Lithium-ion
batteries have become a leading energy source, and researchers are actively
seeking ways to nudge their performance toward ever-higher levels. Now, a new
analysis by researchers at the Massachusetts Institute of Technology (MIT) and
the University of California at Los
Angeles (UCLA) has revealed why one widely used
compound works particularly well as the material for one of these batteries’
two electrodes—an understanding they say could greatly facilitate the process
of searching for even better materials.
Lithium-ion
batteries’ energy and power density are determined mostly by the material used
for the cathode. When such batteries are being used, lithium atoms are stored
within the crystal structure of the cathode; when the battery is being
recharged, lithium ions flow back out of it. Many different materials are currently
used for these cathodes.
But
one of those materials has been a bit of a mystery. Lithium iron phosphate
(LiFePO4) performs well as a cathode, but this performance has been hard to
explain because unlike other cathode materials, lithium ion phosphate seemed to
require a two-phase process to store lithium—something that should
theoretically reduce its efficiency, but for some reason does not.
That
anomaly has now been explained. A more detailed analysis showed that, in fact,
the compound was following a single-phase process after all, but doing so in an
unusual way—which might point the way to discovery of many other such compounds
that had previously been overlooked. The new analysis was carried out by
Gerbrand Ceder, the Richard P. Simmons (1953) Professor of Materials Science
and Engineering at MIT, his graduate student Rahul Malik, and postdoc Fei Zhou
of UCLA, and published in the journal Nature Materials.
According
to accepted theory, lithium iron phosphate “should have been a low-rate”
cathode material, Ceder says—meaning that it could produce electricity only at
a very low current, suitable for use with very-low-power devices. Instead, “it
has become one of the highest-rate materials in use,” something that “always
puzzled us,” he says.
Most
cathode materials are porous, absorbing lithium ions during charging like water
going into a sponge. But it was thought that lithium iron phosphate required a
two-phase process, first forming one compound, which then morphed into a final,
stable compound. The extra step was expected to add complexity and reduce the
reaction’s speed.
But
the new experiments, which were able to probe the activity of the material as
it absorbed the lithium, found that even though the material ends up reaching
an equilibrium where it has two separate phases, in operation it actually
undergoes a single-phase process. “The way it actually absorbs lithium is not
two-phase,” Ceder says, “but it separates into two phases when it’s done.”
That
makes it much more similar to conventional single-phase cathode materials than
had been thought—and means that it makes sense to look at a wide range of other
candidate materials that had been ignored because they were also assumed to
require a two-phase process. This analysis makes it possible to “understand
better which of these two-phase materials will actually work,” Ceder says,
opening up thousands of new candidate materials to be studied. “Now we have a
way of evaluating which materials may have potential,” Ceder says. “It broadens
the possibilities.”
Previously,
he says, it had been known that “some two-phase materials do zilch, some do
very well,” but nobody knew why. Now it is likely that the ones that work well
are actually using a single-phase reaction, as turns out to be the case with
lithium iron phosphate. Ceder and his colleagues have been developing computer
algorithms that incorporate a wide variety of known properties of materials so
that large numbers of candidates can be screened quickly and efficiently to
search for very specific combinations of properties needed for a particular
application.
Understanding
the dynamics of how lithium ions get incorporated into different molecular
structures “was the missing piece in the high-throughput screening process,”
Ceder says. “Hopefully we’ll be able to do that better now.”