Structural Understanding of Superior Battery Properties of Partially Ni-doped Li 2 MnO 3 as Cathode Material

Transcription

1 Structural Understanding of Superior Battery Properties of Partially Ni-doped Li 2 MnO 3 as Cathode Material Toshiyuki Matsunaga, *a Hideyuki Komatsu, a Keiji Shimoda, a Taketoshi Minato, a Masao Yonemura, b Takashi Kamiyama, b Shunsuke Kobayashi, c Takeharu Kato, c Tsukasa Hirayama, c Yuichi Ikuhara, c, d Hajime Arai, a Yoshio Ukyo, a Yoshiharu Uchimoto, e and Zempachi Ogumi a a Office of Society-Academia Collaboration for Innovation, Kyoto University, Gokasho, Uji, Kyoto 611-1, Japan b Neutron Science Laboratory, Institute of Materials Structure Science, High Energy Accelerator Research Organization, 1-1 Oho, Tsukuba, Ibaraki , Japan c Nanostructures Research Laboratory, Japan Fine Ceramics Center, Mutsuno, Atsuta, Nagoya , Japan d Institute of Engineering Innovation, University of Tokyo, Yayoi, Bunkyo-ku, Tokyo , Japan e Graduate School of Human and Environmental Studies, Kyoto University, Yoshida-nihonmatsu-cho, Sakyo-ku, Kyoto , Japan S1

2 Table S-I: Results of quantitative analyses of the Li 2 (Ni x Mn 1-x )O 3(-δ) specimens. The metal sites in these crystals were assumed to have no vacancies. The structure of Li 2 MnO 3 is better described by the Li[Li 1/3 Mn 2/3 ]O 2 formula with respect to its layered structure. Then, as shown in the second and third columns, the compositions of these Li 2 (Ni x Mn 1-x )O 3(-δ) materials can also be expressed using this formula. # targeted composition obtained composition 1 Li 2 MnO 3 Li[Li 1/3 Mn 2/3 ]O 2 Li[Li Mn ]O Li 2 (Mn 19/20 Ni 1/20 )O 3 Li[Li 1/3 Mn 19/30 Ni /1/30 ]O 2 Li[Li Mn Ni ]O Li 2 (Mn 9/10 Ni 1/10 )O 3 Li[Li 1/3 Mn 3/5 Ni /1/15 ]O 2 Li[Li Mn Ni ]O Li 2 (Mn 5/6 Ni 1/6 )O 3 Li[Li 1/3 Mn 5/9 Ni 1/9 ]O 2 Li[Li Mn Ni ]O Li 2 (Mn 3/4 Ni 1/4 )O 3 Li[Li 1/3 Mn 1/2 Ni 1/6 ]O 2 Li[Li Mn Ni ]O S2

3 Li 1/3 Mn 2/ Li 1/3 Mn 1/2 Ni 1/6 Mn Li Ni 0.0 Mn 1/2 Ni 1/2 Fig. S-1. Metal atom present in the TM layer in Li[Li m Mn n Ni 1-m-n ]O 2(-ε) as represented using a ternary diagram. The open squares represent the preparation compositions; on the other hand, the red dots represent the compositions obtained from the quantitative analyses. The star in the graph represents the composition Li[Li 0.2 Mn 0.6 Ni 0.2 ]O 2. 1 The Li 2 (Ni x Mn 1-x )O 3(-δ) (x=0, 1/20, 1/10, 1/6, and 1/4) specimens lie not only away from the Li 2 MnO 3 Li 2 NiO 3 pseudobinary line and but also away from the Li(Li 1/3 Mn 2/3 )O 2 Li(Ni 1/2 Mn 1/2 )O 2 tie line. (found that the Mn in Li 2 (Ni x Mn 1-x )O 3(-δ) (x=1/4) is exactly tetravalent, while the mean valence of Ni is an intermediate value between 2 and 3. This result is to be reported elsewhere in the near future.) (1) 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, S3

4 Fig. S-2: Stereographic view of ideally ordered Li 2 MnO 3 as depicted by the intersection technique (Li: dark green, O: aqua, Mn: pink). The unit cell of this crystal (contoured with thin lines) is composed of cubic-stacked, TM, O, Li, and O layers. These layers are parallel to the a-b plane (base plane) of the cell. The TM layer consists of Li atoms at the 2b site and Mn atoms at the 4g site; the Li layer is composed exclusively Li atoms located at the 2c and 4h sites. S4

5 Table S-II: Structural parameters of Li 2 (Ni x Mn 1-x )O 3(-δ) (x=0, 1/10, 1/6, and 1/4) refined by the Z-Rietveld program using both neutron and synchrotron diffraction data. The standard deviations are shown within the parentheses. The stacking probabilities, α 11, α 12, and α 13, were determined by the FAULTS program using the synchrotron diffraction data only, while using the structural parameters, g, x, y, z, and B obtained from the above-mentioned joint analyses. x=0: a = (1) Å, b = (2) Å, c = 5.3(1) Å, β = (1) ; R wp = 7.90%, R e = 1.09% (for ND+SR), R wp = 4.20%, R p = 3.00%, R B = 9.29%, and R Fobs = 9.87% (for ND), and R wp = 10.46%, R p = 7.46%, R B = 6.45%, and R Fobs = 6.12% (for SR). atom site g x y z B (Å 2 ) Li (1) 2b / Mn (1) 2b / Ni (1) 2b Li (2) 2c / Li (3) 4h (2) 1/2 1.52(1) Li (4) 4g (2) Mn (2) 4g (2) (6) Ni (2) 4g O (1) 4i 0.968(1) (1) (9) 0.511(6) O (2) 8j (7) (4) (5) R =12.1%, χ 2 = α 11 α 12 α (7) (1) x=1/10: a = (2)Å, b = (3)Å, c = (2)Å, β = (4) ; R wp = 4.70%, R e = 1.21% (for ND+SR), R wp = 3.56%, R p = 2.70%, R B = 7.81%, and R Fobs = 8.13% (for ND), and R wp = 5.97%, R p = 4.31%, R B = 4.71%, and R Fobs = 5.47% (for SR). atom site g x y z B (Å 2 ) Li (1) 2b / Mn (1) 2b 0.227(1) 0 1/ Ni(1) 2b 0.127(1) 0 1/ Li (2) 2c / Li (3) 4h (3) 1/2 1.62(1) Li (4) 4g Mn (2) 4g (5) (9) Ni(2) 4g O (1) 4i 0.976(1) (2) (1) 0.602(8) O (2) 8j (1) (6) (8) R =7.46%, χ 2 =41.6. α 11 α 12 α (10) (1) S5

6 x=1/6: a = (1)Å, b = (2)Å, c = (1) Å, β = (4) ; R wp = 4.05%, R e = 1.02% (for ND+SR), R wp = 3.25%, R p = 2.44%, R B = 7.09%, and R Fobs = 7.49% (for ND), and R wp = 5.58%, R p = 4.18%, R B = 4,99%, and R Fobs = 5.53% (for SR). atom site g x y z B (Å 2 ) Li (1) 2b / Mn (1) 2b 0.261(1) 0 1/ Ni(1) 2b 0.174(1) 0 1/ Li (2) 2c / Li (3) 4h (2) 1/2 1.39(1) Li (4) 4g Mn (2) 4g (5) (7) Ni(2) 4g O (1) 4i 0.969(1) (1) (1) 0.537(6) O (2) 8j (1) (4) (7) R =7.63%, χ 2 =41.0. α 11 α 12 α (9) (1) x=1/4: a = (2) Å, b = 8.558(3) Å, c = 5.028(2) Å, β = (4) ; R wp = 4.72%, R e = 1.08% (for ND+SR), R wp = 3.58%, R p = 2.68%, R B = 9.04%, and R Fobs = 9.10% (for ND), and R wp = 6.05%, R p = 4.56%, R B = 5.35%, and R Fobs = 6.32% (for SR). atom site g x y z B (Å 2 ) Li (1) 2b / Mn (1) 2b 0.236(2) 0 1/ Ni(1) 2b 0.259(1) 0 1/ Li (2) 2c / Li (3) 4h (2) 1/2 0.70(2) Li (4) 4g Mn (2) 4g (6) (1) Ni(2) 4g O (1) 4i 0.991(2) (2) (2) 0.52(1) O (2) 8j (2) (6) (9) 0.52 R =7.88%, χ 2 =54.8. α 11 α 12 α (8) (1) S6

7 Intensity (arb. unit) ND Intensity (kcounts) Fig. S-3-1. Observed (+) and calculated (red line) neutron (ND) and synchrotron (SRD) diffraction profiles acquired at room temperature for Li 2 (Ni x Mn 1-x )O 3(-δ) (x=0), whose structure was refined by the joint Rietveld method (q = 4πsinθ/λ). The difference curves (observed minus calculated, blue lines) appear at the top of the figures; under the profiles, the reflection markers are indicated by spikes. The,,,, and peaks were excluded from SRD data for the analysis. SR S7

8 Intensity (arb. unit) ND Intensity (kcounts) Fig. S-3-2. Observed (+) and calculated (red line) neutron (ND) and synchrotron (SRD) diffraction profiles acquired at room temperature for Li 2 (Ni x Mn 1-x )O 3(-δ) (x=1/10). SR S8

9 Intensity (arb. unit) ND Intensity (kcounts) Fig. S-3-3. Observed (+) and calculated (red line) neutron (ND) and synchrotron (SRD) diffraction profiles acquired at room temperature for Li 2 (Ni x Mn 1-x )O 3(-δ) (x=1/6). SR S9

10 Intensity (arb. unit) ND Intensity (kcounts) Fig. S-3-4. Observed (+) and calculated (red line) neutron (ND) and synchrotron (SRD) diffraction profiles acquired at room temperature for Li 2 (Ni x Mn 1-x )O 3(-δ) (x=1/4). SR S10

11 Fig. S-4. HAADF STEM image of the Li2(NixMn1-x)O3(-δ) (x=1/4) compound synthesized at 900 C. The presence of the cation-mixed structures and stacking faults was confirmed from the image of this quaternary compound. HAADF STEM can yield images in which the brightness can be directly correlated to the atomic number Z (it is approximately proportional to the square of Z).1 The bright spots indicate the positions of the Mn/Ni atomic columns in Li2(NixMn1-x)O3(-δ) (x=1/4) viewed along the [100], [], and [] zone axes, which are indicated by the solid, dashed, and dashed-dotted lines, respectively. The coexistence of these three crystal orientations in a single image indicates that this compound definitely contains stacking faults, as also reported previously.2, 3 In addition to the stacking faults, defective contrasts, highlighted in color, were observed in the TM layers in this image. These disordered contrasts can be ascribed to the intense mixing of Li and Mn/Ni. Different contrast disorders can be attributed to different degrees of cation mixing. The neutron and X-ray diffraction analyses allowed us to determine the averaged structure of the ordered layers and the mixing layers. Note that one hardly make out the dot sequences along the TM layers where the contrasts of the Li atomic columns and those of the Mn/Ni atomic columns are identical; this 3 structure is not present in the crystal. strongly indicates that the LiMO2 (1) Pennycook, S. J.; Jesson, D. E. High-resolution Z-contrast imaging of crystals. Ultramicroscopy 1991, 37(1), (2) Boulineau, A.; Croguennec, L.; Delmas, C.; Weill, F. Structure of Li2MnO3 with different degrees of defects. Solid State Ion. 2010, 180, (3) Wang, R.; He, X.; He, L.; Wang, F.; Xiao, R.; Gu, L.; Li, H.; Chen, L. Atomic structure of Li2MnO3 after partial delithiation and re-lithiation. Adv. Energy Mater. 2013, 3(10), S11

12 z = 3/2 z = 5/4 z = 1 z = 3/4 [0, 0] [1/2, -1/6] [1/6, -1/6] z = 1/2 z = 1/4 z = 0 z = -1/4 Fig. S-5-1. Layer stacking structures of Li 2 MnO 3 (Li: dark green, O: aqua, Mn: pink). Slabs, which consist of four atomic layers, can slip by the translation vector, [0, 0], [1/2, -1/6], or [1/6, -1/6], producing stacking faults in the crystal. The thin lines represent the unit cells while the red dots indicate the alignment points of the upper and lower slabs. S12

13 50 (a) 50 (b) 50 (c) Fig. S-5-2. Stacking structures of the upper (z=1) and lower (z=0) TM layers in Li 2 MnO 3 (Li: dark green, O: aqua, Mn: pink; the lower layers are depicted as half tone images). The upper layer is shifted by the translation vector, a: [0, 0, 1], b: [1/2, -1/6, 1], or c: [1/6, -1/6, 1], which produces three kinds of atomic configurations. Notice that, when the 2b and 4f sites on the TM layer are occupied completely at random by Li and TM atoms (see Fig. S-2), the atomic configurations of (a) and (b) coincide with each other. S13

14 Intensity (kcps) Fig. S-6-1. Observed (+) and calculated (red line) X-ray diffraction profiles of Li 2 (Ni x Mn 1-x )O 3(-δ) (x=0) synthesized at 900 C for 12 h on a logarithmic scale. A difference curve (observed minus calculated, blue line) appears at the top of the figure; under the diffraction profiles, the reflection markers are indicated by spikes. The background profile is shown by a green line. Some of the peaks on the low-angle side are labeled with their Miller indices Intensity (kcps) Fig. S-6-2. Observed (+) and calculated (red line) X-ray diffraction profiles of Li 2 (Ni x Mn 1-x )O 3(-δ) (x=1/10) synthesized at 900 C for 12 h on a logarithmic scale. A difference curve (observed minus calculated, blue line) appears at the top of the figure; under the diffraction profiles, the reflection markers are indicated by spikes. The background profile is shown by a green line. Some of the peaks on the low-angle side are labeled with their Miller indices. S14

15 Intensity (kcps) Fig. S-6-3. Observed (+) and calculated (red line) X-ray diffraction profiles of Li 2 (Ni x Mn 1-x )O 3(-δ) (x=1/6) synthesized at 900 C for 12 h on a logarithmic scale. A difference curve (observed minus calculated, blue line) appears at the top of the figure; under the diffraction profiles, the reflection markers are indicated by spikes. The background profile is shown by a green line. Some of the peaks on the low-angle side are labeled with their Miller indices Intensity (kcps) Fig. S-6-4. Observed (+) and calculated (red line) X-ray diffraction profiles of Li 2 (Ni x Mn 1-x )O 3(-δ) (x=1/4) synthesized at 900 C for 12 h on a logarithmic scale. A difference curve (observed minus calculated, blue line) appears at the top of the figure; under the diffraction profiles, the reflection markers are indicated by spikes. The background profile is shown by a green line. Some of the peaks on the low-angle side are labeled with their Miller indices. S15

16 1. additional explanation about replacement Mn with reactive transition metals In Li 2 NiO 3 in which non-reactive Mn 4+ of Li 2 MnO 3 is all replaced with Ni 4+, its capacity is expected to be almost the same to Li 2 MnO 3. As expected, the specific capacity of Li 2 NiO 3 is no more than 240mAh/g 1). While on the other hand, in Li 2 MoO 3 in which Mn is all substituted with reactive Mo, its capacity is expected to be significantly increased compared with Li 2 MnO 3. But actually, the specific capacity of Li 2 MoO 3 is no more than 230mAh/g 2). Furthermore, in Li 2 RuO 3 in which Mn is all exchanged with reactive Ru, its capacity is expected to be considerably increased. But actually, the specific capacity is no more than 270mAh/g 3). These mean that, even if non-active Mn is replaced by active transition metals, the resultant materials do not show much improvement in capacity. Assuming that only the two-electron reaction of Ni 2+ Ni 4+ contributes to increase of the specific capacity, it corresponds to increase of 125mA/g, for example, for Li 2 (Mn 0.75 Ni 0.25 )O 3 ; however, as can be seen in Fig. 1, the capacity increases more than expected. This difference is deemed to be provided by the formation of Li percolation paths as shown in the main text. (1) (2) 3 rd Euro-Mediterranean Conference on Materials and Renewable Energies EMCMRE-3, Marrakech-Morocco, November 2-6, 2015 (3) Yao, Y.; Bie, X.; Wang, C.; Wei, Y.; Chen, G.; Du, F. J. Mater. Chem. A, 2015, 3, S16

17 Fig. S-7. Nearest O atoms (aqua) and second-nearest metal atoms (TM:pink/Li:green) around a Li atom in the TM layer in Li 2 (Ni x Mn 1-x )O 3(-δ) crystals as depicted using the intersection technique; the metal atoms are linked by bars. All the other metal atoms in the crystals are also coordinated by 18 atoms (six oxygen and twelve metal atoms). S17

18 (a) (b) Fig. S-8. Li-Li chains (connected by bars) located (a) in the Li layer and (b) across the Li and TM layers with their obstructing atoms along the chains in the perfectly ordered Li 2 MnO 3 crystal (cross-eyed viewing / Li: dark green, O: aqua, and Mn: pink). These chains can be smooth Li percolation paths in the Li 2 (Ni x Mn 1-x )O 3(-δ) crystal as well, when the Mn 4+ obstructing ions are substituted by Li + or Ni 2+ ions. We presume that higher capacities can be obtained by replacing Mn not only with Ni but also, for example, with Fe, Mo, or Ru, even if it is Ti, than that of Li 2 MnO 3 by the same topological mechanism, as long as these ransition metals provide the random arrangements in the TM layers together with Mn. S18

19 2. Brief comments on the structural evolution and charge compensation during the charge-discharge processes Assuming that the partially delithiated Li 2 MnO 3, Li 2 y (Ni x Mn 1 x )O 3, keeps the C2/m monoclinic structure with delithiation, Li atoms in the lattice are simply substituted by vacancies. This replacement is expected to facilitate the Li diffusion in the lattice, because the vacancy decreases the Coulomb potential on the percolation paths compared with Li +. It is reported that the spinel structures are generated in the charging process of Li 2 MnO 3, 1 which are presumed to decrease the capacity. Therefore, it is very likely that such unnecessary structures are generated also in Li 2 y(ni x Mn 1 x )O 3. Suppressing the spinel formation and stabilizing the original monoclinic structure will be an essential issue to obtain higher capacities and cyclabilities. Very recently, we reported that the Ni and O ions contribute to the charge compensation for the delithiation from a Li 2 y (Ni x Mn 1 x )O 3 electrode sample (x = 1/4). 2 By introducing Ni ions into Li 2 MnO 3 structure, the additional Ni redox reactions can be available, which would contribute to the observed capacities for Li 2 (Ni x Mn 1 x )O 3( δ). We stress that in addition to these redox reaction mechanisms, the disordered arrangement within the TM layer, which fosters the formation of longer and smoother Li percolation path, contributes to the better electrochemical performances. (1) Phillips, P. J.; Bareno, J.; Li, Y.; Abraham, D. P.; Klie, R. F. On the Localized Nature of the Structural Transformations of Li 2 MnO 3 Following Electrochemical Cycling. Adv. Energy Mater. 2015, 5, (2) Shimoda, K.; Minato, T.; Nakanishi, K.; Komatsu, H.; Matsunaga, T.; Tanida, H.; Arai, H.; Ukyo, Y.; Uchimoto, Y.; Ogumi, Z. Oxidation behaviour of lattice oxygen in Li-rich manganese-based layered oxide studied by hard X-ray photoelectron spectroscopy. J. Mater. Chem. A 2016, 4, S19