INSIGHTS | PERSPECTIVES
sciencemag.org SCIENCEWhen cobalt is introduced into LNO,
both during synthesis and delithiation,
the intermixing of lithium for nickel is
deterred statistically because of the lower
nickel content. More importantly, it allevi-
ates magnetic frustration ( 4 ) because the
Co3+ cation does not have a magnetic mo-
ment and serves as a buffer atom in the
transition-metal layer (see the figure).
Because added cobalt in LNO easily pre-
vents nickel-lithium mixing and subse-
quent phase transitions, the desired lay-
ered structure forms ( 5 ).
Directly decreasing cobalt content can
be effective in achieving acceptable perfor-
mance but only to some minimum cobalt
fraction. For example, in nickel-rich NMC
compositions, thermal stability, which
is crucial for avoiding catastrophic fail-
ures, as well as cycle stability drastically
dropped in comparison to equal nickel-co-
balt fractions. The NMC systems with pro-
gressively higher nickel content from NMC
111, 532, 622, and 811 (where 111 represent
1 part nickel, 1 part manganese, and 1 part
cobalt mass composition, respectively) fol-
low a steady trend in decreasing cycle sta-
bility and safety ( 6 ).
By contrast, partial substitution of cobalt
with other elements such as titanium has
been shown to produce reasonable per-
formance ( 7 ). Although other metals can
limit lithium-nickel mixing, typically poor
kinetics and lower capacities result. Binary
composition with ultralow O 2 gas, such as
LiNi0.94Co0.06O 2 , has exhibited severe surface
reconstruction when compared to NMC 811
( 8 ) that pulverizes the particles after pro-
longed cycling.
Other systems, such as the lithium- and
manganese-rich materials, consist of a mix
of 0.5 Li 2 MnO 3 and 0.5 NMC. This layered-
layered structure offers increased capac-
ity at the cost of severe phase transitions.
The defects introduced limit the capacity
of cathode and the voltage it can produce,
an effect called voltage fade. Moving toward
completely cobalt-free systems has led re-
searchers to pursue disordered rock-salt
materials in hopes of harnessing their in-
creased capacity through the use of anionic
redox couples, such as O2− ( 9 ). However, an-
ionic redox systems have limited cyclability
because of the formation of O 2 gas. Recent
work has yielded some pure anionic redox
systems where the oxidation state of oxygen
is maintained below the superoxide charge
of −0.5 ( 10 ).
In the original LiNiO 2 system, the role
of cobalt may not have been as critical to
performance as initially presumed. Often,
its importance is only apparent when com-
pounded by the effects of another element.
Some combinations of noncobalt transition
metals, such as aluminum, manganese, and
magnesium, can outperform the cobalt-
containing equivalent, although the levels
of performance are still lower than the com-
mercially viable ratios ( 11 ). Even cobalt-free
LiNiO 2 showed surprisingly good cycling
stability when made under synthesis condi-
tions that carefully controlled temperature,
sintering time, and O 2 gas ( 12 ). A mostly
layered Li0.98Ni1.02O 2 structure formed with-
out the need for other transition-metal ad-
ditives. This surprising result reopens the
question of the optimal cathode composi-
tion and its implications for the effect of
magnetic frustration of this system. Simply
tuning the composition of cobalt-free sys-
tems will likely require the substitution of
cobalt with another third transition metal.
A brief performance comparison between
the various commonly studied cathode
materials is given in table S1 in the supple-
mentary materials.
Identifying optimal composition and
synthesis conditions of new cathodes will
likely require rigorous and extensive facto-
rial experimental design that must incorpo-
rate compositions with elemental fractions
decreased to doping levels (<1%). Machine
learning techniques might decrease the
search path for optimal elements and com-
positions. The complete elimination of
cobalt is an important research goal, but
lower-cost cathodes with less cobalt must
maintain performance. This trade-off will
depend on the future supply of cobalt from
both mining and recycling. jREFERENCES AND NOTES- B. K. Sovacool, Extr. Ind. Soc. 6 , 915 (2019).
- J. Zheng et al., J. Phys. Chem. Lett. 8 , 5537 (2017).
- W. Li, J. N. Reimers, J. R. Dahn, Phys. Rev. B Condens.
Matter 46 , 3236 (1992). - Y. Xiao et al., Nano Energy 49 , 77 (2018).
- A. Ueda et al., J. Electrochem. Soc. 141 , 2010 (1994).
- T. Li et al., Electrochem. Energy Rev. 2019 , 1 (2019).
- S. Wolff-Goodrich et al., Phys. Chem. Chem. Phys. 17 ,
21778 (2015). - J. Li, A. Manthiram, Adv. Energy Mater. 9 , 1902731
(2019). - E. Hu et al., Nat. Energy 3 , 690 (2018).
- Z. Zhu et al., Nat. Energy 1 , 16111 (2016).
- H. Li et al., J. Electrochem. Soc. 166 , A429 (2019).
- H. Li, N. Zhang, J. Li, J. R. Dahn, J. Electrochem. Soc. 165 ,
A2985 (2018).
ACKNOWLEDGMENTS
We thank T. Liu for providing consultation on this topic. The
work at Argonne National Laboratory was supported by the
U.S. Department of Energy (DOE), Office of Energy Efficiency
and Renewable Energy, Vehicle Technologies Office. Argonne
National Laboratory is operated for the DOE Office of Science
by UChicago Argonne, LLC, under contract no. DE-AC02-
06CH11357. M.L. acknowledges financial support from the
National Sciences and Engineering Research Council of
Canada, University of Waterloo, and Waterloo Institute for
Nanotechnology.SUPPLEMENTARY MATERIALS
http://www.sciencemag.org/content/367/6481/979/suppl/DC110.1126/science.aba9168PLANETARY SCIENCEA deep dive
into the abyss
The flyby of the Kuiper
Belt object Arrokoth
provides quick and
tantalizing observations
By David C. JewittS
ince 1992, astronomers have un-
veiled a vast population of solid bod-
ies in orbit beyond Neptune. Known
as the Kuiper Belt, this region of the
Solar System is a dynamical fossil
preserving a record of the planet-
formation epoch ( 1 ). The belt is also a re-
pository of the Solar System’s most primor-
dial material and the long-sought nursery
from which most short-period comets orig-
inate. Most of what we know about the
belt was determined using ground-based
telescopes, and studies were limited to ob-
jects larger than about 100 km because the
smaller ones are too faint to easily detect.
Now, 5 years after its flyby of the 2000-km-
diameter Kuiper Belt object Pluto ( 2 ),
NASA’s New Horizons spacecraft has pro-
vided the first close-up look at a small, cold
classical Kuiper Belt object. On pages 998,
999, and 1000 of this issue, Spencer et al.
( 3 ), Grundy et al. ( 4 ), and McKinnon et al.
( 5 ) show one of these objects to be lightly
cratered, ultrared, and binary, respectively.
The scientific impact of the Kuiper Belt
has been huge, in many ways reshaping our
ideas about the formation and evolution of
the Solar System. For example, and quite in-
credibly, Kuiper Belt objects are a thousand
times more plentiful than the (much closer
and more familiar) main-belt asteroids that
orbit between Mars and Jupiter. The Kuiper
Belt objects’ orbital distribution shows that
the planets formed closer to the Sun and
then migrated outward, a finding with pro-
found dynamical consequences. The objects
themselves are divided into dynamically
distinct groups, one of which (the so-called
cold classicals) appears to be the most pri-
mordial population in the Solar System.
The object that New Horizons flew by is
known formally as (486958) Arrokoth (pro-
visional designation 2014 MU 69 ). An earlierDepartment of Earth, Planetary, and Space
Sciences, University of California, Los Angeles, CA.
Email: [email protected]980 28 FEBRUARY 2020 • VOL 367 ISSUE 6481
Published by AAAS