IEEE Spectrum November, 2007 - 33

chemical reaction. Thus a capacitor has an almost unlimited lifetime. It's also fast. Depending on its physical structure, typical
charge and discharge times are on the order of a microsecond;
sometimes they are as quick as a picosecond.
Three main factors determine how much electrical energy a
capacitor can store: the surface area of the electrodes, their distance from each other, and the dielectric constant of the material
separating them. However, you can push conventional capacitor
designs only so far. What the Standard Oil engineers did was
to develop a capacitor that functions differently. They coated
two aluminum electrodes with a 100-micrometer-thick layer of
carbon. The carbon was first chemically etched to produce many
holes that extended through the material, as in a sponge, so that
the interior surface area was about 100 000 times as large as the
outside. (This process is said to "activate" the carbon.)
They filled the interior with an electrolyte and used a porous
insulator, one similar to paper, to keep the electrodes from shorting out. When a voltage is applied, the ions are attracted to the
electrode with the opposite charge, where they cling electrostatically to the pores in the carbon. At the low voltages used in
ultracapacitors, carbon is inert and does not react chemically with
the ions attached to it. Nor do the ions become oxidized or reduced,
as they do at the higher voltages used in an electrolytic cell.
This approach allowed the engineers at Standard Oil to build
a multifarad device. At the time, even large capacitors had
nowhere near a farad of capacitance. Today, ultracapacitors can
store 5 percent as much energy as a modern lithium-ion battery.
Ultracapacitors with a capacitance of up to 5000 farads measure
about 5 centimeters by 5 cm by 15 cm, which is an amazingly
high capacitance relative to its volume. The D-cell battery is
also significantly heavier than the equivalently sized capacitor,
which weighs about 60 grams.

hundreds of thousands of ultracapacitors are manufactured
each year, for applications that require rapid recharging, high
power output, and repetitive cycling. In 2005, the ultracapacitor
market was between US $272 million and $400 million, depending
on the source, and it's growing, especially in the automotive sector.
Though ultracapacitors have generally remained a niche player, the
situation may soon change.
My laboratory at MIT-the Laboratory for Electromagnetic and
Electronic Systems-works with several automobile manufacturers
to investigate ways to improve vehicle performance. About four
years ago, I assisted on a project to evaluate commercial ultracapacitors for use in cars. While on a flight from Boston to Detroit,
I read an article describing a way to grow vertically aligned carbon
nanotubes on a flat surface. This is a truly amazing process. A sheet
of silica is covered with a nanometer-thick layer of an iron catalyst.
The sheet is placed in a vacuum, heated to 650 ºC, and exposed
to a thin hydrocarbon gas, perhaps ethanol or acetylene. The heat
causes the iron to form tiny droplets, which steal carbon molecules
from the gas. The carbon molecules then begin to self-assemble
into tubes, which grow upward from each of the droplets.
By virtue of their dimensions, it struck me that those nanotubes held the promise of even higher porosity than the activated
carbon used in commercial ultracapacitors. Together the nanotubes have an enormous surface area, and their dimensions are
more uniform than those of the activated-carbon pores, making
them more like a paintbrush than a sponge.
There are two major limitations to the conductivity of activated carbon-the high porosity means there isn't much carbon
material to carry current, and the material must be "glued" to
the aluminum current collector using a binder, which exhibits
www.spectrum.ieee.org

how to ultracaP a car
Ultracapacitors can power a number of a car's
functions locally. The orange arrows show how an
ultracapacitor discharges to power acceleration,
while the blue arrows show energy flowing
back during braking. The red squares
indicate places where ultracapacitors
can be used.

ical
Electrrator
e
n
e
g

ent
djustm ism
Seat a mechan

ry
Batte celeration
Ac braking
and

e
Engin
,

ring
r stee
Powe le
modu

l
Whee r
moto
ing
r brak
Powe module

s
, seat
locks ws
r
o
o
D
indo
and w
r
Powe itioning
cond le
modu

a somewhat high resistance. If my colleagues and I replaced the
activated carbon with billions of nanotubes, we predicted we
could make an ultracapacitor that could store at least 25 percent-
and perhaps as much as 50 percent-of the energy in a chemical
battery of equivalent weight. (To get that much improvement,
we'd have to make a number of other changes, as well, such as
increasing the number of ions in the electrolyte to reflect that
new-found storage space.)
Another advantage of nanotubes over activated carbon is that
their structure makes them less chemically reactive, so they
can operate at a higher voltage. And certain types of nanotubes,
depending on their geometry, can be excellent conductors-
which means they can supply more power than ultracapacitors
outfitted with activated carbon.
Even better, this nanotube-enhanced ultracapacitor would
retain all the advantages ordinary ultracapacitors have over
batteries: they would deliver energy in quick bursts, they
would perform well in cold weather, and they would have much
longer life spans. If this ultracapacitor could be developed, it
would be revolutionary.
It was clear from the outset that a lot of know-how would
be needed to make an ultracapacitor according to our design-
knowledge of chemical-vapor deposition, electron microscopy,
material science, quantum chemistry. And it's a challenge to
get people with all those skills together. One of the strengths
of a research university is its incredible diversity of expertise
and equipment, plus there's the willingness of faculty to collaborate. Nobody in my lab had experience fabricating carbon
nanotubes, but much of the early research in that area at MIT
was done in the building next door, at a laboratory under the
direction of Mildred Dresselhaus. Using those facilities and
aided by Dresselhaus and her lab colleagues, we succeeded in
synthesizing a nanotube forest on a small piece of silica in only
a few months.

November 2007 | IEEE Spectrum | NA 45

http://www.spectrum.ieee.org

Table of Contents for the Digital Edition of IEEE Spectrum November, 2007