A Search for Low-Irreversible Capacity and High-Reversible Capacity

Subscriber access provided by CMU Libraries - http://library.cmich.edu

Article

A search for low irreversible capacity and high reversible capacity positive electrode materials in the Li-Ni-Mn-Co pseudo-quaternary system Ramesh Shunmugasundaram, Rajalakshmi Senthil Arumugam, Kristopher J. Harris, Gillian R. Goward, and J. R. Dahn Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/acs.chemmater.5b02104 • Publication Date (Web): 02 Dec 2015 Downloaded from http://pubs.acs.org on December 12, 2015

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 45

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

A search for low irreversible capacity and high reversible capacity positive electrode materials in the Li-Ni-Mn-Co pseudo-quaternary system Ramesh Shunmugasundarama, Rajalakshmi Senthil Arumugamb , Kristopher J. Harrisc, Gillian R. Gowardc and J.R. Dahn*,a,b a

Dept. of Chemistry and b Dept. of Physics and Atmospheric Science, Dalhousie University, Halifax, NS, Canada, B3H 3J5

c

McMaster University, Department of Chemistry and Chemical Biology, 1280 Main St. West, Hamilton, Canada, L8S 4L8

*corresponding author - [email protected]. Tel.: 001-902-494-2628; fax: 001-902-494-5191

1 ACS Paragon Plus Environment


Table of content graphic

0 1.333 1 1.31

0.2 1.286 1.259

Co

0.4 1.231 1.2

0.6 1.167 1.13

0.8 1.091

1.2

1.091

1.2

1.167

1.13

1.091

1.231 0.8

1.259

1.231

1.167

1.13

1.286

1.167

1.13

1.091

1.091 0.6

1.048

Mn

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

Page 2 of 45

0.4

1.048

1.048

0.2

1.048

1.048

1

0

0

0.2

0.4

0.6

Ni


0.8

1

Page 3 of 45

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


Abstract A comprehensive search for Li-ion battery positive electrode materials that can simultaneously exhibit low irreversible capacity loss (IRC) (~ 10 % or less) and high reversible capacity (> 240 mAh/g) was performed in the Li-Ni-Mn-Co-O pseudo-quaternary system. An array of high-capacity Li-rich layered oxides, most of which show an “oxygen release” plateau during the first charge, was synthesized with a wide-range of Ni, Mn and Co compositions and their first cycle electrochemical properties were investigated. Low-IRC materials (IRC ~ 10 % or less) could be synthesized at many Ni-Mn-Co combinations by synthesizing with less lithium than required by site occupation and oxidation state rules. Many of these “Li-deficient” low-IRC materials were found to be single phase layered materials with inherent metal-site vacancies in their pristine state. For such single phase materials, the amount of IRC depends on the concentration of metal-site vacancies in their pristine state.

Increasing the Li deficiency

eventually caused the appearance of the spinel phase, which, when it appears, lowers the IRC, irrespective of the Ni-Mn-Co precursor composition. The number of metal-site vacancies that can be incorporated into the single-phase layered materials depends on the overall metal composition, especially the Co concentration. Low-IRC behaviour is correlated to the fraction of metal-site vacancies in the layered phase in both the single and two-phase materials. 7Li NMR studies on low-IRC materials revealed the relative population of Li between the Li and TM layer. Formula unit calculation based on 7Li NMR results suggests that metal-site vacancies preferably occupy the sites in the Li layer, which could provide room for the intercalation of extra Li into the structure, hence reducing irreversible capacity.



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

Introduction Li-ion batteries for applications such as electric vehicles require high capacity positive electrode materials1. Li-excess materials are considered to be next generation candidates for such applications because they can deliver, with excess Li, specific capacities as high as 250 mAh/g2,3. Li-excess materials, such as Li[Li0.2Ni0.13Mn0.54Co0.13]O2, are single-phase layered Li[M]O2 type structures, with more lithium atoms than transition metal atoms in their pristine state. Such materials normally exhibit a plateau at ~ 4.5 V (vs Li/Li+) during the first charge (delithiation), and normally have first cycle irreversible capacities (IRC) close to 20 %4,5,6 of their first charge capacities. However, high capacity Li-rich positive electrode materials that exhibit the typical 4.5 V plateau but having IRC as low as 6.5 % have recently been reported7. Figure 1 shows a ternary composition diagram of Ni, Mn and Co, in which every point represents a NixMnyCo1-x-y precursor composition (e.g. Ni0.4Mn0.2Co0.4CO3). Layered oxides, Li[M]O2, can be made from such precursors by heating with an appropriate lithium source like Li2CO3. Layered oxides without vacancies and subject to the oxidation state rules: Mn4+, Co3+, Ni2+ or Ni3+ and Li1+ must have more lithium atoms than transition metal atoms when synthesized from precursors lying above the line drawn from Co to Ni0.5Mn0.5 in Figure 1. This triangular area bounded by Co, Ni0.5Mn0.5 and Mn in Figure 1 is called the “Li-rich region”. The red points within the Li-rich region in Figure 1 are the Ni-Mn-Co precursor compositions that have been studied in this work. The low-IRC materials (labelled as A2, A3Q, B2 and B3) reported previously7 were made from a pair of precursors (labelled A and B), which are within a yellow ellipse in Figure 1. Those low-IRC materials were single-phase layered materials with inherent metal-site


Page 4 of 45

Page 5 of 45

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


vacancies7. Recently Rowe et al. have shown a Li-Ni-Mn-Co-O based positive electrode material with metal-site vacancies that exhibits promising properties such as high stability against electrolytes at high potential8.

Most recently, McCalla et al. have also reported positive

electrode materials with metal-site vacancies9. It is important to search for low-IRC materials that may contain metal-site vacancies by scanning through a wide range of compositions in the Li-Ni-Mn-Co-O pseudo-quaternary system as is done in this report The Li/TM ratio and the synthesis conditions are very crucial in the synthesis of pure layered LiMO2 type material. If a LiMO2 material is synthesized with less Li than required for stoichiometric charge balance with excess oxygen available in the atmosphere, the co-existence of a spinel phase (LiM2O4) along with the layered phase is normally expected. The layeredspinel coexistence region in the Li-Ni-Mn-O pseudoternary system has been mapped by McCalla et al10. McCalla et al. have also shown, in a separate report11, the existence of single-phase layered materials such as Li[□1/6Ni1/6Mn2/3]O2, which contain metal site vacancies and have a considerable departure from

a 1:1 cation:anion ratio.

In the latter case, quenching after

calcining was required to synthesize single-phase Li[□1/6Ni1/6Mn2/3]O2. Attempts to synthesize Li[□1/6Ni1/6Mn2/3]O2 with slow cooling resulted in a layered-spinel composite. Compared to Li[□1/6Ni1/6Mn2/3]O2 that did not show low-IRC behavior, the low-IRC materials previously reported7 had metal-site vacancies, also contained Co and were synthesized with slow cooling. It appears that the presence of Co is important for low-IRC and also favors the formation of a single layered phase during slow cooling. This raises a question about the contribution of the overall metal composition, especially Co, to promote low-IRC behavior and to promote the formation of pure layered single phases under stoichiometrically Li deficient conditions. Hence, another focus of this article is to evaluate the combined effect of Li-deficiency and overall metal 5 ACS Paragon Plus Environment


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

composition on the structural and electrochemical properties of materials in the Li-Ni-Mn-Co-O system.


Page 6 of 45

Page 7 of 45

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


Experimental Carbonate precursors Ni(II)xMn(II)yCo(II)1-x-yCO3 were synthesized using a coprecipitation method in a continuously stirred tank reactor (CSTR). The co-precipitation was carried out by mixing aqueous solutions of mixed metal sulfate (Ni(II)xMn(II)yCo(II)1-x-ySO4) and sodium carbonate (Na2CO3). An aqueous solution of 0.1 M NH4OH was used as a coordinating agent for the transition metals. The temperature during co-precipitation was controlled at 60οC and the pH was controlled at 8.0 by adding appropriate amounts of acid (H2SO4) or base (NaOH). After the co-precipitation reaction, the resulting suspension was collected, washed and filtered. The wet MCO3 precipitate was washed with deionized water at least five times to remove any residual Na2SO4 in the solid. The resulting Ni(II)xMn(II)yCo(II)1-xyCO3

precipitate was collected and dried in an oven at 100-120oC in air for about 12 hours. Lithium-transition-metal oxides were synthesized from Ni(II)xMn(II)yCo(II)1-x-yCO3

powder and Li2CO3. The well-mixed powders were fired at 900◦C in air. The heating and cooling profile used was the same as in reference7. Li loss can occur during synthesis at very high temperature12 so 5 wt % excess of Li2CO3 was added before firing to compensate for the Li loss. For this reason, the compositions of the products were not assumed, but instead, the elemental composition of every as-prepared sample was obtained using inductively coupled plasma optical emission spectroscopy (ICP-OES). ICP-OES analysis uses high precision standards of known concentration in order to create a calibration curve to determine compositions of measured samples. The apparatus used in this study was a Perkin Elmer Optima 8000 ICP-OES Spectrometer. The detection limit of lithium and TM were 0.1 µg/L, or 0.1 parts per billion (ppb) and approximately 2 - 5 ppb respectively. The detector was known to saturate at lithium concentrations of over 2 ppm. Due to the varying 7 ACS Paragon Plus Environment


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

mass ratios of TM to Li, careful consideration was taken to design a concentration window in which measurements were performed. A 0 ppm “blank” was used, as well as two standards: one containing 0.5 ppm Li and 1 ppm TM, another containing 1 ppm Li and 2 ppm TM. Standards were prepared by accurately pipetting 1000 ppm single-element standards (Sigma-Aldrich) and diluting with 2% HNO3. The 2% HNO3 used for sample preparation was prepared using deionized water (18.2 MΩ cm). In order to stay approximately in the 0.5 – 2 ppm bracketed region, ~10 mg of each sample was added to 2 mL aqua-regia and left overnight to dissolve fully. Approximately 10 µL was then pipetted and diluted with ~12 mL of 2% HNO3. Samples containing air bubbles had an additional ~5 µL added. The target sample concentrations were designed such that measurement errors (such as those caused by air bubbles) were not likely to cause sample concentrations to lie outside of the measurement concentration window. All calibration curves were linear, with an R2 value of 0.999411 - 1.00000. All the studied samples (of different compositions) were analyzed in a single run, for which the calibration was performed with same standard solutions. Some samples were also analyzed for Na and S as possible contaminants from the precursor synthesis step. Table S1 (supplementary information) shows that the Na and S contents are less than about 0.5 mol.% of the total cation content in the samples. A Siemens D5000 diffractometer equipped with a copper target X-ray tube and a diffracted beam monochromator was used to collect the powder X-ray diffraction patterns. The collected XRD patterns were refined (Rietveld method) using “Rietica”13. A Hitachi S-4700 field emission scanning electron microscope with an accelerating voltage of 10 kV and an emission current of 15 µA was used to record the scanning electron micrograph (SEM) images.


Page 8 of 45

Page 9 of 45

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


An AccuPyc II 1340 gas displacement pycnometer using He gas (Ultra high purity 99.999 %) was used to measure the true density of the samples. The pycnometer was calibrated with the given sample cup and the appropriate standard, for which the volume was known (given by the supplier). The calibration was repeated until the measured volume of the standard and the given volume agreed within the variance. The sample under investigation was heated up to 400oC for at least about 3 hours to remove any impurities at the surface. Then the sample was weighed accurately (around 2 g) and transferred to the pycnometer chamber with minimum exposure to air. The measurement was initiated by purging the sample with helium gas 5 times. The purge-fill pressure and the cycle-fill pressure were set as 19.5 psig. Each run involved 5 cycles with an equilibration end time of 1 min. The average volume from the 5 cycles was extracted to deduce the density of the samples. All 7Li NMR experiments were performed under a 4.7 T applied field using a Bruker DRX console. Each spectrum was collected on a mg size sample undergoing 60 kHz magicangle spinning using a Bruker 1.3 mm probe in less than one hour. Spectra are referenced to 1M LiCl(aq) at 0 ppm, and all experiments used 100 ms recycle delays and 1.5 us π/2 pulses. Spectra containing only isotropic shifts were generated using the projection magic-angle-turning phase-alternating spinning-sideband, pj-MATPASS,14 method of separating sidebands into different slices of a 2D spectrum, which were then aligned and summed to yield the presented spectra. Working electrodes were made from the synthesized powders as described in detail by Marks et al.15. The components of electrodes were as follows: 90. wt % of the positive electrode material, 5. wt % of Super C45 carbon black (commercially available from TIMCAL), 5. wt % of polyvinylidene difluoride (PVDF) binder (commercially available from ARKEMA) and the 9 ACS Paragon Plus Environment


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

Page 10 of 45

appropriate quantity of N-methyl pyrrolidone (NMP) solvent. All the ingredients were thoroughly mixed in a mixer (Mazerustar) and a uniform slurry was made. The freshly prepared slurry was spread into a film on an Al foil using a notch bar spreader. The electrode spread was dried for at least 3 hours at 120ºC to completely remove the NMP. After complete drying, the dried electrode sheet was calendared at 200 Bar. The compressed electrode sheet was punched into several circular disks, which were used as working electrodes in the coin cells. Coin-cells were assembled in an argon-filled glovebox. Each coin cell was a half-cell with a positive electrode and a Li foil negative electrode with two layers of separator (Celgard #2300) in between. 1M lithium hexafluorophosphate (LiPF6) in 1:2 ethylene carbonate (EC)/diethyl carbonate (DEC) (BASF) was used as the electrolyte. The half-cells were galvanostatically cycled using a specific current of 10 mA/g at 30.ᵒC using a computer-controlled charger system (Maccor 4000). The voltage window was set to 4.8 V and 2.0 V for the first charge-discharge cycle and then cells were cycled between 4.6 V and 2.8 V using the same specific current and temperature.


Page 11 of 45

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


Results and Discussion The stoichiometric amount of Li required to make a layered phase, without vacancies, from any NixMnyCo1-x-yCO3 precursor can be calculated assuming that the total metal occupancy and total metal charge must be 2 and 4, respectively, in a LiMO2 formula unit. In Figure 1, the number (in blue) displayed above each point representing NixMnyCo1-x-yCO3 precursor is the calculated number of moles of Li (u) required to be mixed with (2-u) moles of NixMnyCo1-x-yCO3 precursor to make one mole of stoichiometrically balanced single-phase LiMO2 type material. For example, the number of moles of Li (u) required to make one mole of LiMO2 type material (Li[Li0.66Ni0.083Mn0.416Co0.333]O2) from 0.833 moles of Ni0.1Mn0.5Co0.4CO3 precursor is 1.166. The calculation has been done assuming Ni, Mn and Co adopt the +2, +4 and +3 oxidation states, respectively, in LiMO2. Based on the recent report7, a search for low-IRC materials (IRC < 10%) in the Li-Ni-Mn-Co-O pseudo-quaternary system was made by synthesizing LiMO2 type materials with less Li than required for a stoichiometric balance. For example, less than 1.166 moles of Li has been mixed with 0.833 moles of Ni0.1Mn0.5Co0.4CO3 precursor to make a lowIRC material. In this work, a total of 15 Ni-Mn-Co carbonate precursors (red circles in Figure 1) were made and their lithiated oxides were studied. As such, each precursor composition forms a series of lithiated materials, in which the first member is stoichiometrically balanced whereas the rest of the samples in each series are Li-deficient but their Li/TM ratio is not necessarily less than 1. Figure 2 shows scanning electron micrograph (SEM) images of the precursors of A1, C1, H1, J1, L1 and O1 (labelled as A, C, H, J, L and O), which have different Ni:Mn:Co ratios.



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

Page 12 of 45

Although the same reaction conditions (pH, temperature, ammonia concentration, solution feed rates, stirring) were maintained during the synthesis of these precursors, their morphology varied somewhat, which is attributed to their varying composition. Van Bommel et al. showed how the pH and ammonia concentration affects the solubility of TM[NH3]n chelated ions and how it is essential to optimize these conditions for each transition metal ratio to get spherical particles of the same size16. However, reference 7 clearly shows that morphology changes had a minimal impact in reducing the IRC. Therefore it is believed that the variations in precursor morphology (Figure 2) do not significantly impact the measurements of IRC in this work. Table 1 shows the nominal composition of the precursors, the number of moles of Li (u) to be mixed with 2-u moles of precursor required to make one mole of stoichiometric LiMO2, the observed number of moles of lithium (from ICP-OES) per mole of total metals in the synthesized LiMO2 composition (v) and their relative difference [(u-v)/u x 100] which is called the “lithium deficiency” here. Table 1 also lists the first charge capacity, the first discharge capacity, the observed IRC and indicates whether the sample was single phase or not. Table 1 shows that the IRC decreases, in general, as the Li deficiency increases.

The low-IRC materials are broadly

distributed in the Li-rich region of the Ni-Mn-Co ternary system (Figure 1) suggesting that lowIRC behavior is not just confined to a particular Ni-Mn-Co composition. Single phase low-IRC materials Table 1 shows that Li-deficient materials were either single-phase layered materials or layered-spinel composites depending on the degree of Li deficiency and the overall metal composition. The single-phase low-IRC materials were found to be solid solutions between Li2MnO3 and LiMO217,18,19 containing “metal-site vacancies”7,11.


The existence of vacancies

Page 13 of 45

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


will be addressed in the next paragraph. Figure 3 shows the XRD patterns of a few single phase low-IRC materials in the angular range of 10o to 70o. They are all layered O3 structures (similar to α-NaFeO2), which can be defined in the R-3m spacegroup. The set of superstructure peaks between 20ᵒ and 30o are due to the ordering (C2/m) of Li or vacancies7,11 with Ni, Mn and Co to form the √3ahex x √3ahex superlattice in the TM layer. Superlattice formation does not necessarily require Li in the TM layer as was shown by McCalla et al for the case of Li[□1/6Ni1/6Mn2/3]O2 11 where vacancies order with TM atoms to form a superlattice. Table 2 shows the results of the XRD Rietveld refinement results of samples in selected series (D and K) including the lattice parameters. In every series, the lattice parameters increase as the Li deficiency increases, which is expected based on an increase in the TM/Li ratio. As reported recently, the single-phase low-IRC materials contain metal-site vacancies7. The total cationic charge of the low-IRC materials calculated from the as-obtained ICP-OES composition (Li:Ni:Mn:Co) was above 4 when the oxidation states of the cations are Li+, Ni2+, Co3+ and Mn4+. Hence the formula unit of any single-phase low-IRC compound would have excess oxygen (> 2). However, the total oxygen in a formula unit can be adjusted to 2 by introducing metal-site vacancies. The concentration of metal-site vacancies required in each composition was calculated based on the amount of excess cationic charge. That is, the ICP-OES composition formula unit was renormalized for a total oxygen content equal to 2 to get the formula unit with metal-site vacancies. For example, the formula unit of sample L2: Li0.982Ni0.203Mn0.402Co0.413O2.117 has been renormalized to Li0.9280.111Ni0.192Mn0.38Co0.39O2, which includes metal-site vacancies. Table 3 shows the metal atom ratios observed from ICPOES analysis, the calculated vacancy concentration per mole of synthesized LiMO2 and the proposed formula unit with metal-site vacancies included in the single-phase low-IRC materials. 13 ACS Paragon Plus Environment


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

Page 14 of 45

The existence of metal-site vacancies can be verified by comparing measured and calculated densities. The densities of the samples were measured using a helium pycnometer, and they were calculated from the lattice parameters (XRD) and chemical compositions that include metal-site vacancies. Table 3 shows the experimentally measured densities (pycnometer) and the calculated densities. For comparison purposes, the densities of the materials were also calculated for corresponding LiMO2 units in which the ICP-based metal composition was substituted for a LiM unit with no vacancies. In the latter case some TM ions must exist in lower oxidation states than normal in order to compensate for the total anionic charge of 4. For example, the zero metal-site

vacancy

counterpart

of

Li1.010.084Ni0.179Mn0.450Co0.277O2

(sample

B2)

is

Li1.054Ni0.187Mn0.470Co0.289O2. Table 4 shows the calculated densities of LiCoO2 and Li2MnO3 which essentially contain no vacancies. LiCoO2 was obtained from E-one Moli Energy and Li2MnO3 was synthesized by heating MnCO3 and Li2CO3 at 900oC for 12 hours in air. Table 4 shows the densities, measured using the helium pycnometer, of the LiCoO2 and Li2MnO3 samples which agree within 0.024 g/mL to the calculated densities.

This shows that the

pycnometer methods used here are reliable. Figure 4 shows a comparison of the measured densities with the calculated densities. In Figure 4, the densities calculated assuming metal-site vacancies are shown as black circles and the densities assuming no vacancies are shown as red circles. The measured densities and the calculated densities assuming metal site vacancies agree very well (black circles) resulting in data with a slope of 1.00 and an intercept of zero in Figure 4. By contrast, the measured densities do not agree well with densities calculated assuming no metal site vacancies (red circles) except for a few points where, it turns out, that the vacancy content is very small. The close agreement between the measured densities and the densities calculated assuming metal site 14 ACS Paragon Plus Environment

Page 15 of 45

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


vacancies strongly suggests the presence of vacancies in the pristine single-phase materials. Moreover, simultaneous metal-site vacancies and oxygen vacancies would result in even lower density and this possibility was discarded. The calculated compositions with metal-site vacancies have the same the metal atom ratios as obtained from the ICP-OES results. Hence the mass difference between the formula units of the two structures (with and without vacancies) should be the weighted average of cation atomic mass multiplied by the vacancy concentration. This means the density difference should be approximately proportional to the vacancy content, neglecting small variations in unit cell size because all compositions have approximately a 1:1 Li:TM ratio. Figure 5 shows a plot of the difference between the measured density (pycnometer) and the calculated densities that assume no vacancies versus the vacancy content. Figure 5 shows that the expected linearity is observed. If the vacant sites had been occupied by equal amounts of Li and TM (i.e. a 1:1 ratio), then the slope of the plot is expected to be very close to 1.5 g/mL shown by the red line in Figure 5. The measured slope of a best fit line through the data is equal to 1.69 g/mL and is shown as a black line in Figure 5. The measured slope agrees quite well with the “back of the envelope” estimate. Estimating the distribution of metal-site vacancies between the Li and TM layer is also essential to get more insight into the structural features of low-IRC materials. One way to predict the vacancy distribution is to probe the distribution of Li between the two layers. 7Li NMR spectroscopy was used to investigate the local environments around the lithium atoms. It is well known that lithium NMR spectra in this family of materials are dominated by Fermi contact interactions with any paramagnetic metals that are bonded to the same oxygen atom as a given lithium20. The amount of unpaired-electron-spin density that is shared to the lithium site across the intervening oxygen bonds by a particular metal is relatively consistent as the stoichiometry is 15 ACS Paragon Plus Environment


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

Page 16 of 45

changed, making at least partial assignment of the spectra straightforward20. In particular, peaks from Li in the TM layer are shifted strongly to high frequency with respect to those in the lithium layer, because of the different geometry of the bonds and the fact that there is significant clustering of Mn atoms around Li (to balance the Li+1 charge, which differs significantly from the average +3 ionic charge for metals in the TM layer). This effect is most obvious in the 7Li NMR spectrum of Li2MnO3, i.e., Li[Li1/3Mn2/3]O2, shown in Figure 6. Here, the simple structure has two distinct environments: Li in the TM layer fully surrounded by Mn at 1420 ppm, and Li in the lithium layer between two Mn/Li metal-oxide layers at ca. 700 ppm20. 7Li NMR spectra collected from the series of samples studied herein follow this same pattern (see Figure 6 and S1), differing only in that each region of the spectrum is broadened by the larger array of local metal clusters. Similar features have been reported for a wide range of stoichiometries20 using the more time-consuming 6Li NMR spectroscopy, and appear as expected here when the much more rapid 7Li magic-angle turning method14 is employed. The simple two-region basic format of the 7Li NMR spectra allows facile measurement of the ratio of Li atoms in the TM layer to those in the lithium layer (see Table 5), and a straightforward measurement of the locations of the vacancies. Given an ICP-measured formula LipqNixMnyCozO2, there are 1-(x+y+z) sites in the TM layer (and 1-p sites in the Li layer) that can be occupied by either Li or a vacancy. The NMR spectrum reports the percentage of p Li atoms in each layer, and these may be compared to the number of available sites to measure the location

of

the

vacancies.

For

example,

A2

has

an

ICP-measured

formula

Li0.9980.098Ni0.153Mn0.443Co0.309O2, which has 1-(x+y+z) = 0.095 sites in its TM layer that may be occupied by either Li atoms or vacancies. The NMR spectrum shows that 9.9% of the 0.998 Li atoms in the unit formula, i.e. 0.098, are inside the TM layer; see Figure 6 and Table 5. 16 ACS Paragon Plus Environment

Page 17 of 45

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


Clearly, all non-TM atom sites in the TM layer are occupied by lithium atoms, and all vacancies in Sample A2 occur in the Li layer. Similar behaviour is observed in all of the vacancy containing samples, Table 5, in that none show evidence of vacancies in the TM layer. In all samples, the NMR/ICP measured values for Li populations within the TM layer are approximately equal to the calculated number of sites available for Li atoms or vacancies. In some cases, slightly more than the numbers of slots predicted to be available are filled, likely because of some transition metals (presumably Ni to Li exchange) in the Li layer, and small errors in the measurements. The fact that the fully occupied structures have the more highly charged ions in the TM layer suggests that this layer is more nucleophilic and likely explains why vacancies are less favoured in the TM layer than in the Li layer. The density of the low-IRC material is independent of the location (Li layer or TM layer) of the vacancies. Hence, the assignment of vacancies in the Li layer does not mean they were simply substituted for Li atoms. Figure 5 shows that a typical vacancy has a mass that is approximately equal to that of the average cation mass. The NMR results suggest that the vacancies, which could be located on any cation site, actually move predominantly to the sites in the Li layer.

Li atoms in the TM layer have an immense effect on the extent of Li

reintercalation after the first charge, as shown by Arunkumar et al21. The initial (during first cycle) mobility of Li ions in the TM layer and consequences such as subsequent interlayer mixing of TM atoms, irrespective of the oxygen release, could be the origin of an irreversible structural change. This can lead to issues such as large IRC5 and voltage fade22. Jiang et al. have done 6Li MAS spectra studies on Li-excess materials by recovering active materials after the first cycle and showed that a reduced amount of Li could be reinserted back into the TM layers, 17 ACS Paragon Plus Environment


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

Page 18 of 45

contributing to irreversible cpacity23. 7Li NMR results on low-IRC materials suggest that the metal-site vacancies prefer to populate the Li layer. As a consequence, reintercalation of extra Li during first discharge could fill or partially fill the metal-site vacancies in the Li layer and contribute to reduced irreversible capacity. Figure 7 shows a representative set (samples C4, D4, K2 and L2) of first cycle chargedischarge voltage profiles (voltage vs. specific capacity) of cells with electrodes of several single-phase low-IRC materials in the range between 4.8 and 2.0 V. Most of the samples display two distinct regions on their first charge profile: (a) a sloping region corresponding to the oxidation of TM ions and (b) a ~ 4.5 V plateau region corresponding to the electrochemical reactions dictated by oxygen anions (oxygen release24 or oxygen oxidation25) similar to other reported Li-excess materials. The observed first charge capacities of all the samples were between 260 to 280 mAh/g. The charge capacity increases with increasing Li content whereas the sloping region (redox) capacity increases with an increasing proportion of TM redox ions. Table 1 shows the first charge and discharge capacities of the single-phase materials. Figure 8 shows the first cycle differential capacity versus potential profiles of the cells described by Figure 7. Every sample exhibits an oxidation peak at ~ 3.8 V corresponding to Ni and Co oxidation. Figure 8 also shows the difference in the shape of the 4.5 V oxidation peaks between the samples. The intensity of the peaks varies with the involvement of the oxygen anions in the electrochemical reaction, which in turn depends on the available redox TM ions. For example, sample K2 has almost no oxygen-based redox peak implying that nearly all the Li ions were extracted with the aid of redox TM ions similar to traditional layered materials26. Sample K2 has a Li/TM ratio less than 1 and hence it is not a Li-rich material. Hence samples such as K2 are traditional layered materials but notably contain metal-site vacancies. While the 18 ACS Paragon Plus Environment

Page 19 of 45

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


focus of this study is finding high capacity low-IRC material, discovering materials such as K2 is a bonus as they represent another category of positive electrode materials. Thus the low-IRC materials with metal-site vacancies can be categorized into two types. The first (e.g. K2) are similar to traditional layered materials, such as LiNi0.333Mn0.333Co0.333O2, but with metal-site vacancies. The complete characterization of such materials will be the subject of an upcoming article. The second type of materials have Li/TM >1 (e.g. B2), contain metal site vacancies and exhibit high capacity with oxygen-based redox activity such as is observed in a typical Li-rich material. Figure 8 shows that other single phase low-IRC materials exhibit the oxygen-based redox reaction. The redox capability of the TM ions, which primarily depends on the overall composition, and the initial lithium content affect the extent of reversible Li deintercalation and intercalation as will be discussed after the next section on two-phase materials. Composite Low-IRC materials Figure 9 shows the XRD patterns of series H materials in the angular range between 10o to 70o. Figure 8 shows that the Li-deficient members (samples H2 and H3) are two-phase composites with a spinel component. The green arrows in Figure 9 indicate the Bragg peaks corresponding to a spinel phase. Figure 10 shows the first cycle charge-discharge profiles of the H series, in which the low-IRC counterparts are layered-spinel composites. Their first cycle electrochemical behavior is similar to that of any high capacity layered-spinel composite reported elsewhere27. Sample H3 exhibits an IRC as low as 3%, the lowest IRC ever reported for such materials. The reduced IRC in spinel-layered composites may be attributed to the combined effect of spinel and layered components – the ability of the spinel component to intercalate Li into its, initially empty, 16c sites and the presence of metal-site vacancies in the layered component. The vacancy concentration of two-phase materials were calculated as if they were 19 ACS Paragon Plus Environment


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

Page 20 of 45

pure layered LiMO2 materials even though the other phase (e.g. spinel) may contribute to the reduced IRC behavior. The key condition for realizing low-IRC in single-phase LiMO2 type materials is sufficient Li deficiency7. By comparing the electrochemical behavior of all the low-IRC materials discussed in this paper, correlations between materials that show low IRC can be observed. Figure 11a shows the correlation between the IRC and the vacancy concentration. The black and red circles or diamonds in Figure 11 represent the single phase and two phase material respectively. As the amount of metal-site vacancies increase, the IRC is also reduced7. Figure 11b also shows the first discharge capacity of the materials studied. Some of the low-IRC materials exhibit relatively high capacity. IRC is determined by the ability or inability of the Li ions to intercalate back into the host structure after the first charge28. For conventional Li-excess materials Li[LiαM1-α]O2, the relative number of transition metal atoms versus oxygen atoms increases after the irreversible oxygen loss leading to structures such as M1-αO2-j at the end of the first charge. Therefore reaccommodating all the deintercalated Li may not be possible after the first charge and a large irreversible capacity results. By contrast, the low-IRC materials reported here have metal-site vacancies in the pristine state and can compensate for the irreversible oxygen loss. Figure 12 summarizes the data collected in this paper. Figure 12a shows the irreversible capacity plotted versus the “Li excess” in the materials. The Li excess is the difference of the lithium content in column 4 of Table 3 from 1.00.

If the Li excess is greater than 0.00, then

there are certainly Li atoms in the transition metal layer and the irreversible capacity increases.


Page 21 of 45

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


If the Li excess is less than 0.00, then there are probably vacancies in the Li layer, as well as possible vacancies in the TM layer and the IRC decreases. Figure 12a separates the results according to whether samples are single phase (black symbols) or two phase (red symbols). The appearance of the spinel phase when the lithium excess is less than 0.00 is not unexpected.

However, it is important to understand the

compositions that are single phase and also have large negative values of lithium excess (i.e. lithium deficiency). Figures 12b and 12c show the variation of the fraction of IRC and the phase purity versus the lithium excess and the Co content in a formula unit. Li-deficient materials (Li excess less than 0.00) with higher Co content can tolerate a higher concentration of vacancies and remain in a single layered phase. By contrast, Ni-rich materials readily form a co-existing spinel phase when they have the same degree of Li deficiency. For example, the Li contents of samples A2 (Li0.9980.098Ni0.153Mn0.443Co0.309O2) and I2 (Li1.000.094Ni0.266Mn0.547Co0.093O2) are the same but A2 (Co content is 0.309) is single phase while I2 (Co content is 0.093) is two phase. The reason an increased Co content favors a single layered phase material containing metal-site vacancies needs further investigation. Co hinders the interlayer mixing of Li and Ni ions between Li and TM layers in LiMO2 materials29,30. Increased Ni in the Li layer may promote spinel phase formation. Rietveld refinements of the XRD patterns given in Table 2 shows that, in the case of Ni-rich samples, the probability of finding Ni in the Li layer is higher. Spinel formation when the Li/TM ratio is low may be a consequence of increased Ni in the Li layer which can be suppressed by addition of Co. Thus a material with an appropriate amount of Li, Co, Ni, Mn and metal-site vacancies can be prepared as a single layered phase even when cooled slowly.



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

Page 22 of 45

Conclusion A search for low IRC (~ 10 % or low) materials was performed in the Li-Ni-MnCo-O pseudo-quaternary system by synthesizing materials with a variety of Li, Ni, Mn, and Co compositions. Low-IRC materials (~ 10 % or less) can be synthesized at most Ni-Mn-Co combinations by forcing sufficient Li deficiency (Figure 12). The as-prepared materials were characterized structurally using XRD. Most of the low-IRC materials were single-phase layered (Figure 3), contained metal site vacancies (Figures 4, 5 and Table 3) and exhibited an O3 type structure (Figure 3, Table 2). If the Li deficiency was increased beyond a certain limit, depending on the Ni:Mn:Co ratios in the material, a spinel phase appeared in the samples (Figure 8) and the materials which contained a spinel phase component generally had low IRC (Figures 9 and 12a). In general, materials with high cobalt content appeared to tolerate more Li deficiency in the layered single phase and delayed the appearance of spinel (Figure 12c). Among the single phase low-IRC materials, some can be regarded as traditional layered materials without a 4.5 V oxygen loss plateau, but containing metal-site vacancies (e.g. sample K2, Figures 6 and 7). Other single phase low-IRC materials with metal-site vacancies have a 4.5 V oxygen loss plateau that appears in typical Li-rich materials (e.g. samples C4, D4 and L2, Figures 6 and 7). Low irreversible capacity in the single-phase layered materials is well-correlated to the presence of metal-site vacancies as shown by Figure 11. 7Li NMR studies suggest that most of the metal-site vacancies must be in the Li Layer (Figure 6).


Page 23 of 45

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


Acknowledgement The authors thank NSERC and 3M Canada for funding this work under the auspices of the Industrial Research Chair program. Supporting Information 7

Li pj-MATPASS NMR spectra (Figure S1) of all the samples described in Table 5 and the ICP-

OES results (Table S1) performed on select oxide samples after sintering to check for Na and S contamination from the precursor step. Corresponding Author *E-mail: [email protected] Tel.: 001-902-494-2991. Fax: 001-902-494-5191.

Notes The authors declare no competing financial interest



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

Page 24 of 45

References (1) (2)

(3)

(4) (5) (6)

(7)

(8)

(9)

(10)

(11) (12) (13)

(14) (15) (16)

(17)

Goodenough, J. B.; Kim, Y. Challenges for Rechargeable Li Batteries†. Chem. Mater. 2009, 22, 587–603. Zhou, F.; Zhao, X.; Bommel, A. van; Xia, X.; Dahn, J. R. Comparison of Li[Li1∕9Ni1∕3Mn5∕9]O2, Li[Li1∕5Ni1∕5Mn3∕5]O2, LiNi0.5Mn1.5O4, and LiNi2∕3Mn1∕3O2 as High Voltage Positive Electrode Materials. J. Electrochem. Soc. 2011, 158, A187–A191. Thackeray, M. M.; Kang, S.-H.; Johnson, C. S.; Vaughey, J. T.; Benedek, R.; Hackney, S. A. Li2MnO3Stabilized LiMO2 (M = Mn, Ni, Co) Electrodes for Lithium-Ion Batteries. J. Mater. Chem. 2007, 17, 3112–3125. Lu, Z.; MacNeil, D. D.; Dahn, J. R. Layered Cathode Materials Li [NiXLi(1/3−2x/3)Mn(2/3−X/3)]O2 for Lithium-Ion Batteries. Electrochem. Solid-State Lett. 2001, 4, A191–A194. Bommel, A. van; Krause, L. J.; Dahn, J. R. Investigation of the Irreversible Capacity Loss in the Lithium-Rich Oxide Li[Li1/5Ni1/5Mn3/5]O2. J. Electrochem. Soc. 2011, 158, A731–A735. Koga, H.; Croguennec, L.; Ménétrier, M.; Mannessiez, P.; Weill, F.; Delmas, C. Different Oxygen Redox Participation for Bulk and Surface: A Possible Global Explanation for the Cycling Mechanism of Li1.20Mn0.54Co0.13Ni0.13O2. J. Power Sources 2013, 236, 250–258. Shunmugasundaram, R.; Senthil Arumugam, R.; Dahn, J. R. High Capacity Li-Rich Positive Electrode Materials with Reduced First-Cycle Irreversible Capacity Loss. Chem. Mater. 2015, 27, 757–767. Rowe, A. W.; Camardese, J.; McCalla, E.; Dahn, J. R. High Precision Coulometry Studies of SinglePhase Layered Compositions in the Li-Mn-Ni-O System. J. Electrochem. Soc. 2014, 161, A1189– A1193. McCalla, E.; Abakumov, A.; Rousse, G.; Reynaud, M.; Sougrati, M. T.; Budic, B.; Mahmoud, A.; Dominko, R.; Van Tendeloo, G.; Hermann, R. P.; et al. Novel Complex Stacking of Fully-Ordered Transition Metal Layers in Li4FeSbO6 Materials. Chem. Mater. 2015, 27, 1699–1708. McCalla, E.; Rowe, A. W.; Shunmugasundaram, R.; Dahn, J. R. Structural Study of the Li–Mn–Ni Oxide Pseudoternary System of Interest for Positive Electrodes of Li-Ion Batteries. Chem. Mater. 2013, 25, 989–999. McCalla, E.; Rowe, A. W.; Camardese, J.; Dahn, J. R. The Role of Metal Site Vacancies in Promoting Li–Mn–Ni–O Layered Solid Solutions. Chem. Mater. 2013, 25, 2716–2721. McCalla, E.; Carey, G. H.; Dahn, J. R. Lithium Loss Mechanisms during Synthesis of Layered LixNi2 − xO2 for Lithium Ion Batteries. Solid State Ion. 2012, 219, 11–19. Hill, R. J.; Howard, C. J. Peak Shape Variation in Fixed-Wavelength Neutron Powder Diffraction and Its Effect on Structural Parameters Obtained by Rietveld Analysis. J. Appl. Crystallogr. 1985, 18, 173–180. Hung, I.; Zhou, L.; Pourpoint, F.; Grey, C. P.; Gan, Z. Isotropic High Field NMR Spectra of Li-Ion Battery Materials with Anisotropy >1 MHz. J. Am. Chem. Soc. 2012, 134, 1898–1901. Marks, T.; Trussler, S.; Smith, A. J.; Xiong, D.; Dahn, J. R. A Guide to Li-Ion Coin-Cell Electrode Making for Academic Researchers. J. Electrochem. Soc. 2011, 158, A51–A57. van Bommel, A.; Dahn, J. R. Analysis of the Growth Mechanism of Coprecipitated Spherical and Dense Nickel, Manganese, and Cobalt-Containing Hydroxides in the Presence of Aqueous Ammonia. Chem. Mater. 2009, 21, 1500–1503. Genevois, C.; Koga, H.; Croguennec, L.; Ménétrier, M.; Delmas, C.; Weill, F. Insight into the Atomic Structure of Cycled Lithium-Rich Layered Oxide Li1.20Mn0.54Co0.13Ni0.13O2 Using HAADF STEM and Electron Nanodiffraction. J. Phys. Chem. C. 2015, 119, 75–83.


Page 25 of 45

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


(18)

(19) (20) (21)

(22)

(23)

(24)

(25)

(26) (27)

(28) (29)

(30)

Jarvis, K. A.; Deng, Z.; Allard, L. F.; Manthiram, A.; Ferreira, P. J. Atomic Structure of a Lithium-Rich Layered Oxide Material for Lithium-Ion Batteries: Evidence of a Solid Solution. Chem. Mater. 2011, 23, 3614–3621. Lu, Z.; Chen, Z.; Dahn, J. R. Lack of Cation Clustering in Li[NixLi1/3-2x/3Mn2/3-x/3]O2 (0 < X ≤ 1/2) and Li[CrxLi(1-x)/3Mn(2-2x)/3]O2 (0 < X < 1). Chem. Mater. 2003, 15, 3214–3220. Grey, C. P.; Dupré, N. NMR Studies of Cathode Materials for Lithium-Ion Rechargeable Batteries. Chem. Rev. 2004, 104, 4493–4512. Arunkumar, T. A.; Wu, Y.; Manthiram, A. Factors Influencing the Irreversible Oxygen Loss and Reversible Capacity in Layered Li[Li1/3Mn2/3]O2−Li[M]O2 (M = Mn0.5-yNi0.5-yCo2y and Ni1-yCoy) Solid Solutions. Chem. Mater. 2007, 19, 3067–3073. Sathiya, M.; Abakumov, A. M.; Foix, D.; Rousse, G.; Ramesha, K.; Saubanère, M.; Doublet, M. L.; Vezin, H.; Laisa, C. P.; Prakash, A. S.; et al. Origin of Voltage Decay in High-Capacity Layered Oxide Electrodes. Nat. Mater. 2015, 14, 230–238. Jiang, M.; Key, B.; Meng, Y. S.; Grey, C. P. Electrochemical and Structural Study of the Layered, “LiExcess” Lithium-Ion Battery Electrode Material Li[Li1/9Ni1/3Mn5/9]O2. Chem. Mater. 2009, 21, 2733–2745. Armstrong, A. R.; Holzapfel, M.; Novák, P.; Johnson, C. S.; Kang, S.-H.; Thackeray, M. M.; Bruce, P. G. Demonstrating Oxygen Loss and Associated Structural Reorganization in the Lithium Battery Cathode Li[Ni0.2Li0.2Mn0.6]O2. J. Am. Chem. Soc. 2006, 128, 8694–8698. Sathiya, M.; Rousse, G.; Ramesha, K.; Laisa, C. P.; Vezin, H.; Sougrati, M. T.; Doublet, M.-L.; Foix, D.; Gonbeau, D.; Walker, W.; et al. Reversible Anionic Redox Chemistry in High-Capacity LayeredOxide Electrodes. Nat. Mater. 2013, 12, 827–835. Lu, Z.; MacNeil, D. D.; Dahn, J. R. Layered Li [NiXCo1 − 2xMnX]O2 Cathode Materials for Lithium-Ion Batteries. Electrochem. Solid-State Lett. 2001, 4, A200–A203. Cabana, J.; Johnson, C. S.; Yang, X.-Q.; Chung, K.-Y.; Yoon, W.-S.; Kang, S.-H.; Thackeray, M. M.; Grey, C. P. Structural Complexity of Layered-Spinel Composite Electrodes for Li-Ion Batteries. J. Mater. Res. 2010, 25, 1601–1616. Mueller-Neuhaus, J. R.; Dunlap, R. A.; Dahn, J. R. Understanding Irreversible Capacity in Li XNi1 − YFe YO2 Cathode Materials. J. Electrochem. Soc. 2000, 147, 3598–3605. Yabuuchi, N.; Koyama, Y.; Nakayama, N.; Ohzuku, T. Solid-State Chemistry and Electrochemistry of LiCo1∕3Ni1∕3Mn1∕3O2 for Advanced Lithium-Ion Batteries II. Preparation and Characterization. J. Electrochem. Soc. 2005, 152, A1434–A1440. Choi, J.; Manthiram, A. Role of Chemical and Structural Stabilities on the Electrochemical Properties of Layered LiNi1∕3Mn1∕3Co1∕3O2 Cathodes. J. Electrochem. Soc. 2005, 152, A1714–A1718.



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

List of Figure Captions Figure 1: Ni-Mn-Co ternary diagram with every point representing NixMnyCo1-x-y precursor composition (e.g. Ni0.4Mn0.2Co0.4CO3). Black and red points (samples studied here) represent precursor compositions that can be used to make Li-excess materials. Points within the yellow ellipse represent the precursor compositions that were previously reported7. Figure 2: Scanning electron micrograph images of precursors of A1, C1, H1, J1, L1 and O1. Figure 3: XRD patterns of few single phase low-IRC materials (samples C4, D4, K2 and L2) between 20ο to 70ο Figure 4: Calculated densities assuming metal-site vacancies (black) and no vacancies (red) vs measured densities (using a He-pycnometer) Figure 5: Density difference between the calculated (no vacancy case) and the measured densities (from He pycnometer) vs the calculated metal-site vacancy content Figure 6: 7Li pj-MATPASS NMR spectra of Samples A1, A2, and Li2MnO3 collected at 4.7 T under 60 kHz magic-angle sample spinning. All spinning sidebands are co-added to produce spectra equivalent to an infinite-spinning-rate experiment, which therefore contain only isotropic shifts. Figure 7: First cycle charge-discharge voltage vs specific capacity profiles of few single phase low-IRC materials represented in Figure 3 Figure 8: First cycle charge-discharge differential capacity profiles of few single phase lowIRC materials represented in Figure 3


Page 26 of 45

Page 27 of 45

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 9: XRD patterns of samples H1, H2 and H3. Green arrows in H2 and H3 indicate Bragg peaks from the co-existing spinel phase. Figure 10: First cycle charge-discharge voltage vs specific capacity profiles of H series. H1 and H2 are two-phase low-IRC materials. Figure 11. (a) Irreversible capacity vs. vacancy concentration and (b) First discharge capacity vs. vacancy concentration showing single phase (black circles or diamonds) and two phase (red circles or diamonds) samples. Figure 12: (a) Fraction IRC vs. Li excess showing single (black points) and two phase (red points) samples, (b) Co content vs. Li excess (fraction of IRC represented as size of the data points) and (c) Co content vs. Li excess showing whether a sample is single phase (Y) or not (N).



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

Page 28 of 45

List of Tables Table 1. Sample, nominal precursor composition, Calculated [Li] to make one mole of the corresponding stoichiometric LiMO2, Observed [Li] in one mole of synthesized LiMO2 derived from ICP-OES, % Li deficiency, first charge specific capacity, first discharge specific capacity, % IRC relative to the first charge specific capacity and Layered single phase purity of the sample Sample

Nominal Precursor Composition

Calculated [Li] to make one mole of the corresponding stoichiometric LiMO2 (u)

Observed [Li] in one mole of synthesized LiMO2 (v) (ICP-OES)

Li deficiency = [(u-v)/u] (%)

First charge capacity (mAh/g)

First discharge capacity (mAh/g)

Observed IRC (% first cycle charge capacity)

Pure layered single phase? (based on XRD)

A1

Ni0.167Mn0.5Co0.333

1.143

1.1260

1.4787

319

247

22.5

Yes

A2

Ni0.167Mn0.5Co0.333

1.143

1.0492

8.1993

295

272

7.7

Yes

A3

Ni0.167Mn0.5Co0.333

1.143

1.0135

11.3192

285

266

6.5

No

B1

Ni0.2Mn0.5Co0.3

1.130

1.1550

-2.1731

315

246

22

Yes

B2

Ni0.2Mn0.5Co0.3

1.130

1.0538

6.7833

295

267

9.5

Yes

B3

Ni0.2Mn0.5Co0.3

1.130

1.0455

7.5143

280

259

7.5

Yes

C1

Ni0.3Mn0.5Co0.2

1.091

1.0786

1.1259

295

228

22.7%

Yes

C2

Ni0.3Mn0.5Co0.2

1.091

1.0822

0.7969

283

228

19.5%

Yes

C3

Ni0.3Mn0.5Co0.2

1.091

1.0573

3.0805

275

235

14.5%

Yes

C4

Ni0.3Mn0.5Co0.2

1.091

1.0212

6.3885

265

237

10.7%

Yes

C5

Ni0.3Mn0.5Co0.2

1.091

1.0083

7.5702

257

232

9.8%

No

D1

Ni0.333Mn0.5Co0.167

1.077

1.0761

0.0803

276

208

24.5%

Yes

D2

Ni0.333Mn0.5Co0.167

1.077

1.0715

0.5000

274

212

22.4%

Yes

D3

Ni0.333Mn0.5Co0.167

1.077

1.0648

1.1256

273

218

20.3%

Yes

D4

Ni0.333Mn0.5Co0.167

1.077

1.0159

5.6620

270

249

7.6%

Yes

D5

Ni0.333Mn0.5Co0.167

1.077

1.0030

6.8666

258

240

6.9%

No

E1

Ni0.05Mn0.5Co0.45

1.184

1.1552

2.4312

327

206

37.0%

Yes

E2

Ni0.05Mn0.5Co0.45

1.184

1.1031

6.8065

317

241

23.9%

No

F1

Ni0.16Mn0.4Co0.44

1.108

1.0534

4.8532

300

233

22.5%

Yes

F2

Ni0.16Mn0.4Co0.44

1.108

1.0076

8.9953

274

233

15.0%

Yes

F3

Ni0.16Mn0.4Co0.44

1.108

0.9934

10.2710

268

233

13.2%

Yes

G1

Ni0.2Mn0.6Co0.2

1.167

1.1770

-0.8857

332

281

15.5%

Yes

G2

Ni0.2Mn0.6Co0.2

1.167

1.0749

7.8650

253

215

14.9%

No


Page 29 of 45

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


H1

Ni0.4Mn0.5Co0.1

1.046

1.0492

-0.1464

279

235

15.7%

Yes

H2

Ni0.4Mn0.5Co0.1

1.046

0.9657

7.8242

256

243

5.0%

No

H3

Ni0.4Mn0.5Co0.1

1.046

0.9429

10.0007

240

233

3.0%

No

I1

Ni0.3Mn0.6Co0.1

1.13

1.1506

-1.7825

314

240

23.5%

Yes

I2

Ni0.3Mn0.6Co0.1

1.13

1.0491

7.1952

279

259

7.3%

No

J1

Ni0.2Mn0.7Co0.1

1.2

1.1922

0.6530

311

218

30.0%

Yes

J2

Ni0.2Mn0.7Co0.1

1.2

1.0891

9.2434

299

258

13.8%

No

K1

Ni0.3Mn0.4Co0.3

1.048

1.0285

1.8273

290

220

24.0%

Yes

K2

Ni0.3Mn0.4Co0.3

1.048

0.9261

11.5984

243

218

10.4%

Yes

L1

Ni0.2Mn0.4Co0.4

1.091

1.0870

0.3575

308

231

25.0%

Yes

L2

Ni0.2Mn0.4Co0.4

1.091

0.9819

9.9932

263

241

8.5%

Yes

M1

Ni0.1Mn0.4Co0.5

1.13

1.1210

0.8309

320

218

32.0%

Yes

M2

Ni0.1Mn0.4Co0.5

1.13

1.0163

10.0952

274

246

10.1%

Yes

N1

Ni0.2Mn0.3Co0.5

1.048

1.0140

3.2044

280

207

26.2%

Yes

N2

Ni0.2Mn0.3Co0.5

1.048

0.9263

11.5851

242

222

8.2%

Yes

O1

Ni0.1Mn0.3Co0.6

1.091

1.0734

1.6069

308

213

30.9%

Yes

O2

Ni0.1Mn0.3Co0.6

1.091

0.9725

10.8535

252

233

7.6%

Yes



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

Page 30 of 45

Table 2. Rietveld refinement results of samples from the D and K series

Sample

a (Å)

c (Å)

D1

2.861(4)

14.24(8)

D2

2.863(1)

14.25(4)

D3

2.864(3)

14.25(7)

D4

2.865(3)

14.270

K1

2.854(6)

14.22(2)

K2

2.859(6)

14.24(7)

z(O)

n Ni in Li sites

Calculated metal-site vacancy concentration

RB

0.258(0)

0.03(8)

0.000

2.02

0.257(9)

0.03(8)

0.257(7)

0.04(5)

0.252(7)

0.06(3)

0.063

4.05

0.259(1)

0.01(1)

0.014

2.07

0.256(2)

0.03(8)

0.119

2.40


0.003 0.012

1.66 2.01

Page 31 of 45

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


Table 3. Sample, ICP-OES metal ratio, Calculated no. of moles of vacancies in one mole of synthesized LiMO2, Calculated composition of a formula unit with metal-site vacancy, Layered single phase purity, Measured density from a He Pycnometer, Calculated density assuming metal-site vacancies and Calculated density assuming No vacancies for low-IRC materials

ICP-OES Sample Li : Ni : Mn : Co

Calculated no. of moles of vacancies in one mole of synthesized LiMO2 (q)

Calculated composition of a formula unit with metal-site vacancies Lip

qNixMnyCozO2

Pure layered Phase? (based on XRD)

Calculated density assuming No vacancy (g/cm3)

Measured Density (Pycnometer) (g/cm3)

Calculated density based on Metal-site Vacancies (g/cm3)

4.5304 ± 0.0082

4.5063 ± 0.0323

4.6567 ± 0.0340

N/A

N/A

4.5118 ± 0.0165

4.5006 ± 0.0320

4.6279 ± 0.0334

4.5313 ± 0.0252

4.5096 ± 0.0321

4.6491 ± 0.0337

4.5061±0.0320

4.5178±0.0321

4.5348±0.0114

4.5207±0.0323

4.5736±0.0329

4.5313±0.0213

4.5534±0.0329

4.6648±0.0340

A2

1.049:0.161:0.466:0.325

0.098

Li0.9980.098Ni0.153Mn0.443Co0.309O2

Yes

A3

1.013:0.167:0.484:0.336

0.135

Li0.9450.135Ni0.156Mn0.451Co0.313O2

No

B2

1.0538:0.187:0.470:0.289

0.084

Li1.010.084Ni0.179Mn0.45Co0.277O2

Yes

B3

1.046:0.190:0.472:0.292

0.091

Li0.9980.091Ni0.181Mn0.451Co0.279O2

Yes

C2

1.082:0.275:0.455:0.187

0.008

Li1.0780.008Ni0.274Mn0.454Co0.186O2

Yes

C3

1.057:0.282:0.468:0.193

0.035

Li1.0380.035Ni0.277Mn0.46Co0.19O2

Yes

C4

1.021:0.294:0.486:0.198

0.072

Li0.9850.071Ni0.284Mn0.469Co0.191O2

Yes

C5

1.008:0.295:0.495:0.201

0.088

Li0.9640.087Ni0.282Mn0.474Co0.192O2

No

N/A

N/A

N/A

D2

1.072:0.312:0.460:0.157

0.003

Li1.070.003Ni0.311Mn0.459Co0.157O2

Yes

4.5226±0.0305

4.5382±0.0328

4.5419±0.0329

D3

1.065:0.311:0.465:0.158

0.012

Li1.0590.01Ni0.309Mn0.463Co0.157O2

Yes

4.5410±0.0201

4.5344±0.0329

4.5527±0.0330

D4

1.016:0.328:0.489:0.167

0.063

Li0.9840.062Ni0.318Mn0.474Co0.162O2

Yes

4.5426±0.0160

4.5681±0.0336

4.6652±0.0346

D5

1.003:0.330:0.499:0.168

0.079

Li0.9640.078Ni0.317Mn0.479Co0.161O2

No

N/A

N/A

N/A

E2

1.103:0.046:0.440:0.411

0.090

Li1.0530.09Ni0.044Mn0.42Co0.393O2

Yes

4.4282±0.0076

4.4479±0.0339

4.5807±0.0355

F2

1.008:0.163:0.387:0.443

0.099

Li0.9570.1Ni0.155Mn0.368Co0.421O2

Yes

4.6097±0.0071

4.6375±0.0339

4.7963±0.0357

F3

0.993:0.162:0.394:0.450

0.116

Li0.9360.115Ni0.153Mn0.371Co0.424O2

Yes

4.5954±0.0131

4.6367±0.0341

4.8230±0.0362

G2

1.075:0.185:0.550:0.190

0.102

Li1.020.102Ni0.176Mn0.522Co0.18O2

No

N/A

N/A

N/A

H2

0.966:0.395:0.532:0.107

0.098

Li0.9190.098Ni0.376Mn0.506Co0.102O2

No

N/A

N/A

N/A

H3

0.943:0.403:0.545:0.110

0.120

Li0.8860.122Ni0.378Mn0.512Co0.103O2

No

N/A

N/A

N/A

I2

1.049:0.279:0.574:0.098

0.093

Li1.000.094Ni0.266Mn0.547Co0.093O2

No

N/A

N/A

N/A

J2

1.089:0.179:0.639:0.093

0.132

Li1.0170.132Ni0.167Mn0.597Co0.087O2

No

N/A

N/A

N/A

K2

0.926:0.320:0.424:0.330

0.119

Li0.8710.119Ni0.301Mn0.399Co0.31O2

Yes

4.7056±0.0134

4.7168±0.0336

4.9146±0.0357

L2

0.982:0.203:0.402:0.413

0.111

Li0.9280.111Ni0.192Mn0.38Co0.39O2

Yes

4.6984±0.0124

4.6576±0.0335

4.8374±0.0355

M2

1.016:0.099:0.383:0.502

0.118

Li0.9560.119Ni0.093Mn0.36Co0.472O2

Yes

4.5952±0.0112

4.6192±0.0355

4.8078±0.0377


N/A

4.4914±0.0103


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

N2

0.926:0.217:0.310:0.547

0.114

Li0.8730.114Ni0.205Mn0.292Co0.516O2

Yes

O2

0.973:0.103:0.300:0.625

0.118

Li0.9150.119Ni0.097Mn0.282Co0.588O2

Yes

Page 32 of 45

4.8342±0.0053

4.8285±0.0370

5.0230±0.0392

4.7335±0.0062

4.7648±0.0394

4.9626±0.0418

Table 4. Calculated and measured densities of LiCoO2 and Li2MnO3

Sample

Calculated density (g/mL)

Measured density (Helium Pyconometer) (g/mL)

Commercial LiCoO2

5.0576

5.0569 ± 0.0098

Li2MnO3

3.888

3.9137 ± 0.0041


Page 33 of 45

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


Table 5. Combined NMR and ICP derived measurements of lithium populations

Sample

NMRa

NMRa

% of Li

% of Li

in Li

in TM

Layer

Layer

NMR/ICPb

NMR/ICPb

[Li]

[Li]

in Li

in TM

Layer

Layer

per

per

Formula

Formula

Unit

Unit

Formula Unit (NMR/ICP)

74.5

25.5

0.99

0.34

Li[Li1/3Mn2/3]O2

A1

90.4

9.6

1.010

0.107

Li[Li0.118Ni0.146Mn0.426Co0.296]O2

A2

90.1

9.9

0.899

0.099

Li0.8990.098Ni0.003[Li0.099Ni0.150Mn0.443Co0.309]O2

A3Q

90.3

9.7

0.858

0.092

Li0.8580.132Ni0.010[Li0.092Ni0.145Mn0.450Co0.313]O2

B2

90.4

9.6

0.913

0.097

Li0.9130.084 Ni0.003[Li0.097Ni0.177Mn0.450Co0.277]O2

C4

91.7

8.3

0.902

0.082

Li0.9020.072Ni0.026[Li0.082Ni0.258Mn0.469Co0.191]O2

D1

89.9

10.1

0.968

0.109

Li0.968Ni0.032[Li0.109Ni0.277Mn0.459Co0.157]O2

D4

92.0

8.0

0.905

0.079

Li0.9050.063Ni0.032[Li0.079Ni0.286Mn0.474Co0.162]O2

K1

92.3

7.7

0.942

0.079

Li0.9420.014Ni0.044[Li0.079Ni0.250Mn0.378Co0.293]O2

K2

96.5

3.5

0.841

0.030

Li0.8410.118Ni0.041[Li0.030Ni0.261Mn0.399Co0.310]O2

L1

93.3

6.7

1.013

0.073

Li[Li0.0860.002Ni0.180Mn0.359Co0.372]O2

L2

94.4

5.6

0.875

0.052

Li0.8750.111Ni0.015[Li0.052Ni0.177Mn0.379Co0.390]O2

N1

96.3

3.7

0.963

0.037

Li0.9630.028Ni0.009[Li0.037Ni0.187Mn0.280Co0.496]O2

N2

98.7

1.3

0.862

0.011

Li0.8620.114Ni0.024[Li0.011Ni0.180Mn0.292Co0.516]O2

O1

95.7

4.3

1.019

0.046

Li[Li0.0650.015Ni0.092Mn0.268Co0.560]O2

O2

96.6

3.4

0.884

0.031

Li0.8840.116[Li0.0310.003Ni0.097Mn0.282Co0.588]O2

Li[Li1/3Mn2/3]O2 (Li2MnO3)

a

Integrated intensities from the 7Li NMR spectra shown in Figure 6 and S1 measuring the percentage of Li observed in the shift region for the TM or Li layers in these structures; see text for further details. b Separation of the ICP-measured Li atoms per LiMO2 formula unit (Table 4), into those in the Li layer and those in the TM layer, using the NMR-measured percentages shown in the first two columns.



Figures

0 1.333 1 1.31

0.2 1.286 1.259

Co

0.4 1.231 1.2

0.6 1.167 1.13

0.8 1.091

1.2

1.091

1.231 0.8

1.259

1.231

1.167

1.13

1.286

1.2

1.167

1.13

1.091

1.167

1.13

1.091

1.091 0.6

1.048

Mn

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

Page 34 of 45

0.4

1.048

1.048

0.2

1.048

1.048

1

0

0

0.2

0.4

0.6

Ni Figure 1


0.8

1

Page 35 of 45

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


A

C

20 μm

J

H

20 μm

20 μm

L

20 μm

O

20 μm

20 μm

Figure 2



20

30

40

50

60

Intensity (a.u.)

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

Page 36 of 45

70

C4

D4

K2

L2 20

30

40

50

Scattering Angle (deg.) Figure 3


60

70

Page 37 of 45

5.2 N2

Calculated density (g/mL)

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


O2 K2

5 F3 M2 F2

L2

4.8 N2

D4 C4

4.6

J2

E2 I2 C3 D3 D2 G2 C2 D4 C4 D2 D3 C2 C3

4.4

E2 I2 G2

4.2 4.2

O2 K2 F2 F3 M2

L2

Vacancies No vacancies

J2

4.4 4.6 4.8 Measured density (g/mL) Figure 4


5


0.3 Density difference (g/mL)

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

Page 38 of 45

Slope = 1.69 g/mL

0.2

0.1

0

0

0.04 0.08 0.12 Vacancy content

Figure 5


0.16

Page 39 of 45

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 6



4.8 4.4

Voltage (V)

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

Page 40 of 45

4 3.6 3.2

C4 D4 K2 L2

2.8 2.4 2

0

50

100

150

200

Capacity (mAh/g) Figure 7


250

300

Page 41 of 45

900

dQ/dV (mAh/gV)

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


C4 D4 K2 L2

600

300

0

-300

2

2.4

2.8

3.2

3.6

4

Voltage (V) Figure 8


4.4

4.8


H1

Intensity (CPS)

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

Page 42 of 45

H2

H3

35

40

45

50

60

Scatte Scattering angle (deg.)

Figure 9


65

70

Page 43 of 45

4.8 4.4

Voltage (V)

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


4 3.6 3.2 H1 H2 H3

2.8 2.4 2

0

50

100

150

200

Capacity (mAh/g) Figure 10


250

300

Page 44 of 45

300

b 280

260

240

220

200

a 30

IRC %

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

1st Discharge Capacity (mAh/g)


20

10

0

-0.04

0

0.04

0.08

Vacany concentration(q) Figure 11


0.12

0.16

Page 45 of 45

0.4 Fraction IRC

a) 0.3

Single phase Two Phase

0.2 0.1 0

Fraction IRC ?

Co Content

0.6 b) 0.4

0.37

0.2 0.03

0

Single phase?

0.6 c)

Co content

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


Y Y

Y

Y YY Y

0.4

Y

Y 0.2 N N 0 -0.2

-0.1

Y Y Y Y

Y

Y YY N Y N Y YY Y NN Y N 0 0.1

Excess Li Figure 12


Y

Y YY 0.2

A Search for Low-Irreversible Capacity and High-Reversible Capacity

Recommend Documents