Morphological and Structural Design Through Hard-Templating of PGM-Free Electrocatalysts for AEMFC Applications

This study delves into the critical role of customized material design and synthesis methods in influencing the performance of electrocatalysts for oxygen reduction reaction (ORR) in anion exchange membrane fuel cells (AEMFCs). It introduces a novel approach to obtain platinum-free electrocatalysts (PGM-free), based on the controlled integration of iron active sites onto the surface of silica nanoparticles (NPs) by using nitrogen-based surface ligands. These NPs are used as hard templates to form tailored nanostructured electrocatalysts with an improved iron dispersion into the carbon matrix. Utilizing a wide array of analytical techniques including infrared and X-ray photoelectron spectroscopies, X-ray diffraction and surface area measurements, this work provides insight into the physical parameters that are critical for the ORR electrocatalysis with PGM-free electrocatalysts. The new catalysts showed an hierarchical structure containing large portion of graphitic zones which contribute to the catalyst


Introduction
Green hydrogen is expected to be used in many hard-to-abate sectors, such as the metals' industries, fertilizers, automotives, and residential purposes 1 , where fuel cell technologies will play a pivotal role in producing electricity directly from hydrogen.
The most developed low-temperature fuel cell technology today is the proton exchange membrane fuel cell (PEMFC).These fuel cells rely on platinum group metals (PGM) supported on carbon (Pt/C) to catalyze both the anodic hydrogen oxidation reaction (HOR) and the cathodic oxygen reduction reaction (ORR).The ORR is considered the bottleneck reaction due to its sluggish kinetics and complex pathways; therefore, high loadings of Pt/C are required.This, in turn, translates into a high cost and large dependence on PGMs that are also considered as critical raw materials (CRMs), and hinder the PEMFC's wide commercialization. 2 One way to mitigate this issue is to develop anion exchange membrane fuel cells (AEMFC), recently proposed as a costeffective and practical alternative to PEMFCs. 3,4 ][7][8][9] Particularly, atomically dispersed transition metals coordinated with nitrogen embedded in carbonaceous support (M-N x -Cs) are the most promising active sites for the ORR in AEMFCs. 10One drawback of using PGM-free ORR electrocatalysts is their relatively low turnover frequency (TOF) when compared with PGMs. 11,12 [15][16] Different synthetic routes have been pursued over the years, mainly based on the mechanical mixing of various precursor species followed by pyrolytic processes in which the temperature and atmosphere are finely controlled. 17The influence of temperature increments during pyrolysis has been elucidated, shedding light on the temperature's impact on the metal coordination, as well as activity and durability of the derived electrocatalysts. 18- 20][23] To deal with their intrinsically low TOF, which is approximately two orders of magnitude lower than that of PGMs, 24,25 soft and hard templating strategies have been adopted to increase the porosity of electrocatalyst materials, and thus the active sites' utilization [26][27][28] as well as the electrochemically active site density (EASD). 11,29 ft templating consists of a templating agent that disappears during pyrolysis, forming the porosity of the material. 30For tuning the porosity, for example, Zn or ZnO can be used, and later decomposed to create porous structures. 31Another soft templating agent can be urea which decomposes at a relatively low temperature (350°C), increasing the porosity 32,33 and can be used as a source of nitrogen for the M-N x -C. 34Even hard templating methods have been widely used.][62][63] The only commercially available PGM-free ORR electrocatalysts is commercialized by Pajarito powder 64 Precious-Metal-Free™ (PMF) electrocatalysts.These electrocatalysts are produced via the silica hard templating method and have achieved high electrocatalytic activity, and nowadays are considered as the benchmark for ORR electrocatalysis.With the aim of achieving higher electrochemically active site density (EASD), to mitigate one of the intrinsic issues relating to PGM-free ORR catalysts: low TOF.The best way to do this is to increase the EASD to compensate for the relatively low reaction kinetics, which is app. 2 orders of magnitude lower than that of Pt.And better control of the iron coordination and its distribution within the electrocatalyst, herein we propose a new synthetic method for synthesizing ORR Fe-based electrocatalysts to substitute the more conventional mechanical mixing approach Amino-based functionalizing agents were tethered to silica NPs and used to coordinate iron ions onto the surface of the NPs, which were used as a hard templating sacrificial support to create porous carbonaceous materials through a combined pyrolytic and etching method.Such functionalized silica NPs at the same instance acted as the template to induce porosity and Fe source that can reduce the undesired coalesce of Fe into the evolution of metallic NPs, and ensure the homogeneous distribution of active sites.Later, the effect of the synthesis conditions on the physical properties and eventually the performance were investigated.The remarkable performance in AEMFC is discussed in terms of the synthesis protocol itself and the pyrolysis conditions, which resulted in an increase in the EASD.Experimental Materials All the chemicals and reagents used in the study were of high-purity analytical grade and used as it is without any further processing.For the synthesis of silica templates functionalized with amino-based ligands and iron: tetraethylorthosilicate (TEOS), ammonium hydroxide 25% (NH 4 OH), and iron sulfate hydrate Fe 2 (SO 4 ) 3 •7H 2 O were purchased from Sigma Aldrich.

Silica
Templates Fabrication and Functionalization SiO 2 NPs (70 ± 5 nm average diameter) were synthesized through a classical Stöber method. 65The procedure is based on the quick addition of TEOS (0.24 M) to a basic ethanol solution (NH 4 OH 0.36 M) at room temperature (RT) to form colloidal silica NPs.After 24 h, the powder is recovered through centrifugation (9000 rpm, 30 min), washed twice with fresh EtOH and with distilled water and finally dried at 80 °C overnight.
SiO 2 NPs were then functionalized with APTES or EDTMS to obtain surface amino groups attached to the SiO 2 surface suitable for iron coordination.The procedure was adapted from previous publication, 66 by employing Fe 2 (SO 4 ) 3 •7H 2 O as a suitable iron precursor.The functionalization step was carried out in EtOH at 60°C and 24 h, by adding NH 4 OH (0.0042 M) to favor the hydrolysis and condensation of the silanes over the SiO 2 surface.The amounts of APTES or EDTMS used were equal to a molar ratio between the silane and surface OH groups of SiO 2 of 1:2 (calculated according to the Thermogravimetric Analysis, TGA in the supplementary information).The powder was again recovered through centrifugation, washed twice with EtOH, and dried at 80°C overnight.The samples are called SA and SE, respectively, where A and E mean APTES or EDTMS.
In the last step, iron was dispersed over SiO 2 NPs by exploiting the amino groups of APTES or EDTMS.The synthesis was carried out in EtOH at reflux, by using a slight iron excess (Fe: silane molar ratio equal to 1.5:1, as determined by TGA analysis, Figure S1b).The final material was recovered through centrifugation, washed with fresh EtOH, dried at 80 °C overnight and finally labeled as SAFe and SEFe, respectively.

Electrocatalyst Development
Using iron-functionalized silica templates i.e.SAFe and SEFe, atomically dispersed Fe-N-Cs were developed.This part of the synthesis utilizes the simultaneous opening of the materials' pores and incorporation of Fe using the hard templating method.In brief, the functionalized silica NPs (SAFe or SEFe), simultaneously acting as Fe source and templating agent, were mixed with 1-methylimidazole (M, an organic precursor containing both nitrogen and carbon) in a 3:7 weight ratio in deionized water for 24 h under continuous stirring, followed by overnight drying at 80°C.The obtained dried slurry was ground into a fine powder using a mortar grinder and subsequently subjected to a first pyrolysis (P1) at 900°C for 1 h under N 2 at a heating and cooling rate of 300°C/h.The obtained black powders were named SAFe_M_P1 and SEFe_M_P1, respectively.Next, the materials were treated with an acidic solution (2:1 mixture of 25 wt.%HF and 35 wt.% HNO 3 ) for three days at RT to remove the silica templates.The materials obtained after the acid washing were labeled as SAFe_M_P1A and SEFe_M_P1A.Finally, a second pyrolysis treatment (P2), like the previous one with the only difference in atmosphere (a slightly reducing atmosphere of N 2 balanced with 5 wt.%H 2 ) was conducted.The samples obtained in the final stage were labeled as SAFe_M_P1AP2 and SEFe_M_P1AP2.The electrocatalysts obtained at different levels of experimental design are summarized in Table S1.
Thermogravimetric analysis (TGA) was performed by a Mettler Toledo StarE System TGA/DSC1 (scan range 30-1000 °C, heat rate 10 °C min -1 , airflow 50 mL min -1 ).TGA curves were registered on the as-prepared SAFe and SEFE NPs to confirm the silica functionalization with APTES and EDTMS, as well as after the acid treatment (P1AP2) to verify the effective silica removal.
Inductively coupled plasma-optical emission spectroscopy (ICP-OES) was used to measure the Fe amount during the preparation steps, using an ICP-OES Optima 7000 DV PerkinElmer instrument.The acid digestion was carried out in a microwave Milestone Ethos mineralizer instrument.
X-ray photoelectron spectroscopy (XPS) analysis was conducted using a Nexsa spectrometer equipped with a monochromatic, microfocused, lower Al Ka X-ray (photon energy 1486.6 eV).
BET measurements were conducted using a Quantachrome Autosorb iQ instrument with nitrogen as the adsorbate at 77 K.The degassing step included heating the sample to 130°C (heating rate 5°C min -1 ), until the apparatus detected no gas emission from the sample.The pore size distribution was calculated using the DFT method for slit pores.
Further investigation of the morphological structure of the catalysts was done by using a High-Resolution Transmission electron microscopy (HRTEM), JEOL JEM 2100, LaB6 filament, 200 kV.

Electrochemical Measurements
To carry out electrochemical measurements, an electrochemical ink was prepared according to our previous reports. 67,68 irst, 4.5 mg of the electrocatalyst powder was dispersed in 985 µL of isopropanol and then 15 µL of 5 wt.%Nafion ® D-520 (Alfa Aesar) was added using a precision pipette (to serve as a binder).The solution was probesonicated for 10 mins and then the vials containing inks were placed in an ultrasonic bath for 30 mins for further homogenization.The working electrode (glassy carbon) was loaded with 0.6 mg cm -2 catalyst.Eventually, the ORR electrocatalysis in a half-cell arrangement was analyzed using a typical threeelectrode system (with Pine RDE connected to a Pine bipotentiostat) comprising of RRDE (glassy carbon disk and Pt ring) as the working electrode, glassy carbon rod as a counter electrode and for the reference saturated calomel electrode (SCE, °  = 241 ) was used. 69The measurements were carried out in both O 2 -saturated acidic and alkaline media of 0.5 M H 2 SO 4 and 0.1 M KOH, respectively.The measured potentials were then converted to a pH-independent scale of reversible hydrogen potential (RHE) using Eq. 1 while the potential window was maintained in the potential window of 1.2 to 0.0 V vs RHE.
Eq. 1 By measuring the disk and ring currents (I disk and I ring , respectively), the peroxide yield (%) and the number of electrons transferred during ORR (n) were calculated using Eq. 2 and Eq. 3, respectively.
An accelerated stability test was performed using 2000 voltage cycles to evaluate the stability profiles of the best-performing electrocatalysts.Briefly, 0.6 mg cm -2 were drop-casted on a GC disk in an RRDE electrode and cycled 2000 times at the scan rate of 50 mV s -1 between 1.20 to 0.32 V vs RHE in the O2-saturated 0.1 M KOH at a rotation speed of 1600 rpm.For comparison, LSVs in the beginning and at every subsequent 500 th cycle was recorded at the slow scan rate of 5 mV s -1 .

Fuel Cell Testing AEM
The electrocatalyst ink for the cathode (∼2 mg cm -2 loading) was prepared by mixing 11 mg of electrocatalyst with 3.66 mg of ionomer (Aemion, HNN-5).After grinding the solids with a mortar and pestle, 1.5 mL of methanol: H 2 O solution (3:1 w/w) was added, and the ink was sonicated for 1 h.The ink was sprayed using Sono-Tek ultrasonic spray on a 5 cm 2 gas diffusion layer (GDL) (BC-29) until the desired loading was reached.The anode was prepared in the same way, using 40 wt% Pt/C (HiSPEC 4000) as an electrocatalyst (0.6 mg Pt cm −2 ).Before the assembly, the membrane (AF1-HNN8-25) and GDEs were activated by soaking in 0.1 M KOH solution for 1 h.The KOH solution was replaced every 20 min.Fuel cell measurements were conducted using the Scribner 850e station.The anode and cathode gases were H 2 and O 2 , respectively.The flow rate was 0.3 L min -1 at the anode and 0.5 L min -1 at the cathode, with 100% RH and running at 60/60/60°C at the anode gas inlet, cell, and cathode gas inlet, respectively.The backpressure was 150 KPa.

PEM
Both electrocatalysts developed in this work were also tested in a PEM-FC.The cathode (∼5 mg cm -2 loading) was prepared by mixing 26 mg of electrocatalyst with 152 mg of Nafion solution (D1021), 0.566 mL isopropanol and 0.477 mL of DI water.The ink was sonicated for 1.5 h in ice.The electrocatalyst was sprayed on 5 cm 2 GDL (BC-29) using Sono-Tek ultrasonic spray.At the anode side, a commercial gas diffusion electrode (GDE) (Fuel Cells Etc., CST-GDE-01) of 0.2 mg Pt cm −2 Pt/C.The anode GDE was hot pressed with NR-211 Nafion membrane (120 °C) for 2 min.Then, the cathode GDE was assembled without hot pressing the MEA.Fuel cell measurements were conducted using the Scribner 850e station.The anode and cathode gases were H 2 and O 2 , respectively.The flow rate was 0.5 L min -1 both at the anode and cathode, with 100% RH and running at 80/80/80°C at the anode gas inlet, cell, and cathode gas inlet, respectively.The backpressure was 150 KPa.

FTacV
After IV polarization and cyclic voltammetry (CV) were measured, Fourier-transformed alternating current voltammetry (FTacV) was used to evaluate the electrochemically active site density (EASD) of the electrocatalysts in situ working fuel cell.The operation conditions were similar to the conditions described in the previous section.The FTacV measurements were conducted using Biologic SP-300 potentiostat with the reference and counter shortened at the anode.The cell temperature was 80°C and the anode and cathode gases were H 2 and N 2 , respectively.Both H 2 and N 2 gases were fully humidified and with a flow rate For the simulations of each electrocatalyst the parameters were: SEFe_M_P1AP2: R = 0.333 Ohm; Cdl = 0.139 F cm -2 ; number of sites = 5.72*10 16 ; E 0 = 0.76; k = 7.88 s -1 ; C= 1.9*10 -8 mol cm -2 .SAFe_M_P1AP2: R = 0.144 Ohm; Cdl = 0.09 F cm -2 ; number of sites = 6.62*10 16 ; E 0 =0.76; k=6.97 s -1 ; C = 2.2*10 -8 mol cm -2 .The uncompensated resistance was calculated from EIS (electrochemical impedance spectroscopy) conducted at 0.5V, by scanning the frequency from 10 kHz to 100 mHz with 20 points per decade, and at a perturbation amplitude of 10 mV.The double-layer capacitance (Cdl) was calculated from the CV.All the data analysis and simulations were performed in the HRview software. 62

Results and discussion
Herein we report on the development of the Fe-N-C ORR electrocatalysts using the hard templating method as described in the experimental section (Scheme 1).In contrast to previous studies, where mixtures of all precursors and templating agents are prepared for the pyrolysis step 71- 73 , in this work iron was integrated through the ligand-tethered silica templates prior to the pyrolysis process, to achieve high EASD.To study this new approach, two different types of surface functionalizing agents (3-aminopropyl)triethoxysilane (APTES), or N-aminoethylaminopropyltriethoxysilane (EDTMS) tethered to SiO 2 -based templating agents were used.
First, SAFe and SEFe (Fe coordinated with SiO 2 -tethered APTES or EDTMS, respectively) samples were characterized to confirm the effective functionalization of silica NPs with both the silanes and iron.FTIR and TGA measurements confirm the presence of surface organic ligands on the SiO 2 surface (Figure S1).The amounts of APTES and EDTMS were calculated according to the previous publication, 66 as described in the supplementary information.The amount of iron anchored on SiO 2 NPs was measured with ICP-OES and was equal to 1.4 wt% ± 0.1 for SAFe and 1.5 wt% ± 0.1 for SEFe, validating the efficient addition of iron to both samples.SAFe and SEFe were later used as hard templating agents for the electrocatalyst, by mixing with 1methyimidazole (M).FTIR spectra were recorded after each preparation step to follow the material evolution (Figure 1, SEFe sample reported as an example).
The FTIR spectrum of SEFe_M is dominated by the typical peaks of M.However, in the SEFe_M_P1 the M pattern disappears due to its decomposition during the first pyrolysis; besides, the peaks associated with SiO 2 NPs appear, with the main broad peak at 1100 cm -1 associated with Si-O-Si stretching.The spectra of both SEFe_M_P1A and SEFe_NC_P1AP2 do not show SiO 2 peaks, confirming the effective removal of SiO 2 NPs after the acid treatment.Similar results were obtained in the presence of SAFe, as shown in Figure S2.Further confirmation of silica removal was obtained by TGA analysis performed before and after the acid treatment (Figure S3).The TGA results indicate that after the acid treatment (P1A) the samples do not show any inorganic residues (< 5 wt%), whereas the inorganic part after the first pyrolysis was equal to about the 70 wt% of the sample weight.This highlights the good performance of the employed etching procedure.
The XRD patterns of the two samples after the entire preparation procedure are presented in Figure 2a.These results show one broad peak at approximately 25° and one minor peak at about 44°, which corresponds to the 002-oriented and 101oriented diffraction peaks of graphite, respectively. 74he comparison with the previous preparation steps (Figure 2b, SEFe_M series reported as exemplary sample) underlines the shift of the main broad peak at 22° in the SEFe_M_P1 sample probably due to SiO 2 NPs, which become imperceptible after silica removal (SEFe_M_P1A).Besides, the crystalline phase due to methylimidazole (SEFe_M) completely disappears after the first pyrolysis (P1) and no crystalline phases were detected in any of the subsequent  preparation steps.This demonstrates that by using this synthetic approach, the formation of the oxide phases is discouraged, whereas additional oxide phases were obtained by using other published methods increasing the electrocatalyst speciation. 32,33 addition, there is a slight difference between the graphitic structure of SEFe compared to SAFe.The broader Lc peak of SEFe suggests smaller crystallite sizes and a possible lower crystallinity or the presence of polycrystalline structures.This distinction in peak width may provide insight into the structural differences between the two materials, which can impact their ORR catalytic activity.
The structures of the new materials were also studied using HRTEM (Figure 3).The images show slight differences in samples morphology.In agreement with XRD results, both SAFe_M_P1AP2 and SEFe_M_P1AP2 display similar graphitic structures constituted by NPs aggregates.However, while in SAFe_M_P1AP2 the particles seem organized in larger pseudospherical aggregates, in SEFe_M_P1AP2 the random orientation of the crystallographic planes may suggest the formation of more elongated aggregates.These features are expected to manifest themselves in the overall performance of these materials in AEMFC.Fewer active edge groups on ordered materials can lead to lower catalytic ORR activity, as edge sites often play a pivotal role in electrocatalysis. 75,76 onversely, the   inclusion of a graphitic structure in electrocatalysts is also important, as it offers a highly conductive framework, facilitating very good electronic conductivity, and increased durability. 77,78 e BET surface area and pore-size distribution were determined from the nitrogen adsorption isotherm for both SEFe_M_P1AP2 and SAFe_M_P1AP2.The SEFe_M_P1AP2 has a surface area of 521 m 2 g -1 , while SAFe_M_P1AP2 has 319 m 2 g -1 .Examination of the cumulative surface area (Figure 4) revealed a very wide pore-size distribution for the SEFe_M_P1AP2, with a pore diameter spanning from 5 to 35 nm, whereas the SAFe_M_P1AP2 has a much narrower pore-size distribution, where most of the pores are of 10-15 nm in diameter.
The analysis of the adsorption data, utilizing the DFT method for pore size distribution calculation indicates that SEFe displayed a less ordered structure.Yet, both exhibit distinct average pore sizes, ranging from micro-to meso--pores.Such morphologies may favour a high distributuion of the active sites and have been found to be instrumental in facilitating the good mass transport of oxygen to the active sites in fuel cells with PGM-free electrocatalysts with potential benefits to their performances. 13PS was used to study the surface composition and coordination chemistry of the newly developed electrocatalysts.All samples were analyzed after each preparation step, to track the structural composition changes during the synthesis.The extensive elemental analysis, depicted in Table S2, emphasizes the predominance of carbon, nitrogen, and oxygen as principal components in both materials, while silicon is initially detected in the first synthetic steps and decreases in quantity after the etching procedure.Highresolution Fe2p and N1s spectra were carried out to study the coordination  chemistry of the metal active site in the materials.Figure S4 shows that both materials have low iron content manifested in their weak Fe2p spectra, consistent with ICP-OES measurements (< 2 wt% iron, Table S3).Early pyrolysis (P1) alters the iron coordination, with Fe2p 3/2 signals at ~711 eV, indicating the formation of more Fe 2+ sites.SAFe_M_P1 and SAFe_M_P1AP2 have similar Fe 2+ and Fe 3+ species, while SEFe_M_P1 retains its original coordination chemistry, confirmed by the N1s spectra (Figure 5).It shows the existence of diverse nitrogen compounds and their changes throughout each stage of the synthesis.The deconvolution of the N1s peak of both SAFe_M and SEFe_M revealed imine-like N, N-Fe, and pyrrolelike N, in both samples.Additionally, a secondary amine is present in the SEFe_M sample, which most probably originates from the pristine SEFe.A noticeable generation of different species was detected during the early pyrolysis stage (P1).The deconvolution analysis revealed the presence of several nitrogen compounds including imine, pyridinic N, N x -Fe, hydrogenated N, graphitic N, and NO x .The SAFe_M_P1 sample remarkably exhibited a composition closely mirroring that found in its ultimate form (SAFe_M_P1AP2), at this stage the main nitrogen species originates from hydrogenated N (401 eV), while the secondary major species is pyridinic N (399 eV).The atomic ratio between them is 1.5:1 in relation to the hydrogenated N, and it remains consistent after the second pyrolysis step (P1AP2).This was in contrast with the SEFe_M_P1 sample, which exhibited different compositional ratio characteristics of its final product (SEFe_M_P1AP2).The composition of SEFe_P1 initially consisted of equal amounts of hydrogenated nitrogen and pyridinic nitrogen, with a ratio of 1:1.However, this ratio changed to 1.5:1 in favor of the hydrogenated nitrogen species during the pyrolysis and acid wash stages (P1AP2).This disparity suggests a more consistent and direct formation of active sites in the SAFe_M_P1AP2 as opposed to the SEFe_M_P1AP2.Subsequently, focusing on the final product (SAFe_P1AP2 and SEFe_P1AP2), it becomes apparent that the composition and ratios of the nitrogen species mentioned earlier are similar in SAFe_MP1AP2 and SEFe_MP1AP2.This indicates that the employed synthesis method and the use of the two different silanes do not significantly alter their composition and active site environment.
The electrocatalytic activity of all samples obtained at various stages of the fabrication process in the ORR reaction was first analyzed by RRDE in alkaline environment (0.1 M KOH).A summary of the ORR activity, selectivity, and kinetics parameters of the electrocatalysts are given in supplementary Figure S5, S7 and summarized in Table S4.
The electrocatalytic activity of the newly developed materials was positively affected by the synthetic procedure, resulting in very active electrocatalysts, exhibiting remarkable ORR performance (Figure 6).This high activity can be attributed to the high active site density obtained after the removal of the templating agent during the acid wash, that could promote the good distribution of atomically dispersed active moieties in the micro-mesoporous architecture during the second pyrolysis.Interestingly, after the final synthesis step, i.e. after the second pyrolysis step, both samples showed very high onset potentials and limiting current densities, hinting on the 4-electron selectivity, as can be seen in Figure 6a (top).The onset potential, here determined at a current density of -0.1 mA cm -2 , is 950 and 965 mV vs. RHE for SAFe_M_P1AP2, and SEFe_M_P1AP2, respectively.Both SAFe_M_P1AP2 and SEFe_M_P1AP2 show comparable ORR activity when compared to the state-of-theart Fe-N-C electrocatalyst (Pajarito powder®), highlighting SXFe_P1AP2 potential as a practical and effective Fe-N-Cs electrocatalyst for ORR.As can be seen in Figure 6b (bottom), the peroxide anion yield remained below 22%, in both cases, it ramps up from 0.6 V vs. RHE and lower potentials.This can be explained by the catalysis of ORR by quinone moieties formed on the carbonaceous material, a well-documented phenomenon in alkaline conditions. 79,80 nterestingly, it is more pronounced with the SAFe_M_P1AP2, which could hint at a more oxidized surface and thus higher concentration of quinone moieties there.The calculated number of electrons in the ORR is almost a perfect 4 for both electrocatalysts at the high potentials' region 0.6 -0.8 V vs. RHE and slightly decreases below that (Figure 6b), in full agreement with the peroxide anion yield.Based on the RRDE measurements, Tafel slopes were used to examine the electrocatalyst' kinetics.
(Figure S7).The final development step provided the lowest value for both electrocatalysts (59 mV dec -1 for SEFe_M_P1AP2, and 62 mV dec -1 for SA_M_P1AP2).The decrescent of the Tafel slope's values through the development process can be correlated with the modifications in the active sites, as evidenced by the XPS results.Linking the enhancement in the electrocatalytic kinetics to the structural changes in the active site.Furthermore, the stability of both electrocatalyst was measured in alkaline conditions (Figure S8 and S9).
The findings indicate that SEFe_M_P1AP2 had a limiting current reduction of 0.6 mA cm -2 after 2000 cycles, whereas SAFe_M_P1AP2 had a decrease of 1.5 mA cm -2 .These results are consistent with SEFe_M_P1AP2 having more graphitic regions, which improve structural robustness and conductivity.
The ORR performance of the developed electrocatalysts were also analyzed in an acidic medium (0.5 M H 2 SO 4 ), and the recorded trends are presented in Figure S6.The activity patterns, on the whole, stayed consistent with those observed in alkaline medium where the samples obtained at the final stage of the second pyrolysis demonstrated optimum performance with an aptitude to carry out ORR following the direct 4-electron route.Notably, the performance in acidic environment was lower than in alkaline, exhibiting onset potentials of 772 mV and 808 mV, for the SEFe_M_P1AP2 and SAFe_M_P1AP, respectively.Durability measurements were conducted for the SEFe_M_P1AP2 and SAFe_M_P1AP2 as discussed in the experimental section and are presented in Figures S8 and S9.Interestingly, even after 2000 cycles, only a negligible decay in the E onset and E 1/2 values was observed, however, the limiting current densities were slightly reduced over the course of the stability test, where the major decrease was observed after the first 500 cycles.This is also manifested in the slight rise in peroxide anion yield by app.2%.However, afterward, the further increase remained insignificant.Nevertheless, the reaction mechanism was still predominantly a 4-electron reduction to hydroxide, even after 2000 cycles, justifying the efficacy of the developed electrocatalysts.It is noteworthy that the electrocatalyst degradation phenomenon is complex and may involve multiple paths.Their elucidation somehow remains out of the scope of the current study and will be pursued independently in detail to exactly reveal the origin of degradation.
Given the impressive performance in RRDE measurements, both catalysts were studied in AEM (Figure 7 and S10) and PEM (Figure S11) fuel cells.The AEMFC measurements were conducted with 100% humidified O 2 feed to the cathode (0.5 L min −1 ) and hydrogen to the anode (0.3 L min −1 ), and operating at 60°C.The measured open circuit voltage (OCV) in the AEMFC was 0.825 V for SAFe_M_P1AP2 and 0.845 for SEFe_M_P1AP2, respectively.Moreover, SAFe_M_P1AP2 reached a peak power density of 219 mW cm −2 at a current density of 500 mA cm −2 , whereas SEFe_M_P1AP2 only reached 103 mW cm −2 at a current density of around 230 mA cm −2 (Figure 7).
The observed discrepancy in fuel cell performance, as depicted in Figure 7 and S12, evidences that SAFe_M_P1AP2 performs much better in fuel cell conditions, in comparison with SEFe_M_P1AP2.
,63 Since no peaks associated with the Fe(II)/Fe(III) transition were observed in the CV in AEMFC, we examined the active site in PEMFC.After measuring the I-V and CV polarizations, FTacV measurements were conducted, in situ in a fuel cell.The potential of the redox peaks is E 0' = 0.76 V vs. RHE, measured by the CV and the FTacV measurements.Figure S12 demonstrates the 5 th , 6 th and 7 th FTacV harmonics and their fitting simulations.The EASD was calculated from the 5 th harmonic peak current for both electrocatalysts.The EASD observed for SAFe_M_P1AP2 is 1.47x10 19 sites g -1 , while SEFe_M_P1AP2 EASD is 1.14x10 19 sites g -1 .These results strongly support the higher fuel cell performance observed with SAFe_M_P1AP2.Besides, a defined and uniform pore size distribution, as was demonstrated by SAFe_M_P1AP2, ensures good accessibility and exposure of the active sites, facilitating improved mass transport of reactants to the active sites, ultimately enhancing the performance in fuel cell.Finally, both the XRD and HRTEM of SAFe_M_P1AP2 showed a higher graphitization that provides a conductive and stable framework for the active sites, facilitating efficient electron transfer and catalytic reaction kinetics during the ORR.
Thus, both the specific pore size distribution and increased graphitic surface area lead to a greater EASD, as was evaluated by FTacV, offering more reactive areas for the ORR to occur.These morphological differences are significantly increasing the availability of overall catalytic activity by SAFe_M_P1AP2, ensuring effective and efficient electrochemical conversion of oxygen in the fuel cell environment.

Conclusions
This study highlights the crucial role of tailored material design, structural characteristics, and synthesis methods in shaping electrocatalytic performance for ORR in AEMFCs.Through a range of analytical techniques, the investigation elucidated the impact of controlled integration of silica and metal, removal of templating agents, and the use of nitrogenrich organic precursors on electrocatalyst morphology, porosity, and chemical composition.
The results indicate substantial improvement in ORR performance as fabrication stages progress, illustrating the potential of PGM-free electrocatalysts in facilitating ORR in AEMFCs.The enhanced ORR activity was notably observed in the final stage (P1AP2) of the development process.The analysis emphasized the significance of specific pore size distribution and increased graphitic surface area in enhancing the EASD and fuel cell performance in AEMFCs.This study also highlights the successful synthesis methods using combined pyrolytic processes and templating strategies to enhance electrocatalytic activity.Overall, this work offers insights into the controlled materials design and tailored synthesis methods necessary to optimize electrocatalysts for improved ORR performance and their potential for green energy applications.

Figure 4 :
Figure 4: Pore size distribution calculated using DFT method for nitrogen adsorption in slit pores of (a) SEFe_M_P1P2; (b) SAFe_M_P1P2

Figure 6 :
Figure 6: ORR measurements in 0.1 M KOH demonstrated by electrocatalysts evolved at final stage of the research design i.e. after second pyrolysis (P1AP2); compared to Pajarito Powder®.(a) ring current densities (top) and LSVs obtained at 5mVs -1 (bottom), (b) number of electrons transferred (top) and peroxide anion yield (bottom) during ORR while keeping the electrocatalyst loading at 0.6 mg cm -2 and rotation of RRDE at 1600 rpm in O 2 rich 0.1 M KOH.