Physical model for recognition tunneling.

Area of Science:

• Nanotechnology
• Molecular Biology
• Computational Chemistry

Background:

• Recognition tunneling (RT) is an emerging technique for single-molecule identification.
• RT relies on specific chemical interactions between target molecules and functionalized electrodes.
• Potential applications include DNA and protein sequencing.

Purpose of the Study:

• To theoretically investigate fundamental aspects of recognition tunneling.
• To validate RT current magnitudes using computational models.
• To explore the role of molecular fluctuations in RT signals and molecular identification.

Main Methods:

• Utilized multiscale theoretical modeling, including all-atom and coarse-grained DNA models.
• Performed non-equilibrium Green's function calculations on solvated all-atom models.
• Applied machine learning algorithms to analyze fluctuation frequency characteristics.

Main Results:

• RT current magnitudes are consistent with all-atom non-equilibrium Green's function calculations.
• Brownian fluctuations in hydrogen bond lengths generate current spikes observed experimentally.
• Frequency characteristics of fluctuations can be used for machine learning-based molecular identification.

Conclusions:

• Provides a theoretical foundation for understanding recognition tunneling mechanisms.
• Demonstrates that molecular fluctuations are key to RT signal generation and identification.
• Establishes a theoretical basis for using machine learning with RT for single-molecule analysis.