Non-equilibrium electron-ion dynamics in thin metal-oxide films

Lead Research Organisation: University of York
Department Name: Physics

Abstract

Recent estimates suggest there are now over 3 billion mobile phones and 1 billion personal computers in use worldwide. The total energy consumption associated with such devices is growing and is predicted to triple by 2030, becoming equivalent to the current residential electricity consumption of the US and Japan combined (Gadgets and Gigawatts - Policies for Energy Efficient Electronics, 2009). Given the environmental costs associated with energy generation and storage, improving the energy efficiency of electronic devices is now an urgent priority.

The key to reducing the energy consumption of electronic devices is better control of the electric currents flowing within them. Crucially, this is often dependent on the properties and robustness of thin metal-oxide (MO) films. For example, insulating MO films are used to separate metallic and semiconducting electrodes in transistors. During operation, the voltage applied between the electrodes causes current to leak through the MO film, causing wasteful energy consumption. Over time, leakage current can grow and lead to a more terminal problem whereby the MO film abruptly becomes highly conducting, a process known as breakdown. These deleterious effects are becoming increasingly important as transistors are ever further miniaturised to meet consumer demand for increasingly powerful devices. On the other hand, the reversible switching of a MO film between insulating and conducting states by applying voltages has recently received interest as the basis for a non-volatile and low-power memory technology. For transistors, memristors and many other oxide-based electronic devices there is speculation that electron trapping by defects, polycrystallinity, electric fields and redox reactions at the electrode, all play important roles, however, there are few theoretical models which take these factors into account.

The main aims of this fellowship are to learn how structure and composition are related to the electrical properties of thin MO films sandwiched between conducting electrodes, and to understand the mechanisms responsible for the transformation of these properties by application of a voltage. This will provide a framework for understanding leakage current and resistive switching in MO films, and allow strategies to control these effects to be investigated. Materials modelling can play a crucial role in addressing these aims by elucidating processes taking place over a wide range of time- and length-scales, and identifying the critical material parameters. The usual modelling approach is first to determine the equilibrium structure, then to calculate the corresponding electronic properties and current. However, this does not allow for the possibility that the non-equilibrium flow of electrons can modify the structure of the material, e.g. by field driven ion diffusion and local heating. Considering such non-equilibrium effects is essential to be able to model breakdown and resistance switching, and is also important for other processes involving correlated electron-ion dynamics, such as radiation damage. Therefore, the development of a new integrated approach is proposed that can describe the feedback between electron and ion dynamics consistently, resulting in dynamically evolving non-equilibrium structure and properties. It will combine several levels of theoretical modelling to describe the polycrystalline film structure, including defects and interfaces, the associated electronic and thermodynamic properties, and the coupled non-equilibrium dynamics of both electrons and ions. Through close collaboration with project partners, models will be tested and refined. Ultimately, this will feed into the electronics industry, leading to the design of more efficient and more reliable devices. In the later stages of the project the methodologies developed will be extended to address related materials challenges for applications including solid oxide fuel cells and batteries.

Planned Impact

One of the main aims of the proposed research is to work with academia and industry in order to learn how to reduce the energy demands of electronic devices. This aim is in strong alignment with UK and international environmental policies. Achieving this aim will ultimately benefit consumers by making electronic devices which cost less to run and are more reliable. Engagement with the general public throughout the project on how scientific research contributes to addressing these problems will also contribute to increased public awareness and understanding of science.

The microelectronics industry will benefit from the proposed research as they can exploit the results to help design new and improved products. Many of the companies who will benefit in this way have a strong presence in the UK, including industry leaders, such as ARM, Intel and IBM, and therefore this will enhance the UK economy. The Department of Physics has existing strong links with Intel, providing opportunities for dissemination and exploitation of results (for example, by partnering with SMEs based in the UK). The relevance of the proposed research to industry is demonstrated by the involvement of SEMATECH as a project partner. Academia will benefit directly from SEMATECH's involvement as they will co-sponsor a PhD studentship at the University of York.

Aside from academic impact in the research area of the proposed project, the wider scientific community will benefit from the development of new methodologies and the fundamental understanding of non-equilibrium processes in metal oxide materials (see Academic Beneficiaries and Case for Support). This will encompass such diverse fields as; (photo-) catalysis, earth science, electrochemistry, bio-engineering, superconductivity, photonics, environmental science and bio-chemistry.

Many technological industries in the UK rely on a constant supply of suitably trained graduate and postgraduate students. The employment and training of two PhD students and a PDRA on this project contributes to meeting this demand. However, if they choose to move into non-technology based employment sectors, businesses can benefit from the advanced analytical and research skills they will posses. Overall, this contributes to the UK's economic and scientific competitiveness.

Publications

10 25 50
 
Description This Fellowship aimed to understand the structure and electrical properties of polycrystalline oxide films, and their transformation by electrical voltages, in order to help improve the energy efficiency and reliability of resistive switching memories (ReRAM) and related oxide-based electronic devices (e.g. MRAM). Over this period KM has published 28 papers (including many in high-impact journals such as Nature Communications, JACS and Nano Letters), acquired > 2000 additional citations and given >25 invited talks demonstrating the sustained impact he is delivering. A brief summary of progress towards the project objectives (RU1-4 in the original proposal) follows:

Interface, extended and point defects (RU1/RU2 - 18 publications): The structure and electronic properties of Cu/TiO2, TiN/HfO2, Pt/HfO2, CoFeB/MgO and SrTiO3/ Fe3O4 interfaces have been characterised at the atomic level with models validated against experimental data. In each of these oxides the properties of extended defects such as grain boundaries (GBs) and dislocations as well as segregation of point defects has also been investigated. Highlights include the discovery that highly stable antiphase boundary defects in Fe3O4 reduce performance in spintronic devices, that TiN and Pt electrodes act as oxygen diffusion barriers in ReRAM, and that control over oxide stoichiometry is the key to ensuring high performance in MgO-based MRAM devices. The structure of GBs near oxide/electrodes interfaces (MgO/CoFe and HfO2/TiN) has also been recently modelled for the first time, providing atomistic insight into forming in ReRAM devices.

Electron-ion transfer processes (RU3 - 6 publications): New first principles approaches for predicting the rates of defect assisted electron transfer have been developed in collaboration with Jochen Blumberger (UCL) and applied to point defects in MgO and HfO2. Methods for simulating the diffusion of polarons in oxides have also been developed, including their interaction with point defects and GBs. In TiO2 it was demonstrated that GBs significantly reduce electron mobility with relevance to both ReRAM and photovoltaic applications. We have also demonstrated how defects in oxides can be generated dynamically in the presence of non- equilibrium electric fields and charge trapping. These methodological advances mean the key parameters needed to model resistive switching can now be computed directly from first principles.

Transformation of electrical properties (RU4 - 4 publications): Combining the theoretical predictions described above with experimental material and device characterisation it has been possible to provide atomic level insight into resistive switching mechanisms. For example, the role of GBs as preferential locations for enhanced leakage and breakdown has been demonstrated. The stoichiometry of HfOx has been identified as key materials property governing the formation of metallic filaments during forming, with x=1.50-1.75 identified as optimal. To date KM has published 15 papers on atomistic modelling of materials with relevance to ReRAM which have been highly cited (>800 citations) demonstrating the impact they are having in the field. Building on the achievements in RU1-3 KM is now developing methods to model the entire resistive switching process in TiO2 and HfO2 from first principles. This will allow strategies to improve the reversibly and reliability of resistance switching to be investigated. Two further publications are in advanced draft stage and due to be submitted in the next few months.

We are fully on course to meet our original objectives and a number of new research directions have also developed which will lead to important discoveries.
Exploitation Route The findings are already being used in academia, as demonstrated by citations. We are aware of a number of cases where our methods and ideas are being applied by other groups. Some of the findings (in particular in relation to resistive switching) are being taken on board by industry in the development of devices.
Sectors Chemicals,Digital/Communication/Information Technologies (including Software),Electronics,Energy

URL http://www-users.york.ac.uk/~km816/index.shtml
 
Description Underpinning research into resistive switching in oxides and the role of polycrystallinity is being taken up by industry in their efforts to develop a low power, effcient, non-volatile and high capacity data storage technology. The project partner SEMATECH and many other citing groups worldwide are using our findings in order to model resistive switching device characteristics and improve their performance.
First Year Of Impact 2012
Sector Digital/Communication/Information Technologies (including Software)
Impact Types Economic

 
Description Retreat on EPSRC future support for early career researchers
Geographic Reach National 
Policy Influence Type Membership of a guidance committee
 
Description COST Action CM1104 - Short term scientific mission
Amount € 1,220 (EUR)
Organisation European Cooperation in Science and Technology (COST) 
Department COST Action
Sector Public
Country Belgium
Start 03/2015 
End 04/2015
 
Description EPSRC Early Career Fellowship extension
Amount £588,608 (GBP)
Funding ID EP/P023843/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 01/2018 
End 12/2020
 
Description EPSRC responsive mode
Amount £798,645 (GBP)
Funding ID EP/P006051/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 01/2017 
End 12/2019
 
Description EPSRC responsive mode
Amount £1,191,641 (GBP)
Funding ID EP/K03278X/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 10/2013 
End 09/2017
 
Description Undergraduate summer student bursary
Amount £1,440 (GBP)
Organisation Institute of Physics (IOP) 
Sector Learned Society
Country United Kingdom
Start 06/2013 
End 08/2013
 
Description Collaboration on fillament structure in HfO2 based resistive switching devices 
Organisation National Institute for Materials Sciences
Country Japan 
Sector Academic/University 
PI Contribution Provided models of the structure of metallic fillaments in HfO2.
Collaborator Contribution International SEMATECH - Production of samples for atom probe analysis National Institute for Materials Science NIMS - 3D atom probe analysis of samples
Impact This collaboration in ongoing.
Start Year 2013
 
Description Collaboration on fillament structure in HfO2 based resistive switching devices 
Organisation University at Albany
Department International SEMATECH
Country United States 
Sector Charity/Non Profit 
PI Contribution Provided models of the structure of metallic fillaments in HfO2.
Collaborator Contribution International SEMATECH - Production of samples for atom probe analysis National Institute for Materials Science NIMS - 3D atom probe analysis of samples
Impact This collaboration in ongoing.
Start Year 2013
 
Description Membership of COST action CM1104: Reducible oxide chemistry, structure and functions 
Organisation European Cooperation in Science and Technology (COST)
Country European Union (EU) 
Sector Public 
PI Contribution Attending working group meeting of COST action CM1104: Reducible oxide chemistry, structure and functions. Presenting results and participating in discussions on key challenges in the field.
Collaborator Contribution Presenting results and participating in discussions on key challenges in the field. Critical discussion and input into my research programme.
Impact Results on Grain Boundary Controlled Electron Mobility in Polycrystalline Titanium Dioxide presented at COST meetings and discussed. Led to publication "Grain Boundary Controlled Electron Mobility in Polycrystalline Titanium Dioxide", S. Wallace and K. P. McKenna, Advanced Materials Interfaces 1, 1400078 (2014).
Start Year 2013
 
Description Centre for Energy Efficient Materials 
Form Of Engagement Activity Engagement focused website, blog or social media channel
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Industry/Business
Results and Impact Led the establishment of the York Centre for Energy Efficient Materials (https://www.york.ac.uk/ceem/) which is focussed on industrial engagement and is holding a launch event on March 28 which is being attended by representatives from business, EPSRC and innovate UK.
Year(s) Of Engagement Activity 2016,2017
URL https://www.york.ac.uk/ceem/