Article
pubs.acs.org/cm
Copper Phosphate as a Cathode Material for Rechargeable Li
Batteries and Its Electrochemical Reaction Mechanism
Guiming Zhong,† Jingyu Bai,†,‡ Paul N. Duchesne,§ Matthew J. McDonald,† Qi Li,† Xu Hou,†
Joel A. Tang,∥ Yu Wang,Δ Wengao Zhao,¶ Zhengliang Gong,¶ Peng Zhang,§ Riqiang Fu,∥
and Yong Yang*,†,¶
†
State Key Lab of Physical Chemistry of Solid Surfaces, Collaborative Innovation Center of Chemistry for Energy Materials and
Department of Chemistry and ¶School of Energy Research, Xiamen University, Xiamen, Fujian 361005, China
‡
Shanghai Institute of Space Power-Sources, No. 2965 Dongchuan Road, Shanghai 200245, China
§
Department of Chemistry, Dalhousie University, Halifax, Nova Scotia B3H4R2, Canada
∥
National High Magnetic Field Laboratory, 1800 East Paul Dirac Drive, Tallahassee, Florida 32310, United States
Δ
Shanghai Synchrotron Radiation Facility, Shanghai Institute of Applied Physics, Chinese Academy of Science, Shanghai 201204,
China
S Supporting Information
*
ABSTRACT: In the search for new cathode materials for rechargeable lithium batteries, conversion-type materials have great
potential because of their ability to achieve high specific capacities via the full utilization of transition metal oxidation states. Here,
we report for the first time that copper phosphate can be used as a novel high-capacity cathode for rechargeable Li batteries,
capable of delivering a reversible capacity of 360 mAh/g with two discharge plateaus of 2.7 and 2.1 V at 400 mA/g. The
underlying reaction involves the formation as well as the oxidation of metallic Cu. The solid-state NMR, in situ XAFS, HR-TEM,
and XRD results clearly indicate that Cu can react with Li3PO4 to form copper phosphate and LixCuyPO4 during the charging
process, largely determining the reversibility of Cu3(PO4)2. This new reaction scheme provides a new venue to explore
polyanion-type compounds as high-capacity cathode materials with conversion reaction processes.
■
INTRODUCTION
Recently, many researchers have investigated several
materials, such as Li2MSiO4 and Cu2.33V4O11, which undergo
multiple-electron reactions, leading to higher capacities for
rechargeable Li batteries.10−16 However, the capacities of these
materials have not yet been entirely realized. Poizot et al. found
that Li2O can react with transition metals, M, in the charging
process, thus enabling nanosized transition metal oxide (MO)
particles (e.g., cobalt oxides) to exhibit high capacities by
utilizing the full valence states of the metal.17 Unfortunately,
among the MxNy family of conversion materials (M = Fe, Co,
Ni, Cu, etc.; N = N, O, F, P, S, etc.), only the MFx family has
Lithium ion batteries, a promising power system for hybrid and
pure electric vehicles, have been widely used in various types of
portable devices. Those important applications have driven the
global search for novel electrode materials with higher specific
capacities and energy densities.1−5 Certain promising anode
materials, such as Si and Sn, have shown a large capacity of up
to 3000 mAh/g.6−9 On the other hand, commercialized
intercalation-type cathode materials exhibit much lower
capacities, for example LiCoO2 (theoretical capacity: ∼ 274
mAh/g) and LiFePO4 (theoretical capacity: ∼ 170 mAh/g).
Thus, cathode materials have become a technical “bottleneck”
in the further development of Li-ion batteries with a high
energy density.
© 2015 American Chemical Society
Received: June 16, 2015
Revised: July 22, 2015
Published: July 24, 2015
5736
DOI: 10.1021/acs.chemmater.5b02290
Chem. Mater. 2015, 27, 5736−5744
Article
Chemistry of Materials
cell had been discharged (or charged) to a certain state, charging was
halted, and XAFS measurements were immediately performed. The
XAFS data at the Cu K-edge were collected in transmission mode at
room temperature, using ion chamber detectors at beamline BL14W1
of the Shanghai Synchrotron Radiation Facility (SSRF) and a Si(111)
double-crystal monochromator. The focal spot size at the position of
the sample was ∼0.25 mm. The monochromator was calibrated to
reject higher harmonics of the selected wavelength (harmonic content
<10−4), and data was collected over a range of energies, from 200 eV
below to 800 eV above the Cu K-edge (8979 eV). The incident
photon energy was calibrated using a standard Cu metal foil just prior
to data collection in all in situ XAFS measurements. Processing and
fitting of the XAFS data were performed using WinXAS software.36
Solid-State NMR. All NMR experiments were performed in a 9.4
T magnetic field with a Bruker Avance III spectrometer. The 7Li NMR
spectra were recorded using 1.3 mm probe-heads at a magic-angle
spinning speed of 60 kHz with a Hahn-echo pulse sequence (90° pulse
− τ − 180° pulse − τ). A recycle delay of 10 s and a 90° pulse length
of 1.3 μs were used for 7Li. A stepwise spin−echo pulse sequence with
a recycle delay of 0.01 s and a 90° pulse length pulse width of 1.6 μs
was applied to acquire 31P NMR spectra. The irradiation frequency
was varied with a step size of 0.15 MHz over the range of 1.8 MHz.
The spin−echo mapping 31P NMR spectra were obtained by summing
the individual spectra at different irradiation frequencies. The chemical
shifts of 7Li and 31P were calibrated using LiCl powder (0 ppm) and
adenosine diphosphate (1 ppm), respectively.
shown sufficiently high conversion voltages to make them
possible candidates for use as rechargeable Li battery
cathodes.5,18−27 However, it is still important to explore new
types of conversion-type cathode materials with different
chemistries.
Many works have demonstrated that CuO can be used as an
electrode material for lithium ion batteries, showing a relatively
large capacity and good cycling performance.17,28 Replacement
of oxide ions by phosphate may increase the conversion
voltage, due to the inductive effect. Therefore, in this work, we
revisit copper phosphate, a well-known cathode for primary
batteries.29,30 To the best of our knowledge, this material has
never been reported in the literature as a rechargeable cathode
material. We observed for the first time that, by utilizing the
Cu2+/Cu0 redox reaction, a carbon-coated amorphous copper
phosphate cathode was capable of delivering a reversible
capacity larger than 360 mAh/g at two discharge plateaus of 2.7
and 2.1 V with a high current density of 400 mA/g within 10
cycles. Additionally, both amorphous and crystalline materials
were studied, showing that the material with an amorphous
state delivered better electrochemical performance. This may
be due to a more uniform distribution of reduction products.
Furthermore, we employed several techniques to understand
the detailed reversible electrochemical reaction mechanism for
this new type of rechargeable material. X-ray diffraction (XRD)
and high-resolution transmission electron microscopy (HRTEM) were used to observe the morphology and fine structure
of the material; high-resolution solid-state 7Li and 31P magic
angle spinning (MAS) nuclear magnetic resonance (NMR)
spectroscopy was used to track the local structure and phase
changes in the material;31−33 while X-ray absorption fine
structure (XAFS) spectroscopy34,35 was implemented to
investigate the oxidation state changes of Cu.
■
■
RESULTS
Electrochemical Performance of the Cu3(PO4)2/C
Cathode. The rechargeability of transition metal phosphate
as a conversion cathode material was demonstrated first. Figure
1a shows the discharge/charge profiles of the crystalline
Cu3(PO4)2/C material operated between 1.5 and 4.8 V at a
specific current of 10 mA/g. A capacity of up to 425 mAh/g
(theoretical capacity: 423 mAh/g) was observed in the first
EXPERIMENTAL SECTION
Materials. Crystalline Cu3(PO4)2 powders were prepared by the
solid-state reaction of stoichiometric amounts of (NH4)2HPO4 and
CuO at 900 °C. The amorphous Cu3(PO4)2 were prepared through
precipitation of (NH4)2HPO4 and Cu(NO3)2 aqueous solutions,
followed by calcining at 500 °C. High-energy ball milling of
Cu3(PO4)2 and acetylene black was utilized to produce the carboncoated material Cu3(PO4)2/C (90 wt % Cu3(PO4)2). The electrodes
were then fabricated using 80 wt % Cu3(PO4)2/C, 10 wt % acetylene
black, and 10 wt % PVDF, with NMP as a solvent. The loading of the
active material was around 2 mg/cm2. The electrochemical performances of the cathode materials were examined using CR2025 coin-type
cells. These cells were assembled in an argon-filled glovebox and
included the prepared electrode as the cathode, lithium metal as the
anode, 1 M LiPF6 in ethylene (EC)/dimethyl carbonate (DMC) as the
electrolyte, and Celgard 2300 as the separator. After the cells had
finished cycling, they were disassembled, and the electrodes were
washed using DMC in an argon-filled glovebox. Finally, the electrodes
were dried at room temperature and used for the ex situ study.
XRD and TEM. The XRD patterns of samples were collected using
a Panalytical X-pert diffractometer (PANalytical, Netherlands)
equipped with a Cu Kα radiation source and operating at 40 kV and
30 mA. In order to prevent oxidation of the samples, electrodes were
covered with Kapton brand tape in the glovebox prior to analysis. The
Kapton tape generated a broad but weak peak between 10° and 35°,
but there was no effect on signal collection from the electrodes. TEM
and high resolution TEM imagery was collected with a Tecnai F20 ST
TEM operating at 200 kV. The samples were loaded on copper grids
in an argon-filled glovebox, sealed in an argon-filled transfer holder,
and then transferred into the TEM column.
XAFS. A standard CR2025 coin cell with small holes in each side
was used to perform the in situ XAFS measurements. Once the in situ
Figure 1. Electrochemical profiles of carbon coated crystalline (a) and
amorphous (b) Cu3(PO4)2 materials under 10 and 400 mA/g,
respectively. The inset figures are the cycling performance of the
materials under 400 mA/g.
5737
DOI: 10.1021/acs.chemmater.5b02290
Chem. Mater. 2015, 27, 5736−5744
Article
Chemistry of Materials
discharge, suggesting that the transfer of two electrons per Cu
atom takes place and Cu2+ is completely reduced to Cu0 upon
completion of the discharge. In the second cycle, the crystalline
Cu3(PO4)2 material displayed a reversible capacity of only 290
mAh/g (with 4.1 electrons per formula unit) in the second
cycle and decayed quickly (Figure S1). Nevertheless, as shown
in Figure 1b, the amorphous copper phosphate material
exhibited much better cycling performance, with a reversible
capacity larger than 360 mAh/g at 400 mA/g between 1.5 and
4.8 V (Figure 1b and Figure S1) within 10 cycles, meaning that
5.1 electrons per Cu3(PO4)2 unit, or 1.7 electrons per Cu atom,
are reversibly transferred. The capacity decayed to 220 mAh/g
after 30 cycles, which may be due to the contact problem of
reduction products and dissolution of Cu.37,38The better
cycling performance of the amorphous sample may be due to
the smaller particle sizes of discharge products, copper and
lithium phosphate in the amorphous sample, providing better
physical contact between Cu and Li3PO4, thus facilitating the
recharge reaction. Therefore, the cycling performance of the
materials may be further improved through optimizing particle
size and morphology.28,39 It should be noted that there were
two clearer discharge plateaus, at 2.7 and 2.1 V, in the second
discharge process, which may be due to the morphology
changes. From the crystal structure of Cu3(PO4)2 (Figure 2), it
Figure 3. (a) XRD patterns of crystalline Cu3(PO4)2 and Cu3(PO4)2/
C. (b) Ex situ XRD patterns of LixCu3(PO4)2 in different lithiation
states.
reoxidized during the charge process. The discharge curve of
the electrode (Figure 1) supports the claim of the complete
conversion of Cu3(PO4)2 to Li3PO4 and Cu metal when the
discharge process is finished. However, no other new Bragg
peaks were visible during cycling other than that at 43.3°
(Cu(111)), indicating that Li3PO4 and any other products
formed on recharging are amorphous and making it very
difficult to obtain further structural information regarding any
possible products and/or reaction intermediates from the
powder XRD patterns. Nevertheless, the ex situ XRD study
confirmed that Cu3(PO4)2 was converted into copper during
the discharge process, while the copper was reoxidized during
the charge process.
Fine Structure and Differences between Amorphous
and Crystalline Materials Observed with TEM. Figure 4
shows TEM images of both amorphous and crystalline
materials at different stages (i.e., before discharge, after first
discharge, and after first recharge). Figure 4a shows that the
pristine amorphous (main image) material exhibits a large
degree of disorder, and the crystalline (inset) Cu3(PO4)2
sample possesses high crystallinity. Figure 4b demonstrates
the submicron level structure of the discharged amorphous
material. This structure consisted of often-interconnected
particles, each several hundred nm in diameter and with a
carbon coating of varied thickness (typically 5−20 nm). These
larger particles were in turn formed from fairly uniform Cu
nanoparticles, approximately 5 nm in diameter, and surrounding crystallites of Li3PO4, which can be seen in the image as
dark “dots” and lighter regions, respectively. The discharged
crystalline samples shared this general structure, but the
constituent Cu nanoparticles exhibited a greater size as well
Figure 2. Crystal structure of Cu3(PO4)2 with the P-1 space group.
can be seen that there is no vacant site in the Cu3(PO4)2
structure into which Li may insert, suggesting that a conversion
reaction occurs even in the initial discharge process. Thus, the
two discharge plateaus in the second discharge process may be
due to different conversion reactions. As a new type of material
for rechargeable Li batteries, the reaction mechanism may give
us some important information. In order to understand the
reactions of this new type of cathode, especially the two
discharge plateaus and the recharging process, the crystalline
material was examined in more detail, because of its high
crystallinity and purity.
Crystal Phase Evolution Detected Using XRD. XRD is a
very useful tool for monitoring crystalline phase changes during
electrochemical reactions. Figure 3a depicts the XRD patterns
of crystalline Cu3(PO4)2 and Cu3(PO4)2/C. A broader half
peak width appeared in the XRD pattern of Cu3(PO4)2/C,
indicating that smaller particles were formed after ball milling,
which brought about the rechargeable property. Figure 3b and
Figure S2 are the XRD patterns of the material during the first
two discharge/charge cycles. It can be clearly seen that the
intensities of the XRD peaks corresponding to Cu3(PO4)2
decreased, and a new and broad Bragg peak at 43.3° (attributed
to Cu (111)) was detected when the x value in LixCu3(PO4)2
reached 2 during the first discharge, indicative of the
decomposition of Cu3(PO4)2 and the formation of metallic
Cu during discharge. In the recharge process, the intensity of
the 43.4° peak decreased, indicating that the copper was being
5738
DOI: 10.1021/acs.chemmater.5b02290
Chem. Mater. 2015, 27, 5736−5744
Article
Chemistry of Materials
Figure 4. (a) TEM image of pristine amorphous material and pristine crystalline material (inset). (b) Submicron structure of amorphous material
upon first discharge. (c) and (d) are nanoscale structures of amorphous and crystalline material upon first discharge and first recharge, respectively.
Local Structure Changes during Cycling Detected by
Solid-State NMR. High-resolution solid-state NMR is a
powerful method for understanding the local structure of
materials, especially those in amorphous states. Figure 5 shows
the mass-normalized 7Li MAS NMR spectra of crystalline
Cu3(PO4)2/C during the first discharge and second charge
processes. The inset is the enlarged view of selected states.
Upon discharge, two peaks at −0.3 and −13 ppm were
observed, as shown in Figure 5a (x = 0.3, 2.7 V). The peak at
−13 ppm can be assigned to the Li close to Cu2+ (3d9), as the
unpaired electrons from paramagnetic Cu2+ causes a large shift
due to Fermi-contact interactions.31 The peak at −0.3 ppm was
due to the Li located in a diamagnetic environment. In addition,
it can be clearly seen that the line-width of the −0.3 ppm peak
has a trend of being reduced during discharge. Since the
material was completely converted to Cu metal and Li3PO4, the
peak at −0.3 ppm in Figure 5a (x = 6.0) could be confidently
assigned to Li3PO4. The −0.3 ppm peak with a relatively
broader line-width for other discharge states (x = 0.3, 1.0, 2.0)
may indicate the formation of another diamagnetic environment during the discharge process, such as Cu+ (3d10), with
strong evidence for this provided by XAFS results (Figure 8a).
Figure 5b presents the 7Li NMR spectra during the charge
process in the second cycle. The decrease in the intensity of the
peak at −0.3 ppm indicated that the Li in Li3PO4 deintercalated
during the charge process. As shown in the inset of Figure 5b,
as a wider size distribution, which are likely due to the larger
contiguous Cu3(PO4)2 regions in the pristine material. Figure
4c shows this structural model, in which the Li3PO4 and Cu
regions are labeled in the amorphous sample that underwent a
single discharge. The dark, circular areas correspond to Cu and
the lighter surroundings, Li3PO4. The Li3PO4 was in the form
of small crystallites for both amorphous and crystalline samples,
although it was difficult to determine if the average Li3PO4
crystallite size differed between these two. Finally, Figure 4d
shows the crystalline sample after the first recharge. Areas of Cu
and Cu3(PO4)2 are seen, demonstrating the incomplete
reaction of Cu with surrounding Li3PO4 and the presence of
substantially larger Cu nanoparticles compared to amorphous
samples, which is in accordance with the XAFS results (Figure
8a). However, the TEM imagery of amorphous samples after
the first recharge did not display any crystal planes of
Cu3(PO4)2, indicating that the recharged material was highly
amorphous. This TEM data is consistent with the previous
observation: the degree of disorder of Cu3(PO4)2 crystals in the
pristine material had an impact on the charge capacity of the
sample. The material with higher crystallinity formed larger Cu
nanoparticles upon discharge and larger regions of Cu3(PO4)2
upon recharge, presumably interfering with Li ion transport
during the reaction process and leaving “islands” of unreacted
material dispersed throughout the structure.
5739
DOI: 10.1021/acs.chemmater.5b02290
Chem. Mater. 2015, 27, 5736−5744
Article
Chemistry of Materials
Figure 5. Mass-normalized 7Li solid-state NMR spectra of crystalline
Cu3(PO4)2/C during the first discharge (a) and second charge (b)
processes. The values on the left axis refer to the x value in
LixCu3(PO4)2. The inset figures are the enlarged views.
Figure 6. Mass-normalized 31P MAS NMR spectra of crystalline
Cu3(PO4)2/C during the first discharge process. The inset figure is an
enlarged view.
when the material was charged to 2.7 V only one peak at −0.3
ppm with decreased intensity and a slightly broader line-width
was observed compared to the discharged sample, which may
be due to the formation of Cu+. Another peak at −13 ppm
could be seen when the voltage reached 3.3 V. As mentioned
before, the 7Li NMR peak at −13 ppm was due to the Li
located around the Cu2+. Thus, we can conclude that Cu was
oxidized to Cu+ when the voltage was below 2.7 V, while Cu2+
was formed afterward. However, the intensity of the peak at
−13 ppm did not increase in the following charge process,
indicating that another compound without Li formed. The
discharge in the second cycle shows reaction processes similar
to the first cycle, as shown in Figure S3. The intensity of the
peak at −13 ppm increased when the material discharged to 2.7
V and decreased when the voltage was 2.0 V, indicating that
Cu2+ and Li+ contained intermediates that primarily formed
when the discharge voltage was larger than 2.7 V and converted
into Li3PO4 when the discharge voltage was lower than 2.5 V.
Figure 6 shows the mass-normalized 31P NMR spectra of
crystalline materials at different discharge states, giving more
detailed information regarding the reaction products. The 31P
peaks between 1500 and 2000 ppm were assigned to crystalline
Cu3(PO4)2. The 31P resonance of paramagnetic Cu3(PO4)2 will
shift under different temperatures, and although the room
temperatures when testing different samples were the same, the
real temperatures of the samples in the high-spinning rotors
may have been different, due to the unavoidable presence of Al
pieces scraped from the electrodes. Differing amounts of Al
cause different temperature rises under a spinning frequency of
60 kHz, resulting in different chemical shifts observed for
Cu3(PO4)2. However, the same 31P shifts were obtained under
a low spinning frequency of 25 kHz, as shown in Figure S4.
Figure 6a depicts the 31P NMR spectra of the crystalline
material when the x value in LixCu3(PO4)2 was less than 2. In
addition to the peak assignable to crystalline Cu3(PO4)2, a peak
at around 9 ppm, whose line-width at half height decreased
with discharge, can be assigned to Li3PO4 and diamagnetic
species, which was in accordance with the 7Li NMR results.
Therefore, we can conclude that the diamagnetic intermediate
formed in the discharge process was LixCu(I)yPO4.40 The
reduction in the line-width may have been due to the
decomposition of LixCu(I)yPO4 and the formation of Li3PO4.
Additionally, a broad 31P peak at around 400 ppm was
observed, which was due to the P close to Cu2+. The intensity
of the broad peak at around 400 ppm increased with discharge
when the x value was less than 2 and then decreased in the
following discharge process. When the x value reached 5, the
peak almost vanished, as shown in Figure 6b.
Figure 7a shows the 31P NMR spectra of the material during
the charge process in the second cycle. It can be seen that the
Figure 7. Mass-normalized 31P MAS NMR spectra of crystalline
Cu3(PO4)2/C during the charge (a) and discharge (b) processes in the
second cycle.
broad peak at around 400 ppm was obtained again when the
charge voltage reached 3.3 V. Considering that almost the same
trend of the 7Li peak at −13 ppm and the broad31P peak at
around 400 ppm can be found during the first and second
cycles, it is reasonable to believe that the Li−Cu(II)−P−O
compound formed during the discharge/charge processes. The
general material formula LixCu(II)yPO441,42 will be used here.
This LixCuy(II)PO4 formed mainly when the charging voltage
was lower than 3.3 V. However, another compound formed in
the following charge process. As shown in Figure 7a, a very
5740
DOI: 10.1021/acs.chemmater.5b02290
Chem. Mater. 2015, 27, 5736−5744
Article
Chemistry of Materials
Figure 8. (a) The first derivative curves of in situ XANES spectra (black) and the linear combination fitting curves (red), obtained using the first
derivative XANES curves of Cu3(PO4)2 (x = 0) and Li6Cu3(PO4)2 as standards. The vertical axis (x) indicates the corresponding value in
LixCu3(PO4)2. (b) Discharge/charge profiles and fitting results (coordination numbers and bond lengths) for Cu−O and Cu−Cu from the in situ
EXAFS data.
broad 31P peak between 500 and 3000 ppm was observed when
the material was charged to 4.0 V. The ex situ 7Li NMR spectra
during the charge process demonstrates that a compound
without Li was possibly formed when the material was charged
to 4.0 V. Additionally, the 31P shift of amorphous copper
phosphate was observed to be between −500 and 4000 ppm
(Figure S5). Therefore, the broad peak between 500 and 3000
ppm can be assigned to the amorphous copper phosphate. In
other words, amorphous copper phosphate formed when the
charge voltage reached 4.0 V. Figure 7b is the 31P NMR spectra
of carbon coated crystalline materials during the second
discharge process. It can be seen clearly that the intensity of
the signal from the amorphous copper phosphate decreased
with discharge, and the peak from LixCu(II)yPO4 was reduced
when the voltage was up to 2.0 V, which was consistent with
the 7Li NMR results. It should be noted that a recycle delay of
0.01 s was applied for the acquisition of 31P NMR spectra,
causing a suppression of the signal of the peak at 9 ppm from
Li3PO4 and LixCu(I)yPO4, which showed a very long spin−
lattice relaxation time (T1), larger than 10 s. However, the
relaxation times of the 31P peaks corresponding to copper
phosphate and LixCu(II)yPO4 were much less than 1 ms,
allowing a full relaxation when using a recycle delay of 0.01 s.
The 31P NMR spectra at different states using a recycle delay of
60 s are shown in Figure S6.
The high-resolution solid-state 7Li, 31P MAS NMR spectra
thus demonstrate that two intermediate compounds, LixCu(I)yPO4 and LixCu(II)yPO4, were formed during the first
discharge process. When x was larger than 2, the conversion
reaction of Cu3(PO4)2 and LixCu(I, II)yPO4 into Li3PO4 was
mainly occurring. The charge process of the material
underwent three reaction processes. First, LixCu(I)yPO4
formed, while the voltage was less than 2.7 V. Second, when
the voltage was between 2.7 and 3.3 V, the formation of
LixCu(II)yPO4 dominated. Finally, when the voltage was above
3.3 V, amorphous copper phosphate was the primary reaction
product. In addition, the second discharge process looks like
the reverse of the charge process. LixCu(I)yPO4 and LixCu(II)yPO4 were formed when the discharge voltage was larger
than 2.7 V, and LixCu(II)yPO4 started to convert into Li3PO4
when the voltage was lower than 2.5 V.
Tracking Cu Oxidation States and Local Structure
Changes Using in Situ XAFS. The XAFS study gives a direct
view of the oxidation of copper. Figure 8a shows first-derivative
curves of in situ X-ray absorption near-edge structure (XANES)
spectra (black) and linear combination fitting curves (red),
obtained using the first-derivative XANES curves of the
crystalline Cu 3 (PO 4 ) 2 electrode and fully discharged
Li6Cu3(PO4)2 as standards. It can be clearly seen that in
addition to the peaks located at 8979 eV (Cu) and 8984 eV
(Cu2+), a signal was also observed at 8981.4 eV in the
discharge/charge spectra, which can be attributed to the Cu+
ion.43,44 Due to the fact that the valence state of copper in
Cu3(PO4)2 is Cu2+, we can deduce that the Cu+ is present in
the intermediate product, LixCu(I)yPO4. These results agree
well with evidence from NMR testing. The first-derivative
spectra and XANES spectra (Figure S7) reveal that the
formation of Cu+ in the first discharge process occurred mainly
before the x value in LixCu3(PO4)2 reached 0.5 and decreased
slowly until disappearing when x increased to 5.5. Combining
these results with the observation of the continuously
increasing intensity of the 7Li peak at −0.3 ppm (Figure 5a)
and 31P peak at 9 ppm (Figure 6a), we can tell that Li3PO4
formed continuously during discharge and was the main
discharge product during the whole discharge process. During
charging, the amount of Cu+ was found to gradually increase as
x decreased, demonstrating the continuous formation of a small
amount of LixCu(I)yPO4 during the charge process. Of course,
5741
DOI: 10.1021/acs.chemmater.5b02290
Chem. Mater. 2015, 27, 5736−5744
Article
Chemistry of Materials
Figure 9. Reaction scheme of a crystalline Cu3(PO4)2/Li battery during the first two cycles. CP, LC(II)P, LC(I)P, and LP refer to copper phosphate,
LixCu(II)yPO4, LixCu(I)PO4, and Li3PO4, respectively.
the main reaction was the oxidation of Cu to Cu2+, which is
shown clearly in Figures S7 and S8. Additionally, the peak at
8979 eV was observed in the XANES spectrum of
Li0.5Cu3(PO4)2, indicating that Cu metal formed as soon as
the battery began discharging; thus, a complex reaction
mechanism involving Cu metal formation could be anticipated.
Additionally, the XAFS data shows that the local environment
of Cu changes during the discharge/charge process. Figure 8b
graphs the bond lengths of Cu−O and Cu−Cu, as well as their
coordination number (CN), during cycling. Both Cu−O and
Cu−Cu bond lengths remained relatively unchanged throughout this process; however, the CN of Cu−O decreased with
discharging, while that of Cu−Cu increased. On charging, these
trends were reversed. The increasing CN of Cu−O
demonstrated that Cu was oxidized to a Cu−O containing
compound, in accordance with the NMR results. Furthermore,
the decreasing CN of the Cu−Cu unit indicated that the
particle size of Cu decreased with charging, which was due to
the reaction of Li3PO4 and Cu to form LixCuyPO4. Of note is
the fact that the Cu−Cu CN for the discharged material was
about 9, which was less than that of bulk Cu (12), indicating
that Cu metal nanoparticles were formed. Furthermore, some
Cu−Cu bonding was retained even after the battery was fully
charged, indicating that some unreacted Cu was still present
even at 4 V, which is consistent with the TEM results.
The XAFS study confirmed the existence and change of Cu+
in the intermediates LixCu(I)yPO4 during the cycling of
crystalline copper phosphate. Additionally, it revealed that a
Cu mass was formed upon discharge, and then was reoxidized
to Cu+ and Cu2+ during the charging process. The analysis
results from XAFS are highly consistent with those from NMR.
capacity of larger than 300 mAh/g within 20 cycles at high
current density. The reversibility of the material is very sensitive
to the particle size; the as-prepared crystalline sample shows
almost no cycle performance. However, the ball-milled carbon
coated sample, which has smaller particles, displays reversibility.
In addition, the amorphous material has a better rechargeability.
Considering the different morphologies of the discharged
crystalline and amorphous samples as given by TEM results, we
can deduce that the rechargeability of Cu3(PO4)2 depends
strongly on the particle sizes of the discharge products but not
that of the pristine samples (Figures S10 and S11). It is
speculated that the smaller particle size of the Cu and Li3PO4
phases formed during discharge leads to a more uniform
reaction product distribution and provides better contact
between Cu and Li3PO4, thus facilitating the recharge reaction
(oxidation of Cu). In addition, the high diffusity of copper ions
in the polyanion framework could be another factor that affects
the reversibility of the conversion reaction in Cu3(PO4)2.45
Although the capacity of the material currently decays to 220
mAh/g within 30 cycles, it is likely to be improved after
introducing morphology and size control.28,46
Several local structure characterization techniques such as
XAFS, NMR, and XRD have been carried out to probe the
crystal and local structure changes in the nanostructured
Cu3(PO4)2 during the discharge/charge processes and thus to
understand these complex conversion reactions, as summarized
in Figure 9. In the first discharge process, small amounts of
LixCu(I)yPO4 were formed when the x value in LixCu3(PO4)2
was less than 0.5, material that was gradually reduced to Cu and
Li3PO4 in the following discharge process. The result was
strongly supported by in situ XAFS study. However, this is only
a side reaction. Li3PO4, LixCu(II)yPO4, and Cu are the three
main products when the x value in LixCu3(PO4)2 is less than 2.
Moreover, the reduction of both Cu3(PO4)2 and LixCu(II)yPO4
to Li3PO4 and Cu primarily occurred during subsequent
discharge processes. In the charge process, Li3PO4 and Cu were
converted to LixCu(I)yPO4 when the voltage was less than 2.7
V. Subsequently, LixCu(II)yPO4 was primarily formed until the
voltage reached 3.3 V, where amorphous copper phosphate
started to generate. The second discharge curve showed two
clear plateaus at 2.7 and 2.1 V. However, the reactions in the
■
CONCLUSION
The Cu3(PO4)2 material has superior characteristics that make
it suitable, with some additional process tuning, as a cathode
material for lithium ion batteries. Due to the inductive effect of
the phosphate group, the Cu3(PO4)2 shows a higher working
voltage of 2.7 and 2.1 V compared to CuO (1.5 V).
Furthermore, it is possible to control/select the discharge
voltage by changing the transition metal and anion group
materials. Presently, the material can deliver a reversible high
5742
DOI: 10.1021/acs.chemmater.5b02290
Chem. Mater. 2015, 27, 5736−5744
Article
Chemistry of Materials
second discharge process were similar to the first discharge.
Amorphous copper phosphate converted to LixCu(II)yPO4,
Li3PO4, and Cu before 2.7 V, and then LixCu(II)yPO4 and
amorphous copper phosphate were reduced to Cu and Li3PO4.
The discharge plateau at 2.7 V in the second cycle had larger
capacity than that in the first cycle, which may be due to
morphology change after the first discharge. A surface with
higher specific area may form, facilitating the reaction. It is wellknown that the transition metal oxides, MO, can be used as
conversion electrodes for lithium ion batteries. However, the
low conversion voltages of MO have thus far been insufficient
for use as cathode materials. By contrast, polyanion compounds
display a higher conversion voltage because of the stronger
inductive effect of the polyanion groups. Our results indicate
that transition metals, M, such as Cu, and Li3PO4 can be
converted into LixMPO4, which holds important implications
for the further optimization of these materials and the design of
new polyanion-type cathode materials with large specific
capacities.
■
are given for insightful comments made by Prof. Clare P. Grey
about the NMR results.
■
ASSOCIATED CONTENT
S Supporting Information
*
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.chemmater.5b02290.
Detailed information for the whole set of XRD patterns,
NMR spectra, TEM figures, XANE spectra, and fitting
curves of extended k3-weighted EXAFS spectra during
cycling of the material (PDF)
■
REFERENCES
(1) Whittingham, M. S. Lithium batteries and cathode materials.
Chem. Rev. 2004, 104, 4271−4302.
(2) Armand, M.; Tarascon, J.-M. Building better batteries. Nature
2008, 451, 652−657.
(3) Bruce, P. G.; Scrosati, B.; Tarascon, J.-M. Nanomaterials for
Rechargeable Lithium Batteries. Angew. Chem., Int. Ed. 2008, 47,
2930−2946.
(4) Goodenough, J. B.; Park, K.-S. The Li-Ion Rechargeable Battery:
A Perspective. J. Am. Chem. Soc. 2013, 135, 1167−1176.
(5) Cabana, J.; Monconduit, L.; Larcher, D.; Palacin, M. R. Beyond
intercalation-based Li-ion batteries: the state of the art and challenges
of electrode materials reacting through conversion reactions. Adv.
Mater. 2010, 22, E170−92.
(6) Li, H.; Huang, X.; Chen, L.; Wu, Z.; Liang, Y. A high capacity
nano Si composite anode material for lithium rechargeable batteries.
Electrochem. Solid-State Lett. 1999, 2, 547−549.
(7) Chen, H.; Xiao, Y.; Wang, L.; Yang, Y. Silicon nanowires coated
with copper layer as anode materials for lithium-ion batteries. J. Power
Sources 2011, 196, 6657−6662.
(8) Chan, C. K.; Peng, H.; Liu, G.; McIlwrath, K.; Zhang, X. F.;
Huggins, R. A.; Cui, Y. High-performance lithium battery anodes using
silicon nanowires. Nat. Nanotechnol. 2008, 3, 31−35.
(9) Ng, S. H.; Wang, J.; Wexler, D.; Konstantinov, K.; Guo, Z. P.; Liu,
H. K. Highly Reversible Lithium Storage in Spheroidal Carbon-Coated
Silicon Nanocomposites as Anodes for Lithium-Ion Batteries. Angew.
Chem., Int. Ed. 2006, 45, 6896−6899.
(10) Melot, B. C.; Tarascon, J.-M. Design and Preparation of
Materials for Advanced Electrochemical Storage. Acc. Chem. Res. 2013,
46, 1226−1238.
(11) Rangappa, D.; Murukanahally, K. D.; Tomai, T.; Unemoto, A.;
Honma, I. Ultrathin nanosheets of Li2MSiO4(M= Fe, Mn) as highcapacity Li-ion battery electrode. Nano Lett. 2012, 12, 1146−1151.
(12) Lv, D.; Bai, J.; Zhang, P.; Wu, S.; Li, Y.; Wen, W.; Jiang, Z.; Mi,
J.; Zhu, Z.; Yang, Y. Understanding the High Capacity of Li2FeSiO4: In
situ XRD/XANES study combined with first-principles calculations.
Chem. Mater. 2013, 25, 2014−2020.
(13) Cheng, F.; Chen, J. Transition metal vanadium oxides and
vanadate materials for lithium batteries. J. Mater. Chem. 2011, 21,
9841−9848.
(14) Morcrette, M.; Rozier, P.; Dupont, L.; Mugnier, E.; Sannier, L.;
Galy, J.; Tarascon, J.-M. A reversible copper extrusion−insertion
electrode for rechargeable Li batteries. Nat. Mater. 2003, 2, 755−761.
(15) Whittingham, M. S.; Song, Y.; Lutta, S.; Zavalij, P. Y.; Chernova,
N. A. Some transition metal (oxy) phosphates and vanadium oxides
for lithium batteries. J. Mater. Chem. 2005, 15, 3362−3379.
(16) Gong, Z.; Yang, Y. Recent advances in the research of
polyanion-type cathode materials for Li-ion batteries. Energy Environ.
Sci. 2011, 4, 3223−3242.
(17) Poizot, P.; Laruelle, S.; Grugeon, S.; Dupont, L.; Tarascon, J.
Nano-sized transition-metal oxides as negative-electrode materials for
lithium-ion batteries. Nature 2000, 407, 496−499.
(18) Ma, D.-l.; Cao, Z.-y.; Wang, H.-g.; Huang, X.-l.; Wang, L.-m.;
Zhang, X.-b. Three-dimensionally ordered macroporous FeF3 and its
in situ homogenous polymerization coating for high energy and power
density lithium ion batteries. Energy Environ. Sci. 2012, 5, 8538−8542.
(19) Kim, S.-W.; Seo, D.-H.; Gwon, H.; Kim, J.; Kang, K. Fabrication
of FeF3 Nanoflowers on CNT Branches and Their Application to High
Power Lithium Rechargeable Batteries. Adv. Mater. 2010, 22, 5260−
5264.
(20) Badway, F.; Mansour, A. N.; Pereira, N.; Al-Sharab, J. F.;
Cosandey, F.; Plitz, I.; Amatucci, G. G. Structure and Electrochemistry
of Copper Fluoride Nanocomposites Utilizing Mixed Conducting
Matrices. Chem. Mater. 2007, 19, 4129−4141.
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected].
Author Contributions
G.Z. and J.B. contributed equally. G.Z., J.B., and Q.L. carried
out the materials preparation and electrochemical test. G.Z. and
X.H. acquired the NMR data, G.Z., R.F., and J.A.T. analyzed
the NMR data. G.Z., W.Y., and Z.G. carried out the XAFS
experiments. G.Z., P.N.D, and P.Z. processed and analyzed the
XAFS data. M.M. and W.Z. recorded and analyzed the TEM
results. Q.L. acquired and processed the XRD data. Y.Y. and
G.Z. proposed the research, and Y.Y. attained the main financial
support for the research.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
The authors acknowledge the financial support of their research
from the National Natural Science Foundation of China
(Grants No. 21233004, 21474138, 21428303, and 21321062)
and National Basic Research Program of China (973 program,
Grant No. 2011CB935903). In particular, R. Fu is indebted as a
PCOSS fellow supported by the State Key Lab of Physical
Chemistry of Solid Surfaces, Xiamen University, China. P.
Zhang thanks NSERC Canada for supporting the XAFS
analysis program. J. A. Tang and R. Fu are grateful for financial
support from the User Collaboration Grants Program (UCGP)
at the National High Magnetic Field Laboratory (NHMFL)
which is supported by USA National Science Foundation
(NSF) Cooperative Agreement No. DMR-1157490, the State
of Florida, and the U.S. Department of Energy. Special thanks
5743
DOI: 10.1021/acs.chemmater.5b02290
Chem. Mater. 2015, 27, 5736−5744
Article
Chemistry of Materials
(21) Badway, F.; Pereira, N.; Cosandey, F.; Amatucci, G. CarbonMetal Fluoride Nanocomposites Structure and Electrochemistry of
FeF3:C. J. Electrochem. Soc. 2003, 150, A1209−A1218.
(22) Yamakawa, N.; Jiang, M.; Grey, C. P. Investigation of the
Conversion Reaction Mechanisms for Binary Copper(II) Compounds
by Solid-State NMR Spectroscopy and X-ray Diffraction. Chem. Mater.
2009, 21, 3162−3176.
(23) Yamakawa, N.; Jiang, M.; Key, B.; Grey, C. P. Identifying the
Local Structures Formed during Lithiation of the Conversion Material,
Iron Fluoride, in a Li Ion Battery: A Solid-State NMR, X-ray
Diffraction, and Pair Distribution Function Analysis Study. J. Am.
Chem. Soc. 2009, 131, 10525−10536.
(24) Li, T.; Chen, Z. X.; Cao, Y. L.; Ai, X. P.; Yang, H. X. Transitionmetal chlorides as conversion cathode materials for Li-ion batteries.
Electrochim. Acta 2012, 68, 202−205.
(25) Liu, J.-l.; Cui, W.-j.; Wang, C.-x.; Xia, Y.-y. Electrochemical
reaction of lithium with CoCl2 in nonaqueous electrolyte. Electrochem.
Commun. 2011, 13, 269−271.
(26) Amatucci, G. G.; Pereira, N. Fluoride based electrode materials
for advanced energy storage devices. J. Fluorine Chem. 2007, 128, 243−
262.
(27) Zhang, W.; Ma, L.; Yue, H.; Yang, Y. Synthesis and
characterization of in situ Fe2O3-coated FeF3 cathode materials for
rechargeable lithium batteries. J. Mater. Chem. 2012, 22, 24769−
24775.
(28) Zheng, S.-F.; Hu, J.-S.; Zhong, L.-S.; Song, W.-G.; Wan, L.-J.;
Guo, Y.-G. Introducing dual functional CNT networks into CuO
nanomicrospheres toward superior electrode materials for lithium-ion
batteries. Chem. Mater. 2008, 20, 3617−3622.
(29) Dey, A. N.; Needham; Holmes, R. W.; Mass, D. Metal
phosphate and metal arsenate organic electrolyte cells. U.S. Patent
3736184 A, 1973.
(30) Caillaud, T.; Lecerf, A. Positive active material for an electric cell
with a non-aqueous electrolyte. U.S. Patent 4260668 A, 1981.
(31) Grey, C. P.; Dupré, N. NMR studies of cathode materials for
lithium-ion rechargeable batteries. Chem. Rev. 2004, 104, 4493−4512.
(32) Blanc, F. d. r.; Leskes, M.; Grey, C. P. In Situ Solid-State NMR
Spectroscopy of Electrochemical Cells: Batteries, Supercapacitors, and
Fuel Cells. Acc. Chem. Res. 2013, 46, 1952−1963.
(33) Hu, Y.-Y.; Liu, Z.; Nam, K.-W.; Borkiewicz, O. J.; Cheng, J.;
Hua, X.; Dunstan, M. T.; Yu, X.; Wiaderek, K. M.; Du, L.-S. Origin of
additional capacities in metal oxide lithium-ion battery electrodes. Nat.
Mater. 2013, 12, 1130−1136.
(34) Zhang, W.; Duchesne, P. N.; Gong, Z.-L.; Wu, S.-Q.; Ma, L.;
Jiang, Z.; Zhang, S.; Zhang, P.; Mi, J.-X.; Yang, Y. In Situ
Electrochemical XAFS Studies on an Iron Fluoride High-Capacity
Cathode Material for Rechargeable Lithium Batteries. J. Phys. Chem. C
2013, 117, 11498−11505.
(35) Song, M.-K.; Park, S.; Alamgir, F. M.; Cho, J.; Liu, M.
Nanostructured electrodes for lithium-ion and lithium-air batteries: the
latest developments, challenges, and perspectives. Mater. Sci. Eng., R
2011, 72, 203−252.
(36) Ressler, T. WinXAS: a program for X-ray absorption
spectroscopy data analysis under MS-Windows. J. Synchrotron Radiat.
1998, 5, 118−122.
(37) Whitehead, A. H.; Schreiber, M. Current Collectors for Positive
Electrodes of Lithium-Based Batteries. J. Electrochem. Soc. 2005, 152,
A2105−A2113.
(38) Hua, X.; Robert, R.; Du, L.-S.; Wiaderek, K. M.; Leskes, M.;
Chapman, K. W.; Chupas, P. J.; Grey, C. P. Comprehensive Study of
the CuF2 Conversion Reaction Mechanism in a Lithium Ion Battery. J.
Phys. Chem. C 2014, 118, 15169−15184.
(39) Kang, W.; Shen, Q. The shape-controlled synthesis and novel
lithium storage mechanism of as-prepared CuC2O4·xH2O nanostructures. J. Power Sources 2013, 238, 203−209.
(40) Snyder, K.; Raguz, B.; Hoffbauer, W.; Glaum, R.; Ehrenberg, H.;
Herklotz, M. Lithium Copper(I) Orthophosphates Li3‑xCuxPO4:
Synthesis, Crystal Structures, and Electrochemical Properties. Z.
Anorg. Allg. Chem. 2014, 640, 944−951.
(41) Upreti, S.; Chernova, N. A.; Xiao, J.; Miller, J. K.; Yakubovich,
O. V.; Cabana, J.; Grey, C. P.; Chevrier, V. L.; Ceder, G.; Musfeldt, J.
L.; Whittingham, M. S. Crystal Structure, Physical Properties, and
Electrochemistry of Copper Substituted LiFePO4 Single Crystals.
Chem. Mater. 2012, 24, 166−173.
(42) Amine, K. Active material for lithium batteries. U.S. Patent
6319632 B1, Nov. 20, 2001, 1999.
(43) Lamberti, C.; Bordiga, S.; Bonino, F.; Prestipino, C.; Berlier, G.;
Capello, L.; D’Acapito, F.; i Xamena, F. L.; Zecchina, A. Determination
of the oxidation and coordination state of copper on different Cubased catalysts by XANES spectroscopy in situ or in operando
conditions. Phys. Chem. Chem. Phys. 2003, 5, 4502−4509.
(44) Cui, Y.-H.; Xue, M.-Z.; Zhou, Y.-N.; Peng, S.-M.; Wang, X.-L.;
Fu, Z.-W. The investigation on electrochemical reaction mechanism of
CuF2 thin film with lithium. Electrochim. Acta 2011, 56, 2328−2335.
(45) Yu, H. C.; Ling, C.; Bhattacharya, J.; Thomas, J. C.; Thornton,
K.; Van der Ven, A. Designing the next generation high capacity
battery electrodes. Energy Environ. Sci. 2014, 7, 1760−1768.
(46) Gao, X. P.; Bao, J. L.; Pan, G. L.; Zhu, H. Y.; Huang, P. X.; Wu,
F.; Song, D. Y. Preparation and Electrochemical Performance of
Polycrystalline and Single Crystalline CuO Nanorods as Anode
Materials for Li Ion Battery. J. Phys. Chem. B 2004, 108, 5547−5551.
5744
DOI: 10.1021/acs.chemmater.5b02290
Chem. Mater. 2015, 27, 5736−5744