午夜福利1000集合

ELIAS GREENBAUM is waiting for his man. In less than ten minutes a new
consignment is due to arrive and, thanks to a deal with the driver, Greenbaum
will get first choice of the goods. If it鈥檚 high quality鈥攁nd it usually
is鈥攈e鈥檒l buy half a kilo and head back to his office. Greenbaum鈥檚 buying
spinach. And with it, he wants to build a computer.

The thought of a computer made from spinach probably conjures up an image of
a green box that runs on water and sunlight, and wilts while you鈥檙e away on
holiday. But this isn鈥檛 quite what Greenbaum has in mind.

He believes that tiny structures tucked deep inside spinach leaves could be
used to make complex circuitry. If he鈥檚 right, it could revolutionise the way
computers are built, making them smaller, cheaper and, yes, perhaps even a
little greener.

This bizarre idea could be the answer to a computer engineer鈥檚 prayer. To
keep up with Moore鈥檚 Law鈥攚hich states that the density of transistors on a
processor chip will double every 18 months鈥攅ngineers must continue to
carve out ever tinier transistors (New Scientist, 7 November 1998, p
42). But their techniques have almost reached their limit. To go shrink circuits
further they鈥檒l need a radical approach, and Greenbaum may have found it.

Based at Oak Ridge National Laboratory in Tennessee, Greenbaum and his
colleagues spend their time unravelling the finer details of photosynthesis, the
light-fuelled reaction by which plants convert carbon dioxide and water into
oxygen and energy-rich carbohydrates.

When light hits a leaf, its energy is absorbed by arrays of chlorophyll
molecules that are arranged into antennae. These collect and funnel the energy
into structures called reaction centres. It is these tiny protein
clusters鈥攅ach about a million times smaller than a grain of
sand鈥攖hat do the work of photosynthesis. 鈥淭hese are the engines of
creation,鈥� says Greenbaum, 鈥渢he powerhouses that take the Sun鈥檚 light and turn
it into a form of energy.鈥�

Plants have two types of reaction centre, one that converts carbon dioxide
into sugar and another that splits water to produce oxygen. Greenbaum鈥檚 efforts
focus on the first of these, known as photosystem I.

The reaction centres are essentially pumps powered by sunlight. When energy
from a photon is fed into photosystem I, an electron in a pair of specially
aligned chlorophyll molecules known as P700 is excited to a higher energy state
and kicked out on to a neighbouring chlorophyll molecule. From here, the
electron makes a series of jumps, first to a molecule called phylloquinone, then
to a series of iron-sulphur clusters and finally out of the reaction centre
altogether (see Diagram).
The whole process takes just a few billionths of a second.

This intricate mechanism for shifting electrons got Greenbaum thinking. The
important thing about photosystem I is that it is structured in a way that
prevents an electron jumping backwards and recombining with P700. In other
words, it allows the electrons to move in one direction only. And that reminded
Greenbaum of something else that only allows current to flow one way: a
semiconducting diode.

Semiconducting diodes are commonly used to convert alternating currents into
direct currents, a process called rectification. But connected in the right way,
they can also be used to build logic gates, the digital switches that carry out
electronic calculations. String enough logic gates together and you鈥檝e got a
computer.

What Greenbaum realised is that reaction centres are essentially
ready-made diodes, but being just 7 nanometres long, they鈥檙e over 20 times
smaller than existing silicon diodes. And since every leaf contains hundreds of
millions of reaction centres, one spinach plant could give you a lot of raw
computing power.

But it isn鈥檛 as easy as that. Reaction centres are very delicate鈥攅ach a
precisely arranged cluster of proteins. Disrupt the position or shape of the
proteins and the reaction centre stops working.

To make matters worse, they鈥檙e wedged into the internal membranes of
chloroplasts, the organelles that house the entire photosynthetic machinery.
Digging them out in one piece is no mean feat. This is one of the reasons
Greenbaum chose spinach rather than any other plant. 鈥淪pinach is readily
available and has soft, tender leaves鈥攖hat makes it much easier to extract
the reaction centres.鈥�

Thanks to a bit of local knowledge, Greenbaum has a good supply of the green
stuff. 鈥淲e have an arrangement with the guy at the supermarket. We know what
time the spinach gets delivered, so we just go down there and get it right off
the truck. We get the freshest and best spinach that comes in.鈥� Back at the lab,
the spinach is blended with plenty of ice to make a kind of spinach juice. It
must be kept cold because enzymes released when the spinach is torn up can
destroy the reaction centres. By chilling it, you slow that reaction down.

Molecular synergy

After making spinach juice, the next step is to filter and centrifuge it to
separate the chloroplasts from the spinach debris. These chloroplasts are
then mixed with a detergent solution. 鈥淚t鈥檚 like molecular surgery,鈥� says
Greenbaum. 鈥淭he detergent molecules chisel the reaction centres out of the
membranes that hold them in place.鈥� Once the soap has done its job, the reaction
centres can be sifted out by chromatography.

Having worked out how to extract the reaction centres, Greenbaum set about
testing whether they could be used as diodes. Since they are too small to
handle, Greenbaum deposited layers of the reaction centres on gold-coated wafers
by immersing the wafers in a suspension of reaction centres. The wafers, with
the reaction centres clinging to them, could then be removed and looked at under
a scanning tunnelling microscope.

By applying a voltage between the microscope鈥檚 fine tip and the gold wafer,
Greenbaum could measure the electrical properties of individual reaction
centres. He found that current flowed through them only when they were pointing
upwards. 鈥淓ach one of the reaction centres was a molecular diode,鈥� he says.

But a layer of jumbled up bio-diodes is worthless. To make them useful, they
must be aligned and connected to form circuits. In the absence of nano-tweezers,
Greenbaum tried modifying the gold surfaces. He found that coating the gold with
a chemical called mercaptoacetic acid gave the surface a negative charge. When
reaction centres were deposited on this, they arranged themselves horizontally.
Another chemical, 2-mercaptoethanol, made the reaction centres stand up
vertically. Greenbaum believes these are the first tentative steps towards
assembling reaction centres into working circuits.

The challenge is now to make a complete circuit. If dots of mercaptoacetic
acid can be used to anchor reaction centres in specific patterns, then one
hurdle will be overcome. But wiring them all together will be even harder.
鈥淣obody knows how to do it right now,鈥� Greenbaum concedes, 鈥渂ut our goal is to
use carbon nanotubes and organic molecular wires which would transport
肠丑补谤驳别.鈥�

Reaction centres have slight positive charges at each end, so Greenbaum
suggests that giving nanotubes a slight negative charge might be a way to attach
the wires to the biodiodes. 鈥淭his is where the ingenuity of physicists and
chemists will come together to move this forward.鈥�

Potential spinach computer-makers face more prosaic challenges, too. If
they鈥檙e exposed to air, the proteins that make up the reaction centres oxidise,
which would stop the diodes working. So spinach-based circuits would have to be
sealed in oxygen-free boxes.

However, using these tiny spinach structures to make complex circuitry might
be only half the story. It turns out that diodes aren鈥檛 the only components
reaction centres can mimic. In a paper soon to be published in The Journal
of Physical Chemistry B, Greenbaum shows for the first time that each
isolated reaction centre can generate a potential difference of 1 volt across
its length when illuminated by red laser light.

鈥淭hey鈥檙e actually identical to solar cells,鈥� explains Greenbaum. 鈥淲hen you
illuminate them, you get charge separation鈥攖he electron leaves one side,
so it could go around a circuit and come back.鈥� So not only could reaction
centres be used to make computer circuits, they could also be used to provide
computer power supplies that run on sunshine.

On paper, photosystem I could be just as efficient as a silicon solar cell,
but until Greenbaum has developed and tested a working device, it will be
impossible to know for sure.

Just how bright the future of spinach computers is will depend on how tough
circuit-building proves to be. If the problems can be cracked, spinach might
just become a major player in the next generation of computers. It could mean an
unlikely revolution for the computer industry: instead of pouring cash into the
hottest new chip-making facilities, the giants of silicon valley could one day
be investing in spinach farms.