Gangloff, Christian Schimpf, Peter Michler

Published 2026-03-19

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Menotti Markovic, Lucia Oberndorfer, Tobias M. Krieger, Ievgen Brytavskyi, Barbara Lehner, Julia Freund, Vishnu Prakash Karunakaran, Matthias Domke, Dorian Gangloff, Christian Schimpf, Peter Michler, Simone Luca Portalupi, Michael Jetter, Rinaldo Trotta, Javier Martín-Sánchez, Armando Rastelli, Fadi Dohnal, Sandra Stroj

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 have a very special, super-sensitive piece of crystal. This crystal is like a tiny, super-strong spring that can wiggle and move when you zap it with electricity. Scientists call this PMN-PT, and it's the heart of a "piezo actuator."

Think of these actuators as tiny robotic muscles. Their job is to stretch or squeeze other tiny things attached to them. In this specific story, they are being used to squeeze "quantum dots" (which are like microscopic light bulbs that can emit special pairs of light particles called photons). By squeezing these dots just right, scientists can make them sing in perfect harmony, creating the building blocks for future quantum computers.

However, there's a problem. The old way of making these robotic muscles was like trying to carve a delicate watch out of a block of stone using a sledgehammer. It worked, but the pieces were too big, too thick, and the edges were a bit rough. To get the best performance, the scientists needed these muscles to be tiny, thin, and incredibly precise.

Here is how the scientists in this paper solved the problem, explained simply:

1. The Problem: Too Thick, Too Clunky

The original crystals were 300 microns thick (about the width of three human hairs). To make them move efficiently, you need to apply a lot of voltage (electric pressure). If you could make the crystal thinner, you wouldn't need as much voltage, and you could pack more of them onto a single chip.

But, if you try to carve a thin piece of glass with a regular tool, it shatters. These crystals are brittle, like a very hard piece of ice.

2. The Solution: The "Laser Scalpel"

The team used a femtosecond laser. Imagine a laser that is so fast it acts like a time-traveling scalpel. It cuts so quickly that it removes material before the heat can spread and crack the surrounding area. It's like using a super-fast knife to slice through a tomato without squishing it.

They developed a three-step "recipe" to build these mini-muscles:

Step A: The "Sandpaper" Trick (Local Thinning)

Instead of cutting the whole crystal down, they only wanted to thin out the specific spots where the muscle needs to bend.

The Analogy: Imagine you have a thick wooden block, and you want to make a thin, flexible hinge in the middle without breaking the sides. You would sand down just the middle area.
The Innovation: They used the laser to "sand" away the middle of the crystal, leaving a thin 100-micron spot surrounded by thick, sturdy walls. This makes the thin spot very easy to bend with low voltage, while the thick walls keep the whole thing from falling apart.
The Secret Sauce: They found that if you just sand in circles, you get a weird "dip" or crater at the edge. So, they changed their pattern, randomly changing the size of the circles as they sanded. This smoothed out the edges, ensuring the metal wires they add later would stick perfectly without breaking.

Step B: The "Gold Paint" (Electrodes)

Once the crystal was thinned, they needed to paint it with gold to conduct electricity.

The Challenge: Painting a flat table is easy. Painting a bowl with a curved bottom and steep sides is hard; the paint might get too thin or break in the corners.
The Fix: They sprayed a thin layer of gold over the whole thing. Then, they used the laser again to "erase" the gold in specific lines, creating separate electrical tracks. Because the laser is so precise, it could cut the gold without scratching the crystal underneath, creating perfect, isolated electrical paths even over the curved, thinned areas.

Step C: The "UV Razor" (Final Cutting)

This was the biggest breakthrough. In the past, they used a standard laser (green light) to cut the final shape. It was okay, but the edges were a bit fuzzy, and the cuts had to be wide (about 10–15 microns) to avoid breaking the crystal.

The Upgrade: They switched to a UV laser (ultraviolet light).
The Analogy: Think of the green laser as a blunt knife and the UV laser as a razor-sharp scalpel. The UV light is absorbed much better by the crystal, allowing it to cut with incredible precision.
The Result: They could now cut the crystal with a gap of only 5 microns (half the size of before!). The edges were razor-sharp and clean, with no cracks or chips.

Why Does This Matter?

By making these "robotic muscles" smaller, thinner, and sharper:

They use less power: You don't need a big battery to move them; a small voltage works perfectly.
They are stronger: You can squeeze the quantum dots harder, creating better light for quantum computers.
You can pack more: Because they are so small, you can fit many different quantum light sources on a single tiny chip, rather than just one.

The Bottom Line

The scientists took a brittle, hard-to-work-with crystal and used a high-tech, ultra-fast laser to carve it into a microscopic, high-performance machine. They figured out how to thin it out without breaking it, paint it with gold without smudging, and cut it with a razor-sharp edge. This paves the way for smaller, faster, and more powerful quantum devices that could one day power the internet of the future.

1. Problem Statement

Single-crystal Lead Magnesium Niobate-Lead Titanate (PMN-PT) is a superior piezoelectric material for Micro-Electro-Mechanical Systems (MEMS), particularly for strain-tuning semiconductor quantum dots (QDs) to generate entangled photon pairs. However, current fabrication methods face significant challenges:

Brittleness and Complexity: PMN-PT is brittle and requires micrometer-scale resolution, arbitrary geometries, and sharp edges, which are difficult to achieve with traditional methods like wet/dry etching or ion milling.
Size and Voltage Limitations: Previous femtosecond (fs) laser-fabricated actuators were limited to a thickness of 300 µm and a side length of 5 mm. This resulted in a minimum arm spacing of 30 µm and required high operating voltages to achieve sufficient strain.
Integration Density: Emerging quantum photonics applications require higher device densities and the integration of multiple quantum light sources on a single chip, necessitating miniaturized actuators with lower operating voltages.
Edge Quality: Standard laser cutting (using the second harmonic, 523 nm) often results in debris, cracks, and poor edge definition, which compromises mechanical stability and the ability to bond delicate semiconductor membranes.

2. Methodology

The authors developed a complete fabrication route relying exclusively on femtosecond laser ablation, eliminating the need for lithography or chemical etching. The process involves three main stages:

A. Localized Thinning (Cavity Generation)

Goal: Reduce the substrate thickness from 300 µm to 100 µm in specific active regions to lower the required operating voltage ( $V \propto t$ ).
Technique: The authors scanned the laser in circular hatch patterns. They investigated three scanning strategies to manage debris and edge quality:
1. Large-to-Small (LTS): Diameter decreases per repetition.
2. Small-to-Large (STL): Diameter increases per repetition.
3. Random (RND): Diameter varies randomly.
Optimization: The STL and RND strategies were found superior to LTS because they prevent debris accumulation at the cavity edges. The study also demonstrated that shifting the pattern during processing allows for the precise control of side-wall angles.

B. Conductive Layer Structuring

Goal: Create electrical contacts on the non-planar, thinned surfaces.
Process:
1. Deposition: A thin metal layer (Titanium for adhesion, Gold for conductivity) was vapor-deposited over the thinned cavity and the surrounding substrate.
2. Selective Ablation: The fs-laser (523 nm) was used to selectively ablate the metal layer to define individual actuator arms.
3. Verification: The sharp ablation threshold of gold vs. PMN-PT allowed for clean electrical isolation. The STL method was selected for the final device due to its superior edge geometry, which resulted in lower electrical resistance across the metal layer compared to the RND method.

C. UV Laser Cutting

Goal: Define the final actuator geometry with high precision and minimal thermal damage.
Innovation: The authors switched from the previously used Second Harmonic (SH, 523 nm) to the Third Harmonic (UV, 348 nm).
- Physics: The photon energy of 348 nm (3.56 eV) exceeds the bandgap of PMN-PT (3.1–3.3 eV), enabling linear absorption rather than the two-photon absorption used at 523 nm. This allows for more efficient energy deposition and reduced collateral thermal damage.
Strategy: A hybrid approach was used:
1. Pre-thinning: The geometry was pre-thinned using a line-hatch pattern to create a flat, high-quality cavity.
2. Final Cut: Equidistant lines were cut with higher fluence to sever the actuator arms.
Substrate Handling: The substrate was bonded to a holder using Polyvinyl Alcohol (PVA) to trap and remove debris during cutting.

3. Key Contributions

Complete Laser-Only Fabrication: Demonstrated a fully laser-based workflow (thinning, structuring, cutting) for complex 3D piezoelectric structures without chemical processing.
Local Thinning Strategy: Introduced a method to create localized thin regions (100 µm) within a thicker substrate (300 µm), enabling lower operating voltages while maintaining mechanical stability.
UV Laser Cutting: Proved that switching to 348 nm significantly improves edge quality, reduces the minimum cutting width, and eliminates the "cushion" effect and debris often seen with 523 nm cutting.
Scanning Pattern Optimization: Systematically analyzed scanning strategies (LTS, STL, RND) to control side-wall angles and minimize edge defects, identifying STL as the optimal method for electrical continuity over curved surfaces.

4. Results

Miniaturization: The new design achieved a minimum inter-arm spacing of 5 µm on the laser-exit side, a significant reduction from the 30 µm limit of previous 523 nm-based devices.
Edge Quality:
- UV vs. SH: UV-cut actuators exhibited sharp, well-defined edges on both the entrance and exit sides, free of the protrusions and breakouts seen in SH-cut devices.
- Surface Roughness: The STL scanning strategy produced smooth transitions from the bottom surface to the side walls, crucial for the conductivity of the deposited metal layers.
Electrical Performance:
- Resistance measurements confirmed that the metal layer remained conductive over the thinned, non-planar areas.
- Electrical isolation between actuator arms was successfully achieved, with resistance values exceeding 50 MΩ.
Mechanical Stability: The combination of local thinning and UV cutting produced devices capable of withstanding the mechanical stress required for multiaxial strain tuning.

5. Significance

This work establishes a versatile and scalable manufacturing route for next-generation piezoelectric actuators. By enabling miniaturization and lower operating voltages, these devices are critical for:

Quantum Photonics: Facilitating the integration of multiple individually addressed quantum emitters (QDs) on a single chip.
Advanced Strain Engineering: Allowing for higher and more directional strain values, which is essential for tuning the fine-structure splitting in biexciton–exciton cascades to generate highly entangled photon pairs.
Scalability: The all-laser process is compatible with complex geometries and high-throughput manufacturing, paving the way for advanced integrated quantum optical systems.