Silicon Photonics Bonding Interface with Trench Support Structures

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The integration of compound semiconductor materials with silicon waveguides in silicon photonics is challenged by weak bonding sites due to wide trenches and sharp ends at waveguide merges, leading to reduced bond yield and device failures.

Innovation Solution

Automated insertion of support structures within the trenches using EDA software to enhance bonding surfaces, ensuring structural integrity and minimizing optical interference.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If automated processes are used to handle complexity and scale in modern photonic circuits, then productivity and manufacturing efficiency are improved, but the ability to address complex bonding issues like wide trenches and sharp ends at waveguide merges deteriorates

Engineering Contradiction:
Improvemanufacturing efficiencyVSAvoidbond yield
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The EDA software performs preliminary identification and classification of trench types (wide trenches, sharp ends at waveguide merges) during the design phase, before manufacturing occurs. Support structures are automatically inserted in advance at locations where bonding issues are predicted, ensuring that when manufacturing happens, the structures are already in place to prevent bonding failures.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The EDA tool automatically analyzes the photonic circuit layout, identifies problematic trench regions, and inserts appropriate support structures without manual intervention. The system self-evaluates the design, detects bonding-risk areas, and applies corrections autonomously, maintaining high productivity while improving bond yield.

Inventive Principle:
Principle #25Self-service

2Reliability

If support structures are inserted to provide additional bonding surfaces, then bond yield and structural integrity are improved, but device complexity increases

Engineering Contradiction:
Improvebond yieldVSAvoidstructural complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

Instead of uniformly adding support structures throughout the entire device, the EDA software applies local quality by inserting support structures only in specific regions where wide trenches and sharp ends are detected. The support structures are localized to the exact areas needing reinforcement, minimizing overall structural complexity while maximizing bonding reliability where it matters most.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The support structures are segmented into discrete, identifiable elements that can be automatically inserted and managed. Each support structure is a separate feature with defined geometry and location, allowing the EDA tool to handle them individually and systematically rather than dealing with a monolithic complex structure.

Inventive Principle:
Principle #1Segmentation

3Manufacturing precision

If manual methods are used to identify and address bonding issues, then customization and precision can be improved, but productivity and scalability deteriorate

Engineering Contradiction:
Improvebonding precisionVSAvoidmanufacturing throughput
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The manual mechanical process of visually inspecting and manually adding support structures is replaced with an automated computational system. The EDA software uses algorithmic analysis to detect trench geometries and automatically generates support structure placements, substituting human expertise with automated computational methods that scale efficiently.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The EDA software creates a digital copy of the photonic circuit layout and analyzes it computationally to identify bonding issues. Support structures are designed and positioned in the digital model first, then automatically transferred to the manufacturing process, eliminating the need for manual interpretation and adjustment of physical or graphical designs.

Inventive Principle:
Principle #26Copying

Data Source

PatentUS20260079294A1Automated silicon photonics bonding interface enhancement
Publication Date: 2026.03.19 OPENLIGHT PHOTONICS INC
  • US20260079294A1 patent drawing
  • US20260079294A1 patent drawing
  • US20260079294A1 patent drawing

AI summary

A silicon photonic device includes a substrate formed from a silicon-containing material and patterned to comprise a first waveguide and a second waveguide defining a first trench extending between the first waveguide and the second waveguide. The first waveguide and second waveguide have upper surfaces exposed for bonding to an epitaxially grown layer. The first trench being exposed to the epitaxially grown layer. A support structure is formed within the first trench and extends upward to an upper surface at a height of the upper surfaces of the first waveguide and second waveguide. The support structure is optically non-functional.