Re-building Daniell Cell with a Li-Ion exchange Film

Transcription

1 Supplementary Information Re-building Daniell Cell with a Li-Ion exchange Film Xiaoli Dong, Yonggang Wang*, Yongyao Xia Department of Chemistry and Shanghai Key Laboratory of Molecular Catalysis and Innovative Materials, Institute of New Energy, Fudan University, Shanghai , China * Corresponding author. Tel & Fax: address: ygwang@fudan.edu.cn 1

2 Figure S1 Schematic illustration and operating mechanism of conventional Daniell cell in chemistry curricula. 2

3 Figure S2 Self-discharge investigation of conventional Daniell cell with a KCl-based salt bridge. (a) OCV curve and succedent discharge curve. (b) Enlargement of discharge curve after OCV test. (c) Discharge curve without OCV storage. For this investigation, the cathodic electrolyte is 1 ml 2M LiNO 3 solution containing M Cu(NO 3 ) 2 (i.e mg Cu 2+ in the cathodic electrolyte solution), and the anodic electrolyte is 1mL 1 M Zn(NO 3 ) 2 solution. Cu plate and Zn plate were used as cathode and anode, respectively. The cathodic room and anodic room were connected through a KCl-based salt bridge as shown in Figure S1. The cell was kept at the open circuit voltage for 100 hours, and then was discharged with a current of 0.1 ma (Figure S2a). As shown in Figure S2a, the cell voltage keeps at about 0.8 V over the OCV process. However, the cell voltage sharply reduces in the succedent discharge process even at the low current of 0.1 ma (Figure S2a and S2b). The discharge capacity, calculated based on the mass (1.78 mg) of Cu 2+, is just 281 mah g -1 Cu which is much lower than the theoretical capacity (843 mah g -1 Cu ) of Cu 2+ (see Cu e - Cu). Obviously, the capacity loss is owing to the Cu 2+ crossover. In other words, a lot of Cu 2+ ions diffuse from the cathodic room to the anodic room through the salt bridge, 3

4 and then are reduced into metallic Cu on the surface of Zn anode. It also should be noted that the additional discharge plateau at end of discharge should be attributed to the H 2 evolution on Cu electrode at low potential (i.e. 2H + + 2e - H 2 ). Herein, we also carried out a control study as a comparison where a control cell was built in the same way as the one described for Figure S2a, and was discharged without OCV storage at the same current of 0.1mA. Discharge curve of the control cell without OCV storage is given in Figure S2c. It can be observed that the control cell displays a flat discharge voltage of ~ 0.7 V for about 13.6h at the current of 0.1mA. The discharge capacity, calculated based on the mass (1.78 mg) of Cu 2+ -1, is 765 mah g Cu which is close to the theoretical capacity (843 mah g -1 Cu ) of Cu 2+. However, the cell can only deliver a capacity of 281 mah g -1 Cu with an average voltage of about 0.5 V after 100 hours OCV storage (see Figure S2a and 2b). The comparison between the discharge curve with and without OCV test further confirms the serious self-discharge of Daniell cell with a KCl-based salt bridge. 4

5 Figure S3 Charge/discharge cycle of conventional Daniell cell with a KCl-based salt bridge. In this Daniell cell, the Cu-plate in 1mL 2M LiNO 3 solution and the Zn-plate in 1mL 1M ZnNO 3 solution were connected through a KCl-based salt bridge. In this investigation, the cell is charged for 6 hours with a current of 0.25mA to reach a charge capacity of 1.5 mah, and then the battery is discharged at a current of 0.25 ma. As shown in Figure S3, the cell dies only after 3 cycles. 5

6 Figure S4 (a) Schematic illustration of the assembly of the LATSP-based Zn-Cu battery. (b) Photos of the LATSP-based Zn-Cu battery. 6

7 Figure S5 Nyquist plot obtained from EIS measurement of LiNO 3 /LATSP/LiNO 3 electrolyte (a) and Cu(NO 3 ) 2 /LATSP/Cu(NO 3 ) 2 electrolyte (b). In the typical EIS experiment, two 2M LiNO 3 solutions [or two 1M Cu(NO 3 ) 2 ] were separated by the ceramic LATSP film and two Pt-plate electrodes(1 1 cm 2 ) were symmetrically placed into these LiNO 3 [or Cu(NO 3 ) 2 ] solutions. The EIS measurements were performed at the open-circuit voltage of the cell in the frequency range of Hz with an AC signal amplitude of 10mV. As shown in Figure S5, the Nyquist plot typically consists of a high frequency semicircle and a low frequency spike. Typically, the high frequency semicircle is associated with these resistances: the internal resistance of electrode, the resistance of electrolyte and the interface resistance. It can be detected from Figure S5a that the diameter of the semicircle at high frequency is around 170 Ohm, indicating that Li-ion can pass across the LATSP film. However, for Cu(NO 3 ) 2 /LATSP/Cu(NO 3 ) 2 electrolyte, the diameter of the semicircle even reaches the huge value of ~ Ohm. The result from Figure S5b clearly demonstrates that Cu 2+ can not pass across the LATSP thin film. 7

8 ** Further discussion of the phenomenon about the degradation of the charge voltage As shown in Figure 2, there is a clear degradation of the charge voltage curved at the beginning of the charging step. This phenomenon may be ascribed to the formation of Cu 2 O and the growth of dendrite on the surface of cathode. It can be assumed that some Cu atoms at the surface of cathode are electrochemically oxidized into Cu 2 O by following equation: 2Cu + 2OH Cu O + H O + e E O = V (S1) 2 2 [For our experiment, the ph value of the catholyte is about 7 (i.e. concentration of OH - is close to 10-7 ). According to Nernst equation, the theoretical potential of Equation (S1) in the electrolyte with ph value of 7 is about 0.053V, which is lower than the potential for Cu/Cu 2+ conversion (i.e. Cu Cu e - E O =0.34V).] Then, the formed Cu 2 O can be further oxidized into Cu 2+ during the sequent charge, in parallel with the direct Cu/Cu 2+ conversion. Therefore, two plateaus can be detected in the charge curves. Over cycling process, the dendrite growth results in the increase of surface area of electrode, which may increase the electrochemical reaction of equation (S1). Furthermore, the surface variety over cycling process may also result in potential change for dissolution/deposition reaction of metallic electrode (i.e. Zn and Cu electrode). However, it should be noted that the explanation mentioned above is just an assumption. In our future investigation, we will further focus on this point. 8

9 Figure S6 Charge/discharge cycle of Nafion-based Zn-Cu battery. In this battery, the 1cm 2 Cu-plate in 0.25mL 2M LiNO 3 solution and 1cm 2 Zn-plate in 0.25 ml 1M Zn(NO 3 ) 2 solution were separated by a Nafion film. The structure of the battery is similar with that shown in Figure S4. In this investigation, the cell is charged for 6 hours with a current of 0.25mA to reach a charge capacity of 1.5 mah, and then the battery is discharged at a current of 0.25 ma. As shown in Figure S3, the battery dies only after 2 cycles. 9

10 Figure S7 Nyquist plot obtained from EIS measurement of Cu(NO 3 ) 2 /Nafion/Cu(NO 3 ) 2 electrolyte. In the typical EIS experiment, two 1M Cu(NO 3 ) 2 were separated by a Nafion film (117) and two Pt-plate electrodes(1 1 cm 2 ) were symmetrically placed into these Cu(NO 3 ) 2 solutions. The EIS measurement was performed at the open-circuit voltage of the cell in the frequency range of Hz with an AC signal amplitude of 10mV. As shown in Figure S7, the Cu(NO 3 ) 2 /Nafion/Cu(NO 3 ) 2 electrolyte displays an internal resistance of about 120 Ohm without obvious interface resistance, indicating that Cu 2+ can pass across the Nafion film easily. 10