Yeela
Persky
,
Yan
Yurko
,
Rifael Z.
Snitkoff-Sol
,
Noam
Zion
and
Lior
Elbaz
*
Chemistry Department, Bar-Ilan Center for Nanotechnology and Advanced Materials, Bar-Ilan University, Ramat-Gan 5290002, Israel. E-mail: lior.elbaz@biu.ac.il
First published on 12th December 2023
Fe–N–C catalysts are currently the leading candidates to replace Pt-based catalysts for the oxygen reduction reaction in proton exchange membrane fuel cells. To maximize their activity, it is necessary to optimize their structure to allow high active site density on one hand, and hierarchical porous structure that will allow good mass transport of reactants and products to and from the active sites on the other hand. Hence, the hierarchical structure of the catalyst plays an important role in the balance between the electrochemical active site density and the mass transport resistance. Aerogels were synthesized in this work to study the interplay between these two parameters. Aerogels are covalent organic frameworks with ultra-low density, high porosity, and large surface area. The relative ease of tuning the composition and pore structure of aerogels make them prominent candidates for catalysis. Herein, we report on a tunable Fe–N–C catalyst based on an Fe porphyrin aerogel, which shows high electrocatalytic oxygen reduction reaction activity with tunable hierarchical pore structure and studied the influence of the porous structure on the overall performance in proton exchange membrane fuel cells.
The oxygen reduction reaction (ORR) is regarded as the bottleneck reaction in proton exchange membrane fuel cell (PEMFC) technologies,8 and strongly affects their overall performance and efficiency.9 Hence, fuel cells’ research in the past decade has mainly been focused on the search for new ORR catalysts based on Earth-abundant, non-critical, and thus cheap, materials.10,11 One of the most interesting categories of platinum group metal-free (PGM-free) ORR catalysts comprises bio-inspired catalysts,12 composed of macrocyclic complexes of transition metals, such as metallo-porphyrins and other macrocyclic complexes.13–16 A sub-category that evolved from this work is the MNC catalyst composed of a transition metal (usually Fe), a nitrogen source to coordinate the Fe ions, and a carbon source to serve as the framework for this catalytic system.17 The performance of these PGM-free catalysts is relatively high in terms of electrochemical cell measurements but does not necessarily transfer well to fuel cell performance for several reasons such as low electrochemically active site density,18–20 formation of peroxides, overall durability and stability under electrochemical and chemical environments in fuel cells,21,22 and mass-transport issues.23 Hence, further improvements are necessary in order to realize their full potential and make fuel cell technologies more viable alternatives for green energy generation.
One possible way to circumvent some of these issues is to increase the electrochemically active site density.20,24 A promising path to achieve this, is by using highly porous materials with maximal active site density.25,26 One such family of materials are aerogels.26–28 Aerogels are composed of a network of covalently bonded molecules that form a ultra-low density, highly porous, hierarchical, covalent framework.29,30 Previous studies used an iron porphyrin as a monomer and an aldehyde cross-linker to form a porphyrin aerogel with ultra-high mass site density (MSD).31,32 In order to achieve optimal performance, utilize the maximum number of active sites, and reach high electrochemically active site density (EASD), one must consider the tradeoff between the EASD and the control of the hierarchical structure of the aerogel and its pore size distribution, which can be tuned to allow optimal mass transport to and from the active sites and expose a significant portion of them.33 Understanding the tradeoff between these parameters can provide useful distinctions regarding PGM-free catalysts in general, and aerogel-based PGM-free ORR catalysts in particular. The selected aerogel systems can be used as good model systems since they are composed of a well-defined atomically dispersed catalyst.
Herein, the length of the aerogel cross-linker molecule was changed in order to study the interplay between the EASD and mass transport. This change entails several differences between the aerogel structures. These new aerogels were studied in PEMFCs and optimized for best performance.
The morphology of the aerogels was studied using SEM. As can be seen from the images in Fig. 1, the structures of the four aerogels in Scheme 1 are quite different. This is the first indication that the cross-linkers have significant impact on the aerogel structure. The 1-FePA seems to be composed of aggregates of interconnected nanoparticles that build the aerogel, whereas the 2-FePA forms bulky interconnected balls, which resemble cotton balls with large void volumes between them. The 3-FePA seems to resemble the 2-FePA, but is decorated by some wires and flakes, while the 4-FePA seems to have formed dense nanosheets of the porphyrin aerogel.
Rotating ring-disk electrode (RRDE) measurements were conducted both in acidic (0.5 M H2SO4) and alkaline (0.1 M KOH) electrolytes to study the electrocatalytic activity of all four aerogels (Fig. 2). In both electrolytes, the 2-FePA reached the highest limiting current density, indicative of its tendency for higher reaction selectivity for the 4e− (or 2 + 2) reduction to water, when compared to the other aerogels. 4-FePA and 3-FePA exhibit the lowest and the highest onset potentials and half-wave potentials, respectively, which indicate the lowest and highest reaction kinetics as well. These aerogels also reached the lowest limiting current density, which indicates lower selectivity toward 4e− reduction to water and higher peroxides production, according to the Levich equation. The increased distance between the active sites in 3-FePA and 4-FePA might decrease the synergistic effect between adjacent sites and therefore resulted in lower selectivity toward 4e− reduction. All the onset and half-wave potentials, as well as the limiting currents of each measurement, are summarized in Table 1. It is very apparent that all aerogels perform better under alkaline conditions than in acidic electrolyte, with almost 100 mV difference in onset and half-wave potentials, and higher selectivity for the 4e− (or 2 + 2e−) ORR mechanism.
![]() | ||
Fig. 2 RRDE measurements (O2-saturated, 900 rpm, 5 mV s−1) in (A) 0.1 M KOH and (B) 0.5 M H2SO4 of 1-FePA (red), 2-FePA (purple), 3-FePA (orange), and 4-FePA (blue). |
Cross-linker | In alkaline electrolyte (0.1 M KOH) | In acidic electrolyte (0.5 M H2SO4) | ||||
---|---|---|---|---|---|---|
E onset (V vs. RHE) | E 1/2 (V vs. RHE) | Number of electrons | E onset (V vs. RHE) | E 1/2 (V vs. RHE) | Number of electrons | |
1-FePA | 0.95 | 0.84 | 3.9 | 0.874 | 0.601 | 3.3 |
2-FePA | 0.957 | 0.795 | 4.0 | 0.858 | 0.623 | 3.6 |
3-FePA | 0.957 | 0.816 | 3.5 | 0.874 | 0.699 | 2.6 |
4-FePA | 0.905 | 0.750 | 3.4 | 0.849 | 0.542 | 2.3 |
Stability tests were performed in alkaline (0.1 M KOH) electrolytes (Fig. 3). All the aerogels showed very good stability for the duration of the experiments (12 hours), with 1-FePA and 4-FePA exhibiting the highest and lowest stability, respectively. The 1-FePA lost less than 5% of its initial current density during a potentiostatic measurement at 0.5 V vs. RHE, whereas the 4-FePA lost about 12% of its performance. The differences between the stability of these aerogels can be attributed to the less selective 4-electron ORR with the 4-FePA and the formation of peroxide anions at higher concentrations, as can be deduced from Table 1.
![]() | ||
Fig. 3 Stability measurements at 0.5 V vs. RHE in 0.1 M KOH (O2-saturated, 200 rpm) of 1-FePA (red), 2-FePA (purple), 3-FePA (orange), and 4-FePA (blue). |
N2 adsorption isotherm measurements were performed in order to gain insight into the porous nature of the various aerogels and understand how the size of the linker molecule affects it. The calculated BET surface areas of all four aerogels are presented in Table 2. As can be seen in this table, the surface area increases with the increase in the length of the linker up until the length of the third aromatic ring. The increase in the size of the linker can be expected to result in larger pores, and thus larger total void volumes, given the fact that the same amount of porphyrin was used in the synthesis of all aerogels, and thus also in an increase in the surface area. The 4-FePA is an exception, most probably due to the collapse of some of the pores and formation of a dense aerogel, as was also observed in the SEM image, where it formed dense nano-flakes.
Cross-linker | 1-FePA | 2-FePA | 3-FePA | 4-FePA |
---|---|---|---|---|
Surface area (m2 g−1) | 249.17 | 361.96 | 412.01 | 377.07 |
The pore size distribution of the aerogels was also studied to further understand the differences between them. As mentioned earlier, the size of the pores may affect the mass transport and the EASD, and thus an optimum balance between the MSD, the EASD and mass transport should be found to maximize the performance of aerogels in fuel cells. Generally, micropores increase the surface area, mesopores promote wetting inside the pores, and thus increase the EASD, and create the three-phase boundary between the catalyst, the electrolyte, and the oxygen, while macropores help facilitate the mass transport into the pores.34–36 A combination of all of the above, to create a tunable hierarchical structure, is expected to result in high electrocatalytic performance, especially in fuel cells.
The pore size distributions of all four aerogels are presented in Fig. 4A and B. The pore size distributions were normalized according to the surface area of the pores in order to understand the different ratios between the pore sizes for each aerogel. As can be seen in Fig. 4B, each aerogel has different pore size distributions. The 2-FePA structure has more meso- and macro-pores in comparison with the micro-pores than the other aerogels. The majority of 3-FePA's surface area comes from 2 pore sizes: below 1 nm and around 3.8 nm, and it almost does not have any meso-pores, whereas the 4-FePA is composed mostly of micro-pores below 1 nm (61%), which confirms the previous observations regarding its density. The 1-FePA is also mostly composed of micro-pores, but unlike 4-FePA, the rest of its surface area is composed of a wider range of pore sizes, which indicates a more hierarchical structure in the 1-FePA aerogel. An indication for the existence of large macro-pore in these aerogels can be obtained from their N2 adsorption–desorption isotherms, which are presented in Fig. 4C. This is manifested by the rapid increase in the adsorption at P/P0 close to 1, which is associated with pore condensates in macro-pores,37,38 and suggests the existence of macro-pores in all the aerogels, as was also observed clearly in the SEM images, with 3-FePA having the largest portion of macro-pores, and 4-FePA having the lowest. This analysis suggests that 2-FePA has the most hierarchical pore structure of the aerogels studied in this work.
Another important parameter that affects catalytic activity is the site density. Intuitively, increasing the length of the cross-linkers can cause a decrease in the mass site density (MSD), and this was confirmed by ICP measurements (Table 3) that show a decrease in the Fe content with an increase in the cross-linker length. An increase in the length of the cross-linker translates to an increase in the weight percentage of carbon and correspondingly a decrease in the iron weight percentage.
Cross-linker | 1-FePA | 2-FePA | 3-FePA | 4-FePA |
---|---|---|---|---|
ICP Fe (wt%) | 7.52 | 6.99 | 6.06 | 6.12 |
Although the MSD is important, measuring the EASD is the accurate way to report the active site density, since it takes into account only the electrochemically accessible sites, or in simpler words, only those sites that are available to participate in the reaction. To calculate the EASD in a working fuel cell, cyclic voltammetry (CV) measurements were performed in proton exchange membrane fuel cells (PEMFCs) (Fig. 5). The CVs of all the aerogels have a similar peak at around 0.76 V vs. the H2 anode, which is associated with redox peaks of Fe(III)/(II).19,39,40 Since the active sites of this category of PGM-free catalysts are known to be FeN4,41,42 and the catalytic cycle involves the redox reaction of the Fe(III)/(II) couple,43 the integration of the area under the peaks can be used to calculate the EASD. This is not always so clear and easy, and in most cases, more advanced electrochemical techniques, such as Fourier-transformed alternating current voltammetry, are required if the peaks are not well pronounced in the CV.19,20 The EASD was calculated using eqn (1):44
![]() | (1) |
![]() | ||
Fig. 5 Cyclic voltammetry measurement in fuel cells. Anode/cathode, H2/N2, 200 ccm/800 ccm, 80 °C/80 °C/80 °C. 1-FePA (red), 2-FePA (purple), 3-FePA (orange), and 4-FePA (blue). |
Cross-linker | 1-FePA | 2-FePA | 3-FePA | 4-FePA |
---|---|---|---|---|
Peak area (mC cm−2) | 13.45 | 27.64 | 27.90 | 26.05 |
EASD (sites × 1019 g−1) | 2.09 | 4.64 | 4.13 | 4.46 |
All four aerogels showed very similar EASD values, except for the 1-FePA, which showed a lower EASD, although according to the ICP results (Table 3), this aerogel has the highest Fe wt%. This can be explained by the formation of hidden Fe sites as discussed earlier.
The ultimate way to study the effect of the aerogels' hierarchical structure, which seems to be the most significant difference between the aerogels based on the EASD and BET measurements, is the fuel cell measurement. And thus, in order to acquire a more comprehensive picture of the performance of the aerogels as ORR electrocatalysts, which is not necessarily reflected in RRDE measurements,23 the aerogels were tested in an H2–air PEMFC (Fig. 6A and B). The results were compared to those of a Pajarito Powder cathode as well, which was outperformed by the 1-FePA and the 2-FePA. From the I–V curves of all four aerogels, the PEMFC based on 2-FePA as the cathode catalyst exhibited the highest current densities at both high potential and low potential. The open-circuit voltage (OCV) was very close between all four aerogels.
The kinetic current densities were determined from the results shown in Fig. 6A at 0.75 V and are found to be Jk@0.75 V = 12.08, 13.00, 14.14, and 7.9 mA cm−2 for 1-FePA, 2-FePA, 3-FePA, and 4-FePA, respectively. These results in the kinetic activation region agree with the RRDE results (Fig. 2A and B), showing that the 3-FePA is the most active aerogel in this region but is not the overall best catalyst of the four, where from the RRDE measurements it seems to have much lower selectivity for the 4e− (or the consecutive 2 + 2) ORR mechanism, much like the 4-FePA, whereas the 1-FePA and the 2-FePA show better selectivity, both in alkaline and acidic environments, manifested in their higher current densities in RRDE and in higher current densities in the PEMFC polarization curves (Fig. 6A). This could be explained by the distance between active sites, which allows higher selectivity, and eventually higher overall performance in a fuel cell, as the active sites are closer together as was observed with molecular dual-site porphyrinic systems described in the past.45 This is an important indication for the synergistic behavior of these active sites.
The fuel cell performance does not correlate with the BET surface area, nor the EASD, as illustrated in Fig. 6B and C, but it does correlate with the pore size distributions of the aerogels, which might explain the trend in higher current densities. The highest activity was observed with the 2-FePA aerogel, which has the most hierarchical structure (Fig. 4A and B). 4-FePA, which is composed mostly of a large proportion of <1 nm pores, shows the lowest activity, whereas the 3-FePA, which is mainly composed of <1 nm and ∼3.8 nm pores, exhibits reasonable activity at high potentials, but relatively poor activity at lower potentials because of its non-hierarchical structure. 1-FePA, which also has a hierarchical structure, also reached relatively high activity in its PEMFC.
In order to test this hypothesis and to achieve better understanding of the influence of mass transport resistance on the activity, the aerogels were tested in a PEMFC with mixture of N2/O2, with 4% O2. The measurements were carried out in order to test the performance in a low-oxygen environment where the mass transport becomes critical. The I–V polarization curves are presented in Fig. 7.
![]() | ||
Fig. 7 Fuel cell measurements. H2/96% Ar–4% O2, 200 ccm/200 ccm, 80 °C/80 °C/80 °C. 1-FePA (red), 2-FePA (purple), 3-FePA (orange), and 4-FePA (blue). |
In this case, the cell with the 2-FePA aerogel also exhibited the highest performance and reached the mass transport limitation at much higher current densities compared to the other aerogels. The “knee”, which represents the transition from the ohmic region to the mass transport region, started at lower current densities for the aerogels with lower hierarchy in their pore structure, and the performance trend increased according to the following order: 4-FePA < 3-FePA < 1-FePA < 2-FePA. These measurements confirm that the structure of an aerogel does indeed affect the mass transfer, and therefore optimized structure, with hierarchical pore structure, can be critical to obtaining high performance in PEMFCs with this class of catalysts.
The Brunauer–Emmett–Teller (BET) surface area measurements were conducted using an N2 adsorption isotherm at 77 K in Quantachrome Autosorb iQ. The pore size distribution was calculated according to DFT models.
The morphology was studied using an Environmental Scanning Electron Microscope (E-SEM, Quanta FEG 250).
This journal is © The Royal Society of Chemistry 2024 |