Tom Shelley reports on the imminent commercialisation of a solid-state technology set to revolutionise the direct conversion of heat to electricity
Novel solid-state devices that convert heat directly to
electric power have a claimed efficiency of 70-80% of the
theoretical Carnot Cycle heat engine efficiency.
They rely on the thermionic emission of electrons across a very
narrow gap between electrodes. The work functions of the
electrodes have been optimised for maximum transfer by
engineering the emitting surface to take advantage of a
previously obscure quantum effect.
First commercial applications are expected to be in satellites,
aerospace power generation, waste heat recovery from steel and
aluminium mills, generators taking internal combustion engine
waste heat and converting it to electricity, and geothermal power
generation. Ultimately – if the work can be scaled up –
they may replace internal combustion engines and industrial power
generating equipment.
Numerous research groups are developing direct solid-state
methods of converting heat to electric power with much higher
efficiencies than conventional thermocouple-type devices (Seebeck
effect) – and with potentially higher efficiencies than
internal combustion engines and steam-raising systems coupled to
electric generators (see box).
The crucial quantum effect in the latest devices was theorised,
researched and discovered by Avto Tavkhelidze, who is based at
the Tbilisi State University in Georgia. All his inventions are
assigned to Borealis, which started out as a Canadian mining
company, Borealis Exploration, but which is now an assembly of
mainly US-financed technology companies headquartered in
Gibraltar. Other researchers engaged in the project include Dr
Alan Feinerman of the University of Illinois at Chicago, a team
at Portland State University led by Dr Gertrude Rempfer, and Dr
Marty Kordesch at Ohio University.
At the recent Clean Energy Investment Showcase in London,
organised by the Centre for Sustainable Engineering in
Peterborough, a Power Chips spokesman – who described
himself as a “serious shareholder and business development
consultant” – explained that it relies on electrons
moving from a hot material with a low work function, across a gap
less than 1 micron wide, to a cold material with a low work
function. The work function is the amount of energy required to
take an electron out of a solid material. The gap prevents heat
flow and backflow of current, which is a problem that makes the
best conventional thermoelectric devices less than 10% efficient.
The spokesman is a business angel, and asked to remain anonymous
so as not to be ‘bombarded’ with funding enquiries.
He explained: “The Russians did a lot of the basic research
in this field decades ago. We took on the principal scientists
including Avto Tavkhelidze in Tbilisi – where the majority
of the development work is still being done.”
He showed Eureka a sample cell, with an active area about 10mm
across, which appeared to have a textured internal silicon
surface. He mentioned that, run in reverse, the same devices can
be used for cooling – in which application they are referred
to as “cool chips”.
This Avto Metals texture, which consists of ridges and grooves
about 250nm across, is crucial. The grooves reduce the number of
possible standing wave energy states that electrons can occupy
within the solid – making them less likely to stay in the
material, and more likely to leave. The spokesman said that as
regards work function: “We are heading towards 1eV, but we
can play around with the surface structure. Gaps are 500nm to 1
micron and builds are getting down to 2 Angstroms [0.2nm]
flatness and even smaller gaps.”
However, he says there is no point making gaps smaller than 0.1
microns – and in many cases can be up to 5 microns. He
claimed that a Power Chip diode with an Avto layer could be made
in less than 20 steps, which was far fewer than for other
technologies.
“Simpler technology leads to lower costs,” he added.
In their current form, the devices have a projected power output
of 10-40W per square centimetre when the work functions get low
enough and can operate at up to 250ºC. Studies on lifetimes for
the devices suggest that they should be of the order of 2 years
at 1,600K (1,327ºC) and 10 years at 1,200K (927ºC).
The spokesman said: “The initial sale prices for Power Chips
will go as high as tens of thousands of pounds per Watt.”
While he declined to be drawn on which applications he had in
mind, he did mention “applications out of this world”
– and an obvious use would be converting heat from small
reactors in military satellites and deep space probes into
electric power – something currently done very inefficiently
using Seebeck devices. The spokesman claimed that existing
reactors relied on 78kg of plutonium because the power conversion
efficiency is so poor.
He cited a study showing that geothermal resources using existing
technology could support 35-70GW of worldwide electrical
generating capacity. Currently, geothermal power stations use
steam, heat or hot water to rotate turbine generators to produce
electricity. With Power Chips, most of such plant would become
redundant.
“Power Chips will competently make more geothermal resources
viable,” he said. “They won’t introduce any
additional energy or carbon emissions, and will extract energy
from what are currently considered non viable sources of
power.”
Regarding internal combustion engines, he said: “Typically,
they achieve 20% of maximum efficiency. Power Chips will
eliminate the alternator by converting waste exhaust heat
directly into electricity.
He said that a 100kW engine has 200kW of waste heat available.
Power Chips could improve the efficiency and emissions of normal
and hybrid vehicles by 75%, he claimed.
“In the future we expect Power Chips will be able to harness
body heat, to power the likes of pacemakers,” he said.
“Many more uses will become feasible. We are working on
solar power versions, to replace photovoltaics.”
Lots of runners in heat-to-power race
We know of more than a few other developments aimed at producing
solid state devices that can turn heat directly into electricity.
Most are coy about their technological basis, but of those
willing to communicate, JX Crystals in Issaquah, Washington is a
spin-off from Boeing – with licenses for patents on
infrared-sensitive gallium antimonide photovoltaic cells.
The company’s first commercial Thermo Photo Voltaic (TPV)
product was what they called their “Midnight Sun Stove”
which used the cells to produce 100W of electricity as well as
25,000 BTU/hour of heat. The company’s Jim Avery says:
“We are working on a residential TPV micro-CHP product that
would provide all the heat, water and electricity for a home. The
overall efficiency from fuel to heat and power is quite high.
Fuel to electric will be about 10%.”
Electric power output will apparently be 1.5kW. The company is
headed by founder Dr Lewis Fraas. Funding for a new family of
mirror-enhanced solar cells – also being developed by the
company – has been provided by the Shanghai Science and
Technology Committee. A 100kW demonstration unit is under
construction in Shanghai with an additional 300kW now on order.
Also likely to end up in the Far East is a rival ex-Soviet Union
solid state heat to power conversion technology being developed
in St Petersburg. The technology, said to be derived from silicon
photovoltaics, claims an efficiency of up to 80% of Carnot. All
was to be revealed to Eureka until we heard they were in
negotiation with Sharp Electronics, the world’s largest
manufacturer of conventional photovoltaics.
Professor Ted Sargent, at the University of Toronto, on the other
hand, has invented a technology based on quantum dots – nano
particles of semiconductor – combined with polymer that can
be spun coated onto a glass substrate patterned with electrodes.
We do not know what the polymer is but we do know his first
experiments used oleic acid, the main ingredient in olive oil.
The resulting cells respond to infrared light. Efficiencies are
still very low but the manufacturing technology is very cheap. We
have since heard that the same – or very similar – idea
is being pursued by researchers in Thailand using straight olive
oil as the dispersant.
http://www.powerchips.gi
http://www.jxcrystals.com
Pointers
* Devices convert heat directly to electric power at efficiencies
said to be 70-80% of theoretical maxima as defined by the laws of
thermodynamics
* Projected power outputs are 10-40W per square centimetre
* Initial target markets with be high value applications –
probably in space – but the long-term target is to replace
internal combustion engines and steam-raising systems
* Run in reverse, the same devices can be used for cooling
For more technical
developments see www.eurekamagazine.co.uk