Nanoscale

View Article Online

PAPER View Journal | View Issue

Functionalized MXenes as effective polyselenide

immobilizers for lithium–selenium batteries:
Published on 15 June 2020. Downloaded by FAC DE QUIMICA on 2/15/2022 12:41:37 AM.

Cite this: Nanoscale, 2020, 12, 14087

a density functional theory (DFT) study†
Rahul Jayan and Md Mahbubul Islam *

Received 22nd March 2020,
Accepted 15th June 2020
DOI: 10.1039/d0nr02296a
rsc.li/nanoscale
The practical applications of lithium selenium (Li–Se) batteries are impeded primarily due to the dis-
solution and migration of higher-order polyselenides (Li2Sen) into the electrolyte (known as the shuttle
effect) and inactive deposition of lower-order polyselenides. The high electrical conductivity and
mechanical strength of MXenes make them a suitable candidate to provide adequate anchoring to
prevent polyselenide dissolution and improved electrochemical performance. Herein, we used density
functional theory (DFT) calculations to understand the binding mechanism of Li2Sen on graphene and
surface-functionalized Ti3C2 MXenes. We used graphene as a reference material to assess Li2Sen binding
strengths on functionalized Ti3C2X2 (where X = S, O, F, and Cl). We observed that Ti3C2S2 and Ti3C2O2
exhibit superior anchoring behavior compared to graphene, Ti3C2F2, and Ti3C2Cl2. The calculated Li2Sen
adsorption strengths, provided by S- and O-terminated Ti3C2, are greater than those of the commonly
used ether-based electrolyte, which is a requisite for effective suppression of Li2Sen shuttling. Ti3C2X2 and
graphene with adsorbed Li2Sen retain their structural integrity without chemical decomposition. Density
of states (DOS) analysis demonstrates that the conductive behavior of Ti3C2X2 is preserved even after
Li2Sen adsorption, which can provide electronic pathways to stimulate the redox electrochemistry of
Li2Sen. Overall, our unprecedented simulation results reveal superior anchoring behavior of Ti3C2S2 and
Ti3C2O2 for Li2Sen adsorption, and this developed understanding can be leveraged for designing carbon-
free Ti3C2 MXene-based selenium cathode materials to boost the electrochemical performance of Li–Se
batteries.

1. Introduction compared to commercial LIBs, and they are cost effective,

As lithium ion batteries (LIBs) are approaching their theore-
tical limit, alternative battery technologies are being developed
for large-scale applications such as electric vehicles (EVs). The
implementation of rechargeable batteries in EVs requires high
energy density, long cycle life, low cost, and high efficiency.1
The wide abundance and astounding energy density of chalco-
gen group elements, such as O, S, and Se, make them suitable
candidates for future energy storage applications. However,
owing to electrolyte decomposition and the need for excess Li
in the anode, the commercialization of Li–O2 batteries is far
from becoming a reality.2 Rechargeable lithium sulfur (Li–S)
batteries can fulfil their purpose because of their high theore-
tical energy density (2600 W h kg−1) and capacity (1673 mA h
g−1) values, which are approximately five times higher when
Department of Mechanical Engineering, Wayne State University, MI 48202, USA.
E-mail: gy5553@wayne.edu; Tel: +(313) 577-3885
† Electronic supplementary information (ESI) available. See DOI: 10.1039/
d0nr02296a
widely abundant, and low in toxicity.3,4 However, the practical
realization of Li–S batteries is thwarted due to various chal-
lenges such as the dissolution of intermediate polysulfides
into the electrolyte, commonly known as the shuttle effect,
and the low utilization of the active materials due to the for-
mation of insoluble polysulfides.3,5 Although strategies such
as introducing conductive porous matrices,6 additives7 and
suitable electrolytes8 have been developed for Li–S batteries,
their practical application is still restricted. This warrants the
search for new Li–S battery cathode materials that can deliver
high energy density and specific capacity to meet the require-
ments of the next generation energy landscape.
Rechargeable lithium selenium (Li–Se) batteries are analo-
gous to their Li–S counterparts in terms of chemical pro-
perties, as selenium reacts with lithium ions to form lithium
polyselenides (Li2Sen). Although the gravimetric capacity of Se
(678 mA h g−1) is low when compared to S (1672 mA h g−1), its
high volumetric capacity (3253 mA h cm−3) is comparable to
that of S (3467 mA h cm−3).9,10 Moreover, the electronic con-
ductivity of Se (1 × 10−3 S m−1) is at least twenty orders of mag-

This journal is © The Royal Society of Chemistry 2020 Nanoscale, 2020, 12, 14087–14095 | 14087

Paper
nitude higher than that of S (5 × 10−28 S m−1), leading to
better electrochemical activity of Se.8 Therefore, the Li–Se
battery is a prospective candidate to meet the requirements of
higher volumetric energy density for large-scale applications,
and can serve as an effective alternative to Li–S batteries.
However, the dissolution of soluble intermediate Li2Sen
species into the electrolyte results in low battery capacity and
limits the cycle performance of Li–Se batteries.11,12 Carbonate-
based electrolytes have been reported to suppress polyselenide
shuttling due to a single step conversion to Li2Se; however, the
Published on 15 June 2020. Downloaded by FAC DE QUIMICA on 2/15/2022 12:41:37 AM.
performance improvement is quite limited.13 The nucleophilic
nature of Se can cause reactions with carbonate-based sol-
vents, thus detrimentally affecting the battery capacity and
cyclability.14 Furthermore, the large volumetric change of Se
cathodes during charge/discharge cycles results in fracture
and pulverization of the Se active materials.13
Se confinement into carbon mesopores has been reported
to provide the space required to accommodate volume changes
and suppress intermediate polyselenide shuttling during
lithiation/delithiation cycles.15–19 Liu and coworkers incorpor-
ated Se into the microporous carbon nanofibers via vacuum
and heat treatment.20 This process leads to a better anchoring
and homogenous distribution of the selenides on the carbon
matrix.20 Abouimrane et al.9 studied Se and Se–S composite
cathode materials and reported that the high utilization of
bulk Se (ca. 45 wt%) active material upon extended cycling
cannot be achieved with a S bulk cathode, indicating the
better activity and lesser shuttling of Se. A stable capacity of
415 mA h g−1 at 0.2 C for 100 cycles was realized through
homogeneously distributed Se on carbonized polypyrole and
KOH-activated interconnected porous carbon nanofibers.21
CoSe2-porous carbon composites,12 graphene–selenium hybrid
microballs,22 hollow double-shell Se@CNx nanobelts,11 Ag2Se-
coated Se encapsulates, and various metal organic frameworks
have been reported to alleviate the dissolution of polysele-
nides. However, none of these materials have entirely elimi-
nated the dissolution of polyselenides, which is required to
achieve the target performance of Li–Se batteries. Se incorpor-
ated into 3D interconnected porous carbon nanofibers has
been investigated as a possible means to suppress active
material dissolution via uniform Se distribution onto confined
mesopores, which readily act as a cushion for volume expan-
sion and enhance the electrochemical performance.23
Meanwhile, polyselenide migration can be suppressed by con-
straining Se into microporous N-doped carbon.24 The porous
nature of nitrogen-doped carbon structures has been reported
to provide improved Se confinement within the host and to
promote ion transport.25,26 The hierarchical architecture devel-
oped by the fusion of graphene and 3D porous carbon nano-
particles has also been probed to improve electronic and ionic
conductivities and suppress the dissolution of polyselenides.27
He et al.28 introduced a 3D graphene–CNT/Se cathode and
found that the conductive network provided organized chan-
nels for Li diffusion and electron transfer. Furthermore, the
hierarchical structure prevented polyselenide dissolution into
electrolyte. Although selenium confinement in a carbon matrix
14088 | Nanoscale, 2020, 12, 14087–14095
View Article Online
Nanoscale
can reduce migration, the optimization of micropore size in
the carbon structure is a challenging task. As the pore size of
the carbon matrix depends on the overall architecture, a
minimal size could lead to difficulties in accommodating the
electroactive material (Se) and diffusion of Li ions.29
Larger micropores can promote the dissolution of Se upon
charging and drastically affect the electrochemical perform-
ance,13 which in turn prompts further research efforts to inves-
tigate alternate anchoring materials (AMs) for the Li–Se
system.
Researchers have sought various novel materials and strat-
egies to address polyselenide shuttling during the charge/dis-
charge process. Recently, MXenes, a class of 2D materials,
have attracted special attention because of their multiple
attractive features including high electronic conductivity, high
surface area, and catalytically active sites for facilitating Li2Sen
conversion reactions.30 However, bare MXenes cannot be used
directly in the chalcogenide system due to the high reactivity
of polychalcogenides towards transition metals, which causes
the chemical decomposition of polychalcogenides.31,32 Thus,
surface functionalization is required, and commonly, chalco-
genide and halogen groups are employed.31,32,33–36 MXenes
have exhibited promising performances in Li–S batteries.
Wang et al.37 studied the shuttling effect of polysulfides in Li–
S batteries using density functional theory (DFT) calculations
with anchoring materials such as bare V2C, V2CO2, and V2CS2.
They reported that sulfur-functionalized MXenes exhibit ade-
quate thermal and chemical stability and polysulfide binding
strength. The study claimed that the material lowers the
barrier for Li2S decomposition and lithium diffusion, and
retains metallic properties even after the adsorption of polysul-
fides.37 Furthermore, Ti3C2 MXenes functionalized with N, O,
F, S, and Cl have been investigated as cathode materials for Li–
S batteries, and surface functionalization with O and S was
found to facilitate strong binding behavior with polysulfides
and superior catalytic activity for the conversion of polysul-
fides.30 However, the role of MXenes in the context of polysele-
nide retention has not been studied as a plausible route to
improve the capacity and cycle performance of Li–Se batteries.
The excellent electronic properties of MXenes and their
promising performance in Li–S batteries can be leveraged for
designing MXene-based Se cathode materials. To the best of
our knowledge, no studies have been conducted to elucidate
polyselenide binding characteristics and mechanisms on bare
and functionalized MXenes. Herein, we performed first-prin-
ciples-based DFT calculations to understand the structural
stability, adsorption behavior, and electronic properties of
MXenes and graphene. We used graphene as a reference
material to compare the anchoring performance of MXenes.
We started with bare Ti3C2 MXenes to investigate polyselenide
binding behavior and found that bare MXenes cause strong
binding and decomposition of polyselenides. This observation
motivated us to incorporate four different surface functionali-
zations and extensively examine their role on the underlying
characteristics and mechanisms of polyselenide anchoring on
MXenes as host materials for the cathodes of Li–Se batteries.
This journal is © The Royal Society of Chemistry 2020
 View Article Online

Nanoscale Paper

2. Calculation methods
All first-principles calculations were executed using the Vienna
ab initio simulation package (VASP).38 We used the Projector
Augmented Wave (PAW) pseudopotential to investigate the
electron-ion interaction and the Perdew–Bruke–Ernzerhof
(PBE) functional with the Generalized Gradient
Approximations (GGA) to describe the electron–electron
exchange correlations.39 The kinetic energy cut off was selected
as 520 eV for the plane-wave basis calculations. A vacuum
Published on 15 June 2020. Downloaded by FAC DE QUIMICA on 2/15/2022 12:41:37 AM.

space of 30 Å was added to avoid interactions across the peri-
odic boundary. We accounted for van der Waals interactions
using the DFT-D3 method to provide better accuracy in evalu-
ating the binding strength of polyselenides with MXenes.40,41
The conjugate gradient method was employed for geometry
optimization, and the convergence thresholds of energy and
force were fixed at 10−4 eV and 0.025 eV Å−1, respectively. The
calculations were performed with a 4 × 4 supercell of Ti3C2.
The Brillouin zone sampling for the geometric relaxations and
the electronic structure calculations were implemented using 5
× 5 × 1 and 11 × 11 × 1 k-point meshes generated by the
Monkhorst–Pack grid scheme, respectively. Bader charge ana-
lysis42 was performed to study the charge transfer between the
AMs and Li2Sen, and the charge density difference was calcu-
lated using the equation:
ρb ¼ ρadsorbed state  ðρadsorbent þ ρAM Þ
where ρadsorbed state, ρadsorbent, and ρAM refer to the charge
transfer of Li2Sen adsorbed on AMs, isolated Li2Sen and AMs,
respectively. The charge density difference data were visualized
using VESTA code.43 The electronic behavior of bare and
surface-functionalized MXenes was probed using density of
states (DOS) analysis.
Fig. 1 (a) The optimized geometric structures of Li2Sen; (b) top view of
bare Ti3C2; (c) total density of states (DOS) of Ti3C2. Color codes: green
= selenium; purple = lithium; blue = titanium; gray = carbon. The partial
DOS of Ti3C2 is provided in Fig. S1 of the ESI.† The Fermi level is indi-
cated with a vertical line.
higher- to lower-order polyselenides, and the behavior can be
ascribed to the stronger Li–Se covalent interactions.
3.2. Graphene
Carbon-based selenium composite materials are widely used in Li–
Se batteries.16,23,24,27 As a representative model carbon material, we
considered graphene to compare the Li2Sen anchoring perform-
ance of MXenes. We aimed to understand the structural stability
and binding strength of Li2Sen adsorbed on graphene. The gra-
phene monolayer used for our calculations contains a 5 × 5 super-
cell, and the lattice parameter is 12.338 Å. The Li2Sen molecules
were allowed to adsorb on the graphene by placing two Li atoms
on top of C atoms. We performed simulations to assess the
anchoring behavior of Li2Sen on graphene, and the adsorption
energies were calculated using the formula:
Eads ¼ ELi2 Sen þ EAM  ELi2 Sen þ AM

where ELi2Sen, EAM, and ELi2Sen+AM denote the total energies of polyse-
lenides, AMs, and polyselenide adsorbed on AMs, respectively. The
calculated binding energies are presented in Fig. 2, which compare
3. Results and discussion well with other works.44 The configurations of the most energeti-
3.1. Li2Sen structures
The underlying chemistry in Li–Se and Li–S batteries is quite
analogous. During the discharge process, the lithium ions
released from the anode react with the Se cathode to generate
intermediate polyselenides, leading to the final discharge pro-
ducts such as Li2Se2 and Li2Se. The most stable structures of
Li2Sen are obtained via relaxation simulations, and the opti-
mized Li2Sen structures are shown in Fig. 1. The observed
three-dimensional geometries of Li2Sen molecules arise due to
the tendency of Li+ cations to chemically bond with electron-
rich terminal Se atoms. The puckered-ring-structured Se8 pos-
sesses D4d symmetry and the shortest Se–Se bond length is
2.353 Å. Higher-order soluble polyselenides such as Li2Se8,
Li2Se6, and Li2Se4 exhibit C2 symmetry with minimum Li–Se
bond distances of 2.542, 2.581, and 2.524 Å, respectively. The
insoluble Li2Se2 and Li2Se exhibit C2v symmetry and the corres- Fig. 2 Adsorption energies of Se8 and Li2Sen on graphene and Ti3C2X2
ponding shortest bond lengths are 2.387 and 2.238 Å. We (X = S, O, F, and Cl) and the binding energies of polyselenides with elec-
observed a decreasing trend in the Li–Se bond lengths from trolyte solvent molecules (DME and DOL).

This journal is © The Royal Society of Chemistry 2020 Nanoscale, 2020, 12, 14087–14095 | 14089
 View Article Online

Paper Nanoscale

Table 1 The differences in Li–Se bond distance for Li2Sen adsorbed on AMs and isolated Li2Sen (ΔdLi–Se) and the minimum distance between the Li
atoms in Li2Sen and the AMs (dLi-AM)

Se8 Li2Se8 Li2Se6 Li2Se4 Li2Se2 Li2Se

ΔdLi–Se (Å) Graphene — 0.000 −0.081 −0.025 0.027 0.071

Ti3C2S2 — 0.045 −0.01 0.095 0.157 0.233
Ti3C2O2 — 0.066 0.038 0.087 0.157 0.190
Ti3C2F2 — 0.045 −0.034 0.022 0.065 0.106
Ti3C2Cl2 — 0.005 −0.075 −0.013 0.027 0.106
dLi-AM (Å) Graphene 3.474 2.667 3.580 3.153 2.535 2.439
Ti3C2S2 3.256 2.374 2.355 2.344 2.328 2.413
Ti3C2O2 3.163 1.867 1.908 1.864 1.876 1.866
Published on 15 June 2020. Downloaded by FAC DE QUIMICA on 2/15/2022 12:41:37 AM.

Ti3C2F2 3.466 2.064 2.050 1.931 1.875 1.846

Ti3C2Cl2 3.459 2.826 2.666 2.602 2.357 2.459

cally favored graphene-Li2Sen systems are displayed in Fig. S2.†
From the figure, we can observe that the cyclooctaselenium (Se8)
molecule is orientated parallel to the graphene sheet, with a
minimum bond distance of 3.474 Å. We observed an increase in
the Li–C bond length with increasing Se concentration in Li2Sen.
The Li–C interactions retain the structural conformation of Li2Sen;
however, slight changes in the average intramolecular Li–Se bond
lengths are observed, as presented in Table 1. Overall, the calcu-
lated binding energies for all Li2Sen are found to be below 0.80 eV.
Such weaker interactions between Li2Sen and graphene indicate
the inefficient containment of Li2Sen within the cathode materials.
3.3. Bare Ti3C2
The ineffectiveness of apolar carbon materials in preventing
Li2Sen shuttling entails the search for alternative AMs. MXenes
are promising candidates to provide the required anchoring
effect, which possibly stems from their polar characteristics.
The chosen Ti3C2 monolayer was formulated with three Ti
atoms and two C atoms arranged in a sequence of Ti–C–Ti–C–
Ti with the Ti and C atoms forming edge-sharing octahedral
structures (Ti6C).35 The relaxed lattice constant of the 4 × 4
Ti3C2 supercell is 12.370 Å, and this value is in good agree-
ment with other reports.30 The density of states (DOS) calcu-
lations reveal that the Ti3C2 material is electronically conduc-
tive (see Fig. 1) and the projected DOS (PDOS) analysis explains
that the metallic behavior arises primarily due to the 3d state
of Ti, as displayed in Fig. S1.†
The binding energy values for bare Ti3C2 with adsorbed
Li2Sen are 28.72 eV for Se8, 22.61 eV for Li2Se8, 11.19 eV for
Li2Se6, 10.91 eV for Li2Se4, 8.23 eV for Li2Se2 and 5.49 eV for
Li2Se. The optimized configurations are displayed in Fig. S3.†
It was noted that the adsorption energy decreases with
decreasing Se count in Li2Sen. The values indicate a strong
chemical interaction between Li2Sen and surface Ti atoms.
Ostensibly, the derived adsorption energies are sufficient to
bind the polyselenides to prevent their dissolution into the
electrolyte; however, the stronger binding results in the chemi-
cal decomposition of the intermediate polyselenides, which in
turn hinders the reversibility of Se. This observation directly
corroborates the experimentally observed decreased capacity
when bare MXenes are used as AMs.45,46 Additionally, strong
Ti–Se bonds can be readily formed during the preparation of
Ti3C2/Se composites or in the course of electrochemical
cycling of Se-cathodes. The decomposition of Li2Sen found
during adsorption precludes the use of bare Ti3C2 in selenium
cathodes as AMs. Thus, in order to improve the polyselenide
binding behavior with Ti3C2 and improve battery performance,
it is essential to implement surface modifications in Ti3C2.
3.4. Ti3C2X2 (X = S, O, F, and Cl)
To achieve adequate binding of polyselenides with the Ti3C2
MXene, we employed various surface functional groups, such
as X = S, O, F, and Cl. These surface-terminated MXenes were
selected based on their reported superior performance in
terms of thermodynamic stability, electronic properties, inter-
calation mechanisms, and energy storage capacity.30,47 We
obtained the optimized structures of Ti3C2X2 via relaxation
simulations with the target functional groups. The functional
group bridge between the Ti atoms at the top surface can be
observed from the optimized geometries displayed in Fig. S4.†
The resulting surface architectures are in accordance with
other reports.30 The lattice parameters of the surface-functio-
nalized Ti3C2X2 have slight differences from those of bare
Ti3C2, and the values are tabulated in Table S1.†
After obtaining stable configurations for both Li2Sen and
the various surface-functionalized Ti3C2 MXenes, we explored
the interactions between them. The Li2Sen adsorption energies
on the AMs were calculated from relaxation simulations, and
are displayed in Fig. 2.
For Ti3C2O2, the relaxed structural configuration with
Li2Sen is presented in Fig. 3. From the figure, one can see that
the Se8 molecule remains intact and parallel to the Ti3C2O2
structure without any notable deformation, and the nearest
bond distance between the Se8 molecule and the AM is
3.163 Å. However, we observed structural deformations in the
case of adsorbed higher-order Li2Sen. From Fig. 2, it can be
seen that aside from Li2Se8, the adsorption energy values
increase with lithiation and the formation of higher- to lower-
order polyselenides, and this trend is opposite to that of bare
Ti3C2. Among the Li2Sen species, Li2Se (3.07 eV) and Li2Se6
(1.40 eV) hold the highest and lowest binding energies,
respectively. The absence of Li-atoms in the Se8 molecule
results in a lower binding energy (1.03 eV) than that of the
polyselenides, since the binding is predominantly due to van

14090 | Nanoscale, 2020, 12, 14087–14095 This journal is © The Royal Society of Chemistry 2020

Nanoscale
Published on 15 June 2020. Downloaded by FAC DE QUIMICA on 2/15/2022 12:41:37 AM.
Fig. 3 Side views of the most stable configurations of Li2Sen adsorbed
on Ti3C2S2 and Ti3C2O2. The top views are provided in Fig. S5 of the
ESI.†
der Waals interactions. Furthermore, the adsorption energies
of Li2Se8, Li2Se4, and Li2Se2 were calculated to be 1.68 eV, 2.30
eV, and 2.86 eV, respectively. The shortest bond distance and
the variation in the Li–Se bond distance for all the surface-
functionalized groups are presented in Table 1.
Next, we explored Li2Sen–Ti3C2S2 interactions, and the most
stable relaxed configurations are shown in Fig. 3. Similar to
Ti3C2O2, in the case of Ti3C2S2, the Se8 molecule remains par-
allel to the surface, with the shortest bond distance between
the Se–S atoms of 3.256 Å. This value is greater when com-
pared to Ti3C2O2. The Se8 adsorbed on Ti3C2S2 retains its geo-
metric configuration; however, deformations are observed for
the polyselenides. Likewise, for Ti3C2O2, we observed that the
binding energies for Li2Sen on Ti3C2S2 are in the order of 1.12
eV for Se8, 1.41 eV for Li2Se8, 1.31 eV for Li2Se6, 1.69 eV for
Li2Se4, 3.63 eV for Li2Se2, and 3.30 eV for Li2Se, as shown in
Fig. 2. The binding energy data indicate that Li2Se possesses
the highest binding energy, while Li2Se6 exhibits the lowest
among other Li2Sen species. For both Ti3C2O2 and Ti3C2S2, Se8
exhibits the lowest binding energy.
Similarly, the adsorption energies for Ti3C2F2 and Ti3C2Cl2
were calculated from the optimized geometric configurations
(see Fig. S6 and S7†) and the data are displayed in Fig. 2.
However, the adsorption energy values of Se8 are found to be
larger when compared to other higher-order polyselenides,
indicating weaker Li2Sen adsorption behavior that could result
in ineffective anchoring. With lithiation, both classes of
materials exhibit a decrease in the adsorption energy values
for Li2Se6, and the value increases as lithiation further pro-
ceeds. The shortest bond distances between Se8, Li2Sen, and
AMs were calculated from the most energetically favorable con-
figurations and are tabulated in Table 1. From the overall
This journal is © The Royal Society of Chemistry 2020
View Article Online
Paper
adsorption energy data of Ti3C2F2 and Ti3C2Cl2, we observed
poor adsorption behavior when compared to other functiona-
lized Ti3C2X2. Additionally, the polyselenide adsorption ener-
gies on graphene were found to be the lowest among all the
materials considered.
Next, in order to assess the anchoring effectiveness of the
functionalized Ti3C2 MXenes—to evaluate adequate binding
strength for preventing dissolution of higher-order polysele-
nides (Li2Sen ≥ 4) into the electrolyte—we calculated the
binding strength of Li2Sen ≥ 4 with the commonly used 1,3-
dioxolane (DOL) and 1,2-dimethoxymethane (DME) electrolyte
solvent molecules and compared them with the corresponding
binding energies of surface-functionalized MXenes and gra-
phene. Typically, the binding energies of polychalcogenides
with solvent molecules are calculated considering the inter-
action of one Li atom of polychalcogenide through a Li–O
covalent bond.37,48 The binding energies of Li2Sen ≥ 4 bonded
with the solvent molecules are represented in Fig. 2, and the
most favorable configurations are displayed in Fig. S8.† It has
been reported that stronger interactions between the higher-
order polyselenides and AMs than those of solvent molecules
are required to prevent polyselenide dissolution into the
electrolyte.30,37 Fig. 2 clearly indicates that out of all the
materials considered, only Ti3C2S2 and Ti3C2O2 exhibit higher
binding energies for Li2Sen ≥ 4 compared to the DOL/DME
cases. Therefore, it can be concluded that the higher thermo-
dynamic stability of Li2Sen ≥ 4 adsorbed on Ti3C2S2 and
Ti3C2O2 structures can effectively constrain the higher-order
polyselenides from dissolving into the electrolytes.
Next, we analyzed the geometric features of Li2Sen adsorbed on
AMs. From Table 1, one can see that for all the cases of graphene
and Ti3C2X2, the largest distance between the AMs and adsorbate
are found for Se8, and Fig. 3 and Fig. S4–S7† show no structural
deformation of the molecules. In the case of Li2Sen, the Li atoms
are found closer to the AMs than the Se atoms, and the shortest
distance between Li in Li2Sen and the surface increases with
increasing number of Se atoms. This is predominantly because of
the electronegative nature of Se and its stronger chemical inter-
actions. The trend of increasing Li2Sen and AM interaction strength
with decreasing Se number is correlated with the weakening of the
intermolecular Li–Se bond strength, which is evident from the
increasing intramolecular Li–Se bond lengths in Li2Sen.
Finally, by observing the polyselenide adsorption energies
of various functionalized Ti3C2X2, we conclude that both S and
O functionalization results in superior anchoring behavior
compared to F and Cl functionalization and graphene. The
moderate polyselenide binding energies of O and S can be
deemed as adequate for the containment of Li2Sen within the
Se-cathode material without chemical decomposition.
Furthermore, we probed the relative strength of the chemi-
cal and van der Waals (vdW) interactions, employing the vdW
ratio calculated using the formula:
 
no vdw
Ebvdw  Eb
R ¼
Ebvdw
Nanoscale, 2020, 12, 14087–14095 | 14091

Paper
Published on 15 June 2020. Downloaded by FAC DE QUIMICA on 2/15/2022 12:41:37 AM.
Fig. 4 The ratio of vdW interactions of Se8 and Li2Sen adsorbed on gra-
phene and Ti3C2X2.
where Evdw
b and Enovdw
b are the binding energies of Li2Sen with
and without van der Waals interactions, respectively. The ratio
of vdW interactions for graphene and various AMs are shown
in Fig. 4. From the figure, we can observe that for all AMs, the
ratio of vdW interactions is nearly 100% in the case of Se8,
which clearly demonstrates that the Se8 interactions with the
AMs are predominantly physical, not chemical. However, as
the lithiation process continues, the vdW ratio drops and the
trend is found to be opposite to that of the binding energy
values. The interaction of Li atoms with the AMs results in
stronger chemical interactions, which is evident from the
reduction of the vdW ratio, and the highest chemical inter-
actions are found in the cases involving Li2Se. For Ti3C2S2 and
Ti3C2O2, we observe a systematic decrease in the vdW inter-
actions with lithiation; however, in the case of Ti3C2F2 and
Ti3C2Cl2, the vdW ratio is greater than 50% except for Ti3C2F2–
Li2Se, which indicates that the lithiation process is dominated
by vdW interactions and these values are consistent with other
work.30
3.5. Bader charge analysis
In order to obtain further insights into the Li2Sen binding
mechanisms, Bader charge and charge density difference ana-
lyses were performed for graphene and various surface-termi-
nated Ti3C2X2. The charge transfer values are displayed in
Fig. 5. From the figure, it is evident that for all variations of
Ti3C2X2, with the discharge process, electron transfer from the
polyselenides to Ti3C2X2 increases, indicating stronger chemi-
cal interactions between them. However, in the case of gra-
phene, the charge transfer values are minimal, indicating the
weaker adsorption of Li2Sen on the surface due to the nonpolar
characteristics of graphene. All of the materials exhibit a
similar trend in the charge transfer values, that is, the values
increase from higher- to lower-order polyselenides, except in
the case of Li2Se6 species. The lower charge transfer value of
Li2Se6 is correlated with the lower adsorption energies. The
amounts of charge transfer in Ti3C2S2 and Ti3C2O2 are greater
when compared to F and Cl functionalization and graphene,
which demonstrates the stronger chemical interactions of
14092 | Nanoscale, 2020, 12, 14087–14095
View Article Online
Nanoscale
Fig. 5 Charge transfer between Se8 and polyselenides with graphene
and Ti3C2X2.
Li2Sen with S- and O-terminated surfaces. However, the lower
charge transfer values for F and Cl can be attributed to the
weaker chemical interaction between Li2Sen and the AMs. The
calculated charge transfer values for Ti3C2S2-adsorbed struc-
tures are 0.274 |e| for Se8, 0.864 |e| for Li2Se8, 0.949 |e| for
Li2Se6, 0.840 |e| for Li2Se4, 0.980 |e| for Li2Se2 and 1.110 |e| for
Li2Se. For Ti3C2O2, the values are 0.159 |e| for Se8, 1.031 |e| for
Li2Se8, 0.967 |e| for Li2Se6, 1.027 |e| for Li2Se4, 0.998 |e| for
Li2Se2 and 0.949 |e| for Li2Se. The positive charge transfer
values observed in all of the Ti3C2X2 materials indicate that
the charge is transferred from the Li2Sen species to the
Ti3C2X2. The results illustrate that the charge transfer value of
Se8 is the lowest for all functionalized groups when compared
to other Li2Sen species. The softening of the adsorbed intra-
molecular Li–Se bonds can be explained by considering the
charge transfer process from the polyselenides to Ti3C2X2. The
electrons are bereaved from the Li2Sen as they are absorbed on
the Ti3C2X2. The higher electron deficiency in Li2Sen at the
adsorbed state results in enhanced chemical interactions of Li
with the AMs and causes greater elongation of Li–Se bonds in
Li2Sen. The charge transfer primarily occurs from the Li atom
to the AMs, because of the smaller differences in electro-
negativity between Se and surface atoms and the distance of Se
atoms from the surface. To visualize the charge transfer
characteristics, we performed differential charge density ana-
lysis on representative cases, including Li2Se, Li2Se4, and
Li2Se8, and the results are displayed in Fig. 6, S9 and S10.† The
charge depletion and accumulation observed from the analysis
are in accordance with the charge transfer values obtained
from the Bader charge analysis.
3.6. Density of states (DOS)
We further studied the bonding characteristics between the
Li2Sen and Ti3C2X2 by examining the electronic structures of
the adsorbed systems. We calculated the total and projected
DOS (PDOS) to probe the conductive behavior of the Ti3C2X2.
Total DOS analyses of pristine Ti3C2X2, Li2Se8, Li2Se4, and
Li2Se adsorbed on Ti3C2X2 systems (Fig. S11–S14†) were per-
This journal is © The Royal Society of Chemistry 2020

Nanoscale
Published on 15 June 2020. Downloaded by FAC DE QUIMICA on 2/15/2022 12:41:37 AM.
Fig. 6 Side views of charge density differences of Li2Se8, Li2Se4 and
Li2Se on Ti3C2S2 and Ti3C2O2. The iso-surface level is set at 0.0020 e
Å−3. The red and green colors denote charge accumulation and
depletion, respectively.
formed as representative cases of higher, medium, and lower-
order polyselenides, to understand the change in electronic
structure due to charge transfer from Li2Sen to the functiona-
lized MXenes before and after adsorption. However, in all
Li2Sen cases, there is no significant change in the DOS for all
the functional groups regardless of Li2Sen adsorption. From
Fig. S11–S14,† we observe that, compared to pristine AMs, the
systems with adsorbed Li2Sen result in negligible changes in
the electronic structures, as the electronic conductivity of the
AMs is retained even after Li2Sen adsorption. Furthermore, to
understand the influence of corrected on-site coulombic inter-
actions on the electronic properties of the MXenes, we per-
formed a few benchmark calculations using GGA + U with U =
4.2 eV.49,50 The comparisons of the TDOS calculated using the
PBE/GGA and GGA + U are presented in Fig. S15.† The figure
clearly demonstrates that both methods predict metallic
characteristics of the functionalized and polysulfides adsorbed
MXenes. Since we are interested in understanding the elec-
tronic properties of the AMs in a qualitative manner, the
chosen PBE/GGA method is adequate. The electronic conduc-
tive properties facilitate electron transfer directly to the
involved redox reactions of Li2Sen intermediates. In order to
further investigate the influence of Li2Sen on Ti3C2X2, pro-
jected PDOS was analyzed for both pristine and adsorbed
systems of Ti3C2S2 and Ti3C2O2, and the results are displayed
in Fig. 7. From the figure, we observe that compared to pristine
AMs, a considerable amount of electron states crosses the
Fermi level in the case of Li2Sen adsorbed on Ti3C2S2 and
Ti3C2O2, indicating that the host materials can effectively cir-
cumvent the insulating behavior of Se8. Moreover, the metallic
behavior of Li2Sen adsorbed on Ti3C2S2 and Ti3C2O2 is main-
tained, as is evident from the abundance of electronic states in
the Fermi energy level stemming from the 3d state of Ti.
This journal is © The Royal Society of Chemistry 2020
View Article Online
Paper
Fig. 7 Projected density of states of representative polyselenides in the
case of Ti3C2S2 (left) and Ti3C2O2 (right). The Fermi level is indicated with
vertical lines.
Moreover, a strong or partial polar covalent bond is formed
from the hybridization of the Ti-3d and 2p/3p of S/O, resulting
in Ti–S/O bonds. Hence all the adsorbed systems possess good
electrical conductivity owing to the contribution of electrons
from Ti atoms, which is essential to expedite electrochemical
reactions in Li–Se battery systems.
4. Conclusion
In summary, we employed DFT calculations to demonstrate
the functionality of graphene and Ti3C2 MXenes as host
materials for Se-cathodes. We elucidated the mechanistic
details of the inadequate polyselenide (Li2Sen) binding ener-
gies on graphene and strategies to improve the Li2Sen retention
capability within the Se-cathode using a new class of materials,
Ti3C2 MXenes. The calculated weak adsorption energies,
resulting from Li2Sen binding on graphene, clearly indicate
that graphene is incapable of providing requisite anchoring to
prevent polyselenides from dissolving. Furthermore, we found
that bare Ti3C2 cannot be used as an AM because of its stron-
ger binding and subsequent dissociation of polyselenides on
the surface, which can debilitate the reversibility of the battery.
We leveraged Ti3C2 surfaces modified with various functional
groups to overcome the limitations of bare Ti3C2. We identified
Ti3C2S2 and Ti3C2O2 as AMs with superior performance, which
can provide adequate binding to Li2Sen without chemical
decomposition, compared to graphene and other functiona-
lized MXenes. We consider S- and O-functionalized Ti3C2 to be
an appropriate choice for achieving improved performance of
Se-based cathodes for Li2Sen adsorption. The computed
binding energies of polyselenides with Ti3C2S2 and Ti3C2O2
Nanoscale, 2020, 12, 14087–14095 | 14093

Paper
are higher than those of commonly used ether-based electro-
lyte solvents, which indicates the effectiveness of the materials
to suppress Li2Sen migration. From the vdW ratio analysis, we
observed that the chemical interactions primarily originate
from the intercalated Li+ ions, and the van der Waal inter-
actions were contributed by Se. From the charge transfer ana-
lysis, we found that the charge is transferred from the Li2Sen
structure to Ti3C2X2. The density of states analysis of Ti3C2X2
revealed that the metallic properties of all the materials are
retained even after adsorption with polyselenides, which is
Published on 15 June 2020. Downloaded by FAC DE QUIMICA on 2/15/2022 12:41:37 AM.
beneficial for promoting the electrochemical process. Overall,
S- and O-functionalized Ti3C2 can be used to achieve two-fold
objectives of: (i) improving performance for polyselenide
anchoring; and (ii) facilitating the electronic conduction path
to overcome the limitations of the insulating nature of Se, as
well as to eliminate the use of carbon. Thus, the utilization of
the Ti3C2S2 and Ti3C2O2 MXenes in Se-based cathodes can be
deemed as a sizeable step forward towards designing carbon-
free high-performance Li–Se batteries.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
M. M. I. acknowledges the start-up funds from Wayne State
University and the support of the Extreme Science and
Engineering Discovery Environment (XSEDE)51 for providing
the computational facilities (Start-up Allocation: DMR190089
and Research Allocation: DMR200009).
Notes and references
1 M. Armand and J.-M. Tarascon, Nature, 2008, 451, 652–657.
2 P. G. Bruce, S. A. Freunberger, L. J. Hardwick and
J.-M. Tarascon, Nat. Mater., 2011, 11, 19–29.
3 X. Ji and L. F. Nazar, J. Mater. Chem., 2010, 20, 9821–9826.
4 L. F. Nazar, M. Cuisinier and Q. Pang, MRS Bull., 2014, 39,
436–442.
5 S.-Y. Chung, J. T. Bloking and Y.-M. Chiang, Nat. Mater.,
2002, 1, 123–128.
6 X. Ji, K. T. Lee and L. F. Nazar, Nat. Mater., 2009, 8, 500–
506.
7 X. Liang, Z. Wen, Y. Liu, M. Wu, J. Jin, H. Zhang and
X. Wu, J. Power Sources, 2011, 196, 9839.
8 D. Rezan, Li-s Batteries: The Challenges, Chemistry,
Materials, And Future Perspectives, 2017.
9 A. Abouimrane, D. Dambournet, K. W. Chapman,
P. J. Chupas, W. Weng and K. Amine, J. Am. Chem. Soc.,
2012, 134, 4505–4508.
10 L. Liu, Y. Hou, X. Wu, S. Xiao, Z. Chang, Y. Yang and
Y. Wu, Chem. Commun., 2013, 49, 11515–11517.
14094 | Nanoscale, 2020, 12, 14087–14095
View Article Online
Nanoscale
11 Q. Cai, Y. Li, L. Wang, Q. Li, J. Xu, B. Gao, X. Zhang,
K. Huo and P. K. Chu, Nano Energy, 2017, 32, 1–9.
12 J. Yang, H. Gao, D. Ma, J. Zou, Z. Lin, X. Kang and S. Chen,
Electrochim. Acta, 2018, 264, 341–349.
13 A. Eftekhari, Sustainable Energy Fuels, 2017, 1, 14–29.
14 C.-P. Yang, Y.-X. Yin and Y.-G. Guo, J. Phys. Chem. Lett.,
2015, 6, 256–266.
15 W. Yuan, Y. Feng, A. Xie, X. Zhang, F. Huang, S. Li,
X. Zhang and Y. Shen, Nanoscale, 2016, 8, 8704–8711.
16 C. Luo, Y. Xu, Y. Zhu, Y. Liu, S. Zheng, Y. Liu, A. Langrock
and C. Wang, ACS Nano, 2013, 7, 8003–8010.
17 T. Liu, C. Dai, M. Jia, D. Liu, S. Bao, J. Jiang, M. Xu and
C. M. Li, ACS Appl. Mater. Interfaces, 2016, 8, 16063–16070.
18 S.-F. Zhang, W.-P. Wang, S. Xin, H. Ye, Y.-X. Yin and
Y.-G. Guo, ACS Appl. Mater. Interfaces, 2017, 9, 8759–8765.
19 T. Liu, Y. Zhang, J. Hou, S. Lu, J. Jiang and M. Xu, RSC Adv.,
2015, 5, 84038–84043.
20 Y. Liu, L. Si, Y. Du, X. Zhou, Z. Dai and J. Bao, J. Phys.
Chem. C, 2015, 119, 27316–27321.
21 J. Zhang, Z. Zhang, Q. Li, Y. Qu and S. Jiang, J. Electrochem.
Soc., 2014, 161, A2093–A2098.
22 H.-C. Youn, J. H. Jeong, K. C. Roh and K.-B. Kim, Sci. Rep.,
2016, 6, 1–8.
23 L. Zeng, W. Zeng, Y. Jiang, X. Wei, W. Li, C. Yang, Y. Zhu
and Y. Yu, Adv. Energy Mater., 2015, 5, 1401377.
24 Z. Yi, L. Yuan, D. Sun, Z. Li, C. Wu, W. Yang, Y. Wen,
B. Shan and Y. Huang, J. Mater. Chem. A, 2015, 3, 3059–
3065.
25 Z. Li and L. Yin, Nanoscale, 2015, 7, 9597–9606.
26 H. Lv, R. Chen, X. Wang, Y. Hu, Y. Wang, T. Chen, L. Ma,
G. Zhu, J. Liang, Z. Tie, J. Liu and Z. Jin, ACS Appl. Mater.
Interfaces, 2017, 9, 25232–25238.
27 K. Han, Z. Liu, J. Shen, Y. Lin, F. Dai and H. Ye, Adv. Funct.
Mater., 2015, 25, 455–463.
28 J. He, Y. Chen, W. Lv, K. Wen, P. Li, Z. Wang, W. Zhang,
W. Qin and W. He, ACS Energy Lett., 2016, 1, 16–20.
29 K. Mi, Y. Jiang, J. Feng, Y. Qian and S. Xiong, Adv. Funct.
Mater., 2016, 26, 1571–1579.
30 D. Wang, F. Li, R. Lian, J. Xu, D. Kan, Y. Liu, G. Chen,
Y. Gogotsi and Y. Wei, ACS Nano, 2019, 13, 11078–11086.
31 M. Naguib, M. Kurtoglu, V. Presser, J. Lu, J. Niu, M. Heon,
L. Hultman, Y. Gogotsi and M. W. Barsoum, Adv. Mater.,
2011, 23, 4248–4253.
32 M. Hu, Z. Li, T. Hu, S. Zhu, C. Zhang and X. Wang, ACS
Nano, 2016, 10, 11344–11350.
33 J. Hu, B. Xu, C. Ouyang, S. A. Yang and Y. Yao, J. Phys.
Chem. C, 2014, 118, 24274–24281.
34 Y.-X. Yu, J. Phys. Chem. C, 2016, 120, 5288–5296.
35 Q. Tang, Z. Zhou and P. Shen, J. Am. Chem. Soc., 2012, 134,
16909–16916.
36 X. Liang, Y. Rangom, C. Y. Kwok, Q. Pang and L. F. Nazar,
Adv. Mater., 2017, 29, 1603040.
37 Y. Wang, J. Shen, L.-C. Xu, Z. Yang, R. Li, R. Liu and X. Li,
Phys. Chem. Chem. Phys., 2019, 21, 18559–18568.
38 G. Kresse and J. Furthmüller, Phys. Rev. B: Condens. Matter
Mater. Phys., 1996, 54, 11169.
This journal is © The Royal Society of Chemistry 2020

Nanoscale
39 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,
1996, 77, 3865–3868.
40 S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem.
Phys., 2010, 132, 154104.
41 S. Grimme, S. Ehrlich and L. Goerigk, J. Comput. Chem.,
2011, 32, 1456–1465.
42 G. Henkelman, A. Arnaldsson and H. Jónsson, Comput.
Mater. Sci., 2006, 36, 354–360.
43 K. Momma and F. Izumi, J. Appl. Crystallogr., 2011, 44,
1272–1276.
Published on 15 June 2020. Downloaded by FAC DE QUIMICA on 2/15/2022 12:41:37 AM.
44 Y. Qiu, W. Li, W. Zhao, G. Li, Y. Hou, M. Liu, L. Zhou, F. Ye,
H. Li, Z. Wei, S. Yang, W. Duan, Y. Ye, J. Guo and Y. Zhang,
Nano Lett., 2014, 14, 4821–4827.
45 W. Bao, C. E. Shuck, W. Zhang, X. Guo, Y. Gogotsi and
G. Wang, ACS Nano, 2019, 13, 11500–11509.
This journal is © The Royal Society of Chemistry 2020
View Article Online
Paper
46 Y. Song, Z. Sun, Z. Fan, W. Cai, Y. Shao, G. Sheng,
M. Wang, L. Song, Z. Liu, Q. Zhang and J. Sun, Nano
Energy, 2020, 70, 104555.
47 H.-W. Wang, M. Naguib, K. Page, D. J. Wesolowski and
Y. Gogotsi, Chem. Mater., 2016, 28, 349–359.
48 B. Wang, S. M. Alhassan and S. T. Pantelides, Phys. Rev.
Appl., 2014, 2, 034004.
49 S. Kajiyama, L. Szabova, K. Sodeyama, H. Iinuma,
R. Morita, K. Gotoh, Y. Tateyama, M. Okubo and
A. Yamada, ACS Nano, 2016, 10, 3334–3341.
50 D. Wang, Y. Gao, Y. Liu, D. Jin, Y. Gogotsi, X. Meng, F. Du,
G. Chen and Y. Wei, J. Phys. Chem. C, 2017, 121, 13025–13034.
51 J. Towns, T. Cockerill, M. Dahan, I. Foster, K. Gaither,
A. Grimshaw, V. Hazlewood, S. Lathrop, D. Lifka and
G. D. Peterson, Comput. Sci. Eng., 2014, 16, 62–74.
Nanoscale, 2020, 12, 14087–14095 | 14095