3.1 structure of electrocatalysts
TEM analysis of the synthesized electrocatalysts revealed the presence of PANI is also widely distributed on the surface of the carbon support, which is different from the 40Pt/C-JM catalyst. It can be seen from Fig. 1 (b) that the polymer fibers are cross-linked and distributed on GO, while (c), (d) and (e) also have similar polymer fibers, and it can also be found that Pt is distributed on the carrier and PANI. According to the statistics of the size of Pt particles in the picture, it can be seen that the average size of Pt particles in 40Pt/PANI-GO, 40Pt/PANI-CNT, 40Pt/PANI-Cabot and 40Pt/PANI-CB is 2.9nm, which is significantly smaller than that of commercial catalysts 4.62nm. Interestingly, according to the above figure, it can be found that the conductive polymer PANI can be attached to the carbon support with less surface defects and higher graphitization degree, and play a certain dispersive role in the active particles.
In addition, in order to further investigate the Pt particles structure, HR-TEM was utilized to characterize the microstructure. In Fig. 2 (a-e), by measuring the lattice spacing, the Pt (111) planes of 40Pt/C-JM is 0.23nm, while that of the modified catalysts are average only about 0.214nm, which may be attributed to the presence of polyaniline causes compression of the Pt (111) planes and Pt particles strongly interact with the catalyst support.
XRD technology was used to obtain the crystal structural information of the carbon substrate and the corresponding platinum-supported catalyst. As the Fig. 2 (f) shown, the characteristic diffraction peaks of 40Pt/C-JM catalyst appear at 39.80, 46.20 and 67.80, corresponding to the (111), (200) and (220) crystal planes of Pt (PDF#04-0802) respectively, with the Pt (111) crystal face being the most exposed. However, for modification catalysts, the three characteristic diffraction peaks shift to high angles relative to JM-Pt/C, which is due to the reduction of Pt lattice spacing due to the incorporation of N into Pt lattice to form the alloy, and this contributes to the activity and stability of the catalyst ORR[42, 43].
The molecular structure of the conductive composite materials was characterized by Fourier transform infrared (FT-IR). As Fig. 3 (a) shown that the curves of the Pt-loaded catalysts are smoother than that of the unloaded modified carrier. This may indicate that Pt interacts strongly with the functional groups on the modified carrier, thus making the corresponding functional group peaks less obvious. The peaks of wave number from 3000 to 3500 cm− 1, which mainly caused by amino and -OH groups on the surface[44]. For PANI, the absorption peak is concentrated at 3410 and 3110cm− 1, which corresponds to the free N–H stretching vibration of primary (-NH2) and secondary (-NH-) groups[45]. This may account for the –NH stretching frequencies of PANI at 3445 cm− 1 is completely diminished in PANI-Pt, indicating strong interaction which influences the movement of polaron through nitrogen[46].
As shown in Fig. 3 (b), PANI-CB exhibited several absorption peaks corresponding to the vibration absorption peaks of amide groups appear in the absorption vibration (1677 cm− 1), C = C stretching of quinonoid (1,618 cm− 1) and C-N stretching of the secondary aromatic amine (1,300 cm− 1)[47]. Noteworthily, it can be clearly seen that the absorption peak of C-N stretching of the secondary aromatic amine (1,300 cm− 1) becomes very weak with the loading of Pt. Meantime, the peaks of C-O are shifted to a higher wavenumber, which indicates that PANI is dispersed on the CB surface and interacts with each other.
To better understand its role in the resulting cathode, the elemental analysis of these 40Pt/PANI-carbon support catalysts are carried out and compared with 40Pt/C-JM by XPS. Figure 4 (a) is used to analyze the valence state of Pt in catalysts support. Compared with 40Pt/C-JM, the two peaks of Pt (4f7/2 and 4f5/2) in 40Pt/PANI-carbon support catalysts are shifted in the direction of high binding energy, it can be attributed to the fact that the electrons are transferred from the Pt nanoparticles to the PANI-carbon supports, indicating that the Pt electronic structure is changed. In the ORR process, the electron deletion on Pt weakens of O2 adsorption, which may enhance the catalytic activity and stability of the ORR. In addition, according to the Pt binding energy of all catalysts, the Pt 4f peaks were divided into Pt0 4f7/2, Pt2+ 4f7/2, Pt0 4f5/2 and Pt2+ 4f5/2 by peak fitting, respectively. As Table 1 shown, compared with 40Pt/C-JM catalyst, the binding energy of Pt0 4f7/2 on PANI-carbon support are also shifted in the direction of high binding energy. Subsequently, by comparing the content of metallic Pt and oxidation state Pt in Table 1, it is found that the percentage metallic platinum content of the 40Pt/PANI-carbon support catalysts are was higher than that of the 40Pt/C-JM. Metal platinum plays a vital role in the catalytic reaction. Therefore, this result indicated that the surface Pt in 40Pt/PANI-carbon support catalysts have more active sites.
XPS analysis was also conducted to investigate the detailed nitrogen state in these samples, the corresponding N 1s spectra are shown in Fig. 4 (b), respectively. The N 1s peak of these samples contains two main nitrogen species in terms of pyridine-N (398.3 eV), graphitic-N (401.1 eV)[48]. It is widely accepted that the pyridinic N state, which is recognized as an effective active site can absorb the charges and participate in redox reactions providing additional capacitance[49]. As the Fig. 4 (b) shown, the presence of shifts in higher binding energies observed in pyridinic N state. This aspect can be attributed to the synergistic effect between PANI-carbon supports and Pt nanoparticles. However, graphite nitrogen is generally relatively stable and does not participate in any reaction, so the peak position does not shift significantly. According to some recently literatures[50–52], the presence of pyridine-N and graphite-N in the catalyst is beneficial for the ORR performance, because both have lone electron pairs, which can contribute to Pt atoms or can change the electrons in adjacent atoms distributed. Therefore, PANI-carbon support prepared as a catalyst support is advantageous for the ORR reaction.
Table 1
Fitted parameters of XPS data for Pt 4f region of catalysts and N 1S region of modification catalysts and PANI-CB.
Samples | Binding energy of Pt0 4f7/2 (eV) | Pt0 4f7/2/Pt2+ 4f7/2 | Binding energy of pyridine-N (eV) |
PANI-CB | - | - | 398.68 |
40Pt/PANI-GO | 70.58 | 2.088 | 399.18 |
40Pt/PANI-CNT | 70.68 | 3.118 | 399.38 |
40Pt/PANI-Cabot | 70.58 | 4.979 | 400.28 |
40Pt/PANI-CB | 70.68 | 0.866 | 399.08 |
40Pt/C-JM | 70.38 | 0.855 | - |
3.2 performance of electrocatalysts
The electrocatalytic ORR performance of the as-prepared 40Pt/PANI-carbon support catalysts and reference 40Pt/C-JM catalysts were evaluated by cyclic voltammetry (CV) and linear sweep voltammetry (LSV) in 0.1M HClO4 solution at room temperature of ca. 25 oC. The CV curves of all catalysts recorded in O2-saturated electrolyte at a scan rate of 10 mVs− 1, as shown in Fig. 5 (a). It is evident that the charge for H2 adsorption/desorption peaks (0.05–0.4 V) of 40Pt/PANI-carbon support catalysts are larger than that of 40Pt/C-JM. In addition, the oxide reduction peaks (0.7–0.9 V) for the 40Pt/PANI-GO and 40Pt/PANI-Cabot have shifts positively by 20 mV compared to that of 40Pt/C-JM, and the oxide reduction peaks of other 40Pt/PANI-carbon support catalysts also have shifts positively compared with that of 40Pt/C-JM, suggesting weaker affinity of the OH chemisorption energy on the surface of 40Pt/PANI-carbon support catalysts. Because of adsorbed OH species have negative impact on ORR, the weaker affinity of the OH chemisorption energy of 40Pt/PANI-carbon support catalysts are beneficial to catalytic activity.
As Table 2 shown, the electrochemical surface areas (ECSA) of 40Pt/PANI-GO, CNT, Cabot and CB catalysts are calculated as 194.4, 71.6, 92.2 and 42.3 m2gPt − 1, which are larger than that of 40Pt/C-JM one (39.8 m2gPt − 1). Especially 40Pt/PANI-GO has such large ECSA could be directly correlated to the greater number of available catalytic sites. This result suggests that more active sites exist in 40Pt/PANI-carbon support catalysts, which can be attributed primarily to Polyaniline dispersed on carbon support has the ability to disperse and anchor platinum particles. In addition, the limiting current density and half-wave potential (E1/2) of 40Pt/PANI-carbon support catalysts are greater than that of 40Pt/C-JM, and almost all the onset-potential of 40Pt/PANI-carbon support catalysts are more positive than that of 40Pt/C-JM catalysts. It reveals the 40Pt/PANI-carbon support catalysts exhibit superior catalytic activity due to the effective dispersion of Pt particles by PANI, resulting in more active sites, and the pyridinic N configuration with the lowest overpotential in ORR.
Table 2
Electrochemical results of 40Pt/C-JM 40Pt/PANI-GO, 40Pt/PANI-CNT, 40Pt/PANI-Cabot and 40Pt/PANI-CB
Samples | ECSA(m2/g) | E1/2(V) | Eonset(V) | [email protected](A/mg) |
40Pt/C-JM | 39.8 | 0.834 | 0.951 | 0.058 |
40Pt/PANI-GO | 194.4 | 0.865 | 0.982 | 0.142 |
40Pt/PANI-CNT | 71.6 | 0.849 | 0.983 | 0.103 |
40Pt/PANI-Cabot | 92.2 | 0.873 | 0.981 | 0.123 |
40Pt/PANI-CB | 42.3 | 0.841 | 0.951 | 0.064 |
In addition to the activity of catalysts, the stability of the electrochemical catalyst is also a key factor to evaluate the performance of the catalysts. The stability of the catalyst was evaluated by accelerated stability test (ADT). Figure 6 shows the CV curves of 40Pt/C-JM and 40Pt/PANI-GO before and after different potential cycles durability test. It is noted that the OH adsorption peaks of 40Pt/C-JM shifted positively by 32 mV after 2000 potential cycles, but after 3000 potential cycles, the OH adsorption peaks shifted another 28 mV negatively. And the integrated area of hydrogen adsorption between 0.05 and 0.4 V is sharply reduced compared with original CV curve. This may imply that the structure of the entire catalyst has changed in an unfavorable direction. While, there is no significant change in the integrated area of hydrogen adsorption for 40Pt/PANI-GO, indicating no obviously loss of ECSA, and the OH adsorption peaks shifted positively by 11 mV after 5000 potential cycles. This is probably due to metal nanoparticles were stably anchored in PANI-carbon support, which hinder the nanoparticles migration and aggregation and PANI covers the surface defects of the carbon support, and because the conductive polymer PANI connects the isolated catalytic particles, the local voltage of the catalytic particles is more average.
It is well known that carbon nanotubes are one dimensional carbon materials, graphene is two dimensional carbon materials, Conductive black and Cabot carbon are three dimensional materials[53]. The catalysts made after those carbon materials were combined with PANI, and the experiment in this research can be simplified into the mechanism as shown in the Fig. 7. The main results show that PANI can be stabilized on the surface of carbon supports, including defects of carbon supports, and can play a role in dispersion and anchoring of platinum particles, ensuring enough active sites, so as to have excellent activity and stability. Meanwhile, PANI can avoid the occurrence of carbon corrosion on catalysts, then ensure the original appearance of the catalyst, and ensure the long life of the catalysts.