IEEE Spectrum February, 2010 - 43

the infrared, because outside of certain
wavelength bands in this range, glass
fibers tend to absorb or distort the light
sent through them. Infrared is also useful for conducting secure line-of-sight
optical communications.
My colleagues and I at the University
of Toronto have made great progress in
recent years building devices using what
are essentially paints that respond well
to infrared wavelengths. This work is
still in the research stage-products
remain between one and five years away,
depending on the application-but the
pace of advance has been so swift that
it's not too soon to look forward to the
many exciting possibilities.

early sales of photovoltaic panels now amount to tens of billions
of dollars, and the overall energy
market is measured in the trillions. The
ideal that solar-cell developers are seeking is a device that is both efficient and
inexpensive. Solar cells constructed
from costly semiconductor wafers have
yielded the greatest efficiencies-upward
of 40 percent-but because they are so difficult to manufacture, such high-efficiency
cells are too pricey for all but the most
demanding applications, such as for the
solar panels attached to spacecraft.
Photovoltaic cells made out of organic
polymers cost far less, but the best efficiencies they've shown have typically
been around 5 or 6 percent. That's stunningly good for something that can be
manufactured so cheaply, but it's still
less than the 10 percent figure experts
say will be needed for this technology to
take off commercially.
One common strategy to boost the
efficiency of solar cells of any kind might
be called the layer-cake method. The top
layer of the cell absorbs photons of relatively short wavelengths, and thus
of high energy, turning them into electricity. These wavelengths include visible light and some of the ultraviolet as
well. Photons of lower wavelengths pass
through this layer into a second one
below, which is designed to absorb them
and transform their energy into electric
power. Some of these layer-cake designs
include a third stratum at the bottom to
capture the even lower-energy photons
that penetrate the top two layers.
Companies making paint-on or
print-on solar cells have been unable
to take advantage of this strategy, however. The reason is that for years the only
paintable photovoltaic materials have
50

Na * ieee spectrum * february 2010

quantum dots: this electron-microscope image shows close up the nanometer-scale
quantum dots used by the author and his research team to fashion infrared-sensitive
optoelectronic devices. Image: eDWarD h. sargent

been based on organic molecules that
are sensitive to visible wavelengths or to
infrared wavelengths that are very close
to the red end of the visible spectrum. So
manufacturers had nothing that could be
used for the lower layers.
Fortunately, the researchers on my
team have lately made good progress
in devising paint- or print-on solar
cells sensitive to infrared wavelengths
that are well separated from the visible
spectrum, which is to say wavelengths
of 1 micrometer or longer. Five years
ago, when we first proved the concept,
the efficiencies at these wavelengths
for our pioneering devices were less
than 1 percent. But in 2008 we showed
how to boost efficiencies to just under
4 percent. That's only about a third of
the efficiency figure you'd need for commercialization, but it represents a huge
step forward, and we expect further
progress as we continue to devise new
and better designs.
Our infrared solar cells contain something called quantum dots-tiny bits of
semiconducting materials that absorb or
emit light. For our work, we used particles of a lead-sulfur compound. We and
others are also experimenting with compounds of bismuth, tin, and indium with
sulfur, selenium, and oxygen. Whereas

typical optoelectronic devices operate at
fixed wavelengths defined by the nature
of their constituent chemistry, quantum
dots can be tuned to absorb or emit light
of different wavelengths simply by varying their size.
Quantum dots work this way because
the electrons moving within them "feel"
the nearby boundaries of the semiconductor. That's because the quantummechanical waves associated with
these electrons are constrained by the
margins of the dots. As is the case with
sound vibrations reverberating in a box
or microwaves reflecting back and forth
in a cavity with conductive walls, the
size of the container determines which
wavelengths can exist within. For semiconductor quantum dots, increasing
the diameter of the particles from, say,
1 to 10 nanometers shifts the action from
the visible portion of the spectrum well
into the infrared.
Early research in this area mostly
involved embedding quantum dots relatively sparsely in a semiconducting polymer. Investigators believed that if they
didn't keep the dots spaced well apart
this way, they wouldn't remain tuned
to the desired wavelength. But in 2007,
my colleagues and I omitted the polymer entirely and merely glued the dots
spectrum.ieee.org

http://spectrum.ieee.org

Table of Contents for the Digital Edition of IEEE Spectrum February, 2010