Toward three-dimensional microelectronic systems: directed self-assembly of silicon microcubes via DNA surface...

Area of Science:

• Materials Science
• Nanotechnology
• Biotechnology

Background:

• Biological processors like the brain exhibit intelligent processing via massive parallel interconnectivity.
• Current artificial silicon microprocessors lack intelligent processing due to limited interconnectivity.
• Three-dimensional (3D) structures are essential for realizing massively parallel interconnectivity in artificial systems.

Purpose of the Study:

• To achieve self-assembly of silicon-based building blocks into 3D structures for artificial intelligence.
• To demonstrate the feasibility of creating 3D artificial processors through bottom-up self-assembly.
• To electrically characterize the self-assembled 3D structures.

Main Methods:

• Functionalization of silicon substrates and microcubes with DNA oligonucleotides via self-assembling monolayers.
• Fabrication of 3 μm silicon microcubes using silicon-on-insulator wafers, reactive ion etching, and wet etching.
• Characterization using contact angle measurements, ellipsometry, atomic force microscopy, and fluorescence microscopy.
• Demonstration of microcube self-assembly on patterned surfaces and into aggregates via DNA hybridization.

Main Results:

• Successful self-assembly of DNA-functionalized silicon microcubes on patterned surfaces with 88% correct placement.
• Formation of microcube aggregates (dimers) with a yield of approximately 44%.
• Electrical characterization of the formed dimers using probe tips within a scanning electron microscope.

Conclusions:

• Self-assembly of silicon microcubes using molecular recognition (DNA) is a viable first step towards 3D artificial processors.
• This approach enables the creation of massively parallel interconnectivity, crucial for artificial intelligence.
• The study lays the groundwork for fabricating artificially intelligent systems through bottom-up 3D assembly.