Fast and reversible thermoresponsive polymer switching materials for safer batteries

Transcription

1 ARTICLE NUMBER: DOI: /NENERGY Fast and reversible thermoresponsive polymer switching materials for safer batteries Zheng Chen, Po-Chu Hsu, Jeffrey Lopez, Yuzhang Li, John W. F. To, Nan Liu, Chao Wang, Sean Andrews, Jia Liu, Yi Cui and Zhenan Bao Supplementary Figure 1. Low (a) and high (b) magnification SEM images of native spiky Ni particles, which show clear nanoscale extrusions on the particle surface. NATURE ENERGY 1

2 DOI: /NENERGY Supplementary Figure 2. DSC curve of pure LDPE. The melting point is measured to be about 95 C. 2 NATURE ENERGY

3 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Figure 3. A digital photograph of PE/GrNi-based TRPS coated on an Al foil. The film size is about 25 cm*15 cm. NATURE ENERGY 3

4 DOI: /NENERGY Supplementary Figure 4. SEM (a) and TEM (b) images of spherical Ni particles with featureless surface. 4 NATURE ENERGY

5 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Figure 5. DSC plots of PE/GrNi-based TRPS films at different volume ratios of GrNi particle. All different samples have melting points at ~95 C, which is similar to the pure LDPE. NATURE ENERGY 5

6 DOI: /NENERGY Supplementary Figure 6. DSC plots of pure PP and PP/GrNi-based TRPS. All different samples have melting points at ~150 C. This allows PP/GrNi-based TRPS to be operated at higher temperature (> 100 C) than PE/GrNi-based TRPS. 6 NATURE ENERGY

7 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Figure 7. DSC plots of pure PVDF (a) and PVDF/GrNi-based TRPS (b). All different samples have melting points at ~170 C. This allows PVDF/GrNi-based TRPS to be operated at further increased temperature (e.g C, (c)). NATURE ENERGY 7

8 DOI: /NENERGY Supplementary Figure 8. Reversible thermal switching behavior of a TRPS (PE/GrNi, 30 vol% GrNi) film upon heating and cooling over 20 repeating cycles. The heating was performed by blowing hot air with a hot gun set at 157 C and about 2 cm away from the sample surface, which allowed the TRPS film to reach 80 C (measured by an IR gun). The resistance of the film kept stable during repeating operation. 8 NATURE ENERGY

9 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Figure 9. Charge/discharge curves and capacity summary of LiCoO 2 /graphite full cells made from GrNi- (a) and bare Ni-based (b) TRPS current collectors. The designed capacity of both cells is ~ 2.2 mah. The battery with GrNi showed a capacity of 2.1 mah after 3 activation cycles, and the coulombic efficiency reached > 99%. For battery with bare Ni, the initial charge capacity reached ~3 mah, and a large part of the capacity came from voltage range below 4 V, indicating strong oxidation-dissolution of Ni. The battery can only be cycled for 2 times and the later coulombic efficiency was very low because of continuous dissolution of Ni. Some of such batteries cannot even cycle to 4.3 V (a commonly used charge cut-off voltage for LiCoO 2 /graphite cell) at the first cycle. The possible failing reasons could be the continuous oxidation-dissolution of Ni and gradually increased cell internal resistance. NATURE ENERGY 9

10 DOI: /NENERGY Supplementary Figure 10. Cyclic voltammograms of normal (using Al current collector) and safe LiCoO 2 batteries (after 3 initial cycles) at a scan rate of 0.5 mv s -1. Both batteries show similar redox characteristics, further confirming similar electrochemical activity and stability. 10 NATURE ENERGY

11 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Figure 11. Rate capability of normal and safe LiCoO 2 -based batteries (after 5 initial galvanostatic cycles at 0.2 C). Both batteries show similar rate performance. The relatively low capacity of LiCoO 2 in this experiment is due to a moderate quality of such cathode material obtained commercially. NATURE ENERGY 11

12 DOI: /NENERGY Supplementary Figure 12. Cyclic voltammograms of normal (using Cu current collector) and safe graphite-based batteries (after 5 initial cycles) at a scan rate of 0.2 mv s -1. Both batteries show similar redox characteristics, indicating similar electrochemical activity and stability. 12 NATURE ENERGY

13 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Figure 13. Cycling stability of safe (top) and normal (bottom) LiFePO 4 batteries at 25 and 50 C, respectively. The batteries were first cycled at room temperature for 50 cycles and then the environmental temperature was increased to 50 C (controlled by temperature chamber). To show intrinsic battery stability, a commercial available LiFePO 4 (which has better intrinsic cycling stability than LiCoO 2 at high temperature) was used as the active material. The result shows that PE/GrNi (30 v% of GrNi) TRPS allows battery to cycle at good performance in a broad range of temperature. This temperature can be further tuned by changing the composition of TRPS, as discussed in the main text. NATURE ENERGY 13

14 DOI: /NENERGY Supplementary Figure 14. Electrochemical impedance spectra (Nyquist) of normal and safe LiCoO 2 -based batteries after increasing the temperature to 70 C. Both batteries show similar EIS at room temperature (Figure 4e in main text), while the equivalent series resistance (ESR) of normal LiCoO 2 battery decreased to ~15 Ω due to de-lithiation and increased charge transfer at increased temperature. By comparison, the safe battery showed an ESR of ~1600 Ohm due to the turn-on of TRPS electrode. 14 NATURE ENERGY

15 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Figure 15. The second (a) and third (b) shut-down of the same battery as shown in Figure 4g. After the second shut-down, the battery resumed again and continued with the stable cycling. The battery was ramped to high temperature and then subjected to the third shutdown (b). Figure (c) shows magnified region of the shut-down cycles in (b). NATURE ENERGY 15

16 DOI: /NENERGY Supplementary Figure 16. Shut-down of graphite-based safe battery. Figure (a) shows the normal cycling at 25 C and then the shut-down after temperature was increased to 70 C. Figure (b) shows magnified region of the shut-down cycles in (a). 16 NATURE ENERGY

17 DOI: /NENERGY SUPPLEMENTARY INFORMATION Supplementary Table 1. Parameters used for thermal simulation. All the physical properties of the materials listed in the table are adapted from references. 1,2,3,4,5,6 Material Density (kg/m 3 ) Electrical resistivity (W*m) Thermal conductivity (W/(m*K)) Heat capacity (J/(kg*K)) Temperature coefficient of resistance (K -1 ) (T ref = 20⁰C) Ionic resistivity (W*m) Cu foil ** Graphite ** ** 30 electrode Separator ** 30 LiCoO ** ** 30 electrode Al foil ** Stainless steel ABS resin ** ** ** **Not used in simulation Supplementary References 1. Chen, S., Wan, C. & Wang, Y. Thermal analysis of lithium-ion batteries. J. Power Sources 140, (2005) Serway, R., Jewett J. W., Physics for Scientists and Engineers with Modern Physics (9th ed.), 2013, chapter 24, page 814, Brooks/Cole, 20 Channel Center Street, Boston. ISBN 10: NATURE ENERGY 17

18 DOI: /NENERGY Supplementary Movie 1. Demonstration of the fast thermal switching behavior of TRPS devices made by PE/GrNi. A LED is connected to a TRPS film in a circuit and lights up at room temperature. The LED is shut off soon after applying heat with a hot gun on blowing air, which results in a rapidly increased temperature of TRPS. After removal of the hot gun, the TRPS film cools down and LED lights up again. The shut-down response time is less than 1 sec upon applying heat source, confirming an ultra-fast switching behavior. The shut-down and resuming operation can be repeated by many times without obvious change of sensitivity and response time. Supplementary Movie 2. Demonstration of slicing properties of TRPS film. A free-standing PE/GrNi film is hand-cut by a razor blade at a moderate force, which leaves clean and smooth cross-section on both sides of the film. 18 NATURE ENERGY