Molecular electronic devices using quantum tunneling could achieve integration densities 1,000 times greater than silicon chips by combining atomic-precision assembly with three-dimensional manufacturing techniques.
(Nanowerk Spotlight) Making transistors smaller has powered computing progress since the 1950s, but the strategy is nearing its end. The most advanced chips now in mass production, including Apple’s A17 Pro and M4 processors built on TSMC’s 3 nm process, pack transistors with physical gate lengths below 15 nm. At these dimensions, electrons tunnel through barriers meant to block them, causing current to leak even when transistors are switched off. This wastes power, generates heat, and erodes the efficiency gains that once accompanied each new generation of smaller transistors. The economics have grown equally daunting: fabrication plants for 3 nm chips cost more than $20 billion to construct.
These pressures have revived interest in a radically different approach. What if individual molecules served as electronic components? The idea originated in 1974 when theorists Arieh Aviram and Mark Ratner proposed that a single organic molecule with an electron-donating region at one end and an electron-accepting region at the other could function as a rectifier (Chemical Physics Letters, “Molecular rectifiers”).
Because electrons flow more easily from donor to acceptor than the reverse, such a molecule would conduct current preferentially in one direction. The proposal launched an entire field, but testing it demanded the ability to contact, position, and measure objects barely one nanometer across. Reliable experiments became possible only after decades of painstaking technical development.
A review article published in Microsystems & Nanoengineering (“Molecular electronic devices based on atomic manufacturing methods”) now synthesizes this progress. The authors examine fabrication methods, functional devices, and integration strategies, concluding that molecular electronics has matured from speculation to candidate technology. Potential integration densities reach 10¹⁴ devices per square centimeter, three orders of magnitude beyond current silicon chips.
Fabrication-Functionalization-Integration-Application process schematic of molecular electronic devices. (Image: Reproduced from DOI:10.1038/s41378-025-01037-8, cc BY)
The underlying physics differs fundamentally from conventional electronics. Charge moves through molecular junctions via quantum tunneling rather than drifting through continuous material. Conductance decays exponentially with molecular length: G = G0e−βl, where G reflects contact conductance and β characterizes tunneling efficiency.
Quantum interference adds another control mechanism. In benzene-based molecules, electrons traveling through different pathways can reinforce or cancel each other depending on connection geometry. Para-connected configurations, where anchoring groups attach at opposite ends of the ring, produce constructive interference and high conductance. Meta-connected configurations produce destructive interference, dropping conductance by several orders of magnitude. These quantum effects enable behaviors unachievable in bulk semiconductors.
Building functional junctions requires electrode gaps below 3 nm. Two fabrication strategies have emerged.
Static junctions use fixed electrodes with controlled spacing. Electromigration offers one approach: current pulses force metal atoms to migrate through a narrow wire until fracture creates atomic-scale gaps. Alternatively, self-assembled molecular layers deposited on substrates can be contacted using eutectic gallium-indium alloy, a liquid metal that forms gentle interfaces without damaging delicate molecules. Carbon electrodes, including carbon nanotubes and graphene, couple more effectively to organic molecules than metal electrodes.
Dynamic junctions repeatedly form and break contacts to build statistical datasets. The mechanically controllable break junction technique bends a flexible substrate to stretch a thin metal bridge until rupture. Molecules in solution span the resulting gap, forming junctions whose conductance researchers record before breakdown.
The scanning tunneling microscope break junction method uses a sharp tip repeatedly approaching and withdrawing from a surface. Microelectromechanical system break junctions, a newer variant, integrate the process onto chips for automated, high-throughput measurements. Thousands of cycles yield histograms revealing characteristic conductance values for specific molecules.
Anchoring chemistry determines how molecules attach to electrodes. Thiol groups dominate because sulfur bonds strongly to gold, though these bonds can restructure electrode surfaces unpredictably. Nitrogen-based anchors like amines and pyridines offer alternatives. Carbon-based connections using protected alkynes form robust metal-carbon bonds. Non-covalent approaches exploit π-π stacking, the attractive interaction between parallel aromatic rings that lets molecules self-organize without chemical bonding.
Several device classes now function reliably. Molecular switches change conductance under external stimuli. Diarylethene molecules toggle between ring configurations when exposed to alternating ultraviolet and visible light, shifting conductance by a factor of approximately 100.
Molecular diodes conduct preferentially in one direction. A 2017 demonstration using silicon electrodes achieved rectification ratios above 4 × 10³ under ambient conditions, matching conventional semiconductor performance.
Molecular transistors modulate current through a gate electrode that shifts molecular orbital energies. The first example, reported in 2000, used C₆₀, the soccer-ball-shaped carbon molecule known as buckminsterfullerene, achieving on-off ratios near 300.
Integration remains the steepest challenge. Two-dimensional arrays using self-assembled layers reach densities between 10⁹ and 10¹² devices per square centimeter. A 160-kilobit molecular memory demonstrated in 2007 validated the concept but suffered from degradation and switching ambiguities.
The review advocates adopting three-dimensional integration techniques from conventional semiconductor manufacturing. Through-silicon vias, vertical channels passing signals through wafer substrates, could connect stacked molecular device layers. Redistribution layers would handle horizontal routing using copper or ruthenium interconnects with tantalum nitride barriers to prevent metal diffusion.
Thermal incompatibility presents a core obstacle. Organic molecules degrade above 200 °C, while standard semiconductor processes involve annealing above 400 °C and chemical vapor deposition between 300 °C and 600 °C. The proposed solution: fabricate conventional interconnects first, then introduce molecules only during final manufacturing stages.
Precise placement demands new techniques. DNA origami folds long DNA strands into predetermined nanoscale shapes using short complementary sequences as guides. Researchers have positioned components with nanometer accuracy on such templates, offering a path toward directed assembly of molecular circuits at predefined sites.
Two applications show particular promise. Molecular memristors, devices whose resistance depends on current history, could enable brain-inspired computing. A 2022 study achieved room-temperature bistable switching in single metallofullerene molecules, demonstrating memory storage and basic Boolean logic operations at the single-molecule level (Nature Materials, “Room-temperature logic-in-memory operations in single-metallofullerene devices”).
Molecular sensors exploit junction sensitivity to chemical environment. One experiment tracked individual catalytic cycles of a single enzyme by monitoring conductance, with plateau currents near 98 pA revealing mechanistic details invisible to bulk techniques.
Substantial obstacles persist. Device reproducibility varies with electrode surface conditions. Organic molecules tolerate limited thermal and chemical stress. Manufacturing infrastructure optimized for silicon demands significant adaptation.
The review nonetheless documents concrete progress. Fabrication has advanced from crude mechanical methods to atomic-precision assembly. Measurement techniques extract single-molecule properties with statistical rigor. Switches, diodes, and transistors perform as designed. The physics supports integration densities impossible for silicon. As conventional scaling confronts quantum limits and escalating fabrication costs, molecular electronics offers a distinct path forward, one built not by carving features into bulk material but by assembling circuits molecule by molecule.
For authors and communications departmentsclick to open
Lay summary
Prefilled posts
Nanowerk Newsletter
Get our Nanotechnology Spotlight updates to your inbox!
Thank you!
You have successfully joined our subscriber list.
Become a Spotlight guest author! Join our large and growing group of guest contributors. Have you just published a scientific paper or have other exciting developments to share with the nanotechnology community? Here is how to publish on nanowerk.com.
