Boya Venugopal1,2,3, Zeru Syum1, Sheng-Yu Yu4, Amr Sabbah1, Indrajit Shown1,5, Chih-Wei Chu6, Li-Chyong Chen4,7, Chih-Hao Lee3, Heng-Liang Wu4,7, Kuei-Hsien Chen1,4. 1. Institute of Atomic and Molecular Sciences, Academia Sinica, Taipei 10617, Taiwan. 2. Nanoscience and Technology Program, Taiwan International Graduate Program, Academia Sinica, Taipei 11529, Taiwan. 3. Department of Engineering and System Science, National Tsing Hua University, Hsinchu 30013, Taiwan. 4. Center for Condensed Matter Sciences, National Taiwan University, Taipei 10617, Taiwan. 5. Department of Chemistry, Hindustan Institute of Technology and Science, Chennai 603103, India. 6. Research Center for Applied Sciences, Academia Sinica, Taipei 11529, Taiwan. 7. Center of Atomic Initiative for New Materials, National Taiwan University, Taipei 10617, Taiwan.
Abstract
The widespread use of energy storage technologies has created a high demand for the development of novel anode materials in Li-ion batteries (LIBs) with high areal capacity and faster electron-transfer kinetics. In this work, carbon-coated Cu2ZnSnS4 with a hierarchical 3D structure (CZTS@C) is used as an anode material for LIBs. The CZTS@C microstructures with enhanced electrical conductivity and improved Li-ion diffusivity exhibit high areal and gravimetric capacities of 2.45 mA h/cm2 and 1366 mA h/g, respectively. The areal capacity achieved in the present study is higher than that of previously reported CZTS-based materials. Moreover, in situ X-ray diffraction results show that lithium ions are stored in CZTS through the insertion reaction, followed by the alloying and conversion reactions at ∼1 V. The structural evolution of Li2S and Cu-Sn/Cu-Zn alloy phases occurs during the conversion and alloying reactions. The present work provides a cost-effective and simple method to prepare bulk CZTS and highlights the conformal carbon coating over CZTS, which can enhance the electrical and ionic conductivities of CZTS materials and increase the mass loading (1-2.3 mg/cm2). The improved stability and rate capability of CZTS@C anode materials can therefore be achieved.
The widespread use of energy storage technologies has created a high demand for the development of novel anode materials in Li-ion batteries (LIBs) with high areal capacity and faster electron-transfer kinetics. In this work, carbon-coated Cu2ZnSnS4 with a hierarchical 3D structure (CZTS@C) is used as an anode material for LIBs. The CZTS@C microstructures with enhanced electrical conductivity and improved Li-ion diffusivity exhibit high areal and gravimetric capacities of 2.45 mA h/cm2 and 1366 mA h/g, respectively. The areal capacity achieved in the present study is higher than that of previously reported CZTS-based materials. Moreover, in situ X-ray diffraction results show that lithium ions are stored in CZTS through the insertion reaction, followed by the alloying and conversion reactions at ∼1 V. The structural evolution of Li2S and Cu-Sn/Cu-Zn alloy phases occurs during the conversion and alloying reactions. The present work provides a cost-effective and simple method to prepare bulk CZTS and highlights the conformal carbon coating over CZTS, which can enhance the electrical and ionic conductivities of CZTS materials and increase the mass loading (1-2.3 mg/cm2). The improved stability and rate capability of CZTS@C anode materials can therefore be achieved.
The increasing demand
for electric vehicles and portable electronic
devices has prompted new research into the development of Li-ion batteries
(LIBs) with higher energy density. Although graphite anode materials
have been widely used in LIBs, their low theoretical capacity (372
mA h/g) and poor rate capability limit their potential in the development
of high-energy density LIBs.[1−3] Therefore, it is highly desirable
to develop new anode materials for LIBs with high energy and power
density, as well as good stability.[4] Various
anode materials such as transition-metal oxides (Fe2O3 and SnO2) and metals (Si and Sn) have been explored.[1,3,5,53] Among
these materials, Si has received considerable attention because of
its high gravimetric capacity and stability at low mass loadings.
However, anode materials with high mass loading and high areal capacity
are needed for practical applications. Although high mass loadings
of the Si anode lead to significantly improved areal capacities (1–3.5
mA h/cm2), the Si material suffers from the low Coulombic
efficiency (CE) and large volume expansion of the active material,
resulting in poor reversibility and fast capacity fading. On the other
hand, materials with 2D-layered structures and large interlayer spacing,
such as SnS2 and MoS2, improve the Li-ion diffusivity
and suppress the volume expansion.[6−8] Unfortunately, these
anode materials suffer from low conductivity and poor rate capability.[9]Recently, several research groups demonstrated
the application
of multimetallic (Cu, Zn, and Sn) sulfides, such as Cu2SnS3 and Cu2ZnSnS4 (CZTS), as anode
materials for LIBs, wherein Sn and Zn can easily form alloys with
lithium and provide high capacity.[10,11] The earth-abundant
transition metals (e.g., Cu2SnS3 or CZTS) act
as a buffer matrix, enabling alloying and conversion reactions that
result in high capacity.[12] Moreover, discharge
products such as the Cu–Zn alloy formed during lithiation can
also enhance the conductivity of the electrode during cycling.[13] Thus, CZTS has been regarded as the advanced
anode material with enhanced energy and power densities in next-generation
LIBs. Yang et al. demonstrated the first CZTS-based
anode for LIBs with initial discharge and charge capacities of 857
and 288 mA h/g, respectively.[14] This system
maintained a capacity around 234 mA h/g after 30 cycles, and it exhibited
a low initial CE and poor electrochemical performance. Later, Jiang et al. reported the electrochemical performance of CZTS
nanoflowers with high capacity and stability at low current rates
(100 mA/g).[15] Another study showed that
CZTS thin films exhibited a high capacity of about 668 mA h/g at a
current rate of 100 mA/g. However, the CZTS thin films exhibited a
low areal capacity of ∼0.45 mA h/cm2 with a limited
mass loading of 0.5 mg/cm2, and the material processing
approach was also expensive.[16] Recently,
Chiu et al. demonstrated that CZTS thin films exhibited
high discharge capacity (1400 mA h/g) and stability over 400 cycles,
but a low mass loading of 0.3 mg/cm2 was obtained.[13] Bree et al. also reported that
CZTS thin films with mass loadings of 0.1–0.2 mg/cm2 delivered a stable discharge capacity of about 600 mA h/g at a current
rate of 500 mA/g.[12] Although CZTS anode
materials have shown promising electrochemical performance, they are
far from satisfying the commercial requirements, mainly due to their
low areal capacity (<0.5 mA h/cm2) and limited mass
loadings. In addition, the mechanistic studies of CZTS were rare.In this work, we proposed a cost-effective and simple process to
develop advanced CZTS anode materials with a bulk structure coated
with different amounts of carbon (hereafter named CZTS@C) and high
mass loadings. The CZTS@C material used as an anode in LIBs exhibits
a high gravimetric capacity of 1366 mA h/g and areal capacities of
1–2.45 mA h/cm2 upon cycling. The stability of CZTS@C
is improved during cycling, which could result from the formation
of a stable solid–electrolyte interphase (SEI). Electrochemical
impedance spectroscopy (EIS) results show that the carbon coating
improves the conductivities of CZTS anode materials during cycling. In situ X-ray diffraction (XRD) is employed to examine the
reaction mechanism of the CZTS@C material. The formation of Li2S at low potential and the evolution of various Cu–Sn/Cu–Zn
alloys are obtained during the charge–discharge process, suggesting
that multiple electron-transfer process results in a high theoretical
capacity. The present study provides new insights into the use of
quaternary sulfides as high-performance anode materials with higher
energy density for the development of advanced LIBs.
Results and Discussion
Morphology
and Crystal Structure
The desired carbon
content in the CZTS@C samples was controlled from 5 to 20 wt %, as
determined by elemental analysis (EA). The CZTS@C samples with carbon
contents of 4.4, 6.9, 9.7, 13.6, and 20.3 wt % are denoted as CZTS@C-5,
CZTS@C-7, CZTS@C-10, CZTS@C-15, and CZTS@C-20, respectively (Supporting Information, Table S1). The synthesis
process of CZTS@C is illustrated in Scheme . The field-emission scanning electron microscopy
(FE-SEM) images of pristine CZTS show a 3D flower-like morphology
composed of petal-like interconnected CZTS nanosheets with a thickness
of few nanometers (Figure a). The FE-SEM images in Figure b–f show that the morphology of CZTS
remains the same even after coating with different amounts of carbon,
indicating that carbon could successfully encapsulate the CZTS nanosheets
as a thin film. Even at higher carbon coatings (13.6 wt %, CZTS@C-15),
the CZTS nanosheet surface was uniformly covered with a thin carbon
layer, and no local agglomeration was observed. Previous studies suggest
that the uniform carbon coating layer on the material avoids direct
contact with the electrolyte and controls the SEI film thickness.
The stability of the material can be improved.[17]
Scheme 1
Synthesis Process of CZTS@C
Figure 1
FE-SEM images of (a) CZTS, (b) CZTS@C-5, (c) CZTS@C-7, (d) CZTS@C-10,
(e) CZTS@C-15, and (f) CZTS@C-20 samples.
FE-SEM images of (a) CZTS, (b) CZTS@C-5, (c) CZTS@C-7, (d) CZTS@C-10,
(e) CZTS@C-15, and (f) CZTS@C-20 samples.The morphology and microstructure of CZTS@C-15 were investigated
using high-resolution transmission electron microscopy (HRTEM). Figure a,b shows that the
CZTS surface was coated with amorphous carbon uniformly and the thickness
of the carbon layer was ∼10 nm. The lattice fringe with an
interplanar spacing of 0.315 nm is consistent with the (112) d-spacing of kesterite CZTS (PDF no. 26-0575).[11,18] The selected-area electron diffraction (SAED) pattern of CZTS highlights
the polycrystalline nature of the sample, and all diffraction rings
are associated with the (112), (220), and (312) lattice planes of
kesterite CZTS (Figure c).[11]Figure d–k shows the high-angle annular dark-field
(HAADF) image and energy-dispersive X-ray spectroscopy (EDX) elemental
pattern as well as the elemental mapping of CZTS@C-15. The elemental
composition measurements of CZTS@C-15 show that the Cu, Zn, Sn, S,
and C elements are evenly distributed within the 3D nanostructure
of the sample.
Figure 2
(a,b) HRTEM images, (c) SAED pattern, (d) HAADF image,
(e) EDX
elemental pattern (f–j), EDX elemental mappings, and (k) EDX
file of Cu, Zn, Sn, S, and C elements in the CZTS@C-15 sample.
(a,b) HRTEM images, (c) SAED pattern, (d) HAADF image,
(e) EDX
elemental pattern (f–j), EDX elemental mappings, and (k) EDX
file of Cu, Zn, Sn, S, and C elements in the CZTS@C-15 sample.We next employ XRD measurements to determine the
crystal structure
of the as-synthesized CZTS and CZTS@C samples (Figure a). The XRD peaks at 2θ = 28.5, 32.9,
47.3, and 56.4° are associated with the (112) (200), (220), and
(312) lattice planes, respectively, which is consistent with previous
studies (JCPDS no. 26-0575) (Figure S1).[18] In addition, the XRD of CZTS@C materials shows
the peaks at 32.5 and 52.4°, which can be attributed to the formation
of a secondary Cu2–S phase upon
the annealing process.[19−21] The XRD results of CZTS@C exhibit the sharper diffraction
pattern with stronger intensity than that of pristine CZTS, suggesting
that the grain size and crystallinity of CZTS are changed after the
coating process. The calcination process was performed at 400 ○C, which could further improve the crystallinity of
CZTS. Also, the coating process could result in the larger grain size
of CZTS (Table S2).
Figure 3
(a) XRD and (b) Raman
spectra of CZTS and CZTS@C samples.
(a) XRD and (b) Raman
spectra of CZTS and CZTS@C samples.Raman spectroscopy was used to study the phase of carbon-coated
CZTS. Figure b shows
the Raman spectra of the CZTS-based materials. The peaks at 287–288
and 335 cm–1 are associated with kesterite CZTS.[20,21] Moreover, the Raman peaks at 1355 and 1585 cm–1 are assigned to the D- and G bands originating from carbon, respectively.[22] Herein, the intensity of D/G ratio is 0.9 for CZTS@C with the carbon content below
10 wt % and 0.8 for CZTS@C with a carbon content above 10 wt %, respectively.
The D/G ratio of the CZTS@C samples
is close, suggesting that the degree of graphitization is similar
in different carbon contents. In addition, the peak with weak intensity
at ∼1750 cm–1 is assigned to the M band,
which is an overtone of the out-of-plane phonon and has been observed
in graphite and single-walled nanotube samples.[23,24]
Electrochemical Behavior of As-Prepared Electrodes
Figure a,b shows
the charge/discharge voltage profiles of the CZTS and CZTS@C-15 materials
in different cycles at a current density of 1000 mA/g. The discharge
profile of CZTS shows a broad discharge plateau at ∼1 V in
the first discharge process and is attributed to the SEI layer formation
with Li2S as one of the main products. In contrast, this
broad discharge plateau diminishes remarkably with the controlled
SEI layer, mainly due to the carbon coating (see Figure b). The subsequent plateaus
located below 1 V could be attributed to the generation of LiZn and LiSn
alloys and Li2S in both CZTS and CZTS@C materials. After
100 cycles, the charge–discharge curve of CZTS exhibits no
plateaus, which results from the severe volume expansion during lithiation
and delithiation processes. The volume expansion also results in poor
electrical and ionic pathways during cycling. In contrast, the CZTS@C
material exhibits better stability over cycling. A similar kind of
information is obtained in other carbon-coated CZTS materials (please
see Figure S2). Figure S3 shows that the CZTS-based materials with different carbon
coating amounts exhibited different performances, suggesting that
the performance of CZTS can be optimized by adjusting the carbon coating
level.
Figure 4
Electrochemical charge–discharge curves for (a) CZTS and
(b) CZTS@C-15 with mass loading 1.1 mg/cm2 at a current
density of 1000 mA/g, from the first to 100th cycle; (c) cycling performance
at a current density of 1000 mA/g of CZTS and CZTS@C samples with
various carbon contents; (d) cycling performance of CZTS@C-15 at various
current densities; (e) rate performances of CZTS and CZTS@C-15 from
500 to 5000 mA/g; and (f) areal capacity of CZTS@C-15 samples with
various mass loadings at a current density of 0.5 mA/cm2.
Electrochemical charge–discharge curves for (a) CZTS and
(b) CZTS@C-15 with mass loading 1.1 mg/cm2 at a current
density of 1000 mA/g, from the first to 100th cycle; (c) cycling performance
at a current density of 1000 mA/g of CZTS and CZTS@C samples with
various carbon contents; (d) cycling performance of CZTS@C-15 at various
current densities; (e) rate performances of CZTS and CZTS@C-15 from
500 to 5000 mA/g; and (f) areal capacity of CZTS@C-15 samples with
various mass loadings at a current density of 0.5 mA/cm2.Figure c shows
that the electrochemical performance of the both CZTS and CZTS@C-15
material with a mass loading of 1.1 mg/cm2. The CZTS material
exhibits a discharge capacity of 2148 mA h/g in the initial cycle
and the capacity fades drastically during cycling. The CZTS@C-15 shows
a discharge capacity of 1802 mA h/g in the first cycle and reaches
a discharge capacity of 815 mA h/g after 100 cycles, which represents
the best performance among the CZTS-based materials. Although the
initial discharge capacity of CZTS varies with different carbon contents,
the initial discharge capacity of CZTS is 1200–1600 mA h/g
when the carbon content is ranging from 0 to 15%. The CZTS@C-15 exhibits
better electrochemical stability and the poor electrochemical performance
of CZTS@C-20 is obtained. Moreover, CZTS-based materials exhibit an
initial CE nearly 70% and increase to 100% in the following cycles
(Figure S4). The low CE suggests that an
irreversible reaction and side reaction, such as the formation of
SEI, take place in the initial cycles.[25]The thin carbon coating layer on the CZTS@C-15 could act as
a buffer
layer to prevent the aggregation of CZTS particles during cycling
and thus mitigate the pulverization problems. As compared to other
CZTS materials with a carbon content below 10%,[26] CZTS could be fully covered by 15% carbon and exhibits
the better electrochemical performance. Upon a further increase of
the carbon content (∼20 wt %), the CZTS@C-20 gives a lower
initial capacity, which could result from the excess carbon on the
CZTS.Figure d shows
that CZTS@C-15 cycled at current densities of 500, 1000, 2000, and
5000 mA/g exhibits the capacities of 1366 (after 50 cycles), 815,
600, and 447 mA h/g after 100 cycles. CZTS@C-15 shows a capacity of
∼220 mA h/g after 450 cycles at a high current rate of 5000
mA/g (Figure S5). Figure e shows the performance rate capability tests
of CZTS and CZTS@C-15 from 500 to 5000 mA/g. CZTS@C-15 exhibits high
discharge capacities of 1385, 1080, 825, and 460 mA h/g at current
densities of 500, 1000, 2000, and 5000 mA/g, respectively. CZTS@C-15
also shows good reversibility cycled at different current densities.
Moreover, we compare the cycling performance and mass loading of CZTS@C
materials with other CZTS-based materials reported in the literature
(Table S3). CZTS materials with different
structures have been investigated in LIBs.[11−13,16], However, the mass loading (0.2–0.3
mg/cm2) and areal capacity (0.2–0.5 mA h/cm2) of these anode materials are insufficient to meet the requirements
of advanced LIBs. Thus, the cycling performance and the stability
of CZTS@C materials with high mass loading obtained at high current
rates should be taken into consideration. Figure f shows that CZTS@C-15 with mass loadings
of 0.49, 0.77, 1.12, and 2.31 mg/cm2 exhibits areal capacities
of 0.61, 0.83, 1.45, and 2.45 mA h/cm2 at a current density
of 0.5 mA/cm2, respectively, suggesting that large areal
capacity and higher mass loading of CZTS@C-15 can be achieved. The
battery performance obtained at various current rates is promising
in the comparison of those reported silicon and germanium anodes.[27−29] The CZTS@C-15 anode material is also examined in the full cell with
a NMC532 cathode at a current density of 100 mA/g in the potential
window ranging from 1.5 to 4.5 V (Figure S6).To investigate the electrochemical behavior of CZTS-based
materials, Figure a,b shows cyclic
voltammetry (CV) measurements of the CZTS and CZTS@C electrodes during
five cycles. The potential window is between 0.005 and 3 V and the
scan rate is 0.2 mV s–1. The CV of the CZTS material
shows two reduction peaks at ∼1.5 and ∼1 V in the first
cycle (Figure a).
The reduction peak around 1.5 V can be attributed to lithium insertion
in CZTS to form a partial LiCZTS phase,
while the peak at 1.04 V may be related to the further conversion
of LiCZTS, which results in the formation
of Li2S, LiSn alloys as well
as the SEI layer, as previously reported.[10,11,30,31]
Figure 5
CV curves of
(a) CZTS and (b) CZTS@C-15 materials cycled at a potential
sweep rate of 0.2 mV s–1 in the potential range
ranging from 0.005 to 3 V.
CV curves of
(a) CZTS and (b) CZTS@C-15 materials cycled at a potential
sweep rate of 0.2 mV s–1 in the potential range
ranging from 0.005 to 3 V.XRD results of CZTS and CZTS@C-15 show that the CZTS has a smaller
particle size (Figure a). The current density of reduction peak at 1.04 V in CZTS is higher
than that of in CZTS@C-15 at the first cycle, which could result from
the higher surface area of CZTS. The CZTS material with a higher surface
area offers more reaction sites for more lithium-ion storage, which
is consistent with electrochemical behavior of CZTS and CZTS@C-15
during the first cycle (Figure a,b). The broad peak around 0.7 V could be associated with
the Li insertion process in various alloys such as Cu–Sn/Zn
and Li2S. The delithiation process of CZTS shows four oxidation
peaks at 0.5, 1.45, 1.9, and 2.3–2.4 V, which are associated
with conversion and delithiation processes.[32−34] The following
cycles show similar redox features, suggesting that the electrochemical
behavior of CZTS becomes stable. The redox features with larger current
density in the first cycle could result from the irreversible reactions.[35] The CV curves of CZTS@C and CZTS show similar
features, suggesting that reaction behavior of CZTS could be similar
to the CZTS@C and carbon coating has less contribution in the redox
reactions.We next examine the Li+-ion diffusion
kinetics of CZTS
and CZTS@C-15. Figure a,b shows CV curves of CZTS and CZTS@C-15 after 100 cycles at different
potential sweep rates. The relationship between the square root of
the scan rate (v1/2) and the anodic peak
current (ip) is illustrated in Figure c. The linear relationship
and corresponding slope of the ipversus v1/2 plots provide insights into the charge
storage kinetic behavior. In particular, the Randles–Sevcik
equation can be used to analyze the Li+-ion diffusion kinetics
in CZTS anode materials[36]
Figure 6
CV curves of (a) CZTS and (b) CZTS@C-15 after
100 cycles at different
potential sweep rates; (c) linear relationship between the anodic
peak current (ip) and square root of scan
rate (v1/2) of CZTS and CZTS@C-15 after
100 cycles.
CV curves of (a) CZTS and (b) CZTS@C-15 after
100 cycles at different
potential sweep rates; (c) linear relationship between the anodic
peak current (ip) and square root of scan
rate (v1/2) of CZTS and CZTS@C-15 after
100 cycles.In equation , n is the electron-transfer
number, S is
the electrode area (cm2), ip is the anodic peak current, D is the diffusion
coefficient of Li+, C is the bulk concentration
of Li+ in the electrode, and v is the
scan rate. The electrode area and bulk concentration used in this
study are 0.945 cm2 and 1, respectively. Based on the current
density of reduction peaks (R1 and R2), the slope obtained for the
CZTS@C-15 anodic current is higher than that of CZTS, indicating that
CZTS@C has a higher lithium-ion diffusion coefficient and faster electrochemical
kinetics.The electrochemical tests show that carbon-coated
CZTS with high
areal capacity exhibits a promising performance. To further understand
the stability of the CZTS material, the morphologies of CZTS and CZTS@C-15
after 100 cycles were analyzed by FE-SEM (Figure S7). The morphology of CZTS changes substantially after 100
cycles and the morphology of carbon-coated CZTS shows less changes
after cycling.
Electrochemical Reaction Mechanism
In order to study
the reaction mechanism of CZTS@C-15 during the reversible cycles, in situ XRD experiments were conducted for the cycled CZTS@C-15
material. Figure shows
the in situ XRD patterns of CZTS@C-15, which exhibits
an obvious peak shift and undergoes phase decomposition during the
charge–discharge process. A variety of species, such as Li2S and Cu–Zn/Cu–Sn alloy phases, is formed at
∼1 V during the lithiation process and the CZTS-associated
peak completely disappears at the end of the insertion process. During
the delithiation process, in situ XRD results show
that Li2S and Cu–Zn/Cu–Sn alloys are converted
to CZTS. Interestingly, the crystallinity of CZTS is improved after
cycling. The residual metal alloys remain after cycling, which correspond
to the irreversible reaction during the initial cycles.[37]
Figure 7
In situ XRD spectra of the CZTS@C-15
sample during
the discharge–charge process between 0.005 and 3 V.
In situ XRD spectra of the CZTS@C-15
sample during
the discharge–charge process between 0.005 and 3 V.
Formation of Li2S and Cu–Zn/Cu–Sn Metal
Alloys
The reaction mechanism of CZTS@C can be divided into
four regions. Figure shows the representative in situ XRD patterns of
CZTS@C taken from Figure . First, the early stage at ∼2.1 and ∼1.5 V
is attributed to the electrochemical activation of CZTS@C during the
lithiation process. The main diffraction peak corresponding to the
(112) plane slowly shifted from 28.54 to 28.49° degree during
the lithiation process (3.0 V-2.1 V), reflecting the volume expansion
of CZTS as well as the increase in lattice parameters. Thus, the insertion
process is obtained above ∼1.3 V. Interestingly, the (112)
diffraction peak became more intense and shifted toward higher angles
below 1.3 V (Figure a,b). This unexpected behavior is due to Li+ ions with
a larger covalent radius (1.33 Å) replacing smaller copper ions
(1.27 Å) in CZTS.[38,39] A higher bond distance for Li–S
(2.45 Å) compared to Cu–S (2.37 Å) is obtained, which
causes the increase in the lattice constant and particle size of CZTS
from 3 to ∼1.2 V (Table S4). In
addition, Figure c
shows that the broad peaks at ∼43°are attributed to the
formation of various metal alloys. Ex-situ XRD spectra
of CZTS@C obtained at 1.05 V and 0.5 V during the lithiation process,
suggesting that two broad peaks at ∼43° are attributed
to the Cu–Sn/Cu–Zn metal alloys (Figures S1 and S8).
Figure 8
Representative in situ XRD
results of the CZTS@C-15
taken from Figure during the (a–f) lithiation (3–0.005 V) and (g–i)
delithiation (0.005–3 V) processes.
Representative in situ XRD
results of the CZTS@C-15
taken from Figure during the (a–f) lithiation (3–0.005 V) and (g–i)
delithiation (0.005–3 V) processes.Second, the intensity of the characteristic CZTS@C peaks significantly
decreases when CZTS@C is further lithiated ∼1 V. Meanwhile,
the peaks associated with the Li2S phase appear at 27.1
and 31.1°, suggesting that a conversion reaction is taking place
at the low potential region.[40] Also, the
Li2S phase with a face-centered cubic structure becomes
more prominent. Figures d–f and S9 show the formation of
Cu6Sn5 and CuZn5 alloys during the
discharge process. Cu6Sn5 and CuZn5-associated peaks shift to lower angles, suggesting that the insertion
process of Cu6Sn5 and CuZn5 alloys
is proceeded. At below 0.5 V, the (111) Li2S-associated
peak becomes broader. Thus, in situ XRD results suggest
that CZTS can initially store lithium ions through insertion reactions
followed by the formation of Cu–Sn/Cu–Zn metal alloys
and Li2S through the conversion process. During the conversion
process, the insertion process of Cu–Sn/Cu–Zn metal
alloys seems take place at lower potential region. Because the formation
of Li4.4Sn has been proposed in the previous study,[11,41] some XRD peaks with weak intensity could be associated with the
Li22Sn5 structure.
Reformation of CZTS during
the Charge Process
Third, Figure g shows that the
Li2S-associated peak slowly shifts from lower angles (26.9°)
to higher angles (27.4°) during the charge (delithiation) process
from 0.15 to 0.9 V, indicating a volume contraction of Li2S. Fourth, Figure h,i shows that Li2S and the alloy-associated peaks disappear
and CZTS is formed from 1.1 to 3 V, suggesting that the conversion
and alloying reactions overlap upon cycling. The intensity of the
CZTS peaks became stronger and sharper, suggesting that the CZTS with
good crystallinity can be reformed during cycling.[13] In addition, in order to provide the insights into the
reaction mechanism of CZTS@C after cycling, ex situ XRD measurement of CZTS@C was performed. The representative ex situ XRD results of CZTS@C-15 show the formation of the
discharged species including alloys and Li2S after 50 cycles
(Figure S10). The XRD results suggest that
the reaction mechanism of CZTS@C is similar after cycling.High-resolution
XPS measurements were carried out to determine the valence states
of the elements in CZTS@C, as well as to confirm the lithium storage
mechanism proposed on the basis of the in situ XRD
data. Figure a shows
the Cu 2p, Zn 2p, Sn 3d, and S 2p XPS peaks of CZTS@C. The high-resolution
XPS spectra reveal that the CZTS sample contains Cu+, Zn2+, and Sn4+ states, which are consistent with the
previous studies.[42,43] XPS spectra of CZTS@C-15 show
the formation of Cu+/Cu0, Zn2+/Zn0, and Sn4+/Sn0 after the lithiation
process, suggesting that metal alloy species are formed after many
charge/discharge processes (Figure b).[13,42,44] Moreover, the binding energy in the sulfur spectrum shifts from
162 to 160 eV, which is attributed to the formation of Li2S.[45] After the delithiation process, XPS
results of CZTS@C-15 show that the oxidation state of Cu, Zn, and
Sn increases, which corresponds to the formation of CZTS (Figure c). Thus, the XPS
results are in a good agreement with XRD measurements. We next examine
the effect of carbon coating on the SEI composition. Li 1s, O 1s,
and F 1s XPS results show that Li2CO3, Li2O, and LiF are the main products in the SEI layer (Figure S11).[45,46] Previous studies
suggest that LiF is beneficial for the stable SEI layer. However,
XPS results of CZTS@C-15 and CZTS show that LiF is formed on both
CZTS@C-15 and CZTS materials. The SEI layer of CZTS@C-15 exhibits
a Li2CO3–rich layer, which could be beneficial
for enhancing the ionic conductivity of the SEI layer in subsequent
cycles, and thereby it results in good stability and rate capability.[45,47,48] The EIS results of CZTS and CZTS@C-15
show that the ionic conductivity of the SEI is improved on the CZTS@C-15
material (Figure S12 and Table S5).
Figure 9
High-resolution
Cu 2p, Zn 2p, Sn 3d, and S 2p XPS spectra of (a)
pristine CZTS@C-15 and (b,c) lithiated and delithiated CZTS@C-15.
High-resolution
Cu 2p, Zn 2p, Sn 3d, and S 2p XPS spectra of (a)
pristine CZTS@C-15 and (b,c) lithiated and delithiated CZTS@C-15.The possible reasons for the CZTS@C nanostructure
with higher areal
capacity and improved electrochemical performance are discussed. First,
the electronic conductivity of the electrode is improved. Second,
the carbon layer can protect the CZTS electrode from the electrolyte.
The side reaction could be suppressed and stable SEI can be formed.
Third, carbon coating could act as a superior conducting medium for
Li-ion migration and improve the stability performance even at a high
current rate.[51,52]
Reaction Mechanism of CZTS@C
Previous studies suggested
that the CZTS can be converted into Cu, Li1.5Zn, Li4.4Sn, and Li2S during the lithiation process.[11] The discharge capacity of CZTS is over 1100
mA h/g[11,15] and high discharge capacity (∼950
mA h/g) can be even achieved after 500 cycles.[49,50] Thus, the detailed mechanism is not fully understood yet. Because
the CV results of these CZTS materials show different redox features,
the CZTS materials synthesized by different methods could exhibit
different chemical formulas and electrochemical behaviors. In the
present study, the reaction mechanism of CZTS was examined using the in situ XRD measurement. Scheme shows that the CZTS@C electrode underwent
a multistep lithiation process through insertion, conversion, and
alloying reactions, leading to a LiCZTS/Li2S mixture and Cu–Sn/Zn alloy as intermediate phases.
The XRD results shows the formation of LiCZTS, suggesting that the insertion process takes place at the initial
stage. At the second stage, Li2S and various alloys, including
Cu–Zn/Cu–Sn, are formed during the alloying and conversion
processes. These alloys can store the lithium through an insertion
process below 1 V. Finally, the CZTS can be reformed during the delithiation
process. Because the formation of various products suggests that the
multielectron-transfer process takes place during cycling, we propose
that CZTS can accommodate 35 Li ions, which results in a theoretical
capacity of more than 2000 mA h/g. The detailed reactions are shown
in the Supporting Information. We should
note that other products with amorphous features or small quantity
could not be observed by using XRD. The formation of these products
could also show the multistep electron-transfer process. Thus, the
theoretical gravimetric capacity of CZTS could be even higher. The
CZTS material with uniform carbon coating allows rapid lithium ion
and electron transport, leading to a high lithium storage capacity
and improved stability of CZTS at high current rates.
Scheme 2
Schematic
Diagram for the Lithium-Ion Storage Reaction Mechanism
in CZTS@C
Conclusions
We
prepared carbon-coated CZTS nanoflowers with a 3D architecture
by a simple and cost-effective solvothermal method. The electrochemical
performance of the CZTS@C samples shows high gravimetric (1366 mA
h/g) and areal (2.45 mA h/cm2) capacities. Moreover, the
CZTS@C-15 sample exhibits a highly stable behavior over 100 cycles
even at high current rates. In situ XRD results reveals
that CZTS undergoes the insertion process at the initial stage, followed
by the alloying reaction and conversion reaction, which take place
at around 1 V. The crystallinity of CZTS can be further improved after
cycling. The CZTS nanoflower structure with an optimum amount of amorphous
carbon shows improved electronic/ionic conductivity and lithium-ion
diffusivity of the CZTS electrode material, which open up new opportunities
for developing high-performance anode materials with high rate capability.
Experimental
Section
Preparation of CZTS@C
The CZTS@C samples were prepared
by a two-step process. First, 2 mM copper (II) acetate dehydrate [Cu(OAc)2·xH2O], 1 mM zinc (II) acetate
[Zn(OAc)2], 1 mM tin (IV) chloride (SnCl4),
and 5 mM thiourea (CH4N2S) were dissolved in
70 mL of ethylene glycol. After constant stirring for 30 min, the
transparent homogeneous mixture was transferred into a Teflon-lined
stainless-steel autoclave and heated at 180 °C for 12 h, followed
by natural cooling to room temperature. The obtained CZTS nanoflowers
were collected by centrifugation, washed three times with deionized
(DI) water and ethanol, and then dried under vacuum at 70 °C
overnight. Then, different amounts (5–20 wt %) of glucose were
mixed with the CZTS nanoflower powder in a 1:1 DI water–ethanol
mixture solvent and dried at 70 °C under vacuum overnight. Finally,
the as-prepared samples were annealed at 100 °C for 1 h, followed
by 400 °C annealing for another hour at a heating rate of 10
°C/min under a nitrogen atmosphere, to obtain different carbon
coating thicknesses over the CZTS surface. The as-prepared CZTS nanoflowers
and the 5, 7, 10, 15, and 20 wt % carbon-coated CTZS samples were
denoted as CZTS, CZTS@C-5, CZTS@C-7, CZTS@C-10, CZTS@C-15, and CZTS@C-20,
respectively. The tap density of the CZTS-based materials is 2.5–3
g/cm3.
Materials Characterization
The structural
properties
of the CZTS and CZTS@C samples before and after cycling were analyzed
by XRD using a Bruker D2 diffractometer with Cu Kα radiation.
The surface morphology of the samples was investigated using FE-SEM
(JSM-6500). The microstructure, elemental distribution, and carbon
coating thickness of the CZTS@C samples were analyzed using a high-resolution
transmission electron microscope (JEOL2000FX), together with an FEI
Tecnai G2 electron microscope coupled to an energy-dispersive X-ray
spectrometer (Bruker AXS). To analyze the elemental composition and
oxidation state of the carbon-coated CZTS samples, XPS measurements
were performed with a monochromatic Al Kα source (1486.6 eV)
operated under ultrahigh vacuum. The peak positions of the obtained
XPS spectra were calibrated with respect to the C 1s peak position.
Finally, all spectra were analyzed by the XPSPEAKS41 software. EA
was carried out to determine the carbon content of the CZTS@C samples.
The Raman spectra of the CZTS and CZTS@C samples were measured by
a Horiba Jobin Yvon UV800 spectrometer with a 633 nm He–Ne
laser as the excitation source.
Electrochemical Measurements
Coin-type 2032 cells were
assembled in an argon-filled glovebox (LABstar, MBRAUN, Germany).
The water and oxygen levels in the glovebox were less than 1 ppm.
Metallic lithium (99.90%, Sigma-Aldrich) was used as the counter as
well as reference electrodes, while Celgard 2325 was employed as the
separator. The electrolyte used in the present work was 1 M LiPF6 in a mixture of ethylene carbonate, dimethyl carbonate (DMC),
and ethylmethylcarbonate with 1% vinyl carbonate (1:1:1 weight ratio,
LP-30, Merck, Germany). To prepare the working electrodes, the active
materials, super-P carbon black, and poly(vinylidene fluoride) (Kynar
2801) binder were added in a mass ratio of 75:15:10 in N-methyl-2-pyrrolidone (NMP) solvent and stirred overnight to form
a homogeneous slurry. The resulting paste was coated onto Cu foil
and then dried at 70 °C for 12 h under vacuum. The loading amount
of active materials was in the range of 1–2.3 mg/cm2. Electrochemical performance measurements were conducted by galvanostatic
charge/discharge tests performed in the potential range of 5 mV–3
V. Cyclic voltammetry (CV) measurements were carried out using a Solatron
analytical 1470E cell test system, in the potential range of 0.005–3
V at a scan rate of 0.2 mV/s. EIS measurements of CZTS-Li metal cells
were performed using a PARSTAT MC multichannel electrochemical workstation
(Solatron analytical 1470E cell test system) in the frequency range
from 1 MHz to 0.01 Hz. In all cells, the discharge capacity was calculated
based on the mass of the active material, including CZTS and carbon.
After cycling, the cells were disassembled in an argon-filled glovebox
and the CZTS electrodes were rinsed with a DMC solution to remove
residual salts for further characterization.
Authors: Nian Liu; Zhenda Lu; Jie Zhao; Matthew T McDowell; Hyun-Wook Lee; Wenting Zhao; Yi Cui Journal: Nat Nanotechnol Date: 2014-02-16 Impact factor: 39.213