Science 28Feb2020

SCIENCE sciencemag.org

GRAPHIC: N. CARY/

SCIENCE

Although several studies have found dif-

ferences in the gut microbiota of lean or

obese humans ( 7 ), the study of Kimura et

al. suggests that early exposure to micro-

bial products may be causal in later obesity

even if there are no differences in the sub-

sequent gut microbiota of the offspring.

The offspring of the high-fiber–fed ani-

mals were heavier at birth, but protected

against obesity later in life. This is consis-

tent with humans, where low birth weight

is associated with future obesity ( 8 ).

Supplementation of pregnant mice with

propionate protected offspring against fu-

ture disease. However, whether propionate

supplementation had any effects on the

mother was not assessed. Propionate sup-

plementation in humans is associated with

weight loss and improved metabolic func-

tion ( 9 ), although studies are conflicting,

with propionate also reported to increase

insulin resistance ( 10 ).

Given the goal of reducing human met-

abolic disease, it is crucial to determine

whether similar mechanisms control hu-

man development. There is an urgent need

to better understand how, why, and under

what circumstances should attempts be

made to modulate gut microbiota or SCFAs

in pregnancy. Diets high in fiber are consis-

tent with existing nutritional recommen-

dations, including during pregnancy, but

whether these would be effective in increas-

ing SCFAs and protecting offspring from

future metabolic disease remains unclear.

There is large interindividual variability in

the microbial composition of humans, with

different microbiota having differing ca-

pacities for SCFA production ( 11 ). Thus, dif-

ferences in the SCFA-producing capacity of

the microbiota may affect risk of offspring

obesity despite high consumption of fiber

during pregnancy. Although supplementa-

tion with propionate may be a convenient

option, the safety and efficacy during preg-

nancy remain to be determined. Further

studies in humans are warranted to under-

stand whether modulation of this pathway

could be an avenue to improving the meta-

bolic health of the next generation. j

REFERENCES AND NOTES

1. H. L. Blackmore, S. E. Ozanne, J. Mol. Cell. Cardiol. 83 ,
   122 (2015).


2. I. Kimura et al., Science 367 , eaaw8429 (2020).


3. A. Koh et al., Cell 165 , 1332 (2016).


4. E. E. Canfora, J. W. Jocken, E. E. Blaak, Nat. Rev.
   Endocrinol. 11 , 577 (2015).


5. D. J. Morrison, T. Preston, Gut Microbes 7 , 189 (2016).


6. B. McDonald, K. D. McCoy, Science 365 , 984 (2019).


7. V. K. Ridaura et al., Science 341 , 1241214 (2013).


8. P. H. Whincup et al., JAMA 300 , 2886 (2008).


9. E. S. Chambers et al., Gut 68 , 1430 (2019).


10. A. Tirosh et al., Sci. Transl. Med. 11 , eaav0120 (2019).


11. P. Louis, H. J. Flint, Environ. Microbiol. 19 , 29 (2017).

10.1126/science.aba7673

BATTERIES

Cobalt in lithium-ion batteries

Replacements are sought for cobalt, a costly element used

in lithium-ion battery cathodes

By Matthew Li,1,2and Jun Lu^1

T

he use of cobalt in lithium-ion bat-

teries (LIBs) traces back to the well-

known LiCoO 2 (LCO) cathode, which

offers high conductivity and stable

structural stability throughout charge

cycling. Compared to the other tran-

sition metals, cobalt is less abundant and

more expensive and also presents politi-

cal and ethical issues because of the way

it is mined in Africa ( 1 ). Cheaper cathodes

have been developed that substitute some

of the cobalt with nickel and manganese,

and LiNi0.80Co0.15Al0.05O 2 (NCA) and LiNi1-x-y

CoxMnyO 2 (NMC, where x and y <1) are

used in the majority of the LIBs in electric

vehicles. Nonetheless, in NCA and NMC,

cobalt enables high-rate performance and

to some extent, enhances cycle stabil-

ity. We outline research efforts that could

further decrease or even eliminate cobalt

content in LIBs to lower their cost while

maintaining high performance.

Efforts to replace cobalt have to start

with an understanding of what makes co-

balt so crucial within the NMC and NCA

compositions. Originally, cobalt and man-

ganese were introduced into LiNiO 2 (LNO)

to stabilize the material itself. Although

LNO has a high theoretical energy density,

it also has very poor cycling stability and

presents potential safety hazards because of

lattice instability. For these reasons, cobalt

was added as a stabilizer. In comparison to

LCO, it is difficult to synthesize pure lay-

ered LNO, which facilitates Li+ ion trans-

port, and more often the undesired rock salt

structure forms.

Also, nickel is inherently unstable by itself

in the transition-metal layer of the oxide as

it has a relatively strong magnetic moment.

Three triangularly placed Ni2+ cations will

always have two opposing magnetic mo-

ments, creating “magnetic frustration” ( 2 ).

Because Li+ ions do not have a magnetic

moment, they preferentially exchange with

some of the nickel ions. The loss of a spin at

one site relieves magnetic frustration (see

the figure). The strong interlayer antifer-

romagnetic coupling between nickel in the

transition-metal layer and the migrated

nickel in the lithium layer creates a super-

exchange interaction that further stabilizes

the Li+ ion ( 2 ). Overall, this lithium-nickel

mixing deteriorates performance because

the lithium-deficient LiO 2 interslab layer

decreases in thickness. This thinner layer

severely hinders transport of lithium ions

and ultimately results in very rapid degra-

dation of the LNO composition ( 3 ).

(^1) Chemical Sciences and Engineering Division, Argonne
National Laboratory, Lemont, IL 60439, USA.^2 Department of
Chemical Engineering, Waterloo Institute of Nanotechnology,
University of Waterloo, Waterloo, ON N2L 3G1, Canada.
Email: [email protected]
Magnetic
frustration
Any arrangement
of Ni3+ creates
aligned spins
that lead to an
unstable, higher-
energy state.
Lithium mixing
A Li+ ion has
no spin. When it
replaces Ni3+,
magnetic
frustration is
alleviated,
but the process
disorders the
cathode lattice.
Cobalt addition
Because Co3+ is
also nonmagnetic,
its addition
to the transition-
metal layer
relieves magnetic
frustration and
creates a stable
cathode.
Magnetic frustration relieved
Li+
Li+
Co3+
Co3+
Ni3+
Ni3+
28 FEBRUARY 2020 • VOL 367 ISSUE 6481 979
Instability of nickel
Nickel (Ni) as a replacement for cobalt (Co) in lithium
(Li) ion battery cathodes suffers from magnetic
frustration. Discharging mixes Li ions into the Ni layer,
versus just storing them between the oxide layers.
Published by AAAS