Enhancement of li–ion battery capacity using nickel doped lifepo4 as cathode material - La Thi Hang

Journal of Science and Technology 55 (1B) (2017) 276–283 ENHANCEMENT OF Li–ION BATTERY CAPACITY USING NICKEL DOPED LiFePO4 AS CATHODE MATERIAL La Thi Hang1, 2, 3, *, Le My Loan Phung4, Nguyen Thi My Anh5, Hoang Xuan Tung5, Doan Phuc Luan5, Nguyen Nhi Tru5 1Vinh Long University of Technology Education (VLUTE) 73 Nguyen Hue Street, Vinh Long City, Vinh Long Province, Vietnam 2 Institute of Applied Materials Science – VAST 1 Mac Dinh Chi Street, Ben Nghe Ward,, District 1, Ho Chi Minh City, Vietnam 3 Graduate University of Science & Technology – VAST 18 Hoang Quoc Viet Street, Nghia Do Ward, Cau Giay District, Hanoi, Vietnam 4Applied Physical Chemistry Lab, University of Science – VNU HCM 227 Nguyen Van Cu Street, Ward 4, District 5, Ho Chi Minh City, Vietnam 5 Faculty of Materials Technology, University of Technology – VNU HCM 268 Ly Thuong Kiet Street, Ward 14, District 10, Ho Chi Minh City, Vietnam *Email: hanglt@vlute.edu.vn Received: 30 December 2016; Accepted for Publication: ABSTRACT Selective supervalent cations (M = Ni, Mn, La etc.) doped LiFePO4 (LFP) is an effective route to enhance its electrical conductivity, thereby improving electrochemical performances of lithium–ion batteries. For this purpose, nickel doped LiFePO4 based cathode material was investigated at different substitution amounts (xNi = 0.05, 0.10). LiFe1–xNixPO4 (LFNP) was synthesized from Ni(NO3)2.6H2O, LiOH.H2O, FeSO4.7H2O, H3PO4, and ascorbic acid precursors via solvothermal technique, followed by calcination in nitrogen atmosphere at 550 °C for 5 h. The structure and morphology of synthesized materials were examined by X–ray diffraction, Scanning electronic microscopy and Raman vibrational micro–spectroscopy. The electrochemical performances of doped materials were studied in Swagelok–type cell using LiPF6/EC–DMC (1:1) as electrolyte. LiFe1–xNixPO4 was shown to exhibit homogenous particles size of 50 ÷ 150 nm. The doped materials were titrated to quantify iron and nickel contents in samples. As anticipated, the electrochemical performances of LiFe1–xNixPO4were significantly enhanced compared to those of undoped LiFePO4. Keywords: LiFe1–xNixPO4 (LFNP), solvothermal, lithium–ion, conductivity, capacity. 1. INTRODUCTION LiFePO4(LFP) is a promising cathode material for manufacturing lithium–ion batteries with prolonged life time, good cycle stability, environmental friendliness, and relatively low cost. La Thi Hang, et al. 277 However, due to its rigid orthorhombic olivine structure, the intrinsic electronic conductivity of LFP and Li+ diffusion rate is considerably low to reach full theoretical capacity during battery operation [1–3]. Different technical approaches have been used to overcome these drawbacks, such as: doping LFP with supervalent cations [3–5], reduction of particle size to nanoscale [3, 6] and/or particle coating with carbon based materials [7, 8]. By doping way, the enhancement of LFP electrochemical capacity has been reported with promising results. By titanium doping, Wu L. et al. [1] observed LFP capacity increase from125 to 150 Ah/kg at 0.1C discharge rate. Göktepe H. et al. [8] reported the electronic conductivity of 1.9×10–3 S/cm and capacity of 146 Ah/kg obtained by combining Yb3+ doped and carbon coated LFP to form LiYb0.02Fe0.98PO4/C composite. Similarly, through the variation of lanthanum amounts doped in LFP, Cho et al. [9] found out that with 1% lanthanum content (LiFe0.99La0.01PO4/C), the LFP structure remained unchanged, meanwhile, its capacity increased from 104 to 156 Ah/kg at 0.2C discharge rate and exhibited good cycling stability. Furthermore, Li et al. [10] examined manganese doped LiMnxFe1–xPO4 with 0.1 ≤ x ≤ 0.9, the highest energy density of 595 Wh/kg was obtained at x = 0.75. Finally, LFP doping with supervalent cations can be conducted via solvothermal method to form relatively pure product with controllable particle sizes [6]. Selection of non–toxic, inexpensive, and naturally affordable (abundant) doping metals is a common strategy to enhance performance of cathode materials. Currently, nickel has been usually applied based on these requirements, comparing to widely used cobalt in lithium ion battery [4, 5, 7, 8]. Furthermore, in LFP structure, P–O bond is flexible compared to the Fe–O bonding. Replacement of iron (rFe = 140 nm) by nickel with relatively smaller atomic radius (rNi = 135 nm) in the crystalline network, can narrow the band gap and depress the rigidity structure to favor the lithium ion diffusion pathway [4]. Besides, nickel (Z = 28: 3d84s2) as d sub–orbital element with a number of unpaired electrons can easily participate in chemical bonding. Finally, the redox potentials shifting to the higher value is also revealed after doping, which can be explained by changes in the Me–O covalent bonding such as: change in electronegativity of Me ion, Me–O bond length, by the influence of the Me–O–Me interactions in the solid composition [4, 11]. Nickel doping at 1–10% range was mentioned in various research works [10, 12]. According to [10], with > 10% doping, the iron replaced olivine structure tends to form defects; meanwhile, with < 5% doping, the XRD analysis could appear in a low detectable range, hinder the following structure interpretation. Thereby, a range of 5–10% nickel doping could be considered acceptable to examine the possibility of LFNP solvothermal synthesis and enhancement of lithium–ion battery performance. In this paper, the nickel doped LFP is prepared via solvothermal reaction with different nickel amounts; the effect of nickel doping on LFP structure and electrochemical properties are reported and discussed. 2. MATERIALS AND METHODS 2.1. Synthesis of LiFe1–xNixPO4 The starting reagents as Ni(NO3)2.6H2O (Merck, Germany), LiOH.H2O (Fisher, Belgium), FeSO4.7H2O (Merck, Germany), H3PO4 (Merck, Germany), C6H8O6 (Fisher, Belgium), ethylene glycol (Merck, Germany) of analytical grade were used for synthesis. The Li:Fe:Ni:P mass ratios were calculated, respectively: 3:0.9:0.1:1 and 3:0.95:0.05:1 for LiFe0.95Ni0.05PO4 and Enhancement of Li–ion battery capacity using nickel doped LiFePO4 as cathode material 278 LiFe0.9Ni0.1PO4. Firstly, LiOH.H2O was dissolved in 40 mL ethylene glycol under ultra– sonication until clear solution was obtained. The FeSO4.7H2O and Ni(NO3)2.6H2O mixture was then separately dissolved in 60 mL ethylene glycol under nitrogen atmosphere for 15 min. Then, the ascorbic acid was slowly added to totally reduce Fe3+ into Fe2+ to form a homogenous light green solution. The H3PO4 solution (85 wt%) was continuously dropped into light green solution and kept stirring in 30 min. The solution was finally transferred into autoclave under the argon atmosphere to perform the solvothermal reaction in 180 °C for 5 h. After reaction, the greyish precipitate was centrifuged, rinsed repeatedly with ethanol and dried in vacuum at 70 °C for 7–9 h. The sample was finally calcined at 550–600 °C in nitrogen atmosphere for 5 h to remove impurities. 2.2. Material analysis and characterization Iron contents in LFP and nickel doped LFP were analyzed by volumetric titration method using KMnO4 in concentrated H2SO4 medium to fully oxidize Fe2+ into Fe3+. The LFP and LFNP samples were stirred with 50 mL H2SO4 until formation of clear green solution. The solution was then titrated with 0.0125N KMnO4. Raman measurements were performed using Horiba Jobin Yvon LabRAM HR300 system (at Institute of Nanotechnology – VNUHCM) with 514.5 nm laser radiation. With 1 µm penetration, the vibration of LFNP bonds was determined and its structure was identified. The LFNP crystalline structure, phase purity and the particles size were characterized using a Rigaku/max 2500Pc and D8 Brucker X–ray diffractometer (XRD) with Cu–Kα radiation (λ=1.5418 Å, 2θ: 10° to 170° at a scan rate 0.25°/s – 1.00°/s). Analysis was performed at Center for Molecular and Nanoarchitecture, Ho Chi Minh city. Scanning electron microscopy (SEM) with Hitachi SEM S4800–NHE equipment was used to characterize the composite morphology. The measurements were conducted in the National Institute of Hygiene and Epidemiology. The LFNP powder was mixed with acetylene black and copolymer binder (PVdF–HFP) (weight ratio 80:10:10) in N–methyl pyrrolidone (NMP). The ink solution was pasted on the aluminum foil with 0.1 mm thickness and dried in vacuum for 24 h. A charge/discharge cycling test for Swagelok–type battery was carried out in liquid electrolyte LiPF6/EC–DMC (1:1) at room temperature. Cells were assembled in a glove box under argon atmosphere with < 2 ppm H2O. Electrochemical studies were carried out using a MPG2 Galvano/Potentiostat (Bio–Logic, France; Applied Physical Chemistry Laboratory, Univesity of Science, VNUHCM) in the potential window of 3.0 – 4.2 V versus Li/Li+ in the galvanostatic mode at the C/10 regime. 3. RESULTS AND DISCUSSION 3.1. Determination of iron content in LFP and LFNP samples The volumetric titration results of Fe2+ content in the LFP and LFNP shows in equation (1). The iron content (in gram) was calculated using equation (2). 5Fe2+ + MnO4– + 8H+5Fe3+ + Mn2+ + 4H2O (1) ࢓ࡲࢋ ൌ ૞ ൈ ࡯ࡷࡹ࢔ࡻ૝ ൈ ࢂࡷࡹ࢔ࡻ૝ ൈ ૞૟ ൈ ૚૙ି૜ (2) Suppose that all iron content is converted equivalently to Fe2+, theoretically, with LFP sample, an initial stoichiometry of Li: Fe: P = 3:1:1, the percentage of Fe content (%) in LFP samples can be calculated by following equation (3): La Thi Hang, et al. 279 %ࡲࢋሺࡸࡲࡼሻ ൌ ૞૟ࢄૠ࢞૜ࢄା૞૟ࢄାૢ૞ࢄ ൈ ૚૙૙ ൌ ૜૛. ૞૟% ሺ࢝࢏࢚ࢎ ࢄ:࢓࢕࢒ሻ (3) Similarly, iron contents were calculated for LiFe0.9Ni0.1PO4 and LiFe0.95Ni0.05PO4 as 29.30% and 30.28% respectively (Table 1). In Table 1, the correlation between the calculated and experimentally analyzed results is clearly observed for LFP and LFNP with acceptable deviation. The iron amount in LiFePO4 is higher than in LiFe0.9Ni0.1PO4 and LiFe0.95Ni0.05PO4 samples due to its 7÷10 wt% replaced by nickel doped in the structure. Table 1. Iron content in LFP and LFNP determined by volumetric titration. Sample KMnO4/H2SO4 titrated volume (mL) Iron content in the sample (g) Iron content in the samples (%) Analyzed Calculated Deviation LiFePO4 23.0 0.0805 30.53 32.56 6.2 LiFe0.95Ni0.05PO4 19.0 0.0665 25.21 29.30 14.0 LiFe0.9Ni0.1PO4 18.5 0.0648 24.55 30.28 19.0 3.2. Structural characterization The XRD patterns of LFP and LFNP solvothermally synthesized with 5 and 10% nickel contents presented in Figure 1 are similar. The XRD patterns indicate an orthorhombic structure with space group Pnma (JCPDS 96–101–1112). Thereby, the nickel insertion did not influence on LFP olivine structure. However, the red shift of XRD patterns for doped samples indicates the contraction of lattice parameters. A considerable shift of XRD peaks after nickel doping can be explained as follow: As a multivalent element, iron is easily oxidized during solvothermal synthesis, result a LFP product with low crystalline peaks [6]. Nickel replacement can eliminate iron oxidation, favoring the product formation with peaks of higher intensity, compared to undoped LFP. Furthermore, the particle sizes were estimated as 20÷60 nm using the Scherrer formula with βcosθ = kλ/D, where β is the full–width–at–half–maximum length of the diffraction peak on a 2θ scale and k is a close to unit constant. The lattice parameters can be estimated using XRD patterns data and the cell volume can also be deduced from orthorhombic (olivine) structure (Eq. 4). Within orthorhombic or cubic lattice, the α, β, γ values are assigned to 90o. ࢂ ൌ ࢇ࢈ࢉඥ૚ ൅ ૛܋ܗܛሺࢻሻ ܋ܗܛሺࢼሻ ܋ܗܛ ሺࢽሻ െ ܋ܗܛ ሺࢻ૛ሻ െ ܋ܗܛ ሺࢼ૛ሻ െ ܋ܗܛ ሺࢽ૛ሻ (4) It is clearly from Table 2, the lattice parameters (values of a, b, c) of undoped LFP is slightly decreased by nickel doping. The decrease of cell parameters is explained by replacement of iron Fe2+ (rFe2+ = 125 Å) by smaller Ni2+ ion (rNi2+ = 121 Å). It proves that nickel has been successfully doped into structure and partly replaced in the M1 sites (Li) or M2 sites (Fe) without impact on the olivine structure. Furthermore, the electronic conductivity of the materials could indirectly be determined from equations Q = I.t and L= 1/R (with Q: capacity, I: current; t: time (h), L: conductivity (S/cm), R: impedance (Ohm) [4] and experimental derived from electrochemical measurement data and calculated results in Table 2 show significant increase of conductivity for synthesized LFNP comparing to LFP (approx. 2 times). En 28 Sam LF LiF LiF do mo an int sym ant PO the vib 90 1 a ob cm hancement 0 Figure 1. Table 2. ples P (synthesize e0.95Ni0.05PO e0.9Ni0.1PO4 The vib cumented in des. The int d the extern ernal vibrati metric O– isymmetric 4 3– and M2+ vibrations rational ban 0–1200 cm– s well as in served in the –1 is assigne of Li–ion bat Comparative Data of lattic d) 4 ration featu the literatu ernal modes al ones are ons in term P–O bond O–P–O bon . It should be may be co ds, respectiv 1 (ν3 modes) ternal mode doped sam d to the D, G tery capacity XRD patterns e parameters a Lattice p a 10.335 10.087 10.080 res of oliv re [11]. Th (ν1, ν2, ν3 pseudo–rot s of symme bending, d bending. mentioned upled. In ely 150–30 . Nickel dop (ν4). The s ple. In addit bands of gr using nicke of undoped a nd conductiv arameters (Å b 6.004 4 5.885 4 5.875 4 ine–type co ose vibratio , ν4) refer to ations and try species F2(ν3) an The transla that this sep Figure 2, R 0 cm–1 (trans ed LFP supp eparation o ion, the appe aphite coati l doped LiFe nd nickel dop ity calculated ) V c .698 .682 .678 mpound (s ns are clas vibrations translations are A1(ν1) s tisymmetric tions includ aration is a aman spect lation mode ressed the t f ν4 modes arance of n ng in case of PO4 as cath ed (5 and 10 for LFP and olume (Å3) 291.5 277.9 277.0 pace group sified into i occurring in of the units ymmetric P P–O stret e motions o guide to disc roscopy sh s, T); 570–6 ranslation m at 995 cm–1 ew vibration nickel dope ode materia %) LFP samp LFNP sample Condu (S/c 1.0× 2.3× 2.5× D2h 16Pnm nternal and the PO43– te . Among th –O stretchin ching, and f the center ussion only owed typica 50 cm–1 (ν4 odes at 150– and 1067 c bands at 13 d LFP. l les. s. ctivity m) 10–3 10–3 10–3 a) were external trahedral ese, the g, E(ν2) F2(ν4) of mass , because lly LFP modes), 300 cm– m–1 was 40–1610 3.3 Figure 2. . Particle si Figure 3. SE Comparison o zes and mor M images of f Raman spec phology ch LFNP with 1 so tra between L aracterizati 0% of Ni (a; b lvothermal te FNP and LF on ) and 5% of N chnique. P (after calcin i (c; d) samp La Thi Ha ation at 550 ° les synthesiz ng, et al. 281 C). ed via Enhancement of Li–ion battery capacity using nickel doped LiFePO4 as cathode material 282 The morphology and particles sizes were determined by SEM images. Figure 3 shows relatively homogenous particle shapes and size distribution of LFNP particles ranging from 50 to 150 nm; The increase of nickel doping amount suppressed partly the agglomeration of particles. 3.4 Characterization of electrochemical performances Figure 4 compares the initial charge – discharge voltage profile of LFP and LFNP electrodes (xNi = 0.05; 0.10) at scan rate of C/10. All the samples have similar charge – discharge curve with flat plateaus corresponding to the lithium intercalation/de–intercalation in/out the olivine structure. The nickel doped LFP enhanced the charge – discharge voltage, typically 3.6 V towards 3.5 V for pure LFP in charge, and 3.4 V towards 3.5 V in discharge. Additionally, the significant increase of specific capacity was observed for LFNP samples. Thus, the doped nickel could enhance electronic conductivity and lithium diffusion rate which improve the capacity and cycle life of batteries. Figure 4. Initial charge – discharge curve at C/10 of LFP and LFNP samples. 4. CONCLUSIONS LiFePO4 and nickel doped samples were successfully synthesized using solvothermal technique. The nickel impact on LFP structure clearly exhibited by the decrease of volume and lattice cell parameter as well as suppression of some vibration bands (translation and internal modes). The capacity and voltage of doped samples enhanced significantly due to increase of electric conductivity and lithium diffusion rate. The results would be further on the aspect of structure evolution during charge – discharge to understand the diffusion pathway. Acknowledgements. This work was financially supported by Ho Chi Minh City University of Technology & Vietnam National University of Ho Chi Minh City through the Science and Technology Funds granted for C2015–20–25 Project. La Thi Hang, et al. 283 REFERENCES 1. Wu L., Wang Z. X., Li X. H., Li L. J., Guo H. J., Zheng J. C., Wang X. J. – Electrochemical performance of Ti4+–doped LiFePO4 synthesized by co–precipitation and post–sintering method, Transactions of Nonferrous Metals Society of China 20 (2010) 814–818. 2. Long Y., Shu Y., Ma X. H., Ye M. X. – In–situ synthesizing superior high–rate LiFePO4/C nanorods embedded in graphene matrix, Electrochimica Acta 117 (2014) 105– 112. 3. Hu J. Z., Xie J., Zhao X. B., Yu H. M., Zhou X., Cao G. S., Tu J. P. – Doping effects on electronic conductivity and electrochemical performance of LiFePO4, Journal of Materials Science & Technology 25 (3) (2009) 405–409. 4. Dahlin G. R., E. Strom K. E. – Lithium batteries, Research and Applications, Publishers: Nova Science. Inc. New York, 2016, pp. 226. 5. Ziółkowska D., Korona K. P., Kamińska M., Grzanka E., Andrzejczuk M., Wu S. H., Chen M. – Raman spectroscopy of LiFePO4 and Li3V2(PO4)3 prepared as cathode materials, Acta Physica Polonica 120 (5) (2011) 973–975. 6. La T. H., Nguyen N. T., Nguyen T. M. A., Le M. L. P., Doan L. V., Doan P. L. – Microwave–assisted solvothermal synthesis of LiFePO4/C nanostructure for lithium ion batteries, Proceedings of the 5th Asian Materials Data Symposium, Hanoi – Vietnam (2016) 343–352. 7. Satyavani T. V. S. L., Srinivas Kumar A. S., Subba Rao P. S. V. – Methods of synthesis and performance improvement of lithium iron phosphate for high rate Li–ion batteries: A review, Engineering Science and Technology, an International Journal 19 (2016) 178– 188. 8. Göktepe H. – Electrochemical performance of Yb–doped Li– LiFePO4/C composites as cathode materials for lithium–ion batteries, Research on Chemical Intermediates 39 (3) (2013) 2979–2987. 9. Cho Y. D., Fey G. T. –K., Kao H. M. – Physical and electrochemical properties of La– doped LiFePO4/C composites as cathode materials for lithium–ion batteries, Journal of Solid State Electrochemistry 12 (2008) 815–823. 10. Li G., Azuma H., Tohda M. – Optimized LiMnyFe1–yPO4 as the cathode for lithium batteries, Journal of the Electrochemical Society 149 (2002) A743–A747. 11. Baddour–Hadjean R., Pereira–Ramos J. P. – Raman microspectrometry applied to the study of electrode materials for lithium batteries, Chemical Reviews 110 (2010) 1278– 1319. 12. Ge Y. C., Yan X. D., Liu J., Zhang X. F., Wang J. W., He X. G., Wang R. S., Xie H. M. – An optimized Ni doped LiFePO4/C nanocomposite with excellent rate performance, Electrochimica Acta 55 (2010) 5886–5890.