Evaluation of the Specific Activity of M−N−Cs and the Intrinsic Activity of Tetrapyrrolic FeN4 Sites for the Oxygen Reduction Reaction

Abstract M−N−C electrocatalysts are considered pivotal to replace expensive precious group metal‐based materials in electrocatalytic conversions. However, their development is hampered by the limited availability of methods for the evaluation of the intrinsic activity of different active sites, like pyrrolic FeN4 sites within Fe−N−Cs. Currently, new synthetic procedures based on active‐site imprinting followed by an ion exchange reaction, e.g. Zn‐to‐Fe, are producing single‐site M−N−Cs with outstanding activity. Based on the same replacement principle, we employed a conservative iron extraction to partially remove the Fe ions from the N4 cavities in Fe−N−Cs. Having catalysts with the same morphological properties and Fe ligation that differ solely in Fe content allows for the facile determination of the decrease in density of active sites and their turn‐over frequency. In this way, insight into the specific activity of M−N−Cs is obtained and for single‐site catalysts the intrinsic activity of the site is accessible. This new approach surpasses limitations of methods that rely on probe molecules and, together with those techniques, offers a novel tool to unfold the complexity of Fe−N−C catalyst and M−N−Cs in general.


Physical characterizations
SEM images were taken with a JEOL JSM-IT200 equipped with a EDX detector. N2-sorption porosimetry measurements were performed on a Quantachrome Autosorb iQ2 after outgassing the samples at 250 °C under vacuum overnight prior to the measurements. Brunauer-Emmett-Teller (BET) theory was employed to determine the specific surface area using the Micropore BET Assistant supplied by Quantachrome ASiQwin software. Pore size distributions were calculated with the quenched-solid density functional theory (QSDFT) method (slit/cylindrical pores, adsorption branch). Mössbauer measurements at T = 4.2 K were performed on a standard transmission spectrometer using a sinusoidal velocity waveform with both the source of 57 Co in rhodium and the absorber in the liquid He bath of a cryostat. In order to refer the measured isomer shifts to α-Fe at ambient temperature, 0.245 mm s −1 was added to the measured values. Both X-ray absorption near edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) measurements at the Fe K-edge (7112 eV) were carried out at the BAMline [3] located at BESSY-II (Berlin, Germany), operated by the Helmholtz-Zentrum Berlin für Materialien und Energie. Due to the low concentration of Fe (< 1 wt. %) the measurements were performed in fluorescence mode, with a 4-Element silicon drift detector, using backscattered geometry. For a good S/N ratio, the measurement time was optimized to have at least a total of 500 kcts at the Fe-Kα fluorescence peak. The scans were performed in 10 eV steps until 20 eV before the edge, followed by 0.5 eV steps until 50 eV above the edge, then in 1 eV steps until 200 eV, and from then on in 0.04 Å equidistant k-steps. XANES data evaluation and treatment was performed by using ATHENA program from Demeter package. [4] This includes background removal, energy calibration with Fe metal foil spectrum, and pre-and post-edge normalization. Further information on the local coordination environment was achieved by EXAFS. EXAFS curves were Fourier Transformed between 2-9 Å with a Hanning-type window, to obtain a radial distribution-like information. These were fitted with a model adapted from our previous publication [2] based on a (OH)2FeN4C52H20 cluster consisting of a Fe atom binding two OH groups in the axial positions and coordinated to four pyrrolic nitrogen atoms embedded in a planar carbon plane. The model was used to calculate the scattering paths by FEFF to be able to quantify the coordination number and bondlength. The goodness of the fit is determined by evaluating the reduced-chi2 test and R-factor. The fiting range was kept between 1-4 Å that includes all the scattering paths displayed in the tables S3 and S4. The degeneracy of the scattering path (which corresponds to the coordination number, N, in the case of single scattering paths) was varied until an amplitude reduction factor (S0 2 ) of about 1 was achieved. The best results are displayed the tables S3 and S4. Following amplitude reduction factors were obtained for the two samples:

Electrochemical measurements
Catalyst inks were prepared by dispersing 5 mg of catalyst in 1.68 mL of N,N-dimethylformamide and 50 μL of 5 wt% Nafion suspension, followed by sonication. To obtained a catalyst loading of 145 μg/cm 2 , 10 μL of ink was drop-cast onto a well-polished glassy carbon electrode and dried under an infrared heater for 60 min. The obtained electrodes were measured in a three-electrode glass cell using 0.1 M HClO4 as electrolyte, Au wire as the counter electrode and a freshly calibrated RHE as the reference electrode. The solution resistance was determined by electrochemical impedance spectroscopy. The ORR curves were corrected for capacitive contribution by subtracting from the curves recorded in O2-saturated electrolyte the ones recorded in Arsaturated electrolyte. For each curve at least two separate measurements were averaged to give the shown polarization curve, and the standard deviation is illustrated with error bars. Kinetic currents were calculated based on limiting current correction as in reference [5], but may also be evaluated based on Koutecký-Levich analysis.

Calculation of TOF values
For calculation of TOF values from the pristine Fe-N-C and the longest leached Fe-N-C, equation (1) from the manuscript was used. The pristine Fe-N-C contains 8 % of oxidic iron, which is considered in the calculation.
Graphical determination of the TOF value from linear interpolation of the pristine Zn-N-C, the pristine Fe-N-C and two extracted Fe-N-Cs reveals TOF = 0.24143 ± 0.01848 ≈ 0.24 s -1 ( Figure S3).

Calculation of utilization factor
The utilization factor is defined as the fraction of accessible Fe active sites and the total amount of iron. For convenience both values are usually expressed as molar concentrations. If the catalyst has a homogeneous composition with a known specific activity (mass activity or TOF), it may also be defined as the fraction of the measured activity and the theoretical activity, e.g. the fraction of kinetic current density and the theoretical kinetic current density.
Herein the utilization factor can be derived from measured kinetic current density and the product of the Faraday constant, the molar iron concentration and the calculated TOF for tetrapyrrolic Fe-N4 sites.
The theoretical kinetic current can be calculated as follows: From the calculated kinetic current density, finally the utilization factor can be calculated:      Table S3. Structural information obtained from EXAFS by fitting the nearest coordination shells around Fe atoms in [FeN4]N32C520 with model (OH)2FeN4C52H20: degeneracy of the scattering path (N), interatomic distance from the fit (R) and from the model (Reff), and Debye-Waller factor (σ 2 ). The goodness-of-fit parameter is indicated by the R-factor.  Table S5. Surface area (SA) and pore volume (PV) of the sample in its pristine state and after partial Feextraction. For the SA, both the value obtained by applying the BET theory (SABET) and the one from QSDFT calculation with slyt and cylindrical pores (SAQSDFT) are reported. For the PV, both the value from QSDFT (PVQSDFT) and the one measured at P/P0 ~ 0.99 (TPV) are reported. Micro indicates pores ≤ 2 nm and meso pores between 2 nm and 33 nm (upper value of the employed model).