Preparation of synthetic rutile and metal-doped LiFePO4 from ilmenite

Powder Technology 199 (2010) 293–297

Contents lists available at ScienceDirect

Powder Technology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / p ow t e c

Short Communication

Preparation of synthetic rutile and metal-doped LiFePO4 from ilmenite Ling Wu, Xinhai Li ⁎, Zhixing Wang, Xiaojuan Wang, Lingjun Li, Jie Fang, Feixiang Wu, Huajun Guo School of Metallurgical Science and Engineering, Central South University, Changsha 410083, PR China

a r t i c l e

i n f o

Article history: Received 23 November 2009 Received in revised form 3 January 2010 Accepted 26 January 2010 Available online 4 February 2010 Keywords: Ilmenite Synthetic rutile Cathode material LiFePO4 Doping

a b s t r a c t The synthetic rutile and metal-doped LiFePO4 are prepared from the high-titanium residue and iron-rich lixivium, which are obtained from the ilmenite by a mechanical activation and leaching process. ICP results show that the rutile contains 92.01% TiO2, 1.59% Fe2O3, 0.034% MnO2 and 0.60% (MgO + CaO), which meet the requirement of the titanium dioxide chlorination process. The results also reveal that small amounts of Al3+, Ca2+ and Ti4+ precipitate in the FePO4·xH2O precursor. XRD and Rietveld-refine results show that the metal-doped LiFePO4 is single olivine-type phase and well crystallized, and Ti4+ occupy M1 site, Ca2+ occupy M2 site and Al3+ occupy both sites, which indicates the formation of cation-deficient solid solution. The sample exhibits a capacity of 123 mAh g−1 at 5C rate, and retains 94.3% of the capacity after 100 cycles. © 2010 Elsevier B.V. All rights reserved.

1. Introduction Ilmenite (FeTiO3) is the primary global source of titanium dioxide which is commonly used in the manufacture of paints, paper, rubber, ceramics and Li–Ti–O anode materials, etc. Titanium dioxide is commercially manufactured by two main processes, namely the sulfate process and the dry chlorination process. Nowadays about 60% of the world's titanium dioxide is manufactured by the chlorination route [1], in which natural or synthetic rutile is used as raw material. The shortage of natural rutile has encouraged research efforts to convert ilmenite into synthetic rutile for the chlorination route. There are several processes for the production of synthetic rutile from ilmenite, such as smelting process [2], Becher process [3], MURSO process [4], ERMS process [5], etc. However, all these processes depend mainly on reductive and/or oxidative thermal pretreatment of ilmenite, which is an extensive energy consuming stage. And most of the subsequent acid leaching requires pressurized conditions, which make the processes complicated and costly. Therefore, a simple and economical route should be researched urgently. On the other hand, as a promising cathode material for lithium-ion batteries, LiFePO4 is typically prepared from the chemically pure or analytically pure iron salts, such as Fe(II)-oxalate [6–9], Fe(II)-acetate [10], Fe(III)-nitrate [11], Fe(II)-sulfate [12], FeCl2 [13], Fe2O3 [14], FePO4 [15], etc. Generally, these highly pure iron salts are prepared from iron-containing ores via complex processes of removing impurities. However, a proper amount of the impurities (Mg, Mn, Al,

⁎ Corresponding author. Tel./fax: + 86 731 88836633. E-mail address: [email protected] (X. Li). 0032-5910/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.powtec.2010.01.018

Ti, etc.) can improve the electrochemical performance of LiFePO4 remarkably, and many researchers also added metallic dopants when they prepared LiFePO4 from the highly pure iron salts [6–8,16–18]. Therefore, if we can prepare the high-performance metal-doped LiFePO4 directly from natural ores, a lot of needless reduplicate processes could be avoided and the cost of production would be reduced. In this study, we propose a simple, efficient and economical method to prepare synthetic rutile and lithium iron phosphate from natural ilmenite. Through this method, almost all of the titanium and iron of ilmenite are utilized, and the ultimate products are high value-added. 2. Experimental Natural ilmenite from Titanium Company of Panzhihua Steel and Iron Corporation, Sichuan, China, was used as raw material. Its particle size is 100–200 μm, and the chemical composition is as follows (wt.%): 47.60 TiO2, 32.81 FeO, 7.25 Fe2O3, 5.64 MgO, 3.35 SiO2, 1.66 Al2O3, 0.70 CaO and 0.663 MnO2. A planetary ball mill with a rotation speed of 200 rpm was employed for the mechanical activation. Four milling cells were fixed on its platform, and each cell was 500 ml stainless steel vessel filled with 250 g Φ20 mm, 200 g Φ10 mm and 50 g Φ5 mm steel balls. The milling was operated in air with ball/ ilmenite mass ratio of 20:1 for 2 h. The milled ilmenite was leached in a 500 ml round-bottomed flask attaching to a refluxing condenser. Firstly, 192 g 25 wt.% hydrochloric acid was heated to 100 °C, and then 40 g milled ilmenite was added to the solution under vigorous stirring. After 2 h, the slurry was rapidly cooled and filtered, and then high-titanium residue and iron-rich filtrate were obtained. The residue was washed with 5 wt.% HCl, dried and calcined at 900 °C for 4 h to obtain synthetic rutile.

294

L. Wu et al. / Powder Technology 199 (2010) 293–297

Metal-doped LiFePO4 was prepared by using the as-obtained lixivium (filtrate) as starting material. FePO4·xH2O precursor was synthesized by the following procedures: (1) The lixivium was boiled for 10 min to vaporize the residual HCl, then diluted with de-ionized water to obtain 0.25 M (Fe) solution; (2) sufficient concentrated hydrogen peroxide (30 wt.%) was added to the solution under vigorous stirring; (3) H3PO4 (85 wt.%) was added to the solution, an equimolar solution of FeCl3 and H3PO4 was obtained; (4) then NH3·H2O (2 M) was dropped slowly into the solution to control the pH = 2.0 ± 0.1, subsequently a white precipitate formed immediately; (5) after being stirred for 30 min, the precipitate was filtered, washed three times with de-ionized water and dried in an oven at 80 °C. Thus, FePO4·xH2O precursor was obtained. Crystalline metal-doped LiFePO4 was prepared by heating an amorphous LiFePO4 in a tubular furnace at 500 °C with flowing argon (99.999%) for 12 h. The amorphous LiFePO4 was obtained through lithiation of FePO4·xH2O precursor with oxalic acid as reducing agent at ambient temperature. Further experimental details can be found in Ref [12]. The elemental content of samples was analyzed using inductively coupled plasma emission spectroscopy (ICP, IRIS intrepid XSP, Thermo Electron Corporation), and the carbon concentration by C–S analysis (Eltar, Germany). The SEM image and elemental mapping of the particles were observed with scanning electron microscopy (SEM, JEOL, JSM-5600LV). The powder X-ray diffraction (XRD, Rint-2000, Rigaku) using CuKα radiation was employed to identify the crystalline phase of the synthesized material. X-ray Rietveld refinement was performed by FULLPROF. The electrochemical performance was performed using a twoelectrode coin-type cell (CR2025) of Li LiPF6 (EC:EMC:DMC = 1:1:1 in volume) LiFePO4. The working cathode is composed of 80 wt.% LiFePO4 powders, 10 wt.% acetylene black as conducting agent, and

10 wt.% poly (vinylidene fluoride) as binder. After being blended in N-methyl pyrrolidinone, the mixed slurry was spread uniformly on a thin aluminum foil and dried in vacuum for 12 h at 120 °C. A metal lithium foil was used as anode. Electrodes were punched in the form of 14 mm diameter disks, and the typical positive electrode loadings were in the range of 1.95–2 mg/cm2. A polypropylene micro-porous film was used as the separator. The assembly of the cells was carried out in a dry argon-filled glove box. The cells were charged and discharged over a voltage range of 2.5–4.1 V versus Li/Li+ electrode at room temperature. 3. Results and discussion 3.1. Synthetic rutile Fig.1(a), (b) and (c) shows the SEM images of ilmenite, 2 h milled ilmenite and synthetic rutile, respectively. The starting ilmenite exhibits large grains in the range of 100–200 μm, after ball milling for 2 h, the particle size was reduced to 0.2–1 μm. Ball milling could enhance the dissolution of ilmenite and, at the same time, accelerate hydrolysis of the dissolved titanium [1]. The particle size of the synthetic rutile is about 4–6 μm, which is finer than the required specification for the titanium dioxide chlorination process (N50 μm) [19]. Thus, an additional granulation process is needed to increase the particle size. Fig.1(d) displays the XRD patterns of ilmenite, hydrolysate and synthetic rutile. The diffraction peaks of ilmenite match well with the standard XRD pattern of hexagonal structure FeTiO3 (JCPDS no. 750519), and the milled sample displays weak and broad peaks, but no impurity phase appears. The XRD pattern of hydrolysate displays the diffraction peaks of rutile, which indicates that the rutile phase begins to form during the coupled dissolution and hydrolysis process.

Fig. 1. (a) SEM image of unmilled ilmenite; (b) SEM image of 2 h milled ilmenite; (c) SEM image of synthetic rutile. (d) XRD patterns of ilmenite, hydrolysate and synthetic rutile. Ilmenite: 1—unmilled, 2—2 h milled; hydrolysate: 3—2 h milled and leached; synthetic rutile: 4—2 h milled, leached and 4 h calcined at 900 °C.

L. Wu et al. / Powder Technology 199 (2010) 293–297

295

However, all the peaks are weak and broad, indicating low crystallinity. The hydrolysate was calcined at 900 °C for 4 h to give a TiO2 concentrate. Its XRD patterns match both the Bragg-position and intensity of rutile (JCPDS no. 71-0650), which demonstrates that a synthetic rutile with high crystallinity and purity was prepared. ICP results show that the synthetic rutile contains 92.01% TiO2, 1.59% Fe2O3, 0.034% MnO2 and 0.60% (MgO + CaO). The contents of MnO2 and (MgO + CaO) are low enough to meet the requirement of chlorination process. The Fe2O3 is slightly higher than the standard value (≤1.5%), which is generally accepted in industry, and it could be reduced by prolonging the leaching time or enhancing the acid/ore ratio. Furthermore, it is very satisfactory that the utilization rate of titanium reaches 99.1%. 3.2. LiFePO4 The chemical composition of starting material (lixivium) and FePO4·xH2O precursor is analyzed by ICP (Table 1). As shown, the lixivium contains large quantity of Mg and small amounts of Mn, Al, Ca and Ti. In order to precipitate the impurities selectively, H3PO4 is selected as a precipitator which is owing to the various solubility products of the corresponding metallic phosphates. It is noted that small amounts of Al, Ca and Ti are detected in FePO4·xH2O, whereas Mg and Mn are not detected. It is reported that Al3+, Ca2+ and Ti4+ ions doping could improve the electrochemical properties of LiFePO4 [6–8,16–18], therefore, a good performance can be expected. Additionally, the recovery rate of iron from lixivium reaches 99.8%; however, the total recovery rate of iron from ilmenite is 95.5%, which is ascribed to the residual iron in synthetic rutile. Fig. 2 shows the Rietveld-refined X-ray diffraction pattern of metal-doped LiFePO4, and the crystal parameters are summarized in Table 2. As shown, the sample is single phase and well crystallized, and all diffraction peaks are indexed in an orthorhombic system with the space group Pnma (JCPDS no. 40-1499). C–S analysis confirms that the residual carbon content of the sample is 0.84 wt.%. The carbon comes from the oxalic acid during annealing procedure. However, we cannot find the characteristic diffraction peaks of carbon, which indicates that the residual carbon is amorphous. The metal-doped LiFePO4 has lattice parameters a = 10.3350(6) Å, b = 6.0093(8) Å and c = 4.7009(6) Å, while the un-doped LiFePO4 has slightly varied lattice parameters a = 10.33392 Å, b = 6.00914 Å and c = 4.69484 Å in our previous report [12]. The unit cell of the crystal lattice was slightly expanded along x, y, and z directions due to the metal doping effect. In order to clarify whether the dopants occupy Li (M1) site or Fe (M2) site, the atom positions and occupancy were refined. We attempted all the possible occupying ways, and the best refinement result was only obtained when constrained Ti4+ to occupy M1 site, Ca2+ to occupy M2 site and Al3+ to occupy both sites. As shown, the observed and calculated patterns match well, and the reliability factors (Rp, Rwp and Rexp) are good. It is noted that Al3+ ions substitute on M1 and M2 sites with charge compensating vacancies on M1 site, and Ti4+ ions substitute on M1 site with charge compensating vacancies also on M1 site, as given in Eq. (1) (Kroger–Vink notation [7]). h

′

VLi

i

 ••   •   •••  = 2 AlLi + AlFe + 3 TiLi

ð1Þ

Table 1 The molar ratio of Fe, Mg, Mn, Al, Ca, Ti and Si in lixivium and FePO4·xH2O. Samples

Fe

Mg

Mn

Al

Ca

Ti

Si

Lixivium FePO4·xH2O

100 100

28.195 ∼0

1.688 ∼0

4.076 1.666

2.411 0.345

0.557 0.054

∼0 ∼0

Fig. 2. The Rietveld-refined X-ray diffraction pattern of metal-doped LiFePO4.

Therefore, according to the refinement results, the formula for metal-doped LiFePO4 can be described as (Li0.9742Al0.0038Ti0.0005Δ0.0215) (Fe0.9843Al0.0124Ca0.0033)PO4, where Δ represents vacant lattice sites. Furthermore, the Li, Fe, Al, Ca, Ti and P contents of the sample are also measured by ICP. It is found that the molar ratio of Li, Fe, Al, Ca, Ti and P is 0.973:0.985:0.016:0.003:0.0005:1, which is basically consistent with the Rietveld refinement results. The primary particle size of the sample is estimated to be 44.6 nm, according to the Scherrer formula d = 0.9λ/β1/2cosθ, where λ is the X-ray wavelength, β1/2 is the corrected width of the main diffraction peak (1 3 1) at half height and θ is the diffraction angle. SEM images of metal-doped LiFePO4 with corresponding EDS maps of Fe, P, Al, Ca and Ti are shown in Fig. 3. As shown, the sample consists of roughly spherical particles of about 1 μm diameter. The particle size is much larger than the calculated value by Scherrer formula, which is attributed to the particles aggregation. The distribution areas for Fe, P, Al, Ca and Ti are homogeneous, owing to the co-precipitation, which results in the atom-scale mixed of various elements. Furthermore, Mg, Mn and Si are not detected by EDS, which are consistent with the results obtained from ICP analysis. Fig. 4(a) displays the initial charge/discharge curves of metaldoped LiFePO4 with various C-rates at room temperature. As shown, the sample shows a capacity of 165 mAh g−1 at 0.1C rate, which approaches the theoretical capacity of 170 mAh g−1. By increasing the C-rate, the utilization rate of the active material decreases, and 163, 156, 141 and 123 mAh g−1 are delivered at 0.5C, 1C, 2C and 5C rates, respectively. From the shape of the discharge profiles, LiFePO4 electrode exhibits very long and plat plateau around 3.4 V, indicating little polarization. Cycling performance of the sample is shown in Fig. 4(b). It can be seen that the capacities waved with the room temperature.

Table 2 Results of structural analysis obtained from X-ray Rietveld refinement of metal-doped LiFePO4. Atoms

Site

x

y

z

Occupancy

Li Al Ti Fe Al Ca P O1 O2 O3

4a 4a 4a 4c 4c 4c 4c 4c 4c 8d

0 0 0 0.2826 0.2826 0.2826 0.0954 0.0965 0.4563 0.1645

0 0 0 0.25 0.25 0.25 0.25 0.25 0.25 0.0484 (10)

0 0 0 0.9737 0.9737 0.9737 0.4180 0.7423 0.2058 0.2839

0.9742 0.0038 0.0005 0.9843 0.0124 0.0033 1 1 1 1

(9) (9) (9) (8) (10) (7) (9)

(11) (11) (11) (9) (6) (8) (10)

(7) (11) (8) (9) (12) (15)

Space group: Pnma. Rp =8.94%, Rwp =10.41%, Rexp =6.15%. Cell constant (Å): a=10.3350(6), b=6.0093(8), c=4.7009(6); cell volume (Å3): 291.9546(9).

296

L. Wu et al. / Powder Technology 199 (2010) 293–297

Fig. 3. SEM image of metal-doped LiFePO4, with corresponding EDS maps of Fe, P, Al, Ca and Ti.

Fig. 4. The initial charge/discharge curves (a) and cycling performance (b) of metal-doped LiFePO4 with various C-rates at room temperature.

L. Wu et al. / Powder Technology 199 (2010) 293–297

After 100 cycles, the sample delivers a capacity of 147, 136 and 116 mAh g−1 at 1C, 2C and 5C rate, respectively, and retains 94.2%, 96.5% and 94.3% of its initial discharge capacity at corresponding C-rate. The excellent electrochemical properties of the final product confirm that synthesis high-performance LiFePO4 from ilmenite lixivium is feasible. However, the electrochemical performance of the sample is still not better than the un-doped LiFePO4 synthesized by the same route [12], which indicates that the metal doping effect is not significant for the electrochemical properties of LiFePO4. 4. Conclusions Synthetic rutile and metal-doped LiFePO4 are prepared from ilmenite, and the utilization rates of titanium and iron reach 99.1% and 95.5%, respectively. The impurity content of synthetic rutile meets the requirement of titanium dioxide chlorination process. The results demonstrate that the mechanical activation and leaching process can avoid the traditional high temperature pretreatment and high pressure leaching processes. The metal-doped LiFePO4, which can be described as the formula of (Li0.9742Al0.0038Ti0.0005Δ0.0215)(Fe0.9843Al0.0124Ca0.0033) PO4, shows excellent electrochemical performance at room temperature. It can be concluded that through this work, three goals could be achieved, namely the comprehensive utilization of ilmenite, preparation of high quality synthetic rutile and synthesis of high-performance LiFePO4. Based on the results, we believe that this method is a simple, efficient and economical way for both ilmenite utilization and materials preparation.

297

Acknowledgement The project was sponsored by the National Basic Research Program of China (973 Program, 2007CB613607).

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19]

C. Li, B. Liang, L. Guo, Hydrometallurgy 89 (2007) 1. R.H. Natziger, G.W. Elger, US Bureau of Mines, Report Invest No. 9065, 1987. R.G. Becher, Australian Patent 247110, 1963. H.N. Sinha, Proceedings of the Eleventh Commonwealth Mining and Metallurgical Congress, Institute of Mining and Metallurgy, London, 1979, p. 669. E.A. Walpole, Heavy Minerals, 169, SAIMM, Johannesburg, 1997. S.Y. Chung, J.T. Bloking, Y.M. Chiang, Nat. Mater. 1 (2002) 123. N. Meethong, Y.H. Kao, S.A. Speakman, Y.M. Chiang, Adv. Funct. Mater. 19 (2009) 1060. R. Amin, C. Lin, J. Maier, Phys. Chem. Chem. Phys. 10 (2008) 3524. B. Kang, G. Ceder, Nature 458 (2009) 190. Y. Hu, M.M. Doeff, R. Kostecki, R. Finones, J. Electrochem. Soc. 151 (2004) A1279. J. Ying, M. Lei, C. Jiang, C. Wan, X. He, J. Li, L. Wang, J. Ren, J. Power Sources 158 (2006) 543. J. Zheng, X. Li, Z. Wang, H. Guo, S. Zhou, J. Power Sources 184 (2008) 574. M. Konarova, I. Taniguchi, Powder Technol. 191 (2009) 111. H. Liu, Z. Wang, X. Li, H. Guo, W. Peng, Y. Zhang, Q. Hu, J. Power Sources 184 (2008) 469. Y. Wang, J. Wang, J. Yang, Y. Nuli, Adv. Funct. Mater. 16 (2006) 2135. L. Li, X. Li, Z. Wang, L. Wu, J. Zheng, H. Guo, J. Phys. Chem. Solids 70 (2009) 238. G.X. Wang, S. Bewlay, S.A. Needham, H.K. Liu, R.S. Liu, V.A. Drozd, J.-F. Lee, J.M. Chen, J. Electrochem. Soc. 153 (2006) A25. C.-J. Yan, G.T.-K. Fey, Y.-C. Lin, Abs. no. A01135-01967, ICMAT 2009 & IUMRS-ICA 2009 Meeting, 2009, Singapore. M.H.H. Mahmoud, A.A.I. Afifi, I.A. Ibrahim, Hydrometallurgy 73 (2004) 99.