Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2207.08723 · Machine Learning Science and Technology

Realistic mask generation for matter-wave lithography via machine learning

Johannes Fiedler, Adrià Salvador Palau, Eivind Kristen Osestad, Pekka Parviainen, Bodil Holst

Published May 26, 2023• 0 citations
arXivHTMLPDF

Abstract

Abstract Fast production of large-area patterns is crucial for the established semiconductor industry and enables industrial-scale production of next-generation quantum devices. Metastable atom lithography with binary holography masks has been suggested as a higher resolution/low-cost alternative to the current state of the art: extreme ultraviolet lithography. However, it was recently shown that the interaction of the metastable atoms with the mask material (SiN) leads to a strong perturbation of the wavefront, not included in the existing mask generation theory, which is based on classical scalar waves. This means that the inverse problem (creating a mask based on the desired pattern) cannot be solved analytically, even in 1D. Here we present a machine-learning approach to mask generation targeted for metastable atoms. Our algorithm uses a combination of genetic optimisation and deep learning to obtain the mask. A novel deep neural architecture is trained to produce an initial approximation of the mask. This approximation is then used to generate the initial population of the genetic optimisation algorithm that can converge to arbitrary precision. We demonstrate the generation of arbitrary 1D patterns for system dimensions within the Fraunhofer approximation limit.

DOI
10.1088/2632-2153/acd988

Related papers

Nanometer-Resolution Mask Lithography with Matter Waves: Near-Field Binary Holography

Torstein Nesse, Ingve Simonsen, Bodil Holst

Mask-based pattern generation is a crucial step in microchip production, but there are many technical challenges in scaling current techniques down to nanometer resolution. Lithography using metastable atoms has been suggested as a cost-effective, less complex alternative to conventional techniques, including extreme ultraviolet (EUV) photolithography. This study presents a method based on binary holography that can be used to shape atom beams into arbitrary patterns. Simulations demonstrate the potential for state-of-the-art helium sources to produce patterns with nanometer resolution.

arXiv:1804.06910•Physical Review Applied•Feb 5, 2019•14 citations

An atom passing through a hole in a dielectric membrane: impact of dispersion forces on mask-based matter-wave lithography

Johannes Fiedler, Bodil Holst

Abstract Fast, large area patterning of arbitrary structures down to the nanometre scale is of great interest for a range of applications including the semiconductor industry, quantum electronics, nanophotonics and others. It was recently proposed that nanometre-resolution mask lithography can be realised by sending metastable helium atoms through a binary holography mask consisting of a pattern of holes. However, these first calculations were done using a simple scalar wave approach, which did not consider the dispersion force interaction between the atoms and the mask material. To access the true potential of the idea, it is necessary to access how this interaction affects the atoms. Here we present a theoretical study of the dispersion force interaction between an atom and a dielectric membrane with a hole. We look at metastable and ground state helium, using experimentally realistic wavelengths (0.05–1 nm) and membrane thicknesses (5–50 nm). We find that the effective hole radius is reduced by around 1–7 nm for metastable helium and 0.5–3.5 nm for ground-state helium. As expected, the reduction is largest for thick membranes and slow atoms.

arXiv:2201.03845•Journal of Physics B Atomic Molecular and Optical Physics•Jan 13, 2022•14 citations

Optimal Design of Grid-Based Binary Holograms for Matter-Wave Lithography

Torstein Nesse, Jean-Philippe Banon, Bodil Holst +1 more

Due to their very short de Broglie wavelengths, matter waves could be used for high-resolution patterning, competing with $e.g.$ e-beam or helium-ion lithography for producing features smaller than 10 nm. Binary holography is presently the only method available for mask-based atom lithography, but its theoretical framework has remained quite limited. The authors present an optimal method for making binary holograms that can shape atom beams arbitrarily, allowing for free selection of the geometrical properties of the generated masks, and the resulting structures.

arXiv:1703.09941•Physical Review Applied•Aug 17, 2017•8 citations

Multiparameter optimisation of a magneto-optical trap using deep learning

Aaron D. Tranter, Harry J. Slatyer, Michael Hush +7 more

Machine learning based on artificial neural networks has emerged as an efficient means to develop empirical models of complex systems. Cold atomic ensembles have become commonplace in laboratories around the world, however, many-body interactions give rise to complex dynamics that preclude precise analytic optimisation of the cooling and trapping process. Here, we implement a deep artificial neural network to optimise the magneto-optic cooling and trapping of neutral atomic ensembles. The solution identified by machine learning is radically different to the smoothly varying adiabatic solutions currently used. Despite this, the solutions outperform best known solutions producing higher optical densities.

arXiv:1805.00654•Nature Communications•Oct 15, 2018•105 citations

A Proposed Nanofabrication Technique Using Optical Masks for Metastable Atom Lithography

Mark A. Baker, A.J. Palmer, Robert T. Sang

We present an overview of our progress towards using optical fields as 'light masks' for metastable atomic beam lithography. Numerical simulations have been undertaken to model the interaction of a beam of metastable neon atoms with a standing wave of 640.2 nm laser light. The current experimental setup and the proposed scheme for focusing the atoms is detailed

Conference on Optoelectronic and Microelectronic Materials and Devices (COMMAD)•Jan 25, 2006•0 citations

Quantum Machine and Deep Learning Models for Industry Automation

Pawan Whig, Arun Velu, Rahul Reddy Nadikattu +1 more

Quantum machine learning unites two of the most contentious subfields in science: quantum computing and classical machine learning. Quantum machine learning is the study of how to solve machine learning problems using insights from quantum mechanics. As more and more information accumulates, the capacity of existing computing machines to successfully train it onto a conventional computation model is stretched to its limits. Quantum computing could be helpful for continuous training with large datasets. When developing learning algorithms, quantum machine learning aims to do so much more quickly than traditional methods. Finding patterns in data and using those patterns to predict the future is at the heart of traditional machine learning. The data can be mined for insights like these. Quantum systems, on either extreme, generate unusual patterns those conventional systems cannot, implying that quantum computers might be able to outpace traditional CPUs in ML tasks. This chapter discusses the past research on quantum machine learning and an update on its current state.

River Publishers eBooks•May 30, 2024•1 citation
Abstract Fast production of large-area patterns is crucial for the established semiconductor industry and enables industrial-scale production of next-generation quantum devices. Metastable atom lithography with binary holography masks has been suggested as a higher resolution/low-cost alternative to the current state of the art: extreme ultraviolet lithography. However, it was recently shown that the interaction of the metastable atoms with the mask material (SiN) leads to a strong perturbation of the wavefront, not included in the existing mask generation theory, which is based on classical scalar waves. This means that the inverse problem (creating a mask based on the desired pattern) cannot be solved analytically, even in 1D. Here we present a machine-learning approach to mask generation targeted for metastable atoms. Our algorithm uses a combination of genetic optimisation and deep learning to obtain the mask. A novel deep neural architecture is trained to produce an initial approximation of the mask. This approximation is then used to generate the initial population of the genetic optimisation algorithm that can converge to arbitrary precision. We demonstrate the generation of arbitrary 1D patterns for system dimensions within the Fraunhofer approximation limit.
Mask-based pattern generation is a crucial step in microchip production, but there are many technical challenges in scaling current techniques down to nanometer resolution. Lithography using metastable atoms has been suggested as a cost-effective, less complex alternative to conventional techniques, including extreme ultraviolet (EUV) photolithography. This study presents a method based on binary holography that can be used to shape atom beams into arbitrary patterns. Simulations demonstrate the potential for state-of-the-art helium sources to produce patterns with nanometer resolution.
Abstract Fast, large area patterning of arbitrary structures down to the nanometre scale is of great interest for a range of applications including the semiconductor industry, quantum electronics, nanophotonics and others. It was recently proposed that nanometre-resolution mask lithography can be realised by sending metastable helium atoms through a binary holography mask consisting of a pattern of holes. However, these first calculations were done using a simple scalar wave approach, which did not consider the dispersion force interaction between the atoms and the mask material. To access the true potential of the idea, it is necessary to access how this interaction affects the atoms. Here we present a theoretical study of the dispersion force interaction between an atom and a dielectric membrane with a hole. We look at metastable and ground state helium, using experimentally realistic wavelengths (0.05–1 nm) and membrane thicknesses (5–50 nm). We find that the effective hole radius is reduced by around 1–7 nm for metastable helium and 0.5–3.5 nm for ground-state helium. As expected, the reduction is largest for thick membranes and slow atoms.
Due to their very short de Broglie wavelengths, matter waves could be used for high-resolution patterning, competing with e.g.e.g.e.g. e-beam or helium-ion lithography for producing features smaller than 10 nm. Binary holography is presently the only method available for mask-based atom lithography, but its theoretical framework has remained quite limited. The authors present an optimal method for making binary holograms that can shape atom beams arbitrarily, allowing for free selection of the geometrical properties of the generated masks, and the resulting structures.
Machine learning based on artificial neural networks has emerged as an efficient means to develop empirical models of complex systems. Cold atomic ensembles have become commonplace in laboratories around the world, however, many-body interactions give rise to complex dynamics that preclude precise analytic optimisation of the cooling and trapping process. Here, we implement a deep artificial neural network to optimise the magneto-optic cooling and trapping of neutral atomic ensembles. The solution identified by machine learning is radically different to the smoothly varying adiabatic solutions currently used. Despite this, the solutions outperform best known solutions producing higher optical densities.
Quantum machine learning unites two of the most contentious subfields in science: quantum computing and classical machine learning. Quantum machine learning is the study of how to solve machine learning problems using insights from quantum mechanics. As more and more information accumulates, the capacity of existing computing machines to successfully train it onto a conventional computation model is stretched to its limits. Quantum computing could be helpful for continuous training with large datasets. When developing learning algorithms, quantum machine learning aims to do so much more quickly than traditional methods. Finding patterns in data and using those patterns to predict the future is at the heart of traditional machine learning. The data can be mined for insights like these. Quantum systems, on either extreme, generate unusual patterns those conventional systems cannot, implying that quantum computers might be able to outpace traditional CPUs in ML tasks. This chapter discusses the past research on quantum machine learning and an update on its current state.