Researchers 3D Print Ultra-Low Loss Photonic Chip Packaging — Achieves Record 0.78 dB Total Loss

Researchers from Heidelberg University have demonstrated a groundbreaking 3D printed packaging solution for photonic integrated circuits (PICs), achieving record-low optical losses. The work, led by Erik Jung and collaborators including Wolfram Pernice from University of Münster, addresses one of the major challenges in scaling photonic computing technologies.

The Challenge of Photonic Packaging

Photonic Integrated Circuits offer transformative potential for computing systems, enabling high-bandwidth neuromorphic processors and facilitating low-decoherence quantum computing on a chip scale platform. However, the development of robust and scalable optical packaging solutions has remained a major challenge.

Efficient fiber-to-chip coupling is essential for minimizing loss and enabling high optical bandwidth - key requirements for photonic computing. Traditional packaging methods struggle to achieve both low loss and high bandwidth simultaneously.

Plug-and-Play 3D Printed Solution

The research team introduced a novel plug-and-play solution for fiber-to-PIC connections using female multi-fiber termination push-on cables. They additively fabricated the alignment counterpart on the circuit via two-photon polymerization, a high-resolution 3D printing technique.

Key achievements:

• Peak transmission: -0.41 dB
• Broadband performance: Losses below 0.55 dB across the 1500-1600 nm range
• Total packaging loss: 0.78 dB (record for passive out-of-plane packaging)
• Bandwidth: Greater than 100 nm (-0.5 dB bandwidth)

Validated With 17-Port Photonic Circuit

The team characterized the reproducibility of this out-of-plane packaging solution by interfacing with a 17-port photonic circuit designed for incoherent photonic computing. The high bandwidth of the packaging is crucial to couple the full 100 nm bandwidth that modern photonic chips can utilize.

What This Means for Computing

This breakthrough has significant implications for the future of computing. Low-loss photonic packaging is essential for:

• Neuromorphic computing: Brain-inspired processors that could outperform traditional AI hardware
• Quantum computing: Scaling quantum systems requires efficient optical connections
• Data center optics: Higher bandwidth and lower power optical interconnects

The use of additive manufacturing (3D printing) for this critical component demonstrates how the technology is enabling new frontiers in electronics packaging beyond traditional 3D printing applications.