ÐÓ°ÉÔ­´´

The man from Disney World wanted to buy some immediately. His dancers
wore clothes that lit up: they had to have 200-volt battery packs strapped
to their backs, and he wanted to lighten their load if he could. He was
calling Frank Karasz, a materials scientist at the University of Massachusetts
in Amherst, who had just announced that he had made a piece of plastic that
could shine blue when only 6 volts was applied to it. Karasz had to explain
that his light-emitting plastic, with a working life of only a few seconds,
was nowhere near commercial quality. But while the blue polymer isn’t ready
for the market, the 60 phone calls Karasz received in a single week show
that the market is ready for it.

Producing a plastic that can emit light would have seemed outlandish
just two or three years ago. Even the idea that plastics can conduct electricity
comes as a surprise to most people. After all, plastics are familiar as
insulating sheaths round wires and as the casings around all manner of electrical
appliances. But in the mid-1970s, a group of scientists discovered that
plastics do not have to be insulators, when they stumbled across a polymer
that behaved as a semiconductor. This discovery opened up the prospect of
materials that might combine the electrical properties of metals with the
ease of processing and versatility that have made plastics a runaway success
over the past half-century. These pioneers predicted that a world of plastic
wires and batteries was just around the corner.

It hasn’t happened quite as planned. The plastics that conducted electricity
were hard to fabricate and they were pretty poor conductors when compared
with metals. But interest in these materials continued to grow. The past
few years have seen the long-predicted technology creeping closer to reality,
as scientists have learnt how to make plastics that can conduct electricity
as well as metals. Along the way, researchers have learnt how to make these
plastics emit light, and the materials have been seized on by people who
see in them a way of making plastic television screens and computer displays.
There are still plenty of obstacles blocking the road that could, one day,
lead to these applications, but chemists, physicists and engineers are daily
learning much more about how the materials are made, the way they work,
and strategies for improving them.

Like conducting plastics, the new light-emitting materials are conjugated
polymers. A polymer is a giant molecule that is formed from hundreds, or
even thousands of smaller molecules called monomers. The archetypal polymer
is polyethylene (polythene), which consists of long chains of carbon atoms,
with two hydrogen atoms attached to each carbon. The molecules are held
together by the sharing of the electrons that belong to their constituent
atoms.

A carbon atom has four electrons available for bonding, and a hydrogen
atom has one. When a carbon atom bonds to hydrogen, one of its electrons
starts to do double duty by orbiting both atoms. The electron from the hydrogen
atom also orbits both partners. The two shared electrons between them constitute
a single covalent bond.

In a polyethylene chain, each carbon atom shares one of its four electrons
with each of the carbons next to it, and one each with the two hydrogens
attached to it. It is these bonds that hold the polymer together, but the
electrons that form them are fixed. They cannot leave their prescribed
orbits, and because an electric current is nothing more than the free
movement of electrons, polyethylene cannot conduct electricity.

The key to making conducting polymers is to provide electrons which
are not fixed in their positions. Remove one hydrogen from each carbon,
and you leave the atoms with an extra electron to play with. The extra electron
will form a second bond with just one of the neighbouring carbons on the
chain, creating a double bond with that atom, while the other neighbouring
carbon atom remains attached by a single bond. The result is a sequence
of alternating single and double bonds all the way down the carbon backbone.
In this regular ‘conjugated’ sequence, the extra electrons can be induced
to move along the polymer chain, forming an electric current, which is why
conjugated polymers conduct electricity.

But the first such material discovered, by the pioneering group at the
University of Pennsylvania in 1975, had an electrical resistance several
million times as high as that of copper. Moreover, the material was brittle
and difficult to process using standard production techniques, in which
the polymer is dissolved and then spun to form fibres or cast into a thin
film. But the pioneers persevered: Alan MacDiarmid, along with Hideki Shirakawa,
now at the University of Tsukuba in Japan, and Alan Heeger, now at the University
of California in Santa Barbara, kept working with the new material until
they discovered that by ‘doping’ the polymers – adding a small quantity
of another material – they could increase the conductivity.

MacDiarmid likens doping a polymer to adding a single drop of red dye
to a pale cake mix. ‘The whole cake becomes pink,’ he says. ‘The one drop
of red food colouring is a tiny, minuscule volume of the whole cake, but
it has enormous effect on one property, the property of colour. The red
dye is now an intrinsic part of the cake, you cannot undye it.’

Changing properties

Similarly, it is possible to change a particular property of a conducting
polymer by doping it. Through trial and error, the group discovered that
soaking the polymer in a solution of iodine removes about 10 percent of
the conducting electrons, leaving a polymer with some positive charges,
and negatively charged iodide ions lying alongside. With fewer electrons
in the polymer, those that remain can move more freely – just as a few marbles
will roll freely around a box, while marbles jam-packed into a full box
will barely move at all.

Doping raised polymer conductivity by a factor of a million or so, but
the polymers were still nothing like as good as copper at conducting electricity.
Two things are essential for good conductivity: having a lot of movable
‘carriers’ in the form of electrons (or their opposite numbers, holes in
an array of electrons), and making sure that they can travel through the
polymer easily. The number of carriers did not pose a problem, because conducting
polymers have thousands of times as many available electrons as familiar
inorganic semiconductors such as gallium arsenide. On the other hand there
was plenty of room to improve mobility.

Polymer chains typically coil up like a mass of tangled string. Electrons
travel along the molecules more easily than they travel between the chains
of molecules, so straightening out each skein provides the electrons with
an easier path. To create such a path, the researchers focused on ways
of increasing molecular order. They found that one simple way of doing
this was to stretch the polymers out into a thin sheet, which tended to
pull the molecules into alignment with each other.

In 1992, MacDiarmid and his team discovered an additional strategy:
with the skeins straightened out, packing them together as tightly as possible
further increases conductivity. So they increased the polymer’s crystallinity,
because two molecules can be no closer than when they are part of a crystal.
Clearly, a densely packed polymer would allow electrons to jump almost effortlessly
from chain to chain. Such insights brought this polymer’s conductivity up,
though still behind copper’s conductivity at 600 000 siemens per centimetre.

Then, about a year ago, MacDiarmid’s group, in a single step, managed
to improve the conductivity of their polymer a thousandfold, bringing its
conductivity into the range typical of metals – though still lagging behind
copper by a factor of 10 000. MacDiarmid’s trick was to change the solvent
used to dissolve the polymer when forming it into a thin plastic sheet.
Instead of chloroform he used fluorophenol, which worked almost as a ‘secondary
dopant’ to improve the polymer’s conductivity. At the time, the precise
mechanism was a mystery, but MacDiarmid has since come up with a theory.
He thinks that the solvent coats the negative iodide ions created by the
primary dopant, shielding the molecules of the polymer backbone from their
influence. In the absence of the neutralising effect of the negative ions,
the positive molecules begin to repel each other, though they remain bound
together. Under these constraints, the only way they can move farther apart
is by straightening themselves out.

Despite their imperfections as conductors, conjugated polymers have
from the outset inspired interest in possible commercial applications. MacDiarmid
holds a patent for plastic rechargeable batteries, and other people are
studying the possibility of using the material to ward off static. Normally,
the static charge that builds up on medical equipment is conducted away
to earth as the user makes contact from time to time with a metal earthing
plate. A conducting polymer shield could continuously drain away electric
charge, maintaining a static-free environment.

Electrical conductivity is not the only application of conjugated polymers.
Certain materials can change their properties – such as transparency – when
a voltage is applied. A plastic that can be darkened electrically could
be used on windows to keep heat out when the sun shines brightly on hot
days, and made transparent when the weather is colder. Medical scientists
are also considering the polymers as a way of releasing controlled quantities
of a drug. The idea would be to infuse a polymer with a particular drug
and attach it to a patient’s skin. Applying a small voltage would change
the polymer’s conductivity, releasing a precisely measured quantity of the
drug.

Despite this plethora of possibilities, most of the recent interest
in conjugated polymers can be attributed to just one thing: the possibility
that they could be used to manu-facture light-emitting diodes (LEDs) out
of plastic. A major driving force for this research is the dream of building
ultra-slim screens, made entirely of plastics, for TVs and computers.

The first steps on the road to this goal came in 1990, when a multidisciplinary
group at the University of Cambridge worked out how a semiconducting polymer
could be made to give off light. To make the plastic LED, the researchers
sandwiched a 200-nanometre film of the semiconducting polymer poly(p-phenyleneviny-lene),
or PPV, between a transparent positive electrode made of indium/tin oxide
and a negative electrode made of aluminium. The negative electrode injects
electrons into the PPV film, while the positive electrode pulls them out
– a process equivalent to injecting positively charged holes. When an electron
and a hole meet, they give off energy, which is transferred to the polymer
chain. This extra energy kicks an electron into what chemists call an ‘excited’
or high-energy state. It does not stay there long, however, but quickly
releases its energy and returns to its original ground state.

Normally, the electron releases its energy as heat, but it can return
to the ground state by releasing a photon, which we see as light. Unlike
other applications of conjugated polymers, LEDs do not require ever higher
levels of conductivity. On the contrary, if it is made to be conducting,
it will not emit light.

The devices built initially by the Cambridge group gave off a faint
yellow-green light. Now the race is on to make LEDs that can produce colours
across the spectrum, with improved energy efficiency. Hardest of all, they
will have to last as long as their inorganic counterparts. The Cambridge
group and Heeger’s group at Santa Barbara are leading the way. In theory,
producing a rainbow of LEDs appears a relatively simple goal. The wavelength
of the light produced by an LED, and hence its colour, varies with the energy
of the photon, which in turn is exactly equal to the energy that kicked
the polymer up to an excited state. By adjusting the structure of the polymer
backbone, it should be possible to tune the amount of energy the polymer
requires to reach this excited state.

True blue

In the past three years, chemists have managed to modify the polymer
chain to produce most colours of the spectrum. The one that remains most
difficult is a true blue, according to Richard Friend, the physicist leading
the Cambridge group. Several blue polymer LEDs have been made, but none
has survived more than a few seconds’ use. However, scientists are confident
that the problem will be solved. ‘Frankly, I’m not worried about getting
blue,’ says Andrew Holmes, the chemist also leading the Cambridge team.
There is no intrinsic property of conducting polymers that forbids blue
and, through trial and error, everyone agrees that it should be created
soon.

Efficiency and longevity are likely to prove more difficult to achieve.
The first devices consumed large quantities of electrical power and generated
only a faint glow in return: for every 10 000 electrons injected, only one
photon came out. Improving efficiency by reducing this ratio is crucial
if LEDs are to be of any practical use. The main problem is that pushing
electrons into the PPV is much more difficult than pulling them out. The
electrons tend to bump up against the electrode/PPV interface as though
it were an impenetrable wall.

The problem stems from a characteristic known as the ‘work function’
of a material, which roughly describes the amount of energy that an electron
requires to jump from the inside of the material to outside its surface.
The PPV has a higher work function than the negative electrode, so it is
an uphill task for the electron to accumulate enough energy to leap across
the barrier from the electrode to the PPV. For holes the positive electrode
is well matched to the work function of PPV, and the ride in the opposite
direction is much smoother. So scientists have sought ways to help the
electron jump into the polymer.

A key breakthrough came in 1992, when the Cambridge group added another
tier to the LED sandwich. This electron transport layer likes electrons
more than the light-emitting layer, and so eases the electrons’ transition
as they move from the negative electrode. While such a layer lowers the
slope for the electron’s passage, it does so enough to make the holes’ reverse
trip an uphill climb. Since the change in work function impedes the progress
of the holes, they must wait patiently at the PPV interface until they meet
their electron partner. The upshot is that no holes are wasted, and adding
an electron transport layer brought the efficiency up from 0.3 percent to
1.5 per cent.

Last October, the Cambridge group announced a jump to 4 per cent efficiency
– comparable to the performance of the inorganic LEDs that are already used
in commercial information displays and blinking lights on control panels.
The improvement came through modifications to the PPV layer itself. The
chemists at Cambridge created a new cyano-substituted polymer called poly(cyanoterephthalylidene)
in which cyano groups were attached to the PPV chain (see New ÐÓ°ÉÔ­´´,
Science, 30 October 1993). The added cyano groups increased the polymer’s
affinity for electrons and enabled it to serve as both light-emitting and
electron-transporting layer. The Cambridge researchers retained a film
of standard PPV as a hole transport layer, however, because that has always
done the hole absorption job so well.

Efficiency drive

The most efficient commercial tungsten light bulbs convert as much as
10 per cent of the electrical energy they consume into light, but the researchers
working with conducting polymers hope their LEDs will surpass this. There
is some disagreement, however, on exactly how high they will be able to
go. According to one theory, the efficiency is limited to 25 per cent, because
the electron and hole can recombine in four different ways, only one of
which will result in light generated by fluorescence. But Heeger, among
others, holds on to the hope that this theory is too simple and that there
may be ways to get 100 per cent efficiency – one photon out for every electron
in.

The biggest drawback to the polymer LEDs built so far remains their
short working lifetime. To be useful in flat panel displays, a polymer LED
would need a lifetime of at least 10 000 hours. There have been some reports
of LEDs lasting as long as 1000 hours – that is more than a month of continuous
operation – but most live for about 100 hours. Even this modest performance
has only been achieved by running the devices in a vacuum. ‘We can extend
the lifetime if we cool the whole thing,’ says Karasz. ‘But when you’re
trying to make a commercial device, cooling it and putting it in a vacuum
are not exactly designed to make people cry out with joy.’

So little is known about why LEDs die young that most people are hesitant
even to hazard a guess. ‘It’s going to take some hard work,’ says Heeger.
‘We’re just going to have to kill a bunch of these things and open them
up to examine the details of the failure.’ There are as yet no established
protocols for testing devices or for determining mean lifetimes. There is
not even an accepted definition of what constitutes a ‘failed’ device. What
is becoming clear is that the polymer layer no longer conducts electricity
when the LED stops working. ‘The resistance goes through the roof,’ says
Mark Thompson of Princeton University, where a large-scale campaign is being
mounted to tackle the problem of device failure. ‘You can’t get any current
²¹³¦°ù´Ç²õ²õ.’

Under a simple optical microscope, tiny black dots can be seen in failed
devices. These spots represent places where electrons and holes recombine
but emit their energy only as heat. The reason the spots form is a mystery,
but it is clear that once they appear they usually grow. The suspicion is
that each spot forms a kind of sink hole that attracts more and more electrons
without any of the crucial light emission taking place. So a dark spot on
an LED not only ruins its display potential, it also stops electrons travelling
far enough to light up other areas of the LED as well.

All the scientists in the field have high hopes that the middle, light-emitting
layer of the LED sandwich will not prove to be inherently unstable. The
culprit could well turn out to be the electrodes, which are made out of
highly reactive metals such as calcium or magnesium. ‘At this point I am
leaning towards saying it’s the contact,’ says Thompson.

Electrodes made of reactive metals with a low work function are best
for injecting electrons in polymer LEDs. But oxygen and water from the
atmosphere can react with such metals, changing their composition in the
process in much the same way as iron is oxidised to form rust. As the electrodes
degrade, their ions may ‘bleed’ into the light-emitting layer, damaging
the LED. It is not clear whether it is water or oxygen, or perhaps a combination
of the two, that is responsible for the damage. But putting the LEDs in
a vacuum, which protects them from both water and oxygen, extends their
life by as much as a factor of 10.

In an attempt to discover what causes the damage, the Princeton group
is testing the polymer LEDs in environments with varying amounts of water
and oxygen. They light up the LED for a fixed time and then check for the
ominous black spots.

A technique from the world of conventional semiconductors, called encapsulation,
has been tried out on organic LEDs. It relies on building a thin barrier
into the device to protect it from oxidation. Several groups have tried
encapsulating an entire device by dipping it in a transparent epoxy resin.
This didn’t prevent the device from failing, although it did prolong its
lifetime. Karasz thinks that a built-in barrier of some sort may do the
trick. As an alternative to encapsulation, he suggests that the electrodes
could be allowed to react with the air if an extra layer was added to the
LED sandwich to prevent corrosion products diffusing into the light-emitting
polymer. As well as being impermeable to anything damaging, such a protective
layer would have to be able to transport electrons effectively.

Trouble spots

Another possibility is that the cause of the damage does not lie with
the electrodes but within the thin light-emitting layer itself. Some fear
that the light it creates might also be destroying it. Photons pack an
intense burst of energy, which might break chemical bonds. Heeger also suggests
that if the film is not of a uniform thickness, a thinner spot might attract
too much current and break down before the rest of the device. Likewise,
a defect or kink in the polymer chain could become a trouble spot. Consequently,
many groups are searching for techniques to make the films as pure as possible.
With semiconducting polymers, however, this can be tricky. ‘One just isn’t
sure which impurities are good or bad,’ says Karasz. ‘In some ways it’s
going to be trial and error.’

Dust particles accidentally built into the LED are also under suspicion
of being the guilty party. A speck of dust could have a radius equal to
10 times the thickness of the polymer film. Holmes wonders whether something
as simple as creating the devices in a clean room could be the solution.
Not all researchers take such precautions when fabricating their LEDs, he
says.

The polymer could also be protected by limiting the heating effect of
the electric current passing through it. A pulsed current would be one way
of doing this as it produces less heat than a steady current of the same
magnitude, yet if the pulse is fast enough the light it produces still appears
continuous. The Santa Barbara group has suggested building cooling devices
into the LED. This is another idea from the world of inorganic semiconductors,
which often incorporate tiny blocks of copper to absorb heat, so cooling
the rest of the material.

If the mechanical solutions of protecting the LED from air, strong electric
fields and heat fail to do the job, then chemists will have to try making
more durable polymers. The Cambridge group’s cyano-substituted PPV has
already shown that desirable qualities can be achieved by tinkering with
the polymer backbone. ‘There are a hundred more compounds out there to be
discovered,’ says Holmes.

A lot of money has been staked on polymer LEDs becoming a commercial
success. The University of Cambridge, along with a local venture company
called CRIL, has formed a company called Cambridge Display Technology to
exploit this SERC-funded science. Heeger has started a company called Uniax
which he hopes will capitalise on the conducting polymers’ potential. ‘This
is a big deal,’ he says. Larger companies such as AT&T Bell Laboratories,
Eastman Kodak and Allied Signal appear to agree. They too are investing
in what Heeger calls ‘a whole new class of materials’

Karen Celia Fox is a science writer based in Washington, DC.