High redox potential transition metals incorporated olivine structure: LiFe0.7(Mn1/3Co1/3Ni1/3)3x−yVyPO4 (x = 0.1, 0.0 ≤ y ≤ 0.10) cathode for Li-ion battery

A series of LiFe0.7 (Mn1/3Co1/3Ni1/3)3x−yVyPO4 (x = 0.1, y = 0, 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.08, and 0.10) compositions doped with high concentration (30 mol%) of transition metals having higher redox potential are synthesized by sol-gel process. In the composition, assumption is made that equal contents of Mn, Co, and Ni are replaced simultaneously by vanadium (V), and changes in structural and electrochemical characteristics are systematically investigated. The X-ray diffraction analysis confirms that for y > 0.04, a secondary phase Li3V2(PO4)3 formation takes place. The Rietveld refinement performed on XRD data shows continuous change in lattice parameters and cell volume with increasing y. X-ray photoelectron spectroscopy study confirms the oxidation state of Fe, Mn, Co, and Ni in + 2, whereas V in + 4 state. The electrochemical characteristics show the positive contribution of Li3V2(PO4)3 in capacity as well as cycle life among doped samples.


Introduction
Among the various cathode materials, olivine LiMPO 4 (M = Mn, Fe, Co, Ni) is considered a potential candidate for Li-ion batteries.Particularly, LiFePO 4 has attracted the most attention due to its low cost, nontoxicity, thermal stability, and relatively high theoretical capacity (170 mAh g −1 ).However, LiFePO 4 suffers from poor electronic conductivity (~ 10 −9 S cm −1 ) and lower lithium ion diffusivity (< 10 −14 cm 2 s −1 ), and relatively low redox potential of Fe +2 /Fe +3 with respect to Li/Li + (3.45 V).Several methods such as coating with conducting layers, improving kinetics by reducing defects, and doping with other elements have been applied to overcome the limitations of LiFePO 4 , [1,2].
The preheated sample was further ground in mortar using pestle and calcined at 700 °C for 5 h in Ar gas atmosphere to get crystalline powder.

Characterization
Phase identification of prepared samples was done by X-ray diffractometer (XRD) (Smart Lab, Rigaku, Japan) using CuK α radiation (λ = 1.54 Å).XRD data were collected at a scanning rate of 0.5°/min with a step size of 0.02° in the range of 2θ = 15-80°.The Rietveld refinements were performed to determine lattice parameters using FullProf Suite with pseudo-Voigt profile function of Thompson et al. [35].For the Rietveld refinements, initial cell parameters and atomic positions were taken from the reported work [36].Morphology and particle size of the samples were determined by field emission scanning electron microscope (FESEM) (FEI Quanta 200 FEG, Netherland).The morphology and elemental mapping of doped LFP samples were determined by using transmission electron microscope (TEM) (JEOL FE, 3200FS HRTEM, Japan) with an acceleration voltage of 300 kV.The surface analysis for chemical composition and oxidation states of prepared samples was done by X-ray photoelectron spectrometer (PHI 5000 Versa Probe III, Japan) using Al-K α (hν = 1486.6eV) X-ray source.MULTIPAK software was used for peak analysis and de-convolution of the peak patterns, and the correction of binding energies was done with reference to C 1s peak at 284.8 eV.The charge-transfer resistance and diffusion coefficient of Li-ions of freshly prepared electrodes were determined using electrochemical impedance spectroscopy (EIS) (Gamry Instruments, INTERFACE 1000) over the frequency range of 0.1 Hz-1 MHz and at voltage amplitude of ± 10 mV.

Electrode preparation and cell fabrication
The slurry for electrode was prepared by mixing active material, carbon black, and PVDF binder in the weight ratio of 80:10:10 in N-methyl-2 pyrrolidone (NMP) solvent.The prepared slurry was pasted onto aluminum foil and dried in air (80 °C for 12 h) and vacuum (120 °C for 8 h).For electrode preparation, electrodes in circular shape of diameter 10 mm were punched from the prepared electrode sheet.The electrolyte 1M LiPF 6 + EC/EMC (50:50 by volume) (Sigma-Aldrich, Germany) was used, whereas separator was made of polypropylene for cell fabrication in Teflon cell of cylindrical geometry.
Arbin Cycler (BT 2000, Arbin Instruments, USA) was used for determining galvanostatic charging/discharging cycles in the voltage range of 2-4 V and 2-4.5 V at different current (C) rates (assuming 1C = 170 mA g −1 ).Cell fabrication was carried out inside an Ar gas-filled glove box (MBRAUN, MB 200G Unilab II, Germany) which has oxygen and moisture levels less than 0.1 ppm.

Results and discussion
XRD patterns of all prepared samples are shown in Fig. 1a.The pattern of undoped sample (LFP) shows single olivine phase of LiFePO 4 without secondary phase(s) as confirmed with ICSD No. 72545.Sample LiFe 0.70 Mn 0.10 Co 0.10 Ni 0.10 PO 4 doped with equal 10 mol% of each element (Mn, Co, Ni) having high redox potential also shows similar XRD pattern as that of pristine LFP.It means Mn, Co, and Ni have successfully been able to replace Fe atoms in the host structure of LiFePO 4 .In addition, XRD patterns of V doped samples according to composition LiFe 0.7 (Mn 1/3 Co 1/3 Ni 1/3 ) 3x−y V y PO 4 for (0.01 ≤ y ≤ 0.04) show pure phase without observable impurity peaks.However, for 0.05 ≤ y ≤ 0.10 composition, a small peak of Li 3 V 2 (PO 4 ) 3 secondary phase is observed as shown by arrow (↓) in the magnified view of XRD patterns in Fig. 1b.
The Li 3 V 2 (PO 4 ) 3 is a common secondary phase in LiFePO 4 with excess V doping.However, Li 3 V 2 (PO 4 ) 3 is an electrochemically active phase and also a cathode material for Li-ion battery.Figure 1c shows the XRD patterns over a smaller range of 2θ to demonstrate the clear peaks shift to higher angles with increasing y in LiFe 0.7 (Mn 1/3 Co 1/3 Ni 1/3 ) 3x−y V y PO 4 composition.
A full Rietveld refinement was carried out on all XRD patterns to determine the lattice parameters and cell volumes.The refinement graphs of four samples LFP, LFP-0V, LFP-4V, and LFP-10V are shown in Fig. 2. The refinement results demonstrate that lattice parameters a and b, and cell volume (V) decrease while lattice parameter c increases with increasing concentration of V doping as shown in Fig. 3.The decrease in the unit cell volume with increasing V-doping concentration is due to formation of cation vacancies to balance the higher charge (V +3 or V +4 ) of vanadium; similar results are also reported earlier [28,37,38].Furthermore, the reduction in unit cell volume of doped samples is also due to smaller atomic radius of doping elements V +4 (0.72 Å), Mn +2 (0.70 Å), Co +2 (0.75 Å), and Ni +2 (0.69 Å) as compared to Fe +2 (0.78 Å) [1,39].In addition, the change in the lattice parameters and continuous shift in XRD peaks with increasing V-doping concentration also confirm the incorporation of V into host lattice [37,38].
Along with the lattice parameters, the site occupancies of transitional metals were also refined to check the charge compensation during V doping.The Li occupancy was fixed at 100%, and refinements was carried out on Fe, Mn, Co, Ni, and V sites only.Initially, occupancies of Fe, Mn, Co, Ni, and V were fixed according to estimated formula LiFe 0.7 (Mn 1/3 Co 1/3 Ni 1/3 ) 3x−y V y PO 4 (x = 0.1, y = 0, 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.08, and 0.10), in that case, slightly higher values of R F and  2 were found.When the occupancies of the transition metals are allowed to refine one by one without constraint, the values of R F and χ 2 were found lower and better fitting was observed.
However, only minor deviation was observed in the estimated occupancies and refined occupancies for all transition  metals (Fe, Mn, Co, Ni, and V) as shown in Table 1.An attempt has also been made to refine all the transition metal sites simultaneously, but in that case, convergence was not reached.However, due to poor resolution between the vacancies and Li-ions in the XRD data, the vacancy formation at both Fe and Li sites cannot be totally excluded.The samples LFP-8V and LFP-10V have shown higher χ 2 and R F values due to higher intensity peaks of Li 3 V 2 (PO 4 ) 3 secondary phase.Furthermore, we have performed two-phase refinement to determine the wt.% of Li 3 V 2 (PO 4 ) 3 in the samples LFP-8V and LFP-10V, and refinement profile of LFP-10V is shown in Fig. 2d.For the best fitted profile, wt.% of Li 3 V 2 (PO 4 ) 3 was found 4.28% and 5.81% for samples LFP-8V and LFP-10V, respectively.However, very small variation was observed in the occupancies of vanadium in both samples.The oxidation states of the transition metals Fe, Mn, Co, Ni, and V are determined by XPS, and results are shown in Fig. 4. The XPS spectra of Mn, Co, and Ni are recorded from the two samples LFP-0V and LFP-10V.
XPS spectra of Fe are recorded for LFP, LFP-0V, and LFP-10V.The XPS spectra of Fe 2p, Mn 2p, Co 2p, and Ni 2p core level show two peaks of 2p 3/2 and 2p 1/2 due to spin-orbit coupling, and for each element, satellite peaks appear along with main peaks.The main peaks of Fe centered at about 709.7 eV and 723.3 eV correspond to 2p 3/2 and 2p 1/2 , respectively, while satellite peaks are observed at about 714.0 eV and 727.6 eV.The observed peaks are well matched with Fe +2 oxidation state of Fe [40].Similarly, Mn spectra show two main peaks at about 641.8 eV (2p 3/2 ) and 653.3 eV (2p 1/2 ) corresponding to Mn +2 and two satellite peaks at 646.4 eV and 658.3 eV [41][42][43].The spectra peaks of Co are observed at about 781.7 eV (2p 3/2 ) and 797.9 eV (2p 1/2 ) with two satellite peaks at about 787.5 eV and 802.4 eV corresponding to Co +2 [44,45].The XPS spectra of Ni show main peaks at about 855.0 eV (2p 3/2 ), 872.4 eV (2p 1/2 ), and one satellite peak at about 860.3 eV which match well with Ni +2 oxidation state of Ni [40,46].Therefore, it is found from the XPS results; there is no change in the oxidation states of Fe, Mn, Co, and Ni.At the end, the oxidation states of V in the composition LiFe 0.70 (Mn x Co x Ni x ) 3x−y V y PO 4 were also studied.The XPS spectra of three samples LFP-1V, LFP-4V, and LFP-10V were recorded to examine the effect of doping concentration on oxidation state of V.The spectra of V 2p core level show two main peaks at 516.6 eV and 523.9 eV corresponding to 2p 3/2 and 2p 1/2 spin-orbit coupling, respectively.The binding energy position of V matches well with that appeared in VO 2 [47,48], suggesting the oxidation state + 4 of V in doped samples which is independent of doping concentration.However, in the composition LiFe 0.70 (Mn x Co x Ni x ) 3x−y V y PO 4 for 0.05 ≤ y ≤ 0.10, a small trace of secondary phase Li 3 V 2 (PO 4 ) 3 is observed in the XRD pattern.But the oxidation state V +3 for Li 3 V 2 (PO 4 ) 3 is not observed in the present study.Moreover, distinguishing between V +3 and V +4 peaks is difficult by XPS due to its surface sensitivity as data is collected from surface, charge compensation used during data collection for non-conducting samples like LiFePO 4 , and calibration of the data to C or O peak [26,47].In addition, the binding energy difference between V +3 and V +4 peaks are very small (0.1-0.5 eV), and comparatively large variations in the binding energy (~ 1.9 eV) of the peak V 2p 3/2 for Li 3 V 2 (PO 4 ) 3 are reported [47,[49][50][51].Morphology and particle size analysis were performed by FESEM, and results are shown in Fig. 5a-i, whereas histograms of particle size distribution are shown in Fig. 5j-r.There is no direct relation found between morphology and doping concentration from the FESEM images.Moreover, samples show morphology variation from circular, elongated to mixed shapes.However, one can notice from the micrographs and histograms that average particle size increases for V-doped samples as compared to without V-doped sample.The average particle size of LFP-0V was found 175.51 ± 5.1 nm as shown in Fig. 5a, j, whereas the particle size of LFP-8P was found maximum 1068.0 ± 77 nm.The results suggest that the multi-element doping can affect particle morphology including the particle size.The previous studies also have reported the increased particle size of LFP on Mn, Co, Ni, and V doping [23,38].
The morphology and particle size were further confirmed by TEM analysis of samples LFP-0V and LFP-4V, and results are shown in Fig. 6.The TEM-EDX elemental mapping of Fe, Mn, Ni, and V was also determined to show distribution of elements in the doped samples.The LFP-4V samples show bigger particle size as compared to LFP-0V which supports the FESEM results.The lattice fringes of samples LFP-0V and LFP-4V in high resolution mode show lattice spacing of 3.34 Å and 2.12 Å corresponding to (111) and (112) planes, respectively, as shown in Fig. 5b, h, and SAED patterns are shown in inset of respective image.From the elemental mapping results, both samples show homogeneous distribution of Fe, Mn, Ni, and Co elements.However, V mapping was performed only on LFP-4V sample.
The electrochemical analysis is performed to understand the redox couples and presence of multi-phase in the multi-doped LiFePO 4 .The charge-discharge characteristics of all samples are shown in Fig. 7 in two different potential ranges 2V-4V and 2V-4.5 V at current rate of 0.1C.The samples chargeddischarged in the voltage range 2V-4V show only one redox couple at about 3.45V/3.48Vfor Fe +2 /Fe +3 vs. Li.The electrochemical study of undoped pristine LFP has already been reported in our previous papers [32,33]; therefore, in this work, we have mainly focused on electrochemical characteristics of doped LFP.The initial capacity was found 112 mAh g −1 for sample LFP-0V sample which is much lower than the reported capacity 150 mAh g −1 of LFP [32].Since redox couples of Mn (4.1V), Co (4.8V), and Ni (5.1V) are higher than the experimental voltage range 2V-4V, these elements behave as inactive materials in electrochemical reaction.So, the capacity contribution would be only from Fe, i.e., 70% (105 mAh g −1 ) of capacity reported for pristine LFP (150 mAh g −1 ); however, sample shows higher capacity 112 mAh g −1 (which is about 75% of reported LFP capacity) than predicted (105 mAh g −1 ).Therefore, the higher capacity can be attributed to the enhancement in the electronic and ionic conductivity of LFP due to multidoping as also reported earlier [22][23][24]52].A fall in the capacity was noticed on low concentration V doping and showed lowest capacity about 78 mAh g −1 for samples LFP-1V and LFP-2V.However, improvement in the capacity was found with increasing V doping concentration and reached maximum value of about 105 mAh g −1 (70% of LFP) for sample LFP-10V.The improvement in the capacity with higher V concentration may relate to the V doping weaken the Li-O bond and enhanced the Li-ion diffusion [2,53].
Charge-discharge characteristics were further observed for doped samples in larger potential window 2V-4.5V at current rate of 0.1C.The capacity increment in the sample LFP-0V is recorded about 11 mAh g −1 above 4V due to contribution of Mn which becomes redox active at 4.1V [22].The capacity drops on low concentration of V doping are also observed here as found in voltage range 2V-4V.However, apart from the main plateau at about 3.45V of LiFePO 4 , a second plateau is also observed at about 4.1V of Mn +2 in sample LFP-0V and V-doped samples for more than 4 mol% (0.05 ≤ y ≤ 0.10).The second plateau with higher V-doping concertation is due to the presence of secondary phase Li 3 V 2 (PO 4 ) 3 for 0.05 ≤ y ≤ 0.10 (as shown in the XRD pattern) which shows redox activity at 4.1V for second Li + -ion extraction (V +3 /V +4 ) [54].The Li 3 V 2 (PO 4 ) 3 phase also shows redox activity at about 3.5V and 3.8V for first Li + -ion extraction (V +3 /V +4 ) [54], which is not observed in the present study, and this may be due to overlapping with LiFePO 4 curves.However, it is difficult to separate the capacity contribution of Mn and V in V-doped samples, because both show plateaus at 4.1V.The improvement in the capacity is reached maximum for samples LFP-8V and LFP-10V which shows capacity higher than that of LFP-0V.The sample LFP-10V shows maximum charge and discharge capacities of 133 mAh g −1 and 129 mAh g −1 , respectively, in which charge and discharge capacity contribution above 4V is about 13 mAh g −1 and 9 mAh g −1 , respectively, as shown by dotted square in Fig. 7.
Furthermore, it can be noticed in Fig. 7a, b that the charge capacities of the samples LFP-0V, LFP-8Vm and LFP-10V are not in the same order at 4V.However, other samples follow the same order of charge capacities in both cases.The higher capacities of samples LFP-10V and LFP-8V as compared to LFP-0V in Fig. 7b at 4V could be possible due to the presence of secondary phase Li 3 V 2 (PO 4 ) 3 or due to unknown factors.According to hypothesis, it is true that contribution of secondary phase Li 3 V 2 (PO 4 ) 3 should be similar at 4V; however, during the experimentation, influence of other factors such as concentration polarization due to limited diffusion of active species to and from the electrode surface for replacing the reacted material to sustain the reactions of cathode cannot be neglected.To confirm the actual reason of capacity difference, in situ XRD analysis of all samples needs to be done at various stages of chargedischarge cycles.
The CV characteristics of all samples are shown in Fig. 8a, b in voltage range 2V-4V and 2V-4.5V,respectively.The all-doped samples show broad redox peaks.The CV curves recorded in the 2V-4.5Vpotential window show also a secondary redox peak at 4.1V which is due to Mn +2 / Mn +3 or/and V +3 /V +4 as described above.
The cycle life of all samples is shown in Fig. 8c.The sample LFP-0V shows highest capacity retention of 94% after completion of 300 cycles at 0.5C rate.The lowest capacity retention of 59% was observed for sample LFP-1V which improved with increasing V-doping concentration and reached 91% for sample LFP-10V.Therefore, based on the electrochemical results, one can infer that high concentration of multi-element doping in LiFePO 4 is not feasible with presently used carbonate-based electrolytes.At the high redox potential also, these doping elements (Co and Ni) are redox active; however, the available carbonate-based LiPF 6 electrolyte is decomposed at that high potential [7].Therefore, a new electrolyte is required for doped LFP to observe complete electrochemical characteristics.
The contribution of multi-element doping in electrochemical reaction was evaluated by EIS study through ).A Nyquist plot exhibiting a semicircle in the high-frequency region and an inclined line in the low-frequency range represents charge-transfer resistance at the electrode/electrolyte interface and the diffusion kinetics of lithium ions in the bulk electrode materials, respectively [55].The corresponding equivalent circuit is shown in the inset of Fig. 9a which is composed of series resistance (R s ) of cell, a constant phase element (CPE), a charge-transfer resistance (R ct ), and a Warburg impedance (Z w ) [2].The Warburg impedance slope of the inclined line is used to determine D Li + .The coefficient D Li + is calculated from the following equation [56]: where R, T, A, n, F, and C represent the gas constant (8.314J mol −1 K −1 ), absolute temperature (300 K), area of electrode flat surface, number of electrons per molecule during oxidation (for LiFePO 4 , n = 1), Faraday constant (96485 Coulomb mol −1 ), and Li-ion concentration in LiFePO 4 , respectively.
The Warburg coefficient (σ ω ) is determined by the following equation [57]: The σ ω is calculated from the slope of Z real vs.ω −1/2 plot in the low-frequency region as shown in Fig. 9b, where ω is the angular frequency of the applied alternating current.
The charge-transfer resistance of pristine LFP was found lowest 158.7 Ω with highest D Li + (1.72 × 10 −12 cm 2 s −1 ) which may be due to lower particle size facilitating easy charge transfer.The values of R ct and D Li + are found 455.8 Ω and 1.47 × 10 −12 cm 2 s −1 , respectively, for sample LVP-0V.The R ct value is higher while D Li + is lower of LFP-0V as compared to reported values of pristine LFP [32].This could be due to high concentration of inactive elements (Co and Ni) and very low electronic (1) (2) and ionic conductivity of olivine phase LiMnPO 4 as compared to LiFePO 4 [58,59].The R ct further increases with low V-doping concentration and is found to be highest at 470.9 Ω for LFP-1V while D Li + found to be 9.97 × 10 −13 cm 2 s −1 .The R ct decreases and D Li + increases with increasing V-doping concentration due to improved kinetics of LiFePO 4 by V doping [2,53].The R ct and D Li + of 4 mol% V-doped sample (LFP-4V) shows 303.8 Ω and 1.37 × 10 −12 cm 2 s −1 , respectively.The kinetics of the cell further increases with higher doping level and R ct (280.1 Ω) and D Li + (1.57× 10 −12 cm 2 s −1 ) are found for LFP-10V.This may be due to contribution of electrochemically active secondary phase Li 3 V 2 (PO 4 ) 3 because D Li + of Li 3 V 2 (PO 4 ) 3 phase was reported 10 −9 to 10 −10 cm 2 s −1 which is higher than D Li + of LiFePO 4 [60,61].In Fig. 9c, d, we have compared the variation of discharge capacity and D Li + of all samples with V concentration in LiFe 0.7 (Mn 1/3 Co 1/3 Ni 1/3 ) 3x−y V y PO 4 (x = 0.1) compounds.The plot c clearly shows minimum of capacity at vanadium (V) = 0.01 which increases further with higher concentration of V doping.The similar trend is observed for D Li+ except for the sample LFP-5V showing a dip in the curve, which is perhaps due to experimental error or unknown factors.Therefore, one can conclude that improvement in the capacity with higher V concentration is due to weakening of Li-O bond in olivine structure which also enhances the Liion diffusion in the structure.However, reason of reduction in capacity and D Li+ at lower V concentration (0.01) is still not fully clear.
We have also performed charge-discharge cycling of LFP-2V and LFP-4V electrodes at high temperature of 60 °C and post-mortem analysis of electrodes by XRD and SEM for which results are shown in Fig. 10.The chargedischarge cycles are recorded at 0.1C (1C = 170 mA g −1 ) in the potential range of 2-4.5V as shown in Fig. 10a, b.The charge-discharge cycling was done in Swageloktype cells, and heating arrangement was done on heating plate inside the glove box.So, evaporation of electrolyte and inhomogeneous heating cannot be ruled out during measurement.
Both cells show higher polarization at high temperature as compared to room temperature due to high reactivity of electrolyte with electrodes.Furthermore, there are significant differences in charge and discharge capacities of both samples.It is possible due to electrolyte degradation/decomposition at high temperature.During the charge, Li + ions are released from cathode and deposited on Li metal (used as anode) which again inserted into cathode during discharge.Higher charge capacity means additional Li + ions are deposited on Li metal which can only be possible from electrolyte decomposition.However, a small increment in the discharge capacities is also noticed at high temperature as compared to room temperature in both the cells.The charge-discharge operation of the cells is found to be terminated after running of a few cycles.After charge-discharge cycles, cells were dismantled inside the glove box, and we noticed that cells were completely dried, and colour of Li-metal is also turned black.It means, at high temperature, electrolyte evaporates very fast and reacts not only with electrolyte but also with cathode.
To further explore the origin of excess capacity, weather from electrolyte or structural change of LFP during hightemperature charge-discharge, XRD analysis was carried out with recovered electrode materials.As shown in Fig. 10c, all XRD peaks are found matching with reference pattern of LFP (ICSD No. 72545), and no additional peak is noticed.This signifies that the possible generation of excess capacity is from decomposition of electrolyte.However, a broad amorphous hump can be seen in between 20 and 25° which is due to carbon black and PVDF used in electrode preparation [34,62].Intensity of XRD peaks is low due to very small quantity of recovered electrode material.
We have further observed the morphology of LFP-2V electrode before and after (inset of image) charge-discharge cycling as shown in Fig. 10d.Before charge-discharge, we can see LFP (bigger size) and carbon black (smaller size) particles clearly in the electrode.However, after cycling, a gel-type layer can be seen over the LFP and carbon black particles which is due to reactivity of electrolyte and electrode materials.

Conclusion
In the present study, a series of multi-element-doped LiFePO 4 samples were synthesized according to the composition LiFe 0.7 (Mn 1/3 Co 1/3 Ni 1/3 ) 3x−y V y PO 4 (x = 0.1, y = 0, 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.08, and 0.10).In the composition, iron (Fe) concentration was kept fixed at 70 mol%, and assumption was made that equal amount of Mn, Co, and Ni elements is replaced simultaneously by V doping.The effect of simultaneous doping of V, Mn, Co, and Ni on the performance of LiFePO 4 has systematically been investigated with X-ray diffraction, X-ray photoelectron spectroscopy, electrochemical measurements, and electrochemical impedance spectroscopy.The XRD analysis confirms that 4 mol% V can be doped fully into the Fe sites, whereas partial incorporation of V was observed beyond 4 mol% which results in secondary-phase Li 3 V 2 (PO 4 ) 3 formation.A linear change in lattice parameters and cell volume was observed by Rietveld refinement on doping.The oxidation states of Fe, Mn, Co, and Ni were found in + 2 and V found in + 4 as confirmed by XPS.The capacity of the doped samples was found lower in the voltage range 2-4 V due to high content of inactive dopant elements.However, the increased capacity was found when cells were charged/discharged in the larger voltage window 2-4.5 V due to additional electrochemical activity of Mn and V in the doped compounds.

Fig. 1 aFig. 2
Fig. 1 a XRD patterns of all prepared samples, b magnified view of samples showing Li 3 V 2 (PO 4 ) 3 impurity peaks shown by arrow (↓) with high V doping, and c XRD peak shifting due to doping

Fig. 3
Fig. 3 Variation in lattice parameters and cell volume of pristine and doped LiFePO 4 sample as a function of vanadium doping content

Fig. 6
Fig. 6 TEM analysis: a, b TEM micrograph and high resolution fringes of LPV-0V, c, d, e, f TEM-EDX elemental mapping of LFP-0V, g, h TEM micrograph and high-resolution fringes of LPV-4V, and i, j, k, l, m TEM-EDX elemental mapping of LFP-4V ◂

Fig. 9
Fig. 9 EIS characteristics of freshly prepared cells: a Nyquist plots and b Z real vs.ω −1/2 plots for calculating Warburg factor of all samples prepared in the study, c capacity vs. V concentration plots, and d D Li+ vs. V concentration plot

Table 1
Prepared samples according stochiometric ratios, occupancies of transition metals obtained by Rietveld refinement, and abbreviation used for samples Sample according to LiFe 0.7 (Mn 1/3 Co 1/3 Ni 1/3 ) 3x−y V y PO 4