Too Big, Too Small, Too $O_2$: The Pandoro Effect from Oxygen Gradients in Tomographic Volumetric Additive Manufacturing

This study identifies and resolves the "Pandoro effect"—a truncated-cone printing artifact in Tomographic Volumetric Additive Manufacturing caused by vertical oxygen gradients in thermoreversible hydrogels—by combining a novel differentiable photochemical optimization model with process interventions like eliminating the air-resin interface to enable reproducible, cell-laden bioprinting.

The Gist

The Big Idea: 3D Printing Without Layers

Imagine you want to build a complex 3D sculpture, like a delicate birdcage.

• Old Way (Layer-by-Layer): You build it like a cake, stacking one thin layer of frosting on top of another. It's slow, and you can't easily make overhangs without support structures.
• The New Way (TVAM): Imagine putting your sculpture inside a jar of liquid jelly. You spin the jar and flash a light from different angles. The light hits the jelly just right, and poof—the whole object solidifies instantly in mid-air, like a ghost appearing. This is Tomographic Volumetric Additive Manufacturing (TVAM). It's fast, layerless, and amazing for making things like artificial tissues.

The Problem: The "Pandoro" Effect

The researchers found a weird glitch when using a special type of "jelly" (gelatin-based resin) that hardens when cooled.

Sometimes, instead of printing a perfect cylinder, the object comes out looking like a Pandoro (a famous Italian star-shaped Christmas cake that is wider at the bottom and narrower at the top).

• What happens? The bottom of the object gets too hard and grows too big. The top stays soft and doesn't grow enough.
• The Result: The object is truncated, distorted, and useless.

The Culprit: The "Oxygen Ghost"

Why does this happen? It's all about Oxygen.

Think of oxygen as a "spoiler" for the printing process. In this type of printing, light is supposed to turn the liquid jelly into solid plastic. But oxygen acts like a bodyguard that stands in front of the light, blocking the reaction. If there is too much oxygen, the jelly won't harden.

The Recipe for Disaster:

1. Heating: To mix the ingredients, the researchers heat the resin to 40°C (104°F). Heat makes oxygen "flee" the liquid (like bubbles escaping a warm soda). The resin becomes "oxygen-poor."
2. Cooling: They pour this warm, oxygen-poor liquid into a cold jar and let it cool down to 0°C (32°F) to turn it into a gel.
3. The Sneak Attack: As the liquid cools, it wants to soak up oxygen again. But since the jar is sitting still, the oxygen can only sneak in from the top (where the air touches the liquid).
4. The Gradient:
   • Top of the jar: Soaked in fresh oxygen (The "Oxygen Ghost" is strong here). The light can't harden the jelly easily.
   • Bottom of the jar: Still oxygen-poor (The "Ghost" is weak here). The light hardens the jelly too fast.

The Analogy: Imagine trying to paint a wall.

• At the bottom, the paint dries instantly because the air is dry.
• At the top, the paint stays wet because someone is constantly spraying water on it.
• If you try to build a tower, the bottom grows tall and thick, while the top stays short and weak. That's the Pandoro effect.

The Three Solutions

The team came up with three clever ways to fix this "Oxygen Ghost" problem.

1. The "Smart Software" Fix (Optimization)

Instead of using the same light pattern for the whole jar, they used a computer to calculate exactly how much oxygen is at the top vs. the bottom.

• How it works: The computer tells the printer: "Hey, the top has a lot of oxygen blockers! Give it extra bright light to punch through them. The bottom has few blockers, so keep the light dimmer so it doesn't overcook."
• Result: The object hardens evenly, just like a perfect cylinder.

2. The "No Air" Fix (Engineering)

If oxygen comes from the air, why not remove the air?

• How it works: They filled the jar completely to the very brim and sealed it tight with a special cap, leaving zero air bubbles.
• Result: With no air to sneak in oxygen from the top, the oxygen levels stay low and even everywhere. No Pandoro cake!
• Note: This wastes a lot of expensive liquid resin, so they made special tiny caps to use less.

3. The "Argon Bubble" Fix (Atmosphere Control)

If you can't remove the air, change the air.

• How it works: Before sealing the jar, they sprayed it with Argon gas (a gas used to keep wine fresh). Argon pushes the oxygen out.
• Result: The jar is now filled with "safe" gas instead of "spoiler" oxygen. This buys them about an hour of perfect printing time before the oxygen slowly leaks back in.

Why This Matters

This isn't just about fixing a weird cake shape.

• For Medicine: This technology is used to print living tissues (like skin or cartilage) with cells inside. If the shape is wrong, the tissue won't work.
• For the Future: The researchers updated their open-source software (Dr.TVAM) so anyone can use these "smart" corrections. They also proved that these fixes don't hurt the living cells inside the resin.

In a nutshell: They found out that heat and cold create an invisible oxygen gradient that ruins 3D prints. By using smart math, sealing jars tight, or swapping the air, they fixed the problem, making it possible to print perfect, complex biological structures quickly and reliably.

Technical Summary

1. Problem Statement: The "Pandoro Effect"

The authors identify a recurrent printing artifact in Tomographic Volumetric Additive Manufacturing (TVAM) when using thermoreversible hydrogels (specifically gelatin-based resins like GelMA).

• Phenomenon: Printed objects exhibit a truncated-cone distortion (resembling the Italian Christmas cake "Pandoro"), where the bottom of the structure is over-polymerized (too wide) and the top is under-polymerized (too narrow or missing).
• Root Cause: The artifact arises from a vertical oxygen concentration gradient within the resin vial, driven by the thermal history of the resin preparation:
  1. Heating: Resins are heated (~40°C) to dissolve gelatin. According to Henry's Law, oxygen solubility decreases at higher temperatures, causing dissolved oxygen to outgas.
  2. Cooling: The resin is cooled (~0°C) to induce thermal gelation. Oxygen solubility increases significantly at lower temperatures.
  3. Diffusion: Because the resin is now a gel, oxygen cannot mix; it must diffuse from the air-resin interface (headspace) into the bulk. This creates a time-dependent gradient where oxygen concentration is highest at the top and lowest at the bottom.
• Impact: Oxygen acts as a radical scavenger (inhibitor) in free-radical polymerization. The gradient causes delayed polymerization at the top (high oxygen) and premature polymerization at the bottom (low oxygen), leading to geometric failure. This effect worsens over time (peaking around 90 minutes) until equilibrium is reached (approx. 24 hours), at which point the entire volume is over-inhibited.

2. Methodology

The study employs a multi-faceted approach combining experimental validation, numerical modeling, and three distinct mitigation strategies.

A. Experimental Validation

• Materials: Used Gelatin-methacryloyl (GelMA) and Thiol-norbornene functionalized gelatin (Gel-SH/NB).
• Protocol: Resins were heated to 40°C, mixed, poured into vials, and cooled on ice. Prints were attempted at various time intervals (5 min to 24 h) post-cooling.
• Observation: Measured structural heights and geometries confirmed the progressive development of the truncated-cone shape as storage time increased.

B. Numerical Modeling

• Physics: Modeled oxygen diffusion using Fick's Second Law in a 1D vertical domain.
• Parameters: Used diffusion coefficients for water (approx. 960 µm²/s) and Henry's law constants to simulate the transition from 40°C (low saturation) to 0°C (high saturation).
• Goal: To predict the spatiotemporal evolution of the oxygen gradient and correlate it with polymerization kinetics.

C. Mitigation Strategies

The authors propose and validate three solutions:

1. Algorithmic Correction (Optimization-based):
   • Developed a coupled ray-optical and photochemical optimization model integrated into the open-source Dr.TVAM platform.
   • Unlike standard threshold models, this framework explicitly simulates spatially heterogeneous inhibitor concentrations and dynamic diffusion.
   • The inverse solver predicts local inhibition gradients and compensates by projecting higher light doses to oxygen-rich regions (top) and lower doses to oxygen-poor regions (bottom).
2. Engineering Intervention (Interface Elimination):
   • Eliminated the air-resin interface by filling vials completely and sealing them with custom caps to prevent oxygen diffusion from the headspace.
3. Atmospheric Control (Inert Gas Flushing):
   • Flushed the vial headspace with Argon (99%) before sealing to reduce the partial pressure of oxygen, thereby slowing the diffusion gradient formation.

3. Key Results

• Mechanism Confirmation: Experiments confirmed that the Pandoro effect is driven by oxygen gradients, not light inhomogeneity. Even "oxygen-insensitive" thiol-ene systems (Gel-SH/NB) exhibited the effect, though less severely, because oxygen still competes for radicals and delays propagation.
• Model Accuracy: The numerical simulation of oxygen diffusion accurately predicted the time-dependent gradient and the resulting print distortions.
• Strategy Efficacy:
  • Optimization: Successfully corrected prints at 60 and 120 minutes, restoring target geometry without physical changes to the setup.
  • Engineering: Completely eliminated the effect for up to 24 hours by removing the air interface.
  • Atmosphere: Extended the defect-free printing window to ~60 minutes. (Note: Prolonged storage in Argon led to a "reverse Pandoro effect" due to oxygen undersaturation).
• Bioprinting Compatibility: All strategies were tested with cell-laden resins (2 × 10⁶ cells/mL).
  • All three methods produced defect-free constructs.
  • Cytocompatibility: Cell viability remained high (>97% at Day 0, >80% at Day 5) across all conditions, with no significant difference compared to non-mitigated controls.

4. Key Contributions

1. Identification of the Pandoro Effect: Systematically characterized a previously undocumented artifact in TVAM caused by thermal cycling and oxygen diffusion.
2. Advanced Computational Framework: Introduced a differentiable, coupled ray-optical and photochemical optimization model. This is a significant advancement over conventional threshold-based approaches, as it accounts for spatially varying inhibitor concentrations and reaction-diffusion dynamics.
3. Open-Source Implementation: Integrated these capabilities into the Dr.TVAM platform, making advanced polymerization modeling accessible to the community.
4. Practical Guidelines: Provided actionable, tiered solutions (software, engineering, and atmospheric) for reliable bioprinting with thermoreversible hydrogels.

5. Significance

• Reproducibility: This work resolves a major barrier to the reproducibility of TVAM with gelatin-based bioresins, which are widely used in tissue engineering due to their biocompatibility and thermal reversibility.
• Scalability: The proposed strategies allow for the reliable printing of centimeter-scale, complex 3D constructs containing living cells, which is critical for clinical translation.
• Generalizability: While focused on gelatin, the "Pandoro effect" applies to any photoresin undergoing thermal cycles between preparation and printing. The mitigation strategies (especially the computational model) offer a general framework for handling inhibitor gradients in volumetric additive manufacturing.
• Future Directions: The explicit modeling of reaction-diffusion kinetics opens new avenues for enhancing spatial resolution and accounting for metabolic oxygen consumption by encapsulated cells in future bioprinting workflows.