Elucidation of Fe-N-C electrocatalyst active site functionality via in-situ X-ray absorption and operando determination of oxygen reduction reaction kinetics in a PEFC

Highlights

• ORR kinetics was determined for a PGM-free catalyst in a PEFC environment.

• Exchange current density, activation energy, and reaction order were determined.

• Kinetics are a function of PGM-free active site redox potential and oxidation state.

• XANES for a PGM-free catalyst were examined in a PEFC environment.

• Model values for redox potential and oxidation state agree with XANES measurements

Abstract

In the past decade the notable effort placed on improving intrinsic electrochemical kinetics of platinum group metal (PGM)-free electrocatalysts for the oxygen reduction reaction (ORR) has led to a significant improvement in both performance and understanding of this class of electrocatalysts. However, a limited amount of this development and understanding has been undertaken using operando electrochemical diagnostics at the membrane electrode assembly (MEA) level. In this work, the operando ORR kinetics on an atomically dispersed iron-nitrogen-carbon ((AD)Fe-N-C) PGM-free electrocatalyst have been examined to extract the reaction order and the activation energy of the ORR. The experiments were carefully designed to ensure the stability/predictability of the electrocatalyst during the data collection process and thus validate the relevance of the values obtained for the aforementioned parameters. A kinetic model that considers a potential-dependent availability of active sites (θ) is proposed. Active site availability is shown to be a function of both the change in the oxidation state (n_R) and the redox potential at which the metal center transitions from a higher oxidation state to a lower one (). The resulting model fitting parameters for n_R and (0.71 and 0.788 V, respectively) obtained from the analysis of operando data correlate well with those from in situ X-ray absorption near edge structure measurements (n_R = 0.57) and in situ cyclic voltammetry measurements (0.75 V <   < 0.8 V) in the MEA environment. The resulting model provides an excellent fit of MEA performance across the range of pressures, temperatures, and potentials under which the data were collected.

Introduction

Polymer electrolyte fuel cells (PEFCs) are electrochemical devices that directly convert the chemical energy of a fuel (most often hydrogen) into electrical energy. The high efficiency of this energy conversion, together with the emission of only water, if H₂ is produced using renewable energy, make PEFCs an attractive technology for clean energy applications. In addition, PEFC operating temperatures are close to ambient conditions such that start-up and shutdown are quick. All these features make the PEFC an attractive candidate to replace internal combustion engines for powering vehicles [[1], [2], [3]].

Unfortunately, PEFCs also have several challenges hindering their widespread application in the automotive sector. One of the main drawbacks is their high cost, mainly dictated by the use of Pt (a costly and rare noble metal) to catalyze both the anodic hydrogen oxidation reaction (HOR) and the cathodic oxygen reduction reaction (ORR) [4]. Due to the slower kinetics of the ORR compared to the HOR, the Pt loading in a PEFC is considerably higher on the cathode than on the anode [[5], [6], [7]]. Thus, an enormous amount of research effort has been dedicated to reducing the Pt cathode loading and to replacing Pt with a less costly, platinum group metal-free (PGM-free), earth-abundant material.

The fundamental understanding of PGM-based electrocatalysts has improved as a result of several decades of research dedicated to understanding the ORR mechanism and its limitations. This research has utilized systems ranging from single crystals to dispersed Pt-based electrocatalysts measured using rotating disk electrodes (RDE) [[8], [9], [10], [11], [12], [13]], and single PEFC membrane-electrode assemblies (MEAs) [5,[14], [15], [16]]. While PGM-free ORR catalysts have also been the subject of intensive research for over half a century [2,3,[17], [18], [19]], the heterogeneous nature of these materials has limited the depth of the knowledge of the active site and the ORR mechanism versus that of the PGM electrocatalysts. To that end, several consortia focused on PGM-free electrocatalyst / electrode development for PEFCs were recently established to enable a more integrated and comprehensive analysis of PGM-free-based MEAs [[20], [21], [22]].

For PGM-based electrocatalysts, MEA-level operando studies have enabled the determination of relevant ORR kinetic parameters (i.e., reaction order, activation energy, and exchange current density), which has improved the understanding of potential-driven phenomena, such as the impact of Pt oxide coverage as well as the identification, assessment, and subsequent improvement of both proton and local Pt transport losses [16,[23], [24], [25], [26], [27], [28], [29], [30]]. Despite the increasing interest and recent developments in the field of PGM-free ORR catalysis, similar systematic efforts to obtain operando kinetic information have not yet been undertaken. This may in part be due to the variety of PGM-free catalysts (stemming from a variety of synthesis precursors and approaches) [3,[17], [18], [19]], the resulting distribution of electrochemically- active site turnover frequencies and chemical compositions [18,31,32], difficulty in clearly identifying the precise ORR active site(s) [33,34], and most critically, the inability to obtain steady-state performance data [[35], [36], [37]].

A majority of recent studies have come to a consensus that the ORR active sites in materials derived from a transition metal precursor and a source of carbon and nitrogen, such as a carbon-nitrogen-based polymer (e.g., polyaniline) or metal-organic framework, are comprised of an ionized transition metal atom (e.g., Fe, Co, Mn) coordinated by N, embedded in a carbonaceous matrix [[38], [39], [40], [41], [42]]. Assuming that these N-coordinated ionized transition metal atoms act as the ORR active site(s), various research groups have tried to elucidate the relationship between active site availability and the resulting kinetics. One observation that has been repeatedly reported is the concomitant positions of the redox peak, obtained from cyclic voltammetry, and the ORR onset potential [[43], [44], [45], [46]]. This correlation has been probed by investigating the aforementioned electrochemical characteristics of various Fe-based PGM-free electrocatalysts [47,48].

In this regard, the theory of the “asymmetric volcano” trend was proposed [39,49]. Here, the redox potential of the N-coordinated transition metal atoms was related to the ORR onset potential and the intrinsic ORR activity in terms of turn over frequency (TOF) and O₂ binding energy [39,49]. RDE results have shown a good agreement with this theory, indicating that the transition metal itself, along with its coordination and incorporation in the catalyst’s carbonaceous matrix, play a fundamental role in properly catalyzing the ORR [[50], [51], [52], [53]]. However, while RDE is a powerful investigative tool, it suffers from certain fundamental limitations, namely low dissolved O₂ concentration in the aqueous electrolyte, which limit the ability to acquire critical information at high current density and low cell potential [54]. Hence, it becomes evident that any experiment designed to probe the electrochemical performance across a wide range of potential and current density must be done in an operating PEFC. While operando kinetics have been measured previously for Pt-based catalysts [15,16], to the best of our knowledge such a comprehensive study does not yet exist for PGM-free catalysts.

Recent efforts in PGM-free catalysts development at Los Alamos National Laboratory have yielded a relatively robust atomically dispersed iron ((AD)Fe-N-C) [55,56] PGM-free ORR catalyst, with polarization curves exhibiting little variation within a several hour-long test window.

In this work, we leveraged the limited degradation of the (AD)Fe-N-C electrocatalyst to enable the collection of operando kinetic information, producing, for the first time, MEA-level kinetic parameters (exchange current density, reaction order, activation energy, and active site(s) availability) for a PGM-free electrocatalyst. The operando data was utilized in combination with in situ X-ray absorption near edge structure (XANES) experiments in an MEA to produce a better understanding of PGM-free active site availability/activity as a function of both cathode potential and the change in oxidation state for atomically-dispersed Fe-based PGM-free electrocatalysts, elaborating on previously proposed active site availability models [39,43,57]. The resulting PGM-free kinetic model produced an excellent fit of the experimental data over the complete range of O₂ pressure, temperature, and cell potential for which the experimental data are available.