volume 45 number 39 21 october 2016 pages 15263–15686

Dalton Transactions

An international journal of inorganic chemistrywww.rsc.org/dalton

ISSN 1477-9226

PAPER Lu-Lu Zhang, Xue-Lin Yang et al. Investigation of Co-incorporated pristine and Fe-doped Li

3 V

2 (PO

4 ) 3

cathode materials for lithium-ion batteries

Volume 45 Number 39 21 October 2016 Pages 15263–15686

DaltonTransactions

PAPER

Cite this: Dalton Trans., 2016, 45,15317

Received 23rd May 2016,Accepted 31st July 2016

DOI: 10.1039/c6dt02058e

www.rsc.org/dalton

Investigation of Co-incorporated pristine andFe-doped Li3V2(PO4)3 cathode materials forlithium-ion batteries

Hua-Bin Sun,a Lu-Lu Zhang,*a Xue-Lin Yang,*a Gan Liangb and Zhen Lia

Monoclinic Li3V2(PO4)3/C (LVP/C) and Li3V1.95Fe0.05(PO4)3/C (LVFP/C) composites were successfully

modified by cobalt incorporation. The effects of cobalt incorporation on the structure, morphology and

electrochemical performance of the LVP/C and LVFP/C composites were systematically investigated. The

results show that most Co exists in the form of CoO and forms a hybrid layer with the carbon coating on

the surface of the LVP and LVFP particles; moreover, a small part of Co enters into the LVP or LVFP lattices

due to atomic diffusion. Compared with LVP/C and LVFP/C, Co-incorporated samples exhibit better

electrochemical performance. In particular, under the common effect of doping and a hybrid layer

(carbon and metal oxides) coating, the LVFP/C-Co electrode displays a prominent initial capacity of

124.7 mA h g−1 and a very low capacity fading of ∼0.04% per cycle even after 500 cycles at 20 C. This

novel co-modification method with cation doping and a hybrid layer (carbon and metal oxide) coating is

a highly effective way to improve the electrochemical performance and has great potential to be easily

used to modify other cathode materials with poor electrical conductivity.

Introduction

As a cathode material, monoclinic Li3V2(PO4)3 (LVP) exhibitshigh energy and power density, thermal stability, high safety,low cost and a large theoretical capacity of 197 mA h g−1 whencharged up to 4.8 V vs. Li/Li+.1–3 Consequently, it has beenproposed as one of the most promising cathode materials forcommercialization. However, due to the electronically insulat-ing phosphate groups isolating the valence electrons of tran-sition metals within the lattices, a low electrical conductivity(2.4 × 10−7 S cm−1)4,5 and poor ionic conductivity (10−13–10−8

cm2 s−1)6,7 result. This drawback also hampers LVP’s practicalapplication. Moreover, when charged above 4.6 V, LVP haspoor cycling performance due to vanadium dissolution in theelectrolyte and decomposition of the electrolyte. Many efforts,such as coating (i.e. carbon and oxide)8–15 and doping,16–32

have been made to overcome these problems. Carbon coatingis the simplest and most effective method to improve the elec-trical conductivity. However, excessive carbon incorporationcan reduce the tap density of LVP. Hence, metal oxide coating

is also gradually being applied to modify LVP.10–14 On the onehand, a metal oxide coating can depress vanadium dissolutionin the electrolyte;11,12,14 while on the other hand, to a certainextent, the metal oxide coating can also improve the electricalconductivity and reduce the charge-transfer resistance.10,12,14

In particular, a carbon and oxide co-coating has been provedto be an effective method to remarkably enhance the electro-chemical performance of LVP,10,12–14 because the hybrid layercoating can effectively improve the electrical conductivity ofLVP and provide more effective protection for the activematerials against direct contact with the electrolyte. However,it is still difficult to improve the intrinsic electronic conduc-tivity of LVP only by applying a carbon and metal oxidecoating, thus the capacity performance is not competitive.Besides this method, doping is another effective method toimprove intrinsic electrical conductivity due to the decrease inthe band gap of LVP and the formation of some lattice defectsin LVP.18–20,22,28 Here, transition metal ions are commonlyused as doping elements, such as Fe, Cr, Y, Bi, Nb, Mn, Ni, Zr,Ti and Co.16–20,22,23,25–32 However, if LVP is modified only bycoating or doping, it is not easy to simultaneously realize bothan increase in capacity performance and an improvement incycle performance when charged above 4.6 V. Therefore,doping combined with carbon coating is widely used tomodify LVP and to a certain extent can also help obtain anobviously improved electrochemical performance.17–20 But in

aCollege of Materials and Chemical Engineering, Hubei Provincial Collaborative

Innovation Center for New Energy Microgrid, China Three Gorges University,

8 Daxue Road, Yichang, Hubei 443002, China. E-mail: [email protected],

[email protected] of Physics, Sam Houston State University, Huntsville, Texas 77341, USA

This journal is © The Royal Society of Chemistry 2016 Dalton Trans., 2016, 45, 15317–15325 | 15317

Publ

ishe

d on

01

Aug

ust 2

016.

Dow

nloa

ded

on 0

8/10

/201

6 03

:39:

01.

View Article OnlineView Journal | View Issue

www.rsc.org/dalton

http://crossmark.crossref.org/dialog/?doi=10.1039/c6dt02058e&domain=pdf&date_stamp=2016-09-28

http://dx.doi.org/10.1039/C6DT02058E

http://pubs.rsc.org/en/journals/journal/DT

http://pubs.rsc.org/en/journals/journal/DT?issueid=DT045039

reality, the cycle performance is still unsatisfactory.17,29 Forinstance, both Kuang31 and Nathiya32 reported Co as a dopingelement and carbon as a coating material for modifying LVP,whereby the capacity was greatly increased but the capacity reten-tion was not desirable. In fact, there are also some reports aboutCo as a coating element to reduce the capacity fading ofo-LiMnO2 at 55 °C cycling caused by Mn3+ dissolution.33,34

However, there are no reports of Co as a coating element for LVP.In this work, we intended to obtain a novel LVP composite

co-modified with doping and a hybrid layer coating (carbonand metal oxide). Here, we selected Fe as the doping elementand Co as the coating element together with carbon. For thisco-modified LVP composite (LVFP/C-Co), doping leads to animproved intrinsic electronic conductivity, while the hybridlayer coating can give more effective protection for the activematerials against direct contact with the electrolyte; moreover,the metal oxide in the hybrid coating layer can provide somedefects in the lattice, which is conducive not only to increasingthe electrical conductivity but also to Li+ absorption anddiffusion. Therefore, under the common effects of doping anda hybrid layer coating, LVFP/C-Co presents outstanding high-rate capacity performance and excellent cycle stability, i.e.LVFP/C-Co presented an excellent high-rate performance of∼80% capacity retention and a capacity fading of ∼0.04% percycle after 500 cycles at 20 C.

ExperimentSample synthesis

The Fe-doped Li3V2(PO4)3/C composites were prepared by asolid-state method as follows: first, stoichiometric lithiumcarbonate (Li2CO3), ammonium metavanadate (NH4VO3),ferric oxide (Fe2O3) and ammonium dihydrogen phosphate(NH4H2PO4) were mixed and ball-milled for 10 h in ethanol.Here, the molar ratio of Li, V, Fe and P was 3.06 : 2 − x : x : 3(x = 0 and 0.05). After drying in an oven at 50 °C overnight, themixture was pre-calcined at 350 °C for 6 h at a heating rate of3 °C min−1 in N2 atmosphere and then cooled down to roomtemperature. Second, 15 wt% glucose, which acts as not onlythe carbon source but also a reductive agent, was added intothe resulting precursor and further ball-milled for 6 h. Third,the dried powder was sintered at 700 °C for 10 h in N2 atmo-sphere to finally achieve the single carbon-coated Li3V2(PO4)3and Li3V1.95Fe0.05(PO4)3 composites (denoted as LVP/C andLVFP/C, respectively).

To obtain Co-incorporated LVP/C and LVFP/C composites,the as-prepared LVP/C and LVFP/C powders were first dis-persed in (CH3COO)2Co·4H2O ethanol solution (3 wt% of LVP/C and LVFP/C powders, respectively) by an ultrasonic methodfor 2 h. Then, this above mixture was stirred by a magneticforce stirrer until dry. Finally, the precursor was calcined at600 °C for 5 h in N2 atmosphere, and the Co-incorporated LVP/C and LVFP/C composites were obtained, respectively. The twocorresponding Co-incorporated samples were labelled as LVP/C-Co and LVFP/C-Co, respectively.

Sample analysis

The phase and crystalline structure of the four samples (LVP/C, LVP/C-Co, LVFP/C and LVFP/C-Co) were studied by powderX-ray diffraction (XRD, Rigaku RINT-2000) with Cu-Kα radi-ation. All the diffraction patterns were obtained in a 2θ rangeof 10–70° at a slow step of 0.4° min−1. The contents of Fe andCo in the final products were tested by atomic absorption spectro-scopy (AAS, AA240FS, Agilent), and the carbon content wasmeasured by an IR carbon/sulfur system equipped with a highfrequency induction combustion furnace (HW2000B, China).Electrical conductivity was measured with a standard four-probe method by a RTS resistivity measurement system (RTS-8,China) on disc-shaped pellets with diameters of 8 mm andthicknesses of about 1.0 mm. Both a field emission scanningelectron microscope (FESEM, JSM-7500F, JEOL) and a high-resolution transmission electron microscope (HRTEM,JEM-2100, JEOL) were used to observe the morphology of thesamples. X-ray photoelectron spectroscopy (XPS, PHIQuantera, U-P) combined with Ar-ion sputtering was used toanalyze the oxidation state and distribution of the keyelements in the samples.

Electrochemical measurements

The CR2025-type coin cells were assembled to study theelectrochemical properties of the as-prepared materials (i.e.LVP/C, LVP/C-Co, LVFP/C and LVFP/C-Co). The active material,acetylene black conductor and PVDF binder were mixed in aweight ratio of 75 : 15 : 10 and then dissolved in N-methylpyrrolidinone solvent to obtain a slurry. Subsequently, the mixedslurry was cast on Al foil to form a uniform film. After dryingat ∼ 60 °C, the film was cut into a disc of 14 mm diameter andthen pressed at a pressure of about 6 MPa. Finally, the discwas dried at 120 °C for 8 h in vacuum and then transferredinto an argon-filled glove box (Super 1220/750, Mikrouna) toassemble the CR2025 coin cells. Here, Celgard 2400 wasused as a separator, lithium foil as the counter and refer-ence electrode, and 1.1 M LiPF6/(EC + PC + EMC + DEC)(15 : 20 : 25 : 40 wt%) as the electrolyte. Constant currentcharge/discharge measurements at various rates were per-formed on a battery measurement system (LAND CT2001A,China). Cyclic voltammetry (CV) and electrochemical impe-dance spectroscopy (EIS) measurements were performed on anelectrochemical working station (CHI614C, China).

Results and discussion

Fig. 1a shows the XRD patterns of the LVP/C, LVP/C-Co, LVFP/C and LVFP/C-Co composites. All the diffraction peaks areindexed to a monoclinic Li3V2(PO4)3 phase with the spacegroup P21/n (JCPDS, no. 72-7074) and agree well with the pre-vious reports,35,36 indicating that Fe and Co incorporationdoes not change the monoclinic structure of Li3V2(PO4)3.Furthermore, Rietveld refinements were also carried out on allthe four samples by using the Maud program software.37–41

The refinement results are shown in Fig. 1(b–e), Tables 1 and 2.

Paper Dalton Transactions

15318 | Dalton Trans., 2016, 45, 15317–15325 This journal is © The Royal Society of Chemistry 2016

Publ

ishe

d on

01

Aug

ust 2

016.

Dow

nloa

ded

on 0

8/10

/201

6 03

:39:

01.

View Article Online


Because the reliability factor of s is less than 2, and Rw is lessthan or about 15%, the Rietveld refinement results are con-sidered reliable. As seen in Table 1, compared with pristineLVP/C, both LVP/C-Co and LVFP/C show a slightly decreasedunit cell volume due to the slightly smaller ionic radius of Fe2+

(0.076 nm) and Co2+ (0.074 nm) than V3+ (0.078 nm).16,39,42 Incontrast, LVFP/C-Co has the least cell volume because of thesynergistic effect of Fe2+ and Co2+. To some extent it was thusdemonstrated that both Fe and Co had been doped into theLVP lattice. To support this point, we investigated the distri-bution of key elements in LVP/C-Co and LVFP/C-Co using XPSmeasurements assisted by Ar-ion sputtering, and describe thisin the next part. The Fe content in the LVFP/C and LVFP/C-Cosamples was 0.66 wt% and 0.82 wt%, respectively, while the Co

content in LVP/C-Co and LVFP/C-Co was 0.75 wt% and 0.67 wt%,respectively. We noted that no diffraction peaks for crystal-line carbon could be observed in all the XRD patterns, whichindicated that the residual carbon was amorphous or/and thecarbon content was too low to be detected. The results of theIR carbon/sulfur system further showed that the carboncontent in LVP/C, LVP/C-Co, LVFP/C and LVFP/C-Co sampleswas 3.44 wt%, 3.16 wt%, 2.74 wt% and 2.71 wt%, respectively(Table 3). The significantly reduced carbon content in Fe-doped samples (LVFP/C and LVFP/C-Co) could be attributed tothe carbon consumption for the reduction of Fe3+ in Fe2O3.Table 3 also shows the electrical conductivities of LVP/C, LVP/C-Co, LVFP/C and LVFP/C-Co, which were 0.79 × 10−3, 1.27 ×10−3, 1.32 × 10−3 and 1.48 × 10−3 S cm−1, respectively.Obviously, the carbon content in LVFP/C is lower than that inLVP/C, but the electrical conductivity of LVFP/C is still higherthan that of LVP/C. Moreover, the Co-incorporated samples(i.e. LVP/C-Co and LVFP/C-Co) had lower carbon content thanthe samples without Co-incorporation (i.e. LVP/C and LVFP/C),but the former still exhibited higher electrical conductivitythan the latter. The results demonstrate that both Fe-dopingand Co-incorporation can improve the electrical conductivity.

Fig. 2(a–d) show the SEM images of LVP/C, LVP/C-Co,LVFP/C and LVFP/C-Co samples. It can be observed that all thesamples show irregular shapes, together with some agglo-meration and a wide size distribution ranging from tens ofnanometres to a few microns, indicating that Fe and Co in-corporation have no obvious influence on the morphology.The EDX elemental mappings of LVFP/C-Co (Fig. 2e) were usedto further investigate the distribution of each element (V, Fe,P, C and Co). The bright dot in the mappings infers a highcontent of the corresponding element. Apparently, the V, Fe, P, Cand Co elements were uniformly distributed in the LVFP/C-Coparticles.

XPS analysis was used to investigate the valence states anddistribution of the key elements (V, Fe and Co) in the LVP/C,LVP/C-Co and LVFP/C-Co samples (Fig. 3). The binding energyof all samples were calibrated by carbon (C 1s = 284.5 eV). Dueto the chemical reduction of Ar-ion sputtering,43 all the XPSpeaks in the interior were shifted slightly towards low bindingenergy compared to those on the surface. It could be clearlyseen from the high-resolution XPS spectra of LVP/C-Co andLVFP/C-Co (Fig. 3b2, b3, c2 and c3) that the V 2p3/2 peaks onthe surface are weaker but the Co 2p3/2 peaks on the surfaceare stronger, which confirms that most Co exists on thesurface of LVP or of the LVFP particles, while the weaker Co2p3/2 peaks in the interior also reveal that a small part of Coenters into the LVP or LVFP lattices. The part of Co in the lat-tices can be attributed to the atomic diffusion at high sinteringtemperature. In addition, as shown in Fig. 3c4, the Fe 2p3/2peaks appear not only in the interior but also on the surface ofthe sample, indicative of Fe-doping and FeO-coating.16 Thispart of FeO on the surface can also be attributed to atomicdiffusion at high sintering temperature. To further analyze thevalence of V, Fe and Co, the asymmetrical XPS peaks weredecomposed by the curve-fitting approach. As shown in

Fig. 1 (a) XRD patterns and (b–e) Rietveld refinement results for LVP/C,LVP/C-Co, LVFP/C and LVFP/C-Co, respectively.

Table 1 Lattice parameters of LVP/C, LVP/C-Co, LVFP/C and LVFP/C-Co

Sample a (Å) b (Å) c (Å) V (Å3)

Reliabilityfactor

s Rw

LVP/C 8.6016 8.6038 12.0286 890.19 1.55 14.95LVP/C-Co 8.5998 8.5995 12.0317 889.79 1.46 9.91LVFP/C 8.5944 8.5958 12.0241 888.29 1.66 15.94LVFP/C-Co 8.5912 8.5832 12.0150 885.99 1.78 16.94

Dalton Transactions Paper


Publ

ishe

d on

01

Aug

ust 2

016.

Dow

nloa

ded

on 0

8/10

/201

6 03

:39:

01.

View Article Online


Fig. 4a1, b1 and c1, the V 2p3/2 main peak position for LVFP/C-Co is very close to that for LVP/C-Co and LVP/C, which indi-cates that Co-incorporation does not change the valence ofV3+.10,12,16,19 The Fe 2p3/2 main peak for LVFP/C-Co located at

∼710.7 eV verified that the valence of Fe for LVFP/C-Co is+2.16,19 Thus, combined with the results from Fig. 3, we canreasonably suggest that the part of Fe on the surface exists inthe form of FeO. Furthermore, the Co 2p3/2 asymmetricalsignal at ∼781.0 eV and the satellite peak at ∼786.3 eV (shownin Fig. 4b2 and c2) confirm the valence of Co in LVP/C-Co andLVFP/C-Co is +2, corresponding to Co2+ in CoO,44,45 whichmeans that the part of Co on the surface exists in the form ofCoO and composes a hybrid coating layer with carbon. It isworth mentioning that for the LVFP/C-Co sample, the twokinds of metal oxides (FeO and CoO) in the hybrid coatinglayer may give more effective protection for active materials

Table 2 Atomic fractional coordinates of LVP/C, LVP/C-Co, LVFP/C and LVFP/C-Co

Atom

LVP/C LVP/C-Co LVFP/C LVFP/C-Co

x/Å y/Å z/Å x/Å y/Å z/Å x/Å y/Å z/Å x/Å y/Å z/Å

V1(Fe) 0.2534 0.4643 0.1119 0.2559 0.4549 0.1108 0.2570 0.4534 0.1097 0.2450 0.4575 0.1133V2(Fe) 0.7446 0.4667 0.3898 0.7496 0.4790 0.3905 0.7466 0.4739 0.3895 0.7512 0.4798 0.3901P1 0.1249 0.1063 0.1444 0.1064 0.1026 0.1482 0.1180 0.1096 0.1405 0.1067 0.1115 0.1389P2 0.6020 0.1121 0.3457 0.5976 0.1254 0.3525 0.5848 0.1101 0.3478 0.5954 0.1322 0.3443P3 0.0271 0.2276 0.5079 0.0377 0.2339 0.4939 0.0376 0.2328 0.4935 0.0191 0.2410 0.4964O1 0.9300 0.1371 0.1662 0.9139 0.1463 0.1426 0.9061 0.1527 0.1555 0.9238 0.1344 0.1488O2 0.1678 0.9804 0.2278 0.1851 0.9861 0.2347 0.1609 0.9951 0.2350 0.1792 0.9841 0.2410O3 0.1624 0.0418 0.0373 0.1665 0.0778 0.0416 0.1715 0.0576 0.0352 0.2251 0.0465 −0.0149O4 0.1803 0.2818 0.1846 0.1631 0.2995 0.1950 0.1650 0.2993 0.1939 0.1339 0.2604 0.1944O5 0.4128 0.0985 0.3711 0.4140 0.0891 0.3542 0.4164 0.1102 0.3529 0.4097 0.0839 0.3527O6 0.6915 −0.0071 0.2426 0.6858 −0.0052 0.2812 0.6787 −0.0064 0.2806 0.6665 0.0022 0.2812O7 0.6306 0.0767 0.4761 0.6480 0.0686 0.4807 0.6285 0.0682 0.4894 0.6496 0.0938 0.4831O8 0.6356 0.2913 0.3151 0.6590 0.2956 0.3092 0.6649 0.2892 0.3206 0.6798 0.3149 0.3199O9 0.8677 0.1194 0.5834 0.9404 0.1387 0.5756 0.9336 0.1403 0.5703 0.9531 0.1064 0.5813O10 0.9445 0.3498 0.4130 0.9453 0.3764 0.3891 0.9141 0.3650 0.3991 0.9411 0.3165 0.3863O11 0.1248 0.1955 0.4344 0.1471 0.1711 0.4420 0.1429 0.1755 0.4294 0.1671 0.1868 0.4336O12 0.1600 0.3686 0.5855 0.1595 0.3649 0.3649 0.1351 0.3669 0.5855 0.1023 0.3414 0.5688Li1 0.2107 0.7601 0.1984 0.1494 0.8227 0.8227 0.2388 0.4919 0.1294 0.1823 0.8287 0.1792Li2 0.9695 0.2670 0.2421 0.9283 0.2785 0.2785 0.9756 0.2727 0.2334 0.8968 0.2989 0.1847Li3 0.6616 0.3783 0.1561 0.5576 0.5975 0.5975 0.5911 0.4131 0.1874 0.5032 0.4388 0.2051

Table 3 Electrical conductivity and carbon content of all the samples

Sample Electrical conductivity/S cm−1 Carbon content/%

LVP/C 0.79 × 10−3 3.44LVP/C-Co 1.27 × 10−3 3.16LVFP/C 1.32 × 10−3 2.74LVFP/C-Co 1.48 × 10−3 2.71

Fig. 2 SEM images of (a) LVP/C, (b) LVFP/C, (c) LVP/C-Co, (d) LVFP/C-Co, and (e) EDX mapping of LVFP/C-Co powders.



Publ

ishe

d on

01

Aug

ust 2

016.

Dow

nloa

ded

on 0

8/10

/201

6 03

:39:

01.

View Article Online


against direct contact with the electrolyte,11,12 which thenleads to better cycling stability. Moreover, FeO and CoO cangenerate some lattice defects during the sintering process,which is beneficial to Li+ absorption and diffusion in thehybrid coating layer. As a result, LVFP/C-Co is expected toexhibit better electrochemical performance.

Fig. 5 shows the TEM and HRTEM images of the LVFP/Cand LVFP/C-Co samples. It can be clearly seen that both theLVFP/C and LVFP/C-Co particles show some agglomeration. Asseen in Fig. 5b, the LVFP/C particles are covered by an amor-phous hybrid layer (C and FeO) with a thickness of ∼7 nm,and the lattice fringes exhibit interplanar distances of

0.236 nm and 0.407 nm, which correspond to the (214) and(021) planes of LVP, respectively. The LVFP/C-Co particles arealso coated with an amorphous hybrid layer (carbon, CoO andFeO), but the corresponding thickness (∼3 nm) is obviouslythinner due to the lower carbon content in the LVFP/C-Cosample. In Fig. 5d, the marked lattice fringes with a d-spacingof 0.306 nm correspond to the (220) plane of LVP. The selectedarea electron diffraction (SAED) and the fast Fourier transform(FFT) patterns in the inset in Fig. 5b and d indicate the highlycrystalline character of both samples.

In order to evaluate the electrochemical performance, allthe samples were constant current charged and discharged in

Fig. 3 XPS spectra of (a) LVP, (b) LVP/C-Co and (c) LVFP/C-Co.

Fig. 4 XPS fitting patterns of V 2p3/2, Co 2p3/2 and Fe 2p3/2: (a) LVP/C, (b) LVP/C-Co and (c) LVFP/C-Co.



Publ

ishe

d on

01

Aug

ust 2

016.

Dow

nloa

ded

on 0

8/10

/201

6 03

:39:

01.

View Article Online


the potential range of 3.0–4.8 V. Fig. 6a and c show the initialcharge–discharge curves of LVP/C, LVP/C-Co, LVFP/C andLVFP/C-Co electrodes at 1 C and 5 C (1 C = 197 mA g−1). In the

charge curves, four typical plateaus around 3.62, 3.70, 4.10 and4.59 V can be observed, corresponding to the four step-extrac-tion of all three Li+ ions in the phase transition processes(i.e. Li3V2(PO4)3 → Li2.5V2(PO4)3 → Li2V2(PO4)3 → LiV2(PO4)3 →V2(PO4)3).

10,46 In the discharge curves, there is an S-shapedcurve followed by two plateaus (about 3.65 and 3.57 V). TheS-shaped curve corresponds to a solid solution process fromV2(PO4)3 to Li2V2(PO4)3.

10 The two discharge plateaus arerelated to the insertion of the first Li+ in two steps (Li2V2(PO4)3→ Li2.5V2(PO4)3 → Li3V2(PO4)3). At a higher current density of5 C, the charge/discharge plateaus are becoming increasinglyblurred due to the growing electrode polarization, which alsoindicates that the electrochemical process is controlled by thediffusion of Li+ ions. As shown in Fig. 6 and Table 4, at 1 Cand 5 C, LVFP/C-Co can deliver an initial capacity as high as170.2 and 148.1 mA h g−1, respectively. Moreover, the corres-ponding discharge capacity still remains 138.3 and 127.1 mA h g−1

even after 100 cycles, which means that the capacity reten-tion of LVFP/C-Co is ∼81% (1 C) and ∼86% (5 C). However,LVFP/C only exhibits an initial capacity of 163.9 and 143.3mA h g−1 with a capacity retention of ∼75% and ∼84% at 1 C and5 C, respectively. Meanwhile, LVP/C-Co also shows a higherdischarge capacity and better cycling stability than LVP/C.Fig. 6e displays the long-life cycling performance of LVFP/C-Coat a high rate of 20 C. The initial specific discharge capacity is124.7 mA h g−1. After 500 cycles, LVFP/C-Co still owns a dis-charge capacity of 99.5 mA h g−1 with a capacity retention of∼80% and a very low capacity fading of ∼0.04% per cycle.Obviously, compared with single carbon coating, the hybridlayer coating further improves the electrochemical propertiesof LVP, which is attributed to the following factors: first, Feand/or Co entering into the lattice can effectively improve theintrinsic electronic conductivity of LVP, resulting in a highercapacity; second, the metal oxide (FeO and CoO) can improve heelectron and Li-ion transport, also leading to a higher capacity;third, the hybrid layer (C + FeO, C + CoO, and C + CoO + FeO)coating can prevent the active materials from dissolution in theelectrolyte better to stabilize the LVP structural, thus improvingthe cycle performance.11,12 Therefore, this novel co-modificationmethod with a hybrid layer (carbon and metal oxide) coatingand cation doping could be expected to be able to modify othercathode materials with poor electrical conductivity.

To further study the effect of co-modification on the struc-ture stability of Li3V2(PO4)3 electrodes, XRD patterns of all thefour electrodes were checked after 100 cycles at 5 C, as shown

Fig. 5 TEM and HRTEM images for (a, b) LVFP/C and (c, d) LVFP/C-Cosamples.

Fig. 6 The capacity and cycling performance for the as-preparedsamples: (a, c) the initial charge/discharge profiles, (b, d) the cycle per-formance profiles at 1 C and 5 C and (e) the discharge capacity of LVFP/C-Co at a high rate of 20 C for 500 cycles.

Table 4 Discharge capacity and capacity retention ratio of the samples

Sample

1 C 5 C

1st 100th R* (%) 1st 100th R* (%)


* Capacity retention ratio compared to the first cycle.



Publ

ishe

d on

01

Aug

ust 2

016.

Dow

nloa

ded

on 0

8/10

/201

6 03

:39:

01.

View Article Online


in Fig. 7. Compared with Fig. 1a, it was found that themain diffraction peaks of all the samples show no differencebefore and after 100 cycles, indicating that the monoclinicLi3V2(PO4)3 phase still remains stable even after 100 cyclesbetween 3.0–4.8 V.

CV measurements were conducted in the potential range of3.0–4.8 V at a slow scan rate of 0.05 mV s−1 in order to investi-gate the electrochemical behaviour of all the samples duringthe charge–discharge process. Here, we should point out thatwhen considering the electrolyte penetration into the elec-trode, the structural change and the solid electrolyte interface(SEI) film formation, we chose the second cycle for the ana-lysis.47,48 As shown in Fig. 8a and b, it is apparent that all theCV curves are very similar. There are four oxidation peaks andthree reduction peaks, corresponding to the multi-phase trans-formations in LixV2(PO4)3 (x = 3, 2.5, 2.0, 1.0).49,50 The reasonfor the difference between the oxidation and reduction peaksis attributed to the solid solution of two phases from V2(PO4)3

to Li2V2(PO4)3 during the Li-intercalation process.10 Moreover,as shown in Fig. 8a and b and Table 5, LVP/C-Co and LVFP/C-Co show stronger current peaks and smaller potential differ-ences between the anodic and cathodic peaks than LVP/C andLVFP/C, indicating that Co-incorporation can help obtainlower electrode polarization and better electrochemical reversi-bility of Li+ ion extraction/insertion. More interestingly, fromthe enlarged patterns shown on the right side of Fig. 8b, otheroxidation/reduction peaks around 3.4 V can be detected withboth the LVFP/C and LVFP/C-Co electrodes. The oxidation/reduction peaks around 3.4 V are characteristic of the electro-chemical reactions of Fe2+/Fe3+ redox couples in LiFePO4,which is consistent with the results of XPS.

EIS measurements of LVP/C, LVP/C-Co, LVFP/C and LVFP/C-Co composites were performed over a frequency range from0.01 Hz to 100 kHz, and the results are shown in Fig. 8c and d.All the EIS curves are composed of a depressed semicircle anda straight line, which correspond to the charge-transfer resist-ance (Rct) and Warburg impedance, respectively. After fittingall the EIS curves by an equivalent circuit composed of“R(C(RW))” using the ZSimpWin program,44,45 it could beclearly seen that both LVP/C-Co and LVFP/C-Co show a moregreatly decreased charge-transfer resistance (14.44 and 12.65Ω, respectively) than LVP/C (47.08 Ω) and LVFP/C (21.86 Ω)(Table 6). Furthermore, the lithium-ion diffusion coefficient

Fig. 7 XRD patterns of the LVP/C, LVP/C-Co, LVFP/C and LVFP/C-Coelectrodes after 100 cycles at 5 C.

Fig. 8 (a, b) CV profiles of the as-prepared samples, (c) EIS spectra, (d) the relationship between the Z’ and ω−1/2 in the low frequency region ofLVP/C, LVP/C-Co, LVFP/C and LVFP/C-Co electrodes and (e) schematic of the structure of the hybrid layer (C + CoO + FeO).

Table 5 Peak potentials and the difference between reduction and oxi-dation potentials for the as-prepared samples

Sample EA1 EA2 EC1 EC2 ΔEA1–C1 ΔEA2–C2




Publ

ishe

d on

01

Aug

ust 2

016.

Dow

nloa

ded

on 0

8/10

/201

6 03

:39:

01.

View Article Online


could be calculated from the straight line in low frequencyaccording to the following equation:51,52

DLiþ ¼ R 2T 2=2A 2n 4F 4C 2δ2;

where R is the gas constant, T is the absolute temperature, A isthe surface area of the cathode, n is the number of electronsper molecule during oxidation, F is the Faraday constant, C isthe concentration of lithium ions and δ is the Warburg coeffi-cient, which is relative with Z′:

Z′ ¼ Rc þ Rct þ δω�1=2;

where ω is the frequency at low frequency. To obtain theWarburg coefficient (δ), the Z′–ω−1/2 relation curves of all thesamples were obtained and the results are shown in Fig. 8d.The linear fitting results show that the lithium-ion diffusioncoefficients of LVP/C, LVP/C-Co, LVFP/C and LVFP/C-Co are0.17 × 10−11, 1.42 × 10−11, 1.07 × 10−11 and 10.5 × 10−11 cm2

s−1, respectively (Table 6). Clearly, the Co-incorporatedsamples had a higher Li+ diffusion coefficient, especially,LVFP/C-Co, which shows the highest Li+ diffusion coefficient,indicating the best electrochemical performance, which is con-sistent with the results of Fig. 6.

For a better understanding of the excellent electrochemicalperformance of LVFP/C-Co, the structural schematic represen-tation is shown in Fig. 8e. Obviously, the novel hybrid layer(C + CoO + FeO) contains two parts: one is amorphous carbon,which acts as the main body of the coating layer and con-structs a conductive network to improve electrical conductivity;the other is the CoO and FeO particles embedded in thecarbon layer, which can improve conductivity not only of theelectrons but also of the Li ions, resulting in the high capacity.Besides, this hybrid layer (C + CoO + FeO) coating can alsobetter prevent V3+ from dissolution in the electrolyte, thusleading to better cycling performance. Besides, the parts ofFe and Co entering into the lattice improve the intrinsic elec-tronic conductivity of LVP, also bringing about the highcapacity. Therefore, under the common effect of doping and ahybrid layer coating, a high capacity and good cycle perform-ance of LVP are easily obtained.

Conclusions

In summary, both LVP and LVFP composites were modified byCo-incorporation by an ultrasonic-assisted solid-state method.The results of XPS and HRTEM measurements confirmed thattrace Fe exists in the form of FeO on the surface of LVP orLVFP due to atomic diffusion at high temperature. Both the

LVP/C-Co and LVFP/C-Co particles are wrapped with an amor-phous hybrid coating layer of carbon and a metal oxide (CoOand CoO + FeO, respectively). This hybrid layer can effectivelybetter prevent the active materials from dissolution in the elec-trolyte; moreover, FeO and CoO can generate some latticedefects to improve both the electrical conductivity and Li-iontransport. Besides, a small part of Co enters into the LVP orLVFP lattices due to atomic diffusion. The existence of Co andFe in the lattice can improve the intrinsic electronic conduc-tivity of LVP and LVFP. Therefore, under the common effect ofdoping and a hybrid layer coating, LVFP/C-Co exhibits the bestelectrochemical performance, delivers the highest initialcapacity of 170.2 mA h g−1 at 1 C with a capacity retentionratio of 81.3% after 100 cycles and still retains a desirablecapacity retention ratio of ∼80% after 500 cycles at 20 C.

Acknowledgements

This work was supported by the NSFC (no. 51302153,51572151 and 51272128), the Outstanding Youth Science andTechnology Innovation Team Project of Hubei EducationalCommittee (no. T201603), and the Opening Project of CAS KeyLaboratory of Materials for Energy Conversion (no. CKEM131404).

Notes and references

1 C. Wei, W. He, X. Zhang, F. Xu, Q. Liu, C. Sun and X. Song,RSC Adv., 2015, 5, 54225.

2 J. Kim, J. K. Yoo, Y. S. Jung and K. Kang, Adv. Energy Mater.,2013, 3, 1004.

3 C. Wei, W. He, X. Zhang, S. Liu, C. Jin, S. Liu andZ. Huang, RSC Adv., 2015, 5, 28662.

4 X. Zhang, N. Böckenfeld, F. Berkemeier and A. Balducci,ChemSusChem, 2014, 7, 1710.

5 Y. Li, D. Xie, Y. D. Zhang, D. Zhou, X. Q. Niu, Y. Y. Tong,D. H. Wang, X. L. Wang, C. D. Gu and J. P. Tu, J. Mater.Chem. A, 2015, 3, 14731.

6 S. C. Yin, P. S. Strobel, H. Grondey and L. F. Nazar, Chem.Mater., 2004, 16, 1456.

7 Y. Li, W. Q. Bai, D. H. Wang, X. Q. Niu, Y. D. Zhang,H. Tang, X. L. Wang, C. D. Gu and J. P. Tu, Electrochim.Acta, 2015, 161, 252.

8 J. Su, X. L. Wu, J. S. Lee, J. Kim and Y. G. Guo, J. Mater.Chem. A, 2013, 1, 2508.

9 Y. Luo, X. Xu, Y. Zhang, Y. Pi, Y. Zhao, X. Tian, Q. An,Q. Wei and L. Mai, Adv. Energy Mater., 2014, 4, 1.

10 R. Zhang, Y. Zhang, K. Zhu, F. Du, Q. Fu, X. Yang, Y. Wang,X. Bie, G. Chen and Y. Wei, ACS Appl. Mater. Interfaces,2014, 6, 12523.

11 J. Zhai, M. Zhao, D. Wang and Y. Qiao, J. Alloys Compd.,2010, 502, 401.

12 L. L. Zhang, G. Liang, G. Peng, F. Zou, Y. H. Huang,M. C. Croft and A. Ignatov, J. Phys. Chem. C, 2012, 116, 12401.

13 M. Choi, I. H. Yoon, I. H. Jo, H. S. Kim, B. S. Jin, Y. M. Leeand W. K. Choi, Mater. Lett., 2015, 160, 206.

Table 6 EIS parameters of the samples

Sample Rct (Ω) δ (Ω s−1/2) DLi (cm2 s−1)

LVP/C 47.08 94.73 0.17 × 10−11

LVP/C-Co 14.44 32.55 1.42 × 10−11

LVFP/C 21.86 37.47 1.07 × 10−11

LVFP/C-Co 12.65 11.95 10.5 × 10−11



Publ

ishe

d on

01

Aug

ust 2

016.

Dow

nloa

ded

on 0

8/10

/201

6 03

:39:

01.

View Article Online


14 C. Lai, J. Wei, Z. Wang, Q. Xu, Y. Lu and H. Li, Solid StateIonics, 2015, 272, 121.

15 L. Zhang, X. L. Wang, J. Y. Xiang, Y. Zhou, S. J. Shi andJ. P. Tu, J. Power Sources, 2010, 195, 5057.

16 L. L. Zhang, G. Liang, G. Peng, Y. H. Huang, L. Wang,L. Qie, M. C. Croft, A. Ignatov and J. B. Goodenough,J. Electrochem. Soc., 2012, 159, A1573.

17 Y. Chen, Y. Zhao, X. An, J. Liu, Y. Dong and L. Chen,Electrochim. Acta, 2009, 54, 5844.

18 S. Zhong, L. Liu, J. Jiang, Y. Li, J. Wang, J. Liu and Y. Li,J. Rare Earth., 2009, 27, 134.

19 M. Ren, Z. Zhou, Y. Li, X. P. Gao and J. Yan, J. PowerSources, 2006, 162, 1357.

20 J. S. Huang, L. Yang, K. Y. Liu and Y. F. Tang, J. PowerSources, 2010, 195, 5013.

21 J. Dang, F. Xiang, N. Gu, R. Zhang, R. Mukherjee, I. K. Oh,N. Koratkar and Z. Yang, J. Power Sources, 2013, 243, 33.

22 S. Zhong, B. Zhao, Y. Li, Y. Liu, J. Liu and F. Li, J. WuhanUniv. Technol., 2009, 24, 343.

23 Y. Xia, W. Zhang, H. Huang, Y. Gan, C. Li and X. Tao,Mater. Sci. Eng. B, 2011, 176, 633.

24 G. Bai, Y. Yang and H. Shao, J. Electroanal. Chem., 2013,688, 98.

25 Y. Cheng, K. Feng, W. Zhou, H. Zhang, X. Li and H. Zhang,Dalton Trans., 2015, 44, 17579.

26 C. Lai, Z. Chen, H. Zhou, Y. Lu and H. Li, J. Mater. Sci.,2015, 50, 3075.

27 Y. Zhang, Q. Y. Huo, Y. Lv, L. Z. Wang, A. Q. Zhang,Y. H. Song, G. Y. Li, H. L. Gao, T. C. Xia and H. C. Dong,J. Alloys Compd., 2012, 542, 187.

28 J. Xu, G. Chen, H. Zhang, W. Zheng and Y. Li, J. Appl.Electrochem., 2015, 45, 123.

29 C. Deng, S. Zhang, S. Y. Yang, Y. Gao, B. Wu, L. Ma,B. L. Fu, Q. Wu and F. L. Liu, J. Phys. Chem. C, 2011, 115,15048.

30 W. Wang, J. Zhang, Z. Jia, C. Dai, Y. Hu, J. Zhou andQ. Xiao, Phys. Chem. Chem. Phys., 2014, 16, 13858.

31 Q. Kuang, Y. Zhao, X. An, J. Liu, Y. Dong and L. Chen,Electrochim. Acta, 2010, 55, 1575.

32 K. Nathiya, D. Bhuvaneswari, D. Nirmala and N. Kalaiselvi,RSC Adv., 2012, 2, 6885.

33 J. Cho, Chem. Mater., 2001, 13, 4537.34 J. Cho, T. J. Kim and B. Park, J. Electrochem. Soc., 2002, 149,

A288.35 C. Wang, Z. Guo, W. Shen, A. Zhang, Q. Xu, H. Liu and

Y. Wang, J. Mater. Chem. A, 2015, 3, 6064.36 H. Huang, S. C. Yin, T. Kerr, N. Taylor and L. F. Nazar, Adv.

Mater., 2002, 14, 1525.37 J. F. Zhang, X. W. Wang, B. Zhang and H. Tong,

Electrochim. Acta, 2015, 180, 507.38 Z. Liu, Y. Fan, W. Peng, Z. Wang, H. Guo, X. Li and

J. Wang, Ceram. Int., 2015, 41, 4267.39 L. L. Zhang, H. B. Sun, X. L. Yang, Y. W. Wen, Y. H. Huang,

M. Li, G. Peng, H. C. Tao, S. B. Ni and G. Liang,Electrochim. Acta, 2015, 152, 496.

40 S. Zhong, L. Wu and J. Liu, Electrochim. Acta, 2015, 74, 8.41 L. Yang, Y. Xia, L. Qin, G. Yuan, B. Qiu, J. Shi and Z. Liu,

J. Power Sources, 2016, 304, 293.42 R. T. Shannon, Acta Crystallogr., 1976, 32, 751.43 T. Choudhury, S. O. Saied, J. L. Sullivan and A. M. Abbot,

J. Phys. D: Appl. Phys., 1989, 22, 11855.44 G. Fierro, M. L. Jacono, M. Inversi, R. Dragone and P. Porta,

Top. Catal., 2000, 10, 39.45 Y. Liu, H. Dai, J. Deng, L. Zhang, Z. Zhao, X. Li, Y. Wang,

S. Xie, H. Yang and G. Guo, Inorg. Chem., 2013, 52, 8665.46 W. Zhang, Q. Liu, L. Feng, S. Wang, Y. Yang and H. Wei,

New J. Chem., 2014, 38, 2265.47 L. L. Zhang, G. Liang, A. Ignatov, M. C. Croft, X. Q. Xiong,

I. M. Hung, Y. H. Huang, X. L. Hu, W. X. Zhang andY. L. Peng, J. Phys. Chem. C, 2011, 115, 13520.

48 M. Choi, K. Kang, H. S. Kim, Y. M. Lee and B. S. Jin, RSCAdv., 2015, 5, 4872.

49 X. J. Zhu, Y. X. Liu, L. M. Geng and L. B. Chen, J. PowerSources, 2008, 184, 578.

50 X. Zhang, R. S. Kühnel, H. Hu, D. Eder and A. Balducci,Nano Energy, 2015, 12, 207.

51 L. L. Zhang, S. Duan, X. L. Yang, G. Liang, Y. H. Huang,X. Z. Cao, J. Yang, M. Li, M. C. Croft and C. Lewis, J. PowerSources, 2015, 274, 194.

52 L. L. Zhang, S. Duan, X. L. Yang, G. Liang, Y. H. Huang,X. Z. Cao, J. Yang, S. B. Ni and M. Li, Sci. Rep., 2014, 4,5064.



Publ

ishe

d on

01

Aug

ust 2

016.

Dow

nloa

ded

on 0

8/10

/201

6 03

:39:

01.

View Article Online


volume 45 number 39 21 october 2016 pages 15263–15686

Documents