Nano Energy 23 (2016) 97–104

Contents lists available at ScienceDirect

Nano Energy
journal homepage: www.elsevier.com/locate/nanoen

Borophene: A promising anode material offering high specific capacity

and high rate capability for lithium-ion batteries
H.R. Jiang 1, Ziheng Lu 1, M.C. Wu, Francesco Ciucci, T.S. Zhao n
Department of Mechanical and Aerospace Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China

art ic l e i nf o a b s t r a c t

Article history: In this work, we adopt first-principles calculations and ab-initio molecular dynamics simulations to in-
Received 31 January 2016 vestigate the potential of borophene as an anode material for lithium-ion batteries. It is found that
Received in revised form borophene has an adsorption energy to lithium atom of
1.12 eV, which is large enough to ensure a good
13 March 2016
lithium-borophene stability during the lithiation process. The fully lithiated phase of borophene is Li0.75B,
Accepted 15 March 2016
Available online 16 March 2016
corresponding to a theoretical specific capacity of 1860 mA h g
1, which is about 4 times larger than that
of the commercial graphite anode (372 mA h g
1). More excitingly, it is found that the energy barrier
Keywords: along the furrow of corrugated borophene is only 2.6 meV, which is much lower than those of other
Borophene widely investigated anode materials such as phosphorene (80 meV) and Ti3C2 (70 meV). The finding
Specific capacity
suggests that lithium diffusion on borophene can be extremely fast. In the meantime, a strong directional
Diffusion barrier
anisotropy is observed for lithium diffusion, with a 325.1 meV barrier perpendicular to the furrow of
Lithium-ion batteries
First-principles calculations borophene. This phenomenon is further proved by ab-initio molecular dynamics simulations at 300 K
Ab-initio molecular dynamics simulations and the result shows the lithium atom can freely drift along the furrow, but seldom jumps to the
neighboring furrows. Finally, borophene is found to exhibit metallic characteristics during the whole
lithiation process, indicating that the material has an excellent electronic conductivity. The findings re-
ported in this work suggest that borophene, as an anode material for lithium-ion batteries, has potential
to drastically boost batteries' energy density and power density.
& 2016 Elsevier Ltd. All rights reserved.

1. Introduction known example is graphene, which has been extensively in-

To enable widespread adoption of electric vehicle technology, li-
thium-ion batteries (LIBs) with high energy density, high power
density and long cycle life are in urgent demand [1–6]. Particularly,
the performance of LIBs relies greatly on the properties of constituent
electrodes [7]. In terms of anode material, graphite is commercially
used owing to its good cyclic durability, high coulombic efficiency
and low cost [8]. However, the relatively low theoretical specific
capacity (372 mA h g
1) impedes its further application. Other
widely investigated anode materials such as metal oxides, oxysalts
and silicon also suffer from critical issues including large irreversible
capacity, low lithium diffusivities and large volume change [9–11].
Therefore, there is an imperative need to search for new anode
materials to further improve the performance of LIBs.
In this regard, two-dimensional (2D) materials have attracted
much attention due to their high surface–volume ratio and unique
electronic properties compared with bulk counterparts. One well-
n
Corresponding author.
E-mail address: metzhao@ust.hk (T.S. Zhao).
1
These authors contributed equally to this work.
vestigated as an anode material and great success has been
achieved with large reversible capacity, high rate capability and
long cycle life [12–15]. Meanwhile, other novel 2D materials are
also gaining increasing interests. For example, MoS2 nanoplates
were prepared as the anode material for LIBs by Hwang et al. using
a solvothermal method and exhibited a reversible capacity of
912 mA h g
1 at 1 C and 554 mA h g
1 at 50 C [16]. MXene ‘paper’
with delaminated Ti3C2 was synthesized by Mashtalir et al. and a
capacity of 110 mA h g
1 at 36 C was obtained [17]. Besides ex-
periments, theoretical calculations also play an important role in
finding new anode materials. Jing et al. found that VS2 monolayer
had a high theoretical capacity (466 mA h g
1), low diffusion
barrier (0.22 eV) and low average open circuit voltage (0.93 V),
suggesting its potential used as the anode material in LIBs [18].
Tritsaris et al. investigated the interaction of lithium with silicene,
and showed that silicene had a high theoretical capacity of
954 mA h g
1 as well as a low lithium diffusion barrier within
0.6 eV [19]. Li et al. calculated the binding and diffusion behavior
of lithium on phosphorene and an ultrafast lithium diffusion with
a 0.08 eV barrier along zigzag direction was found [20].
Very recently, a new 2D material known as borophene (crys-
talline 2D boron sheets) was successfully synthesized on the

http://dx.doi.org/10.1016/j.nanoen.2016.03.013
2211-2855/& 2016 Elsevier Ltd. All rights reserved.
98 H.R. Jiang et al. / Nano Energy 23 (2016) 97–104
Ag (111) surface under ultrahigh-vacuum conditions [21]. Unlike
the predicted structures in previous works [22–27], the fabricated
borophene was found to have a planar, anisotropic structure by
scanning tunneling microscope (STM). More importantly, bor-
ophene shows a puckered morphology and metallic character-
istics, indicating its potential applications for metal ion storage,
fast atomic diffusion and electric conduction, which are all re-
quirements for an anode material in LIBs.
In this work, density functional theory (DFT) based first-prin-
ciples calculations and ab-initio molecular dynamics (AIMD) si-
mulations were used to investigate the adsorption and diffusion of
lithium atoms on freestanding borophene to evaluate its potential
as an anode material in LIBs. We first optimized the structure of
borophene and obtained the valence electron density. Then we
calculated the adsorption energies and Bader charges of lithium on
four high-symmetry adsorption sites. To clarify the effect of li-
thium concentration, the most stable structures of lithiated bor-
ophene with increasing lithium concentration were obtained and
the maximum theoretical charge capacity was predicted. The dif-
fusion process of lithium atom in two representative directions
was characterized using both climbing-image nudged elastic band
(CI-NEB) theory and AIMD simulations. Finally, by calculating the
electron density of states (DOS) with different lithium contents,
the electronic property of lithium-stuffed borophene was further
investigated.
2. Computational methods
All DFT calculations were carried out using Vienna ab-initio
simulation package (VASP) [28] adopting the Perdew-Burke-Ern-
zerhof (PBE) generalized gradient approximation [29]. Projector-
augmented-wave (PAW) potentials [30] were used to take into
account the electron–ion interactions. An energy cutoff of 500 eV
was used for the plane wave-basis expansion and the convergence
criteria of the electron self-consistent loop was set to 10
7 eV. The
structural relaxation was carried out with a conjugate gradient
algorithm until the Hellmann–Feynman forces acting on the atoms
were less than 10
4 eV/Å. A vacuum layer of 20 Å was added to
avoid image-image interaction in all calculations. To calculate the
lattice parameters of pristine borophene, a unit cell containing
2 boron atoms was used and the Brillouin zone was sampled on a
25  15  1 Monkhorst-Pack grid [31]. After a convergence test
(Fig. S1), a 4  3 supercell was used for the calculation of single
lithium adsorption/diffusion to avoid the Li–Li interaction. Smaller
supercells were adopted when dealing with the effect of lithium
concentration to maintain a reasonable computational cost. For all
supercell calculations, k-point meshes with the same density as
the unit cell were used. The adsorption energy of lithium atoms on
borophene was obtained from:
Ead = ( ELi + B − nμLi − EB )/n
where Ead and EB are the DFT total energies with and without
lithium adatoms, μLi is the chemical potential of lithium leveled
from lithium metal, and n is the number of adsorbed lithium
atoms.
For the calculation of lithium diffusion barrier, the climbing-
image nudged elastic band method [32,33] was adopted. 7 images
were used and the convergence criteria was set to be 0.01 eV/Å. To
evaluate the temperature-dependent molecular transition rate
[18,20,34], diffusion constant was calculated by Arrhenius equa-
tion:
⎛ −E ⎞
D ∼ exp ⎜ a ⎟
⎝ kB T ⎠
where D is diffusion constant, Ea and kB are diffusion barrier and
Boltzmann's constant, respectively, and T is the temperature.
Charge transfer between lithium and borophene was estimated
using the grid-based Bader charge method [35].
The AIMD simulations were used to investigate the lithium
migration dynamics on borophene. The simulations were per-
formed on a 4  3 borophene supercell containing one lithium
atom at 300 K using a NVT ensemble. The time step was set to be
1 fs. The system was first equilibrated for 1 ps and the following
4 ps were used for trajectory sampling. A 3  3  1 k-point mesh
was adopted during the AIMD simulation.
3. Results and discussion
3.1. Atomic structure of monolayer borophene
The optimized structure of freestanding borophene and valence
electron densities are presented in Fig. 1. It is confirmed that the
borophene has a puckered surface with two atomic boron layers
and there exists repeated corrugations along a-direction, which is
different from the planar surface of graphene [36]. The optimized
lattice parameters are a ¼1.614 Å and b¼ 2.861 Å, in agreement
with the experimental and theoretical results reported earlier
[21,37]. Each boron atom in borophene bonds to six neighbors,
with two of them in the same layer and the other four in the
neighboring layer. The calculated B–B bond lengths are 1.615 Å for
the atoms within the same layer and 1.883 Å for the atoms from
different layers. The two kinds of bond angles for each boron atom
are 64.63° and 50.79°, respectively.
3.2. Lithium adsorption on borophene
A fundamental requirement of a potential anode material for
LIBs is that the material must have a relatively large adsorption
energy to lithium. To this end, we first investigate the adsorption
behavior of a single lithium atom on borophene. Four high-sym-
metry adsorption sites are considered: i) TF sites: above the top of
boron atoms in the furrow; ii) BF sites: above the middle-point of
B–B bonds in the furrow; iii) TR sites: above the top of boron atoms
in the ridge; iv) BR sites: above the middle-point of B–B bonds in
the ridge.
The optimized structures (top and side views) of single lithium
atom adsorbed on borophene and charge difference plots are
shown in Fig. 2. It is found that TF site is the most stable one with
an adsorption energy of
1.12 eV, and a much larger value

2.75 eV is obtained if the chemical potential of lithium is leveled
from lithium atom instead of lithium metal. In this case, the dis-
tances between the lithium atom and neighboring boron atoms
are 2.386 Å. Interestingly, the adsorption energy on BF site is

1.11 eV and the difference between this value and that of the
most favorable TF site is extremely small, suggesting the lithium
(1)
adsorption energy along the furrow is similar and may lead to a
low diffusion barrier along this direction. This unique phenom-
enon is further explained by Bader charge analysis. The results
show that the electrons transfer from lithium to borophene at the
TF and BF sites have comparable values, i.e., 0.888 and 0.887 |e|,
respectively. In the meantime, the adsorption energies of lithium
atom on the two ridge sites TR and BR are
0.81 eV and
0.79 eV,
smaller than those on furrow sites TF and BF. This is consistent
with fewer Li–B bonds on the ridge sites. To better understand the
nature of lithium-borophene bonding, the charge difference plots
of these four high-symmetry sites are shown in Fig. 2c. It can be
seen that electrons tend to accumulate around boron atoms and
their density surrounding lithium atom decreases. The observed
(2) electrons transfer agree with the larger electronegativity of boron
 H.R. Jiang et al. / Nano Energy 23 (2016) 97–104 99

Fig. 1. (a and b) Side views, (c) top view and (d) general view of optimized structure of borophene; valence electron density in (e) a- and (f) b-directions. To facilitate the
visualization, the boron atoms in different atomic layers are colored in either red or yellow.

Fig. 2. (a) Top and (b) side views of four high-symmetry lithium adsorption sites (TF, BF, TR, BR), (c) corresponding charge density difference plots. The yellow and blue areas
represent electron gains and loses. The charge difference plot is calculated by ρ = ρtotal − ρborophene − ρLi , where ρtotal , ρborophene and ρLi are the total charge of the system,
borophene and lithium atom, respectively. The green ball represents lithium atom.

(2.04) than lithium (0.98). Also, the low electron localization consistently with the results obtained from Bader charge analysis.
function (ELF) value between lithium and boron cores means the It is known that the clustering of lithium limits the application of
ionic nature of Li–B bonds is strong, as shown in Fig. S2, graphene as anode material [38,39]. Therefore, we compare the
100 H.R. Jiang et al. / Nano Energy 23 (2016) 97–104

adsorption energy of lithium atom on borophene (
2.75 vs. li-

thium atom) with the binding energy of Li2 (
0.86 eV) dimer and
cohesive energy of lithium bulk (
1.63 eV). The result suggests
that lithium atoms would form 2D layers instead of clusters on the
borophene surfaces at low concentrations [40], which is beneficial
to the safety of LIBs.

3.3. Effect of lithium concentration and theoretical specific capacity

During lithiation/delithiation process, the lithium concentra-

tion will increase/decrease, which will induce compositional and
structural change of the anode material. Thus, it is of great sig-
nificance to investigate the effect of lithium concentration. The
lithium adsorption at different concentrations (LixB) is firstly in-
vestigated by gradually increasing the number of lithium atoms on
borophene surface using a 3  2 supercell. The x values of LixB are
0.08, 0.17, 0.25, 0.33, 0.42, 0.50, 0.58 and 0.67, corresponding to
Fig. 4. Adsorption energies as the function of lithium concentration x in LixB.
LiB12, LiB6, LiB4, LiB3, Li5B12, LiB2, Li7B12 and Li2B3, respectively. The
most stable structures at each x value are shown in Fig. 3a–h and
increases, the adsorption of lithium atoms continuously weakens
the corresponding adsorption energies are summarized in Fig. 4.
due to the enhanced Li–Li repulsion. The highest lithium con-
Other calculated structures and adsorption energies are presented
centration in a 3  2 supercell is x ¼0.67 and the adsorption energy
in supporting information (Figs. S3–S8). It is shown that the li-
thium atom tends to adsorb on TF site when x ¼0.08, which is is
0.30 eV/atom, as shown in Fig. 3h. In this case, both sides of
identical to the results obtained using a larger 4  3 supercell. the borophene surface are fully covered with lithium atoms, which
When x¼0.17, the adsorption energy decreases from
0.94 eV to are all adsorbed in the furrows. If more lithium atoms are added

0.88 eV, and the two lithium atoms prefer to adsorb on different into the furrows, the inserted atoms are pushed out of the surface
sides of borophene to decrease the repulsion between them, as of borophene and form the second lithium layer on surface of
shown in Fig. 3b. During the relaxation process, lithium atoms adsorbed lithium. Therefore, the highest lithium concentration
spontaneously move from the TF to the BF site, consistently with using a 3  2 borophene supercell corresponds to an x value of
the results that the adsorption energy difference along the furrow 0.67. It is interesting to note that, unlike other 2D materials that
is extremely small. With the lithium concentration further specific lithium adsorption sites exist during lithiation process

Fig. 3. The most stable structures at different x values of LixB, the x values of are 0.08, 0.17, 0.25, 0.33, 0.42, 0.50, 0.58, 0.67 and 0.75 for (a)–(h), respectively.
 H.R. Jiang et al. / Nano Energy 23 (2016) 97–104
(e.g., the center of the hexagonal ring in silicene [19], H site in
phosphorene [41] and T site in VS2 [18]), lithium atoms can adsorb
almost anywhere in the furrows of borophene due to the ex-
tremely small adsorption energy difference (o 0.01 eV). Therefore,
the highest lithium concentration is dependent on the minimum
distance between lithium atoms in the furrows and is affected by
the supercell size.
To exclude the influence of supercell size and find the fully li-
thiated phase, we add 2  (N
1) lithium atoms into the furrows of
borophene in N  1 supercells (N4 2) and relax the structures. The
fully lithiated phase of borophene needs to satisfy several re-
quirements: i) it has the highest lithium concentration; ii) the
adsorption energy to lithium atoms is negative; iii) there is no li-
thium pushed out of the borophene surface; iv) the structure of
borophene does not have irreversible deformation [19,38,42,43].
By examining the relaxed structures, the maximum N value could
be obtained and the fully lithiated phase is found. The optimized
structure of 4 lithium atoms in a 3  1 supercell representing
x ¼0.67 is shown in Fig. S9 and it has a similar adsorption energy
as the result obtained from 3  2 supercell, validating our ap-
proach. When 6 lithium atoms are added into a 4  1 supercell, a
higher concentration of x ¼0.75 is obtained with a lithium ad-
sorption energy of
0.06 eV/atom, as shown in Fig. 3i. However,
when adding 8 atoms in a 5  1 supercell and 10 atoms in a 6  1
supercell, some lithium atoms are pushed out of the surface and
the structure of borophene is destroyed (Fig. S11), which suggests
the fully lithiated phase is Li0.75B, corresponding to the N value of
4 and an ultra-high theoretical specific capacity of 1860 mA h g
1.
Furthermore, the stability of borophene during repeated lithiation
and delithiation was evaluated by adopting the method in Tritsaris
et al.'s work, in which all lithium atoms were removed and the
structure was relaxed [19]. Result shows that after removal of all
the lithium atoms, the structure of borophene is not destroyed and
can recover its initial form. To further validate this result, a fully
lithiated 4  3 supercell at x ¼0.75 is also calculated (Fig. S10).
Identical lithium adsorption energies and optimized structures are
obtained compared with that obtained from the 4  1 supercell.
3.4. Diffusion of lithium on the surface of borophene
In addition to the high specific capacity, a good rate capability is
another requirement for a promising anode material for LIBs,
especially in high power density application such as electric ve-
hicles. It is known that the rate capability is mainly dependent on
Fig. 5. (a) Energy profile of lithium diffusion along TF
BR
TF and TF
BF
TF pathways; (b) top and side views of lithium diffusion along TF
BR
TF path; (c) top and side
views of lithium diffusion along TF
BF
TF path.
101
the transport properties of lithium ions and electrons [9], corre-
sponding to the lithium diffusion barrier and electronic con-
ductivity of borophene. Here the lithium diffusion kinetics is in-
vestigated using CI-NEB and AIMD simulations. Considering the
highly symmetric structure of borophene, two representative dif-
fusion pathways are calculated: one is path TF
BF
TF which is
along the furrow and the other is path TF
BR
TF which is per-
pendicular to the furrow, as presented in Fig. 5b and c, respec-
tively. The energy profiles of lithium diffusion along two pathways
are also illustrated in Fig. 5a. It is shown that the lithium diffusion
barrier in the TF
BR
TF path is 325.1 meV, where the lithium
atom migrates from the most stable TF site towards another TF site
in the neighboring furrow through a transition state BR site be-
tween them. In contrast, a diffusion barrier of only 2.6 meV for
lithium is found along the direction of the furrow in the TF
BF
TF
path, much lower than that in TF
BR
TF path. This ultra-low
diffusion barrier indicates the lithium atoms are extremely easy to
diffuse along the furrow and all sites tend to have a relatively si-
milar lithium adsorption state, which is consistent with the find-
ings in previous parts. The calculation of lithium diffusion barrier
is based on the 0 K results, but thermal vibrations in the Li-bor-
ophene nanostructure are inevitable at the working temperature
of a LIBs. Hence, the diffusion constants derived from Arrhenius
equation are taken into account to evaluate the temperature de-
pendent molecular transition rates along two representative
pathways. The calculated diffusion constant along the furrow in
TF
BF
TF path is 105 times larger than that perpendicular to the
furrow in TF
BR
TF path at 300 K, showing a highly anisotropic
behavior for lithium diffusion in borophene. In other words, due to
the much lower diffusion barrier (125 times) and larger diffusivity
(105 times), lithium atoms tend to diffuse along the furrow on
borophene surface and migration to neighboring furrows is rela-
tively difficult.
To further explicitly prove the ultra-fast, highly anisotropic li-
thium diffusion on borophene, AIMD simulations are performed at
300 K. The three coordinates of lithium during the last 4 ps si-
mulation are extracted from the trajectory and shown in Fig. 6b. It
can be seen that the b- and c-coordinates of the lithium atoms are
limited within a small range, which means along the direction
perpendicular to the furrow, the motion of lithium atoms is hin-
dered by the relatively large diffusion barrier. In contrast, along the
furrow direction (a-direction), the lithium atoms are allowed to
drift almost freely. A more direct visualization of lithium diffusion
is presented in Fig. 6a where lithium trajectories are shown. From
102 H.R. Jiang et al. / Nano Energy 23 (2016) 97–104

Fig. 6. (a) The calculated lithium trajectories and (b) extracted coordinates in three directions during the last 4 ps by AIMD simulations.

the 4 ps lithium trajectories, the lithium atom is able to travel

through the entire region in the furrow but there is no observation
of lithium jumping towards the neighboring one. Another inter-
esting point to mention here is that, unlike most diffusion pro-
cesses, which consist of long lattice vibrations followed by abrupt
jumps, the drift of lithium on borophene is nearly ballistic. Pre-
viously, similar behavior has only been observed in systems where
van der Waals interaction dominates [44,45]. Although more
thorough investigations should be conducted to validate this
phenomenon and some fundamental understanding is still miss-
ing, here we phenomenologically attribute this to the nearly
negligible migration barrier along the furrow direction.

3.5. Electronic properties of pristine and lithiated borophene

The electronic conductivity is another property that affects the

rate capability of the anode material in LIBs, electronic properties
of pristine and lithiated borophene are therefore studied by cal-
culating their DOS, as shown in Fig. 7. Unlike many other 2D
materials that exhibit semi-conducting/semi-metallic character-
istics [46,47], borophene is intrinsically metallic, which is also
reported by previous work [21]. Therefore, it is expected that
pristine borophene has enough electronic conductivity as anode
material even with no further structural modification. The DOS of
the structures during lithiation process are also calculated because
the insertion of lithium into the anode materials could affect the
electronic property of the pristine phase [20]. As we can see, the
electronic structure undergoes some changes upon lithiation as a
consequence of electron transfer from lithium to borophene. Fig. 7. The density of states of (a) pristine borophene, (b) Li0.08B, (c) Li0.67B and (d)
However, the metallic nature of borophene remains unchanged Li0.75B. The Fermi levels have been shifted to zero.
during the whole process and an excellent electronic conductivity
is maintained.
borophene are 4.4 and 107 times larger than the commercialized
3.6. Comparison with other anode materials anode material graphite. When compared with other 2D mate-
rials, the specific capacity of borophene is almost 2 times that of
To have a more comprehensive view of the specific capacity silicene and more than 3 times larger than others. Moreover, the
and diffusion barrier of borophene, the properties of some most lithium diffusion on borophene is estimated to be 103 and 10
investigated anode materials are given for reference in Table 1. times faster than that on silicene and Ti2C3 at room tempera-
It is found that borophene has the second largest theoretical ture. We also notice that Banerjee et al. [26] investigated the
specific capacity and the lowest diffusion barrier among the
possibility of computationally predicted 2D-boron sheets as an
listed anode materials. Although the theoretical specific capa-
anode material in LIB. Their results showed a specific capacity of
city of borophene is 1.2 times lower than silicon, the lithium
383 mA h g
1 and lithium diffusion barrier of 0.39 eV. The bet-
diffusivity is 109 times faster than that of silicon at room tem-
perature, suggesting a much better rate capability. In addition, ter performance achieved in our work is attributed to the un-
the theoretical specific capacity and lithium diffusivity on ique puckered surface of synthesized borophene.
 H.R. Jiang et al. / Nano Energy 23 (2016) 97–104
Table 1
Theoretical specific capacity and diffusion barrier of some current widely investigated anode materials.
Species Theoretical specific capacity (mA h g-1)
Borophene 1860
Graphite 372
VS2 466
Phosphorene 433
Ti3C2 320
Silicene 954
Silicon 4200
Sn 994
Li4Ti5O12 175
4. Conclusion
In summary, DFT based first-principles calculations and AIMD
simulations were used to investigate the potential of borophene as
an anode material in LIBs. It is found that TF is the most favorable
adsorption site for a single atom and a large adsorption energy
with the value of
1.12 eV is obtained. The adsorption energies
per lithium atom gradually decrease with the increased lithium
concentration due to the enhanced Li–Li repulsion. In addition, the
fully lithiated phase of borophene is Li0.75B, corresponding to a
high theoretical specific capacity up to 1860 mA h g
1, which is
much larger than other 2D materials. More excitingly, lithium
diffusion on borophene is extremely fast with an energy barrier of
only 2.6 meV along the furrow of corrugated borophene. This is
much lower than those of other investigated anode materials such
as silicon (580 meV), phosphorene (80 meV) and Ti3C2 (70 meV).
In the meantime, a strong directional anisotropy is observed for
lithium diffusion, with a 325.1 meV barrier perpendicular to the
furrow of borophene, which is further supported by AIMD simu-
lations at 300 K and the result shows the lithium atom can freely
drift along the furrow, but rarely jumps to the neighboring fur-
rows. Finally, borophene is found to exhibit metallic characteristics
during the whole lithiation process, indicating that the lithium-
borophene nanostructures have an excellent electronic con-
ductivity. Given these advantages, it is expected that borophene
can be a promising anode material in LIBs with high energy den-
sity and high power density. In addition, we speculate that bor-
ophene is also an excellent anode candidate for other metal-ion
(e.g., Na-ion, K-ion, Mg-ion, Ca-ion, Al-ion) batteries and the
method used here is applicable. We strongly encourage both
theoretical and experimental works to follow up.
Acknowledgments
The work described in this paper was fully supported by a grant
from the Research Grants Council of the Hong Kong Special Ad-
ministrative Region, China (Project no. 16213414).
Appendix A. Supporting information
Supplementary data associated with this article can be found in
the online version at http://dx.doi.org/10.1016/j.nanoen.2016.03.013.
References
[1] Y. Idota, T. Kubota, A. Matsufuji, Y. Maekawa, T. Miyasaka, Science 276 (1997)
1395–1397.
[2] J. Tarascon, M. Armand, Nature 414 (2001) 359–367.
[3] K.T. Nam, D.W. Kim, P.J. Yoo, C.Y. Chiang, N. Meethong, P.T. Hammond, Y.
M. Chiang, A.M. Belcher, Science 312 (2006) 885–888.
103
Diffusion barrier (meV) References
2.6 This work
450  1200 [2,48,49]
220 [18]
80 [20,50]
70 [51]
230 [19]
580 [52]
390 [53,54]
300 [53,55]
[4] F. Ning, Y. He, B. Li, H. Du, D. Zhai, F. Kang, J. Alloy. Compd. 513 (2012) 524–529.
[5] Y. Shi, Y. Wang, J.I. Wong, A.Y.S. Tan, C. Hsu, L. Li, Y. Lu, H.Y. Yang, Sci. Rep. 3
(2013) 2169.
[6] Q. Zhang, J. Wang, J. Dong, F. Ding, X. Li, B. Zhang, S. Yang, K. Zhang, Nano
Energy. 13 (2015) 77–91.
[7] M. Obrovac, V. Chevrier, Chem. Rev. 114 (2014) 11444–11502.
[8] M. Endo, C. Kim, K. Nishimura, T. Fujino, K. Miyashita, Carbon 38 (2000)
183–197.
[9] V. Etacheri, R. Marom, R. Elazari, G. Salitra, D. Aurbach, Energy Environ. Sci. 4
(2011) 3243–3262.
[10] M. Reddy, G. Subba Rao, B. Chowdari, Chem. Rev. 113 (2013) 5364–5457.
[11] M.T. McDowell, S.W. Lee, W.D. Nix, Y. Cui, Adv. Mater. 25 (2013) 4966–4985.
[12] E. Yoo, J. Kim, E. Hosono, H. Zhou, T. Kudo, I. Honma, Nano Lett. 8 (2008)
2277–2282.
[13] P. Guo, H. Song, X. Chen, Electrochem. Commun. 11 (2009) 1320–1324.
[14] Z. Wu, W. Ren, L. Xu, F. Li, H. Cheng, ACS Nano 5 (2011) 5463–5471.
[15] W. Ai, J. Jiang, J. Zhu, Z. Fan, Y. Wang, H. Zhang, W. Huang, T. Yu, Adv. Energy
Mater. 5 (2015) 1500559.
[16] H. Hwang, H. Kim, J. Cho, Nano Lett. 11 (2011) 4826–4830.
[17] O. Mashtalir, M. Naguib, V.N. Mochalin, Y. Dall’Agnese, M. Heon, M.
W. Barsoum, Y. Gogotsi, Nat. Commun 4 (2013) 1716.
[18] Y. Jing, Z. Zhou, C.R. Cabrera, Z. Chen, J. Phys. Chem. C 117 (2013) 25409–25413.
[19] G.A. Tritsaris, E. Kaxiras, S. Meng, E. Wang, Nano Lett. 13 (2013) 2258–2263.
[20] W. Li, Y. Yang, G. Zhang, Y. Zhang, Nano Lett. 15 (2015) 1691–1697.
[21] A.J. Mannix, X.F. Zhou, B. Kiraly, J.D. Wood, D. Alducin, B.D. Myers, X. Liu, B.
L. Fisher, U. Santiago, J.R. Guest, M.J. Yacaman, A. Ponce, A.R. Oganov, M.
C. Hersam, N.P. Guisinger, Science 350 (2015) 1513–1516.
[22] X. Yang, Y. Ding, J. Ni, Phys. Rev. B 77 (2008) 041402.
[23] E.S. Penev, S. Bhowmick, A. Sadrzadeh, B.I. Yakobson, Nano Lett. 12 (2012)
2441–2445.
[24] X. Yu, L. Li, X. Xu, C. Tang, J. Phys. Chem. C 116 (2012) 20075–20079.
[25] W. Li, Q. Chen, W. Tian, H. Bai, Y. Zhao, H. Hu, J. Li, H. Zhai, S. Li, L. Wang, J. Am.
Chem. Soc. 136 (2014) 12257–12260.
[26] S. Banerjee, G. Periyasamy, S.K. Pati, J. Mater. Chem. A 2 (2014) 3856–3864.
[27] B. Zheng, H. Yu, Y. Xie, Y. Lian, ACS Appl. Mater. Interfaces 6 (2014) 19690–19701.
[28] G. Kresse, J. Hafner, Phys. Rev. B 47 (1993) 558–561.
[29] J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77 (1996) 3865–3868.
[30] P.E. Blöchl, Phys. Rev. B 50 (1994) 17953–17979.
[31] H.J. Monkhorst, J.D. Pack, Phys. Rev. B 13 (1976) 5188–5192.
[32] G. Henkelman, B.P. Uberuaga, H. Jónsson, J. Chem. Phys. 113 (2000)
9901–9904.
[33] P. Maragakis, S.A. Andreev, Y. Brumer, D.R. Reichman, E. Kaxiras, J. Chem. Phys.
117 (2002) 4651–4658.
[34] W. Li, Y. Mu, Nanoscale 4 (2012) 1154–1159.
[35] G. Henkelman, A. Arnaldsson, H. Jónsson, Comp. Mater. Sci. 36 (2006)
354–360.
[36] X. Fan, W. Zheng, J. Kuo, ACS Appl. Mater. Interfaces 4 (2012) 2432–2438.
[37] B. Peng, H. Zhang, H. Shao, Y. Xu, R. Zhang, H. Zhu, arXiv preprint ar-
Xiv:1601.00140, 2016.
[38] X. Fan, W. Zheng, J. Kuo, D.J. Singh, ACS Appl. Mater. Interfaces 5 (2013)
7793–7797.
[39] M. Liu, A. Kutana, Y. Liu, B.I. Yakobson, J. Phys. Chem. Lett. 5 (2014) 1225–1229.
[40] V.V. Kulish, O.I. Malyi, C. Persson, P. Wu, Phys. Chem. Chem. Phys. 17 (2015)
13921–13928.
[41] T. Hu, J. Hong, J. Phys. Chem. C 119 (2015) 8199–8207.
[42] Q. Li, C. Duan, X. Wan, J. Kuo, J. Phys. Chem. C 119 (2015) 8662–8670.
[43] G. Guo, X. Wei, D. Wang, Y. Luo, L. Liu, J. Mater. Chem. A 3 (2015) 11246–11252.
[44] S. Jakobtorweihen, M. Verbeek, C. Lowe, F. Keil, B. Smit, Phys. Rev. Lett. 95
(2005) 044501.
[45] I.V. Lebedeva, A.A. Knizhnik, A.M. Popov, O.V. Ershova, Y.E. Lozovik, B.
V. Potapkin, Phys. Rev. B 82 (2010) 155460.
[46] K.F. Mak, C. Lee, J. Hone, J. Shan, T.F. Heinz, Phys. Rev. Lett. 105 (2010) 136805.
[47] H. Liu, A.T. Neal, Z. Zhu, Z. Luo, X. Xu, D. Tománek, P.D. Ye, A.C.S. Nano, 8 (2014)
4033–4041.
[48] K. Toyoura, Y. Koyama, A. Kuwabara, F. Oba, I. Tanaka, Phys. Rev. B 78 (2008)
214303.
[49] P. Ganesh, J. Kim, C. Park, M. Yoon, F.A. Reboredo, P.R. Kent, J. Chem. Theory
Comput. 10 (2014) 5318–5323.
104 H.R. Jiang et al. / Nano Energy 23 (2016) 97–104

[50] Q. Li, C. Duan, X. Wan, J. Kuo, J. Phys. Chem. C 119 (2015) 8662–8670.
[51] Q. Tang, Z. Zhou, P. Shen, J. Am. Chem. Soc. 134 (2012) 16909–16916.
[52] W. Wan, Q. Zhang, Y. Cui, E. Wang, J. Phys. : Condens. Matter 22 (2010) 415501.
[53] W. Zhang, J. Power Sources 196 (2011) 13–24.
[54] C. Chou, H. Kim, G.S. Hwang, J. Phys. Chem. C 115 (2011) 20018–20026.
[55] B. Ziebarth, M. Klinsmann, T. Eckl, C. Elsässer, Phys. Rev. B 89 (2014) 174301.
Haoran Jiang was born in Shandong, China. He re-
ceived his Bachelor's degree in School of Energy and
Power Engineering from Huazhong University of Sci-
ence and Technology, China in 2014. He is currently
pursuing his Ph.D. degree at the Department of Me-
chanical and Aerospace Engineering, The Hong Kong
University of Science and Technology with Prof. T.S.
Zhao. His research focuses on first-principles study of
metal-air and metal-ion batteries.
Francesco Ciucci is an assistant professor in Mechan-
ical and Aerospace Engineering and in Chemical and
Biomolecular Engineering at the Hong Kong University
of Science and Technology. He received his Ph.D. in
Engineering and Applied Sciences from the California
Institute of Technology where he was a Rotary Am-
bassadorial Scholar and a Bechtel Fellow. Francesco did
postdoctoral work at the Institute of Scientific Com-
puting at the University of Heidelberg where he re-
ceived a Marie Curie Grant from the European Union
and a fellowship from the Heidelberg graduate school.
Prior to that he graduated with a dual M.Sc. degree in
Applied Physics at Ecole Centrale Paris, France and in
Aerospace Engineering at Politecnico di Milano, Italy. His current research focuses
on fuel cells, batteries, and the modeling solid state ionic devices using continuum
and atomistic simulations.

T.S. Zhao is currently the Chair Professor of Mechanical

& Aerospace Engineering at HKUST, the Director of the
HKUST Energy Institute, and a Senior Fellow of the
HKUST Institute for Advanced Study. He is an elected
Fellow of the American Society Mechanical Engineers
(ASME), Fellow of the Royal Society of Chemistry (RSC),
Ziheng Lu was born in Nanjing, China. He received his and a Highly Cited Researcher in Engineering by
B.E. degree at Nanjing University of Science and Tech- Thomson Reuters (2014 and 2015).
nology, China in 2014. After that, he joined Ciucci's
group at Hong Kong University of Science and Tech-
nology to pursue his Ph.D. degree. His current research
interests include molecular dynamics simulations and
density functional theory calculations applied to ma-
terials for lithium and sodium batteries.

Maochun Wu was born in Guangdong, China. He re-

ceived his Bachelor's degree in School of Chemistry and
Chemical Engineering from South China University of
Technology, China in 2014. He is currently pursuing his
Ph.D. degree at the Department of Mechanical and
Aerospace Engineering, The Hong Kong University of
Science and Technology with Prof. T.S. Zhao. His re-
search focuses on lithium-air batteries.