Subscriber access provided by HARVARD UNIVERSITY
Article
General Observation of Fe3+/Fe2+ Redox Couple Close to 4 V in Partially Substituted Li2FeP2O7 Pyrophosphate Solid-Solution Cathodes Tian Ye, Prabeer Barpanda, Shin-ichi Nishimura, Naoya Furuta, Sai Cheong Chung, and Atsuo Yamada Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/cm401547z • Publication Date (Web): 28 Aug 2013 Downloaded from http://pubs.acs.org on September 4, 2013
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Chemistry of Materials is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 17
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Chemistry of Materials
Table of Content Figure The general observation of unusually high Fe3+/Fe2+ redox potential close to 4.0 V vs. Li/Li+ in mixed-metal Li2MxFe1–xP2O7 (M = Mn, Co, Mg) phases is reported with stabilized edge sharing polyhedra. Such a high voltage Fe3+/Fe2+ operation over 3.5 V has long been believed to be possible only by existence of much more electronegative but hygroscopic anions as SO43- or F-.
ACS Paragon Plus Environment
Chemistry of Materials
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
General Observation of Fe3+/Fe2+ Redox Couple Close to 4 V in Partially Substituted Li2FeP2O7 Pyrophosphate Solid-Solution Cathodes
Tian Ye, Prabeer Barpanda, Shin-ichi Nishimura, Naoya Furuta†, Sai-Cheong Chung, and Atsuo Yamada* †
Department of Chemical System Engineering, School of Engineering, The University of Tokyo, 7-3-1 Hongo, Bunkyo-Ku, Tokyo 113-8656 , Japan
*
[email protected] Phone: +81-3-5841-7295 Fax: +81-3-5841-7488
Keywords: Lithium-Ion Battery, Pyrophosphate, Redox Potential Tunability, Structural Stabilization
ACS Paragon Plus Environment
Page 2 of 17
Page 3 of 17
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Chemistry of Materials
Abstract Exploring the newly unveiled Li2MP2O7 pyrophosphate cathode materials for lithiumion batteries, the current study reports the general observation of unusually high Fe3+/Fe2+ redox potential close to 4.0 V vs. Li/Li+ in mixed-metal Li2MxFe1–xP2O7 (M = Mn, Co, Mg) phases with a monoclinic structure (space group P21/c). Such a high voltage Fe3+/Fe2+ operation over 3.5 V has long been believed to be possible only by existence of much more electronegative but hygroscopic anions as SO43- or F-. Thereby, this is the first universal confirmation of >3.5 V operation by stable simple phosphate material. High voltage (close to 4V) operation of Fe3+/Fe2+ couple was stabilized by all dopants, either by larger Mn2+ or smaller Co2+ and Mg2+ ions, where Mg2+ is redox inactive, revealing the high voltage is induced neither by reduced Fe-O bond covalency nor by contamination by the redox couple of other transition metals. The cause of higher Fe3+/Fe2+ redox potential is argued and rooted in the stabilized edge-sharing local structural arrangement and the associated larger Gibbs free energy in charged state.
ACS Paragon Plus Environment
Chemistry of Materials
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
1. Introduction Since 1962, alkali metal pyrophosphates have been extensively studied, but mostly limited to mono-alkali (AMP2O7) type materials.[1-5] In 2008, Adam et al. [6] reported the synthesis of bi-alkali Li2MnP2O7 pyrophosphate. Realizing the possibility of electroactive A2MP2O7 class of cathode, we introduced Li2FeP2O7 as the first bi-alkali pyrophosphate cathode capable of delivering a reversible capacity of 100 mAh/g with 3.5 V (Fe3+/Fe+2 vs. Li/Li+) redox potential.[7] Following, electrochemically active Li2CoP2O7 as a 4.9 V (Co3+/Co+2 vs. Li+/Li) cathode was reported. [8] With a robust three-dimensional (P2O7)-4-framework, Li2FeP2O7 offers excellent thermal/ chemical stability, easy synthesis, highest redox potential of 3.5 V among PO4-based materials and a two-dimensional channel enabling efficient Li-diffusion along with possibility of 2-electron redox reaction. It inspired us to investigate the underlying electrochemical reaction mechanism in Li2FeP2O7 pyrophosphate,[9] isostructural electroactive Li2MnP2O7 [10] as well as their solid-solution compounds Li2MnyFe1-yP2O7 for battery usage.[11,12] In all cases, we could obtain complete Fe3+/Fe2+ redox activity during Li (de)insetion reaction. Nevertheless, the presence of Mn significantly modified the Fe3+/Fe2+ redox couple by ‘stabilizing’ and gradual ‘upshifting’ of the initial two redox peaks. Resulting, without any aid of more electronegative ions such as SO43- or F-, we observed partial Fe-redox activity at a very high redox potential of 3.9 V, [11] which equals to the highest ever-reported redox potential associated with any Fe-based compounds.[13] It attests the rich structural and electrochemical tunability of the pyrophosphate systems. In order to generalize this issue, we investigated the novel Fe–Co and Fe–Mg binary solid-solutions. Here, we present the solid-state synthesis, structural and electrochemical properties of novel Li2MxFe1–xP2O7 (M = Co, Mg) family of compounds. Although gradual Co or Mg substitution into Fe sites reduces the lattice dimension and thereby increases the covalency of M-O bonds, which is opposite to Mn substitution,[11] we observed the higher Fe3+/Fe2+ redox couple at around 3.9 V in both cases. Here, we explain the common Fe-redox tunability in Li2MxFe1–xP2O7 (M = Co, Mg, Mn) phases taking into account the stabilized atomic arrangement and higher Gibbs free energy in the charged state associated with the edge-shared geometry. Contrary to popular argument on inductive effect, the thermodynamic principle has more significant effect on tunability of metal redox potential in the present system. This insight can be a key in designing high-voltage Fe-based cathodes materials. 2. Experimental Solid-state Synthesis: The Li2CoxFe1-xP2O7 binary compounds were produced by conventional solid-state method by using stoichiometric mixture of Li2CO3 (Wako, 99%), FeC2O4.2H2O (Junsei, 99+%), CoC2O4 (Kojundo, 99+%) and (NH4)2HPO4 (Wako, 99%). For Li2MgxFe1–xP2O7, stoichiometric amounts of Mg(OH)2 (Wako, 99.9%) precursor was used. Precursors were homogenously mixed by wet planetary milling in acetone media for 3 h (600 rpm) employing Cr-hardened stainless steel (CrSS) milling media and container. After drying out the acetone, the precursor mixture was ground in an agate mortar, pressed into pellets and were annealed at 600 °C (heating rate = 10 °C/min) for 12 h inside a tube furnace (under Ar flow) to obtain the desired product. Structural Characterization: Powder X-ray diffraction patterns of synthesized samples were collected by a Bruker AXS D8 ADVANCE diffractometer equipped
ACS Paragon Plus Environment
Page 4 of 17
Page 5 of 17
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Chemistry of Materials
with a Co-Kα source (λ1 = 1.78897 Å) (operating at 35 mA/ 40 kV) and a Vantec-1 linear position sensitive detector. The scan was performed in the 2θ range of 10~80° (at 0.02°/s). Rietveld refinement was performed using TOPAS-Academic V4.1 software. Mössbauer spectroscopy was performed with a Topologic System Inc. spectrometer equipped with a 57Co γ-ray source duly calibrated with a standard α-Fe foil. Typically, ~0,1 g powder samples were sealed in Pb-sample holder by polyethylene films and the spectra were collected for ~5 h. The model fitting was performed with Moss Winn 3.0 software. X-ray absorption spectroscopy (XAS) analysis was conducted at beamline 9C at Photon Factory, High Energy Accelerator Research Organization (KEK). A water-cooled Si(111) double-crystal monochromator was employed for energy selection. The Fe/Co K-edge absorption spectra were collected in transmission mode at room temperature and the intensities of incident and transmitted X-rays were monitored by ionization chambers. Morphology Characterization: The particle morphology was observed on powder samples mounted on conducting carbon paste by a field emission type scanning electron microscope (SEM) unit (Hitachi S-4800) operating at 2-5 kV. Electrochemical Measurement: To improve the electronic conductivity, 85wt% of the synthesized materials were mixed with 8wt% ECP and 2wt% of VGCF in shaking milling for 30 min to obtain the Li2MxFe1–xP2O7-carbon-composite. For coin cell batteries which were used for normal galvanostatic cycling, the working electrodes were formulated by mixing 95wt% of the Li2MxFe1–xP2O7/C composite and 5wt% of polyvinylidene fluoride (PVdF) binder in minimal amount of NMP (Nmethylpyrrolidone) solvent. The slurry was cast on an Al-film acting as current collector. After drying at 60 °C in vacuum overnight, circular disks (φ = 15.95 mm) were punched out. The active material was ~20 µm thick with a cathode loading of 3 mg/cm2. The cathode disk as well as Al mesh and cathode can was dried at 120 °C in glass tube oven overnight to remove residual NMP and water. Then these cathode disks were used to assemble 2032-type coin cells inside an Ar-filled glove box (Miwa, Japan) with Li metal foil as anode, polypropylene film as separator and 1 M LiPF6 dissolved in a mixture of ethylene carbonate/diethyl carbonate (EC/DEC, 3/7 v/v) as electrolyte. These coin cells were subjected to galvanostatic cycling with a TOSCAT3100 battery tester (Toyo system, Japan) (at C/20 rate, 25 °C), with constant voltage relaxation down to C/200 applied at the end of the charging segment. For batteries which were used for ex situ measurements, the working electrodes were prepared in another way. 95wt% of Li2MxFe1–xP2O7/C composite and 5wt% of Polytetrafluoro - ethylene (PTFE) binder were mixed in an agate mortar to get a thin sheet. Punch the sheet to get circular electrode (φ = 10 mm) with a cathode loading of 40 mg/cm2 and press it to the Al-film (φ = 15.95 mm). After ex situ measurements finished, the batteries were decomposed inside the Ar-filled glove box (Miwa, Japan), the electrodes were rinsed by Dimethyl Carbonate to remove the electrolyte on the surface and sealed into poly flex bag for next measurement (XANES, Mössbauer). All of experiments of Li2MnxFe1–xP2O7 are reported by Furuta et al. [11] 3. Results and Discussion 3.1. Binary Solid-solution Compounds Figure 1 shows that a complete family of Li2MxFe1-xP2O7 (M = Co, Mn) solidsolution was synthesized successfully with 0≤ x ≤1, while for Li2MgxFe1-xP2O7, the target samples were obtained limited to 0≤ x ≤0.5. It was found that with the content
ACS Paragon Plus Environment
Chemistry of Materials
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
of Mg further increasing, impurity phases emerged. The corresponding lattice parameters are plotted in Figure 2. Both of Co and Mg substitution induced a continuous lattice contraction with more substitution while Mn showed an expansion. Anisotropy was small for all cases. This should attribute to different ionic radius: Mn2+ (83 pm, high spin), Co2+ (74.5 pm, high spin), Mg2+ (72 pm) vis-à-vis Fe2+ (78 pm, high spin).[14] In all of these three cases, the lattice parameters showed a linear change, which also confirmed the formation of solid-solution. (All of the corresponding lattice parameters are summarized in Table S1 for Li2MxFe1-xP2O7 (M = Mn, Co, Mg) respectively.) The morphology of representative sample is shown in Figure S1. The collected Mössbauer spectra and respective parameters of Li2MxFe1xP2O7 (M = Co, Mg) are shown in Figure S2. We have already published the data for the case of M = Mn[11]. Each spectrum confirms the constituent Fe to be majorly in Fe2+ oxidation state independent of composition, restricting the Fe3+ impurity below 5 %. Similar to Li2MnxFe1–xP2O7 system, the Fe−Co, Mg solid-solution compounds led to broad, non-resolved peaks, which could be fitted by assuming two distinct sites occupancy by Fe2+ species. The outer doublet was attributed to the FeO6 (MO6 site, site 1) octahedral site, whereas the inner doublet was assigned to the combination of FeO5 (MO5/LiO5 site, site 2) . The inner doublet contains Fe signals coming from LiFe anti-sites, but it was impossible to distinguish or quantify the signals from FeO5 and LiO5 polyhedra. It confirms an equal and random distribution metals (Fe and M) between site 1 (MO6) and site 2 (MO5/LiO5) with no preferential occupancy. 3.2. Electrochemistry 3.2.1. Galvanostatic Cycling The electrochemical performance of Li2MxFe1-xP2O7 (M = Co, Mn, Mg) compounds was tested in standard half-cell coin-type architecture. The results of galvanostatic cycling are shown in Figure 3. The theoretical capacity for one-electron reaction of Li2FeP2O7 is 110mAh/g, and each case in Figure 3 just showed the capacity close to one-electron reaction of corresponding Fe’s ratio. This is a hint that only Fe3+/Fe2+ had redox reaction in all of these galvanostatic cycling tests. Because there are two kinds of metal in the solid-solution Li2MxFe1-xP2O7, it is necessary to confirm these two peaks correspond to which metal element’s reaction. For Mn, it has been proved in the report of Fututa et al.[11] that no Mn’s contribution in all of reaction behaviors below 4.5V. For Mg, due to its redox inactivity, it can be affirmed that all of redox peaks of Li2MgxFe1-xP2O7 are arisen by Fe’s reaction. And for Co, with the Co3+/Co2+ redox couple located at 4.9 V,[8] the redox potentials in the potential range of 3.5~3.9 V can be solely ascribed to the constituent Fe-species. To further ascertain the Feredox activity at 3.9 V, ex situ XANES (X-ray Absorption Near-Edge Structure) had been performed on the representative material Li2Co0.8Fe0.2P2O7 (Figure 4). The Fe K-edge shifted to low energy side (black spectrum to red to green) by electrochemical lithium intercalation, while the Co K-edge exhibited no significant change, affirming only Fe3+/Fe2+ redox is involved with negligible contamination by Co3+/Co2+ redox couple. Similar results were obtained for Mn substitution case, which means that contamination by redox couple by other transition meals has negligible impact to Fe redox potential tunability .[11] 3.2.2. Differential Voltage Profiles (dQ/dV) To get more information of the electrochemical properties, differential voltage profiles (dQ/dV) were obtained by calculating the differential of capacity over potential. In Figure 5, for all of the three doping cases, with substitution of M in the system of Li2MxFe1-xP2O7 (M = Co, Mn, Mg), two major common aspects for
ACS Paragon Plus Environment
Page 6 of 17
Page 7 of 17
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Chemistry of Materials
Fe3+/Fe2+ redox couple were found in the dQ/dV plots. First, the top figures are dQ/dV plots for pristine Li2FeP2O7. In the first charging process, there are two distinct anodic peaks. However, in the first discharging and all of subsequent cycling, the two peaks merged into one centered around 3.5 V. On the contrary, after doping some foreign metal M, these original two peaks, one of which positions at higher-voltage rigion, were stabilized even after repeated cycling. Second, with the amount of M doping increasing, both of two redox potentials slightly upshift to higher potential side. Especially for the high potential peak, it can reach a value as high as 3.9 V, which is the highest Fe3+/Fe2+ redox potential reported till date. Such a high voltage Fe3+/Fe2+ operation over 3.5 V has long been believed to be possible only by existence of much more electronegative but hygroscopic anions as SO43- or F- [13.15]. Thereby, this is the first universal confirmation of >3.5 V operation by stable simple phosphate material. 3.3 Mechanisms 3.3.1. Two Redox Peaks Stabilization In the study of the Fe-Mn binary Li2MnxFe1-xP2O7 pyrophosphates,[11] Furuta et al. considered that the two peaks in the first charge process are from two distinct sites of Fe’s reaction, one octahedral and another being trigonal bipyramid with edge-sharing configuration. The two peaks merged into one in the first discharge can be due to some irreversible structural disorder, which can be suppressed by introducing Mn. In this study, substitution of Co and Mg led to the similar observation and also supported the above general consideration. Irrespective of M’s size and redox activity, the amount of lithium extraction from the lattice is decreased at the voltage region lower than 4.2 V. In any host-guest intercalation materials where guest specie originally occupy specific crystallographic site, the host structure tends to be more stable in smaller extraction of guest species. We are now exploring the details of the irreversible structural disorder induced during the first charge process of Li2-xFeP2O7 and how it is suppressed by introducing other elements. Preliminary analysis showed it includes irreversible Fe migration into interstitial sites, as often observed in other Fe-based intercalation compounds such as LixFe2O3,[16] Li2-xFeSiO4[17] and Li1+xFe5O8,[18] of which driving force is considered strong local coulombic interactions among adjacent cations. The details of this phenomenon will be reported elsewhere.[19] Pristine Li2FeP2O7 has two distinct sites of trigonal-bipyramidal FeO5 and octahedral FeO6 with edge-sharing configuration. During the delithiation reaction, the energetically more unfavorable Fe at FeO5 site can migrate into interstitial site, where the driving force for Fe migration was thought to be the strong coulombic interaction between two adjacent Fe3+. Figure 6 shows a schematic energy diagram for the delithiation reaction. The difference of free energy between the extracted and the pristine phases determines the reaction potential. For Li2FeP2O7 case (right portion of Figure 6), the spontaneous structural rearrangement decreases the free energy of delithiated state, and thus, decreases the reaction potential corresponding to ᇞE1 in right portion of Figure 6. On the other hand, the coulombic interaction for Fe3+ - Fe3+ in the stabilized original structure can result in a larger energy difference ᇞE2 compared to ᇞE1 (Figure 6). This means that the intrinsic reaction potential of hypothetical rigid Li2FeP2O7 structure is higher than experimental one, if there is no structural rearrangement at the delithiated state. In summary, the strong local Fe3+ Fe3+ interaction induced both the higher voltage generation and structural rearrangement (lower voltage generation) in a competitive way.
ACS Paragon Plus Environment
Chemistry of Materials
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Then, in the doping case, there are three kinds of edge-sharing structures for Fe in the starting material (Figure 6, left portion). The foreign metal M can substitute both FeO6 and FeO5 to form the MO6 – FeO5 and FeO6 – MO5 edge-sharing configurations. The third one is the original FeO6 – FeO5 without substitution. After full delithiation, the Li concentration is higher than that in the un-doped Li2FeP2O7 case, because all of M ions remain inert below 4.2 V. The remaining Li can block the Fe migration and can stabilize the Fe’s original local structure and whole crystal structure, which means that the redox potential of some portion of Fe3+/Fe2+ in Li2MxFe1-xP2O7 should be higher than that in pristine Li2FeP2O7 case. Extremely high redox potential of 3.9 V for Fe3+/Fe2+ in Li2-yMxFe1-xP2O7 can be attributed to the edge-sharing structure. This value of redox potential for Fe3+/Fe2+ is very close to a recently reported system, triplite LiFeSO4F,[13] which can reach the redox potential as high as 3.9 V for Fe3+/Fe2+, while it is limited to 3.6 V for tavorite structure.[15] Chung et al. [20] proposed the main reason is that two octahedral sites of Fe have edge-sharing configuration in triplite structure, whereas they have corner-sharing configuration in tavorite structure. So in this study, we can also assume that the redox potential of 3.9 V is from the octahedral site of Fe, which has the edge-sharing structure. 3.3.2. Additional Redox Peaks Upshifting In the work of Li2MnxFe1-xP2O7, [11] it was argued that one possible reason behind the additional Fe3+/Fe2+ upshifting by substitution is due to the inductive effect.[11] The Mn-substitution increases the unit cell volume owing to larger ionic radius of Mn. It reduces the hybridization between Fe and O, lower the Fe 3d - O 2sp antibonding state and thus giving higher Fe redox potential. However, in the present case, similar degree of Fe3+/Fe2+ upshifting is observed in both Fe-Co binary Li2CoxFe1-xP2O7 and Fe-Mg binary Li2MgxFe1-xP2O7 pyrophosphates having exactly opposite trend in the lattice dimensions (Figure 2). Gradual Co substitution shrinks the lattice volume from 1034 Å3 (Co = 0%) to 1016 Å3 (Co = 100%), thereby reducing the Fe-O bond length (S-Table 1). As inductive effect argument, it should lead to stronger hybridization between Fe-O and hence decrease the redox potential of Fe-metal center. Based on the inductive effect theory, another factor which may result in voltage upshifting is importing M with stronger electronegativity than Fe. If so, in Fe-O-M, the covalency between Fe and O will be lowered and the redox potential of Fe will increase. However, only Co2+ has stronger electronegativity than Fe2+ while Mn2+ and Mg2+ are weaker. So the observation of similar Fe3+/Fe2+ upshifting independent of lattice volume (expansion/contraction) and electronegativity in binary compounds proves the Fe-redox tunability in pyrophosphate structure cannot be explained solely by inductive effect. Except the classical inductive effect, Malik et al. [21] claimed that the redox potential upshifting in binary olivine phosphates can be explained by considering the change in the relative energy of the intermediate compounds. This concept can also be extended to the binary pyrophosphate system. For one electron reaction of Li2FeP2O7, after fully charging to LiFeP2O7, half of Li sites changed to vacancies (VLi’) and all of Fe2+ changed to Fe3+, so the VLi’ is mostly surrounded by Fe3+ because of the strong attractive interaction between VLi’ and Fe3+. However, in case of substituted Li2MxFe12+ xP2O7, M is inert and does not take part in any reaction below 4.2 V. Thus, after full delithiation, some VLi’ will be surrounded by M2+. The unfavorable VLi’-M2+(MFe×) and/or LiLi×-Fe3+(FeFe•) interactions will increase the energy of this intermediate composition Li2-yMxFe1-xP2O7, so the difference of standard Gibbs energy of formation between before/after delithiation is larger than Li2FeP2O7 case, and result in
ACS Paragon Plus Environment
Page 8 of 17
Page 9 of 17
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Chemistry of Materials
higher voltage in this region where Fe3+/Fe2+ redox reaction occurs, via thermodynamic voltage definition, Voltage = E(Li) + E(charged) – E(discharged), ignoring TS and PV terms. 4. Conclusions The potential tunability for Fe3+/Fe2+ redox couple at unusually high voltage region of 3.5 – 3.9 V vs. lithium are general for any metal M doping in Li2MxFe1-xP2O7 system. The phenomena include two aspects; (1) two redox reactions at different potentials are stabilized with doping of foreign metal M, (2) with more dopant, both of two redox reactions upshift to higher potential, one even goes close to 4 V. Substitution of M into Fe sites may suppress migration of Fe from the FeO5 site upon charging, and original two distinct Fe sites become robust to stabilize the edge sharing geometry of FeO5 and FeO6 polyhedral with large Fe3+ - Fe3+ coulombic repulsion energy, leading to the two distinct redox reactions with inherently high potentials. The change in the relative energy of the intermediate compounds, which is induced by the unfavorable VLi’-M2+(MFe×) and/or LiLi×-Fe3+(FeFe•) interaction in the doping case, may be a reason for the further potential upshifting. The classic inductive effect cannot explain the redox potential upshifting phenomenon in this case. ASSOCIATED CONTENT Supporting Information. Lattice parameters of Li2MxFe1-xP2O7 (M = Mn, Co, Mg) , SEM image of a representative pyrophosphate powder, Li2FeP2O7, and Mössbauer spectra and their analysis results. This material is available free of charge via the Internet at http://pubs.acs.org.” AUTHOR INFORMATION Corresponding Author *
[email protected] Present Addresses †Mitsubishi Gas Chemical Co.. Author Contributions The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Funding Sources This work was financially supported by the Cabinet Office, Government of Japan, the Funding Program for World-Leading Innovative R&D on Science and Technology, and Mitsubishi Motor Company. ACKNOWLEDGMENT PB is grateful to the Japan Society for the Promotion of Sciences for a ‘JSPS Fellowship’ at the University of Tokyo. XAS experiments were performed by KEKPF User Program No. 2011G063. The crystal structures were drawn by using the VESTA software.[22]
ACS Paragon Plus Environment
Chemistry of Materials
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
References [1] F. d’Yvoire, Bull. Soc. Chim. Fr.1962, 1, 1224. [2] J.P. Gamondes, F. d’Yvoire, A. Boulle, C.R. Acad. Sci. Paris. 1969, C269, 1532. [3] J. Belkouch, L. Monceaux, E. Bordes, P. Courtine, Mater. Res. Bull. 1995, 30, 149. [4] P. Barpanda, S. Nishimura, A. Yamada, Adv. Energy Mater. 2012, 2, 841. [5] A. K. Padhi , K. S. Nanjundaswamy , C. Masquelier , S. Okada , J. B. Goodenough , J. Electrochem. Soc. 1997, 144, 1609. [6] L. Adam , A. Guesdon , B. Raveau , J. Solid State Chem. 2008, 181 , 3110. [7] S. Nishimura , M. Nakamura , R. Natsui , A. Yamada , J. Am. Chem. Soc. 2010, 132 , 13596. [8] H. Kim , S. Lee , Y.-U. Park , H. Kim , J. Kim , S. Jeon , K. Kang , Chem. Mater. 2011, 23, 3930. [9] D. Shimizu, S. Nishimura, P. Barpanda, A. Yamada, Chem. Mater. 2012, 24, 2598. [10] M. Tamaru, P. Barpanda, Y. Yamada, S. Nishimura, A. Yamada, J. Mater. Chem. 2012, 22, 24526. [11] N. Furuta, S. Nishimura, P. Barpanda, A. Yamada, Chem. Mater. 2012, 24, 1055. [12] Zhou, H.; Upreti, S.; Chernova, N.A.; Hautier, G.; Ceder, G.; Whittingham, M.S. Chem. Mater. 2011, 23, 293. [13] P. Barpanda, M. Ati, B. C. Melot, G. Rousse, J-N. Chotard, M-L. Dupont, M. T. Sougrati, S. A. Corr, J-C. Jumas and J-M. Tarascon, Nat. Mater. 2011, 10, 772. [14] R.D. Shannon, Acta Crystallogr. 1976, A32, 751. [15] N. Recham, J-N. Chotard, L. Dupont, C. Dellacourt, W. Walker, M. Armand and J-M. Tarascon, Nat. Mater. 2010, 9, 68. [16] Y. Huang, Z. Dong, D. Jia, Z. Guo, W. Cho, Solid State Ionics, 2011, 201, 54. [17] Anton Nytén, Saeed Kamali, Lennart Häggström, Torbjörn Gustafsson and John O. Thomas, J. Mater. Chem. 2006, 16, 2266. [18] Y. T. Lee, C.S. Yoon, Y. S. Lee, and Y. Sun, J. Power Sources, 2004, 134, 88. [19] S. Nishimura, T. Kurita, and A. Yamada, unpublished. [20] S. C. Chung, P. Barpanda, S. Nishimura, Y. Yamada, A. Yamada, Phys. Chem. Chem. Phys.2012, 14, 8678 [21] R. Malik, F. Zhou and G. Ceder, Phys. Rev. B, 2009, 79, 214201. [22] K. Momma, F. Izumi, J. Appl. Crystallogr. 2011, 13, 1272.
ACS Paragon Plus Environment
Page 10 of 17
Page 11 of 17
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Chemistry of Materials
Figure Captions Figure 1. X-ray diffraction patterns of isostructural binary compounds Li2MxFe1-xP2O7 (M = Co, Mn, Mg). The very similar patterns prove that complete families of solid-solution for Co and Mn substitution (0≤ x ≤1) were synthesized successfully, while substitution content of Mg is limited to 0≤ x ≤0.5 to form the solid-solution.
Figure 2. The linear change in all lattice parameters of Li2MxFe1-xP2O7 (M = Co, Mn, Mg) pyrophosphates, proving the formation of solid-solution, retaining the same structure.
Figure 3. Galvanostatic cycling profiles of coin cell batteries with (a) Li2CoxFe1-xP2O7 [x = 0~1], (b) Li2MnxFe1-xP2O7 [x = 0~1], (c) Li2MgxFe1-xP2O7 [x = 0~0.5] as cathode, cycled between 2 to 4.5 V at the rate of C/20 (1 Li in 20 h) in coin cell settings. In each case, complete Fe3+/Fe2+ redox activity is realized, which solely contributes to the observed capacity.
Figure 4. Ex situ XANES spectra of Li2Co0.8Fe0.2P2O7 obtained for Fe K-edge and Co K-edge at fully charged state (4.7V), partially discharged state (3.7V), and fully discharged state (2.0V). Complete inactivity of Co species and sole activity of Fe-species is clearly observed during electrochemical cycling, further proving the sole presence of Fe3+/Fe2+ redox activity.
Figure 5. The differential galvanostatic voltage profiles (dQ/dV) of coin cell batteries with (a) Li2CoxFe1-xP2O7 [x = 0~1], (b) Li2MnxFe1-xP2O7 [x = 0~1], (c) Li2MgxFe1-xP2O7 [x = 0~0.5] as cathode. In all of these three doping cases, two redox potential tunability phenomena were found: 1) Original two Fe3+/Fe2+ redox peaks were stabilized with more substitution; 2) Both of two redox peaks upshifted to high potential sites with more substitution. Particularly, the high potential site peak is gradually upshifted close to 4V, matching with the highest ever Fe3+/Fe2+ redox potential.
Figure 6. Schematic description of free energy difference between starting and delithiation materials. The right portion is the pristine Li2FeP2O7 system. The spontaneous structural rearrangement (Fe’s migration ) destroys the edge-sharing configuration and decreases the free energy of delithiated state, which results in an energy difference of ᇞE1. The left portion is the doping system Li2MxFe1xP2O7. After full delithiation, the Li concentration is higher than that in the Li2FeP2 O7 case, because all of M ions remain inert. The remaining Li can block the Fe migration and can stabilize the Fe’s original local structure and whole crystal structure, which means that the energy difference ᇞE2 should be higher than ᇞE1.
ACS Paragon Plus Environment
Chemistry of Materials
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 1
ACS Paragon Plus Environment
Page 12 of 17
Page 13 of 17
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Chemistry of Materials
Figure 2
ACS Paragon Plus Environment
Chemistry of Materials
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 3
ACS Paragon Plus Environment
Page 14 of 17
Page 15 of 17
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Chemistry of Materials
Figure 4
ACS Paragon Plus Environment
Chemistry of Materials
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 5
ACS Paragon Plus Environment
Page 16 of 17
Page 17 of 17
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Chemistry of Materials
Figure 6
ACS Paragon Plus Environment