Microwave-Assisted Solvothermal Synthesis and Characterization of

Article pubs.acs.org/cm

Microwave-Assisted Solvothermal Synthesis and Characterization of Various Polymorphs of LiVOPO4 Katharine L. Harrison and Arumugam Manthiram* Electrochemical Energy Laboratory & Department of Mechanical Engineering, University of Texas at Austin, Austin, Texas 78712, United States ABSTRACT: All three polymorphs of LiVOPO4 have been synthesized, for the first time, by a microwave-assisted solvothermal (MW-ST) method by adjusting the reaction media and conditions. The triclinic polymorph (αLiVOPO4) was obtained as the most stable and stoichiometric product and was thus chosen for optimization. Varying solvent mixtures consisting of water and alcohols/glycols proved to have significant effects on the particle size/morphology, with the water and glycol mixtures generally producing smaller particles. Particle size was also reduced with decreasing reaction time, decreasing reactant concentration, and CTAB surfactant addition. The electrochemical performance was analyzed in two potential ranges: 3.0−4.5 V corresponding to the insertion/extraction of one lithium and 2.0−4.5 V corresponding to the insertion/extraction of two lithium into/from the LiVOPO4 structure. Synthesis in a mixture of water and ethylene glycol in particular led to a high capacity of 134 mAh/g during the first cycle in the 3.0−4.5 V region. Initial attempts at coating the particles with PEDOT:PSS improved the cycling performance.

KEYWORDS: lithium-ion batteries, LiVOPO4 cathodes, chemical synthesis, polymorphic modifications, electrochemical properties

■

INTRODUCTION Most portable devices use layered Li[Mn,Ni,Co]O2 cathodes in lithium-ion cells, and there is immense interest to develop safe, less expensive cathodes for advanced transportation and grid energy storage. In this regard, polyanion cathodes have become appealing, following the initial investigation of the polyanion cathodes Fe2(SO4)3, Fe2(MoO4)3, and Fe2(WO4)3 by Manthiram and Goodenough.1,2 More recently, phosphates and silicates have become the subject of particular interest. Among them, the most promising and actively studied polyanion material is the olivine LiFePO4.3 It exhibits a high capacity of 170 mAh/g, but the energy density is limited due to its lower operating voltage of 3.45 V. LiFePO4 is further limited by its ability to insert/extract only one lithium ion per transition metal ion, as is the case with most cathode materials. There are only a few cathodes such as the Li2MSiO4 that have the potential to exhibit the reversible extraction of two lithium ions per transition metal ion. LiVOPO4 is another candidate that offers the potential to reversibly extract two lithium ions per vanadium ion (for a theoretical capacity of 318 mAh/g). Even if only one lithium ion is inserted/extracted, the theoretical capacity is 159 mAh/g at 4 V, leading to a higher energy density than LiFePO4. However, the poor electronic and ionic © XXXX American Chemical Society

conductivities require novel synthesis approaches to realize this goal. Accordingly, we present here the synthesis of various LiVOPO4 polymorphs via a novel MW-ST approach. There is much literature dedicated to delithiated VOPO4, which forms in seven polymorphic modifications, but those modifications can only be used with lithiated anode materials.4−11 LiVOPO4 forms in three different polymorphic modifications including triclinic (α), orthorhombic (β), and tetragonal (α1),10,12−27 which could be used with lithium-free anodes. There has been significant emphasis in the literature on synthesizing LiVOPO4 by many different methods because the various methods lead to widely varying electrochemical performances. The triclinic phase has been synthesized by a solution-based reaction followed by solid-state heating,12,13,19,26 chemical lithiation of VOPO4,18 glass ceramic processing,21 and hydrothermal22 and solvothermal methods.24 The orthorhombic phase has been synthesized by a solution-based reaction followed by solid-state heating,12,14 carbothermal reduction of VOPO4,15 chemical lithiation,16 hydrothermal,20 glass ceramic Received: January 19, 2013 Revised: March 11, 2013

A

dx.doi.org/10.1021/cm400227j | Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials

Article

processing,21 sol−gel,23 and microwave-assisted solid-state heating methods.25 The tetragonal phase has been synthesized by chemical lithiation10 as well as hydrothermal synthesis followed by conventional heating.17 The processing conditions are critical since favorable morphology and small particle size can improve electrochemical performance in materials like LiVOPO4 with poor electronic and ionic conductivities.15,20,26 Previous reports have demonstrated the importance of small particle size for obtaining good electrochemical performance with LiVOPO4 even at very slow rates.13,14 Similar to LiFePO4, it has also been shown that the electrochemical performance of LiVOPO4 can be improved by conductive coating of the particles.23 Microwave-assisted hydrothermal (MW-HT) and microwave-assisted solvothermal (MW-ST) synthesis methods have been employed by our group previously to produce olivine LiMPO4 (M = Mn, Fe, Co, and Ni) with small particle size and interesting morphologies within a very short reaction time of 5−10 min at less than 300 °C.28−32 Microwave synthesis techniques are used primarily because of their ability to dramatically reduce processing time, improve yields, or improve properties through control of particle size, morphology, purity, and crystallinity.33−35 Conventional solvothermal heating is slow because it depends on thermal conduction/convection of heat from the outside of a reaction vessel to the solution. It is also inefficient because an entire oven must be heated to heat the vessels and, in turn, the solution. Microwave heating is fundamentally different because the heating mechanism involves raising the temperature of the solution directly and uniformly through volumetric dielectric and ohmic heating.33−35 With an aim to obtain LiVOPO4 with controlled particle size and morphology, we present here, for the first time, the synthesis of three polymorphs of LiVOPO4 by the MW-ST method by varying the reaction medium and conditions. The triclinic polymorph (α-LiVOPO4) formed with the most stoichiometric elemental ratios, so this work focuses on optimizing the electrochemical performance of the triclinic phase. The orthorhombic (β-LiVOPO4) and tetragonal (α1LiVOPO4) polymorphs will be further optimized in the future.

■

was equipped with a stir bar, and the turntable was spun during synthesis. Four vessels were always used in a given synthesis run, and the power during the temperature ramp time was set to 600 W. The temperatures and pressures of each vessel were monitored during the reactions. In order to screen many different combinations of reaction conditions quickly, four different solutions were run simultaneously with a maximum pressure set to 40 bar and the total reaction time set to 50 min. This included approximately 25 min of ramping time to the final temperature. The ratios of precursors and solvents were varied to test many different conditions, as will be subsequently discussed. After each synthesis run, the microwave convectively cooled the vessels to 50 °C. The precipitates were washed with water and acetone and centrifuged several times. The products were then dried overnight in a vacuum oven at 150 °C. The α-LiVOPO4 (triclinic) and α1-LiVOPO4 (tetragonal) polymorphs were green in color, and the β-LiVOPO4 (orthorhombic) polymorph was brown. Material Characterization. X-ray diffraction (XRD) was performed on Philips and Rigaku Ultima IV X-ray diffractometers with filtered Cu Kα radiation. Lattice parameters were obtained by Rietveld refinement36 of the XRD patterns in Fullprof/WinPLOTR.37,38 Elemental ratios were determined with a Varian 715ES ICP optical emission spectrometer. Four standards were prepared for each element by diluting concentrated commercial ICP standards. Errors between calibration standard data points and the calibration curve were always less than 2−3%. Scanning electron microscopy (SEM) images were obtained with JEOL JSM-5610 and FEI Quanta 650 SEMs. Fourier transform infrared spectroscopy (FTIR) was also utilized here on a PerkinElmer BX FTIR. Pellets for FTIR analysis were prepared by grinding and pressing samples with dried KBr powder. Electrochemical Characterization. Electrodes were prepared by grinding the as-prepared active material with conductive carbon and teflonated acetylene black (TAB) in a mortar and pestle; TAB consists of polytetrafluoroethylene (PTFE) and acetylene black. It should be noted that the samples were not subjected to any carbon coating. Electrodes were fabricated with 70 wt % active material, 15 wt % carbon, and 15 wt % TAB. The resulting composites (∼5 mg active material) were rolled into thin sheets and cut into 0.64 cm2 area circles using a punch (loading of ∼7.8 mg/cm2 or ∼1.24 mAh/cm2 based on one Li+ insertion/extraction or ∼2.48 mAh/cm2 based on two Li+ insertion/extraction). They were then pressed onto Al mesh to be used as a current collector. The electrodes were dried overnight in a vacuum oven at 115 °C before constructing the cells. CR2032 coin cells were assembled in an argon filled glovebox with a metallic lithium anode, Celgard polypropylene separator, and 1 M LiPF6 in 1:1 diethyl carbonate/ethylene carbonate as the electrolyte. The coin cells were cycled at a C/20 rate on Arbin battery cyclers. All capacities are specified in reference to the weight of the active material.

EXPERIMENTAL SECTION

Synthesis. Various ratios of vanadium (V) oxide (Alfa Aesar), oxalic acid (Fisher), and phosphoric acid (Fisher) precursors were dissolved in water and stirred overnight until the solutions became transparent. Oxalic acid was added according to reaction 1 as V2O5 + 3H 2C2O4 → 2VOC2 O4 + 3H 2O + 2CO2

■

RESULTS AND DISCUSSION Synthesis of LiVOPO4 Polymorphs. Conventional hydrothermal synthesis methods have been demonstrated in the literature for all three polymorphs of LiVOPO4; thus, it seemed likely that microwave-assisted hydrothermal methods for LiVOPO4 could be developed. The conventional hydrothermal method for β-LiVOPO4 consists of a Li:V:P = 1:1:3.3 mixture of precursors heated at 450 °C for 7 days.20 The hydrothermal method in the literature for synthesizing α-LiVOPO4 consists of a Li:V:P mixture of 9:3:1 heated at 250 °C for 48 h with the addition of N2H4.H2O as a reducing agent.22 The α1 polymorph of LiVOPO4 has been synthesized by a two-step process. First, LiVOPO4·2H2O was synthesized by a conventional hydrothermal process at 120 °C for 48 h with stoichiometric precursors and oxalic acid as a reducing agent. The hydrothermal step was followed by dehydration at 300 °C under vacuum.17 Since the existing methods indicate that nonstoichiometric ratios of precursors are often required to obtain

(1)

Then lithium hydroxide monohydrate (Fisher) was added, and the solutions were stirred for an additional hour. The overall reaction that occurs is then VOC2 O4 + LiOH + H3PO4 → LiVOPO4 + H 2O + H 2 + 2CO2

(2)

In some syntheses, various amounts of alcohols (methanol, ethanol, isopropanol, butanol, octanol, ethylene glycol (EG), tetraethylene glycol (TEG), and polyethylene glycol (PEG)) were added just before transferring the solutions to 100 mL polytetrafluoroethylene (PTFE) microwave reaction vessels. The total amount of solution was always 15 mL, and the concentration was kept at 0.067 M in V unless otherwise noted. These vessels were inserted into ceramic liners and sealed such that autogenous pressure was generated during the reaction. The vessels were secured on a rotor, which was placed on a turntable in a microwave reaction system (Anton Paar Synthos 3000). To ensure uniform microwave heating and reaction mixing, each vessel B



Article

Table 1. Dominant Products from the MW-ST Synthesis in Water or Mixed Water and Ethanol Solvent with Various Ratios of Li:V:Pa water:ethanol Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P Li:V:P a

= = = = = = = = = = = = = = = = = = = =

1:1:1 1:1:2 1:1:3 1:1:4 2:1:1 3:1:1 4:1:1 5:1:1 2:1:2 3:1:3 4:1:4 5:1:5 4:1:3 6:1:3 9:1:3 12:1:3 3:1:2 4:1:2 5:1:2 6:1:2

1:0

3:1

1:1

1:3

amorphous VPO4·H2O HVOPO4·0.5H2O HVOPO4·0.5H2O no precipitate Li3PO4 Li3PO4 Li3PO4 amorphous no precipitate unknown amorphous no precipitate Li3PO4 Li3PO4 Li3VO4 no precipitate Li3PO4 Li3PO4 Li3PO4

no precipitate HVOPO4·0.5H2O HVOPO4·0.5H2O HVOPO4·0.5H2O amorphous no precipitate Li3PO4 Li3PO4 amorphous α1-LiVOPO4 α1-LiVOPO4 α1-LiVOPO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4

amorphous HVOPO4·0.5H2O HVOPO4·0.5H2O + α-LiVOPO4 HVOPO4·0.5H2O α1-LiVOPO4 Li3PO4 Li3PO4 Li3PO4 α-LiVOPO4 α-LiVOPO4 α-LiVOPO4 α-LiVOPO4 α-LiVOPO4 HVOPO4·0.5H2O Li3PO4 HVOPO4·0.5H2O α-LiVOPO4 Li3PO4 Li3PO4 Li3PO4

β + α1-LiVOPO4 β-LiVOPO4 β-LiVOPO4 β-LiVOPO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4 unknown unknown unknown Li3PO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4 Li3PO4

All reactions were run with a maximum pressure of 40 bar for a total reaction time of 50 min and a V concentration of 0.067 M.

was between 200 and 225 °C. In contrast to the reactions run in water, which led to no LiVOPO4 formation at 40 bar, many different conditions in the mixed solvent reactions produced LiVOPO4, and the products formed at lower temperatures in the mixed solvent solutions at the fixed pressure chosen here. Since impurities also formed for the reactions summarized in Table 1, the reaction conditions (temperature and precursor ratios) had to be further varied to obtain more phase-pure materials. Figure 1 shows the XRD patterns for the resulting materials compared to the patterns in the database.10,20,27 The synthesis conditions and elemental ratios determined by ICP are shown in the insets of the XRD patterns. It is clear that only the α phase gives stoichiometric elemental ratios. Patterns for the α1 and α polymorphs show no impurities, but the peak intensity ratios for the α1 polymorph are different than expected. Specifically, the peak at ∼20° has a lower intensity than is typically observed. The varied peak intensity could be due to defects, since the ICP data indicate something of the form Li1.16(VO)0.91PO4, which is approximately charge balanced assuming all V4+, but the deviations from the expected stoichiometry could alternatively indicate amorphous impurities. The β polymorph also showed unexpected peak intensity ratios in addition to lithium deficiency and some slight impurities, including a small amount of α-LiVOPO4. Conversely, the α polymorph is phase pure, stoichiometric, and shows no obvious peak intensity ratio variations from the expected patterns. Lattice parameters obtained with the Rietveld refinement of the three polymorphs are presented in Table 2. The lattice parameters for the α-LiVOPO4 polymorph agree well with the literature values,27 and the refinement fit is very good. The fits show more error for the β-LiVOPO4 and α1-LiVOPO4 polymorphs, which is not surprising because of the nonstoichiometry of the samples. The unit cell volumes were larger than most literature values for both the β-LiVOPO4 and α1-LiVOPO4 polymorphs, although there is a range of values in the literature.10,15−17,20,21,25

LiVOPO4, synthesis attempts were made with stoichiometric precursors, excess P, excess Li, and both excess P and excess Li. The effectiveness of the volumetric dielectric and ohmic heating employed during a microwave reaction is dependent on the nature of the solvent and the dissolved species in solution, so adjustment of the precursor ratios causes significant variance in the temperature and pressure relationships during the reactions. It is difficult to predict the maximum possible temperature that can be reached before actually running a reaction. On account of this difficulty and the desire to test many different reaction conditions, initial experiments to screen for promising conditions were all performed with a maximum pressure set to 40 bar rather than with a maximum temperature. Once 40 bar was reached, the reaction was held at this final pressure until 50 min of total reaction time had passed. The maximum temperature reached typically varied between 215 and 235 °C for these reactions with the exception of one sample that only achieved 180 °C. Under these conditions, no LiVOPO4 polymorphs formed, as shown in the first column of Table 1. The temperature was then increased to 250 °C for the reactions with the same precursor ratios used in the conventional hydrothermal methods,17,20,22 but LiVOPO4 still did not form (although the precursors were not all exactly the same as in conventional hydrothermal reactions in the literature). Because obtaining LiVOPO4 in water did not appear promising from the initial results, the solvent was varied. V2O5 does not dissolve in alcohols easily even with the addition of oxalic acid, so water and ethanol were mixed in 3:1, 1:1, and 1:3 ratios, and microwave reactions were performed for the various precursor ratios with a maximum pressure of 40 bar. The ethanol was added as the last step in precursor preparation just before transferring the solutions to the microwave vessels. The results are outlined in Table 1, which shows the dominant phase that formed (impurities also formed) during each reaction. The maximum temperature reached in these reactions C



Article

Table 3. Examples of Transition to α Polymorph with Increased Pressure (Which Equates to Increased Temperature)a precursor ratio

water:ethanol ratio

dominant product at 40 bar

Li:V:P = 1:1:1

1:3

Li:V:P = 1:1:2 Li:V:P = 1:1:3

1:3 1:3

β-LiVOPO4 + α1-LiVOPO4 β-LiVOPO4 β-LiVOPO4

Li:V:P = 1:1:4

1:3

β-LiVOPO4

Li:V:P = 5:1:5 Li:V:P = 5:1:5

3:1 1:3

α1-LiVOPO4 Li3PO4

dominant product at 50 bar α1-LiVOPO4 α-LiVOPO4 β-LiVOPO4 + α-LiVOPO4 β-LiVOPO4 + α-LiVOPO4 α-LiVOPO4 α-LiVOPO4

a

All tests were run to a maximum of either 40 or 50 bar with a V concentration of 0.067 M for a total reaction time of 50 min.

Figure 1. XRD patterns for the three polymorphs of LiVOPO4 compared to the patterns in the database.10,20,27 All three samples were prepared with a V concentration of 0.067 M and a 50 min total reaction time. Slight impurities for β-LiVOPO4 are indicated with red arrows.

Figure 2. FTIR absorbance spectrum of α-LiVOPO4 synthesized with a V concentration of 0.067 M, Li:V:P = 5:1:5, water:ethanol = 3:1, and a total reaction time of 50 min (∼25 min ramp time to 230 °C and ∼25 min hold time at 230 °C).

Upon varying the temperatures in an attempt to obtain phase pure polymorphs, it was common for the orthorhombic and tetragonal phases to transition into the triclinic phase as the temperature and pressure was increased, which indicates that the other phases form as intermediates and may be metastable under these reaction conditions. Several examples that show the transition to the triclinic polymorph with increasing temperature are shown in Table 3. Because the triclinic phase was the easiest to synthesize and formed the most stoichiometric product, it was selected for further optimization. FTIR was also used to characterize the α-LiVOPO4 sample, as shown in Figure 2. The FTIR spectrum matches the literature well with no detectable impurities,26,39 further confirming the phase purity of the sample. The VO bond peak is clearly visible near 900 cm−1. Although VO bond peaks generally occur at higher wavenumbers near 1000 cm−1, there are many examples of vanadium-based phosphates in the literature that exhibit VO bonds near 900 cm−1, as is the case for LiVOPO4.25,26,39−42 The ν1 and ν3 peaks arise from symmetric and asymmetric bending vibrations of the PO4 tetrahedra, respectively, and the ν2 and ν4 peaks arise from bending vibrations of PO4. Effects of the Solvent. Phase-pure α-LiVOPO4 was synthesized at 230 °C in 3:1 water:ethanol, so it was of interest to determine whether a phase-pure material could be

synthesized in the other solvent ratios at the same temperature and with the same precursor ratios. Because a maximum pressure was used as the limit in the initial reactions described in the previous section, the temperatures were not constant for all of the samples. Thus, the reactions in pure water, 1:1 water:ethanol, and 1:3 water:ethanol were repeated all at 230 °C with Li:V:P = 5:1:5 to see how critical the solvent mixture was to obtain the pure phase sample (at a constant temperature); the XRD results are shown in Figure 3. It is clear that impurities form for all conditions except the 3:1 mixture of water:ethanol, so the ratio of ethanol to water is critical. It was possible to obtain phase-pure α-LiVOPO4 with 100% water as the solvent and Li:V:P = 5:1:5, but the temperature had to be elevated to 240 °C. ICP data for the sample synthesized in water at 240 °C revealed nearly stoichiometric elemental ratios, as shown in the inset on the XRD pattern in Figure 3. Because α-LiVOPO4 could be synthesized in mixtures of ethanol and water, we were curious whether it could also be synthesized in other water/alcohol mixtures. To see whether the specific alcohol had an effect on the products, α-LiVOPO4 was synthesized at 230 °C with Li:V:P = 5:1:5 and 3:1 mixtures of water and methanol, ethanol, isopropanol, butanol, octanol, polyethylene glycol (PEG), ethylene glycol (EG), and tetra-

Table 2. Lattice Parameters of the LiVOPO4 Polymorphs polymorph

a (Å)

b (Å)

c (Å)

α

β

γ

V (Å)

χ2

Rwp

RBragg

α α1 β

6.7872(3) 6.3166(8) 7.467(2)

7.2152(2) 6.3166(8) 6.338(2)

7.8861(3) 4.4337(4) 7.172(2)

89.904(2) 90 90

88.578(2) 90 90

62.835(3) 90 90

343.46(2) 176.90(4) 339.38(15)

1.35 5.19 5.25

18.9 32.3 33.8

6.85 41.8 41.1

D



Article

Single-phase α-LiVOPO4 formed for all of these samples, except for the samples synthesized in mixtures of water with octanol, EG, and methanol. Synthesis in a mixture of octanol and water did not lead to a decipherable phase, and synthesis in the methanol or EG and water mixtures led to just a very slight impurity which we believe to be β-LiVOPO4. Elevating the temperature to 235 °C for these two samples led to phase pure materials (but increasing the reaction time did not). Elemental ratios for the samples which formed pure or nearly pure phases are indicated in the insets of the XRD patterns. The samples generally showed close to stoichiometric elemental ratios, with slight Li deficiency in some of the samples, especially for the samples prepared with mixtures of glycols and water. The sample prepared in a mixture of EG and water, specifically, presented 12% Li deficiency. Clearly, a large excess of Li was added to the reaction vessel, so this deficiency is not due to lack of Li+ ions in solution. Instead, this either indicates lithium deficiency such that the vanadium charge is V4.12+ rather than the V4+ that is implied by reactions 1 and 2 or the presence of an amorphous impurity. The XRD patterns were generally highly crystalline, but the sample prepared in a mixture of EG and water presented wider peaks than the other samples, possibly indicating small particles or a lower degree of crystallinity. It is worth noting that preparing LiVOPO4 in a mixture of EG and water at 235 °C led to a phase pure pattern, but the elemental ratios and peak widths remained similar to the sample prepared at 230 °C shown here so the lithium deficiency is not due to the small impurity in the pattern at 230 °C. Lattice parameters are presented in Table 4 for all of the phase-pure and nearly phase-pure samples. Clearly, the lattice parameters are very similar for all samples except for the sample prepared in water and EG, for which the unit cell volume is significantly lower than those of the other samples. This is consistent with the lithium deficiency indicated by ICP. The samples prepared in mixtures of water and other glycols show only a small decrease in the lattice parameters compared to the samples prepared in water and alcohols. SEM images for the materials synthesized in various water and alcohol mixtures are shown in Figure 5. The materials form in microflower-like morphologies with similar sizes for most of the simple alcohols. The samples synthesized in 3:1 mixtures of water to butanol and the various glycols showed smaller particle sizes. The smallest and most uniform particles were synthesized in mixtures of water and EG. The morphology of the sample prepared in a mixture of EG and water also differed; instead of flowers with distinct “petals”, these samples comprised of more loosely agglomerated primary particles with a less regular shape.

Figure 3. XRD of α-LiVOPO4 prepared with a V concentration of 0.067 M, Li:V:P = 5:1:5, and a total reaction time of 50 min (∼25 min ramp time to 230 °C and ∼25 min hold time at 230 °C) in various water:ethanol ratios. All samples were prepared at 230 °C except for the sample prepared in water, which was held at 240 °C.

ethylene glycol (TEG). The XRD patterns are shown in Figure 4 for the products resulting from each of these solvent mixtures.

Figure 4. XRD patterns of α-LiVOPO4 prepared with a V concentration of 0.067 M, Li:V:P = 5:1:5, and a total reaction time of 50 min (∼25 min ramp time to 230 °C and ∼25 min hold time at 230 °C) in various solvent mixtures.

Table 4. Lattice Parameters of α-LiVOPO4 Prepared with a V Concentration of 0.067 M, Li:V:P = 5:1:5, and a Total Reaction Time of 50 min (∼25 min Ramp Time to 230 °C and ∼25 min Hold Time at 230 °C)a

a

solvent

a (Å)

b (Å)

c (Å)

α

β

γ

V (Å)

χ2

Rwp

RBragg

3:1 water:ethanol 3:1 water:methanol 3:1 water:butanol 3:1 water:isopropanol 3:1 water:EG 3:1 water:TEG 3:1 water:PEG water 240 °C

6.7872(3) 6.7759(3) 6.7858(3) 6.7850(2) 6.790(1) 6.7950(4) 6.7967(4) 6.7773(2)

7.2152(2) 7.2135(2) 7.2143(2) 7.2128(2) 7.199(1) 7.2127(3) 7.2109(3) 7.2127(2)

7.8861(3) 7.8965(3) 7.8844(3) 7.8878(3) 7.864(1) 7.8745(4) 7.8720(4) 7.8935(2)

89.904(2) 89.895(2) 89.890(2) 89.897(2) 90.18(1) 89.980(3) 90.014(4) 89.903(2)

88.578(2) 88.631(2) 88.588(2) 88.579(2) 88.741(9) 88.581(3) 88.572(3) 88.619(2)

62.835(3) 62.885(2) 62.830(3) 62.840(2) 62.83(1) 62.799(4) 62.821(4) 62.886(2)

343.46(2) 343.43(2) 343.27(2) 343.33(2) 341.87(9) 343.12(3) 343.08(3) 343.33(2)

1.35 1.45 1.34 1.39 1.21 1.41 1.39 1.45

18.9 18.8 19.9 19.5 22.7 20.1 21.2 21.2

6.85 6.71 6.54 6.53 11.8 6.79 6.55 6.99

All samples were prepared at 230 °C expect for the sample prepared in water, which was held at 240 °C. E



Article

Figure 6. SEM of α-LiVOPO4 prepared with Li:V:P = 5:1:5 and water:ethanol = 3:1 for the various indicated reaction hold times at 230 °C (not including ∼25 min ramp time to 230 °C), V concentrations, and amounts of CTAB solution substituted for water.

samples prepared with 0.5 and 1.0 mL of CTAB solution were phase-pure α-LiVOPO4. The particle size decreased with increasing CTAB for the same reaction time, as shown in Figure 6, and smaller particle sizes were obtained than for the samples prepared with water and glycol mixtures as well as the samples prepared with short reaction times and lower precursor concentrations. The sample synthesized with 0.5 mL of CTAB solution still showed microflower like morphology, but the sample synthesized with 1.0 mL of CTAB showed more irregular particles. All attempts to further reduce the particle size led to amorphous products. ICP analysis was carried to determine the elemental ratios for the samples prepared with shorter reaction times, with lower concentrations, and with CTAB. The results are presented in Table 5. The data reveal close to stoichiometric ratios except for the sample prepared with 1.0 mL of CTAB solution, which shows significant Li deficiency compared to the other samples despite a phase-pure XRD pattern. The elemental ratios suggest that the sample may have defects or impurities.

Figure 5. SEM of α-LiVOPO4 prepared with a V concentration of 0.067 M, Li:V:P = 5:1:5, water:ethanol = 3:1, and a total reaction time of 50 min (∼25 min ramp time to 230 °C and ∼25 min hold time at 230 °C) in various mixed solvents. All samples were prepared at 230 °C except for the sample prepared in water, which was held at 240 °C.

Controlling Particle Size. Initially, a very long hold time at 230 °C was used to ensure that products had sufficient time to form (50 min total reaction time with a ∼25 min initial ramp time), but long hold times lead to particle growth, which contribute to poor charge−discharge capacities. Therefore, the hold time was varied from 5 to 45 min for the sample synthesized in water and ethanol to determine the effect of reaction time on particle growth. A hold time of at least 10 minutes was necessary to obtain phase-pure α-LiVOPO4 in water and ethanol solvent mixtures. The particle size was found to decrease with decreasing reaction time as indicated in Figure 6. Decreasing the precursor concentration also decreased the particle size, as has been previously demonstrated for microwave synthesized LiFePO4.30 The concentration may need to be decreased further to see a more appreciable effect, but attempts to decrease the concentration to significantly lower values led to amorphous products. Furthermore, it is preferable to control particle size with reaction time rather than concentration to maximize product yield. In an attempt to decrease particle size further, a surfactant (cetyl trimethylammonium bromide (CTAB)) was used to prevent particle growth and agglomeration. A total of 0.5−1.5 mL of CTAB extraction solution (Teknova 2% CTAB, 100 mM HCL pH 8.0, 20 mM EDTA pH 8.0, and 1.3 M NaCl) was substituted for water in the precursor solutions such that there was still a 3:1 mixture of water + CTAB solution to ethanol. The sample with 1.5 mL of CTAB was amorphous, but the

Table 5. Elemental Analysis of Samples Prepared with a 5:1:5 Li:V:P Ratio in 3:1 Water:Ethanol with Varying Concentrations, Reaction Times (∼ 25 min Ramp Time to 230 °C Plus the Documented Hold Time at 230 °C), and Amounts of CTAB Solution Substituted for Water

F

hold time

V concentration

CTAB volume

V/P

Li/P

10 15 25 25 25 25 45

0.067 0.067 0.067 0.045 0.067 0.067 0.067

0 mL 0 mL 0 mL 0 mL 0.5 mL 1.0 mL 0 mL

0.98 0.97 0.98 0.98 0.95 0.97 0.97

0.97 0.98 0.97 0.97 0.99 0.86 0.96



Article

Table 6. Lattice Parameters of α-LiVOPO4 Samples Prepared with a 5:1:5 Li:V:P Ratio in 3:1 Water:Ethanol with Varying Concentrations, Reaction Times (∼25 min Ramp Time to 230 °C, Plus the Documented Hold Time at 230 °C), and Amounts of CTAB Solution Substituted for Water if Present solvent

a (Å)

b (Å)

c (Å)

α

β

γ

V (Å)

χ2

Rwp

RBragg

10 min, no CTAB, 0.067 M in V 25 min, no CTAB, 0.067 M in V 45 min, no CTAB, 0.067 M in V 25 min, no CTAB, 0.045 M in V 25 min, 0.5 mL CTAB, 0.067 M in V 25 min, 1.0 mL CTAB, 0.067 M in V

6.7802(3) 6.7872(3) 6.7771(2) 6.7783(3) 6.7579(2) 6.720(1)

7.2099(3) 7.2152(2) 7.2126(2) 7.2108(2) 7.2121(1) 7.1878(8)

7.8932(4) 7.8861(3) 7.8955(3) 7.8917(3) 7.9189(2) 7.949(1)

89.911(3) 89.904(2) 89.901(2) 89.906(2) 89.827(1) 89.997(8)

88.619(3) 88.578(2) 88.614(2) 88.604(2) 88.711(1) 89.122(7)

62.858(3) 62.835(3) 62.887(2) 62.861(3) 63.002(1) 63.156(9)

343.24(3) 343.46(2) 343.40(2) 343.13(2) 343.79(1) 342.55(8)

1.31 1.35 1.48 1.38 2.17 1.84

19.6 18.9 20.9 20.7 12.7 18.3

5.81 6.85 5.74 6.88 4.88 6.64

Lattice parameters are presented in Table 6 for the samples prepared with varying reaction time, concentration, and amount of surfactant. The results show that the unit cell volumes are very similar to those for the α-LiVOPO4 samples presented in Tables 2 and 4, except for the sample prepared with 1.0 mL of CTAB solution, which showed a lower unit cell volume. This again, similar to the sample prepared in EG and water presented in Table 4, agrees with the ICP data, suggesting lithium deficiency for the sample prepared with 1.0 mL of CTAB. Electrochemical Performance. To determine the effects of the solvent on the electrochemical performance, coin cells were fabricated, and the performance at C/20 is shown in Figure 7 for samples prepared in water and various alcohol mixtures. In this work, cells were cycled at the potential ranges of 2.0−4.5 and 3.0−4.5 V. It is typical to cycle LiVOPO4 between 3.0 and 4.5 V, which leads to the removal/insertion of one Li ion per formula unit. The results of cycling between 3.0 and 4.5 V are shown in the dashed line boxes in Figure 7 and will be discussed first, followed by subsequent discussion of the larger potential window between 2.0 and 4.5 V. The highest capacities in the potential range of 3.0−4.5 V are achieved for samples prepared in mixtures of water and glycols. In particular, the initial capacity in the 3.0−4.5 V range for the sample prepared in a mixture of EG and water showed ∼134 mAh/g, which is as high as any studies have reported for this material.13 The improved performance found when preparing the materials in mixtures of water and glycols correlates well with the smaller particle sizes (Figure 5) obtained. Small particle size has been previously demonstrated as a means to improve electrochemical performance for LiVOPO4 due to its poor ionic conductivity.13,14 The sample synthesized in butanol and water also showed slightly higher capacity in the 3.0−4.5 V range, but much lower still than that obtained with the samples synthesized in the glycols, despite similar particle size. Therefore, it is likely that the glycols have an effect on the product beyond simply particle size control. It should also be noted that the sample synthesized in pure water showed lower capacity in the 3.0−4.5 V range than any of the samples prepared in the mixed solvents despite similar particle size as in the mixtures of alcohols and water. The glycols are very reducing, which could be useful since it is necessary to keep V in the V4+ state. However, the ICP ratios shown in Figure 4 indicate slight lithium deficiency for the samples synthesized in water and glycol mixtures. Lithium deficiency suggests that there may be a small amount of V5+ in the samples to maintain charge balance, but this is clearly inconsistent with the glycols being good reducing agents. Therefore, it is unclear how to rationalize these observations that the samples with slight lithium deficiency exhibited the highest capacities. It should be noted that the increased capacity

Figure 7. First charge−discharge curves at a C/20 rate of α-LiVOPO4 prepared with a V concentration of 0.067 M, Li:V:P = 5:1:5, and a total reaction time of 50 min (∼25 min ramp time to 230 °C and ∼25 min hold time at 230 °C) in various mixed solvents. All samples were prepared at 230 °C except for the sample prepared in water, which was held at 240 °C.

for the sample prepared in a mixture of EG and water does not result from the slight impurity (Figure 4) evident in the XRD pattern for this sample; preparing the sample at 235 °C led to similar capacity and elemental ratios but showed no impurity. The high capacity for the sample prepared in a mixture of EG and water may result from the morphology. The SEM images show loosely agglomerated primary nanoparticles for the sample prepared in an EG and water mixture rather than the more typical flower-like morphology with the densely packed surface features evident for most of the samples. It has been suggested in the literature that a second Li can be inserted into LiVOPO4 to form Li2VOPO4.12,43−50 The second Li ion insertion occurs between 2.0 and 2.5 V and increases the G



Article

solution may be due to the nonstoichiometry demonstrated in Table 5. However, the sample prepared in EG and water showed similar nonstoichiometry but presented high capacity, so it is not clear why the sample prepared with 1 mL of CTAB solution (in a mixture of water and ethanol) performed so poorly. As discussed for the samples shown in Figure 7, high capacity at ∼4 V does not correlate well with high capacity at ∼2 V for the samples shown in Figure 8. The capacity at ∼2 V for the sample prepared with 0.5 mL of CTAB solution was quite limited, despite the high capacity in the ∼4 V region. Despite lower capacity in the 3.0−4.5 V potential range, the sample prepared with 1.0 mL of CTAB solution showed higher capacity in the 2.0−4.5 V range than the sample prepared with 0.5 mL of CTAB. The cause of this observed phenomenon is not clear. A stronger understanding of the process involving intercalation of the second Li ion into LiVOPO4 is needed to determine what factors contribute to high capacities at the ∼2.0 V plateau. Studying the structure upon insertion of a second Li ion is the subject of our future work. The cycling performance for the samples prepared with varied solvents is shown in Figure 9a,b for the 3.0−4.5 V and 2.0−4.5 V potential ranges, respectively. Similarly, Figure 9c,d shows the cycling performance for the samples with varied reaction times and CTAB concentration (all with solvent mixtures of water and ethanol). It is clear that the cycle performance is poor for most samples. Furthermore, the cycle performance does not generally correlate well to high initial capacity. For example, the sample prepared in pure water showed the lowest initial capacity of all the samples in the 3.0− 4.5 V range, but the capacity was more stable than the samples prepared in mixed solvents. Similarly, the sample prepared with a 45 min hold time exhibited good cycle performance but poor initial capacity. The samples with the highest initial capacity in the 3.0−4.5 V potential window (samples prepared in EG and water and sample prepared with 0.5 mL of CTAB in ethanol and water) showed the poorest capacity retention. The samples presenting high initial capacity tended to consist of smaller particles. Thus, we can approximately correlate small particle size with performance degradation. Small particles imply high exposed surface area which suggests that the performance degradation during cycling may be related to side reactions with the electrolyte at the surface of the particles, especially since cycling up to 4.5 V pushes the stability boundary for the electrolyte window. Thus, coating the particles holds promise for improving the cycling performance. Preliminary attempts were made to coat the sample prepared in 3:1 water:ethanol (10 min reaction hold time) with PEDOT:PSS (poly(3,4-ethylenedioxythiophene) poly(styrenesulfonate), Clevios P VP AI 4083 with a resistivity of 500−5000 Ω·cm). This was accomplished by mixing 10 mL of water, 0.1 mg of the α-LiVOPO4 sample, and enough PEDOT:PSS solution to incorporate 5 wt % PEDOT: PSS into the final product. The mixture was stirred on a hot plate at 85 °C until the water evaporated. This process is similar to a previously described PEDOT coating procedure for LiFePO4.30 The cycling performance of the PEDOT:PSS coated sample is compared to the other samples in Figure 9c,d (referred to as the “ex situ PEDOT” sample with green plus sign symbols). The ex situ PEDOT:PSS-coated sample shows greatly improved cyclability in the 3−4.5 V potential range compared to the pristine sample prepared in water and ethanol with a 10 min microwave reaction time before coating (orange diamond

theoretical capacity to 318 mAh/g, though this extra capacity is not generally fully realized. First charge−discharge curves are shown in Figure 7 for the voltage range of 2.0−4.5 V to allow access to the V3+/4+ couple corresponding to the second lithium insertion/extraction. Although there are large differences in capacity in the 3.0−4.5 V range for the samples prepared in various solvent mixtures, the differences are not as large or as systematic when the cells are cycled in the range of 2.0−4.5 V (Figure 7). Furthermore, high capacity in the 3.0−4.5 V range does not necessarily correlate to high capacity in the 2.0−4.5 V range. For example, the sample synthesized in pure water shows the lowest capacity for the V4+/5+ couple at ∼4 V (∼60 mAh/g) but shows the highest capacity for the V3+/4+ couple at ∼2 V (∼134 mAh/g). Similarly, the sample prepared in ethylene glycol and water shows high capacity at ∼4 V (∼134 mAh/g) but shows very low capacity at ∼2 V (∼20 mAh/g). Additionally, the capacities in the 2.0−4.5 V range do not exhibit a correlation to the particle size in contrast to the strong dependence of capacity on particle size in the 3.0−4.5 V range. The electrochemical performance in the 3.0−4.5 V range was also improved by decreasing the reaction time, decreasing the concentration of the reactants, or the addition of 0.5 mL of CTAB solution as shown in Figure 8 (all samples prepared in a

Figure 8. First charge−discharge curves at a C/20 rate of α-LiVOPO4 prepared with Li:V:P = 5:1:5 and water:ethanol = 3:1 for various reported reaction hold times at 230 °C (not including ∼25 min ramp time to 230 °C), V concentrations, and amounts of CTAB solution substituted for water.

mixture of water and ethanol). Generally, smaller particle sizes correlated well with improved electrochemical performance as has been previously demonstrated for microwave-synthesized LiFePO4.30 The performance improvement was very modest for the sample prepared with lower precursor concentration. The sample prepared with 1.0 mL of CTAB solution showed lower capacity than the sample prepared with 0.5 mL of CTAB solution in the 3.0−4.5 V range despite smaller particles. The poor performance of the sample prepared with 1 mL of CTAB H



Article

Figure 10. (a) XRD pattern of the α-LiVOPO4 sample coated ex situ (inset shows pattern after drying under vacuum), (b) SEM image of the α-LiVOPO4 sample coated ex situ, (c) SEM of the α-LiVOPO4 sample coated in situ, (d) first charge−discharge curves at a C/20 rate of the α-LiVOPO4 sample coated ex situ, and (e) first charge− discharge curves at a C/20 rate of the α-LiVOPO4 sample coated in situ. For both samples, α-LiVOPO4 was prepared with Li:V:P = 5:1:5, water:ethanol = 3:1, a 10 min hold time at 230 °C (after an initial ∼25 min ramp time to 230 °C) and coated in situ or ex situ with PEDOT:PSS.

consist of 5 wt % PEDOT: PSS. On the first cycle, this coating attempt (blue triangles in Figure 9c,d) led to lower capacity at the ∼4 V plateau than the corresponding uncoated sample (orange diamond symbols), but higher capacity for the plateau at ∼2 V. XRD revealed no impurities for the in situ coated sample, so it is not shown here. The first charge−discharge curve for the sample coated in situ showed no distinct differences in shape from the uncoated samples and also showed similar morphology (Figure 10c,e), although the particle size was large despite the short reaction time. The “in situ PEDOT” sample also showed very good capacity retention upon cycling in both potential ranges. Overall, coating shows promise for improving the cycling performance for these samples, but the coating procedure needs to be optimized. For the ex situ strategy, a different solvent may improve the results since the α-LiVOPO4 phase appears to be unstable in water under the coating conditions. The in situ coating strategy may be more effective with smaller particles which can be achieved by decreasing reaction time, decreasing concentration of the reactants, changing the solvent mixture, or adding a surfactant like CTAB. Such optimization will be the subject of future work.

Figure 9. Cycle performance of α-LiVOPO4 at a C/20 rate prepared with Li:V:P = 5:1:5: (a) V concentration of 0.067 M and a total reaction time of 50 min (∼25 min ramp time to 230 °C and ∼25 min hold time at 230 °C) in various mixed solvents and cycled at 3.0−4.5 V, (b) same as (a) but cycled at 2.0−4.5 V, (c) water:ethanol = 3:1 for various indicated reaction hold times at 230 °C (not including ∼25 min ramp time to 230 °C), V concentrations, amounts of CTAB solution substituted for water, and coating strategies cycled at 3.0−4.5 V, and (d) same as (c) but cycled at 2.0−4.5 V. All samples were prepared at 230 °C expect for the sample prepared in water, which was held at 240 °C. Note that the legends in (a) and (b) are common to both plots (a) and (b) and that the legends in (c) and (d) are common to both plots (c) and (d).

symbols). However, lower initial capacity is exhibited for the coated sample. XRD (Figure 10a) reveals that the sample is no longer phase pure after the attempted ex situ coating and the particle morphology and charge−discharge curves change as well (Figures 10b,d). However, upon drying the sample further in a vacuum oven at 150 °C before making electrodes, some of the impurity peaks (indicated with black outlined green triangles) disappear, indicating that some of the impurity peaks result from a hydrated phase (inset in Figure 10a). The impurity phase likely explains why there is a change in the shape of the charge−discharge profile with PEDOT:PSS addition (Figure 10d). Attempts were also made to coat the material in situ by adding PEDOT:PSS to the reaction vessel with a 3:1 mixture of water:ethanol (10 min reaction hold time at 230 °C). Again PEDOT:PSS was added such that the final product would

■

CONCLUSIONS Three polymorphs of LiVOPO4 were synthesized with a facile, low-temperature, microwave-assisted solvothermal approach. By varying the ratios of ethanol and water as the solvent as well as the precursor ratios and temperature, the three known polymorphs of LiVOPO4 could be stabilized. Formation of the triclinic phase (α-LiVOPO4) was favored at high temperature and formed the most stoichiometric product, whereas the other I



Article

(16) Gaubicher, J.; Le Mercier, T.; Chabre, Y.; Angenault, J.; Quartona, M. J. Electrochem. Soc. 1999, 146 (12), 4375. (17) Hameed, A. S.; Nagarathinam, M.; Reddy, M. V.; Chowdari, B. V. R.; Vittal, J. J. J. Mater. Chem. 2012, 22, 7206. (18) Kerr, T. A.; Gaubicher, J.; Nazar, L. F. Electrochem. Solid-State Lett. 2000, 3 (10), 460. (19) Kuo, H. T.; Bagkar, N. C.; Liu, R. S.; Shen, C. H.; Shy, D. S.; Xing, X. K.; Lee, J.-F.; Chen, J. M. J. Phys. Chem. B 2008, 112, 11250. (20) Lii, K. H.; Li, C. H.; Cheng, C. Y.; Wang, S. L. J. Solid State Chem. 1991, 95 (2), 352. (21) Nagamine, K.; Honma, T.; Komatsu, T. J. Am. Ceram. Soc. 2008, 91 (12), 3920. (22) Ren, M. M.; Zhou, Z.; Gao, X. P.; Liu, L.; Peng, W. X. J. Phys. Chem. C 2008, 112, 13043. (23) Ren, M. M.; Zhou, Z.; Su, L. W.; Gao, X. P. J. Power Sources 2009, 189, 786. (24) Saravanan, K.; Lee, W. S. L.; Kuezma, M.; Vittal, J. J.; Balaya, P. J. Mater. Chem. 2011, 21, 10042. (25) Wang, L.; Yang, L.; Gong, L.; Jiang, X.; Yuan, K.; Hu, Z. Electrochim. Acta 2011, 56, 6906. (26) Yang, Y.; Fang, H.; Zheng, J.; Li, L.; Li, G.; Yan, G. Solid State Sci. 2008, 10, 1292. (27) Lavrov, A. V.; Nikolaev, V. P.; Sadikov, G. G.; Porai-Koshits, M. A. Sov. Phys. Dokl. 1982, 27, 680. (28) Murugan, A. V.; Muraliganth, T.; Manthiram, A. J. Phys. Chem. C 2008, 112 (37), 14665. (29) Murugan, A. V.; Muraliganth, T.; Manthiram, A. J. Electrochem. Soc. 2009, 152, A79. (30) Murugan, A. V.; Muraliganth, T.; Manthiram, A. Electrochem. Commun. 2008, 10, 903. (31) Murugan, A. V.; Muraliganth, T.; Ferreira, P. J.; Manthiram, A. Inorg. Chem. 2009, 48 (3), 946. (32) Muraliganth, T.; Murugan, A. V.; Manthiram, A. J. Mater. Chem. 2008, 18, 5661. (33) Baghbanzadeh, M.; Carbone, L.; Cozzoli, P. D.; Kappe, C. O. Angew. Chem., Int. Ed. 2011, 50, 11312. (34) Fini, A.; Breccia, A. Pure Appl. Chem. 1999, 71 (4), 573. (35) De la Hoz, A.; Diaz-Ortiz, A.; Moreno, A. Chem. Soc. Rev. 2005, 34, 164. (36) Rietveld, H. M. J. App. Crystallogr. 1969, 2, 65. (37) Rodriguez-Carvaja, J. Physica 1993, 192B, 55. (38) Roisnel, T.; Rodriguez-Carvajal, J. Epdic 7: Eur. Powder Diff. 2001, 378-3, 118. (39) Baran, E. J.; Vassallo, M. B. J. Raman Spectrosc. 1994, 25, 199. (40) Wang, X.; Liu, L.; Jacobson, A. J. J. Am. Chem. Soc. 2002, 124, 7812. (41) Sauvage, F.; Quarez, E.; Tarascon, J.-M.; Baudrin, E. Solid State Sci. 2006, 8, 1215. (42) De, S.; Dey, A.; De, S. K. J. Phys. Chem. Solids 2007, 68, 66. (43) Chernova, N. A.; Roppolo, M.; Dillonb, A. C.; Whittingham, M. S. J. Mater. Chem. 2009, 19, 2526. (44) Davis, L. J. M.; He, K. J.; Bain, A. D.; Goward, G. R. Solid State Nucl. Magn. Reson. 2012, 42, 26. (45) Ren, M. M.; Zhou, Z.; Gao, X. P. J. Appl. Electrochem. 2010, 40, 209. (46) Song, Y.; Zavalij, P. Y.; Whittingham, M. S. J. Electrochem. Soc. 2005, 152 (4), A721. (47) Whittingham, M. S. Chem. Rev. 2004, 104, 4271. (48) Whittingham, M. S. Mater. Res. Soc. Bull. 2008, 33, 411. (49) Whittingham, M. S.; Song, Y.; Lutta, S.; Zavalij, P. Y.; Chernova, N. A. J. Mater. Chem. 2005, 15, 3362. (50) Pozas, R.; Maduefio, S.; Bruque, S.; Moreno-Real, L.; MartinezLara, M.; Criado, C.; Ramos-Barrado, J. Solid State Ionics 1992, 51, 79.

two polymorphs showed nonstoichiometric elemental ratios and atypical XRD peak intensity ratios. The effects of synthesis conditions on the particle size/ morphology and the electrochemical performance of αLiVOPO4 were studied in detail. The ratio between water and alcohol as the solvent was found to be critical for the formation of a phase-pure sample. Several different alcohols and glycols could be used to synthesize this phase, with the smallest particles and best electrochemical performance resulting from synthesis in mixtures of water and glycols. The reaction time and concentration were also varied to show that the particle size could be controlled by these factors. Higher initial capacity was generally found to be linked to reduced particle size when the cells were cycled in the typical potential window for LiVOPO4. To further control the particle size, CTAB was added to the precursor solution, which helped to prevent particle growth and agglomeration, reduced particle size, and improved electrochemical performance. The cycling performance was improved by coating the particles with PEDOT:PSS.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: (512) 471-1791. Fax: 512-471-7681. E-mail: manth@ austin.utexas.edu. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Office of Vehicle Technologies of the U.S. Department of Energy under Contract No. DEAC02-05CH11231. The authors would also like to acknowledge the National Science Foundation for funding the FEI Quanta 650 scanning electron microscope used in this work.

■

REFERENCES

(1) Manthiram, A.; Goodenough, J. B. J. Solid State Chem. 1987, 71, 349. (2) Manthiram, A.; Goodenough, J. B. J. Power Sources 1989, 26, 403. (3) Padhi, A. K.; Nanjundasawamy, K. S.; Goodenough, J. B. J. Electrochem. Soc. 1997, 144 (4), 1188. (4) Azmi, B. M.; Ishihara, T.; Nishiguchi, H.; Takita, Y. J. Power Sources 2003, 119, 273. (5) Azmi, B. M.; Ishihara, T.; Nishiguchi, H.; Takita, Y. Electrochim. Acta 2002, 48, 165. (6) Dupre, N.; Gaubicher, J.; Le Mercier, T.; Wallez, G.; Angenault, J.; Quarton, M. Solid State Ionics 2001, 140, 209. (7) Dupre, N.; Gaubicher, J.; Angenault, J.; Wallez, G.; Quarton, M. J. Power Sources 2001, 97, 532. (8) Girgsdies, F.; Schneider, M.; Bruckner, A. Solid State Sci. 2009, 11, 1258. (9) Lim, S. C.; Vaughey, J. T.; Harrison, W. T. A.; Dussack, L. L.; Jacobson, A. J.; Johnson, J. W. Solid State Ionics 1996, 84, 219. (10) Dupre, N.; Wallez, G.; Gaubicher, J.; Quarton, M. J. Solid State Chem. 2004, 177, 2896. (11) Dupre, N.; Gaubicher, J.; Angenault, J.; Quarton, M. J. Solid State Eletrochem. 2004, 8, 322. (12) Allen, C. J.; Jia, Q.; Chinnasamy, C. N.; Mukerjee, S.; Abraham, K. M. J. Electrochem. Soc. 2011, 158 (12), A1250. (13) Li-Zhi, X.; Ze-Qiang, H. Acta Phys. Chim. March Sin. 2010, 26 (3), 573. (14) Azmi, B. M.; Ishihara, T.; Nishiguchi, H.; Takita, Y. J. Power Sources 2005, 146, 525. (15) Barker, J.; Saidi, M. Y.; Swoyer, J. L. J. Electrochem. Soc. 2004, 151 (6), A796. J


Microwave-Assisted Solvothermal Synthesis and Characterization of

Recommend Documents