Microsphere-assisted hyperspectral imaging: super-resolution, non-destructive metrology for semiconductor devices

Area of Science:

• Materials Science and Engineering
• Optical Metrology
• Semiconductor Manufacturing

Background:

• Shrinking semiconductor devices necessitate advanced metrology for in-cell critical dimensions (CD).
• Traditional methods like test element group (TEG) and optical critical dimension (OCD) lack the resolution for fine, repetitive patterns.
• Conventional OCD's large spot size (~25 μm) limits detailed in-cell structural analysis.

Purpose of the Study:

• To develop a novel small-spot, non-destructive metrology method for semiconductor in-cell CD.
• To address the limitations of conventional metrology in analyzing complex 3D semiconductor structures.
• To demonstrate the efficacy of microsphere-assisted hyperspectral imaging (MAHSI) for enhanced resolution.

Main Methods:

• Implementation of a microsphere-assisted hyperspectral imaging (MAHSI) system.
• Utilizing microsphere-assisted super-resolution imaging to surpass the diffraction limit.
• Integrating hyperspectral imaging (400-790 nm) for spectral capture at each pixel.

Main Results:

• Achieved an optical resolution of 66 nm with a 5.6 μm × 5.6 μm field of view.
• Obtained a pixel resolution (spot size) of 14.4 nm/pixel with 450X magnification.
• Enabled measurement of local uniformity in critical areas like DRAM cell corners and edges.

Conclusions:

• The MAHSI system provides a breakthrough in small-spot, super-resolution optical metrology for semiconductor devices.
• This method overcomes the resolution limitations of conventional OCD, enabling detailed analysis of complex 3D structures.
• Represents the first global implementation of microsphere-assisted hyperspectral imaging for advanced semiconductor metrology challenges.