Synthesis of the Zn-based zeolitic imidazolate framework (ZIF8)In our previous work8, Zn(NO3)2·6H2O (10 mmol, 2.975 g) and 2-mIm (80 mmol, 6.568 g) were each dissolved in 100 mL of methanol using ultrasound for 5 min. The Zn(NO3)2 solution was then quickly added to the 2-mIm solution. The mixture was vigorously stirred for 16 h at room temperature. The resulting white ZIF-8 precipitate was centrifuged, washed multiple times with methanol, and dried under vacuum at 60 ⁰C overnight.Synthesis of NC1.0 g ZIF-8 and 0.3 g 1,10-phenanthroline were dispersed in a 2:1 ethanol and deionized water solution. The mixture was stirred magnetically for 12 h at room temperature, then evaporated in an 80 °C oil bath. The dried powders were ground thoroughly and pyrolyzed under Ar at 1000 °C (5 °C min–1) for 1 h, then cooled naturally to room temperature. The resulting black products were named NC.Synthesis of xFe-NC-T
An appropriate amount of FeCl2·4H2O and 100 mg NC (with Fe feed ratio converted to x) were thoroughly ground in an agate mortar. The xFe-NC-T was synthesized by heating the mixture in an Ar atmosphere at a specific temperature (T) for 1 h (5 °C min–1).Synthesis of 0.015FeOx-NC-900An appropriate amount of FeOx and 100 mg NC (with Fe feed ratio converted to 0.015) were thoroughly ground in an agate mortar. The 0.015FeOx-NC-900 was obtained by heating the mixture in an Ar atmosphere at 900 °C for 1 h (5 °C min−1).In situ heating TEMFor in situ analysis, the as-prepared 0.015Fe-NC-T was firstly dispersed into deionized water. The suspension was sonicated for 20 min at room temperature, and then deposited directly onto a homemade heating chip purchased from CHIP-NOVA company. The in situ heating TEM experiments were performed on TECNAI F20 at a unit acceleration voltage of 200 kV in a vacuum atmosphere and a heating control system. The TEM specimen was heated from 20 to 1000 oC. To observe the temperature range of the variation of 0.015Fe-NC-T, the temperature of 500 oC, 800 oC and 1000 oC was maintained for 10 min–30 at each increment and the image was taken when the sample was stable.In situ heating XRDX-ray power diffraction (XRD) patterns were taken on a Rigaku Ultima IV diffractometer (Rigaku, Japan) with Cu Kα X-ray source. The sample was fully ground and filled to a ceramic sample groove, and then transferred into the quasi in situ reaction cell for treatment under various conditions that were comparable with the real reaction conditions. The range of 30°–50° was the position of the main peak of Fe species, so it was selected to analyze the transformation process of Fe species during heat treatment. The scanning rate was 1° min−1 and the step length is 0.02° min–1. During the test, Ar was continuously injected, the flow rate was 15 mL min–1, the heating rate was 10 °C min–1, and the constant temperature time of each temperature was 5 min.Quasi in situ heating XPSX-ray photoelectron spectroscopy (XPS) measurements were conducted on an Omicron Sphera II Hemispherical electron energy analyzer with monochromatic Al Kα radiation (1486.6 eV) operated at 15 kV and 300 W. The base pressure of the systems was 5.0 × 10–9 mbar. The sample was placed into the sample tank, compressed, secured on the stainless steel sample frame, and then transferred to the quasi in situ reaction pool. It was heated to 500 oC at a rate of 10 oC min–1 in an Ar atmosphere for 10 min. Subsequently, the product is to be transferred to the vacuum chamber for XPS test.In situ heating XASThe in situ heating XAS measurement was performed at beamline 1W1B of the Beijing Synchrotron Radiation Facility (BSRF). The power precursors were fitted into a ceramic groove custom-made with an 8 mm internal diameter and 1.5 mm depth. The Fe K-edge spectra were recorded in fluorescence mode and an iron foil spectrum was measured simultaneously with each sample spectrum for energy calibration. Transmission data was taken from 6950 eV to 7940 eV. Spectra were taken at room temperature, then the quartz tube was heated at a ramp rate of approximately 20 °C min–1 under flowing nitrogen to approximately the following temperature set points during the increasing temperature portion of the temperature profile: room temperature, 500 oC, 700 oC, 800 oC and 900 oC. A XAS scans was collected at each temperature set point. The hold time at each temperature was approximately 10 min. The XAS data were processed and fitted using the Ifeffit−based Athena and Artemis programs. Scans were calibrated, aligned, and normalized with background removed using the IFEFFIT suite.Electrochemical measurementsAll electrochemical measurements were conducted using a CHI 760e electrochemical workstation with a three-electrode setup. The working electrode was the catalyst-modified glassy carbon electrode, with a graphite rod as the counter electrode and a saturated calomel electrode (SCE) as the reference. To prepare a uniform catalyst ink, 6 mg of catalyst was sonicated for 30 min in 1 mL of a mixture containing 600 μL isopropanol, 380 μL ultrapure water, and 20 μL of 5 wt% Nafion solution. For the commercial 20 wt% Pt/C sample, 1 mg of catalyst was dispersed in 1 mL of 0.05 wt% Nafion solution. A precise volume of the catalyst ink was then applied onto the polished glassy carbon rotating disk electrode (RDE, diameter 5 mm, area 0.196 cm²) or rotating ring-disk electrode (RRDE, diameter 5.61 mm, area 0.2475 cm²) to achieve the desired catalyst loading. The catalyst loading was 0.6 mgtotal cm⁻² for M-N-C and 0.012 mgPt cm⁻² for Pt/C.ORR polarization curves were performed at 30 °C in 0.1 M H2SO4 solution. RRDE measurements utilized linear sweep voltammetry (LSV) in the range of 0.1 to 1.1 V (vs. RHE) at 900 rpm with a scan rate of 10 mV s⁻¹ with IRS compensations, while the ring electrode was held at 1.2 V (vs. RHE). The resistance is automatically compensated by 80 %, and the resistance in the impedance spectrum mode is measured by the pine system. The electrolyte was measured by pH instruments to ensure a constant test environment (pH = 0.7 for 0.1 M H2SO4). All potentials were converted to reversible hydrogen electrode (RHE) potentials using the equation:$${E}_{{RHE}}={E}_{{SCE}}+0.2415+0.059\times {pH}-{{IR}}_{S}$$
(1)
The electron transfer number (n) and H2O2 percentage were calculated using the following equations:$$n=\frac{4\times {I}_{D}}{{I}_{D}+{I}_{R}/N}$$
(2)
$${{{{\rm{H}}}}}_{2}{{{{\rm{O}}}}}_{2}\left(\%\right)=\frac{200\times {I}_{R}/N}{{I}_{D}+{I}_{R}/N}$$
(3)
where ID is the disk current, IR is the ring current, and N (0.37) is the current collection efficiency of the Pt ring.The kinetic current densities (jk) involved during the ORR process were determined by analyzing Koutecky-Levich (K-L) Eq. (4):$$\frac{1}{j}=\frac{1}{{j}_{k}}+\frac{1}{{j}_{L}}$$
(4)
where j is the measured current density, jL and jk are the limiting and kinetic current densities.Quantification of active Fe–N4 sitesThe SD was determined using the nitrite reduction method by Kucernak35. Briefly, nitrite forms stable poisoned adducts with Fe metal centers, which can be completely stripped between 0.35 to −0.35 V (vs. RHE). The excess coulometric charge (Qstrip) from the stripping peak is proportional to SD:$${{{\rm{SD}}}}\left({{{\rm{sites}}}}\, {{{{\rm{g}}}}}^{-1}\right)=\frac{{Q}_{{{{\rm{strip}}}}}\left({{{\rm{C}}}}\,{{{{\rm{g}}}}}^{-1}\right)\times {N}_{A}({sites}\, {{mol}}^{-1})}{{n}_{{{{\rm{strip}}}}}{{{\rm{F}}}}({{{\rm{C}}}}\,{{{{\rm{mol}}}}}^{-1})}$$
(5)
$${{{\rm{TOF}}}}\left({{{{\rm{s}}}}}^{-1}\right)=\frac{{n}_{{{{\rm{strip}}}}}\,{\Delta j}_{{{{\rm{k}}}}}\, ({{{\rm{mA}}}}\, {{{{\rm{cm}}}}}^{-2})}{{Q}_{{{{\rm{strip}}}}}\left({{{\rm{C}}}}\, {{{{\rm{g}}}}}^{-1}\right){L}_{{{{\rm{C}}}}}({{{\rm{mg}}}}\,{{{{\rm{cm}}}}}^{-2})}$$
(6)
where nstrip (=5) is the number of electrons transferred per nitrite stripped. NA is Avogadro constant (6.02 × 1023 sites mol−1). F is Faraday’s constant (96485 C mol−1). LC was the catalyst loading (0.242 mg cm−2). The catalysts were tested without further treatment.PEMFC testsMembrane electrode assemblies (MEAs) were made using the hot-pressing method. The cathode ink was prepared by ultrasonically mixing the required amounts of Fe-NC catalysts, deionized water (0.2 mL), isopropanol (0.8 mL), and a 5 wt% Nafion solution in an ice bath for 1 h. This ink was applied to a gas diffusion layer (GDL, PTFE-pretreated Toray 060 carbon paper) with a loading of 3.5 mg cm–2. The Nafion content in the cathode layer was about 50 wt%. The anode catalyst was 40 wt.% Pt/C with a loading of 0.4 mgPt cm-2. The MEA was assembled by hot-pressing the cathode, anode, Nafion membrane (NRE 211), and a gasket at 135 °C and 3 MPa for 2 min. The active area of the MEA was and 2.1 × 2.1 cm2. Polarization curves were obtained at 80 °C using a Model 850e fuel cell test system (Scribner Associates, Inc.) in conjunction with an absolute pressure of 1.5 bar. The H2 and O2 (air) flow rates were 0.3 L min–1 and 0.4 L min−1 at 100% RH during measurements.Computational methodsAll spin-polarized DFT computations were carried out with the VASP37. The PBE functional was applied in combination with the Van der Waals interaction38,39. The kinetic cut-off energy was 400 eV. The Fe13O13 cluster structure was firstly relaxed in the NVT ensemble at the temperature of 1170 K for 5 ps using the Nose-Hoover thermostat40,41. Accordingly, the structure of the Fe13O13 cluster on the N4-C was constructed, and the vacuum region was 12 Å at least between the periodic images. The whole structure was optimized by using a 2 × 2 × 1 k-point mesh in the Monkhorst-Pack scheme. All atoms were allowed to relax during the geometry optimization. The energy convergence criterion was 10-5 eV, and the final force was less than 0.01 eV Å−1 at each atom. The nudged elastic band method was used to study the minimum energy path of the Fe atom transfer and the FeN4-C formation. The transition state was further searched by the dimer method42,43; the force convergence was 0.05 eV Å−1, and a single gamma point was applied. The molecular dynamics simulation was performed to study the reaction between the Fe13O13 cluster and the N4-C site leading to the FeN4-C formation at 1180 K. A time step of 0.5 fs was used, and the structure was preequilibrated for 0.5 ps before sampling. A single gamma point was applied in molecular dynamics simulations.
An in situ exploration of how Fe/N/C oxygen reduction catalysts evolve during synthesis under pyrolytic conditions
