Dr Ned Thaddeus Taylor (he/him/his)

I am a postdoctoral research fellow in the Physics Department at the University of Exeter, UK, currently funded by the Royal Academy of Engineering via their UK Intelligence Community fellowship scheme. I am a theoretical materials scientist, with focuses on materials discovery, material operational limits, and the application of machine learning to these fields. Many of our modern devices rely on the unique physical properties of interfaces (the joins between materials). At the same time, these interfaces are often the limiting factor of device operation - often being the first point of failure when it comes to varying electrical, thermal, chemical, and mechanical conditions.

Research: I study the limits of material interfaces and their impact on electronic devices. I have applied this research to green energy solutions, including solar cells, water-splitters, capacitors, and batteries. I am exploring how machine learning techniques can help to identify material properties and be used to aid in materials discovery. I have worked with various companies to identify viable materials for applications in renewable devices. I am starting to investigate how material interfaces affect operation of devices for space applications. I now use machine learning methods, such as neural networks, to identify links between material structure and properties not previously known. The aim is to provide a shortcut to determining interface properties without the need for computationally expensive modelling.

Current projects: I am currently funded by the RAEng UKIC scheme to explore the operational limits of devices through exploration of their interfaces. Through computational modelling of material interfaces and use of new machine learning methods, the aim is to show how interface properties can be predicted. I am aligned with Dr Hepplestone's research group.

I have developed the ARTEMIS software package to aid users in identifying the most energetically favourable interface to form between any two given materials. I am also currently developing the RAFFLE Fortran and Python library to predict new material phases at interfaces. This library will be publicly available by the end of 2024. 

I am also developing a Fortran-based open-source neural network library, ATHENA, allowing users to easily implement machine learning into their Fortran-based workflows.

Collaborations: I am always looking to work with other academics and industrial partners on projects within, and beyond, my scope of research experience. Collaboration is a key way to improve the transfer and application of knowledge, and ebales the generation of new ideas.