Reliable, Scalable and Affordable Thermoelectrics: Spin Seebeck Based Devices for Energy Harvesting

As part of the Energy Efficiency Directive, the UK has committed to a 20% increase in energy efficiency, a reduction of greenhouse gas emissions by at least 20% and an increased share of renewable energy sources (compared to 1990 levels) by 2020. To address these challenges a stable and diverse range of energy sources will need to be developed and, unsurprisingly, this has been the focus of an intense international research effort. The associated research challenges can be loosely categorised into renewable sources (solar, wind, tidal), sustainable sources (e.g. carbon capture, fusion), and micro generation (e.g. energy harvesting from thermal, light, sound, or vibrational sources). One example of such sources is the harvesting of waste heat with thermoelectric generators (TEGs), a technology that has the advantage of reliability (no moving parts), but is limited by high costs (use of critical elements such as Te) and low efficiencies (<10% for a 200K temperature difference). Given the abundant sources of waste heat in everyday life (boilers, engines, computers, district heat networks), development of low-cost TEGs that could easily be applied to various surfaces could present a significant vector for change. For example, harvesting just 5% of the energy lost as waste heat by car engines in the UK would save the equivalent of 1 hundred thousand equivalent tonnes of oil per year (or ~1% of the UK's total energy usage in 2014).

Conventional TEGs are typically based on the Seebeck effect: a physical process that results in the generation of an electric current when a temperature difference exists between two ends of a material. One of the bottlenecks for improvement of the efficiency of these devices is the co-dependence of two key material properties: the thermal and electric conductivity. Whilst some progress has been made to circumvent this by nano-engineering, there is still some way to go before widespread commercialisation becomes viable. This could, however, be overcome with TEGs based on the spin Seebeck effect, where an additional degree of freedom - the spin of the electrons - results in a device architecture that scales with surface area (unlike conventional thermoelectrics), enables separation of the thermal and electric conductivities that drive the efficiency of the device and boasts active materials that could be sourced from abundant sources (such as iron or copper, rather than bismuth telluride).

The aim of this Fellowship is to investigate the spin Seebeck effect with regards to its application as a TEG. There are 5 key challenges that will be addressed:
(1) precise determination of the efficiency of such spin Seebeck based TEGs;
(2) discovery of new materials (from abundant sources);
(3) development of prototype TEGs;
(4) identifying the controlling factors with regards to the efficiency of the overall device; and
(5) understanding the underlying physics of this effect.

For example, harnessing the maximum spin polarised current generated by the spin Seebeck effect typically requires the use of expensive platinum contacts. For such technology to become economically viable would therefore require discovery of cheaper alternatives, such as the doped metals that will be investigated. In addition, precise characterisation of the spin Seebeck effect is limited by instrumentation that typically only monitors the temperature difference (rather than heat flow), hence instrumentation will be developed to monitor both these parameters so that the power conversion can be determined. There is also, as of yet, no comprehensive coefficient that can be used to compare different material systems (such as the Seebeck coefficient for conventional thermoelectrics), nor a rigorously tested figure of merit. Once this has been established, a comprehensive comparison of different materials and engineering of the overall device can be made.

Thermoelectric effects can be used to convert low grade heat into electricity (and vice versa), however, their widespread application is currently limited by high production costs and low efficiencies. The work highlighted in this Fellowship is targeted at a new breed of thermoelectric generators - spin Seebeck TEGs - that could be used to break through the 'efficiency barrier' that limits conventional devices. This will involve discovery of new materials for the conversion of spin currents, development of instrumentation to accelerate the standardisation of measurements, development of prototype TEGs, and the engineering of such devices to optimise their efficiency.

In principle, development of this technology could be used to siphon electricity from energy that is stored as heat, either from a district heat network, or from waste sources such as engines or boilers. Not only will this complement current activities in securing a resilient energy portfolio (for the UK), but could result in the generation of income through industrial partnerships or spin-outs intended to capitalise on the intellectual property generated. This could therefore have a significant impact on society, the economy, and the computing industry, as outlined below.

1. Society and the public
An increase in efficiency of thermoelectric generators (TEGs) could have a significant impact on society over the next 10-50 years as they become more economically viable. For example, a reduction in energy usage is arguably one of the key societal challenges for the next century and over the next 5-20 years a reasonable improvement of the efficiency of a TEG from 7 to 20% could be expected as a result of this work. Given that up to 60% of the energy in fuel burnt (in car engines for example) is lost as waste heat, such an improvement, coupled with low cost implementation that is the focus of this work, could easily save >3% of the UK's annual energy usage.

Not only would TEGs be useful as energy harvesters, but they could also serve as vectors for energy storage as they allow for conversion of heat to electricity (and vice versa). Any technology that can utilise or direct heat in this way will contribute to the stability of the UK's energy portfolio and thus contribute to stable or reduced energy costs (in the long-term).

Delivering impact from this will require the development of prototype devices in order to demonstrate proof of concept, which will then be followed by discussion with potential industrial partners.

2. Emerging industries and the UK economy
Although TEGs are not a new concept they have been limited by material costs and relatively low efficiencies. Recent improvements in efficiency have required nano-engineering of the basic material that can be difficult to up-scale. Should a new device be developed that could be used to circumvent the problem of low efficiency and fabrication costs then it is highly likely that it will lead to a new (industrial) market. This could, for example, bring in new jobs for production and distribution of the TEGs, or income from patented technology.

3. The computing industry
For the last 40 years the miniaturisation of transistors in computers has closely followed Moore's law but as transistor density increases, the increasing power density leads to instability of the device (due to heat). Spintronics is one research field that aims to move beyond Moore with the manipulation of spin polarised currents and which will see impact from this work.

This will have both economic and technological impact on society as it will significantly reduce the heat barrier faced by downscaling components that is common to charge transport based devices. As a result this could lead to new technology and more efficient computing (technology exports and reduced energy costs). Due to high consumer demand it is realistic to expect commercialisation of this technology within the next 5-20 years.