Nanoelectronics

Nanoelectronics exploits structures and devices whose critical dimensions lie below 100 nm, harnessing quantum and interfacial phenomena to extend the capabilities of conventional microelectronics. At these scales, band structure engineering, Coulomb interactions and tunable charge confinement enable device concepts—such as two-dimensional transistors, single-electron pumps, resistive memories and neuromorphic elements—that transcend the performance limits of bulk silicon. Materials ranging from graphene and transition-metal dichalcogenides to novel covalent organic frameworks provide atomically thin channels or tunnelling barriers, while nanometre-scale gaps in mechanical switches eliminate off-state leakage. Key applications encompass ultra-low-power logic, high-frequency amplification into the terahertz band, non-volatile memories at femtojoule energies, in-memory computing for artificial intelligence and microsystems integration for biomedical and environmental sensing. The convergence of diverse nanoscale platforms and device architectures underpins a rapidly evolving landscape, with emerging hybrid devices promising to bridge photonic, electronic and mechanical domains on a unified chip.

Research from Nature Portfolio

Recent studies have demonstrated a “hot-emitter” transistor in which double mixed-dimensional graphene–germanium Schottky junctions produce stimulated emission of high-energy carriers. The device achieves a subthreshold swing below 1 mV / decade—far below the Boltzmann limit—while delivering a pronounced negative differential resistance with a peak-to-valley current ratio exceeding 100 at room temperature. Multi-valued logic and high inverter gain have been realised by tailoring bias conditions, highlighting a new route to reconfigurable low-power circuits. Complementing this, advances in memristive reservoir computing exploit the intrinsic ionic dynamics of nanoscale resistive switches. A small array of dynamic memristors forms a hardware reservoir that processes temporal data directly in the memory fabric, enabling handwritten-digit recognition and nonlinear time-series prediction without backpropagation. These works underscore the potential of hot-carrier and memristive device paradigms to deliver multifunctional, energy-efficient computation at the nanoscale.

Research from all publishers

Beyond Nature communications, a high-gain graphene-based hot-electron transistor has been fabricated by wet-transferring hexagonal boron nitride and graphene onto a germanium substrate. It delivers a record saturated output current density of 800 A cm⁻² with substantial current gain, demonstrating compatibility with large-scale manufacturing and prospects for high-frequency logic and amplification. In the realm of neuromorphic hardware, ordered imine-linked two-dimensional covalent organic frameworks have been integrated as memristive elements. The long-range ordered nanochannels guide the formation of conductive filaments, yielding an ON/OFF ratio above 10⁶, data retention beyond 10⁵ s and high-precision waveform recognition that rival software benchmarks. These developments illustrate the breadth of nanoelectronic devices—from quantum tunnelling transistors to architected molecular networks—pushing performance frontiers in logic, memory and sensing.

Nanoelectronics publication trend

The graph below shows the total number of articles in nanoelectronics across all publications each year (not limited to Nature Index journals).

Technical terms

Subthreshold swing: The gate-voltage change required to increase drain current by one decade in a field-effect transistor, with lower values indicating sharper switching.

Negative differential resistance (NDR): A regime in which increasing applied voltage results in a decreasing current, enabling multi-valued logic and oscillator circuits.

Memristor: A two-terminal element whose resistance state depends on the history of applied voltage or current, serving as a non-volatile synaptic analogue in neuromorphic systems.

Hot electron: A carrier possessing kinetic energy well above thermal equilibrium, exploited in hot-electron transistors for rapid switching and amplification.

Reservoir computing: A neural-network paradigm in which a fixed, dynamic reservoir projects temporal inputs into a high-dimensional space, training only the read-out layer for efficient temporal processing.

References

1. A hot-emitter transistor based on stimulated emission of heated carriers. Nature (2024).
2. Reservoir computing using dynamic memristors for temporal information processing. Nature Communications (2017).
3. High Gain Graphene Based Hot Electron Transistor with Record High Saturated Output Current Density. Advanced Electronic Materials (2023).