Towards manufacturing high uniformity polysilicon circuits through TFT contact barrier engineering

Area of Science:

• Materials Science
• Electrical Engineering
• Semiconductor Physics

Background:

• The proliferation of Internet of Things (IoT) devices necessitates high-performance polysilicon technology for advanced sensing and control circuits.
• Polycrystalline silicon's inherent non-uniformity poses challenges for scaled transistors, impacting device performance consistency.
• Uniform electrical characteristics across large areas are crucial for the commercial viability of new polysilicon-based applications.

Purpose of the Study:

• To investigate the impact of engineered potential barriers at the source of polysilicon thin-film transistors (TFTs).
• To achieve highly uniform device performance, specifically on-current, in polysilicon TFTs.
• To understand the role of grain boundaries and current injection mechanisms in device behavior for optimized circuit design.

Main Methods:

• Fabrication of polysilicon thin-film transistors with deliberately engineered source potential barriers.
• Characterization of device performance, focusing on on-current uniformity and kink effect reduction.
• Technology Computer-Aided Design (TCAD) simulations to analyze grain boundary effects and current injection mechanisms.

Main Results:

• Achieved highly uniform on-current with less than 8% device-to-device variation, accounting for material and geometric factors.
• Demonstrated significant reduction in kink effect and high intrinsic gain over a broad drain voltage range (2-20 V).
• TCAD simulations identified critical grain boundary positions (~150 nm in the source area) influencing current injection, though with marginal impact due to low probability.

Conclusions:

• Engineered source potential barriers in polysilicon TFTs lead to substantial improvements in electrical characteristic uniformity.
• The contact-controlled architecture effectively mitigates kink effects and enhances intrinsic gain.
• This research provides critical insights for designing reliable signal conversion and gain circuits for flexible, low-power sensors without complex compensation.