WE ALL love stories of serendipity. They seem to hark back to a time when a
fogged photographic plate or a filthy Petri dish could change the world. Even
today, when financial constraints keep the role of chance to a minimum, science
is still sometimes a spontaneous act, a freelance exploration of the unknown. It
often starts in front of a blackboard when one scientist says, 鈥淚 wonder what
would happen if . . .鈥�, and the other one replies, 鈥淟et鈥檚 give it a try.鈥�
The result of one such conversation two years ago in Eric Mazur鈥檚 laboratory
at Harvard University is a new form of silicon, as black as soot. What started
life as an experimental curiosity now has patents pending and papers in
preparation, because silicon鈥檚 dark side offers a surprising array of potential
applications. It turns visible light into electricity with unprecedented
efficiency, so black silicon could soon be at the heart of your camcorder. It
might also prove to be the innovation that makes it economically viable to put a
solar panel on the roof of every house. And its extraordinary structure means
that it could be made into tiny needles able to deliver a carefully
controlled鈥攁nd painless鈥攄ose of whatever medication you need.
The fortuitous creation of black silicon began when Mazur and a graduate
student, Tsing-Hua Her, were talking about etching semiconductors with a
powerful laser. In the early 1990s, Mazur鈥檚 was the first academic lab in the
world to get its hands on a femtosecond laser. This device produces pulses of
light that are hundreds of billions of times brighter than the Sun. Its immense
power is delivered extremely quickly: each pulse lasts a mere fraction of a
trillionth of a second.
Advertisement
These flashes of laser light have provided researchers with a new way to
probe the characteristics of many materials (快猫短视频, 19 February
2000, p 34). Mazur鈥檚 group was using the powerful femtosecond pulses to study
the surface chemistry of metals. But Her, who is now at the Lawrence Livermore
Laboratory in California, had been wondering for years what the laser would do
to semiconductors like silicon. No one had ever tried it, so there was only one
way to find out.
鈥淲e tried a quick and dirty experiment,鈥� Mazur recalls. Along with graduate
student Claudia Wu, they picked up a silvery-grey wafer of silicon that was
lying around the lab, and mounted it in one of their experimental chambers. 鈥淣ot
our $5 million, ultra-high-vacuum chamber,鈥� Mazur said. 鈥淲e didn鈥檛 want
to mess that up.鈥� Then they let in a little sulphur hexafluoride (SF6)
gas. 鈥淭he only reason we put it in there is because we had it lying around the
laboratory,鈥� he claims.
Well, it was almost the only reason. A short laser pulse will break down
SF6 into sulphur and fluorine radicals, which will attack a silicon
substrate. 鈥淗ydrogen fluoride is used to etch silicon. So we thought maybe the
SF6 would decompose and then the fluorine would somehow react with the
silicon,鈥� Mazur explains.
With no clearer idea than this, the researchers began firing 100-femtosecond
pulses of laser light through the window of their chamber, through the SF6
gas and onto the shiny silicon wafer. After firing about 100 pulses they
cracked the seal on the chamber and removed the wafer. They saw a tiny black
spot at the focal point of the laser beam. A burn, perhaps. Except that Mazur
knew that silicon doesn鈥檛 burn. 鈥淵ou can get silicon oxide, but it鈥檚 not black,鈥�
he says. So what was going on?
Intrigued, they put the chip under an optical microscope. Even under the
highest magnification the surface was a flat, dull black, so Her took the sample
and put it into a scanning electron microscope. Looking straight down at the
wafer didn鈥檛 reveal anything, but then he tilted the wafer to obtain a more
oblique view. 鈥淲ow,鈥� he said. 鈥淭hose things are really sharp.鈥� What he saw was a
forest of millions of minute, needle-like spikes, arranged in a regular pattern,
each topped with a tiny ball about a micrometre across.
After hitting the silicon with so much laser energy鈥攕ix times the
energy that should vaporise the surface鈥攊t was hard to believe there was
any structure left, let alone one so delicate. 鈥淵ou鈥檇 imagine everything would
get destroyed because you blast it so hard,鈥� Her said.
The researchers began to play with the process to work out what was going on.
They tried blasting the silicon with increasing numbers of laser pulses. After a
few pulses the surface of the silicon became bumpy, and as the number of blasts
increased, the bumps became spikes. By increasing the number of laser blasts
they could make the spikes, which had a diameter of about 6 microns at the base,
grow to about 40 microns long. They used techniques known as electron
backscattering and ion channelling to probe the structure of the spikes. These
experiments fire charged particles into the spikes and record the way atoms
deflect these particles. The patterns of deflection and reflection revealed a
startling result. The spikes had the same crystal structure as the original
wafer, so they must have been the result of intricate etching.
Spikes also formed when the silicon was blasted in an atmosphere of
chlorine鈥� another halogen gas鈥攂ut did not form when nitrogen or
helium was used. So it is likely that a chemical process is responsible for
spike formation. Mazur and his group have found that the spike height and
separation is determined by factors like laser energy, pulse duration and the
pressure of SF6 gas (Applied Physics A, vol 70, p 383), but
they still do not understand exactly what makes the spikes form.
The researchers are still looking into the fundamental physics and chemistry
behind the formation of black silicon, but they are not letting their incomplete
understanding of the process hold them back from putting the material to work.
Most of the tantalising applications for black silicon centre on its
extraordinary absorption of radiation. Silicon, the backbone of the
microelectronics industry, is also the basic material from which things like
photovoltaic cells and optical detectors are built. Natural silicon absorbs only
about 60 per cent of the visible light that falls on it. A photovoltaic cell
made from silicon could, at best, convert only a little more than half the
sunlight hitting it into electricity. Black silicon, on the other hand, absorbs
almost all of the visible light that strikes it. Absorption increases with the
height of the spikes: at around 20 micrometres high the silicon spikes absorbed
97 per cent of the visible light. So a photovoltaic cell made with black silicon
could be efficient enough to make solar energy economically
attractive鈥攎aybe even competitive with oil or natural gas. The same
property gives black silicon excellent potential as a component for producing
electrical signals from light in devices like video cameras. Its strong light
absorption means that it will help create a sharp image even in relatively dark
conditions. And because you control which part of the silicon is black鈥攖he
region you choose to irradiate鈥攜ou can integrate electronics with areas of
high sensitivity to light. 鈥淚t鈥檚 like taking a chip with electronic circuitry
and giving it much better eyes,鈥� Mazur says.
The increased absorption may have something to do with the spikes. If
radiation is shone on the surface, it bounces back and forth between the spikes
within a small area, giving it more chance of being absorbed. Acoustic engineers
use this principle when they use spiky, foam sound baffles to cut down on echoes
in concert halls and recording studios. The eyes of certain moths also contain
microscopic spikes that cut down on their reflectivity.
But the spiky texture cannot explain all the properties of black silicon,
particularly its astonishing performance in the infrared region of the spectrum.
Normal grey silicon is practically transparent to infrared light. This is an
unfortunate problem because telecommunications and remote sensing rely heavily
on infrared signals. But Mazur and his group have discovered that black silicon
absorbs infrared light almost perfectly: again, around 97 per cent. And early
indications are that this absorption even extends into the microwave region of
the spectrum. Since radiation of this frequency can carry enormous amounts of
data, the microwave band is likely to be used for future high-speed
communications devices. So black silicon might find itself at the heart of the
next-generation Internet, mobile phone and video-on-demand technologies.
By itself, this absorption of radiation is not necessarily useful, of course.
鈥淎bsorption is fine, but a piece of black plastic or a piece of wood also
absorbs light,鈥� says Mazur. To give absorption some technological value it needs
to be translated into an electrical signal. Black plastic or wood do not have a
mechanism for this, but semiconductors like silicon do.
Absorption and electrical conduction in semiconductors depend on the
characteristic energies of the electrons in the material. The electrons are
limited to specific ranges of energies, known as bands. The lowest band, where
they normally sit, is called the valence band. The electrons must receive a
certain minimum energy to promote them from their usual existence in this
region, where they are tied to individual atoms, to the conduction band where
electrons flow freely. The existence of this minimum energy鈥攖he band
gap鈥攊s what makes semiconductors such poor conductors of electricity. But
it also explains why silicon does not normally absorb infrared light: the energy
of an infrared photon is much smaller than the band gap of silicon. Absorbing an
infrared photon does not give a valence electron enough energy to become a
conduction electron. An in-between state is simply not allowed, so the photon is
not absorbed.
Bridging the gap
But treatment with the laser alters the basic band structure of the silicon.
The researchers have determined that sulphur atoms get blasted into the silicon
spikes, and this disrupts the spikes鈥� orderly crystal structure. In a paper just
submitted to Applied Physics Letters they explain how sulphur and
fluorine embedded in the silicon spikes might introduce extra electron states
into the band gap, making it possible for valence electrons to absorb infrared
energy and reach the conduction band via a series of steps. 鈥淪ulphur introduces
levels into the middle of the band gap of silicon: people have measured the
presence of those levels in cases of light doping before,鈥� says Catherine
Crouch, one of the researchers on Mazur鈥檚 team. 鈥淥ur speculation is that we have
introduced a sufficiently high density of impurities that the individual levels
broaden into a band.鈥� That makes it easy for electrons to move between the
valence and conduction bands, allowing very strong absorption of light at almost
all wavelengths.
To check black silicon鈥檚 ability to create conducting electrons from infrared
radiation鈥攑ossibly its most useful immediate application because of the
importance of infrared telecommunications鈥擬azur called a company called
Radiation Monitoring Devices in nearby Watertown, Massachusetts. The company
lent Mazur a photodetector and he placed it in a chamber with SF6 gas
and bombarded its silicon detecting surface with a few hundred laser blasts. He
returned the device to the company, which found that it now worked as well in
the infrared as it did in the visible region of the spectrum鈥攊n both cases
much better than the untreated silicon.
Mazur is extremely excited by his discoveries. 鈥淭here is no
question鈥攖echnologically鈥攖hat this is a great thing,鈥� he says.
However, he and his team aren鈥檛 yet sure that it will be economically viable.
They need to learn a lot more about how the new material works before they know
the full value of the discovery. One thing is for sure, though: black silicon
has transformed Mazur鈥檚 laboratory, becoming the single largest area of research
in the last two years.
And there鈥檚 more to black silicon than improvements in microelectronics. The
forest of silicon whiskers that fur the surface of black silicon may have
medical applications. The fine needles could pierce tough human skin to deliver
continuous small doses of drugs such as nitroglycerine. The human body responds
better to continuous microdosing than it does to a single, large injection. It
also makes the injection painless, as the needles penetrate only the outermost
layer of skin, where there are no nerve endings. Mazur鈥檚 group is collaborating
on this project with scientists at Massachusetts General Hospital in Boston and
a company in Germany.
One of Mazur鈥檚 early ideas was that the sharp spikes could function not only
as absorbers of light but as emitters: ultra-fine electrodes in a plasma display
for use in computers, televisions and mobile phones. 鈥淚 speculated about this
and wrote it up,鈥� Mazur says. And when the team tried it a few months ago, it
worked straight away. Black silicon seems almost like a magical substance, Mazur
feels. 鈥淭his is what has been so exciting about this material. Everything we
speculate about it turns out to be true,鈥� he says. 鈥淭hat has never happened in
my research career. Never on this scale. It鈥檚 stunning.鈥�
