🔬 optics

Variable dose slicer for refractive index engineering in two-photon polymerization

This paper introduces an open-source variable dose slicer and a corresponding calibration methodology to overcome software and measurement barriers, enabling precise refractive index engineering and expanded design freedom in two-photon polymerization through point-wise dose control.

Original authors: Michal Ziemczonok, Koen Vanmol

Published 2026-02-13

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Michal Ziemczonok, Koen Vanmol

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 a chef trying to bake a cake. Usually, when you bake a cake, the entire thing is made of the same batter. If you want a chocolate cake, you use chocolate batter everywhere. If you want a vanilla cake, you use vanilla batter everywhere. To make a cake with both chocolate and vanilla, you have to bake two separate cakes and glue them together, or use a very complicated mold.

Two-Photon Polymerization (TPP) is like a magical 3D printer that builds tiny, microscopic structures (smaller than a human hair) using a special liquid "batter" (resin) and a laser.

This paper introduces a new way to use this printer. Instead of just printing one type of "batter" everywhere, the authors have created a tool that lets them change the "flavor" of the batter at every single point inside the 3D object.

Here is the breakdown of their breakthrough in simple terms:

1. The Problem: The "All-or-Nothing" Printer

Normally, when you use this laser printer, you have to pick a setting: "High Power" or "Low Power."

High Power: The batter hardens completely. It becomes strong, dense, and has a specific "optical personality" (called the Refractive Index, or how much it bends light).
Low Power: The batter only hardens a little bit. It's softer, less dense, and bends light differently.

The problem was that existing software couldn't tell the printer to say, "Harden this tiny spot a little, but this next spot a lot." It was like trying to paint a picture with a brush that could only be dipped in one color at a time. To get different colors, you had to stop, change the paint, and start over, which was slow and messy.

2. The Solution: The "Variable Dose Slicer"

The authors built a new piece of software (an open-source slicer) that acts like a smart conductor for an orchestra.

The Analogy: Imagine a 3D printer is a robot arm holding a paintbrush. Usually, the robot moves the brush and paints the whole wall one color. This new software tells the robot: "Paint this corner with a light touch, this middle with a heavy touch, and this edge with a medium touch."
The Result: The printer can now create a single object where the inside is "soft" and the outside is "hard," or where the top bends light one way and the bottom bends it another way. They call this (3+1)D printing: 3 dimensions of shape, plus 1 dimension of material "flavor" (density).

3. The Secret Sauce: Measuring the "Flavor"

You can't just guess how much power to use to get a specific "flavor." You need a recipe.

The Calibration: The authors created a special test structure (like a step-ladder made of resin) to measure exactly how much the light bends when they use different laser powers.
The Magic Trick: They used a technique called Digital Holographic Microscopy. Think of this as a super-precise ruler that doesn't touch the object. It measures how much the object slows down light passing through it.
The Outcome: They created a perfect map that says: "If you want the light to bend by X amount, use Y amount of laser power." This map is so accurate that they can predict the properties of the material before they even print it.

4. What Can You Do With This? (The Cool Applications)

Once you can control the "flavor" of every tiny pixel in a 3D object, you can do amazing things:

The "Invisible" Lens: You can print a lens where the density changes smoothly from the center to the edge (like a gradient). This creates a perfect lens without any sharp edges, which is great for tiny cameras or fiber optics.
The "Fake" Cell: Scientists often need to test microscopes on fake cells (phantoms) to see if they work. Usually, fake cells look nothing like real ones. With this tool, they printed a fake worm and a fake cell where the inside looks exactly like a real biological cell because they matched the way light bends inside the fake one. It's like creating a "digital twin" of a cell in plastic.
Smoothing the Stairs: 3D printers often leave "staircase" marks on curved surfaces. This new method allows the printer to adjust the size of the "dots" (voxels) as it moves, smoothing out the stairs without needing to print slower.
Robots that Move: By making parts of a structure soft and other parts hard, they can create tiny robots that wiggle or bend when exposed to light or heat, just like muscles.

5. Why This Matters

Before this paper, making these complex, multi-material objects was like trying to write a book where you had to stop and change your pen every few words. It was slow, difficult, and required expensive, custom software.

Now, the authors have given everyone a free, open-source toolkit that lets any researcher:

Design a 3D object.
Tell the computer exactly how "dense" every part should be.
Print it in one go.

They have turned a rigid, black-and-white printing process into a full-color, high-definition experience. This opens the door for better medical imaging, smarter micro-robots, and tiny optical devices that were previously impossible to build.

In short: They taught the 3D printer how to paint with a rainbow of material densities, not just one color, and they gave everyone the map to do it themselves.

1. Problem Statement

Two-photon polymerization (TPP) is a high-resolution 3D printing technique capable of creating complex microstructures. A critical parameter in TPP is the Degree of Conversion (DC) of the photoresin, which is determined by the laser exposure dose. The DC directly influences material properties such as the Refractive Index (RI), stiffness, density, and chemical composition.

Despite the potential of "variable DC" (or (3+1)D printing) to create gradient-index (GRIN) optics, bio-mimetic phantoms, and stimulus-responsive structures, this capability is underutilized due to three main barriers:

Lack of Standardized Formats: Standard 3D printing formats (e.g., .STL, .STEP) only encode geometric information, not point-wise material properties like DC or RI.
Calibration Complexity: There is a lack of accessible, high-accuracy methodologies to calibrate the relationship between laser power, DC, and the resulting RI.
Stability Concerns: Low-DC regions are mechanically unstable and photosensitive, requiring careful handling and validation.

2. Methodology

The authors developed a comprehensive toolkit to overcome these barriers, consisting of three main components:

A. Open-Source MATLAB Slicer

Function: A custom slicer that accepts a 3D array of desired RI values (or voxelized power maps) as input.
Output: Generates a .gwl file (compatible with Nanoscribe TPP systems) containing 4D coordinates $(x, y, z, \text{power})$ . This allows the laser power to be modulated at every point in the 3D volume.
Features:
- Supports arbitrary 3D RI distributions.
- Includes features for structure rotation (aligning hatching directions to minimize staircasing or optimize flow during multi-material printing).
- Enables grayscale lithography by adjusting voxel size and power to approximate continuous surface profiles without increasing slicing density.
- Allows for parametric sweeps and custom scanning trajectories.

B. Refractive Index Calibration Methodology

To accurately map laser power to RI, the authors devised a non-destructive measurement protocol using Digital Holographic Microscopy (DHM):

Sample Design: A "step structure" with a protective high-DC shell and an internal chamber filled with resin polymerized at varying power levels.
Double-Immersion Technique: The sample is measured in two different immersion liquids with known RIs ( $n_1$ and $n_2$ ).
Calculation: Using the phase shift ( $\Delta\phi$ ) measured in both media and the known height difference ( $\Delta d$ ), the resin RI is calculated via:
$RI_{resin} = \frac{\Delta\phi_1 RI_{immersion2} - \Delta\phi_2 RI_{immersion1}}{\Delta\phi_1 - \Delta\phi_2}$
This eliminates the need for precise geometric height measurements of the fragile sample.

C. Metrological Validation

Voxel Calibration: Scanning Electron Microscopy (SEM) was used to correlate laser power with voxel width.
Aging Studies: Samples were monitored over 303 days to assess RI stability under ambient conditions.

3. Key Contributions

Software Tool: The release of an open-source MATLAB slicer that bridges the gap between 3D design and variable-dose TPP fabrication, enabling point-wise control of material properties.
High-Precision Calibration: A robust, non-destructive method to determine RI for any DC with high accuracy ( $\approx 0.1\%$ ) and precision ( $\approx 0.005\%$ ).
Demonstrated Applications:
- Bio-mimetic Phantoms: Creation of 3D RI-matched models of biological cells (eukaryotic cells) and organisms (C. elegans), including internal organelles, for testing optical diffraction tomography.
- Grayscale Lithography: Achieving smoother surface profiles (comparable to 6x finer slicing) without increasing print time.
- Siemens Star Fabrication: Printing high-resolution radial structures by optimizing hatching direction and power to minimize the "staircase effect."

4. Results

RI Calibration: A linear relationship was established between laser power (0–60%) and RI for IP-Dip2 resin. The RI range achieved was approximately 0.03 (from ~1.50 to ~1.53).
Measurement Accuracy: The expanded uncertainty ( $k=2$ ) for RI measurements ranged from 0.0009 to 0.0015.
Stability: The RI of the printed structures showed excellent stability over 303 days, with an average aging rate of $5.66 \times 10^{-6}$ per day and no clear dependency on DC.
Fabrication Success:
- The cell phantom successfully mimicked the 3D RI distribution of a real cell, allowing for benchmarking of imaging systems.
- The Siemens star achieved a half-pitch tangential resolution of 250 nm with an aspect ratio of 6:1.
- Grayscale surfaces showed quality comparable to 50 nm slicing steps while using 300 nm slicing steps, significantly reducing fabrication time.

5. Significance and Impact

This work provides a complete "toolbox" for 3D Refractive Index Engineering, moving TPP beyond simple geometric fabrication into functional material design.

Democratization: By providing open-source software and a standardized calibration method, the authors lower the entry barrier for researchers to utilize variable DC.
Metrology & Imaging: The ability to print precise RI phantoms allows for the rigorous benchmarking of optical imaging systems (e.g., OCT, tomography) and the development of "digital twins" for manufacturing.
New Design Paradigms: It enables the creation of structures where function is derived from density heterogeneity rather than just geometry, opening doors for advanced metamaterials, micro-actuators, and self-assembling robots.
Future Outlook: The authors suggest that while current limitations exist (e.g., sensitivity to stray light, file size constraints), the integration of these tools into commercial TPP workflows will unlock new inventions in micro-optics and biophotonics.