Rational design of novel catalysts, machine learning techniques and advanced scientific databasing services.

Materials Discovery

Our expertise lies in advanced materials discovery. We are at the forefront of theoretical materials discovery, with strengths in lubrication and materials for the automotive and electronics industries.

Machine Learning

We have in-depth intuition and expertise using machine learning algorithms. We develop machine learning techniques for the academic and commercial world for a wide range of topics. Check out our ML driven submarine!

Diverse Team

Coming from academic backgrounds, each member of our team brings their own unique insight to the group. Our collective scientific background encompasses physics, chemistry, tribology, nanoelectronics and more. Go to our team member's Researchgate and Google Scholar profiles to see their most recent publications.

Consult & Train

Whether trying to decide which materials to use or research an entirely new material, we are here for you. We also provide training in the use of advanced materials modelling and machine learning techniques.

News

Click to find out more

Nanolayers is happy to announce their participation in the Horizon funded MAST3RBoost project. A consortium of thirteen partners across eight countries that aims to develop novel ultraporous frameworks for cold hydrogen storage.

Filippo Canova will be giving a talk at the Workshop on Computational Biophysics of Atomic Force Microscopy organized with support from the International Union for Pure and Applied Biophysics (IUPAB). Join the workshop to see an overview of simulating non-contact atomic force microscopes.

The conductivity across nanotube junctions makes up the bulk of resistivity in nanotube thin films. In an international collaboration, Nanolayers systematically investigated the conductivities across nanotube junctions. The results have been published in Physica Status Solidi Rapid Research Letters.

Nanolayers was delighted to present recent progress on the modelling of switching mechanisms in devices that utilise carbon nanotube films. A talk was presented at the Memristor conference at King's College, London.

In conjunction with University College London, Nanolayers is looking for a student to join a multi-disciplinary, cross-Atlantic research project! Nanolayers’ machine learning capabilities will be combined with scanning probe microscopy at UCL to aid in the fabrication of silicon based quantum bits.

Nanolayers is delighted to take part in a UK-Canada collaboration to develop technologies for quantum devices funded by Innovate UK. By combining Nanolayers' priority machine learning techniques with the project partners' experimental capabilities, advanced techniques will be developed to fabricate large number of quantum bits.

Click here to toggle older news

Projects

Quantum Devices

In collaboration with research teams at University College, London, and McGill University at Montreal, Canada, Nanolayers is developing unique new technologies to exploit the multi-billion dollar silicon industry in the development of quantum devices.

Electronic Device Materials

Nanolayers applies a unique combination of theoretical physics and chemistry techniques to predict the properties of novel materials and apply them in a variety of multi-scale models to predict novel electronic device properties.

Scientific Databasing

Nanolayers can help you design and maintain data management systems particularly suited to scientific data. We run projects around the globe creating and maintaining systems to store and analyze scientific data.

Rational Catalyst Design

As part of the CritCat project, Nanolayers aims to provide solutions for the substitution of critical metals, especially rare platinum. Using advanced machine learning techniques, we provide an innovative approach that allows us to rationally design novel catalytic materials.

Demos

Try some of our software

Nanolayers designs various software products for its clients. These apps are tailored to the client's particular needs. However, here we showcase a number of apps that we have made a bit more general for you to try out!

Neural Network Submarine

At Nanolayers, we use machine learning techniques to advance our understanding of materials. Here, we use it in an entirely different way. In this game, you can use your own intelligence to compete against a submarine's artifical intelligence. The submarine's motion is driven by a neural network that has been trained to avoid collisions, despite only being able to see a short distance away. Try it out for yourself and see if you can beat our intelligent submarine.

Team

David Gao

Founder and Director

David Gao is the founder and director of Nanolayers Research Computing, and a recognised authority in the fields of materials design and machine learning development.

In 2014, David earned his doctorate in Condensed Matter and Materials Physics from UCL and has since continued to lecture there as a visiting research fellow in physics and astronomy. He also holds an associate professorship in the Department of Physics at the Norwegian University of Science and Technology.

In addition to his work with Nanolayers, David has played an active leadership role in the European scientific community, most recently contributing to the development of standardized materials modelling terminology for the European Committee for Standardization.

David is based at Nanolayers’ headquarters in London.


Filippo Federici Canova

Co-founder and Scientific Director

Filippo is the co-founder and scientific director for Nanolayers Research Computing, where he leads the business' development for image recognition and machine learning methods.

Filippo holds a doctorate in Computational Physics from the Tampere University of Technology in Finland. With extensive knowledge of advanced experimental and computational systems, Filippo has been instrumental in the development of some of Nanolayers most significant solutions, including the Nature Conservancy image recognition system and Cassandra Scientific databasing system.

Filippo is based in Finland.


Tassem El-Sayed

Co-founder and Scientist

A recipient of a doctorate in Condensed Matter Physics from UCL, Tassem is also a postdoctoral researcher for the Institute for Microelectronics at TU Wein in Vienna. His main research interests are in atomistic modelling techniques and understanding chemical and physical properties of materials. Tassem leads development of Nanolayers materials modelling computation solutions, as well as managing database and storage solutions for scientific data.

Tassem is based in Vienna.


Find Out What We Can Do For You

Do you think we can help you to realise a project? Are you interested in using advanced scientific methods to optimize your products? Why not have a chat with us? Fill in the form below and somebody from our team will be in touch with you as soon as possible.

Get in Touch

Established by the Solar Impulse Foundation, this alliance was established to bring together involved in developing services, process and technologies that protect the environment in a profitable way. As part of the CritCat group, Nanolayers has been integral to its efforts to present a sustainable solution for the replacement of rare Platinum-group metals, enabling further advances for existing and emerging clean and renewable energy.

The event presents an excellent opportunity for Nanolayers to share how leading academic research and processes are presenting powerful new solutions for the extensive ranges of businesses utilising Platinum Group Metals (PGMs), including automotive emissions control systems, and chemical synthesis and energy conversion.

The article, which can be read by this link, is a clear illustration of the advanced machine learning techniques Nanolayers applies in developing solutions for any number of scientific or industrial challenges. The principles outlined in this paper cover some of the machine learning methodology you can see in our neural sub app and our work with California’s Nature Conservancy.

The conference which was held in Evanston, Illinois, brought together experts from academia and industry to discuss various topics on advanced simulations of nanostructures, nanomaterials, and nanodevices. Nanolayers were invited to discuss the state of atomistic simulations of materials for electronic devices and approaches that can be taken to move the field forward.

Nanolayers presented their latest data management and storage solutions designed for scanning probe microscopy at the NC-AFM conference. In a special session, Nanolayers' David Gao took attendees through our LabCore solution for storing experimental data and performing tailored analyses on the same platform. Attendees could try out our LabCore solution which will help automate microscopy image production and pushing scanning probe microscopy into a new era.

The machine learning conference will bring together experts from around the world to discuss how to develop machine learning approaches in the context of microscopy. Nanolayers will be presenting their state-of-the-art solutions to data management and storage for microscopy and will give interactive sessions using their machine learning tools to automate image processing. Come join us to learn more about our solutions!

The package allows one to easily set up atomistic descriptors for the systems they are learning from. Setting up a descriptor is a vital aspect of making sure that the machine learning model used can properly learn from the samples you give it. Various descriptors exist which can overcome issues such as translational and rotationally invariant descriptions. Check out our recent paper to see examples of its use.

A material's complex quantum effects, such as its magnetic or ferroelectric properties, can be at odd with the redox processes that the material undergoes that change its conductance. In a recent Nature Materials article, an atomically controlled mechanism was used to create point defects in the functional oxide hexagonal Er(Mn,Ti)O3. Using conductive atomic force microscopy, the electric field was used to induce the creation of neutral anti-Frenkel defect pairs. This allows one to locally control and enhance its conductive properties while preserving the material's ferroelectric order. Read more here!

Materials informatics and machine learning have been making huge advances in computational materials science. Nanolayers has been involved in developing state of the art machine learning tools for use in materials science. Filippo Canova was invited to give a lecture at the ML4MS conference, discussing advances in materials descriptors, neural networks, kernel regression, among more exciting new areas of research.

Circuitry exploiting the behaviour of quantum particles is currently a hot topic, as it promises ground breaking advances in many fields, including cryptography, finance, materials design, and more. To aid the development of these technologies, Nanolayers has partnered with a number of research and private institutions in the UK and Canada to develop devices with large numbers of quantum bits (qubits). By applying machine learning techniques developed by Nanolayers to the atomically precise microscopy accomplished by our partners, large numbers of qubits can be accurately and rapidly fabricated onto devices. Moreover, the material of choice in this project is silicon, allowing for seamless integration into the more than 50 years of silicon technology that has developed over the decades. This project is funded by Innovate UK. Check out their press release here.

Nanolayer's David Gao will be giving a colloquium at the Norwegian University of Science and Technology on the 14th of May at 14.15 CET. The colloquim will discuss developments in modern computing and how resistive switching can be exploited for memory and neuromorphic devices. Existing phenomenological models will be discussed, but the main focus will be on how atomistic calculations can be used to understand the mechanisms behind resistive switching in order to optimize them. Check out the colloquium here.

Fabricating large numbers of quantum bits made from dopant atoms in silicon requires the fabrication and structural atomic-scale imaging capabilities of UCL, the characterisation of the electronic structure of the dopant atoms at McGill, and the machine learning image recognition and process automation capabilities of Nanolayers. In particular, UCL and Nanolayers will utilise machine learning for rapid, automatic qubit characterisation using scanning tunnelling microscopy. While quantum technologies have captured the imagination of scientists and technologists alike, the ideal candidate material for scalable future quantum computers, namely silicon, remains largely untapped. In this project the student will fabricate many silicon qubits from foreign 'dopant' atoms in silicon. They will adapt a unique dopant placement capability developed at UCL to make large random distributions of single dopant atoms or a few precisely placed dopants, or a combination of both, in a 2D plane in silicon. In order to use these dopant distributions as a source of many qubits, the student will work with Nanolayers to assess scanning tunnelling microscopy images of the surface using machine learning image recognition to locate the dopants. Subsequently they will collaborate with McGill University who will use their (also unique) atomic force microscopy technique that performs single electron spectroscopy on individual dopant atoms. The final stage of fabricating qubit device structures from a known dopant distribution is currently performed completely manually and cannot be scaled up to more than a few qubit bits at most. However, using newly developed methods by Nanolayers to train transferrable automation AI, fabrication will be scaled-up to many qubits, pushing silicon to the forefront of the quantum information revolution. Find out more here!

The Memristor conference was held at King's College London, UK and was organised by the Institute of Physics. The presentations showed the latest developments in the understanding of resistive switching in a wide range of devices, with a focus on academic studies. Nanolayers presented their latest findings on resistive switching devices that use carbon nanotube films. We provided a more commercial focus of resistive switching devices, showing a resistive switching model that can be incorporated into the design of devices. Find out more in the abstract here!

Understanding the behaviour of CNT junctions can lead to better design of a large range of electronics that use CNT films, such as transparent, flexible and wearable electronics, next-generation transistors, gas sensors, and more. However, previous studies only focussed on certain CNT junctions, with no general understanding of their conductivity. Nanolayers collaborated with an international consortium and systematically studied the conductivity of CNT junctions. We found that a simplified tunnelling equation can be used to parameterize the conductivity as a function of the smallest atomic distances between the nanotubes. A better description was obtained by including the effect of all interatomic distances between the two tubes forming the junction. To find out more, head over to the paper at Physica Status Solidi B. This work was presented at the International Carbon Conference 2022 held in London.

We are delighted to present an overview of simulating non-contact atomic force (ncAFM) microscopy images at the Workshop on Computational Biophysics of Atomic Force Microscopy. ncAFMs are capable of imaging samples at extremely high resolution, enabling us to see the very atomic fabric of surfaces. Unfortunately, these microscopes have quite little in common with optical microscopes we are all familiar with. The probe is usually an atomically sharp tip, rather than our eyes, and the sampling mechanism relies on interatomic forces between the tip and the sample, rather than light diffusing from it; the instrument itself is a very complex piece of machinery, with several electronic feedback controls acting on the probe dynamics.

In his talk, Fillipo will give a broad overview of the theoretical methods involved in the interpretation of ncAFM images, including classical and ab-initio techniques to calculate tip-sample interactions, and circuit-level instrument simulators, and show how they were applied to solve research problems.

MAST3RBoost, Maturing the Production Standards of Ultraporous Structures for High Density Hydrogen Storage Bank Operating on Swinging Temperatures and Low Compression, is a European project which aims to provide a solid benchmark of cold-adsorbed H2 storage (CAH2) at low compression (100 bar or below) by maturation of a new generation of ultraporous materials (Activated carbons, ACs, and Metal Organic Frameworks, MOFs) for mobility applications, i.e., H2-powered vehicles, including road and railway, air-borne and water-borne transportation. The goal is to achieve a 30% increase of the working capacity of H2 at 100 bar (vs. MOF-5, one of the current record holders) reaching 10 wt.% and 44 g H2 /l, by turning the lab-scale synthesis protocols into industrial-like manufacturing processes. Reaching these figures would bring significant advances on Hydrogen storage banks and, therefore, to Europe’s decarbonization.

Carbon dioxide emissions are a problem across the world and a big part of them are produced among the transport sector. In Europe they constitute already one third of all CO2 emissions with over 1,000 million tons, representing a big threat for human health as well as one of the largest contributions to the climate change. Decarbonization of the economy and, in this case, of the transport sector is urgent. There have been improvements with the Fuel Cells and Hydrogen (FCH) batteries, which have proven to be a promising solution for the decarbonization of trucks, buses, ships, trains or large cars. With the larger vehicles being potential early adopters, this new industry has the potential to generate a 130 billion € market in the European Union alone.

The problem is that, at the moment, the state-of-the-art technology for Hydrogen storage on board based on compression at 700 bar, has reached 25 g H22/l, a number which is still low considering that the market-entry goal is to fit 5 kg of H2 in a gasoline equivalent tank (80 kg/90 l). In fact, the complexities associated to an efficient H2 storage are causing a very slow penetration of Fuel Cell Electric Vehicles (FCEVs). MAST3RBoost’s goal is to reach at least 40 g H2 /l , which is a significant milestone that would help to provide the market with an actual FCEV replacement to the current internal combustion engines, which are big contributors to the EU’s greenhouse gas emissions.

Based on a new generation of Machine Learning-improved ultraporous materials – such as Activated Carbons (ACs) and high-density MOFs (Metal-organic Frameworks) –, MAST3RBoost project will enable a disruptive path to meet the industry goals by developing the first worldwide adsorption-based demonstrator at the kg-scale. Lightweight vessels –embedding the ultraporous materials– will be created taking advantage of the innovative Wire-Arc Additive Manufacturing, with dedicated shapes to better fit on-board specific transportation spaces.

Recycled raw materials for the manufacturing of the ultraporous materials will be actively pursued, both from waste agroforestry biomass and from solid urban waste. The research and development process will be performed applying Life Cycle thinking strategies to minimise overall environmental impacts

This project is funded in the topic HORIZON-CL4-2021-RESILIENCE-01-17 by the European Health and Digital Executive Agency. It is a Research and Innovation Action project with a budget of 4,638,414.00 €, 100% funded by the EU.

Coordinated by Envirohemp, the project will last for four years and counts on thirteen partners from eight different countries: Envirohemp S.L. (Spain); Contactica S.L. (Spain); Agencia Estatal Consejo Superior de Investigaciones Científicas (Spain); CIDETEC Surface Engineering Institute (Spain); Spike Renewables SRL (Italy); EDAG Engineering GMBH (Germany); Nanolayers OU (Estonia); LKR Leichtmetall Kompetenzzentrum Ranshofen GMBH (Austria); University of Pretoria (South Africa); Council For Scientific And Industrial Research (South Africa); Stellantis (old PSA Groupe) (Portugal); TWI (UK); University of Nottingham (UK).

Follow MAST3RBoost's progress on Twitter and LinkedIn!