Computational Materials Science xxx (2015) xxxxxx

Contents lists available at ScienceDirect

Computational Materials Science

journal homepage: www.elsevier.com/locate/commatsci

Recent advances in computational studies of organometallic sheets:

Magnetism, adsorption and catalysis
Guizhi Zhu a, Qiang Sun a,b,
a
b

Department of Materials Science and Engineering, Peking University, Beijing 100871, China
Center for Applied Physics and Technology, Peking University, Beijing 100871, China

a r t i c l e

i n f o

Article history:
Received 14 May 2015
Received in revised form 1 July 2015
Accepted 4 July 2015
Available online xxxx
Keywords:
2D materials
Organometallic sheets
Computational studies
Review

a b s t r a c t
The unique geometry and novel properties of graphene have tremendously motivated scientists to
explore other monolayer materials, especially those with separately distributed and exposed metal ions
for magnetism, hydrogen storage, CO2 capture and catalysis. The recent successful synthesis of 2D
organometallic sheets has opened a new pathway to design and fabricate such desirable 2D materials
going beyond graphene and other inorganic sheets. This article briey reviews the recent advances in
computational studies of 2D organometallic sheets based on density functional theory, quantum chemistry modeling and Monte Carlo simulation focusing on stability, magnetic coupling, magnetism tuning,
hydrogen storage, CO2 capture and catalysis. Future research directions in this eld are also discussed.
2015 Published by Elsevier B.V.

1. Introduction
The discovery of graphene in 2004 has opened up a new epoch
for two dimensional (2D) monolayer materials [1]. The hitherto
reported monolayers, such as graphene, boron nitride (BN), silicene, and MoS2 [14], exhibit various novel properties and have
potential applications in lithium ion batteries, integrated circuits,
transparent conducting electrodes, photoluminescence and valleytronics [510]. However, neither of these graphene-like inorganic monolayer materials are intrinsically magnetic in their
pristine forms, nor of them show a distinctive property in gas
adsorption and catalysis. It has been found that the metal adatom
migration activation barriers for the lowest energy migration paths
on pristine monolayer, bilayer, and trilayer graphene are smaller
than or within an order of magnitude of kBT (0.026 eV) at room
temperature, implying very high mobilities for the adatoms. For
example, the binding energy of a Cr atom with graphene is about
0.5 eV and the migration energy barrier is only 0.02 eV [11].
This suggests that metal atoms on graphene quickly migrate across
the lattice and bind together forming clusters. One possible way is
to make graphene porous so that the edge state may prevent the
introduced metal atom from clustering [1214]. However, it is very

Corresponding author at: Department of Materials Science and Engineering,

Peking University, Beijing 100871, China.
E-mail address: sunqiang@pku.edu.cn (Q. Sun).

difcult to control the shape, size and distribution of defects

precisely in experiment, and the induced magnetic properties
depend sensitively on experimental conditions and structural
morphologies [15]. While the problem for the intrinsically
metal-containing sheets like MnO2, MoS2 is that in these monolayers, the metal ions are separately distributed but not well
exposed [16,17], therefore they are not good for gas adsorption
and catalysis.
The key challenge becomes how to synthesize a sheet with separately distributed and exposed metal ions? For design and synthesis, phthalocyanine (Pc) and porphyrin (Por) are widely used as
ideal constituent units, where the pores endow the feasibility of
embedding metal atoms or complexes. The successful synthesis
of 2D FePc-based sheet paved the way for exploring 2D
organometallic materials where the metal species could be
replaced with other elements using STM tip, showing exibility
and diversity [18,19]. Furthermore, the diverse organic ligands
with extended p-conjugation are interesting and appealing oligomers due to their versatile functional properties that show a wide
range of applications [20]. The high thermal and chemical stabilities, facile synthesis and robust nature make the organometallic
molecules attractive building blocks for the assembly of monolayer
nanomaterials [2123]. Based on the theoretical and experimental
advances on diverse organometallic sheets, scientists have
obtained high surface area materials with separately distributed
and exposed metal ions for magnetism, adsorption and catalysis
[24,25].

http://dx.doi.org/10.1016/j.commatsci.2015.07.020
0927-0256/ 2015 Published by Elsevier B.V.

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020

G. Zhu, Q. Sun / Computational Materials Science xxx (2015) xxxxxx

This review briey summarizes the recent advances made by

Chinese research groups in computational studies of organometallic sheets focusing on tunable magnetism, hydrogen storage, CO2
capture and catalysis.
2. Tunable magnetism
Previous studies have demonstrated that the magnetism of
nanomaterials distinguishes from that of bulk materials in a fundamental way because both magnetic coupling and magnetic
moments are strongly dependent on coordination number, symmetry, bond-length and dimensionality-related quantum connements [26], which provide the variables for tuning the magnetism.
2.1. Phthalocyanine-based organometallic sheets
The experimental success in embedding transition metal (TM)
atoms in 2D Pc sheets paves the way for achieving the longstanding dream of 2D atomic sheets with regularly and separately distributed TM atoms [18]. To better understand this kind of 2D
sheets, rst-principles calculations have been carried out to investigate the electronic and magnetic properties, as shown in Fig. 1
[27]. It has been found that among the 3d elements from Cr to
Zn, only the 2D MnPc framework is ferromagnetic (FM), while 2D
CrPc, FePc, CoPc, and CuPc are antiferromagnetic (AFM) and 2D
NiPc and ZnPc are nonmagnetic. The difference in magnetic couplings for the studied systems is found to be related to the different
orbital interactions. The calculated projected densities of states of
the central metal ions evince that only MnPc displays metallic
dxz and dyz orbitals that can hybridize with p electrons of Pc moiety, which mediates the long-range FM coupling (Fig. 1). Monte
Carlo simulations based on the Ising model suggested that the

Curie temperature (TC) of the 2D MnPc framework is 150 K,

which is comparable to the highest TC achieved experimentally
in Mn-doped GaAs. The 2D TMPc nanostructures exhibit very interesting magnetic properties in comparison with dilute magnetic
semiconductors (DMSs) such as Mn-doped Si, GaN, or ZnO, where
the doped TM atoms easily form clusters [28,29], thus resulting in
nonintrinsic magnetism.
From above one can see that a freestanding poly-MnPc sheet
has stable FM order, while most of the other poly-TMPc sheets
exhibit weak AFM order and are not directly suitable for device
applications [27]. One question is how to tune the magnetism in
these organometallic sheets. Recall that in some DMSs, multidecker nanowires, and graphene nanoribbons, carriers play an
important role in tuning the magnetic coupling between magnetic
atoms [3033]. Following this idea, Zhou and Sun [34] carried out
detailed studies showing that the magnetic couplings in poly-CrPc
and poly-FePc sheets will be changed from AFM to FM with high
exchange energies under electron doping. When electrons are
injected to the poly-TMPc sheets, the magnetic coupling can be
changed from AFM to FM with large exchange energies. Based on
Monte-Carlo simulations, it was found that electron doping can
induce stable FM states with TC of 130140 K, while hole doping
will enhance the stability of the AFM states. Such changes in magnetic couplings depend on the balance of AFM superexchange and
FM pd exchange.
Since the interatomic separation between the TM atoms often
determines their magnetic moment and coupling, one would wonder whether the magnetic properties of these systems can be further tuned by external strain. In 2012, Zhou et al. have studied
strain-induced spin crossover in poly-TMPc (TM = Mn, Fe, Co, and
Ni) systems [35]. It was found that the magnetic moment of the
central TM atoms could be increased by 2 lB when a strain was
applied. These poly-TMPc systems, however, show different

Fig. 1. (a) (left) Isosurface at a value of 0.01 e/3 and (right) 2D slice of the spin density (q" q;) for FM 2D MnPc. (b) Band structure and corresponding DOS of FM 2D MnPc.
(c) Projected DOS of the d orbitals on the Mn atom. Symbols: solid for dxy; dash for dyz; dot for dz2; dash-dot for dxz; dash-dot-dot for dx2y2. [Reproduced with permission from
Ref. [27]. Copyright (2011) American Chemical Society].

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020

G. Zhu, Q. Sun / Computational Materials Science xxx (2015) xxxxxx

response to the strain, namely, poly-FePc sheet becomes FM while

poly-MnPc and poly-NiPc sheets become AFM. Poly-CoPc, on the
other hand, remains AFM. These results may be observed in suspended poly-TMPc sheets by using scanning tunneling microscope
(STM) tips to manipulate strain.
In reality, the exposed TM atoms in the poly-TMPc sheets are
very reactive and easily to be poisoned under the air, and then their
magnetic behaviors will be changed. In order to prevent such a
problem, scientists have used ligands to passivate the systems
[36]. To further explore the effect of absorption on magnetism,
Zhou et al. studied the absorption of chlorine atoms on
poly-TMPc sheets [37]. They found that when one Cl is absorbed
on the TM ion, the strong square planar crystal eld is reported
to turn weak in a square pyramidal conguration and the TM is
in the +3 oxidized state, resulting in the magnetic moment of 3,
4, and 5 lB for Cr, Mn, and Fe, respectively, with weak AFM couplings. When another Cl is introduced to the TM on the other side,
it extracts one electron from the Pc framework, making the substrate p-doped. The magnetic coupling is AFM for poly-CrPc2Cl
and poly-FePc2Cl, while it becomes FM for poly-MnPc2Cl. This
study suggests that absorption of chlorine atoms can effectively
modulate the bonding environment and tune the magnetic properties of the poly-TMPc systems. The underlying mechanisms are following: The Cl introduction not only attracts electrons from the
poly-TMPc sheets but also changes the crystal eld of the central
TM atoms from square planar to square pyramidal.
2.2. Polyporphyrin half-metallic sheets
Besides Pc as a molecular unit for 2D organometallic sheet,
another option is Por molecule. In 2013, Zhaos group proposed a
novel 2D periodic organic nanomaterial (2D-polyporphyrin), as
shown in Fig. 2, from rst-principles calculations [38]. They
showed that electron spin-polarization can be achieved in both
metal-free and TM-embedded 2D-polyporphyrins (Fig. 2). In

particular, Cr-polyporphyrin (Cr-PP) has stable FM ordering with

a TC of about 187 K as found by the Monte Carlo simulations based
on the 2D Ising model, which is slightly higher than that in the
Mn-Pc sheet as discussed above. Whats more interesting is that
the FM Cr-PP nanosheet can be tuned to half-metallic by electron
doping.
2.3. Bis-dichiolene based nanosheets
In addition to the Pc- and Por-based square-lattice nanosheets,
the spin ordering in kagome lattices has long been studied in the
search for real materials with a spin-liquid ground state. The recent
synthesis of a nickel bis-dichiolene complex (Ni3C12S12) nanosheet
[39] provided a way for realizing the two-dimensional kagome lattices. A FM kagome spin lattice with S = 3/2 on lattice vertices can
be achieved in an Mn3C12S12 monolayer by substituting Ni with Mn
atoms in nonmagnetic Ni3C12S12 [40]. Monte Carlo simulations on
the basis of the Ising model suggest that it has a TC of about 212 K,
as exhibited in Fig. 3. First-principles calculations show that the FM
Mn3C12S12 monolayer is half metallic with a high carrier mobility
in one spin channel and a band gap of 1.54 eV in another spin channel, which is quite promising for spintronic device applications.
Additionally, a small band gap opens up at the Dirac point of the
kagome bands due to the spinorbital coupling effects, which
may be implementable for achieving a quantum anomalous Hall
effect.
Based on the same experimental study on the Ni3C12S12 sheet
[39], calculations were carried out for Mn3C12N12H12 sheet where
sulfur atoms are replaced by NH groups in Ni3C12S12. The resulting Mn3C12N12H12 sheet was found to exhibit strong FM with a TC
of 450 K [41]. Compared with Mn3C12S12 sheet [40], the enhanced
FM in Mn3C12N12H12 sheet is due to the reduced lattice constant
and the more effective magnetic couplings through pd exchange
interactions. Furthermore, it is also conrmed that the
Mn3C12N12H12 sheet is kinetically and thermally stable, and also
displays half metallicity.
2.4. Other molecules based magnetic sheets

Fig. 2. (a) Spin-resolved band structures of the 2D periodic 2H-PP sheet. Spin-up
and spin-down branches are represented by the solid lines and dashed lines,
respectively. The two arrows indicate the positions of two spin-polarized at bands.
(b) Spin-resolved electron density of states. (c) Isosurfaces of the spin polarized
electron density, Dq = q"  q;. The energy at the Fermi level was set to zero in this
gure. [Reproduced from Ref. [38] with permission from The Royal Society of
Chemistry].

Following the typical route of molecular self-assembly (MSA)

approaches, Bieri et al. have used TM-containing rings to build
the organometallic porous sheets by connecting chemical linking
ligands [42]. Kan et al. have systematically studied the magnetic
and electronic behaviors of the free-standing organometallic porous sheets which are assembled by TM ions and benzene molecules (Fig. 4) [43]. Choosing TM atoms ranging from Ti to Zn, the
authors investigated the magnetism of these nanosheets, and estimated the magnetic critical temperature by the mean eld theory
(MFT). Their results show that V and Mn containing porous sheets
are FM and the estimated TC of V-based sheet using the MFT is
found to be as high as 279 K.
Different ways of MSA from the same molecules may lead to
different 2D structures. Dais group have recently performed comprehensive rst principles calculations to study the electronic and
magnetic properties of 2D Co2C18H12 and Ni2C18H12 lattice structure [44]. Although they are also assembled by TMs and benzene
molecules, the two structures display quite different properties
with the abovementioned one [42,43]. The authors demonstrated
that the low-buckled 2D Co2C18H12 lattice structure is
spin-polarized while the high-buckled 2D Co2C18H12 lattice structure is spin-unpolarized. Transition between the two states will
provide high efciency and high sensitivity for device applications
if we view the separated low-buckled and high-buckled states of
Co2C18H12 as two states of memory devices. More interestingly,
the authors predicated that 2D Ni2C18H12 lattice and low-buckled
2D Co2C18H12 structure possess amazing half-metallic Dirac point

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020

G. Zhu, Q. Sun / Computational Materials Science xxx (2015) xxxxxx

Fig. 3. Orbital resolved electron density of states projected onto (a) Mn, (b) C, and (c) S atoms. Spin-up and spin-down channels are plotted in the top and bottom panels,
respectively. The orbitals contributing mostly to the electron states in proximity of the Fermi level are marked by solid lines combined with solid area for the occupied states.
The energy at the Fermi level was set to zero. (d) Monte Carlo simulations of the average magnetic moment in per unit cell of an Mn3C12S12 monolayer as a function of
temperature. The heat capacity (Cv) is shown in the inset of this gure. [Reproduced from Ref. [40] with permission from The Royal Society of Chemistry].

Fig. 4. (a) Geometric structure of a porous graphene sheet, where dashed lines denote the unit cell, and the right panel displays the enlarged view of the structures enclosed
by the red circles. (b) The calculated band dispersion. (c) The plotted PDOS of different chemical carbon atoms in this porous graphene sheet. [Reproduced from Ref. [43] with
permission from The Royal Society of Chemistry]. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020

G. Zhu, Q. Sun / Computational Materials Science xxx (2015) xxxxxx

Fermi surfaces. This might lead to promising novel applications

and shows an important way to extend the applications of 2D
materials in spintronics devices.
Kan et al. have systematically studied the magnetic characteristics of free-standing 2D porous sheets self-assembled from
dimethylmethylene-bridged triphenylamine (DTPA) molecules
[45]. Since the isolated DTPA molecules are magnetic, carrying a
magnetic moment of 1 lB, magnetic-coupling calculations reveal
that the free-standing 2D DTPA covalent network favors FM coupling. Moreover, electronic structure analysis revealed that the
2D DTPA porous sheet was a half-metal and that the
half-metallicity was robust under obvious external stress.
Based on the recent progress in 2D metal-7,7,8,8-tetracyanoqui
nodimethane (TCNQ) coordination networks assembled from
TCNQ molecules [4650], Ma et al. present a systematic theoretical
study on the structural, electronic, and magnetic properties of the
novel tetragonal TM-based TCNQ molecule coordination single
sheets (referred to as TM@TCNQ, TM = CrCo) [51]. Their results

unveiled that, in TM@TCNQ, two valence electrons would transfer

from one TM atom to TCNQ molecules, making them more stable.
Among these structures, Cr@TCNQ, Mn@TCNQ, and Fe@TCNQ exhibit long-range AFM coupling while Co@TCNQ is paramagnetic.
Such long-range magnetic coupling in the studied systems is
related to the modulation via the TCNQ ligands. Besides, to explain
the magnetic moment qualitatively, the authors proposed a model
on d splitting named 4 + 1 splitting [52]. This work provides theoretical insights leading to a better understanding of these novel
frameworks and probes whether they are promising candidates
for wide applications in MSA nanoelectronics.
2.5. Dimetal-based nanosheets
Unlike conventional organic monolayers that contain isolated
metal atoms, the Mo2Pc-based sheet is reported to be a new
member of the porous organometallic family, which is shown in
Fig. 5. [53] Recent experimental advances by Matsushita et al.

Fig. 5. (a) and (b) Geometric structures of 2D Mo2Pc sheet. In the top view (a) the primitive unit cell is marked by blue dotted line, whereas the rectangular unit cell is marked
by red dotted line. In the side view (b) the buckling is about 1 . (c) Projected densities of states of d orbitals for Mo-1 and Mo-2 atoms in the 2D Mo2Pc sheet. The blue and red
arrows denote spin up and spin down, respectively. The Fermi level is marked by the black dashed line. [Adapted with permission from Ref. [53]. Copyright (2014) American
Chemical Society]. (For interpretation of the references to colour in this gure legend, the reader is referred to the web version of this article.)

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020

G. Zhu, Q. Sun / Computational Materials Science xxx (2015) xxxxxx

[54] have provided us an expanded Pc congener containing a Mo2

dimer and four isoindole moieties. Inspired by this interesting
molecular structure, one wonders whether a 2D Mo2Pc can be constructed and what properties would such a 2D sheet have? The
above experimental foundation has facilitated the design of the
Mo2Pc nanosheet. This dimetal-based material is evinced to possess unique mechanical, magnetic, electronic, and optical properties due to its inherent anisotropy in the structure. The Mo2Pc
sheet is also reported to be a semiconductor with a direct band
gap of 0.93 eV and is AFM coupling (Fig. 5) with each Mo site carrying a magnetic moment of 0.88 lB. The strong anisotropy of 2D
Mo2Pc in elasticity and infrared light absorption are likely to open
new doors for potential applications of dimetal-based porous
organometallic sheets.
Motivated by the recent breakthrough in synthesizing
the 2D [Cu2Br(IN)2]n (IN = isonicotinato) single-layer coordination

polymer, [55] Zhous group systematically investigated the structural, electronic, and magnetic properties of this periodic monolayer [Cu2Br(IN)2]n polymer by means of density functional
theory (DFT) computations [56]. Similar to other class III
mixed-valence polymers [57], single-layer [Cu2Br(IN)2]n polymer
is assembled with pairs of CuCu metal bonds, where each Cu
dimer bridged by bromine is coordinated by four oxygen atoms
from two IN units and two nitrogen atoms from the other two IN
ligands, generating a unique (Cu2Br)3+ unit with mixed Cu(I)
Cu(II) metal bond formulation. Such mixed-valence [58] is of
renewed interest after the discovery of binuclear CuA sites
in cytochrome c oxidase [59] and nitrous oxide reductase [60].
The pristine monolayer [Cu2Br(IN)2]n polymer is found to be
ground-state AFM with a band gap of 0.47 eV. These dimetalbased coordination polymers are new 2D members of a big family
of hybrid inorganicorganic nanomaterials.

Fig. 6. (a) and (b) Geometry of ScPc sheet. (c) and (d) Predicted H2 excess adsorption isotherms for TMPc (TM = Fe, Sc, Ti) at T = 77 K and 298 K. [Adapted from Ref. [61] with
permission from AIP Publishing LLC].

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020

G. Zhu, Q. Sun / Computational Materials Science xxx (2015) xxxxxx

3. Gas adsorption
3.1. Hydrogen storage
Motivated by the recent experimental synthesis technique
where 2D FePc forms a periodic single layer sheet with regularly
spaced Fe atoms, the question concerning hydrogen storage is
raised: Can Fe atoms in FePc trap hydrogen effectively? Are other
TMs better suited for the purpose? In 2011, L et al. [61] have carried out a systematic study of 3d TMs (ScZn) embedded Pc porous
sheets (Fig. 6), using rst-principles theory and grand canonical
Monte Carlo (GCMC) simulation, and found that ScPc can store
4.6 wt% hydrogen at 298 K and 100 bar, as plotted in Fig. 6, which
has fullled the revised DOE target set for 2010, namely 4.5 wt%.
Since Por is lighter in mass than Pc, we can expect the former to
store hydrogen with higher gravimetric density when used to
assemble 2D periodic nanostructures. A systematical study of
Por-based nanosheets on the stability, electronic structures, optical
absorbance and hydrogen adsorption capabilities has been
reported recently [62], where the authors explored the design of
a new sheet with improved hydrogen storage capacity by carrying
out a comprehensive study of the doubly bb-linked 2D Por
nanosheets impregnated with divalent alkaline-earth and
rare-earth metals, as shown in Fig. 7. The sheets are found to be
thermally and mechanically stable and their electronic structure
can be tuned from semiconducting to metallic by embedding different metal atoms, and the nanosheets can absorb near infrared
(NIR) light. Furthermore, the authors also calculated the hydrogen
storage capacities of the MPor (M = Mg, Ca, Sc) at 298 K and

100 bar pressure and nd that the hydrogen gravimetric density

of ScPor nanosheet can reach 6.71 wt% (Fig. 7) which represents
a distinct enhancement as compared to the previously reported
organometallic materials [61,63].
Due to its porous structure and light mass, the recently synthesized triazine-based graphitic carbon nitride (g-C3N4) nanosheet
[6467] has been reported to be a promising material for hydrogen
gas storage [68]. First-principles calculations have been used to
study the hydrogen storage capacity of Li doped g-C3N4 under
ambient thermodynamic conditions. Based on the force eld
parameters derived from quantum chemistry calculations, GCMC
simulations were performed to investigate H2 adsorption isotherms on g-C3N4 sheet. It is reported that the adsorption energy
(Eads) of H2 is 3.48 kcal/mol, and the excess uptake of hydrogen
is about 4.5 wt% at 298 K and 100 bar, showing the potential of
the Li doped g-C3N4 as a hydrogen storage material.
3.2. CO2 capture and its selective adsorption
Carbon-based fossil fuels, which supply over three quarters of
the worlds energy needs, are the main source of greenhouse gas
(CO2) contributing to the climate change [69]. Thus, the carbon
capture and sequestration of CO2 emissions from fossil fuel power
plants are urgently pursued [70].
An ideal material for pre-combustion capture of CO2 should
possess high selectivity and good adsorption capacity for CO2.
Recently, many organometallic porous materials for CO2 capture
have been synthesized in the laboratory, including metal organic
frameworks (MOFs) [7175], covalent organic frameworks (COFs)

Fig. 7. (a)(c) The PES of H2 derived from rst-principles calculations and tting force eld parameters using a cluster model for the substrate. (d) The H2 adsorption
isotherms of the MgPor, CaPor and ScPor nanosheets at 298 K from 0 to 100 bar. [Reproduced with permission from Ref. [62]. Copyright (2015) Hydrogen Energy Publications,
LLC].

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020

G. Zhu, Q. Sun / Computational Materials Science xxx (2015) xxxxxx

[76,77], and zeolitic imidazolate frameworks (ZIFs) [78,79].

However, the interactions between CO2 and MOFs, as well as
COFs and ZIFs, are still too weak, thus showing low CO2 capture
capacity. The excellent performance of the ScPc sheet in hydrogen
storage implies that it may also be a good candidate for CO2 capture [61], which motivates researchers to investigate the CO2 selectivity and capacity of TM embedded in Pc sheets in order to explore
more efcient materials for trapping CO2.
Combining rst-principles calculations and GCMC simulations,
TMPc sheets (TM = Sc, Ti, and Fe) have been studied for
pre-combustion CO2 capture by considering the adsorptions of
H2/CO2 gas mixtures, shown in Fig 8 [80]. It is concluded that
ScPc sheet shows a good selectivity for CO2, and the excess uptake
capacity of single-component CO2 on ScPc sheet at 298 K and
50 bar is found to be 2949 mg/g, larger than that of any other
reported porous materials [73,75,77]. This suggests that the ScPc
sheet can be used for separation of CO2 from H2 better than most
known adsorbents. Furthermore, the interaction mechanisms are
investigated through electrostatic potential and natural bond orbital analyses, showing that the stronger electrostatic interactions as
well as the donation and back donation of electrons between the
TM ions and the CO2 molecules play a key role in the capture
and separation. The present study sheds light on better understanding of novel 2D porous nanomaterials for CO2 capture.
Besides the electronic reasons on gas selectivity, the geometric
structures of 2D monolayers, such as pore size, are often tailored to
make nanosheets function as lter membranes. It has been demonstrated by Zhous group in 2010, that 2D polyphenylene is a typical
semiconductor with the porous structure endowing it remarkably

Fig. 8. (a) Geometry of TMPc sheet, (b) a cluster model used in geometry
optimizations, and (c) cluster models for potential energy curve scanning. (d)
Predicted CO2 excess adsorption isotherms for different TMPc (TM = Sc, Ti, Fe)
sheets at T = 298 K. The experimentally measured CO2 isotherms of MOF-200 and
MOF-210 at T = 298 K are also represented for comparison. [Adapted from Ref. [80]
with permission from AIP Publishing LLC].

high selectivity for H2 permeability relative to CO2, CO and CH4

[81]. This experimentally available porous nanosheet is expected
to nd applications in a novel energy society, because the kinetic
diameter of H2 approaches the width of the pore in polyphenylene,
while those of CO2, CO and CH4 are much larger, illustrating the
fact that H2 can pass though the pore with a moderate barrier,
while the diffusion of CO2, CO and CH4, is difcult.

4. Catalysis
4.1. CO oxidation
The catalytic oxidation of CO has attracted extensive attentions
for several decades due to its crucial role in exhaust treatment of
the post-combustion process for automobiles [82] as well as in
alleviating the poisoning effect on methanol fuel cell electrocatalysts [83]. Many noble metals such as Pt, Rh, Pd and Au nanoparticles have been studied for catalyzing CO oxidation [8486].
However, these catalysts are generally costly, and usually encounter CO poisoning problems, impeding durable operation.
According to previous studies [8792], Deng et al. have summarized the following conditions for an eligible metalsubstrate CO
oxidation catalyst: (1) O2 should have a lower Eads with the substrate than CO; (2) the Eads of CO2 with the substrate should be
higher than 0.52 eV; (3) the Eads of O2 or CO with the substrate
should be lower than the negative values of the corresponding
energy barrier of the catalytic reactions; and (4) the overall energy
barrier for the catalytic reactions should be less than 1 eV. Based on
the experimental synthesis of FePc, they investigated the CoPc and
found that it satised all the above four conditions and exhibited a
good catalytic activity for CO oxidation at room temperature [93].
By exploring the two well-established mechanisms for CO
oxidation with O2, namely, the LangmuirHinshelwood (LH)
and the EleyRideal (ER) mechanisms, it is found that the rst step
of CO oxidation catalyzed by CoPc is the LH mechanism
(CO + O2 ? CO2 + O) with energy barrier as low as 0.65 eV. The second step proceeds via both ER and LH mechanisms (CO + O ? CO2)
with small energy barriers of 0.10 and 0.12 eV, respectively. The
electronic resonance among Co-3d, CO-2p, and O2-2p orbitals is
explained to be responsible for the high activity of CoPc. These
results have signicant implications for a novel avenue to fabricate
organometallic sheet nanocatalysts for CO oxidation with low cost
and high activity.
A study on the superior catalytic CO oxidation capacity of a CrPc
porous sheet was also presented and the authors claimed a lower
LH energy barrier [94]. It is identied that the CO oxidation reaction catalyzed by the CrPc sheet proceeds rst via LH mechanism
and then via ER mechanism [95] with the highest energy barrier
of 0.55 eV, much lower than its Fe-containing counterparts, suggesting this reaction is viable to proceed at low temperature, as
explained in Fig. 9. Compared with the noble metal based catalysts
or graphene supported ones, this system eliminates the poisoning
effect and clustering problem, reveals comparable reaction energy
barrier for low-temperature oxidation, and might lower the cost
for large-scale catalytic CO oxidation in industry.
Preparation of single atom catalysts in experiments has been
largely hampered by the high mobility of noble-metal atoms on
substrate and their easiness to sinter under realistic reaction conditions [96,97]. Thus, tremendous efforts in this area have been
paid to nd proper substrates that can effectively anchor the catalyst atoms without degrading their activities [98]. Chen et al. have
performed calculations to study the possibility of monolayer polymeric TMPcs (TM = ScZn) as the substrate in an effect to obtain
evenly distributed single Pt atom catalyst [99]. They found that
TiPc is the most appropriate compound by virtue of high binding

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020

G. Zhu, Q. Sun / Computational Materials Science xxx (2015) xxxxxx

Fig. 9. Electron energy levels and orbital charge density isosurfaces (isovalue 0.02 coulomb/m3) in 2D CrPc (a) and 2D FePc (b) sheets. The arrows denote the spin-up and
spin-down. The solid lines correspond to the lled levels and the dotted lines to the unlled levels. The doubly degenerate levels are marked with upper-case letter D. The
lines in the rectangles denote levels interacting with O2 2p levels and those in the ovals denote levels interacting with CO 5s and 2p levels. [Reproduced with permission
from Ref. [94]. Copyright (2014) Nature Publishing Group].

Fig. 10. Proposed reaction pathways for CO oxidation on Pt/TiPc. The congurations of each step are presented. The inset in the cycle shows the calculated energy prole.
[Reproduced with permission from Ref. [99]. Copyright (2014) American Chemical Society].

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020

10

G. Zhu, Q. Sun / Computational Materials Science xxx (2015) xxxxxx

strength of Pt/TiPc system and high diffusion energy barrier of Pt

catalyst atom, which jointly prevent the formation of Pt clusters.
CO oxidation on Pt/TiPc presents some unexpected behaviors,
where a presorbed CO molecule could activate the catalyst and
reduce the reaction barrier. The reaction pathways for CO oxidation are presented in Fig. 10. These results demonstrate that 2D
TMPcs provide a unique platform to fabricate regularly and separately distributed single atom catalyst with high activity.
4.2. Oxygen reduction reaction
For fuel cells, the oxygen reduction reaction (ORR) is a rather
critical process in generating electricity. But the rate of ORR is usually much slower than that of the anode reaction, which is one of
the key factors limiting the performance of fuel cells [100,101].
To develop low-cost non-Pt catalysts for ORR, Wang et al. have systematically investigated the potential of applying the experimentally available 2D FePc monolayer with precisely-controlled
distribution of Fe atoms as a catalyst of ORR [102]. The adsorption
of O2 molecule and the ORR processed on 2D FePc monolayer under
an acidic environment were explored by means of spin-polarized
DFT computations. The limiting potential for ORR on FePc monolayer is found be as high as 0.68 V. Compared to FePc molecules,
FePc monolayer has a similar (or even better) catalytic
performance, but an much enhanced stability in acid condition,
indicating that FePc monolayer is a quite promising single-atomcatalyst with high efciency for ORR in fuel cells.
5. Conclusions and perspectives
This article reviews the recent advances made by Chinese
research groups in computational studies of organometallic sheets
for tunable magnetism, hydrogen storage capabilities, CO2 captures, selective gas adsorption and catalysis. Compared with other
2D sheets, the outstanding advantages of organometallic sheets are
following: metal atoms are regularly distributed and well exposed;
it is very exible in synthesis by using different metal atoms and
different molecular units for diverse organometallic sheets with
novel functionalities.
More attentions need to be desirably paid in the future to the
following aspects: (1) Hybrid organometallic sheets. The existing
2D organometallic sheets are mainly based on mono-metal units,
which have the following shortcomings: First, the distance
between metal ions are usually quite large and the concentration
of metal ions is not high, so it is not very efcient for catalysis
and gas molecule adsorptions; Second, due to the large distance
the coupling between metal ions is weak, this is not good for the
system to exhibit the cooperative effects like long-ranged FM order
with high TC and bimetallic catalytic effect; Third, there is no
strong planar anisotropy, so the relevant physical and chemical
processes are not selective and direction-dependent. These drawbacks can be overcome by hybridizing different structural units
with different metals, which provides more variables to diversify
the structures for tuning the properties and functionalities. (2)
Going from 2D to 3D. Using 2D organometallic sheets as building
blocks to rationally design 3D porous materials, which would be
very different from the widely studied 3D MOFs where the metal
sites are not well exposed showing poor performance in gas
adsorption and catalysis. We hope that this review will stimulate
further experimental effort in this direction.
Acknowledgements
This work is partially supported by grants from the National
Natural Science Foundation of China (NSFC-21173007 and

11274023), and from the National Grand Fundamental Research

973 Program of China (2012CB921404). The authors thank Yawei
Li for his critical reading of the manuscript.
References
[1] K.S. Novoselov, A.K. Geim, S. Morozov, D. Jiang, Y. Zhang, S. Dubonos, I.
Grigorieva, A. Firsov, Science 306 (2004) 666669.
[2] W.-Q. Han, L. Wu, Y. Zhu, K. Watanabe, T. Taniguchi, Appl. Phys. Lett. 93
(2008) 223103.
[3] P. Vogt, P. De Padova, C. Quaresima, J. Avila, E. Frantzeskakis, M.C. Asensio, A.
Resta, B. Ealet, G. Le Lay, Phys. Rev. Lett. 108 (2012) 155501.
[4] J.N. Coleman, M. Lotya, A. ONeill, S.D. Bergin, P.J. King, U. Khan, K. Young, A.
Gaucher, S. De, R.J. Smith, Science 331 (2011) 568571.
[5] Y. Jing, Z. Zhou, C.R. Cabrera, Z. Chen, J. Mater. Chem. A 2 (2014) 12104
12122.
[6] L. Ponomarenko, F. Schedin, M. Katsnelson, R. Yang, E. Hill, K. Novoselov, A.
Geim, Science 320 (2008) 356358.
[7] X. Wang, X. Li, L. Zhang, Y. Yoon, P.K. Weber, H. Wang, J. Guo, H. Dai, Science
324 (2009) 768771.
[8] X. Wang, L. Zhi, K. Mllen, Nano Lett. 8 (2008) 323327.
[9] G. Eda, H. Yamaguchi, D. Voiry, T. Fujita, M. Chen, M. Chhowalla, Nano Lett. 11
(2011) 51115116.
[10] K.F. Mak, K. He, J. Shan, T.F. Heinz, Nat. Nanotechnol. 7 (2012) 494498.
[11] T.P. Hardcastle, R. Zan, C.R. Seabourne, R.M.D. Brydson, U. Bangert, Q.M.
Ramasse, K.S. Novoselov, A.J. Scott, Phys. Rev. B, Condens. Matter 87 (2013)
24502458.
[12] B. J, Z. X, J. S, H. Y, D. X, Nat. Nanotechnol. 5 (2010) 190194.
[13] J. Zhou, K. Lv, Q. Wang, X.S. Chen, Q. Sun, P. Jena, J. Chem. Phys. 134 (2011)
174701174705.
[14] A. Krasheninnikov, P. Lehtinen, A. Foster, P. Pyykk, R. Nieminen, Phys. Rev.
Lett. 102 (2009) 126807.
[15] Y. Zhou, Z. Wang, P. Yang, X. Zu, L. Yang, X. Sun, F. Gao, ACS Nano 6 (2012)
97279736.
[16] M. Kan, J. Zhou, Q. Sun, Y. Kawazoe, P. Jena, J. Phys. Chem. Lett. 4 (2013) 3382
3386.
[17] M. Kan, J.Y. Wang, X.W. Li, S.H. Zhang, Y.W. Li, Q. Sun, P. Jena, Y. Kawazoe, J.
Phys. Chem. C 118 (2014) 15151522.
[18] M. Abel, S. Clair, O. Ourdjini, M. Mossoyan, L. Porte, J. Am. Chem. Soc. 133
(2010) 12031205.
[19] A. Sperl, J. Krger, R. Berndt, J. Am. Chem. Soc. 133 (2011) 1100711009.
[20] A. Tsuda, A. Osuka, Adv. Mater. 14 (2002) 7579.
[21] T.K. Chandrashekar, S. Venkatraman, Acc. Chem. Res. 36 (2003) 676691.
[22] E. Rose, A. Lecas, M. Quelquejeu, A. Kossanyi, B. Boitrel, Coordin. Chem. Rev.
178 (1998) 14071431.
[23] M.O. Senge, J. Richter, J. Porphyr. Phthal. 8 (2004) 934953.
[24] C. Zou, C.-D. Wu, Dalton Trans. 41 (2012) 38793888.
[25] X. Feng, L. Chen, Y. Dong, D. Jiang, Chem. Commun. 47 (2011) 19791981.
[26] S. Zhang, Y. Li, T. Zhao, Q. Wang, Nat. Sci. Rep. 4 (2014). 8181.
[27] J. Zhou, Q. Sun, J. Am. Chem. Soc. 133 (2011) 1511315119.
[28] A. Wolska, K. Lawniczak-Jablonska, M. Klepka, M. Walczak, A. Misiuk, Phys.
Rev. B 75 (2007) 113201.
[29] L. Zeng, E. Helgren, M. Rahimi, F. Hellman, R. Islam, B. Wilkens, R. Culbertson,
D.J. Smith, Phys. Rev. B 77 (2008) 073306.
[30] K. Sato, P. Dederichs, H. Katayama-Yoshida, J. Kudrnovsky, J. Phys. Condens.
Matter 16 (2004) S5491.
[31] Z. Zhang, X. Wu, W. Guo, X.C. Zeng, J. Am. Chem. Soc. 132 (2010) 10215
10217.
[32] K. Sawada, F. Ishii, M. Saito, S. Okada, T. Kawai, Nano Lett. 9 (2009) 269272.
[33] H. Raebiger, S. Lany, A. Zunger, Phys. Rev. Lett. 101 (2008) 027203.
[34] J. Zhou, Q. Sun, Nanoscale 6 (2013) 328333.
[35] J. Zhou, Q. Wang, Q. Sun, Y. Kawazoe, P. Jena, J. Phys. Chem. Lett. 3 (2012)
31093114.
[36] S.M. Palmer, J.L. Stanton, N.K. Jaggi, B.M. Hoffman, J.A. Ibers, L.H. Schwartz,
Inorg. Chem. 24 (1985) 20402046.
[37] J. Zhou, Q. Sun, J. Chem. Phys. 138 (2013) 204706204707.
[38] J. Tan, W. Li, X. He, M. Zhao, RSC Adv. 3 (2013) 70167022.
[39] T. Kambe, R. Sakamoto, K. Hoshiko, K. Takada, M. Miyachi, J.-H. Ryu, S. Sasaki,
J. Kim, K. Nakazato, M. Takata, H. Nishihara, J. Am. Chem. Soc. 135 (2013)
24622465.
[40] M. Zhao, A. Wang, X. Zhang, Nanoscale 5 (2013) 1040410408.
[41] J. Liu, Q. Sun, ChemPhysChem (2014).
[42] M. Bieri, M. Treier, J. Cai, K. At-Mansour, P. Rufeux, O. Grning, P. Grning,
M. Kastler, R. Rieger, X. Feng, Chem. Commun. (2009) 69196921.
[43] E. Kan, X. Wu, C. Lee, J.H. Shim, R. Lu, C. Xiao, K. Deng, Nanoscale 4 (2012)
53045307.
[44] Y. Ma, Y. Dai, X. Li, Q. Sun, B. Huang, Carbon 73 (2014) 382388.
[45] E. Kan, W. Hu, C. Xiao, R. Lu, K. Deng, J. Yang, H. Su, J. Am. Chem. Soc. 134
(2012) 57185721.
[46] M. Sorai, M. Nakano, Y. Miyazaki, Chem. Rev. 106 (2006) 9761031.
[47] N. Abdurakhmanova, T.-C. Tseng, A. Langner, C. Kley, V. Sessi, S. Stepanow, K.
Kern, Phys. Rev. Lett. 110 (2013) 027202.
[48] M.N. Faraggi, N. Jiang, N. Gonzalez-Lakunza, A. Langner, S. Stepanow, K. Kern,
A. Arnau, J. Phys. Chem. C 116 (2012) 2455824565.

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020

G. Zhu, Q. Sun / Computational Materials Science xxx (2015) xxxxxx

[49] T.-C. Tseng, N. Abdurakhmanova, S. Stepanow, K. Kern, J. Phys. Chem. C 115
(2011) 1021110217.
[50] N. Abdurakhmanova, A. Floris, T.-C. Tseng, A. Comisso, S. Stepanow, A. De Vita,
K. Kern, Nat. Commun. 3 (2012) 940.
[51] Y. Ma, Y. Dai, W. Wei, L. Yu, B. Huang, J. Phys. Chem. A 117 (2013) 51715177.
[52] W.J. Cho, Y. Cho, S.K. Min, W.Y. Kim, K.S. Kim, J. Am. Chem. Soc. 133 (2011)
93649369.
[53] G. Zhu, M. Kan, Q. Sun, P. Jena, The Journal of Physical Chemistry A 118 (2014)
304307.
[54] O. Matsushita, V.M. Derkacheva, A. Muranaka, S. Shimizu, M. Uchiyama, E.A.
Lukyanets, N. Kobayashi, J. Am. Chem. Soc. 134 (2012) 34113418.
[55] P. Amo-Ochoa, L. Welte, R. Gonzlez-Prieto, P.J.S. Miguel, C.J. Gmez-Garca, E.
Mateo-Mart, S. Delgado, J. Gmez-Herrero, F. Zamora, Chem. Commun. 46
(2010) 32623264.
[56] Q. Tang, Z. Zhou, Z. Chen, J. Phys. Chem. C 116 (2012) 41194125.
[57] J.Y. Lu, K.A. Runnels, Inorg. Chem. Commun. 4 (2001) 678681.
[58] C. Harding, V. McKee, J. Nelson, J. Am. Chem. Soc. 113 (1991) 96849685.
[59] S. Iwata, C. Ostermeier, B. Ludwig, H. Michel, Nature 376 (1995) 660668.
[60] P.M. Kroneck, W.A. Antholine, J. Riester, W.G. Zumft, FEBS Lett. 242 (1988)
7074.
[61] K. Lu, J. Zhou, L. Zhou, Q. Wang, Q. Sun, P. Jena, Appl. Phys. Lett. 99 (2011).
163104163103.
[62] G. Zhu, Q. Sun, Y. Kawazoe, P. Jena, Int. J. Hydrogen Energy 40 (2015) 3689
3696.
[63] G. Zhu, Y. Li, K. L, Q. Sun, ChemPhysChem 15 (2014) 126131.
[64] D.M. Teter, R.J. Hemley, Science 271 (1996) 5355.
[65] J. Zhao, C. Fan, Physica B: Condens. Matter 403 (2008) 19561959.
[66] C. Niu, Y.Z. Lu, C.M. Lieber, Science 261 (1993) 334337.
[67] Y. Km, C. Ml, H. Ee, H. Wl, L. Ay, Phys. Rev. B Condens. Matter 49 (1994) 5034
5037.
[68] G. Zhu, K. L, Q. Sun, Y. Kawazoe, P. Jena, Comp. Mater. Sci. 81 (2014) 275
279.
[69] L. Reijnders, M. Huijbregts, J. Cleaner Prod. 16 (2008) 477482.
[70] J.-R. Li, Y. Ma, M.C. McCarthy, J. Sculley, J. Yu, H.-K. Jeong, P.B. Balbuena, H.-C.
Zhou, Coord. Chem. Rev. 255 (2011) 17911823.
[71] H. Li, M. Eddaoudi, T.L. Groy, O.M. Yaghi, J. Am. Chem. Soc. 120 (1998) 8571
8572.
[72] H. Li, M. Eddaoudi, M. OKeeffe, O.M. Yaghi, Nature 402 (1999) 276279.
[73] A.R. Millward, O.M. Yaghi, J. Am. Chem. Soc. 127 (2005) 1799817999.
[74] D. Britt, H. Furukawa, B. Wang, T.G. Glover, O.M. Yaghi, Proc. Natl. Acad. Sci.
106 (2009) 2063720640.

11

[75] H. Furukawa, N. Ko, Y.B. Go, N. Aratani, S.B. Choi, E. Choi, A.. Yazaydin, R.Q.
Snurr, M. OKeeffe, J. Kim, O.M. Yaghi, Science 329 (2010) 424428.
[76] H.M. El-Kaderi, J.R. Hunt, J.L. Mendoza-Corts, A.P. Ct, R.E. Taylor, M.
OKeeffe, O.M. Yaghi, Science 316 (2007) 268272.
[77] H. Furukawa, O.M. Yaghi, J. Am. Chem. Soc. 131 (2009) 88758883.
[78] B. Wang, A.P. Cote, H. Furukawa, Nature 453 (2008) 207211.
[79] R. Banerjee, A. Phan, B. Wang, C. Knobler, H. Furukawa, M. OKeeffe, O.M.
Yaghi, Science 319 (2008) 939943.
[80] K. Lu, J. Zhou, L. Zhou, X.S. Chen, S.H. Chan, Q. Sun, J. Chem. Phys. 136 (2012)
234703234707.
[81] Y. Li, Z. Zhou, P. Shen, Z. Chen, Chem. Commun. 46 (2010) 36723674.
[82] J.T. Kummer, Prog. Energy Combust. 6 (1980) 177199.
[83] J. Greeley, M. Mavrikakis, J. Am. Chem. Soc. 124 (2002) 71937201.
[84] A.D. Allian, K. Takanabe, K.L. Fujdala, X. Hao, T.J. Truex, J. Cai, C. Buda, M.
Neurock, E. Iglesia, J. Am. Chem. Soc. 133 (2011) 44984517.
[85] M.S. Chen, Y. Cai, Z. Yan, K.K. Gath, S. Axnanda, D.W. Goodman, Surf. Sci. 601
(2007) 53265331.
[86] N. Lopez, J.K. Nrskov, J. Am. Chem. Soc. 124 (2002) 1126211263.
[87] P. Zhao, Y. Su, Y. Zhang, S.-J. Li, G. Chen, Chem. Phys. Lett. 515 (2011) 159
162.
[88] Y.-H. Lu, M. Zhou, C. Zhang, Y.-P. Feng, J. Phys. Chem. C 113 (2009) 20156
20160.
[89] Y. Li, Z. Zhou, G. Yu, W. Chen, Z. Chen, J. Phys. Chem. C 114 (2010) 62506254.
[90] E.H. Song, Z. Wen, Q. Jiang, J. Phys. Chem. C 115 (2011) 36783683.
[91] Z.-P. Liu, P. Hu, A. Alavi, J. Am. Chem. Soc. 124 (2002) 1477014779.
[92] A. Demessence, D.M. DAlessandro, M.L. Foo, J.R. Long, J. Am. Chem. Soc. 131
(2009) 87848786.
[93] Q. Deng, L. Zhao, X. Gao, M. Zhang, Y. Luo, Y. Zhao, Small 9 (2013) 35063513.
[94] Y. Li, Q. Sun, Sci. Rep. 4 (2014).
[95] W. An, Y. Pei, X.C. Zeng, Nano Lett. 8 (2007) 195202.
[96] B. Qiao, A. Wang, X. Yang, L.F. Allard, Z. Jiang, Y. Cui, J. Liu, J. Li, T. Zhang, Nat.
Chem. 3 (2011) 634641.
[97] A. Uzun, V. Ortalan, Y. Hao, N.D. Browning, B.C. Gates, Acs Nano 3 (2009)
36913695.
[98] X.-F. Yang, A. Wang, B. Qiao, J. Li, J. Liu, T. Zhang, Acc. Chem. Res. 46 (2013)
17401748.
[99] X.F. Chen, J.M. Yan, Q. Jiang, J. Phys. Chem. C 118 (2014) 21222128.
[100] B.C.H. Steele, A. Heinzel, Nature 414 (2001) 345352.
[101] M.K. Debe, Nature 486 (2012) 4351.
[102] Y. Wang, H. Yuan, Y. Li, Z. Chen, Nanoscale (2015).

Please cite this article in press as: G. Zhu, Q. Sun, Comput. Mater. Sci. (2015), http://dx.doi.org/10.1016/j.commatsci.2015.07.020