REFRIGERATION is a hot topic. But forget the perils of partially defrosted
chicken, the real problem is how to replace CFCs鈥攖he refrigerants largely
responsible for turning the ozone layer into Swiss cheese. The hunt for the
perfect replacement is far from easy. Some candidates turn out to be powerful
greenhouse gases, others are poisonous, and some could explode if your fridge
springs a leak. But such solutions are all based on the same
assumption鈥攖hat fridges will forever require a compressed-gas system that
cools by allowing the gas to expand. If we could do away with compressor-based
cooling, all these nasty chemicals and their problems could go too. In short, we
need a refrigeration revolution.
Now a British-based company called Borealis Technical claims it might have an
answer. Its technology relies on an electrical device called a vacuum diode.
Developed under the name of Cool Chip, it chills in much the same way that a hot
cup of tea cools as the hottest molecules evaporate from its surface. The Cool
Chip contains two thin films, separated by a narrow vacuum layer. Put a voltage
across the gap and the most energetic electrons on the negative side 鈥渂oil鈥� off,
carrying their kinetic energy to the positive side of the device. As the hottest
electrons leave, the negative side or cathode gets cooler. Stick the cathode up
against a metal plate connected to the inside of a fridge and it might be
possible to build a safe, chemical-free cooler. The extra energy in the
electrons reaching the positive anode is simply dissipated as heat.
This phenomenon, known as thermionics, is not a new discovery鈥擳homas
Edison spotted it back in 1883. But the effect was virtually ignored until
Gerald Mahan published a paper on thermionic refrigeration in 1994. 鈥淵ou could
say I founded the field,鈥� says Mahan, a physics professor at the University of
Tennessee at Knoxville, and a senior researcher at the nearby Oak Ridge National
Laboratory.
Advertisement
However, Mahan鈥檚 paper concluded that thermionic refrigeration couldn鈥檛 work.
He believed that you would have to supply so much energy to make the electrons
leave the surface of a material that the process could never be efficient at or
below room temperature. But from what Mahan has seen of the patents granted to
Borealis, he is now beginning to revise his ideas. 鈥淚t looks very interesting,鈥�
he says. 鈥淚f the results are correct, the device should work.鈥�
Mahan鈥檚 original paper envisaged two major obstacles to room-temperature
thermionics. The first would be finding materials that emit electrons easily
enough. The temperature at which an electron can be enticed away is indicated by
a property of the material called its work function鈥攁 measure of how well
the electrons are bound within a material. The lower the work function, the
lower the temperature for electron emission. To emit thermionically at 0 掳C,
a material needs a work function of around 0.3 electronvolts.
To find suitable compounds, Borealis Technical has drawn on the work of James
Dye, a chemistry professor at Michigan State University, who devised two classes
of unusual but related compounds, called electrides and alkalides. Electrides
are salts of alkali metals such as lithium, sodium or caesium, which have a lone
valence electron in their outer shell. In an ordinary salt such as sodium
chloride, the valence electron deserts its metal atom for a neighbouring
chlorine atom, turning it into a negative chloride ion. But in electrides, the
electrons are the negative ions. They remain trapped in cages or channels within
the compound鈥檚 crystalline structure by large ring-like molecules, such as crown
ethers, made from carbon and oxygen or nitrogen atoms. The result is that the
valence electrons are 鈥渓oosened鈥�, making them easier to pull away from the
material. A similar effect occurs in alkalides, but in this case, alkali metal
atoms make up both the positive and negative ions. These unusual classes of
compounds have allowed Dye to achieve work functions as low as 0.2
electronvolts鈥攍ow enough for thermionic emission at 鈥�80 掳C.
鈥淲e don鈥檛 understand the nature of this cold-electron emission,鈥� says Dye. It
might be due to a complex molecular framework which holds the ions apart and
reduces the attraction between the ions and electrons. 鈥淲e do know that if we
make the materials very pure, we鈥檙e not able to reproduce the results. That
makes me think that some kind of defect must be responsible,鈥� he says.
A big drawback with Dye鈥檚 electrides is that they are unstable at room
temperature, although he expects to find more stable versions in the future. 鈥淚n
principle, we know where to go but we haven鈥檛 found the right compounds yet,鈥� he
says. But others are working on the problem too. Isaiah Cox, chief operating
officer of Borealis, says the company has already found compounds that are more
stable than those that Dye has been working on.
The other problem envisaged by Mahan is actually making the devices. Whereas
Mahan based his calculations on a gas-filled gap of a millimetre or so between
the electrodes, Borealis鈥檚 chips will require the electrons to leap across a
vacuum just a micrometre or so wide. Reducing the gap means the electrons need
less energy to make the jump. But Cox says such tight specifications should not
be a problem. 鈥淭he device builds on other semiconductor technologies, such as
flat panel displays,鈥� he says. 鈥淣obody says it will be impossible to build
迟丑别尘.鈥�
Borealis carried out a proof-of-concept experiment in Texas last year. 鈥淲e
were using unoptimised materials and the spacing between the electrodes was so
large that it could never be efficient,鈥� says Cox. 鈥淏ut we turned on the current
and got instant cooling.鈥� Rick de Vos, manager of systems engineering for
Frigidaire鈥檚 home products division has been monitoring Borealis鈥檚 progress: 鈥淚
sent someone down to see the demo and we saw a cooling effect at room
temperature.鈥� This has never been seen before, he says.
鈥淭he whole project hinges on whether they can find a material with a work
function of less than 0.3 electronvolts. I know of two labs that have already
managed it,鈥� says de Vos. 鈥淔rigidaire will become interested once they have a
working prototype. If their equations bear out and the efficiencies are what
they expect, it really could revolutionise the industry.鈥�
Typical refrigerators operate at between 30 and 50 per cent of their
theoretical maximum efficiency. According to calculations, a Cool Chip could
offer up to 80 per cent efficiency, a huge improvement. Cox says each chip
should offer a cooling capacity of around 3 watts per square centimetre, and a
typical domestic fridge would need a panel of 25 chips covering an area 5
centimetres square. In addition, they would be totally silent with no moving
parts to go wrong, unlike today鈥檚 compressor-driven fridges.
But it鈥檚 not just fridges that could benefit. The tiny size of Cool Chips
would make them perfect for cooling down overworked microprocessors. Heat is a
major problem in the design of complex chips, which can generate more heat per
unit area than a halogen lamp. So improved cooling should help in the design of
faster, more powerful computers. Boeing has also shown an interest, as
lightweight coolers might be ideal for use in airliners for chilling food and
electrical equipment.
So how far away is a working Cool Chip system? 鈥淲e feel we can have a
mass-production prototype in 24 months, maybe six months if we find the right
materials,鈥� says Cox.