← Latest papers
🔬 optics

Light coupling to photonic integrated circuits using optimized lensed fibers

This paper presents a comprehensive co-optimization strategy that synergistically refines the geometries of inverse tapers and lensed fibers by incorporating the fiber's non-Gaussian emission profile, achieving coupling efficiencies exceeding 80% and providing scalable manufacturing solutions for silicon nitride photonic integrated circuits.

Original authors: Dengke Chen, Zeying Zhong, Sanli Huang, Jiahao Sun, Sicheng Zeng, Baoqi Shi, Yi-Han Luo, Junqiu Liu

Published 2026-02-05
📖 4 min read☕ Coffee break read

Original authors: Dengke Chen, Zeying Zhong, Sanli Huang, Jiahao Sun, Sicheng Zeng, Baoqi Shi, Yi-Han Luo, Junqiu Liu

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 trying to pour water from a large, wide-mouthed bucket (an optical fiber) into a very narrow, delicate drinking straw (a photonic chip). If you just dump the water in, most of it will splash out and be wasted. In the world of light and computers, this "spillage" is called coupling loss, and it's a huge problem for making fast, efficient devices like those used in data centers and AI.

For years, scientists have tried to solve this by shaping the end of the straw (the chip) into a funnel, called an inverse taper, to catch the water better. They also use a special "lensed fiber" (a fiber with a tiny, curved glass tip) to focus the water stream.

However, there was a major flaw in how people designed these systems.

The "Perfect Shape" Myth

Until now, engineers assumed the lensed fiber acted like a perfect, smooth spotlight (a "Gaussian beam"). They thought the light came out in a neat, symmetrical cone. Based on this assumption, they designed the chip funnels to match this perfect shape.

The Reality Check:
This paper reveals that the assumption was wrong. Because the tip of the lensed fiber is so tiny (smaller than the wavelength of light), the light doesn't come out perfectly smooth. It's actually a bit messy and uneven, with "tails" of light that the old models ignored. It's like assuming a garden hose sprays a perfect circle of water, when in reality, the nozzle creates a slightly irregular, splashing pattern.

The New Strategy: "Co-Optimization"

The researchers realized that if you design the chip funnel based on a fake, perfect light shape, it won't catch the real, messy light very well.

So, they did something new: They stopped guessing and started measuring.

  1. Real Scans: They took super-high-resolution microscope photos of the actual fiber tips to see their true, bumpy shapes.
  2. Real Simulations: They fed these real, messy shapes into their computer models instead of the fake "perfect" ones.
  3. The Match: They then redesigned the chip funnels to perfectly match the actual messy light coming from the fiber.

Think of it like tailoring a suit. Before, they were making suits based on a mannequin's perfect, idealized measurements. Now, they are taking a tape measure to the actual person, seeing exactly where their shoulders are slightly uneven, and cutting the fabric to fit that specific person perfectly.

The Results

By matching the real fiber shape to the real chip shape, they achieved a massive improvement:

  • Efficiency: They managed to get over 80% of the light from the fiber into the chip. That's like getting 8 out of 10 drops of water from the bucket into the straw, with almost no splash.
  • Forgiveness: They also found that using a slightly larger "spot size" on the fiber makes the connection more forgiving. If you bump the fiber slightly (misalignment), the light still gets in. It's like having a wider funnel that catches the water even if your hand shakes a little.
  • Broadband: This works across a wide range of colors (frequencies) of light, which is crucial for high-speed data.

Why It Matters

The paper concludes that this "co-optimization" (fixing both the fiber and the chip together) is the key to making photonic chips that can be mass-produced in standard factories (CMOS foundries).

This isn't just about better light pipes; it's about enabling the next generation of technology. The authors specifically mention that efficient light coupling is critical for:

  • Optical interconnects: Connecting computer chips faster.
  • Neuromorphic computing: Building computer brains that work like human neurons.
  • Quantum information processing: Handling delicate quantum states for future supercomputers.
  • Data centers and AI: Revolutionizing how we process massive amounts of data.

In short, the paper says: "Stop assuming the light is perfect. Measure the real light, build the chip to match it, and you'll get a system that works incredibly well."

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →