Conflicting Roles of Nickel in Controlling Cathode Performance in
Lithium Ion Batteries
Meng Gu,
†
Ilias Belharouak,
‡
Arda Genc,
§
Zhiguo Wang,
⊥
Dapeng Wang,
‡,∥
Khalil Amine,
‡
Fei Gao,
⊥
Guangwen Zhou,
∥
Suntharampillai Thevuthasan,
†
Donald R. Baer,
†
Ji-Guang Zhang,
#
Nigel D. Browning,
⊥
Jun Liu,
⊥
and Chongmin Wang*
,†
†
Environmental Molecular Science Laboratory,
⊥
Fundamental and Computational Science Directorate, and
#
Energy and
Environmental Directorate, Pacific Northwest National Laboratory, 902 Battelle Boulevard, Richland, Washington 99352, United
States
‡
Chemical Sciences and Engineering Division, Argonne National Laboratory, 9700 South Cass Avenue, Argonne, Illinois 60439,
United States
§
FEI Company, 5350 NE Dawson Creek Drive, Hillsboro, Oregon 97124, United States
∥
Department of Mechanical Engineering, Binghamton University, State University of New York, Binghamton, New York 13902,
United States
*
S
Supporting Information
ABSTRACT: A variety of approaches are being made to
enhance the performance of lithium ion batteries. Incorporat-
ing multivalence transition-metal io ns into metal oxi de
cathodes has been identified as an essential approach to
achieve the necessary high voltage and high capacity. However,
the fundamental mechanism that limits their power rate and
cycling stability remains unclear. The power rate strongly
depends on the lithium ion drift speed in the cathode.
Crystallographically, these transition-metal-based cathodes
frequently have a layered structure. In the classic wisdom, it
is accepted that lithium ion travels swiftly within the layers
moving out/in of the cathode during the charge/discharge.
Here, we report the unexpected discovery of a thermodynamically driven, yet kinetically controlled, surface modification in the
widely explored lithium nickel manganese oxide cathode material, which may inhibit the battery charge/discharge rate. We found
that during cathode synthesis and processing before electrochemical cycling in the cell nickel can preferentially move along the
fast diffusion channels and selectively segregate at the surface facets terminated with a mix of anions and cations. This segregation
essentially can lead to a higher lithium diffusion barrier near the surface region of the particle. Therefore, it appears that the
transition-metal dopant may help to provide high capacity and/or high voltage but can be located in a “wrong” location that may
slow down lithium diffusion, limiting battery performance. In this circumstance, limitations in the properties of lithium ion
batteries using these cathode materials can be determined more by the materials synthesis issues than by the operation within the
battery itself.
KEYWORDS: Lithium ion battery, Li
1.2
Ni
0.2
Mn
0.6
O
2
, nickel segregation, STEM, DFT calculation, lithium diffusion barrier
L
ithium ion batteries have been widely used in consumer
electronics and have entered the electrical vehicle market
due to their high energy density.
1−3
However, their power rate
and cycle life still need to be improved for long-term
applications. The power rate of the lithium (Li) ion batteries
strongly depends on the rate of Li
+
diffusion within the cathode
structure.
1,4−6
Significant efforts have been made to improve
the power rate of Li ion batteries by doping cathodes to
increase electronic conductivity,
6
reducing the Li ion diffusion
distances by using nanoscale particles,
6,7
surface coating,
1
and
enhancing surface facets with Li
+
fast-diffusion channels.
8
Lithium transition-metal oxides have been widely used as the
cathode for Li ion batteries. They can be tailored to gain either
high voltage or high capacity by adjusting the relative ratio of
different transition-metal ions and preparation conditions.
9−20
For example, a layered composite based on lithium nickel
manganese oxide Li
1.2
Ni
0.2
Mn
0.6
O
2
(LNMO) has demonstrated
a rechargeable capacity of >250 mAh/g, which is much larger
than that of the conventional LiCoO
2
cathode (<140 mAh/
g).
10,14,16,21,22
This category of material is featured by a layered
composite structure in which the channels within the structure
can act as a low-barrier path for Li ions to move during the
Received: June 15, 2012
Revised: August 22, 2012
Published: September 17, 2012
Letter
pubs.acs.org/NanoLett
© 2012 American Chemical Society 5186 dx.doi.org/10.1021/nl302249v | Nano Lett. 2012, 12, 5186−5191
charge/discharge processes.
5
Here we report our surprising
discovery of a selective surface lattice plane segregation of
nickel (Ni) ions for the case of LNMO as a representative case
for the transition-metal oxide-based cathode and the possible
implications of such a surface segregation on the Li ion
transport behavior in this category of cathode material. What
we have observed is a phenomenon that is far beyond general
expectation and will broadly impact the research effort for
Figure 1. (a) Overview bright-field scanning transmission electron microscopy (STEM) image of the as-obtained LNMO nanoparticles. (b) Crystal
model for the LiMO
2
R3
m parent phase based on a Li(Mn
0.5
Ni
0.5
)O
2
structure with lattice parameters: a = b = 2.887 Å, c = 14.29 Å, α = β =90°, and
γ = 120°.
24
(c) Crystal model for Li
2
MO
3
C2/m parent phase based on Li
2
MnO
3
with lattice parameters: a = 4.926 Å, b = 8.527 Å, c = 5.028 Å, α = γ
=90°, and β = 109.22° .
24
(d) Partially cation-ordered Li
2
MO
3
C2/m phase based on Ni-containing Li
2
MnO
3
.
16
(e) Z-contrast image of one sample
region corresponding to (f) [010] zone projection of the LiMO
2
R3
m model. (g) Crystal region corresponding to (h) [100] zone projection of the
Ni-doped Li
2
MnO
3
C2/m phase.
Figure 2. (a) Overview Z-contrast image of LNMO nanoparticles. (b) Atomic resolution Z-contrast image of the surface region labeled by the red
arrow in (a). (c) Atomic resolution Z-contrast image of the surface region labeled by the white arrow in (a). (d) Higher magnification image of the
surface layer shown in (c). (e) Simulated [010] zone projection Z-contrast image based on LiNi
0.5
Mn
0.5
O
2
crystal model with 20% Ni/Li disorder
corresponding to the region labeled with blue rectangle in (d). (f) Simulated [0−10] zone projection Z-contrast image based on LiNi
0.5
Mn
0.5
O
2
crystal model with 10% Ni/Li disorder corresponding to the region labeled with a white rectangle in (d). TM is transition metal. Letters A and B in
(c) marks two typical Li diffusion paths. The path A is a fast diffusion channel within the layer.
Nano Letters Letter
dx.doi.org/10.1021/nl302249v | Nano Lett. 2012, 12, 5186−51915187
enhancing the rate performance of Li ion batteries and stability
of cathode in the electrolyte.
As shown in the overview image in Figure 1a, the as-
synthesized LNMO nanoparticles assume a plate-like shape and
a significant number of the particles exhibit good surface facets.
As reported, LNMO structure comprises two components as
shown in Figure 1b−d: monoclinic Li
2
MO
3
C2/m and trigonal
LiMO
2
R3
m (α-NaFeO
2
structure) (M = transition
metal).
11,14,23−25
Both phases belong to a layered structure
composed of repeating transition-metal layers, oxygen layers
and Li layers as shown in Figure 1b−d. In the R3
m structure
(Figure 1b), Ni and Li are located at both the 3a and 3b sites,
whereas Mn is located only at the 3a site (using Wyckoff
notation).
24,26
On the other hand, in the cation ordered
Li
2
MO
3
C2/m structure (Figure 1c), the larger Ni
2+
cations
preferably replace the Li
+
cations in the transition-metal layer
(2b sites) and a small portion of Mn cations as shown in Figure
1d. We found that most of the region of the particle adopts the
α-NaFeO
2
R3
m structure in which the transition-metal layer
has the highest contrast in the high angle annular dark field
(HAADF) Z-contrast image in Figure 1e,f, while the oxygen
layers can be clearly visualized on both sides of the transition-
metal column. Due to the much smaller atomic number of Li, it
cannot be clearly seen in the Z-contrast image Figure 1e,f.
Partially ordered Li
2
MO
3
C2/m phase is also observed in part
of the particle. The Z-contrast image in Figure 1g
unambiguously identifies the partially cation-ordered region
corresponding to the [100] z one projection of the N i-
containing Li
2
MnO
3
C2/m structure model in Figure 1h.
This structure is well-characterized by two bright Mn columns
separated by 0.14 nm and the center of two neighboring Mn/
Ni dumbbell columns spaced by 0.42 nm.
16,23,27
In the Z-contrast image, we noticed that for a single particle,
some faceting planes are brighter than the others as typically
shown in Figure 2a, in which two facets adjacent to each other
were labeled with white and red arrows. Figure 2b is the atomic
resolution image of the red arrowed facet, revealing that this
facet corresponds to the (001) plane and is terminated at
transition-metal ions. On the other hand, the white arrowed
facet corresponds to the (104) plane of the particle and is
characterized by a surface layer with a higher contrast than the
interior of the particle as shown in Figure 2c. The higher Z-
contrast of this surface layer indicates a high concentration of
heavy elements, Ni (Z
Ni
= 28 compared to Z
Mn
= 25 and Z
Li
=
3) at the surface. An atomic resolution Z-contrast image
combined with the multislice image simulations as illustrated in
Figure 2d−f reveals that this surface layer corresponds to the
[010] zone axis projection of R3
m crystal structure. This
surface capping layer is oriented such that the otherwise mix of
cations and anions terminated (104) facet of the particle is also
terminated with a transition-metal layer as is similarly the case
of (001) plane. This observation clearly demonstrates that if the
termination surface ends up with a transition-metal layer (such
as (001) surface), there is no surface modification layer as
shown by Figure 2b. On the other hand, if the particle surface is
terminated with a mix of cations and anions, a surface capping
layer is formed by preferential segregation of Ni ions, which
eventually leads to the termination of the particle facet as
transition-metal ions.
The selective surface modification associated with Ni
segregation is consistently supported by the quantitative
Figure 3. (a) Z-contrast image of a nanoparticle with internal grain boundaries and XEDS maps: (b) Mn, (c) Ni, (d) overlaid Ni and Mn, and (e) O
maps (f) 3D XEDS tomography: reconstructed Mn and Ni elemental distribution maps projected at different angles showing that the Ni segregated
to certain surfaces and grain boundaries.
Figure 4. Z-contrast image and XEDS maps of multiple nanoparticle aggregate; (a) Z-contrast image, (b) Ni, (c) overlayed Ni and Mn, and (d) Ni/
(Mn + Ni) atomic percentage quantification maps. (f) Atomic percentage of Mn and Ni along the white line in (f). The scale bar in (a) applies to all
the images.
Nano Letters Letter
dx.doi.org/10.1021/nl302249v | Nano Lett. 2012, 12, 5186−51915188
composition analysis using large-area X-ray energy dispersive
spectroscopy (XEDS) mapping and tomography (Figures 3 and
4). The Z-contrast images at 0° tilt in Figure 3a show that the
grain boundaries and some surface regions exhibit an enhanced
Z-contrast, implying a higher concentration of Ni. Through the
integration of Mn, Ni, and O K
α
peaks, element specific
quantitative maps were obtained and are shown in Figure 3b−e.
Most surprisingly, the Ni distribution is extremely uneven;
while rich in grain boundaries and certain surface regions, it is
deficient in the interior of the particles. The average Ni:Mn
ratio is 1:3 as quantified using peak integrated XEDS spectra of
the whole particle, which corresponds well to the Ni:Mn ratio
in the nomnal composition Li
1.2
Ni
0.2
Mn
0.6
O
2
. A series of XEDS
maps were also acquired at 5° tilt increments from −70° to
+70° and reconstructed to provide a 3D visualization of the Ni
and Mn distribution in the LNMO nanoparticle. The 3D
morphological structure and the spatial distribution of the
segregated Ni of a single LNMO particle are illustrated by the
overlaid Mn and Ni maps projected at different tilt angles of the
reconstructed LNMO nanoparticle (Figure 3f). The animated
3D visualization of the chemical mapping of the nanoparticle is
in the Supporting Information movie. It reveals that the Ni ions
are selectively segregated at certain surface locations and grain
boundaries. This is further supported by the XEDS chemical
analysis in a region with several nanoparticle aggregates as
illustrated in Figure 4. The XEDS quanti fication reveals that the
Ni/Mn ratio ranges from ∼1:4 inside the particle to ∼1:1 in
certain surfaces or grain boundaries in Figure 4f. It should be
pointed out that Xu et al.
28
have studied the surface structure of
the LNMO particle following the charge/discharge cycles. They
noticed that a surface layer was formed on the particle due to
substitution of Li ions by transition metal cations and formation
of spinal structure. As we are focusing on the freshly prepared
sample, we do not see any formation of spinel at the surface
region of the particles.
What we have observed in this work contrasts markedly with
the general observation of other multicomponent nanoparticles
for which the segregating species tend to uniformly cover the
whole nanoparticle surface, leading to a core−shell struc-
ture.
29,30
This intere sting s elective surface modification
phenomenon is related to the surface energies of different
terminations and the diffusion kinetics of Ni cations within the
layered structure. As calculated by Wei et al.,
8
for a similar
comp osition crystal, Li[Li
0.17
Ni
0.25
Mn
0.58
]O
2
, the (001), a
surface ending with a transition-metal layer has a lower surface
energy and, therefore, is thermodynamically more stable. On
the other hand, the surfaces ending with alternating transition-
metal, oxygen, and Li layers are not thermodynamically stable.
We used DFT to calculate the diffusion energy barriers of Li
+
across different paths. It should be pointed out that the
diffusion energy barrier shows dependence on diffusion
mechanisms and overall elemental composition. For example,
Li diffusion in layered intercalation compounds, such as
Li
x
CoO
2
and Li
x
TiS
2
, is a divacancy-mediated mechanism.
31−33
Here, the calculation is based on the LiNi
0.5
Mn
0.5
O
2
R3
m
model as shown in Figure 5a (details of the calculations are
described in the Supporting Information). Two different paths
are considered for Li migration in this layered structure through
cation vacancy migration, as shown in Figure 5b. Path A is
parallel to the Li
+
layer involving an in-plane jump from the
start site to the nearby vacancy site in the Li
+
layer. As Li
+
can
also reside at the transition-metal layer, the jump from Li site to
a Ni vacancy is also considered, which indicates an off-plane
jump (path B). In Figure 2c, these two possible Li diffusion
paths A and B are labeled with arrows in the bulk crystal and
surface layers. Those two diffusion paths are also considered for
Ni migration. As Li or Ni ions move from one octahedral site to
another, it will pass through an intermediate tetrahedral site
where it encounters the repulsion from a nearby transition-
metal cation. The energy barrier for Ni or Li diffusion should be
related to the different degrees of repulsion from the transition-
metal cation. As shown in Figure 5c, the energy barriers for
both Li and Ni diffusion along path A are much lower than path
B, consistent with previous models for diffusion in this system.
Figure 5. (a) Crystal model of the LiNi
0.5
Mn
0.5
O
2
structure defined for the calculation. (b) Different diffusion paths for Li and Ni diffusion. Note
that at the end of paths A and B are a Li and a transition-metal vacancy site. (c) Energy barrier along the diffusion paths A and B.
Nano Letters Letter
dx.doi.org/10.1021/nl302249v | Nano Lett. 2012, 12, 5186−51915189
The calculation in Figure 5c shows that the energy barrier for
Ni to diffuse along the Li channel is ∼0.25 eV; therefore, it is
very easy for Ni to diffuse to the surface once Ni is in the Li
+
layer. In contrast, the diffusion energy barrier for Ni to diffuse
along path B is as high as 1.47 eV, implying that Ni is not very
likely to diffuse along path B. Similarly, the diffusion barrier for
Li
+
along path A parallel with the Li
+
layer is around 0.67 eV
compared to 0.98 eV along path B. This result corresponds well
with earlier reports regarding to the Li
+
fast diffusion paths.
34
Most importantly, the diffusion barrier for Ni along the Li
+
layer is smaller than Li
+
, implying that Ni diffusion along the Li
+
channels is even easier. Taken together, these results explain
the experimentally observed selective Ni segregation and
corresponding surface modification phenomenon described
above in Figures 2 and 3.
The charge/discharge rate of a cathode material is critically
controlled by the Li ion mobility in the lattice. Theoretical
calculations and experimental observation have established that
in the layer structured cathode, Li ion diffusion along the layer
is much easier than across the layer.
5,28,31−33
The channels
within the surface layer formed on the nanoparticle are
approximately perpendicular to the channels inside the particle
(relative rotation angle of ∼109.5°). Therefore, the formation
of the surface layer should lead to a diffusion barrier for Li at
the particle surface region. Consequently, from all we know of
the mechanism of charging/discharging, this surface layer
should affect the charge/discharge rate of the battery. Of
course, to quantify this link, more systematic experimental work
should be performed to examine the effect of nanoscale
chemical inhomogeneities and structural variations within a
single particle on the overall battery performance. The present
observations clearly indicate that this Ni surface modification
phenomenon can occur during material synthesis. As cation
diffusion is highly related to the synthesis temperature and
high-temperature growth t ime, optimizing these material
processing parameters may decrease the level the segregation
related surface layer. Several approaches were employed in this
study (sol−gel, hydrothermal, and coprecipitation) that
incorporated a wide variation in the synthesis conditions for
the LNMO system. All the LNMO synthesized for our analysis
showed some degree of Ni segregation. In a typical example, we
mixed Ni
0.25
Mn
0.75
(OH)
2
and Li
2
CO
3
, and the mixture was
calcined at 900 °C for different times. Microscopically, we
found that with a longer calcination time, the particles are
prone to possess well-defined facets and show Ni segregations.
The charge/discharge rate of batteries using these materials
shows a dependence on the calcination time of the LNMO
particles: the longer of the calcination time, the poorer the rate
performance. It should be noted that the poor rate for the
particles with a longer calcination time is also related to the
large particle size. Clarification of the exact roles of surface
segregation and particle size on the rate of the battery
performance will be the focus of future work. Further and
more broadly, this surface modification phenomenon may also
be present in other composite cathode materials and may alter
the Li ion diffusion chann els. There fore, th e present
observation of selective surface segregation will impact the
general approach for tailoring cathode structures for a high
capacity and a high power rate.
■
ASSOCIATED CONTENT
*
S
Supporting Information
Supp ortin g movie, synthesis procedure of LNMO nano-
particles, experimental pr ocedure for the STEM XEDS
mapping and tomographic reconstruction, multislice STEM
image simulations, and DFT calculation of Li
+
diffusion energy
barrier. This material is available free of charge via the Internet
at http://pubs.acs.org
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: Chongmin.Wang@pnnl.gov
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
M.G. wants to thank Dr. Chengyu Song from NCEM for
technical support on the TEAM 0.5 microscope and Dr. Paul
Plachinda from the FEI Company for his help on 3D XEDS
data processing and visualization. This work was supported by
the Laboratory Directed Research and Development (LDRD)
program of Pacific Northwest National Laboratory (PNNL).
The work was conducted in the William R. Wiley Environ-
mental Molecular Sciences Laboratory (EMSL), a national
scientific user facility sponsored by DOE’sOffice of Biological
and Environmental Research and located at PNNL. PNNL is
operated by Battelle for the DOE under contract DE-AC05-
76RLO1830. Part of the work performed at NCEM is
supported by DOE under contract no. DE-AC02-05CH11231.
■
REFERENCES
(1) Kang, B.; Ceder, G. Nature 2009, 458, 190.
(2) Chiang, Y.-M. Science 2010, 330, 1485.
(3) Tarascon, J. M.; Armand, M. Nature 2001, 414, 359.
(4) Kang, K.; Morgan, D.; Ceder, G. Phys. Rev. B 2009, 79, 014305.
(5) Nishimura, S.-I.; Kobayashi, G.; Ohoyama, K.; Kanno, R.;
Yashima, M.; Yamada, A. Nat. Mater. 2008, 7, 707.
(6) Chung, S.-Y.; Bloking, J. T.; Chiang, Y.-M. Nat. Mater. 2002, 1,
123.
(7) Delaco urt, C.; Po izot, P.; Lev asseur, S.; Masq uelier , C.
Electrochem. Solid-State Lett. 2006, 9, A352.
(8) Wei, G.-Z.; Lu, X.; Ke, F.-S.; Huang, L.; Li, J.-T.; Wang, Z.-X.;
Zhou, Z.-Y.; Sun, S.-G. Adv. Mater. 2010, 22, 4364.
(9) Hinuma, Y.; Meng, Y. S.; Kang, K.; Ceder, G. Chem. Mater. 2007,
19, 1790.
(10) Reed, J.; Ceder, G. Electrochem. Solid-State Lett. 2002, 5, A145.
(11) Lu, Z.; Beaulieu, L. Y.; Donaberger, R. A.; Thomas, C. L.; Dahn,
J. R. J. Electrochem. Soc. 2002, 149, A778.
(12) Wang, J.; Yao, X.; Zhou, X.; Liu, Z. J. Mater. Chem. 2011, 21,
2544.
(13) Peng, Q.; Tang, Z.; Zhang, L.; Liu, X. Mater. Res. Bull. 2009, 44,
2147.
(14) Armstrong, A. R.; Holzapfel, M.; Nova
k, P.; Johnson, C. S.;
Kang, S.-H.; Thackeray, M. M.; Bruce, P. G. J. Am. Chem. Soc. 2006,
128, 8694.
(15) Hong, J.; Seo, D.-H.; Kim, S.-W.; Gwon, H.; Oh, S.-T.; Kang, K.
J. Mater. Chem. 2010, 20, 10179.
(16) Jarvis, K. A.; Deng, Z.-Q.; Allard, L. F.; Manthiram, A.; Ferreira,
P. J.
Chem. Mater. 2011, 23, 3614.
(17) Wang, J.; Xia, Y.; Yao, X.; Zhang, M.; Zhang, Y.; Liu, Z. Int. J.
Electrochem. Sci. 2011, 6, 6670.
(18) Hong, Y.-S.; Park, Y. J.; Ryu, K. S.; Chang, S. H.; Kim, M. G. J.
Mater. Chem. 2004 , 14, 1424.
(19) Zhang, L.; Zhang, Z. C.; Wu, H. M.; Amine, K. Energy Environ.
Sci. 2011, 4, 2858.
Nano Letters Letter
dx.doi.org/10.1021/nl302249v | Nano Lett. 2012, 12, 5186−51915190
![Figure 1. (a) Overview bright-field scanning transmission electron microscopy (STEM) image of the as-obtained LNMO nanoparticles. (b) Crystal model for the LiMO2 R3 ̅m parent phase based on a Li(Mn0.5Ni0.5)O2 structure with lattice parameters: a = b = 2.887 Å, c = 14.29 Å, α = β = 90°, and γ = 120°.24 (c) Crystal model for Li2MO3 C2/m parent phase based on Li2MnO3 with lattice parameters: a = 4.926 Å, b = 8.527 Å, c = 5.028 Å, α = γ = 90°, and β = 109.22°.24 (d) Partially cation-ordered Li2MO3 C2/m phase based on Ni-containing Li2MnO3. 16 (e) Z-contrast image of one sample region corresponding to (f) [010] zone projection of the LiMO2 R3 ̅m model. (g) Crystal region corresponding to (h) [100] zone projection of the Ni-doped Li2MnO3 C2/m phase.](/figures/figure-1-a-overview-bright-field-scanning-transmission-2dfbs07a.webp)
![Figure 2. (a) Overview Z-contrast image of LNMO nanoparticles. (b) Atomic resolution Z-contrast image of the surface region labeled by the red arrow in (a). (c) Atomic resolution Z-contrast image of the surface region labeled by the white arrow in (a). (d) Higher magnification image of the surface layer shown in (c). (e) Simulated [010] zone projection Z-contrast image based on LiNi0.5Mn0.5O2 crystal model with 20% Ni/Li disorder corresponding to the region labeled with blue rectangle in (d). (f) Simulated [0−10] zone projection Z-contrast image based on LiNi0.5Mn0.5O2 crystal model with 10% Ni/Li disorder corresponding to the region labeled with a white rectangle in (d). TM is transition metal. Letters A and B in (c) marks two typical Li diffusion paths. The path A is a fast diffusion channel within the layer.](/figures/figure-2-a-overview-z-contrast-image-of-lnmo-nanoparticles-b-1uakrfs4.webp)
