As shown in Fig. 1a, X-ray diffraction (XRD) revealed that we have successfully synthesized polyhedrons of transition metal organic framework materials. When increasing the carbonization temperatures, the diffraction peaks at 44.216°, 51.522° and 75.853° become more distinct, which matches well with the (111), (200) and (220) planes of the cubic Cobalt (PDF#15-0806). As is known to all, the annealing temperature has significant effect on the physicochemical and electrochemical performance of the samples [28, 29]. Thus, the obtained samples with a series of temperature gradient were conducted electrochemical measurements to examine the optimized temperature. Fig. 1b shows the electrochemical activities of the materials treated at different temperatures. It is obvious to find that the as-prepared catalyst carbonized at 700 °C (Co/MOFs-700) exhibits the best OER performance. The overpotential is around 480 mV to achieve the current density of 10 mA cm-2 in 0.1 M NaOH.
Then the Co/MOFs before carbonized and the best performed Co/MOFs-700 were picked to carry out the SEM measurements. As shown in Fig. 1c and 1d, the morphology of the obtained Co/MOFs-700 has changed greatly after carbonized at 700 °C. Many fold-like lines appear on its surface and is not smoother than the original materials without carbonized. But it still processes the polyhedron morphology, with regular particle dispersion and no collapse sign.
As previous articles reported, doping the phosphor into transition metal organic framework polyhedron can increase the stability of the sample in acid or alkaline solution and can also effectively improve the electrochemical catalytic activity by break the electroneutrality and facilitate of the O2 adsorption to enhance the catalytic activity [30-32]. Therefore, phosphorus-doped samples are synthesized with in-situ doping method and investigated the electrochemical performance. The obtained products were named as Co/P0/MOFs-700, Co/P1/MOFs-700 and Co/P2/MOFs-700, while the P0, P1, P2 represents the phosphorus sources of sodium hypophosphite, triphenylphosphine and o-trimethylphenyl phosphine, respectively.
According to Fig. 2a, the diffraction peaks of the phosphorus-doped samples still possess the pattern of cubic Cobalt (PDF#15-0806), indicating that doping slight amount of phosphor would not change the structure of the MOFs. Then the electrochemical measurements were conducted to explore the influence of different phosphorus sources on electrochemical catalytic activities. As shown in Fig. 2c, the open potential (0.87 V) and half wave potential (0.78 V) both shows that the Co/P1/MOFs-700 possesses the best ORR activity. However, it is slightly weaker than that of the original product Co/MOFs-700 carbonized under the same temperature. Fig. 2d represents the OER performance of different products doped with phosphorus. When the limited current density is 10 mA cm-2, only Co/P1/MOFs-700 owns the lowest over potential of 430 mV, demonstrating that incorporation of phosphorus into the samples can increase the OER activity, which is coincidence with the reported article that incorporation of phosphorus would adjust the electrical conductivity and meanwhile facilitate the rapid electrons transfer[33]. Moreover, Fig. 2b shows a comparison between the sample with triphenylphosphine as phosphorus source and the original sample (Fig 1c) without incorporation of elements. It can be revealed that the incorporation of phosphorus greatly affected the morphology of the material compared with Co/MOFs-700. Therefore, doping phosphor can not only enhance the electrochemical activity but also changed the morphology of the sample.
Subsequently, in order to further figure out the reason why doping phosphorus can enhance the electrochemical activity. XPS analysis were carried out to probe the composition and chemical state of the Co/MOFs-700 and Co/P1/MOFs-700 sample. According to Fig. 3a, XPS spectra survey of Co/MOFs-700 and Co/P1/MOFs-700 both shows the presence of Co 2p, O 1s, N 1s and C 1s. It is noted that the peak of P 2p appears in the XPS spectra survey in Co/P1/MOFs-700 but shows a rather weak signal compared with strong peaks of C 1s. Moreover, Fig. 3b shows the Co 2p spectra of Co/MOFs-700 and Co/P1/MOFs-700. It was found that the Co 2p 3/2 can be fitted into two peaks. The peaks located at 778.2° and 780.7° can be ascribed to the Co (0) and Co (2+). While Co 2p 1/2 can also be displayed into two peaks positioned at 793.3° and 796.7°, which can be ascribed to the Co (0) and Co (2+). The satellites peaks were positioned at 786.2° and 802.7° [34-36]. When compared with phosphorus-doped sample Co/P1/MOFs-700, we can find that the Co (0) were greatly increased while Co (2+) decreased, indicating that doping phosphorus source during the synthesis process can increase the content of Co (0) in the obtained samples. As is known to us all, Co (0) can greatly enhance the conductivity, thus improve the electrochemical performances which also in accordance with the previous report [37].
Afterwards, we continued to investigate influence on the quality of the doped phosphorus source. The obtained products with a different molar ratio of P were named as Co/P/MOFs-700-x (x = 0.25, 0.5, 0.75, 1.0), while P represents triphenylphosphine and x represents the quality of phosphorus source. Fig. 4a shows that when increasing the content of phosphorus sources, the XRD pattern shows that the main diffraction peaks in these samples are still Cobalt (PDF#15-0806). As Fig. 4b shows, Co/P/MOFs-700-0.5 possesses the best ORR activity whose half wave potential was around 0.8 V among these phosphorus-doped products, but the ORR activity of Co/P/MOFs-700-0.5 is not increased significantly compared with the original sample Co/MOFs-700. It can be seen from Fig. 4c that the OER activity of the samples increased significantly with the addition of triphenylphosphine compounds and decreased with the increase mass of phosphorus source. When the limited current density is 10 mA cm-2, Co/P/MOFs-700-0.25 and Co/P/MOFs-700-0.5 both own the minimum overpotential of 450 mV, indicating that only proper amount of phosphorus sources can improve OER activity while the amount of 0.25 and 0.5 exhibit best. However, when compared with commercial platinum carbon (half-wave potential 0.81 V, limiting current density 5.43 mA cm-2) and excellent OER electrocatalysts Iridium oxide (1.61V @ 10 mA cm-2), Co/P/MOFs-700-0.5 still remains significant difference among the limited current density in ORR performance. As the article reported, when the conductivity of the material is small, so is the limited current density [38].
In order to enhance the conductivity, we firstly measured the current carbon content of the synthesized Co/P/MOFs-700-0.5 analyzed by EDS images. According to Fig. 5, it is obvious that the quality of cobalt accounts for the most which takes almost 52.38%, while the quality of carbon is relatively less of 29.13%.
Therefore, in order to improve the conductivity of the material, we further doped the samples with carbon without any phosphorus sources. The obtained products were named as Co/MOFs-CNTs-700, Co/MOFs-CB-700 and Co/MOFs-A-OMCS-700, respectively. Fig. 6a shows that doping carbon will not affect the structure of the samples, which still keeps the same diffraction peaks of Cobalt (PDF#15-0806). As shown in Fig. 6b, it can be seen that the limited current density of the products are greatly increased with the incorporation of carbon source in ORR, while Fig. 6c indicated that the incorporation of carbon source makes no sense to improve the OER properties of the catalysts.
Combined with the previous experimental data and conclusions, we doped the original sample with both phosphorus and carbon elements by adding 0.5 g of triphenylphosphine and a suitable amount of different carbon sources (CNTs, CB and A-OMCS) to the material for comparison. The obtained samples were named as Co/P/MOFs-CNTs-700, Co/P/MOFs-CB-700 and Co/P/MOFs-A-OMCS-700, respectively. According to Fig. 7a, there is no change of the XRD pattern, with all the samples matching well with cubic Cobalt (PDF#15-0806). As shown in Fig. 7b, co-doping with phosphorus and carbon greatly increased the limited current density and ORR performance of the products. The sample of Co/P/MOFs-CNTs-700 exhibits the best ORR activity, which the half wave potential and limiting current density are 0.8V and 4.81 mA cm-2 and is 10 mV lower than that of commercial platinum carbon. Additionally, as can be seen clearly in Fig. 7c, the OER performance of the products have also been greatly improved. The sample of Co/P/MOFs-CNTs-700 exhibits the lowest over potential voltage of 420 mV. Compared with the voltage corresponding to dioxide iridium, Co/P/MOFs-CNTs-700 is only about 40 mV higher than dioxide iridium. Therefore, Co/P/MOFs-CNTs-700 exhibits to be a favorable bifunctional electrocatalyst.
Meanwhile, to access the stability of the best performed Co/P/MOFs-CNTs-700, chronopotentiometry and chronoamperometric responses tests were carried out. As can be seen in Fig. 8a and 8b, the overpotential only increased 1.5 mV and the ORR performance are reduced by 79.5% after 18 hours continuous tests, proving that both OER and ORR activity of Co/P/MOFs-CNTs-700 are rather stable in 0.1 M NaOH.
Scanning electron microscopy, EDS and mapping on the sample of Co/P/MOFs-CNTs-700 have also been carried out. As can be seen from Fig. 9a-c, Co/P/MOFs-CNTs-700 retained the polyhedron morphology with many fold-like lines on the surface. Besides, the incorporation of carbon nanotubes is embedded into the skeleton of the product, which may increase the specific surface area of the product and provides more adsorption sites for electrochemical reaction. Fig.10d-g are the mapping analysis of the sample. It can be seen that the carbon and phosphorus sources are uniformly dispersed in the skeleton of the sample and become a whole.
As EDS shows, the content of phosphorus and carbon of the material is increased compared with the original sample of Co/MOFs-700 by in-situ doping, thus leading to the increased ORR and OER activity. It is laterally demonstrated that the incorporation of two kinds of phosphorus and carbon elements could be beneficial to increase the electrochemical activity of metal organic framework materials containing cobalt [39]. Because the electronegativity of P (2.19) is different from that of carbon atoms (C: 2.55). Co-doping would break the electroneutrality which can facilitate of the O2 adsorption and improve the ORR activity [40]. Meanwhile, more active sites can arise due to the co-doping phosphorus and carbon by changing the asymmetric spin density of heteroatoms and effectively weak the O-O bonding, thus leading to the enhanced ORR activity [41].
The outstanding electrochemical activities can be attributed to the following reasons. Firstly, doping hetero atoms would lead to the redistribution of the charge density on the catalyst surface, which is beneficial to adsorb oxygen and promote the ORR activities [42]. Secondly, codoping different atoms into the MOFs would result in the synergistic effect which also contributes to the enhanced electeochemical performance [43]. Thirdly, it has been proved that the OER mechenism of Co-based catalyst is a dynamic surface self-reconstruction process. The Co atoms on the surface could form a self-assembled metal oxy(hydroxide) active layer of CoOOH which works as real active site [44]. In addition to the composition, the unique hybrid structure combined with its high conductivity could provide large surface area for the fast charge transfer.