Correlation between pattern density and linewidth variation in silicon photonics waveguides.

Area of Science:

• Photonic Integrated Circuits
• Semiconductor Device Fabrication
• Lithography

Background:

• Local pattern density in mask layouts can influence fabrication processes like plasma etching and planarization.
• Variations in waveguide dimensions can affect the performance of photonic integrated circuits.

Purpose of the Study:

• To investigate the correlation between measured silicon waveguide width and local mask pattern density.
• To determine the impact of pattern density on waveguide thickness.
• To identify optimal layout strategies for minimizing width variations.

Main Methods:

• Fabrication of silicon waveguides using 193 nm lithography.
• Measurement of waveguide width and thickness using an optical test circuit.
• Statistical analysis of 5841 sampled sites across a fabricated wafer.
• Correlation analysis as a function of pattern density within varying radii of influence.

Main Results:

• A significant correlation (coefficient of 0.57) was found between intra-die systematic waveguide width variation and pattern density within a 200 µm radius.
• No correlation was observed between pattern density and intra-die systematic waveguide thickness variation.
• The findings highlight the critical role of local pattern density in controlling waveguide linewidth.

Conclusions:

• Local pattern density is a key factor influencing systematic width variations in silicon waveguides fabricated with 193 nm lithography.
• Understanding this correlation allows for prediction of photonic circuit yield.
• Layout optimization based on pattern density can mitigate width variations and improve device performance.