3.2 Effect of Bi Content on Performance of ZnO@C/Bi
(1) Composition and Structure
ZnO@C/Bi composites with different Bi contents were obtained by changing the amount of Bi(NO3)3·5H2O. And ICP results showed that the content of Bi in ZnO@C/Bi composites was 4.14 wt%, 5.66 wt% and 8.01 wt% when the addition amount of Bi(NO3)3·5H2O was 0.1 g, 0.15 g and 0.2 g, respectively.
Figure 2a shows the XRD pattern of ZnO@C/Bi with different Bi contents. The characteristic peaks of ZnO and Bi stand in XRD curves. Meanwhile, it could be observed that the intensity of the characteristic peaks for Bi is enhanced, while that for ZnO is slightly weakened with the boost of Bi content. Additionally, as the local amplification of XRD at 20° to 25° is shown in Fig. 2a, the dispersion peaks at 22.41°, 22.32° and 22.48° are attributed to the amorphous carbon.
The microstructure of ZnO@C/Bi composites with different Bi contents was revealed by TEM (Fig. 2b-d). It could be observed that the distribution of Bi gets more and more uniform in the carbon coating as the Bi content extends from 4.14 wt% to 5.66 wt%. And the carbon embedded with bismuth could also form a perfect coating on the ZnO surface. When the content of Bi(NO3)3·5H2O continues to increase, Bi content in the intermediate product (glu/Bi(NO3)3) augments.40 The excessive bismuth would be crowded in the carbon layer during the heat treatment. Therefore, metal Bi agglomerates and forms balls with larger particle size when Bi content is 8.01 wt%. Although Bi could be fixed in the carbon, the distinctive carbon could hardly wrap ZnO particles (Fig. 2d).
(2) Electrochemical Performance
Corrosion progress of the ZnO@C/Bi composites was analyzed by Tafel plots (Fig. 3a). It is obvious that Bi content shows a significant effect on the inhibition of hydrogen evolution, which the inhibition effect of hydrogen evolution first strengthens and then weakens with the increment of Bi content (Fig. 3b). Bismuth is too little to exert an inhibitory effect when its content possesses 4.14 wt%. Otherwise, it would aggravate the electrode polarization and accelerate the corrosion rate owing to the microcell effect with the active substance in case of excessive bismuth content.41 Therefore, the corrosion current density upshifts to 20.55 mA cm− 2 and 17.22 mA cm− 2, and the corrosion potential downshifts to -1.436 V and − 1.427 V when Bi content is 4.14 wt% and 8.01 wt% (Fig. 3b). While the zinc electrode with the Bi content of 5.66 wt% shows higher corrosion resistance and lower hydrogen evolution corrosion rate (15.13 mA cm− 2 for corrosion current density and − 1.422 V for corrosion potential) (Table S1). It would be conducive to the better electrochemical performance for ZnO@C/Bi composites.
The electrochemical impedance characteristics of ZnO@C/Bi with different Bi contents was also analyzed (Fig. 4a). As the equivalent circuit diagram shows, Rs means the ohmic impedance, Rct represents the charge transfer impedance of the electrode, Zw is Warburg impedance and CPE reflects the capacitance constant phase element of the double electric layer.42 The charge transfer impedance of ZnO@C/Bi is 10.45 Ω, 29.99 Ω and 66.44 Ω when the content of Bi is 4.14 wt%, 5.66 wt% and 8.01 wt% respectively, which Bi is one of the worst conductive metals among all metals increasing the charge transfer impedance and deteriorating the electrode conductivity.43 The Rct is inversely proportional to the Bi content. Moreover, immoderate Bi embedded in the carbon coating impedes the contact between the active substance and electrolyte, which would slow down the electrochemical reaction rate and aggravates the electrode polarization. The result is consistent with the negative shift of corrosion potential in Tafel plots.
Figure 4b shows the CV curves of ZnO@C/Bi with different Bi contents, and the reduction peak positions of the samples are − 1.565 V, -1.548 V and − 1.526 V, respectively, corresponding to Zn(OH)42− being reduced to Zn. The oxidation peak positions appear at -1.308V, -1.285V and − 1.251 V, respectively, indicating that Zn was oxidized to Zn(OH)42−. The redox peak differences are 0.257 V, 0.263 V and 0.275 V (Table S2), expanding with the addition of Bi, revealing the worse and worse electrochemical reversibility of ZnO@C/Bi composites. From the EIS results, it could be seen that the increment of Bi exacerbates the electron transfer impedance of samples and slows down the kinetics of the reaction, leading to the worse electrode reversibility. Moreover, there is an obvious shoulder peak between − 1.3V and − 1.2V when the Bi content is 4.14 wt% and 5.66 wt%, which corresponds to the oxidation of metal Zn to Zn(OH)3− (Eq. (2)). It could be attributed to the fact that the oxidation reaction consumes OH− at a relatively fast rate during the discharge process. In the absence of sufficient OH− supplement, the subsequent reaction will not proceed according to Eq. (1) but according to Eq. (2), thus the second anode peak would appear.27
Zn + 4OH− → Zn(OH)42− + 2e− (1)
Zn + 3OH− → Zn(OH)3− + 2e− (2)
However, the electrode reaction rate slows down due to the increase of electron transfer impedance for ZnO@C/Bi with 8.01 wt% Bi content. The transmission of OH− is sufficient to compensate the consumption of OH−, so there is an absence of second oxidation peak.
The cycling performance and rate performance of ZnO@C/Bi with different Bi content as the active substance of zinc anode are exhibited in Fig. 5a and Fig. 5b, respectively. On the basis of the above analysis, it could be known that the hydrogen evolution corrosion rate of the ZnO@C/Bi composites with Bi content of 4.14 wt% is faster due to the less and non-uniform metal Bi embedded in the carbon coating, which reduces the discharge specific capacity of the battery and aggravates the decay of the cycle life. Therefore, the discharge specific capacity at 1 C decays to 306.8 mAh g− 1 only after 100 cycles when the Bi content is 4.14 wt%. In addition, the average discharge specific capacity is only 423.54 mAh g− 1 at 5 C and restores 414.39 mAh g− 1 when the rate is returned from 5 C to 1 C. However, ZnO@C/Bi with 5.66 wt% Bi content shows excellent cycling performance and rate performance. The average discharge specific capacity reaches 554.08 mAh g− 1 at 1 C. Although the average discharge specific capacity at 5C is merely 432.27 mAh g− 1, it could keep higher level (588.18 mAh g− 1) with the discharge rate declining from 5 C to 1 C. When Bi content continues to increase to 8.01 wt%, its average specific capacity (505.82 mAh g− 1 at 1 C) goes down accordingly and it (440.14 mAh g− 1) keeps the same level as the other two samples under high discharge rate, but it could only restore 442.64 mAh g− 1 with the reduced rate from 5 C to 1 C. Although abundant Bi could restrain hydrogen evolution corrosion, the charge transfer impedance of the electrode gets larger, the electrode polarization is aggravated (Fig. 5c) and the reversibility is worse, resulting in poor cycling performance and rate performance of batteries.
3.3 Effect of Carbon Coating Thickness on Performance of ZnO@C/Bi
(1) Composition and Structure
The carbon coating thickness of ZnO@C/Bi composites was adjusted by changing the glucose addition amount (1.0 g, 1.8 g, 2.5 g and 3.0 g) and named as C-1, C-2, C-3 and C-4 respectively.
The corresponding XRD patterns are shown in Fig. 6a, b. As can be seen from Fig. 6a, the characteristic peaks of ZnO and Bi exhibit in all the four samples. Meanwhile, the intensity of diffraction peaks of ZnO and Bi descends with the growth of glucose addition, because the thickened carbon coating may shield some signals of ZnO and Bi. Furthermore, the dispersion peaks (Fig. 6b) appeared at 22.50°, 22.51°, 22.58° and 22.57° owe the presence of amorphous carbon.
Figure 6c-f are the TEM images of the ZnO@C/Bi composites obtained at 1.0 g, 1.8 g, 2.5 g, and 3.0 g of glucose addition. It could be seen that the carbon coating thickness of C-1, C-2, C-3 and C-4 is 2.28 nm, 5.48 nm, 14.26 nm and 20.38 nm, respectively. Both the coating thickness and the amount of embedded Bi increase with the addition of glucose. For C-1 and C-2, the glucose addition of 1.0 g and 1.8 g, the carbonized coating is too thin to accommodate more and larger metal Bi due to the little addition of glucose (Fig S2). When glucose is continued to be added, the glucose in the precursor (ZnO@glu/Bi(NO3)3) is sufficient to wrap many and large reduction Bi after carbonization, which is reflected in the microstructure of C-3 and C-4 (Fig. 6e, f and Fig Sc, d). Moreover, compared with C-3, the coating of C-4 could embed larger particle size but unevenly distributed Bi due to the thicker carbon coating.
To explore the compactness of the carbon coating, the specific surface area and pore size of the prepared material were analyzed (Fig. 7). Figure 7a shows that the adsorption-desorption isotherm of ZnO is not closed, showing a type Ⅱ profile. At lower relative pressures, its N2 adsorption amount is very small, almost zero, indicating that the specific surface area of ZnO is small and there is no pore structure.44 However, the adsorption and desorption isotherms of C-1, C-2, C-3 and C-4 are mixed type of type Ⅰ and type Ⅳ.45 The characteristic of type Ⅰ is that the adsorption capacity rises rapidly when the relative pressure (P/P0) is lower than 0.1, revealing that the carbon coating has a microporous structure. Moreover, when P/P0 is higher than 0.46, H3-type hysteresis loops appear, indicating that there are a large number of mesopores in the carbon layer. With the thickening of carbon coating of ZnO@C/Bi, nitrogen adsorption increases, indicating that the porosity of the materials improves and the specific surface area enlarges. As can be seen from the pore size distribution diagram in Fig. 7b, the pore size of ZnO@C/Bi composites is mainly composed of micropores and mesopores. With the thickening of the carbon coating, both the total pore volume (Va tot) calculated by single point adsorption method and the micropore volume (Vb micro) calculated by t-plot method rise, and the specific surface area enlarges from 22.97 m2 g− 1 to 103.97 m2 g− 1. The porosity also increases from 36.41–68.10%, revealing that the ZnO@C/Bi composites are mainly microporous (Table S3). However, the pore size gradually decreases with the thickening of carbon coating, due to the limitation of pore volume, adding excess glucose could only adsorb on the outer surface of the material, so it tends to form micropores after carbonization.46
(2) Electrochemical Performance
According to Tafel plots of C-1, C-2, C-3 and C-4 (Fig. 8a), the corrosion characteristic of ZnO@C/Bi with different coating thicknesses was clarified. Obviously, the coating thickness has a great influence on the hydrogen evolution corrosion of the samples. Among all the samples, C-3 with 2.5 g glucose addition carries the best corrosion resistance (14.75 mA cm− 2 for corrosion current density and − 1.421 V for corrosion potential). However, the thinner the coating layer, the lower the embedded Bi content is, resulting in a worse hydrogen evolution corrosion resistance. In contrast, it would also reduce the hydrogen evolution corrosion resistance due to the large specific surface area of active substances and poor homogeneity of embedded Bi. Therefore, the corrosion current density upshifts to 15.20 mA cm− 2, 15.13 mA cm− 2 and 19.05 mA cm− 2 for C-1, C-2 and C-4 (Fig. 8b). Moreover, the charge transfer impedance (Fig. 9a) of the electrode decreases with the thickening of the coating due to the high conductivity of the carbon. The reduction of charge transfer impedance weakens electrode polarization, so the corrosion potential of samples shifts positively (Fig. 8b). For this reason, the corrosion potential is -1.449 V, -1.426 V, and − 1.412 V, respectively (Table S4).
As mentioned above, the charge transfer impedance analyzed from EIS of C-1, C-2, C-3 and C-4 (Fig. 9a), inversely proportional to the coating thickness, is 41.20 Ω, 34.21 Ω, 27.22 Ω and 5.26 Ω, respectively.
Figure 9b shows the CV curves of C-1、C-2、C-3 and C-4. The reduction peak positions are − 1.555 V、-1.561 V、-1.535 V and − 1.528 V, and the oxidation peak positions appear at -1.278 V、-1.262 V、-1.260 V and − 1.231 V, respectively. The redox peak differences are 0.277 V, 0.299 V, 0.275 V and 0.297 V (Table S5), respectively. Combined with the pore size distribution and EIS results, it could be clearly known that both the charge transfer impedance and the average pore size of the coating are reduced when the carbon coating thickens. Besides, the charge transfer rate rises with the decline in charge transfer impedance, and the ion transfer rate reduces owing to the small coating pore size and the long transfer distance in case of the thick carbon coating. Therefore, although the ion transfer rate is high, the excessive low charge transfer rate weakens the reaction kinetics for C-1 and C-2, resulting in poor reversibility. The situation is opposite when C-4 is used as anode, which the charge transfer rate is well but the ion transfer rate is poor. In conclusion, compared with the other samples, the charge transfer rate and ion transfer rate of C-3 as the anode material in the reaction process are moderate, leading to the best reversibility.
Cycling performance and rate performance of Zn-Ni batteries with C-1, C-2, C-3 and C-4 as anodes are shown in Fig. 10a and Fig. 10b. According to the above analysis, it contains less and uneven Bi which could not effectively inhibit hydrogen evolution corrosion when the carbon coating is thin. At the same time, it could also ineffectively alleviate the dissolution of Zn(OH)42− due to large coating pore size, leading to serious electrode deformation and dendrite growth. Therefore, C-1 and C-2 play the poor cycling performance and rate performance. The discharge specific capacity decays at 90 and 150 cycles at low rate (1 C), respectively. And, at 5 C, their average discharge specific capacity remains in a low level (191.20 mAh g− 1 and 273.10 mAh g− 1). Simultaneously, the severe dendrite growth and hydrogen evolution corrosion at high rate greatly damage the structure of active substances, resulting in limited average specific capacity recovery when the rate declines. Moreover, a thicker coating, which has a smaller pore size and contain more Bi, could effectively inhibit hydrogen evolution corrosion and alleviate electrode deformation and dendrite growth. Just like C-3 and C-4, the batteries using them as anode can work stably more than 200 cycles with excellent average specific capacity (634.96 mAh g− 1 and 593.90 mAh g− 1) at 1 C. And the average discharge specific capacity remains at 640.74 mAh g− 1 and 559.24 mAh g− 1 at 5 C. It could also recover higher level (649.70 mAh g− 1 and 586.48 mAh g− 1) with the rate declining from 5 C to 1 C. Nevertheless, C-4 does not perform as well as C-3 due to the limited internal diffusion rate of ions and the deteriorated electrode polarization caused by the thicker carbon coating (Fig. 10c).