Identification of Durable and Non-Durable FeNx Sites in Fe-N-C Materials for Proton Exchange Membrane Fuel Cells

Abstract

While Fe-N-C materials are a promising alternative to platinum for catalyzing oxygen reduction in acidic polymer fuel cells, limited understanding of their operando degradation restricts rational approaches towards improved durability. Here we show that Fe-N-C catalysts initially comprising two distinct FeNx sites (S1 and S2) degrade via the transformation of S1 into iron oxides while the structure and number of S2 were unmodified. Structure-activity correlations drawn from end-of-test 57Fe Mössbauer spectroscopy reveal that both sites initially contribute to the ORR activity but only S2 significantly contributes after 50 h of operation. From in situ 57Fe Mössbauer spectroscopy in inert gas coupled to calculations of the Mössbauer signature of FeNx moieties in different electronic states, we identify S1 to be a high-spin FeN4C12 moiety and S2 a low- or intermediate spin FeN4C10 moiety. These insights lay the ground for rational approaches towards Fe-N-C cathodes with improved durability in acidic fuel cells.<br>

Potential‐Induced Spin Changes in Fe/N/C Electrocatalysts Assessed by In Situ X‐ray Emission Spectroscopy

“…Furthermore, the combination of the low temporal resolution of MS and the low metal contents of M‐N x sites (typically <2 wt. %) in M/N/C catalysts leads to in situ MS‐measurements requiring tens of hours of spectral acquisition, [20] during which the sample's speciation can significantly vary [21, 22] . On the other hand, in situ XAS measurements routinely require much less than 1 h per spectrum and (unlike MS) are not limited to Fe‐based samples, and thus this technique has been extensively applied to the study of M/N/C materials [19, 23] .…”

“…The Fe‐speciation of these two catalysts was first examined by means of MS measurements at room temperature (see Figure 1 a and the Supporting Information, Figure S2 for the spectra of DW21 vs. Fe0.5, respectively, and the Supporting Information, Table S2 and Figure S3 for a summary of the deconvolution constrains and derived Mössbauer parameters). Both samples’ spectra were devoid of any sextets or singlets assignable to nanometric‐sized Fe‐based particles, [20, 24] and could instead be deconvoluted using three doublets whose sample‐specific relative contents are summarized in Figure 1 b. The assignment of these doublets to Fe‐N x sites with different structures and electronic states is a matter of vivid discussion, [18, 36–38] and is further complicated by the recent observation that they may also contain contributions from particulate phases that are only resolved at ultra‐low temperatures (≤5 K) [15, 36] …”

Active Site Identification in FeNC Catalysts and Their Assignment to the Oxygen Reduction Reaction Pathway by In Situ57Fe Mössbauer Spectroscopy

“…Publications on in situ MS on FeNC catalysts of FeN 4 /C model systems (macrocyclic based) are rare. [ 48,49 ] Scherson et al [ 50 ] studied the iron phthalocyanine (FePc) on Vulcan XC‐72 carbon materials by in situ Mösssbauer, in which a new doublet appeared with δ iso = 1.14 mm s −1 and Δ E Q = 2.85 mm s −1 , the origin of which remains not fully understood. It was assumed to be caused by demetalation of the macrocycle and subsequent formation of FeOOH further reduced to Fe(OH) 2 .…”

“…In situ MS applied on an Fe 0.5 metal‐organic frameworks‐based catalyst was performed by Jaouen's group when coupled to a gas diffusion electrode. [ 49 ] At low‐potential conditions, two iron sites were formed: D1L (L assigned to low potential, δ iso = 0.67 mm s −1 , Δ E Q = 1.99 mm s −1 ) that showed a reversible switching behavior and another site, labeled D3 ( δ iso = 1.15 mm s −1 , Δ E Q = 2.5 mm s −1 ) that showed an irreversible change. Based on this, the authors concluded that D1L and D3 both originate from the same initial D1H doublet (H assigned to high potential and the same as the ex situ D1 discussed earlier), but D3 is unstable and would form ferrous iron oxide moieties.…”

“…Herein, we synthesized three catalysts by different preparation routes, namely, FeNC phen , FeNC ppy , and FeNC porph , to see if common conclusions for all three catalysts can be made, and to understand to what extent the findings resemble those made recently for Fe 0.5 . [ 49 ] After a detailed ex situ characterization, we will first deduce what changes can be expected based on Mössbauer theory and then correlate the observed changes to specific iron sites to draw conclusions on the reaction mechanism. This is assisted by density functional theory (DFT) calculations that evaluate the properties of iron ions with different coordination environments as plausible models for the most likely transitions.…”