Quantum Computing Optimises Fabrication, Finding Better Solutions for Atomic Junctions

A new method for fabricating atomic junctions offers potentially transformative implications for nanoscale electronics. Takumi Kanezashi and colleagues at Tokyo University of Agriculture, in a collaboration between Tokyo University of Agriculture and The University of Tokyo, use a gate-based quantum computer to optimise the complex process of feedback-controlled electromigration. The research reveals that noisy intermediate-scale quantum (NISQ) devices outperform quantum annealing systems in finding effective solutions for optimising experimental parameters, achieving lower residual energies and higher-quality approximations for larger problems. This advancement could sharply improve the precision and efficiency of creating atomic-scale structures.

Reduced residual energy enables precise gold atomic junction fabrication

Residual energies achieved with the new system are now below 0.05, a reduction of over 30% compared to previous D-Wave quantum annealers. This significant decrease in residual energy is crucial because it directly correlates with the precision of the fabricated atomic junctions. Electromigration, the process of moving atoms via electrical current, is inherently susceptible to variations and noise. Minimising residual energy ensures that the final atomic configuration closely matches the desired target, leading to more reliable and reproducible junctions. Earlier systems struggled to consistently create junctions with the required precision, hindering the development of functional nanoscale devices. D-Wave’s team successfully used a noisy intermediate-scale quantum (NISQ) device, containing over 100 qubits, to autonomously optimise feedback-controlled electromigration, a technique for building these nanoscale structures. The qubits, the fundamental units of quantum information, were leveraged to explore the vast parameter space of the electromigration process, identifying optimal settings that would otherwise be inaccessible through conventional methods.

Variational quantum algorithms, employed by gate-based quantum computers, delivered more efficient and reliable approximate solutions than quantum annealing for this complex optimisation problem. Unlike classical optimisation algorithms, these quantum algorithms exploit quantum phenomena such as superposition and entanglement to explore multiple potential solutions simultaneously, accelerating the search for the optimal configuration. This advancement promises greater control over atomic-scale materials and opens new possibilities for nanoscale electronic devices. The computer autonomously scheduled feedback-controlled electromigration (FCE) parameters across multiple experimental cycles, allowing for incremental improvements in junction quality and reducing the need for manual intervention. This automated approach is particularly valuable given the sensitivity of FCE to subtle changes in experimental conditions, which are difficult to control manually.

Such automated optimisation represents a significant step towards scalable nanofabrication. The system successfully fabricated single-electron transistors using FCE, a technique reliant on precise atomic control. Single-electron transistors are promising candidates for future electronic devices due to their ultra-low power consumption and potential for high integration density. The quantum computer optimised the critical feedback voltage parameter, incrementally adjusting voltage applied to gold nanowires to induce atomic migration and reduce conductance. The feedback loop, guided by the quantum computer, allowed for the precise shaping of the gold nanowire, ultimately leading to the formation of a stable atomic junction. Analysis revealed the system could evaluate the controllability of normalised conductance, a key metric for junction quality, with greater efficiency than earlier machine learning approaches. Normalised conductance provides a measure of the current flowing through the junction relative to its size, indicating the quality of the atomic connection. While the residual energies achieved are below 0.05, these results are currently limited to optimising single parameters and do not yet demonstrate the ability to simultaneously control all variables required for complex nanoscale device fabrication. Future work will focus on extending the quantum optimisation framework to handle multiple parameters concurrently, paving the way for the fabrication of more intricate nanoscale devices.

NISQ devices surpass quantum annealing for automated atomic junction fabrication

Fabricating atomic junctions, structures with just a few atoms acting as electrical connections, promises breakthroughs in nanoscale electronics and sensing. These junctions exhibit unique quantum mechanical properties that can be exploited for a variety of applications, including highly sensitive sensors, quantum computing components, and novel electronic devices. Traditionally, achieving precise control over these junctions relies on painstakingly adjusting experimental parameters, a process prone to error and severely limiting scalability. The manual tuning of parameters such as voltage, current, and temperature is time-consuming and often yields inconsistent results. Automated optimisation of this process using noisy intermediate-scale quantum (NISQ) devices is now feasible, but its broader applicability remains an important consideration. The challenge lies in translating the theoretical advantages of quantum computing into practical improvements in nanofabrication.

Acknowledging challenges associated with current noisy intermediate-scale quantum (NISQ) devices is sensible, given limitations in qubit numbers, coherence times, and gate availability. Qubit coherence, the duration for which a qubit maintains its quantum state, is particularly critical for performing complex calculations. However, a NISQ device, however, achieved more effective approximate solutions than a previously reported D-Wave quantum annealer in fabricating gold atomic junctions, a noteworthy result. Quantum annealing, while effective for certain types of optimisation problems, is less versatile than gate-based quantum computing and struggles with problems requiring complex parameter tuning. This investigation establishes a method for autonomously fabricating these nanoscale components, potentially aiding progress in areas such as advanced sensors and electronics despite present-day hardware constraints. The gate-based quantum computer successfully optimised the fabrication of gold atomic junctions, eliminating the need for manual adjustment of experimental parameters, a historically difficult aspect of building such small components. Variational quantum algorithms delivered higher-quality approximate solutions than the quantum annealing system for this complex optimisation problem, offering a route to more precise device designs. The ability to automate this process is crucial for scaling up the production of atomic junctions and realising their full potential in nanoscale technologies.

The research demonstrated that a noisy intermediate-scale quantum (NISQ) device successfully optimised the fabrication of gold atomic junctions, removing the need for manual parameter adjustments. This is important because precisely controlling the creation of these nanoscale components is traditionally a time-consuming and inconsistent process. The NISQ device achieved more effective approximate solutions than a D-Wave quantum annealer for this complex optimisation problem, suggesting a pathway towards more reliable nanofabrication. The authors indicate this method could aid progress in areas such as advanced sensors and electronics.