Atomic-Scale Two-Level Conductance Switches

Updated 20 November 2025

Atomic-scale two-level conductance switches are nanoscale systems that toggle between distinct ‘on’ and ‘off’ conductance states by manipulating a single atomic degree of freedom.
They operate via quantum transport principles where fully open conduction channels provide quantized conductance and tunneling gaps produce exponentially suppressed currents.
These switches are promising for ultra-dense non-volatile memory and quantum circuits, with diverse realizations in metallic contacts, molecular junctions, and 2D heterostructures.

Atomic-scale two-level conductance switches are nanoscale systems whose conductance can be toggled reproducibly between two sharply defined states by manipulation of a single atomic-scale degree of freedom. In these systems, the "on" state typically corresponds to a contact supporting electron transmission via one or more fully open conduction channels (often quantized in units of the conductance quantum, $G_0=2e^2/h$ ), while the "off" state is set by a tunneling gap or local potential barrier such that transmission is exponentially suppressed. This behavior arises in diverse classes of devices, spanning monatomic wires, metallic nanocontacts, molecular and ionic junctions, 2D van der Waals heterostructures, and memristive stacks, all united by the capacity for precise atomic-scale reconfiguration between two distinct structural or electronic configurations that modulate transport by orders of magnitude.

1. Quantum Transport Principles and Conductance Quantization

The central mechanism underlying atomic-scale two-level switching is the quantization of conductance in narrow constrictions, where transport is phase-coherent and constrained by the discrete set of available conduction channels defined by the electronic structure and geometry of the junction. In the "on" state, the transmission probability for each open channel approaches unity, yielding a conductance $G = N G_0$ , where $N$ is the integer number of channels. In metallic atomic contacts or quantum point contacts, the conductance steps observed as the atomic configuration evolves directly reflect this quantization (Berg et al., 2016, Sattar et al., 2015, Wagenaar et al., 2016, Deswal et al., 2019). In the "off" state, the conductance is governed by tunneling, with an exponential suppression $G(d) = G_t \exp(-2\kappa d)$ , where $d$ is the gap width and $\kappa = \sqrt{2m\Phi}/\hbar$ encodes the barrier height $\Phi$ and electronic mass.

Molecular-scale junctions and atomically engineered heterostructures extend this paradigm by realizing two discrete conductance states through manipulation of a single atom, ion, proton, or molecular conformation, which acts as a localized barrier or enables destructive quantum interference (Zhong et al., 2016, Zotti et al., 2013, Hofmeister et al., 2014, Tripathi et al., 2021).

2. Physical Realizations and Mechanisms

Atomic-scale two-level switches have been demonstrated in a variety of materials systems and device architectures, each exploiting distinct atomic-control mechanisms:

Magnetostrictive nanocontacts: In Fe $_{73}$ Ga $_{27}$ (Galfenol), the application of low magnetic fields ( $<30$ mT at 10 K) induces elongation of electrodes via positive magnetostriction, reversibly closing or opening an atomic-scale gap. This modulates the junction conductance sharply between $G \sim 0$ (tunneling) and $G \gtrsim 1$ – $15\,G_0$ (metallic contact), with switching hysteresis as low as 1 mT after training cycles (Berg et al., 2016).
Percolating nanoparticle films: Near the percolation threshold, tunnel gaps in Sn nanoparticle networks allow formation and rupture of atomic-scale metallic filaments via van der Waals forces, electric field-induced surface diffusion, and electromigration. The system cycles between quantized conductance levels—often $n\,G_0$ —and the low-conductance (tunneling) state (Sattar et al., 2015).
Ionic solid-state switches (e.g., Ag $_2$ S, Nb $_2$ O $_5$ ): Conductance states reflect the formation and dissolution of atomic-scale metallic filaments (nanowires) by ionic migration and electrochemical reduction/oxidation. Metallic filaments yield high conductance; their breakup causes discrete steps of $\sim G_0$ . In parallel, resistive switching by lattice modification offers slower, analog retention (Wagenaar et al., 2016, Morales-Masis et al., 2011, Deswal et al., 2019).
Memristive oxide devices (HfO $_2$ ): Atomic motion of single (or few) oxygen atoms within a conducting filament modulates device conductance exponentially as the filament is partitioned into quantum wells by the "blocking" O atom. This enables robust switching between $G\sim G_0$ and $G\ll G_0$ by nanometer-scale O displacement (Zhong et al., 2016).
Graphene nanogaps: Electric-field-driven bond rearrangements at the edge of an encapsulated graphene nanogap cause conductance to switch between two tunneling states, corresponding to gap widths differing by $\sim0.3$ nm. Switching thresholds are highly reproducible, and hysteresis is set by field-induced lowering of the reaction barrier (Pyurbeeva et al., 2021).
Single-molecule switches: Several distinct mechanisms exist, including:
- Conformational or tautomeric transitions—for example, field- or photo-induced proton transfer between enol and keto forms, which alters conjugation and hence electron transmission (Hofmeister et al., 2014, Wierzbowska et al., 2014).
- Gate-driven charge addition, as in phthalocyanine molecules gated by individual adatoms, which couples charge states to orientational conformations, yielding two conductance levels with a large gap (conformation-dependent Franck–Condon blockades) (Martínez-Blanco et al., 2016).
- Molecular orbital rehybridization, as in the (SC $_4$ S) $_2$ Pt $_6$ cluster whose unique double-resonance structure yields sequential on–off–on transitions in response to gate or bias (Zotti et al., 2013).

A table summarizing key mechanisms is given below:

Material/Device	Switching Mechanism	Quantized States
Fe $_{73}$ Ga $_{27}$ nanocontact (Berg et al., 2016)	Magnetostriction-driven contact closure	$G \approx 0$ , $G > G_0$
Sn nanoparticle films (Sattar et al., 2015)	Atomic-wire formation/rupture	$G = n\,G_0$ , $G \ll G_0$
Ag $_2$ S (Wagenaar et al., 2016)	Ag filament growth/dissolution	$G \sim 100\,G_0$ , $G_0$ , tunneling
HfO $_2$ filament (Zhong et al., 2016)	Oxygen-ion modulation	$G\sim G_0$ , $G\ll G_0$
Encapsulated graphene (Pyurbeeva et al., 2021)	Field-driven C–C bond rearrangement	Two tunneling states
Single molecules (Hofmeister et al., 2014, Martínez-Blanco et al., 2016, Zotti et al., 2013)	Proton transfer, gating, interference	$G_{\mathrm{enol}}$ , $G_{\mathrm{keto}}$ ; conformer-specific

3. Theoretical and Computational Framework

The unifying theoretical description applies the Landauer–Büttiker formalism, with the two-terminal conductance at low bias given by

$G = G_0\,T(E_F),$

where $T(E_F)$ is the quantum mechanical transmission probability at the Fermi energy. In many realizations, $T$ switches sharply as a function of a single nuclear coordinate, such as atomic displacement, rotation, or identity. For tunneling regimes, $T$ decays exponentially with distance or barrier height, while in metallic or resonant-contact regimes $T$ approaches unity in fully open channels.

First-principles atomistic modeling—typically based on density functional theory (DFT) with non-equilibrium Green's function (NEGF) transport—permits the calculation of the complete transmission spectrum, barrier profiles, and energy surfaces governing the bistability and kinetics of switching (Zhong et al., 2016, Ring et al., 2023, Zotti et al., 2013, Wierzbowska et al., 2014, Martínez-Blanco et al., 2016). In vibrationally driven switches, a generalized Langevin approach models the bias-driven transition between bistable atomic geometries.

For devices involving ionic migration, electrochemical models incorporating field-enhanced drift and dissolution of species (Ag $^+$ , O $^{2-}$ ) describe switching thresholds, retention, and endurance (Wagenaar et al., 2016, Morales-Masis et al., 2011, Zhong et al., 2016).

4. Experimental Techniques and Device Architectures

Atomic-scale two-level switches are fabricated by a variety of techniques:

Mechanically controlled break junctions (MCBJ) provide atomic-scale control over electrode separation, enabling systematic paper of conductance quantization and training effects in metallic and magnetostrictive systems (Berg et al., 2016).
Nanoparticle drop-casting or evaporation allows percolation-controlled formation of tunnel networks, which are subsequently driven at high field or current densities to instigate filament formation (Sattar et al., 2015).
Solid-electrolyte or oxide stacks are fabricated by sputtering or thermal evaporation, with resistive switching probed via voltage sweeps and current compliance (Deswal et al., 2019, Wagenaar et al., 2016, Zhong et al., 2016).
STM junctions enable electric field-induced manipulation of atomic or molecular positions, single-proton transfers, or atomistic gating by manipulated adatoms (Martínez-Blanco et al., 2016, Morales-Masis et al., 2011, Hofmeister et al., 2014).
2D heterostructure assembly exploits exfoliation and transfer of atomically thin layers, with lithographically defined multi-gate architectures for controlled charge transfer and tunneling switching (Tripathi et al., 2021).
Electron beam lithography and electroburning are used for precise graphene nanogap fabrication and encapsulation for environmental stability (Pyurbeeva et al., 2021).

In all platforms, repeated cycling ("training") strongly influences the stability and reproducibility of the bistable conductance states, reducing the effective switching threshold and hysteresis.

5. Switching Kinetics, Hysteresis, and Retention

Switching between conductance levels occurs via:

Field- or bias-induced atomic/ionic migration (e.g., oxygen or silver ions).
Barrier modulation via external fields (electric/magnetic, local gating).
Conformational or chemical-state transitions (proton transfer, rotation, tautomerization).

Kinetics are determined by the activation barrier for the critical atomistic event, which can be modulated by field strength, device history, or gating conditions (Berg et al., 2016, Hofmeister et al., 2014, Ring et al., 2023). Measured switching times span picoseconds (for shell-quantized nanowires and proton transfer (Bürki et al., 2010, Hofmeister et al., 2014)) to milliseconds or seconds (ionic or filament-based systems (Sattar et al., 2015, Wagenaar et al., 2016)). Hysteresis in the transition threshold is typically minimized by cycling, but can remain substantial due to metastability or slow relaxation (conformational memory, lattice modification).

Retention of the switched state is limited by the barrier for spontaneous reversal: metallic filaments in solid electrolytes are often volatile, dissolving within seconds unless a sustaining bias is applied (Wagenaar et al., 2016, Morales-Masis et al., 2011). In systems where switching is tied to an electronic or conformational state with a deep energy minimum, stability exceeding minutes or hours is attainable at low temperature (Zhong et al., 2016, Deswal et al., 2019).

6. Device Performance, Applications, and Limitations

Atomic-scale two-level conductance switches exhibit the following salient performance metrics:

On/off conductance ratios of $10$– $10^{3}$ , depending on the underlying mechanisms and the electronic coupling between atomic configurations.
Quantized conductance plateaus or discrete steps, indicative of single-atom transport events (Sattar et al., 2015, Wagenaar et al., 2016, Deswal et al., 2019).
Remote or field-driven control: Magnetic (magnetostriction), electric (local gating, bias voltage), or chemical (adsorbate manipulations (Tripathi et al., 2021)) enable actuation without direct mechanical contact.
Reproducibility in cycling: After initial "training," many devices exhibit stable bistability over hundreds of cycles (Berg et al., 2016, Deswal et al., 2019).
Integration and scaling: Architectures such as oxide memristors, graphene nanogaps, and 2D heterostructures are compatible with large-area fabrication and integration into logic, sensing, or neuromorphic arrays.

Applications of atomic-scale two-level switches include ultra-dense non-volatile memory (via bistable filaments or molecular switches), quantum threshold devices, single-electron transistors, reconfigurable quantum circuits, and atomic-precision field sensors (Berg et al., 2016, Deswal et al., 2019, Pyurbeeva et al., 2021). Limitations arise from retention volatility (in metallic filaments), device-to-device variability (in percolating or analog networks), and stochastic telegraph noise when barriers are comparable to $k_B T$ (Bürki et al., 2010).

7. Outlook and Comparisons Across Material Systems

The diverse physical realizations outlined above share fundamental transport physics but differ in terms of retention, switching speed, power, fabrication complexity, and integration compatibility. Magnetostrictive and electrochemical devices excel in remote actuation and low actuation thresholds, whereas memristive oxides offer robust non-volatile bistability via atom motion. Molecular and 2D material switches advance ultimate miniaturization and electrical tunability.

Continued development of these systems seeks to optimize barrier heights, minimize stochasticity, and enhance reproducibility, with comprehensive first-principles modeling and precision characterization providing the foundations for next-generation atomic-scale devices (Zhong et al., 2016, Ring et al., 2023, Martínez-Blanco et al., 2016, Tripathi et al., 2021, Wierzbowska et al., 2014).