4.1 Introduction

The increasing demand for energy and serious environmental concerns have motivated the urgent pursuit of clean and sustainable energy conversion technologies as alternatives to traditional thermal engines [1, 2]. Due to the high energy density of hydrogen and its pollution‐free features, fuel cells and water‐splitting electrolyzers that convert the hydrogen energy into electricity are highly efficient and clean energy conversion devices [3–5]. Electrochemical oxygen reduction reaction (ORR) and hydrogen evolution reaction (HER) occur at the cathodes of fuel cells and at water‐splitting electrolyzers, respectively [6–8]. Electrocatalysts play a crucial role in these devices by accelerating the sluggish reaction kinetics of the ORR and HER processes through the reduction of their kinetic energy barriers [9, 10]. To date, platinum (Pt) or Pt‐based alloys have been considered the most active ORR and HER electrocatalysts. Unfortunately, the scarcity and the prohibitively high cost of Pt present a massive barrier for its large‐scale utilization in fuel cells and water‐splitting electrolyzers. Furthermore, Pt‐based electrocatalysts suffer from severe degradation caused by CO poisoning and catalyst agglomeration [11, 12].

The ORR and HER processes in fuel cells and water‐splitting electrolyzers proceed on the surfaces of the electrocatalysts. Two key parameters directly determine the ORR and HER kinetics of the electrocatalysts. First, the mass transfer of reactants (O₂ for the ORR and proton/or water molecules for the HER) from electrolytes to the surfaces of electrocatalysts is an important parameter that is related to the electrocatalyst nanostructure [13, 14]. Second, the adsorption of the reactants and intermediates on the surfaces of electrocatalysts is inherently determined by the electronic structure of the exposed active sites [15]. Because of their excellent electrical conductivity, outstanding chemical and thermal stability, as well as their large specific surface area, carbon materials are an attractive type of electrocatalysts for the ORR and HER processes [16, 17]. The electronic properties of the carbon‐based electrocatalysts can be modulated by the introduction of heteroatoms (e.g. nitrogen, boron, sulfur, and phosphorus) and metal atoms (e.g. nickel, cobalt, and iron) into the carbon skeletons [18, 19]. Currently, porous carbon materials are ubiquitous and indispensable in many important fields including batteries, supercapacitors, membranes, and catalysis [20]. Based on their pore sizes, carbon materials can be classified as microporous (<2 nm), mesoporous (2–50 nm), and macroporous (>50 nm) carbons [21]. In porous carbon‐based nanostructures, micropores and mesopores can greatly increase the amount of the exposed active centers, whereas mesopores and macropores can effectively promote mass transfer during the electrochemical processes [22]. In this respect, the use of hierarchically porous carbon materials featuring mesopores and micropores as alternative electrocatalysts to Pt is highly promising for accelerating the sluggish electrocatalytic reactions in fuel cells and water‐splitting electrolyzers.

In this chapter, we will discuss the design of active sites and porous nanostructures of carbon‐based electrocatalysts, where various metal‐free and noble‐metal‐free, carbon‐based nanostructures have been developed for the electrocatalytic ORR and HER reactions. For the ORR, the molecular structure of the active centers of carbon‐based electrocatalysts (especially nitrogen‐dopant‐based carbon) still remains controversial: (i) it is still unclear whether pyridinic‐N, graphitic‐N, or pyrrolic‐N is the real active site for catalyzing the ORR in the nitrogen‐doped carbon; (ii) the bonding of metal and nitrogen dopants in noble‐metal‐free carbon including metal valence state, number of binding nitrogen dopants, and the type of nitrogen dopants (pyridinic‐N, graphitic‐N, or pyrrolic‐N) is not well understood. In this context, we do not intend to be exhaustive but rather aim to discuss recent progress toward the design and investigation of ORR active sites in porous carbon catalysts. In particular, the construction of the porous carbon electrocatalysts and the correlation between the porous structures and ORR activity are discussed. For the HER, hydrogen adsorption is widely considered the rate‐limited step. Thus, the modulation of the free energy of hydrogen adsorption of the porous carbon catalysts by heteroatoms and metal centers is systematically discussed in this context. We hope that these important discussions will shed light on the exploration of carbon‐based ORR and HER electrocatalysts with a high activity comparable with that of the Pt‐based catalysts.

4.2 Porous Carbon as ORR Electrocatalysts

The introduction of heteroatoms (N, S, B, O, P, and F) and metal atoms (Fe and Co) into carbon frameworks can induce electron modulation and thus offers desirable electronic structures for the ORR process. Consequently, considerable efforts have recently been made for the development of noble‐metal‐free carbon nanomaterials as ORR electrocatalysts. As N dopants are widely considered the most efficient heteroatoms for enhancing the electrocatalytic ORR performance of the carbon‐based catalysts, here we mainly describe some pioneering and outstanding noble‐metal‐free, N‐dopant‐based ORR carbon catalysts. In addition, porous nanostructures in carbon electrocatalysts have been reported for effectively increasing the density of the active sites and facilitating the mass transfer, but controlled synthesis strategies for porous carbons and the relationship between the porous structures and electrocatalytic activities remain elusive.

4.2.1 Metal‐Free Porous Carbon as ORR Catalysts

In the past decade, various nitrogen‐doped carbon materials have been prepared and investigated as metal‐free ORR electrocatalysts. The binding adsorption of oxygen molecules onto the carbon atoms with nitrogen‐dopant neighbors has been broadly accepted as the key step [16, 23]. Unfortunately, the question of which type of nitrogen dopant, such as pyridinic‐N, graphitic‐N, and pyrrolic‐N, is most efficient for activating neighboring carbon remains controversial [24, 25]. Therefore, we will briefly introduce the representative work on the ORR active sites of the metal‐free carbons. Furthermore, the relationship between porous nanostructures and ORR performance of the metal‐free carbon catalysts will be highlighted.

4.2.1.1 Metal‐Free N‐Dopant‐Based Carbon

In 2009, nitrogen‐doped carbon nanotubes were first reported as a metal‐free ORR electrocatalyst that showed enhanced electrocatalytic activity, long‐term stability, as well as better methanol resistance than platinum in alkaline solution [26]. Afterward, nitrogen‐doped ordered mesoporous graphitic arrays (NOMGAs) were synthesized using a metal‐free nanocasting technology [24]. As shown in Figure 4.1a, a nitrogen‐containing aromatic dye (N,N′‐bis(2,6‐diisopropyphenyl)‐3,4,9,10‐perylenetetracarboxylic diimide (PDI)) as the carbon and nitrogen precursor was impregnated into the ordered mesoporous silica (SBA‐15) as a hard template. After pyrolysis at 900 °C in N₂ and subsequent etching‐off of SBA‐15, the NOMGAs were obtained. A mesoporous nanostructure (the specific surface area is as high as 510 m² g⁻¹ and the pore size is approximately 2–5 nm) and a nitrogen‐doped graphitic carbon framework with nitrogen content of ∼2.7 atm% endowed the NOMGAs with high electrocatalytic activity, excellent long‐term stability, and a good methanol resistance for the ORR in 0.1 M KOH solution. Particularly, the electron transfer number of the NOMGAs and the kinetics‐limiting current density reached 3.89 and 9.15 mA cm⁻² at −0.35 V (versus Ag/AgCl), respectively (Figure 4.1b,c). As a result of the metal‐free preparation procedure, the electrocatalytic ORR activity of NOMGAs can be attributed exclusively to the introduction of nitrogen doping into the carbon skeleton. Moreover, the fraction of graphitic nitrogen is as high as 70.9%, 2.5 times higher than 28.1% for pyridinic nitrogen. By comparing NOMGAs derived from different carbon precursors and formed at different pyrolysis temperatures, the existence of graphitic nitrogen atoms was demonstrated to be responsible for the excellent electrocatalytic ORR performance.

To clearly probe the role of the N dopant in the ORR process, tailored graphite model electrocatalysts with well‐controlled doping of N species (defined pyridinic‐N, pyrrolic‐N, or graphitic‐N) were developed to address the ORR active sites under alkaline conditions (Figure 4.2a). A series of electrocatalytic ORR tests, post‐ORR X‐ray photoelectron spectroscopy (XPS) analyses, and CO₂ temperature‐programmed desorption (TPD) measurements demonstrated that carbon atoms neighboring to pyridinic‐N are the ORR active sites of N‐doped carbon materials in alkaline electrolyte (Figure 4.2b) [27]. Nevertheless, considering its low current density (μA cm⁻²), such a nitrogen‐doped graphite model system cannot account for the practical ORR active sites of the reported nitrogen‐doped porous carbon materials with a high kinetic current density (mA cm⁻²). Further modulation of the electronic structure of N‐doped carbon by introducing binary heteroatoms is an effective strategy for improving ORR activity. Accordingly, a porous carbon co‐doped with nitrogen and phosphorus was prepared through the pyrolysis of a polyaniline (PANI) aerogel linked with phytic acid (Figure 4.3a) [28]. The resultant N, P‐co‐doped porous carbon had a large specific surface area reaching ∼1663 m² g⁻¹. The atomic content of N and P in the N, P‐co‐doped porous carbon was approximately 3.2% and 1.1%, respectively (Figure 4.3b). The linear scan voltammogram (LSV) curves confirmed the electrocatalytic performance of the N, P‐co‐doped porous carbon with a positive onset potential of 0.94 V (versus reversible hydrogen electrode (RHE)) and a half‐wave potential of 0.85 V (versus RHE) in a 0.1 M KOH solution, comparable with the values for the commercial Pt/C catalyst (Figure 4.3c). The electron transfer number per oxygen molecule (n) of the N, P‐co‐doped porous carbon was estimated to be ∼4, suggesting a four‐electron transfer ORR process (Table 4.1). Density functional theory (DFT) calculations further revealed that the N, P co‐doping and graphene edge effects are essential for the enhanced ORR activity. The N, P‐doped porous carbon was utilized as an air electrode for the primary and rechargeable Zn–air batteries, delivering the open‐circuit potential of 1.48 V and the specific capacity of 735 mAh gZn⁻¹ (corresponding to the energy density of 835 Wh kgZn⁻¹).

Electrocatalysis	Half‐wave potential (V)	Kinetics‐limited current density (mA cm−2)	Electron transfer number (n)	H2O2 yield (%)
Nitrogen‐doped carbon nanotubes [26]	−0.15 V versus Ag/AgCl	—	3.9 at −0.4 V	—
Nitrogen‐doped ordered mesoporous graphitic arrays [24]	−0.21 V versus Ag/AgCl	9.15 at −0.35 V	3.89 at −0.35 V	—
Graphene‐based carbon nitride nanosheets [29]	−0.25 V versus Ag/AgCl	7.3 at −0.4 V	4.0 at −0.4 V	—
Nitrogen‐doped carbon nanosheets with defined pore size [30]	−0.13 V versus Ag/AgCl	5.68 at −0.28 V	3.67–3.94 at −0.2–0.9 V	—
Nitrogen‐doped carbon nanospheres with tailored pore size [31]	0.78 V versus RHE	5.5 at 0.6 V	3.86 at 0.6 V	0.6 at 0.8 V
Hierarchically porous carbon with optimized nitrogen doping [32]	0.85 V versus RHE	4.32 at 0.8 V	3.92 at 0.8 V	1.5 at 0.8 V
Nitrogen‐doped carbon [33]	0.1 V versus Ag/AgCl	5.6 at −0.35 V	4.0 at −0.35 V	—
Graphene‐based nitrogen‐doped porous carbon sheets [34]	−0.17 V versus Ag/AgCl	6.1 at −0.4 V	3.88 at −0.35 V	10 at −0.35 V
N, P‐co‐doped porous carbon [28]	0.85 V versus RHE	∼26 at −0.65 V	∼4.0	<8%
N, O‐co‐doped mesoporous carbon [35]	∼0.7 V versus RHE	∼29 at −0.63 V	∼4.0 at 0.4 V	—
N, S‐co‐doped mesoporous graphene [36]	−0.3 V versus Ag/AgCl	∼31 at −0.6 V	∼3.6 at −0.6 V	—

4.2.1.2 The Correlation Between Porous Nanostructures and ORR Activity

In addition to the modulation of the nature of nitrogen doping described above, designing the porous structure and pore size of carbon electrocatalysts is another important strategy for enhancing the ORR activity of electrocatalysts. To address the influence of mesopore size on the electrocatalytic ORR activity of carbon electrocatalysts, a series of nitrogen‐doped carbon nanosheets (NDCNs) with uniform and tunable mesopores were prepared and investigated, where the nitrogen‐doping content and the type of nitrogen dopant were kept the same [30]. Porosity control of the NDCN was realized by the electrostatic assembly of colloidal silica nanoparticles with different sizes on the surfaces of previously prepared graphene oxide/silica nanosheets. After surface growth of polydopamine, and following pyrolysis and etching of the silica template, the NDCN with size‐defined mesopores was synthesized. Nitrogen isothermal adsorption/desorption measurements confirmed that the pore size of the resultant NDCNs could be adjusted from 2 to 22 nm (Figure 4.4a,b). Despite the increased pore size, the half‐wave potential of the NDCN showed a positive shift. The NDCN with the mesopore size of approximately 22 nm showed ORR activity with a higher onset potential of −0.01 V than −0.02 V for the Pt/C catalyst in a 0.1 M KOH aqueous solution (Figure 4.4c).

In another example, nitrogen‐doped mesoporous carbon nanospheres (N‐MCNs) were prepared through self‐assembly of colloidal silica in the presence of PANI, allowing for the simultaneous control of morphology and pore size [31]. The prepared N‐MCNs had adjustable mesopores with the size in the 7–42 nm range and the correspondingly varying specific surface area values (7 nm: 1117 m² g⁻¹; 22 nm: 926 m² g⁻¹; 42 nm: 785 m² g⁻¹) (Figure 4.4d,e). When the N‐MCNs were employed as metal‐free electrocatalysts for the ORR in an alkaline electrolyte, the N‐MCNs with 22 nm mesopores exhibited the highest ORR performance with a half‐wave potential of 0.78 V (versus RHE), a high electron transfer number of 3.86, and an excellent long‐term cycling stability (Figure 4.4f). These experimental results clearly revealed that the ORR activity of nitrogen‐doped carbon electrocatalysts is strongly correlated with the pore size of the mesopores, benefiting from rapid mass transport.

The mesopores of nitrogen‐doped carbon electrocatalysts can efficiently promote the mass transport during the ORR process, but the increase in the active site density is limited. Therefore, constructing nitrogen‐doped carbon electrocatalysts with hierarchical porous structures including micropores and mesopores is highly desirable for greatly enhancing the active site density and promoting the mass transfer of the ORR electrocatalysts. In this regard, the nitrogen‐doped, metal‐free carbon electrocatalysts featuring an abundance of both micropores and mesopores were synthesized [32]. The metal‐free mesoporous carbon electrocatalysts (denoted as meso‐PoPD) were first prepared by combining hard templating synthesis with pyrolysis of nitrogen‐enriched poly(o‐phenylenediamine) as precursors. Then, the micropores (meso/micro‐PoPD) were created in the meso‐PoPD utilizing NH₃ activation (Figures 4.5 and 4.6a,b). Clearly, both meso‐PoPD and meso/micro‐PoPD exhibited a steep and high capillary condensation steps, revealing uniform and well‐constructed mesoporosity with large mesopore volume. Following NH₃ activation, the Barrett–Emmett–Teller (BET) surface area increased dramatically from 685 m² g⁻¹ for the meso‐PoPD to 1280 m² g⁻¹ for the meso/micro‐PoPD without substantial deterioration of mesoporosity (Figure 4.6c). The pore size distribution (PSD) plots indicated that the meso/micro‐PoPD exhibited ultramicropores (0.6 nm), micropores (1.4), and mesopores (12 nm), respectively (Figure 4.6d), confirming the hierarchical porous structure where the micropores were created in the walls of the well‐defined mesopores. For comparison, the non‐PoPD prepared without using a colloidal silica template was examined and showed a small specific surface area of only 40 m² g⁻¹ and reaching 277 m² g⁻¹ after NH₃ activation (micro‐PoPD). These unique porous structures of the nitrogen‐doped carbon electrocatalysts led to the enhanced ORR activity with the half‐wave potential of 0.87 V (versus RHE) in alkaline media, which was higher than the corresponding values for the non‐PoPD, micro‐PoPD, and meso‐PoPD electrocatalysts (Figure 4.6e,f). Furthermore, the full cell (Zn–air battery) utilizing the hierarchical porous nitrogen‐doped carbon electrocatalyst exceeded the performance of the state‐of the‐art Pt/C catalyst (20 wt% Pt, BASF).

4.2.2 Noble‐Metal‐Free Porous Carbon Catalysts

In 1964, a transition metal (M) N₄‐macrocyclic compound, cobalt phthalocyanine, was reported for the first time as a cathodic ORR electrocatalyst [37]. Subsequently, it was found that thermal treatment of active carbon‐supported, cobalt‐based macrocyclic compounds (phthalocyanine, tetraphenylporphyrin, tetrabenzoporphyrin, and tetra(p‐methoxyphenyl)porphyrin) in an inert gas atmosphere at 800–900 °C significantly improved the stability of the acidic ORR electrocatalyst in 1978 [38]. In 2000, a series of nitrogen‐doped carbon materials with different Fe contents were synthesized. The time‐of‐flight secondary ion mass spectrometry studies revealed that one ion, FeN₂C₄⁺, was found in all prepared electrocatalysts regardless of the used Fe precursor or the synthesis procedure [39]. Importantly, the catalytic site in these carbon‐based electrocatalysts was first proposed to be similar to the molecular structure of FeN_x in an iron macrocyclic complex in 2002 [40]. Later, the Fe atom in the FeN_x active centers of the carbon matrix was shown to be responsible for the chemisorption of O₂ molecules [41]. Inspired by these pioneering results, the introduction of some earth‐abundant transition metals such as Fe and Co into the metal‐free, nitrogen‐doped carbon matrix has been considered an important strategy for fabricating high‐performance ORR electrocatalysts, especially in acidic media [42].

Accordingly, it was concluded that two vital parameters determine the ORR activity of non‐noble‐metal/nitrogen‐doped carbon electrocatalysts: (i) metal and nitrogen dopants, which constitute the ORR active centers and (ii) porous structure and density of active sites that determine the mass transfer of reactants and intermediates. On the basis of these considerations, a family of non‐noble‐metal/nitrogen‐doped carbon electrocatalysts has been developed in the past few years. Here, we summarize the development of noble‐metal‐free carbon electrocatalysts and the key relationship between the porous structure and the ORR activity of noble‐metal‐free carbon electrocatalysts.

4.2.2.1 Influence of Metal Centers on the ORR Activity

In the last 10 years, some breakthrough results have been achieved by rationally selecting nitrogen/transition‐metal precursors and carbon supports as well as optimizing the synthetic conditions. In 2009, Black Pearls 2000 (Cabot; micropore surface area 934 m² g⁻¹), iron acetate, and perylenetetracarboxylic dianhydride were used as a microporous carbon support, iron precursor, and carbon precursor, respectively [22]. The iron acetate and perylenetetracarboxylic dianhydride were filled into the micropores of the Black Pearls 2000 utilizing a planetary ball milling method. After the heat treatment in NH₃ atmosphere and leaching in an acidic solution, the Fe/N‐doped porous carbon with the micropore surface area of 580 m² g⁻¹ was obtained. In a proton‐exchange‐membrane‐based fuel cell system with the Fe/N‐doped porous carbon as the cathode, the current density at 0.9 V was equal to that of the cell with the Pt‐based cathode. In addition, the initial current densities produced by the Fe/N‐doped porous carbon (0.75 A cm⁻² with H₂/O₂ at 0.5 V; 0.58 A cm⁻² with H₂/air at 0.4 V) were retained throughout a 100‐h period, with final values of 0.33 A cm⁻² with H₂/O₂ at 0.5 V and 0.36 A cm⁻² with H₂/air at 0.4 V. Following this work, the Fe/N‐doped porous carbon was prepared by heating FeCo/PANI complexes on Ketjenblack [43]. The introduction of iron and cobalt dopants into nitrogen‐doped carbon resulted into an enhanced ORR activity and a four‐electron transfer process. The ORR onset potential of the FeCo/N‐doped carbon was as high as ∼0.93 V. The H₂O₂ yield of the FeCo/N‐doped carbon dropped to only 0.6% at 0.40 V. A 700‐h fuel cell performance test at the constant cell voltage of 0.4 V showed a promising performance stability of the FeCo/N‐doped porous carbon at the fuel cell cathode (Figure 4.7). The current density of the cell in a lifetime test remained nearly constant at ∼0.340 A cm⁻². The current density decreased by only 18 mA h⁻¹.

To achieve a high power density for the metal/nitrogen‐doped carbon, a Zn(II) zeolitic imidazolate framework (ZIF‐8) with the large specific surface area of 1800 m² g⁻¹ and high nitrogen content was employed as a host for impregnating iron(II) acetate (Fe precursor) and 1,10‐phenanthroline (N precursors) [44]. Following the high‐temperature pyrolysis of the above mixture and acidic leaching, Fe/N‐doped porous carbon was achieved. As shown in Figure 4.8 and Table 4.2, the specific surface area of the Fe/N‐doped porous carbon reached 964 m² g⁻¹, with the micropore and mesopore surface areas of 814 and 184 m² g⁻¹, respectively. Remarkably, at 0.6 V, the fuel cell cathode made with Fe/N‐doped porous carbon exhibited a power density of 0.75 W cm⁻² (Figure 4.8a). The peak power density for Fe/N‐doped porous carbon is 0.91 W cm⁻² (Figure 4.8b). This achievement of increased power density approaching that of the Pt‐based cathodes emphasized the importance of optimal mass transport properties and numerous active sites of the derived Fe/N‐doped porous carbon catalyst.

Electrocatalyst	N (atm%)	Fe (atm%)	Zn (atm%)	Activity at 0.8 V (A g−1)	BET surface area (m2 g−1)	Micropore surface area (m2 g−1)	Mesopore surface area (m2 g−1)
1/20/80‐Z8‐400 °C	NA	0.26	5.94	0.002	1237	NA	NA
1/20/80‐Z8‐700 °C	16.1	0.32	6.71	0.4	106	0	57
1/20/80‐Z8‐850 °C	9.4	0.38	2.48	10	273	209	94
1/20/80‐Z8‐1050 °C	3.7	0.65	0.06	254	478	504	46
1/20/80‐Z8‐1050 °C‐15 min	5.3	0.78	0.01	1120	964	814	184
1/50/50‐BP‐1050 °C‐60 min	2.4	0.44	0	429	767	605	162

More recently, a type of Zn/Co bimetallic metal‐oxide framework (MOF) (sodalite coordination of Co²⁺ and Zn²⁺ with 2‐methylimidazole) was employed as a precursor for constructing Co/N‐doped porous carbons [45]. The specific surface area was as large as 1426.8 m² g⁻¹. In an O₂‐saturated 0.1 M KOH solution, Co/N‐doped porous carbons showed an extremely high half‐wave potential of 0.881 V (versus RHE), higher than 0.811 V for the Pt/C catalyst. The kinetic current density of the Co/N‐doped porous carbon reached 21.2 mA cm⁻² at 0.8 V. In addition, at 0.8 V, the electron transfer number and H₂O₂ yield of the Co/N‐doped porous carbon were approximately 3.99 and 1%, respectively.

4.2.2.2 The Correlation Between Porous Nanostructures and ORR Activity

To obtain a good understanding of the relationship between the porous nanostructures and ORR activities of the noble‐metal‐free carbon electrocatalysts, colloidal silica (12 nm SiO₂ nanoparticles dispersed in water), ordered mesoporous silica SBA‐15, and montmorillonite (MMT, a layered clay with a 2D open channel) were used as hard templates to construct the mesoporous structures with tailored specific surface areas [46]. In this respect, vitamin B12 (VB12) and a PANI‐Fe complex were selected as precursors for preparing the Co/N‐doped and Fe/N‐doped mesoporous carbon electrocatalysts, respectively.

The Co/N‐doped carbon and Fe/N‐doped carbon electrocatalysts obtained as described above showed well‐defined mesoporous structures, a large BET surface area (up to 572 m² g⁻¹), and a high nitrogen content of 9.5%. The mesoporous nanostructures of the Co/N‐doped carbon electrocatalysts derived from VB12/colloidal silica provided a short mass transport path and a highly accessible surface area, resulting in an enhanced ORR activity with respect to VB12/SBA‐15, VB12/MMT, and VB12/C catalysts in acidic solution (Figure 4.9 and Table 4.3). In particular, VB12/silica‐colloid‐derived, mesoporous, Co/N‐doped carbon electrocatalysts exhibited the half‐wave potential of 0.79 V (versus RHE), only ∼58 mV lower than that of the Pt/C catalyst. Moreover, considering the almost identical nitrogen and cobalt content among the three electrocatalysts, a strong correlation between the ORR activity and the apparent BET surface area of the mesoporous carbon electrocatalysts was revealed and it undoubtedly reflects the significance of the porous nanostructure for the ORR.

Samples	Nitrogen (atm%)	Metal (atm%)	BET surface area (m2 g−1)	Micropore surface area (m2 g−1)	Mesopore surface area (m2 g−1)	ORR activity at 0.75 V (mA cm−2)
VB12/colloidal silica	9.5	1.3	568	81	462	3.9
VB12/SBA‐15	9.3	1.3	387	164	221	1.4
VB12/MMT	9.5	1.4	239	89	148	1.8
VB12/C	2.3	0.79	134	59	67	0.82
PANI‐Fe/colloidal silica	8.8	1.5	572	137	395	1.68
PANI‐Fe/C	6.5	0.98	153	66	71	0.23

4.3 Porous Carbon for HER Applications

In acidic solutions, HER is a two‐electron‐coupled proton reduction reaction (H⁺ + e⁻ → 1/2H₂). When the electrocatalytic HER activity of a catalyst is plotted as a function of the hydrogen–metal bond strength, a volcano‐shaped plot is found [47]. This behavior corresponds to the Sabatier principle that shows that optimal catalytic activity can be obtained on a catalyst surface with suitable adsorption free energies of intermediates. For the HER, the activation of hydrogen will be difficult if the binding affinity of hydrogen to the surface of the electrocatalysts is too weak. In contrast, when the binding affinity of hydrogen is too strong, the release of formed hydrogen molecules from the surface of the electrocatalysts will be blocked. Therefore, the electrocatalytic HER activity of the catalysts can be quantified by analyzing the free energy of hydrogen adsorption (ΔG_H).

4.3.1 Metal‐Free Carbon Electrocatalysts

The introduction of heteroatoms such as nitrogen into the carbon skeleton can change the electronic structure, resulting in heteroatom‐doped carbon with a decreased |ΔG_H|. In 2015, nanoporous nitrogen‐ and sulfur‐co‐doped graphene was prepared by chemical vapor deposition (Figure 4.10a) [48]. The coupling between the chemical dopants and the geometric lattice defects in the resulting 3D nanoporous graphene offered excellent catalytic activities toward HER in the 0.5 M H₂SO₄ solution. The applied overpotential at the current density of 10 mA cm⁻² was measured to be 280 mV for the N/S‐doped porous graphene (Figure 4.10b). To investigate the H* absorption free energy on the N/S‐doped graphene, DFT calculations were carried out. As revealed in Figure 4.10c, N/S‐doped graphene with one carbon defect near the S dopant (CSC or C=S) in the vicinity of graphitic‐N has the smallest |ΔG_H| value of 0.12 eV, close to the 0.09 eV for the Pt catalyst. Thereafter, various heteroatom‐doped graphenes have been systematically investigated using theoretical calculations and experimental analyses [49]. As shown in Figure 4.11a, theoretical calculation results revealed that N/S‐doped graphene had a low |ΔG_H| value of 0.23 eV that is considerably lower than 0.81 eV obtained for N‐doped graphene. N/P‐doped graphene showed a decreased |ΔG_H| of 0.53 eV, whereas N, B‐doped graphene had a large |ΔG_H| of 1.10 eV. Experimental studies were conducted by carefully synthesizing various dual‐doped graphene electrocatalysts and evaluating their acidic HER activities. As shown in Figure 4.11b, compared with 490 mV for the N‐doped graphene, N/S‐doped graphene and N/P‐doped graphene at 10 mA cm² showed dramatically decreased overpotentials of ∼310 and 420 mV, respectively.

4.3.2 Non‐precious Metal/Nitrogen‐Doped Porous Carbon Catalysts

In the past decades, chemists have made great breakthroughs in the exploration of bioinspired, synthetic metal complex HER catalysts such as cobaloxime [50], cobalt diimine–dioxime [51], and nickel phosphane compounds [52]. During the HER process, the chelated metal atoms in the above compounds behave as the active sites for H adsorption through the formation of metal–hydrogen bonds. Unfortunately, most bioinspired catalysts suffer from serious activity degradation and require very high overpotentials to deliver large current densities. To solve these challenges, chelating metal atoms into the nitrogen‐doped carbon frameworks is a promising strategy for constructing MN_x centers resembling those in metal complexes. The bond strength of the formed MH bond (|ΔG_H|) essentially determines the electrocatalytic HER activity of the metal/nitrogen‐doped carbon.

Accordingly, a porous carbon‐based HER electrocatalyst possessing abundant molecular CoN_x active centers has been developed [53]. The Co/N‐doped carbon electrocatalysts were fabricated through the direct pyrolysis method at high temperature where the cobalt/o‐phenylenediamine complexes served as precursors and colloidal silica was used as a hard template. The obtained Co/N‐doped carbon electrocatalysts exhibited a large specific surface area reaching 1074 m² g⁻¹ and well‐dispersed CoN_x active sites. Compared with nitrogen‐doped carbon (N/C) and Co nanoparticles anchored on the carbon (Co/C) electrocatalysts, the Co/N‐doped carbon electrocatalysts showed a dramatically decreased overpotential of only 133 mV at the current density of 10 mA cm⁻² (Figure 4.12a). The influence of acidic leaching of the Co/N‐doped carbon electrocatalysts on the HER activity was investigated to elucidate whether the metallic cobalt nanoparticles were responsible for the electrocatalytic HER. After acidic leaching, a sufficiently improved HER activity was observed, suggesting that the metallic cobalt or cobalt oxide nanoparticles were not responsible for the electrocatalytic hydrogen evolution (Figure 4.12b). Nevertheless, acidic leaching of the inactive inorganic cobalt species could expose a larger accessible surface area, leading to an enhanced HER activity. To further probe the active centers of the Co/N‐doped carbon electrocatalysts, the poisoning experiments of thiocyanate ions (SCN⁻) on their HER activity were conducted. The SCN⁻ is a poisoning reagent that blocks active metal sites due to its strong chemisorption property on the metal sites. After adding KSCN (10 mM) into the acidic electrolyte, the HER overpotential of the Co/N‐doped carbon electrocatalysts rose by >35 mV and the current density decreased substantially from 16.2 to 6.2 mA cm⁻² at 150 mV, which undoubtedly indicated that 60% of cobalt sites had been poisoned by the SCN⁻ ions (Figure 4.12c). In contrast, no obvious decrease in the HER activity was observed for the N‐doped carbon electrocatalyst in the presence of SCN⁻ ions, suggesting that there were no poisoning effects of SCN⁻ on the metal‐free, nitrogen‐doped carbon electrocatalyst. These results clearly noted that the CoN_x species in the Co/N‐doped carbon electrocatalysts were responsible for the electrocatalytic HER. Then, the turnover frequencies (TOF) per cobalt site reflecting the intrinsic activity of the Co/N‐doped carbon electrocatalysts was evaluated. Based on the hypothesis that all cobalt atoms in the Co/N‐doped carbon electrocatalysts are contained in active CoN_x species, the TOFs of Co/N‐doped carbon electrocatalysts were estimated to be 0.39 H₂ per second and 6.5 H₂ per second at overpotentials of 100 and 200 mV, respectively, surpassing the corresponding TOF values for the recently reported scalable molecular or solid‐state HER electrocatalysts (Figure 4.12d). In another example, a self‐supported, 3D porous Co/N‐doped carbon was synthesized on carbon cloth using the in situ carbonization of Co²⁺/PANI complex (Figure 4.13a) [54]. The MN_x centers in the porous Co/N‐doped carbon were identified as highly active sites for the HER. The overpotential required to drive the cathodic current density of 10 mA cm⁻² was as low as 138 mV (Figure 4.13b), which approached the value obtained using the above‐described Co/N‐doped carbon electrocatalysts containing active CoN_x species. DFT calculations revealed that the |ΔG_H| of Co/N‐doped porous carbon was only 0.15 eV, which was close to that for the Pt catalyst (Figure 4.13c).

4.4 Summary and Conclusions

The metal‐free or noble‐metal‐free porous carbon nanostructures have been intensively studied as ORR and HER electrocatalysts to replace the Pt noble metal. On the basis of the achieved results so far, it is concluded that the performance of ORR and HER of the porous carbon‐based electrocatalysts is strongly related to the heteroatom and metal dopants as well as to the porous nanostructures. The intrinsic properties and molecular structures of the heteroatoms and metal dopants determine the chemisorption properties of the intermediates. The pore size is a significant descriptor for the optimization of the ORR activity of the porous carbon electrocatalysts, balancing the requirements of a high number of active sites and a rapid mass transfer. Throughout, we have emphasized recent reports on porous carbon‐based ORR and HER electrocatalysts that are relevant to the understanding of ORR and HER active sites and to the correlation between the porous nanostructure and electrocatalytic activity.

Currently, in addition to the electrocatalytic ORR and HER, porous carbon catalysts are also drawing increasing attentions for application to other electrocatalytic reactions including oxygen evolution reaction (OER), CO₂ reduction, and photoelectrocatalytic water splitting. For instance, the CoN_x|P‐doped carbon electrocatalysts were synthesized by thermally treating dicyandiamide/phytic acid/Co²⁺ supermolecular complexes on flexible exfoliated graphene (FEG) and were found to be active for electrocatalytic OER [55]. In addition, a metal‐free carbon nanofiber (CNF) catalyst was synthesized by the pyrolysis of electrospun nanofiber mats of heteroatomic polyacrylonitrile (PAN) polymer. The resultant CNFs showed outstanding electrocatalytic CO₂ reduction activity comparable with that of noble metal electrocatalysts [56].

Despite the recent progress, to advance the development of the porous carbon electrocatalysts as an alternative to noble metal catalysts for energy‐conversion‐related reactions, significant challenges should be addressed: (i) the molecular structure of active centers responsible for the catalytic processes still needs to be elucidated; (ii) the fundamental reaction mechanisms on the surface of the catalysts must be revealed using advanced characterization technologies, especially operando spectroscopic measurements; (iii) the problem of long‐term stability must be solved with a particular focus on the development of efficient and stable noble‐metal‐free ORR electrocatalysts in acidic solutions; and (iv) evaluation of the catalytic performance in practical energy conversion systems such as electrolyzers and fuel cells. We hope that this chapter will shed light on the newcomers and experts in the energy‐conversion‐related catalysis fields and encourage them to make fascinating contributions to these rapidly developing areas.

Acknowledgments

The authors are grateful for the financial support from the ERC Grant on 2DMATER and EC under Graphene Flagship (no. CNECT‐ICT‐604391). The authors also acknowledge the Cfaed (Center for Advancing Electronics Dresden) and would also like to thank the colleagues and collaborators for their work cited in this chapter.

References

1. 1 Panwar, N., Kaushik, S., and Kothari, S. (2011). Renew. Sust. Energy Rev. 15: 1513–1524.
2. 2 Aricò, A.S., Bruce, P., Scrosati, B. et al. (2005). Nat. Mater. 4: 366–377.
3. 3 Momirlan, M. and Veziroglu, T.N. (2005). Int. J. Hydrogen Energy 30: 795–802.
4. 4 Edwards, P.P., Kuznetsov, V.L., David, W.I., and Brandon, N.P. (2008). Energy Policy 36: 4356–4362.
5. 5 Walter, M.G., Warren, E.L., McKone, J.R. et al. (2010). Chem. Rev. 110: 6446–6473.
6. 6 Borup, R., Meyers, J., Pivovar, B. et al. (2007). Chem. Rev. 107: 3904–3951.
7. 7 Debe, M.K. (2012). Nature 486: 43–51.
8. 8 Carmo, M., Fritz, D.L., Mergel, J., and Stolten, D. (2013). Int. J. Hydrogen Energy 38: 4901–4934.
9. 9 Nørskov, J.K., Rossmeisl, J., Logadottir, A. et al. (2004). J. Phys. Chem. B 108: 17886–17892.
10. 10 Jiao, Y., Zheng, Y., Jaroniec, M., and Qiao, S.Z. (2015). Chem. Soc. Rev. 44: 2060–2086.
11. 11 Li, Q., He, R., Gao, J.‐A. et al. (2003). J. Electrochem. Soc. 150: A1599–A1605.
12. 12 Gasteiger, H.A., Kocha, S.S., Sompalli, B., and Wagner, F.T. (2005). Appl. Catal., B 56: 9–35.
13. 13 Wang, Z., Wang, C., and Chen, K. (2001). J. Power Sources 94: 40–50.
14. 14 Kong, C.S., Kim, D.‐Y., Lee, H.‐K. et al. (2002). J. Power Sources 108: 185–191.
15. 15 Sheng, W., Gasteiger, H.A., and Shao‐Horn, Y. (2010). J. Electrochem. Soc. 157: B1529–B1536.
16. 16 Dai, L., Xue, Y., Qu, L. et al. (2015). Chem. Rev. 115: 4823–4892.
17. 17 Wang, J., Xu, F., Jin, H. et al. (2017). Adv. Mater. 29 (14): doi: 10.1002/adma.201605838.
18. 18 Maldonado, S. and Stevenson, K.J. (2005). J. Phys. Chem. B 109: 4707–4716.
19. 19 Bezerra, C.W.B., Zhang, L., Lee, K. et al. (2008). Electrochim. Acta 53: 4937–4951.
20. 20 Lee, J., Kim, J., and Hyeon, T. (2006). Adv. Mater. 18: 2073–2094.
21. 21 Liang, C., Li, Z., and Dai, S. (2008). Angew. Chem. Int. Ed. 47: 3696–3717.
22. 22 Lefèvre, M., Proietti, E., Jaouen, F., and Dodelet, J.‐P. (2009). Science 324: 71–74.
23. 23 Ikeda, T., Boero, M., Huang, S.‐F. et al. (2008). J. Phys. Chem. C 112: 14706–14709.
24. 24 Liu, R., Wu, D., Feng, X., and Müllen, K. (2010). Angew. Chem. Int. Ed. 49: 2565–2569.
25. 25 Qu, L., Liu, Y., Baek, J.‐B., and Dai, L. (2010). ACS Nano 4: 1321–1326.
26. 26 Gong, K., Du, F., Xia, Z. et al. (2009). Science 323: 760–764.
27. 27 Guo, D., Shibuya, R., Akiba, C. et al. (2016). Science 351: 361–365.
28. 28 Zhang, J., Zhao, Z., Xia, Z., and Dai, L. (2015). Nat. Nanotechnol. 10: 444–452.
29. 29 Yang, S., Feng, X., Wang, X., and Müllen, K. (2011). Angew. Chem. Int. Ed. 50: 5339–5343.
30. 30 Wei, W., Liang, H., Parvez, K. et al. (2014). Angew. Chem. Int. Ed. 53: 1570–1574.
31. 31 Wang, G., Sun, Y., Li, D. et al. (2015). Angew. Chem. Int. Ed. 54: 15191–15196.
32. 32 Liang, H.‐W., Zhuang, X., Brüller, S. et al. (2014). Nat. Commun. 5: 4973–4979.
33. 33 Aijaz, A., Fujiwara, N., and Xu, Q. (2014). J. Am. Chem. Soc. 136: 6790–6793.
34. 34 Zhong, H.‐X., Wang, J., Zhang, Y.‐W. et al. (2014). Angew. Chem. Int. Ed. 53: 14235–14239.
35. 35 Silva, R., Voiry, D., Chhowalla, M., and Asefa, T. (2013). J. Am. Chem. Soc. 135: 7823–7826.
36. 36 Liang, J., Jiao, Y., Jaroniec, M., and Qiao, S.Z. (2012). Angew. Chem. Int. Ed. 51: 11496–11500.
37. 37 Jasinski, R. (1964). Nature 201: 1212–1213.
38. 38 Bagotzky, V.S., Tarasevich, M.R., Radyushkina, K.A. et al. (1978). J. Power Sources 2: 233–240.
39. 39 Lefevre, M., Dodelet, J., and Bertrand, P. (2000). J. Phys. Chem. B 104: 11238–11247.
40. 40 Lefèvre, M., Dodelet, J.P., and Bertrand, P. (2002). J. Phys. Chem. B 106: 8705–8713.
41. 41 Schulenburg, H., Stankov, S., Schünemann, V. et al. (2003). J. Phys. Chem. B 107: 9034–9041.
42. 42 Bashyam, R. and Zelenay, P. (2006). Nature 443: 63–66.
43. 43 Wu, G., More, K.L., Johnston, C.M., and Zelenay, P. (2011). Science 332: 443–447.
44. 44 Proietti, E., Jaouen, F., Lefèvre, M. et al. (2011). Nat. Commun. 2: 416–424.
45. 45 Yin, P., Yao, T., Wu, Y. et al. (2016). Angew. Chem. Int. Ed. 55: 10800–10805.
46. 46 Liang, H.‐W., Wei, W., Wu, Z.‐S. et al. (2013). J. Am. Chem. Soc. 135: 16002–16005.
47. 47 Greeley, J., Jaramillo, T.F., Bonde, J. et al. (2006). Nat. Mater. 5: 909–913.
48. 48 Ito, Y., Cong, W., Fujita, T. et al. (2015). Angew. Chem. Int. Ed. 54: 2131–2136.
49. 49 Jiao, Y., Zheng, Y., Davey, K., and Qiao, S.‐Z. (2016). Nat. Energy 1: 16130–16138.
50. 50 Baffert, C., Artero, V., and Fontecave, M. (2007). Inorg. Chem. 46: 1817–1824.
51. 51 Jacques, P.‐A., Artero, V., Pécaut, J., and Fontecave, M. (2009). Proc. Natl. Acad. Sci. U.S.A. 106: 20627–20632.
52. 52 Helm, M.L., Stewart, M.P., Bullock, R.M. et al. (2011). Science 333: 863–866.
53. 53 Liang, H.‐W., Brüller, S., Dong, R. et al. (2015). Nat. Commun. 6: 7992–7997.
54. 54 Wang, Z.‐L., Hao, X.‐F., Jiang, Z. et al. (2015). J. Am. Chem. Soc. 137: 15070–15073.
55. 55 Hou, Y., Qiu, M., Zhang, T. et al. (2017). Adv. Mater. 29: 1604480.
56. 56 Kumar, B., Asadi, M., Pisasale, D. et al. (2013). Nat. Commun. 4: 2819–2826.