Transition-Metal-Doped M-Li8ZrO6 (M = Mn, Fe, Co, Ni, Cu, Ce)
SUPPORTING INFORMATION (DEC. 9, 2015)
Transition-Metal-Doped M-Li8ZrO6 (M = Mn, Fe, Co, Ni, Cu, Ce) as HighSpecific-Capacity
Li-Ion
Battery
Cathode
Materials:
Synthesis,
Electrochemistry, and Quantum Mechanical Characterization
Shuping Huang,†,1 Benjamin E. Wilson,†,2 William H. Smyrl,3 Donald G. Truhlar,*,1 and Andreas Stein*,2 1
Department of Chemistry, Chemical Theory Center, and Supercomputing Institute, University of Minnesota, 207 Pleasant Street S.E., Minneapolis, MN 55455-0431
2
Department of Chemistry, University of Minnesota, 207 Pleasant Street S.E., Minneapolis, MN 55455-0431
3
Department of Chemical Engineering & Materials Science, 421 Washington Avenue S.E., University of Minnesota, Minneapolis, Minnesota 55455-132, USA
†
These authors (S.H. and B.E.W) contributed equally.
*E-mail: [email protected], [email protected]
S-1
Figure S1. Duplicate galvanostatic tests for (a) Mn-LZO-10.8, (b) Fe-LZO-10.8, (c) Co-LZO10.8, (d) Ni-LZO-10.8, (e) Cu-LZO-10.8, and (f) Ce-LZO-10.8 cycled at C/5 (based on composite mass) for 25 cycles. For each composition, materials were prepared from the same MZrO2/C composite, but reacted separately with lithium benzoate to form two different samples.
S-2
Figure S2. Duplicate galvanostatic tests for (a) Fe-LZO-8, (b) Fe-LZO-10, (c) Co-LZO-8, (d) Co-LZO-10, (e) Cu-LZO-8, and (f) Cu-LZO-10 cycled at C/5 (based on composite mass) for 25 cycles. For each composition, materials were prepared from the same M-ZrO2/C composite, but reacted separately with lithium benzoate to form two different samples.
S-3
Table S1. Relevant data from Figures S1 and S2 comparing the duplicate samples tested by galvanostatic cycling. The values for specific capacity difference were calculated using the average percent difference between the last five discharge capacities of the first and second sample.
sample name
number
carbon
electrode mass
specific
(mass%)
(mg)
capacity difference (%)
Mn-LZO-10.8 Fe-LZO-8 Fe-LZO-10 Fe-LZO-10.8 Co-LZO-8 Co-LZO-10 Co-LZO-10.8 Ni-LZO-10.8 Cu-LZO-8 Cu-LZO-10 Cu-LZO-10.8 Ce-LZO-10.8
1
34.9
1.3
2
35.3
1.4
1
66.5
1.3
2
66.6
1.3
1
62.6
1.5
2
61.3
1.6
1
50.0
1.4
2
49.9
1.5
1
40.4
1.4
2
40.0
1.4
1
36.5
1.4
2
36.7
1.5
1
34.5
1.3
2
35.6
1.3
1
57.3
1.1
2
56.9
1.1
1
38.6
1.3
2
39.1
1.4
1
37.3
1.3
2
37.1
1.3
1
34.9
1.5
2
34.8
1.5
1
38.9
1.3
2
39.2
1.4
S-4
-3.2 -6.1 -8.6 -2.8 -6.0 -7.4 -1.7 +3.8 -1.8 -4.1 -4.4 -7.3
Table S2. The gap at the Γ point for Li8ZrO6, the delithiation energy (voltage) for Li8ZrO6 → Li7ZrO6 + Li, and the magnetic moments of oxygen in Li7ZrO6. U for O
0
2
4
6
7
8
9
gap at Γ
5.43
5.50
5.56
5.62
5.67
5.71
5.75
3.70
3.59
3.38
3.11
2.97
2.81
2.66
magnetic
0.093
0.053
0.037
0.029
0.027
0.026
0.026
moments
0.145
0.123
0.072
0.051
0.046
0.043
0.043
of O (µB)
0.096
0.05
0.036
0.028
0.026
0.025
0.025
0.357
0.558
0.669
0.719
0.733
0.745
0.745
0.094
0.05
0.03
0.02
0.018
0.016
0.016
0.057
0.028
0.019
0.015
0.014
0.013
0.013
point (eV) voltage (V)
S-5
Figure S3. PXRD patterns of electrodes prepared with Co-LZO-10, as made, after one charge, one full cycle, after the second charge, and the second full cycle at a rate of C/20 (on the basis of the composite mass). Peaks for the Co-doped Li8ZrO6 and for Li2O are marked. The broad unmarked peaks and sloping background originate from the Kapton tape used to protect the electrodes from the atmosphere during the PXRD experiment, as well as from the carbon component. The peaks corresponding to the Co-Li8ZrO6 phase do not change in position during the delithiation/lithiation steps, confirming that the phase remains intact during cycling. Specific capacities were 217, 178, 187, and 177 mAh/g Co-LZO-10 for the first charge, first discharge, second charge, and second discharge, respectively.
S-6
Figure S4. Galvanostatic charge and discharge curves of a cell composed of 4:6:1 SiO2:Super P:PVDF (by mass), used to determine the background contributions of the Super P carbon component to the capacity of the M-LZO containing cells. Colloidal silica (Ludox AS-40 from Sigma) was first dried at 110 °C until all visible water was removed and the sample appeared dry (approximately 18 h). The silica was then calcined under static air at 400 °C for 2 h. The calcined silica and Super P carbon were mixed in a 2:3 ratio by mass, then ball milled for 10 minutes to form the final composite. An electrode was prepared by preparing a slurry composed of the composite (100 mg), a 5% PVDF solution in NMP (200 mg), and additional NMP (500 mg). The slurry was formed into an electrode film in the same manner as the other M-LZO samples. The cell was cycled at C/5 based on the composite mass.
S-7
Figure S5. Galvanostatic cycling curves of the first cycle of doped Li8ZrO6 at a rate of C/5 (based on composite mass). For all of the compositions, the first cycle shows no clearly discernable inflection points or other features ascribed to a specific electrochemical event other than a general asymptotic behavior nearing the potential limits. This is attributed to various electrochemical reactions that condition the cell prior to further cycling, where inflection points begin to appear in the doped LZO samples (see Figure 5b).
S-8
Table S3. Relative fractions of crystalline components in the M-LZO/C composites.a sample Mn-LZO-10.8
Li8ZrO6 40
mass% Li2O 41
Fe-LZO-8
64
0
36
Fe-LZO-10
95
2
3
Fe-LZO-10.8
82
11
7
Co-LZO-8
85
1
14
Co-LZO-10
75
10
15
Co-LZO-10.8
75
19
6
Ni-LZO-10.8
80
10
10
Cu-LZO-8
87
4
9
Cu-LZO-10
79
8
13
Cu-LZO-10.8
77
17
6
Ce-LZO-10.8
80
14
6
a
Li6Zr2O7 19
Mass% values were calculated by a Rietveld refinement of the PXRD patterns in Figures 4 and 6 of the
main paper. All refinements were done using X’Pert Highscore Plus software with the built-in refinement tools to a goodness-of-fit below 5.5 for all samples. Crystallography files were obtained from the following literature sources: Li2O (Farley, T. W. D.; Hayes, W.; Hull, S.; Hutchings, M. T.; Vrtis, M. Investigation of Thermally Induced Li+ Ion Disorder in Li2O Using Neutron Diffraction. J. Phys. Condensed Mat. 1991, 3, 4761-4781), Li6Zr2O7 (Caekalla, R.; Jeitschko, W. Preparation and Crystal Structure of Li6Zr2O7 and Li6Hf2O7. Chem. Ber. 1993. 619, 2038-2042.), and Li8ZrO6 (Huang, S.; Wilson, B. E.; Wang, B.; Fang, Y.; Buffington, K.; Stein, A.; Truhlar, D. G. Y-doped Li8ZrO6: A Li-Ion Battery Cathode Material with High Capacity. J. Am. Chem. Soc. 2015, 137, 10992-11003). The CIF files were downloaded from the Crystallography Open Database Search (www.crystallography.net).
S-9
Figure S6a. The total density of states of Li96Zr11CoO71 and Li95Zr11CoO71 by PBE+U and HSE06 calculations.
S-10
Figure S6b. The total DOS of Li94CoZr12O72 and Li93CoZr12O72 by HSE06 calculations. The black curves are for the majority spin; by convention, this is the α spin (spin up). The red curves are for the minority spin; by convention, this is the β spin (spin down). These total DOS diagrams are the same as those shown in Figure 8 of the main text but are shown here without the partial DOEs to make a clearer comparison to Figure S6a.
S-11
Figure S7. The partial charge density of the highest occupied majority-spin orbital for Li96Fe2Zr10O71. The isovalue is 0.02 for the yellow isosurface.
S-12
Transition-Metal-Doped M-Li8ZrO6 (M = Mn, Fe, Co, Ni, Cu, Ce) as HighSpecific-Capacity
Li-Ion
Battery
Cathode
Materials:
Synthesis,
Electrochemistry, and Quantum Mechanical Characterization
Shuping Huang,†,1 Benjamin E. Wilson,†,2 William H. Smyrl,3 Donald G. Truhlar,*,1 and Andreas Stein*,2 1
Department of Chemistry, Chemical Theory Center, and Supercomputing Institute, University of Minnesota, 207 Pleasant Street S.E., Minneapolis, MN 55455-0431
2
Department of Chemistry, University of Minnesota, 207 Pleasant Street S.E., Minneapolis, MN 55455-0431
3
Department of Chemical Engineering & Materials Science, 421 Washington Avenue S.E., University of Minnesota, Minneapolis, Minnesota 55455-132, USA
†
These authors (S.H. and B.E.W) contributed equally.
*E-mail: [email protected], [email protected]
S-1
Figure S1. Duplicate galvanostatic tests for (a) Mn-LZO-10.8, (b) Fe-LZO-10.8, (c) Co-LZO10.8, (d) Ni-LZO-10.8, (e) Cu-LZO-10.8, and (f) Ce-LZO-10.8 cycled at C/5 (based on composite mass) for 25 cycles. For each composition, materials were prepared from the same MZrO2/C composite, but reacted separately with lithium benzoate to form two different samples.
S-2
Figure S2. Duplicate galvanostatic tests for (a) Fe-LZO-8, (b) Fe-LZO-10, (c) Co-LZO-8, (d) Co-LZO-10, (e) Cu-LZO-8, and (f) Cu-LZO-10 cycled at C/5 (based on composite mass) for 25 cycles. For each composition, materials were prepared from the same M-ZrO2/C composite, but reacted separately with lithium benzoate to form two different samples.
S-3
Table S1. Relevant data from Figures S1 and S2 comparing the duplicate samples tested by galvanostatic cycling. The values for specific capacity difference were calculated using the average percent difference between the last five discharge capacities of the first and second sample.
sample name
number
carbon
electrode mass
specific
(mass%)
(mg)
capacity difference (%)
Mn-LZO-10.8 Fe-LZO-8 Fe-LZO-10 Fe-LZO-10.8 Co-LZO-8 Co-LZO-10 Co-LZO-10.8 Ni-LZO-10.8 Cu-LZO-8 Cu-LZO-10 Cu-LZO-10.8 Ce-LZO-10.8
1
34.9
1.3
2
35.3
1.4
1
66.5
1.3
2
66.6
1.3
1
62.6
1.5
2
61.3
1.6
1
50.0
1.4
2
49.9
1.5
1
40.4
1.4
2
40.0
1.4
1
36.5
1.4
2
36.7
1.5
1
34.5
1.3
2
35.6
1.3
1
57.3
1.1
2
56.9
1.1
1
38.6
1.3
2
39.1
1.4
1
37.3
1.3
2
37.1
1.3
1
34.9
1.5
2
34.8
1.5
1
38.9
1.3
2
39.2
1.4
S-4
-3.2 -6.1 -8.6 -2.8 -6.0 -7.4 -1.7 +3.8 -1.8 -4.1 -4.4 -7.3
Table S2. The gap at the Γ point for Li8ZrO6, the delithiation energy (voltage) for Li8ZrO6 → Li7ZrO6 + Li, and the magnetic moments of oxygen in Li7ZrO6. U for O
0
2
4
6
7
8
9
gap at Γ
5.43
5.50
5.56
5.62
5.67
5.71
5.75
3.70
3.59
3.38
3.11
2.97
2.81
2.66
magnetic
0.093
0.053
0.037
0.029
0.027
0.026
0.026
moments
0.145
0.123
0.072
0.051
0.046
0.043
0.043
of O (µB)
0.096
0.05
0.036
0.028
0.026
0.025
0.025
0.357
0.558
0.669
0.719
0.733
0.745
0.745
0.094
0.05
0.03
0.02
0.018
0.016
0.016
0.057
0.028
0.019
0.015
0.014
0.013
0.013
point (eV) voltage (V)
S-5
Figure S3. PXRD patterns of electrodes prepared with Co-LZO-10, as made, after one charge, one full cycle, after the second charge, and the second full cycle at a rate of C/20 (on the basis of the composite mass). Peaks for the Co-doped Li8ZrO6 and for Li2O are marked. The broad unmarked peaks and sloping background originate from the Kapton tape used to protect the electrodes from the atmosphere during the PXRD experiment, as well as from the carbon component. The peaks corresponding to the Co-Li8ZrO6 phase do not change in position during the delithiation/lithiation steps, confirming that the phase remains intact during cycling. Specific capacities were 217, 178, 187, and 177 mAh/g Co-LZO-10 for the first charge, first discharge, second charge, and second discharge, respectively.
S-6
Figure S4. Galvanostatic charge and discharge curves of a cell composed of 4:6:1 SiO2:Super P:PVDF (by mass), used to determine the background contributions of the Super P carbon component to the capacity of the M-LZO containing cells. Colloidal silica (Ludox AS-40 from Sigma) was first dried at 110 °C until all visible water was removed and the sample appeared dry (approximately 18 h). The silica was then calcined under static air at 400 °C for 2 h. The calcined silica and Super P carbon were mixed in a 2:3 ratio by mass, then ball milled for 10 minutes to form the final composite. An electrode was prepared by preparing a slurry composed of the composite (100 mg), a 5% PVDF solution in NMP (200 mg), and additional NMP (500 mg). The slurry was formed into an electrode film in the same manner as the other M-LZO samples. The cell was cycled at C/5 based on the composite mass.
S-7
Figure S5. Galvanostatic cycling curves of the first cycle of doped Li8ZrO6 at a rate of C/5 (based on composite mass). For all of the compositions, the first cycle shows no clearly discernable inflection points or other features ascribed to a specific electrochemical event other than a general asymptotic behavior nearing the potential limits. This is attributed to various electrochemical reactions that condition the cell prior to further cycling, where inflection points begin to appear in the doped LZO samples (see Figure 5b).
S-8
Table S3. Relative fractions of crystalline components in the M-LZO/C composites.a sample Mn-LZO-10.8
Li8ZrO6 40
mass% Li2O 41
Fe-LZO-8
64
0
36
Fe-LZO-10
95
2
3
Fe-LZO-10.8
82
11
7
Co-LZO-8
85
1
14
Co-LZO-10
75
10
15
Co-LZO-10.8
75
19
6
Ni-LZO-10.8
80
10
10
Cu-LZO-8
87
4
9
Cu-LZO-10
79
8
13
Cu-LZO-10.8
77
17
6
Ce-LZO-10.8
80
14
6
a
Li6Zr2O7 19
Mass% values were calculated by a Rietveld refinement of the PXRD patterns in Figures 4 and 6 of the
main paper. All refinements were done using X’Pert Highscore Plus software with the built-in refinement tools to a goodness-of-fit below 5.5 for all samples. Crystallography files were obtained from the following literature sources: Li2O (Farley, T. W. D.; Hayes, W.; Hull, S.; Hutchings, M. T.; Vrtis, M. Investigation of Thermally Induced Li+ Ion Disorder in Li2O Using Neutron Diffraction. J. Phys. Condensed Mat. 1991, 3, 4761-4781), Li6Zr2O7 (Caekalla, R.; Jeitschko, W. Preparation and Crystal Structure of Li6Zr2O7 and Li6Hf2O7. Chem. Ber. 1993. 619, 2038-2042.), and Li8ZrO6 (Huang, S.; Wilson, B. E.; Wang, B.; Fang, Y.; Buffington, K.; Stein, A.; Truhlar, D. G. Y-doped Li8ZrO6: A Li-Ion Battery Cathode Material with High Capacity. J. Am. Chem. Soc. 2015, 137, 10992-11003). The CIF files were downloaded from the Crystallography Open Database Search (www.crystallography.net).
S-9
Figure S6a. The total density of states of Li96Zr11CoO71 and Li95Zr11CoO71 by PBE+U and HSE06 calculations.
S-10
Figure S6b. The total DOS of Li94CoZr12O72 and Li93CoZr12O72 by HSE06 calculations. The black curves are for the majority spin; by convention, this is the α spin (spin up). The red curves are for the minority spin; by convention, this is the β spin (spin down). These total DOS diagrams are the same as those shown in Figure 8 of the main text but are shown here without the partial DOEs to make a clearer comparison to Figure S6a.
S-11
Figure S7. The partial charge density of the highest occupied majority-spin orbital for Li96Fe2Zr10O71. The isovalue is 0.02 for the yellow isosurface.
S-12
