Direct urea/hydrogen peroxide fuel cells (DUHP-FCs) can produce electrical energy by recycling urea-rich wastewater. This study expands the commerciality of DUHP-FC by removing precious metals from their design. Nickel nanorod/nickel foam (NNR/NF) was fabricated using hydrothermal treatment to be used as the anode, and Prussian blue coating was deposited by potentiostatic electrodeposition onto hydrophilic carbon felt at the cathode (PB/CF). The anode exhibited a 7-folds higher current density than bare NF at 0–2 M urea, and lower charge transfer resistance. The cathode reported a high H2O2 reduction current. In addition, fuel cell tests indicated current density dependency on H2O2 concentration and cell voltage dependency on KCl concentration. A competitive maximum power density of 10.6 mW cm−2 was achieved at 0.98 open circuit voltage and 45 mA cm−2 maximum current density, in 0.33 M urea vs 2 M KCl and 2 M H2O2, exclusively via diffusive mass transfer. These findings indicate the practical application of DUHP-FC on a large scale.
Bibliographical noteKAUST Repository Item: Exported on 2021-04-26
Acknowledgements: The presence of C, N, Fe, Cl, K, and O was confirmed by detection with EDX (Fig. 3g). C was the most available element at 54.56 At.%, corresponding to carbon content from CF support and cyanide in PB. C content was followed by N at 17.56, Fe at 16.4, Cl at 4.95, and K at 4.24 At.%, and the least available element was O at 2.28 At. %. PB has two phases: the original insoluble phase, which exists as FeII[FeIII(CN)]·xHO, and a soluble phase (KFeIII[FeII(CN)]), where K+ ions intercalate into PB, as a result of prolonged transformation between PW and PB . However, K had similar atomic presence as Cl, suggesting that K here belongs to a residual KCl rather than a soluble PB. This confirmed the deposition of PB in the insoluble form rather than in the soluble form.The PB/CF content and crystallinity were further examined by XRD. It can be observed from Fig. 4a that a slight shift in the peak position occurred between the PB powder and PB/CF, due to the chemical interaction between PB particles and carbon material as the main component of CF. Another important observation is that PB/CF showed less crystallinity (more amorphous) structure than that observed for PB powder. This decreased crystallinity is attributed to the presence of amorphous CF as a support material for PB powders. The PB/CF pattern showed five peaks, sharp peaks for KCl at 25.9° and 40.5°, amorphous broad peaks for CF at 25.50° and 43.5°, and barely distinguishable peaks of PB at 17.5°, 27.85°, and 43.5°. Clearer peaks were observed for PB powder, from the excess agglomerates on the materials. PB showed low intensity peaks at 17.25°, 24.54°, 35.15°, 39.40°, and 43.44° for (2,0,0), (2,2,0), (4,0,0), (4,2,0), and (4,2,2), respectively. The broad and short peaks proved the limited existence of a low crystallinity, which are centered cubic crystals mixed with a rich amorphous structure. KCl impurities were recorded at 27.1723° and 40.3711° for (2,0,0) and (2,2,0), respectively. Comparing PB/CF to pristine CF in FT-IR, there are distinctive differences owing to functionalization and PB deposition (Fig. 4b). The functionalization attached C[dbnd]N, C[dbnd]O functional groups on the CF surface, as evidenced by the bands at 1610 and between 980 and 1120 cm−1 [35,36]. The effect of PB deposition was apparent in the Fe–CN–Fe band at 608 cm−1, followed by the C[tbnd]N band between 2020 and 2300 cm−1, including the major peak at 2080 cm−1. The two bands confirm the existence of PB and its different phases in Fe3+-CN-Fe3+ (berlin green – BG), Fe2+-CN-Fe2+ (PW), and Fe2+-CN-Fe3+ (PB) [37–42]. Therefore, we can conclude that the deposited material exists in a mixture of different oxidation states. Furthermore, the knee before 2020 cm−1 and the broad band after 3000 cm−1 demonstrated the existence of –OH from free, bonded, and interstitial water, indicating a hydrous form of PB [35,40,43]. PB and its phases easily transform from one form to another; therefore, the the existence of the catalyst is confirmed by the confirmation of C[tbnd]N and Fe–CN–Fe bonds.Differential capacitance measurement (DCM) was used to calculate the electrochemically active surface area (ECSA) of the anodic electrodes, as explained in literature [48–51]. NNR/NF had an electrochemical active surface area of 913 cm2, whereas NF had an ECSA of 12.5 cm2 (Fig. S3). More information on the followed methodology and ECSA is provided in the supplementary material.This project was funded by National Research Foundation of Korea (NRF) grant from the Korean government (MSIT) (No. 2019R1A2C1006356) and partially by the Ministry of Oceans and Fisheries under TRO sensor development project (M01201920180035).
ASJC Scopus subject areas