3.1 Microstructure comparison between α-MnO2 and AMO
3.1.1 XRD characterisation of α-MnO2 and AMO
Figure 3 showed the XRD patterns of α-MnO2 and AMO. As shown in Figure 3-1, the characteristic diffraction peaks of α-MnO2 appeared at 2θ = 28.69°, 37.51° and 56.08°, which corresponded to the (310), (211) and (521) crystal faces of α-MnO2 (PDF#44-0141), respectively. While AMO only had two bun peaks at 2θ = 36.9° and 66.3°, the bun peak around 37° was the characteristic peak of aqueous manganese oxide. This indicated that the sample did not form a long-range ordered crystal structure, had very low crystallinity and would be amorphous manganese dioxide.
3.1.2 TEM characterization of α-MnO2 and AMO
Figure 4 showed the TEM images of the prepared (a) α-MnO2 and (b) AMO. As shown in the figure, the prepared α-MnO2 was a nanorod-like crystal with pores and uniform thickness, while the prepared AMO was a nanoscale spherical product with abundant pores. Through Nanomeasure statistics, it could be determined that the average rod length of α-MnO2 reached 490 nm and the average rod diameter reached 13 nm, while the average particle size of the prepared AMO was 117 nm.
3.1.3 BET characterization of α-MnO2 and AMO
Figure 5 showed the comparison of (a) nitrogen adsorption and desorption curves and (b) pore size distribution of α-MnO2 and AMO, and Table 1 showed the results of BET analysis of α-MnO2 and AMO. As shown in Figure 5, the nitrogen adsorption and desorption curves of the five samples did not completely coincide, and all had H3-type hysteresis lines. And the pore size distribution of both materials was wide, indicating that both materials had porous structure. From Table 3-1, it could be seen that the specific surface area of AMO was 115.69 m2/g, which was 230.06 % higher compared to 34.99 m2/g with α-MnO2. And AMO had larger pore volume and average pore size with narrower pore size distribution, which made AMO materials have more mesoporous pore structures in the 20-50 nm range. This pore structure was more favourable for oxygen transport, thus improving the catalytic properties of the material.
Table 1 BET analysis results of α-MnO2 and AMO
Materials
|
Specific surface area(m2g-1)
|
Pore volume(cm3g-1)
|
Average pore size(nm)
|
α-MnO2
|
34.99
|
0.073
|
8.381
|
AMO
|
115.69
|
0.192
|
6.632
|
3.1.4 Raman characterization of α-MnO2 and AMO
Figure 6 showed the Raman spectra of α-MnO2 and AMO. As shown in the figure, both samples had only one peak between 500-700 nm, which represented the stretching of Mn-O in the [MnO6] octahedron. The Mn-O stretching vibration peak of AMO was at 623.2 cm-1, which was blueshifted compared to 639.5 cm-1 for α-MnO2, suggesting that the structure of AMO was not as stable as α-MnO2. And the AMO material had more oxygen vacancy defects, which led to more catalytically active sites in the AMO cathode. In addition, the Mn-O stretching vibration peaks of AMO were significantly wider and less intense than those of α-MnO2, which indicated that AMO had a low degree of crystallinity, which was consistent with the results in Figure 3-1.
3.1.5 XPS characterization of α-MnO2 and AMO
Figures 7 showed the full XPS spectra, Mn2p spectrograms and O1s spectrograms of α-MnO2 and AMO materials. From the Mn2p spectra in Figure 3-5, it could be seen that the Mn2p peaks of α-MnO2 were centred at 653.88 and 642.18 eV, respectively, with a spin spacing of 11.7 eV. The Mn2p peaks of AMO were centred at 654.08 and 642.18 eV, respectively, with a spin spacing of 11.9 eV, which was slightly larger than that of α-MnO2. The Mn4+ peak area of the AMO material was slightly smaller than that of α-MnO2, suggesting that the structural stability of the AMO material might not be as good as that of α-MnO2. This corresponded to the Raman test above. From the O1s spectrograms, it could be seen that O2-, O vacancies and H2O were present in both materials. O2- was the oxygen ion bound to the metal and H2O was the water adsorbed on the surface of the material.
3.2、Comparison of electrochemical properties of α-MnO2 and AMO
3.2.1 AC impedance (EIS) test of α-MnO2 and AMO
Figure 8 showed the Nyquist curves of α-MnO2 and AMO. As can be seen from the figure, their Nyquist curves could be divided into two parts: the high-frequency region was semicircular, controlled by the kinetics, and formed by the charge-transfer impedance inside the electrodes. The low frequency region was approximated as a straight line and was the Warburg impedance. The AC impedance curves of α-MnO2 and AMO had similar shapes, suggesting that the redox reaction at the solution interface with both catalysts was the same process. The presence of the second arc in the first quadrant indicated that the reaction between the electrolyte and the catalyst was a stepwise process. The process was as follows:
O2+e-→O2-
O2-+3e-+HO2→4OH-
The process of O2 getting an electron to transform into O2- corresponds to the high frequency region in Figure 8. And the process of O2- getting three electrons to convert to OH- corresponds to the low frequency region in Figure 8. From the figure, it could be seen that the AMO material had a larger Warburg impedance and therefore AMO had better catalytic properties.
3.2.2 CV test of α-MnO2 and AMO
Figure 9 showed the CV curves of α-MnO2 and AMO. Both materials showed an oxidation peak around 0.4 V and a reduction peak around -0.5 V. It was more likely that this peak corresponds to redox of the MnO2 itself (as a cation de-insertion reaction). The corresponding reduction peak was likely attributed to some combination of oxygen reduction and MnO2 redox (cation insertion). For further comparison, it could be clearly found that the area enclosed by AMO was larger than the area enclosed by α-MnO2. This indicated that the AMO cathode cell might have a higher discharge specific capacity than α-MnO2 cathode.
3.2.3 Comparison of discharge curves of α-MnO2 and AMO electrodes
Figure 10 showed the comparative discharge curves of zinc-air batteries assembled with air electrodes prepared from α-MnO2 and AMO. From the figure, it could be seen that the open-circuit voltage of the α-MnO2 air electrode was 1.40 V, while the open-circuit voltage of AMO reached 1.63 V, which was significantly increased. The discharge specific capacity of the α-MnO2 air electrode was 517.8 mAh/g. The discharge specific capacity of the AMO air electrode reached 575.2 mAh/g, which was 11.1 % higher than the α-MnO2 air electrode. The analysis was due to the fact that AMO had more catalytically active sites due to more defects (see Figure 6) and therefore had a higher open-circuit voltage. The increase in open-circuit voltage could effectively prolong the discharge time, which led to an increase in its discharge specific capacity.
3.2.4 Comparison of the cycling performance of α-MnO2 and AMO electrodes
Figures 11 showed a comparison of the long cycle performance of zinc air batteries assembled with air electrodes prepared from α-MnO2 and AMO. It could be seen that both material air electrodes were able to perform long cycles. The electrode prepared with α-MnO2 had an initial round trip efficiency of 65.6 %. The initial round-trip efficiency of the AMO-prepared electrode reached 77.4 %, which was 18.0 % higher than that of α-MnO2. The AMO prepared electrodes maintained good round trip efficiency for the first 600 cycles, but the subsequent round trip efficiency decreased dramatically to 30.7 % at 2200 cycles. The analysis might be attributed to the prolonged charging and discharging which resulted in a crystalline transition of the AMO material, leading to a significant decrease in the catalytic stability of the material.
In order to further analysed the reason for the decrease in round trip efficiency of α-MnO2 and AMO materials, the physical phase composition of the electrodes surface after long cycling were tested. Figure 12 showed the XRD patterns of α-MnO2 and AMO electrodes after 2000 cycles. From this figure, it could be seen that ZnCO3, KCO3 and ZnO were present on the surface of both electrodes after long cycles. The presence of ZnCO3 and KCO3 was due to the electrodes involved in the reaction with CO2 from the air. ZnO was formed due to the combination of oxygen in the air with Zn2+. This led to passivation of the material surface and affected the catalytic properties of the material. Mn3O4 was due to the disproportionation of Mn3+ which formed Mn3O4, an electrochemically inert substance. Further analysis showed that diffraction peaks of α-MnO2 existed on the surface of the electrode after 2000 cycles of the AMO electrode. This indicated that some of the AMO had been transformed to α-MnO2 during the long charge/discharge cycles, which led to a significant decrease in its round-trip efficiency after 2000 cycles. This also further proved that the catalytic performance and stability of AMO electrode in zinc-air batteries are higher than that of α-MnO2 electrode.
Figure 13 showed the comparison of the full XPS spectra, Mn2p spectrograms and O1s spectrograms of the α-MnO2 electrode and the AMO electrode after cycling for 2000 cycles. From the full XPS spectra (a, d), it could be seen that the elements K and Zn were also present on the surface of both electrodes after long charging and discharging cycles, indicated the presence of compounds of these two elements. From the Mn2p spectra (b, e), it was obvious that the Mn4+ peak areas of both materials were much reduced compared to the uncharged and discharged electrodes (Figure 7). This indicated that the stability of the material became worse, which was due to passivation and deformation of the material due to prolonged charging and discharging. As can be seen from the O1s spectrograms, O2- and H2O were present in both electrode materials, and O vacancies were absent or present in very small amounts. O2- was the O element in metal oxides and H2O was the water in the air adsorbed by the electrodes. This was also further proved that the abundance of oxygen defects in AMO materials was important to improve the catalytic performance of their electrodes.