Non-PGM M-N-C ORR Catalysts: Structure, Morphology and Reactivity

Among the state-of-the-art non-platinum group metal (non-PGM) catalysts for oxygen reduction reaction (ORR) is a class of nano-structured materials composed of a transition metal, nitrogen and carbon, thus often abbreviated as M-N-C.¹ The synthesis of those materials usually involves pyrolysis, a high-temperature treatment in inert atmosphere, that leads to reactive decomposition of the catalysts precursor and forming of the active material. There are several ways to synthesize those materials, the most traditional one being the pyrolysis of macrocyclic compounds. For quite some time polymeric precursors that contain nitrogen in the backbone or the side chain have been used as a common carbon and nitrogen precursor and have been pyrolyzed with a transition metal precursor (usually a salt). Later, the heat treatment of mixture of transition metal precursor and source of carbon (often a carbon black) and a nitrogen source such as ammonia gas, or impregnated cyanamide, urea or amines of various types has been practiced. In all these synthesis routes, it is hypothesized in this work, that if the optimization is being made with the performance in ORR as a criterion, the resulting material not only displays similar catalytic performance, but also is essentially composed of the same moieties. All these catalysts are materials with composite characteristics comprised mostly of carbon matrixes with various amounts (and dispersion) of ordered graphitic structures and amorphous carbon content. This carbon matrix is doped with nitrogen of several distinct chemical types: pyrrolic, pyridinic, graphitic and quaternary. Most of the catalysts display also a plurality of surface oxygen groups, associated with carbon, nitrogen or the transition metal. The transition metal is present in a form of reduced metallic nanoparticles, often coated with the native oxides, carbides or dispersed metal oxides of various kinds. Small part (usually around 1% wt.) of the transition metal is shown to be associated with the nitrogen defects (moieties) in the carbon matrix and is displayed as a coordinated transition metal ion. Over the last few years, most of the catalysts of this family have been synthesized with an acid leach and thermal post-processing (secondary pyrolysis), a process that significantly improves the ORR performance.

Since 2000 UNM has been developing an original method of catalyst synthesis that is based on templating of mono-dispersed of hierarchically structured silica particles:  the Sacrificial Support Method (SSM). This method allows producing catalysts with an open-frame morphology at the meso-scale and at the same time eliminates the “formal” support, thus the entire catalyst is the “active material”.² During the pyrolysis the support ensures the catalyst dispersion and is responsible for the resulting high-surface area. After the pyrolysis, silica (the sacrificial support) is removed by dissolving in KOH or HF, resulting in self-supported M-N-C catalyst. We have demonstrated catalysts made by SSM derived from porphyrins^2-5, heteroamines^6,8 and polymers.⁷ We have studied these catalysts intensely by XPS² and EXAFS/XANES⁴ and have made suggestions on the ORR mechanism.^5,8

Exceptionally active catalysts were developed based on the SSM and derived from Fe salt and N-containing amines or even charge-transfer organic salts that are not soluble in water or usual organic solvents. The catalyst nitrogen precursors do not form any complexes with the transition metal and thus defy the standard assumption that the transition metal is being bond by the nitrogen-containing species at the precursor stage and “survives” the pyrolysis conditions. In the case of the new generation UNM non-PGM catalyst, the silica support, the N-C precursor and the iron salt are mixed mechanically by ball-milling. There is no dissolution or impregnation step at all. The cathode catalysts, obtained in this mechano-chemical synthesis meet DOE-EERE non-PGM targets of performance in PEMFC single MEA tests: 0.8 V at 0.1 A at 1.5 Bar total pressure (H₂/O₂test).

There is two strictly different classes of M-N-C catalyst depending on the precursor, but even to a larger extend to the “temperature-time trajectory” of the pyrolytic treatment: (i) low-temperature catalysts in which the nitrogen is predominantly of pyrrolic type and (ii) high-temperature catalysts that show abundance of pyridinic nitrogen. The first type of M-N-C is notable for ORR catalyzed by 2x2e^- ORR mechanism (with hydrogen peroxide as detectable intermediate) and the second type can be optimized to attain to a 4e^-, direct ORR mechanism. With increase of the pyrolysis temperature (or time) above certain critical value most of the nitrogen moieties are being transformed into graphitic type and the catalyst loses activity.

Understanding of the nature of active sites in non-PGM catalysts is difficult due to high heterogeneity of such materials. SSM provides a unique opportunity to study the structure and reactivity of M-N-C catalysts as there is no “support effect” and the activity exceeds the state-of-the-art in non-PGM. All the M-N-C catalysts display plurality of active sites that have reactivity in ORR. Some of these sites support oxygen reduction to peroxide, some are active in peroxide reduction to water, some are active in both reactions and some may support a 4e^- reduction of oxygen to water. In the actual catalyst a plurality of such moieties is presents and they all contribute to the integral activity. In this work we introduce the approach of interpreting the structure of the active sites as defects in graphene. Detailed spectroscopic and morphology analysis revealed correlations between chemical species and ORR activity providing sufficient information on the nature of active sites. We will report on the role of the in-plane Fe-N₄ and Fe-N₃ defects; Fe-N₂ and Fe-N₂₊₂edge defects; metal-less N and O moieties as well as reduced metal and metal oxide nano-phases and will show their correlation with the catalyst activity in ORR individual reactions. Relative contributions of these different active sites to the overall cathode process will be discussed.

Acknowledgments: This work is supported by DOE-EERE Fuel Cell Technology Program: “Development of Novel Non Pt Group Metal Electrocatalysts for PEMFC” (S. Mukerjee, PI).

References:

1. F. Jaouen et al., Energy Environ. Sci., 4 (2011) 114
2. K. Artyushkova et al., Topics in Catalysis, 46 (2007) 263
3. S. Pylypenko et al., Electrochimica Acta 53 (2008) 7875
4. J.M. Ziegelbauer et al., J. Phys. Chem. C 112 (2008) 8839
5. T.S. Olson et al., J. Electrochem. Soc. 157 (2010) B54
6. A. Serov et al., Electrochem. Comm. 22 (2012) 53
7. A. Serov et al., Appl. Catalysis B 127 (2012) 300
8. M.H. Robson et al., Electrochim. Acta, 90 (2013) 656