Fe-codoped porous carbon spheres derived from tetrazine-based polypyrrole as efficient electrocatalysts for the oxygen reduction reaction

Applied Catalysis A, General 559 (2018) 102–111

Contents lists available at ScienceDirect

Applied Catalysis A, General journal homepage: www.elsevier.com/locate/apcata

N-doped and N/Fe-codoped porous carbon spheres derived from tetrazinebased polypyrrole as efficient electrocatalysts for the oxygen reduction reaction ⁎

Tao Suna, Mei Yanga, Hongbiao Chena, , Yijiang Liua, Huaming Lia,b,

T

⁎

a

College of Chemistry, Xiangtan University, Xiangtan 411105, Hunan Province, PR China Key Laboratory of Polymeric Materials & Application Technology of Hunan Province, and Key Laboratory of Advanced Functional Polymeric Materials of College of Hunan Province, Xiangtan University, Xiangtan 411105, Hunan Province, PR China

b

A R T I C LE I N FO

A B S T R A C T

Keywords: Oxygen reduction reaction Polymer sphere Carbon sphere Dope Electrocatalyst

N-doped and N/Fe-codoped carbon materials are currently recognized as the most efficient electrocatalysts for the oxygen reduction reaction (ORR) because of their high catalytic activity, excellent stability, and superior tolerance to fuel. In this work, a novel kind of tetrazine-based polypyrrole spheres (PTPys) is prepared by protonic acid catalyzed Friedel-Crafts polymerization of bis(N-pyrrolyl)-1,2,4,5-tetrazine (TPy) with dimethoxymethane in dichloroethane. The resultant PTPys with a diameter of 100–300 nm are directly pyrolyzed at 900 °C to give N-doped porous carbon spheres (N/Cs-900 electrocatalyst). The N/Fe-codoped porous carbon spheres (Fe/N-Cs-900 electrocatalyst) are then constructed by mixing PTPys with Fe(OAc)2 followed by pyrolysis at 900 °C. The as-synthesized N/Cs-900 and Fe/N-Cs-900 electrocatalysts both have a well-defined spherical architecture together with a relatively high N content, high surface area and porosity. Owing to the synergistic influences, Fe/N-Cs-900 electrocatalyst exhibits excellent ORR activities in both alkaline and acidic electrolytes together with superior poisons tolerance in acidic media, whereas N/Cs-900 electrocatalyst displays ORR activity primarily in alkaline electrolyte. This approach can synthesize heteroatom-doped porous carbon spheres dispensing with the need for any templates, which is of great explored potentiality and utilized value.

1. Introduction Fuel cells (FCs) are a kind of very important energy conversion devices with promising development potential since they can transform chemical energy from fuels into electric energy directly [1,2]. Although FCs can guarantee clean and effective energy conversion with almost zero emissions, they seem to be not as powerful as batteries in terms of energy efficiency [3], due primarily to the slow kinetics of the cathodic oxygen reduction reaction (ORR) that is known to be caused by the state-of-the-art platinum (Pt) and its alloy electrocatalysts [4,5]. In addition, the Pt-based electrocatalysts are highly expensive and concurrently suffer from poor stability, low poison tolerance, and high overpotential, which seriously hamper the development and commercial application of FCs [1,4–9]. Recently, significant effort has been devoted toward designing and synthesizing high-performance ORR electrocatalysts. General speaking, the most ideal electrocatalyst for ORR in FCs should have such features as high surface area for maximum exposure of catalytic active sites [9], high electrical conductivity for facile electron transfer [10–12], hierarchical pore structure for efficient

⁎

mass transport [13], high stability for a long lifetime [14], and much more affordable price for high-volume production [15]. From an electrocatalyst point of view, N-doped carbon materials (N/C) as well as transition-metal (M) incorporated N/C (M/N-C, M = Fe, Co, Ni, Cu, etc.) [4,16–24] are at least partially with the above characteristics. Traditionally, both N/C and M/N-C electrocatalysts can be prepared by post-synthesis and direct pyrolysis approaches. The former method involves direct pyrolysis of a carbon matrix (carbon nanotube [6,9,25], graphene [11,26,27], porous carbon [10,13,28]) with a N-rich precursor for the purpose of synthesizing N/C electrocatalyst, while pyrolysis of a carbon matrix with either a mixture of M salt and N-rich compound or a single N-enriched, M-containing complex can afford M/N-C electrocatalyst [29]. Compared to commercial Pt/C electrocatalyst, both N/C and M/N-C electrocatalysts prepared by this method usually exhibit lower ORR activities due to that the carbon matrix cannot be uniformly mixed with those N-rich and/or M-containing precursors, which restricts the formation of a high-density of ORR active sites and leads to a decrease in ORR activity [30]. However, both of the two electrocatalysts have several advantages including low-

Corresponding authors at: College of Chemistry, Xiangtan University, Xiangtan 411105, Hunan Province, PR China. E-mail addresses: [email protected] (H. Chen), [email protected] (H. Li).

https://doi.org/10.1016/j.apcata.2018.04.028 Received 14 February 2018; Received in revised form 20 April 2018; Accepted 21 April 2018 Available online 22 April 2018 0926-860X/ © 2018 Elsevier B.V. All rights reserved.

Applied Catalysis A, General 559 (2018) 102–111

T. Sun et al.

2. Experiment section

cost, high stability, and excellent fuel tolerance [31,32]. In order to facilitate the formation of ORR active sites, N/C and M/N-C electrocatalysts can also be prepared by directly pyrolyzing a single precursor that concurrently contains C and N sources (i.e., polyaniline [22,33], polypyrrole [34], polyacrylonitrile [35], etc.) in the former case and simultaneously contains C, N, and M sources in the latter case (i.e., polyaniline-Fe complex [22], prussian blue [36], etc.). This method guarantees that the ORR active sites can be fully formed because the N or N/Fe atoms are homogeneously distributed and fixed in the precursors. However, their ORR activities still have room to improve because of the undeveloped pore structures that result in a low degree exposure of ORR active sites [37]. That is to say, the available ORR active sites are still lower compared to the in situ formed active sites. In order to expose the ORR active sites to a large extent, it is beneficial to construct N/C and M/N-C electrocatalysts with high surface area together with high porosity, in which ORR active sites can be distributed all over the electrocatalyst bulk including both outer and inner surfaces. At present, both hard-template strategy [9] and porous organic polymers (POPs)-based precursor [38–44] were adopted to realize this goal. Indeed, the hard-template strategy can guarantee finely tuned porosity for the final N/C and M/N-C electrocatalysts. However, the preparation and removal of hard template are generally laborious and tedious, which hinder the practical usage of such electrocatalysts. In order to surmount the obstacle, POPs have been widely adopted to construct these N/C and M/N-C electrocatalysts. The POPsbased precursors possess several merits including that they can be used to synthesize self-supported N/C and M/N-C electrocatalysts with high surface areas and high porosities, along with that they can promise the uniform N-doping and N/M-codoping. More importantly, recent progress in the field of ORR electrocatalysts has verified that ORR activity can be dramatically enhanced by converting the bulk electrocatalyst into nano/microsized ones. Up to now, various N/C and M/N-C electrocatalysts with architectures of tube [45,46], fiber [14,47], sheet [11,48], box [38], and sphere [19,49–52] etc. have been successfully fabricated. In particular, porous sphere-like structure has gained enormous attention because of the much higher surface area and porosity compared to its bulk counterpart. This porous sphere-like structure can lead to a full exposure of ORR active sites and meantime promise effective mass transport for the ORR [49]. Undoubtedly, the structure control of ORR electrocatalysts is also of great significance to promoting their catalytic activities. Herein, we present a facile method to synthesize spherical N/C and M/N-C electrocatalysts by direct pyrolysis of tetrazine-based polypyrrole spheres (PTPys) and the mixture of PTPys and Fe(OAc)2, respectively. We chose bis(N-pyrrolyl)-1,2,4,5-tetrazine (TPy) (Scheme S1, Fig. S1–3, see Supporting Information, SI) as the monomer for the preparation of spherical polypyrrole networks was due to that TPy had a relatively high N content (39.6 wt%). The TPy was polymerized through the Friedel-Crafts reaction in dichloroethane using dimethoxymethane (DMM) and methanesulfonic acid as the cross-linker and catalyst, respectively. The merit of this polymerization system lies in that the influence of metal residues on N/C electrocatalyst can be completely eliminated because no metal-containing precursor is used in the entire electrocatalyst synthesis process. In addition, Fe/N-C electrocatalyst can also be easily synthesized by mixing PTPys with Fe (OAc)2 followed by pyrolysis. The as-synthesized N-doped and N/Fecodoped porous carbon spheres (denoted as N/Cs-900 and Fe/N-Cs900, respectively) were fully characterized and their ORR performances in both alkaline and acidic electrolytes were systematically investigated. It was demonstrated that the metal-free N/Cs-900 electrocatalyst showed ORR activity primarily in alkaline electrolyte, whereas the Fe/N-Cs-900 electrocatalyst displayed promising ORR activities in both alkaline and acidic electrolytes together with superior poisons tolerance in acidic media.

2.1. Synthesis of N/Cs and Fe/N-Cs electrocatalysts For the preparation of N/Cs electrocatalysts, PTPys (400 mg) were directly pyrolyzed at different temperatures (800, 900 and 1000 °C) in a tube furnace in N2 atmosphere for 2 h with a constant heating rate of 5 °C min−1. The pyrolysis products are referred to as N/Cs-x, where x represents the pyrolysis temperature. The carbon yields were found to be 33.6%, 28.6%, and 12.0%, respectively, for the N/Cs-800, N/Cs-900, and N/Cs-1000 electrocatalysts. For the preparation of Fe/N-Cs electrocatalysts, PTPys (400 mg) and Fe(OAc)2 (14.8 mg) were firstly dispersed in ethanol (30 mL) and sonicated at room temperature for 3 h. After evaporating to dryness, the solid was dried in vacuum. Subsequently, the dried mixture was pyrolyzed at 800−1000 °C in a tube furnace in N2 atmosphere for 1 h. After cooling, the pyrolysis products were leached by H2SO4 aqueous solution (0.5 M) at 80 °C for 24 h to remove redundant Fe and its compounds. After further washing with water, the products were then subjected to a second pyrolysis at 800−1000 °C for 1 h under N2 atmosphere. The resultant electrocatalysts were denoted as Fe/N-Cs-x (x = 800, 900, or 1000). 2.2. Electrochemical measurements The electrochemical measurements were carried out on a CHI760D electrochemical workstation (Shanghai Chenhua Co., China) with a three-electrode cell at ambient temperature using 0.1 M KOH or 0.5 M H2SO4 solution as the electrolyte. The glassy carbon (5 mm in diameter) and Ag/AgCl (3 M KCl) electrodes were used as the working electrode and reference electrode, respectively. Platinum wire was used as the counter electrode in alkaline electrolyte, whereas graphite rod was used as the counter electrode in acidic electrolyte. All measured potentials in this study were converted to reversible hydrogen electrode (RHE) according to the following equation: E(RHE) = EAg/AgCl + 0.198 + 0.059 × pH Rotating ring-disk electrode (RRDE, Pine Research Instrumentation) measurements were carried out in O2-saturated 0.1 M KOH or 0.5 M H2SO4 solution to estimate the peroxide yield and the electron transfer number (n) during ORR. 3. Results and discussion 3.1. Preparation and characterization of N/Cs-900 and Fe/N-Cs-900 electrocatalysts In the present study, the tetrazine-based polypyrrole with well-defined spherical architecture (PTPys) was initially synthesized by Friedel-Crafts polymerization of TPy monomer with dimethoxymethane (DMM) cross-linker using methanesulfonic acid as the catalyst (Scheme S2, see SI). The dropwise addition of methanesulfonic acid to the TPy/ DMM solution is the key process in the synthesis of polymer spheres. When the first drops of methanesulfonic acid were added, a very small fraction of TPy was protonated and became insoluble in dichloroethane, which aggregated to form tiny particles as confirmed by dynamic light scattering (DLS) analysis (Fig. S4, see SI). The Friedel-Crafts polymerization therefore occurred within the particles and adsorbed newly protonated TPy molecules on their surface to grow. At the same time, the polymerized particles might congregate together to form small particle nuclei. With the gradual addition of protonic acid catalyst, the continuous growth of polymer particles finally resulted in the formation of polymer spheres. In this synthesis procedure, aggregation issue of polymer spheres can be able to overcome efficiently due to the electrostatic repulsion between these charged particles. Another merit of 103

Applied Catalysis A, General 559 (2018) 102–111

T. Sun et al.

Fig. 1. (a, b) SEM image (a) and TEM image (b) of PTPys. (c) FT-IR spectra of TPy and PTPys. (d) N2 adsorption-desorption isotherms of PTPys, inset shows the corresponding BJH pore size distribution.

linking at the two pyrrole rings in TPy is expected to leave behind the large tetrazine rings within the networks and therefore results in a very low surface area and pore volume. In contrast, DCB is smaller on middle (benzene ring) and larger on two ends (carbazole ring). The crosslinking of carbazole rings in DCB thus leaves behind the smaller benzene rings within the networks and leads to a high surface area (890 m2 g−1) as confirmed by Han and coworkers [56]. To further prove this, we have synthesize a DCB-like monomer, bis(N-carbazolyl)1,2,4,5-tetrazine (TCy). The hypercross-linked PTCy prepared by methanesulfonic acid catalyzed Friedel-Crafts reaction indeed gives a high surface area of 836 m2 g−1 (Scheme S3, Fig. S7, see SI). The non-metallic N/Cs-x (x = 800, 900, or 1000) electrocatalysts were then prepared by directly pyrolyzing PTPys at different temperatures. In a similar way, iron-containing Fe/N-Cs-x electrocatalysts were also constructed by blending PTPys with Fe(OAc)2 followed by pyrolysis at different temperatures. As for Fe/N-Cs electrocatalysts, the mass ratio of PTPys to Fe(OAc)2 was also optimized. Electrochemical test results demonstrated that the optimal pyrolysis temperature for both N/Cs and Fe/N-Cs electrocatalysts was 900 °C (Fig. S8a, b, see SI) and the optimal mass ratio of PTPys/Fe(OAc)2 was 27/1 (Fig. S9, see SI). In view of these, the next focus of study was the N/Cs-900 and Fe/ N-Cs-900 electrocatalysts. The typical SEM images of N/Cs-900 and Fe/N-Cs-900 electrocatalysts are shown in Fig. 2. As can be seen, both N/Cs-900 and Fe/NCs-900 electrocatalysts are composed of discrete carbon spheres with diameters in the range of 100–250 nm (Fig. 2a, b), whose sizes are smaller compared to their PTPys precursor due to that the shrinkage of polymer spheres during the pyrolysis is inevitable. It is noteworthy that the spherical architecture is basically maintained in the process of high temperature pyrolysis even though a few broken spheres are appeared. These pyrolyzed carbon spheres have a rough surface (Fig. 2a, b), which may result in a high specific surface area. The well-defined spherical structures were also confirmed by TEM images as depicted in Fig. 2c, d.

this polymerization system is that the influence of metal residues on N/ C electrocatalyst can be completely eliminated because no metal-containing precursor is used in the entire electrocatalyst synthesis process. The as-synthesized PTPys was thoroughly washed with KOH solution followed by water to afford basic PTPys. Fig. 1a shows the typical SEM image of the PTPys sample. As can be seen, near-perfect polymer spheres are observed although their size distribution is relatively broad, which ranges from 100 to 300 nm (Fig. S5, see SI). To be consistent with the SEM observation, TEM image indicates that the as-synthesized polymer has a solid spherical structure (Fig. 1b). The FTIR spectra of PTPys and its TPy monomer are illustrated in Fig. 1c. Apart from the characteristic peaks for TPy unit, there are a broad –OH stretching at 3432 cm−1, a C–H stretching at 2932 cm−1, and a weak C–O stretching at 1090 cm−1 in the FTIR spectrum of PTPys, indicating that crosslinker DMM is not fully reacted in the Friedel-Crafts polymerization. A small number of DMM fragments such as –CH2OH and –CH2OCH3 are attached to pyrrole rings and existed in the PTPys sample. The 13C CP/ MAS NMR spectrum of PTPys sample further proves the existence of incompletely reacted linker. As shown in Fig. S6 (see SI), the peaks at 159, 105–145, and 39 ppm can be ascribed to the tetrazinyl, pyrrolyl, and methylene linker carbons, respectively [53,54]. Whereas the peaks for –CH2OH and –CH2OCH3 moieties appear at 50–65 ppm [53,55], which can account for the appearance of O 1 s peak in the following XPS analysis. The physisorption isotherm of N2 at 77 K for PTPys is depicted in Fig. 1d, which displays a typical type I sorption with specific surface area and total pore volume of only 29.1 m2 g−1 and 0.053 cm3 g−1, respectively. Note that the surface area as well as pore volume is considerably lower than those of the reported hyper-crosslinked porous polymers, such as poly(1,4-di(9 H-carbazol-9-yl)benzene) (poly(DCB)) for instance [56]. We suspected that it was caused by the molecular topological structure of TPy monomer, which is big in the middle (tetrazine ring) but is small at both ends (pyrrole ring). The cross104

Applied Catalysis A, General 559 (2018) 102–111

T. Sun et al.

Fig. 2. (a, b) SEM images of N/Cs-900 (a) and Fe/N-Cs-900 (b). (c, d) TEM images of N/Cs-900 (c) and Fe/N-Cs-900 (d), inset of d shows single carbon sphere with porous structure. (e, f) HRTEM images of Fe/N-Cs-900, inset shows corresponding lattice spacing. (g–k) Elemental mapping images of Fe/N-Cs-900.

previous researches on N/C and Fe/N-C electrocatalysts have confirmed that their ORR activity depended largely on the types and contents of N/Fe containing species [29,57–59]. The full scanning XPS spectrum proves that the surface elements are comprised of primarily C (87.13 at %) and some N (7.05 at%) and O (5.82 at%) for N/Cs-900 electrocatalyst (Fig. 4a). The relatively high content of O in this electrocatalyst is expected to be originated from the –CH2OH and –CH2OCH3 terminal groups in the PTPys precursor as determined by FTIR (Fig. 1c) and elemental analysis (15.69 wt% O, Table S1, see SI). Deconvolution of N 1s peak (Fig. 4b) reveals that N atoms are mainly doped in the form of pyridinic-N (398.4 eV), pyrrolic-N (400.0 eV), graphitic-N (401.2 eV), and oxidized-N (403.8 eV) [60,61] with relative contents of 38.63%, 14.91%, 37.82%, and 8.64%, respectively. Note that the relatively high N content (7.05 at%) together with the high proportion of pyridinic-N (38.63%) and graphitic-N (37.82%) can endow the electrocatalyst with high ORR activity, which will be discussed later. As for Fe/N-Cs-900 electrocatalyst, the surface elemental composition is found to be C (87.94 at%), N (5.39 at%), O (5.56 at%), and a small amount of Fe (1.11 at%) (Fig. 4c, Table 1). Being different from N/Cs-900 electrocatalyst, the deconvoluted N 1s peak (Fig. 4d) confirms that the N species are mainly comprised of pyridinic-N (398.4 eV, 37.23%), Fe-N (399.6 eV, 11.15%), graphitic-N (401.1 eV, 47.61%) and oxidized-N (403.8 eV, 4.01%). A comparison of the two electrocatalysts reveals that pyridinic-N in the two electrocatalysts is almost at the same level. However, graphitic-N in Fe/N-Cs-900 electrocatalyst is significantly higher than in the N/Cs-900 electrocatalyst, and meanwhile there is no pyrrolic-N in the Fe/N-Cs-900 electrocatalyst. These results demonstrate that the incorporation of iron into PTPys precursor can facilitate the formation of graphitic-N during the high temperature pyrolysis because of Fe-induced catalytic graphitization as well as Fe-catalyzed N-doping [62–64]. High graphitic-N content can endow the Fe/N-Cs900 electrocatalyst with good electrical conductivity, which can be

With careful observation of the outer edge of these carbon spheres, there exist plenty of mesopores (Fig. 2c), confirming that porous carbon spheres (inset of Fig. 2d) have been successfully constructed. The HRTEM image in Fig. 2e demonstrates that Fe/N-Cs-900 electrocatalyst has a relatively high degree of graphitization with a lattice spacing of 0.31 nm (inset of Fig. 2e). In addition, metal Fe nanoparticles are also observed in the spherical Fe/N-Cs-900 electrocatalyst (Fig. 2f, g) with a lattice spacing of 0.26 nm (inset of Fig. 2f). Moreover, elemental mapping in Fig. 2h−k clearly demonstrates that N, O, and Fe elements are uniformly distributed throughout the carbon sphere, suggesting successful doping of N, O, and Fe atoms into the carbon matrix. Compared with N and O elements, Fe doping level in Fe/N-Cs-900 electrocatalyst is much lower. In order to further prove the porous texture, both N/Cs-900 and Fe/ N-Cs-900 electrocatalysts were thus investigated by N2 sorption isotherms at 77 K. As illustrated in Fig. 3a, N/Cs-900 electrocatalyst generates type-I isotherms, confirming the microporous structure of this electrocatalyst. Evidently, the BET surface area (SBET) of N/Cs-900 electrocatalyst was found to be 584 m2 g−1, in which the micropore surface area (SMic = 509 m2 g−1) accounts for a high proportion (87.2%). The pore size distribution (PSD) estimated by BJH method also confirms that its pore size is below 7 nm (Fig. 3b). The Fe/N-Cs-900 electrocatalyst displays type-IV isotherms with an indistinct type-H4 hysteresis loop (Fig. 3c), reflecting the predominantly microporous structure. The PSD of Fe/N-Cs-900 electrocatalyst proves that its pore size is below 8 nm (Fig. 3d). The SBET of Fe/N-Cs-900 electrocatalyst is found to be 610 m2 g−1, of which above 84.8% is contributed by micropore (SMic = 517 m2 g−1). Note that both isotherms are raised obviously near p/p0 = 1, implying that the two electrocatalysts contain a small number of macropores (Fig. S10, see SI). The bonding configurations and contents of doped heteroatom were finally studied by X-ray photoelectron spectroscopy (XPS), because 105

Applied Catalysis A, General 559 (2018) 102–111

T. Sun et al.

Fig. 3. N2 adsorption-desorption isotherms (a, c) and BJH pore size distribution (b, d) of N/Cs-900 (a, b) and Fe/N-Cs-900 (c, d), inset of c shows the corresponding enlarged isotherms.

agreement with that measured by the inductively coupled plasma atomic emission spectrometry (ICP-AES) method (1.31 wt%).

confirmed by electrochemical impedance spectroscopy (EIS) analysis (Fig. S11, see SI). In addition, the appearance of Fe-N peak in the deconvoluted N 1s spectrum also confirms that Fe ion can coordinate with pyridinic-N to form Fe-N4 active site. XRD patterns (Fig. S12, see SI) of these Fe/N-Cs-x electrocatalysts prepared at different pyrolysis temperatures also demonstrated the existence of Fe-N species. It is worthy of note that the Fe content detected by XPS (1.11 at%) is in good

3.2. ORR performance of N/Cs-900 and Fe/N-Cs-900 electrocatalysts The ORR activities of N/Cs-900 and Fe/N-Cs-900 electrocatalysts were first assessed by linear sweep voltammetry (LSV) on RDE in

Fig. 4. XPS survey spectra (a, c) and high-resolution N 1 s spectra (b, d) of N/Cs-900 (a, b) and Fe/N-Cs-900 (c, d). 106

Applied Catalysis A, General 559 (2018) 102–111

T. Sun et al.

Table 1 C, N, O, Fe contents of N/Cs-900 and Fe/N-Cs-900 evaluated from XPS. Samples

N/Cs-900 Fe/N-Cs-900

Elemental content (at%)

N configuration (%)

C

N

O

Fe

Pyridinic-N

Fe-N

Pyrrolic-N

Graphitic-N

Oxidized-N

87.13 87.94

7.05 5.39

5.82 5.56

– 1.11

38.63 37.23

– 11.15

14.91 –

37.82 47.61

8.64 4.01

Fig. 5. (a) ORR polarization plots of N/Cs-900, Fe/N-Cs-900 and Pt/C in O2-saturated 0.1 M KOH. (b, c) HO2− yields and n values of N/Cs-900 and Pt/C (b), and Fe/ N-Cs-900 and Pt/C (c) in O2-saturated 0.1 M KOH. (d) Endurance tests of N/Cs-900, Fe/N-Cs-900 and Pt/C for 10 000 cycles in O2-saturated 0.1 M KOH.

reference electrocatalyst, the Eo, E1/2, and Jl values of N/Cs-900 electrocatalyst were found to be 0.948 V, 0.823 V, and 5.42 mA cm−2, respectively, suggesting that the ORR activities of the two electrocatalysts are almost identical in alkaline media. To gain insight into the ORR mechanism, LSV curves of these electrocatalysts were then recorded at different rotation speeds in O2saturated 0.1 M KOH solution (Fig. S14a, c, e, see SI). The corresponding Koutecky-Levich (K-L) plots showed good linear relationship between J–1 and ω–1/2 (Fig. S14b, d, f, see SI), suggesting the first-order ORR kinetics relating to O2 concentration. According to the slope of K-L plots, the electron transfer number (n) during ORR can be calculated. The n values for Fe/N-Cs-900 electrocatalyst range between 3.99 and 4.04 in the whole potential range (Fig. S15, see SI) and are similar to those of Pt/C reference electrocatalyst (3.98–4.00) (Fig. S15, see SI), implying a high 4e selectivity during ORR. With respect to the N/Cs900 electrocatalyst, the n values vary from 3.96 to 3.99 at all potentials (Fig. S15, see SI). Obviously, such n values are very close to those of Pt/ C reference electrocatalyst, again implying that 4e pathway is predominant for the ORR. The 4e selectivity of ORR on these electrocatalysts can be further proved by RRDE tests. The n values and peroxide yields for these electrocatalysts were calculated from the corresponding disk (id) and ring currents (ir) (Fig. S16a, see SI). As depicted in Fig. 5c, the HO2− yield estimated for Pt/C electrocatalyst stays below 6.3% at all potentials, while the n value keeps above 3.89 in the same potential range. Compared to Pt/C reference electrocatalyst, N/Cs-900 electrocatalyst shows a slightly higher HO2− yield (below 6.6% at all potentials) together with a slightly lower n value (above 3.87 at all potentials), confirming that the 4e selectivity of ORR on N/Cs-900 electrocatalyst is

Table 2 Electrochemical properties of N/Cs-900, Fe/N-Cs-900, and Pt/C electrocatalysts. Electrocatalysts

Loading (mg cm−2)

Eo (V vs. RHE)

E1/2 (V vs. RHE)

Jl (mA cm−2)

N/Cs-900 Fe/N-Cs-900 Pt/C

0.82 0.61 0.10

0.948 0.986 0.964

0.823 0.871 0.829

5.42 5.89 5.30

alkaline media. For comparative purposes, LSV measurement for Pt/C reference electrocatalyst (20 wt%, JM) was also performed under identical conditions. Considering that the ORR catalytic activity is affected by electrocatalyst loading density, the optimum loadings for N/ Cs-900 and Fe/N-Cs-900 electrocatalysts were therefore optimized and found to be 0.82 mg cm−2 and 0.61 mg cm−2, respectively (Fig. S13a, b, see SI). As for the Pt/C reference electrocatalyst, its loading was fixed at 0.10 mg cm−2 because the most common loading for Pt/C electrocatalyst is in the range of 0.05–0.20 mg cm−2 as mentioned in previous studies [26,29,34,49]. In the present study, the ORR activities of these electrocatalysts were quantitatively evaluated and compared by their ORR onset potential (Eo), half-wave potential (E1/2), and limiting current density (Jl) values. As depicted in Fig. 5a and Table 2, Fe/N-Cs-900 electrocatalyst exhibited the highest ORR catalytic activity among the three electrocatalysts with Eo, E1/2, and Jl values of 0.986 V (vs. RHE, the same hereinafter), 0.871 V, and 5.89 mA cm−2, respectively. Note that the Eo and E1/2 values of Fe/N-Cs-900 electrocatalyst are about 22 mV and 42 mV higher than those of Pt/C reference electrocatalyst (Eo = 0.964 V, E1/2 = 0.829 V, Jl = 5.30 mA cm−2). Compared to Pt/C 107

Applied Catalysis A, General 559 (2018) 102–111

T. Sun et al.

Fig. 6. Effects of SCN− ion on the ORR activity for N/Cs-900 and Fe/N-Cs-900 in O2-saturated 0.1 M KOH (a) and 0.5 M H2SO4 (b) at 1600 rpm.

pyridinic-N and the other with 80% of graphitic-N. Besides, Ishikawa et al. [71] have synthesized a series of N-doped carbon nanotubes with different N contents. Their electrochemical test results confirmed that the ORR activity in terms of E1/2 values was linearly increased with pyridinic-N contents. These results proved that pyridinic-N had a higher ORR activity than graphitic-N. In contrast, Müllen et al. [72] had studied the N-doped ordered mesoporous graphitic arrays and confirmed that graphitic-N played a crucial role for the ORR, that is to say, graphitic-N is of greater importance to ORR than pyridinic-N. Lai et al. [57] had also studied N-doped graphene and found that pyridinic-N was related to Eo value and high pyridinic-N level leaded to a more positive Eo value, whereas graphitic-N appeared to be related to Jl value and high graphitic-N level facilitated a 4e pathway process for the ORR. In our case, N/Cs-900 electrocatalyst with 38.63% of pyridinic-N and 37.82% of graphitic-N showed a combined 4e and 2e pathway for the ORR in alkaline media, while Fe/N-Cs-900 electrocatalyst with 37.23% of pyridinic-N, 11.15% of Fe-N, and 47.61% of graphitic-N showed a complete 4e pathway as mentioned previously in the electrochemical experiments. These results indicate that the impact of graphitic-N on 4e ORR pathway cannot be ignored apart from the contribution of Fe-N active sites, because pyridinic-N in both N/Cs-900 and Fe/N-Cs-900 electrocatalysts is almost at the same level. This can be confirmed by SCN−-poisoning experiments as shown in Fig. 6. With respect to N/Cs900 electrocatalyst, SCN−-poisoning has almost no influence on its ORR activity, indicating that this electrocatalyst contains only N-C active sites. In contrast, SCN−-poisoned Fe/N-Cs-900 electrocatalyst shows an obviously negative shift in E1/2 in compassion with unpoisoned ones, i.e., ∼16 mV in alkaline media (Fig. 6a) and ∼83 mV in acidic media (Fig. 6b), implying the existence of Fe-containing active sites. Note that the SCN−-poisoned Fe/N-Cs-900 electrocatalyst still possessed higher ORR activity than the N/Cs-900 electrocatalyst, demonstrating that additional ORR catalytic activity was mainly due to the high graphitic-N content in the Fe/N-Cs-900 electrocatalyst. Based on the above analysis, we believe that both pyridinic-N and graphitic-N are ORR active and graphitic-N content is closely related to the 4e ORR pathway. In addition, O-containing functionalities such as carboxyl, carbonyl, and hydroxyl groups also have positive influence on the ORR activity as confirmed by Ishizaki et al. [73] and Niu et al. [74]. In the present study, the relatively high O-containing species is at least part of the reason for their high ORR performance (N/Cs-900, 5.82 at% O; Fe/ N-Cs-900, 5.56 at% O; Table 1, Fig. S18, see SI). Secondly, both of the two electrocatalysts possessed a relatively high SBET value, i.e., 584 m2 g−1 for N/Cs-900 electrocatalyst and 610 m2 g−1 for Fe/N-Cs900 electrocatalyst, in which the proportion of micropore surface areas were both higher than 85%. The high proportion of micropores can increase the exposure of ORR active sites to a large extent, and meanwhile the small proportion of macropores is benefit for effective mass transport, therefore resulting in enhanced ORR activity. Finally, the spherical structure of both N/Cs-900 and Fe/N-Cs-900 electrocatalysts

not as good as Pt/C reference electrocatalyst. These results suggested that the ORR on N/Cs-900 electrocatalyst might proceed through a combined 4e and 2e pathway but 4e pathway was overwhelmingly predominant. Conversely, Fe/N-Cs-900 electrocatalyst exhibits a much lower HO2− yield (below 2.6% at all potentials) as well as a higher n value (above 3.97 at all potentials), confirming an almost complete 4e pathway for the ORR. The excellent ORR catalytic activity of N/Cs-900 and Fe/N-Cs-900 electrocatalysts can be further confirmed by Tafel plots. The Tafel slopes were determined to be 66, 65, and 61 mV dec–1, respectively, for Pt/C, N/Cs-900, and Fe/N-Cs-900 electrocatalysts in 0.1 M KOH solution (Fig. S17a, see SI). The smaller Tafel slope of Fe/N-Cs-900 electrocatalyst in comparison with Pt/C electrocatalyst again confirms good kinetic process for the ORR. Besides the outstanding ORR performance, N/Cs-900 and Fe/N-Cs-900 electrocatalysts also exhibited excellent cycling stability. The durability measurements for the three electrocatalysts in O2-saturated 0.1 M KOH solution are depicted in Fig. 5d. After 10 000 potential cycles between 0.6 V and 1.0 V, the negative shifts of E1/2 were found to be 48, 19, and 11 mV, respectively, for Pt/C, N/Cs-900, and Fe/N-Cs-900 electrocatalysts. Clearly, the E1/2 negative shifts of both N/Cs-900 and Fe/N-Cs-900 electrocatalysts are considerably lower than that of Pt/C reference electrocatalyst, suggesting that the long-term stabilities of N/Cs-900 and Fe/N-Cs-900 electrocatalysts are much better than Pt/C electrocatalyst in alkaline media. The high ORR catalytic activities of N/Cs-900 and Fe/N-Cs-900 electrocatalysts might be explained by the synergistic influences of their heteroatom-doping, porosity, and spherical structure. Firstly, the relatively high N content in both N/Cs-900 (7.05 at% N) and Fe/N-Cs-900 (5.39 at% N) electrocatalysts may at least partly account for their high ORR activity, because the increase of N-doping level in N-doped/codoped carbon materials is a crucial factor to the elevation of ORR activity, as has been proved by Kim et al. [65], Chen et al. [66], WigginsCamacho and Stevenson [67,68] just a few years ago. With regard to N/ Cs-900 electrocatalyst, the N-containing (N-C) species such as pyridinicN (38.63%) and graphitic-N (37.82%) occupy a large proportion of the total N in this electrocatalyst as mentioned in the XPS analysis (Fig. 4a), which should be considered to be ORR active as proved by previous studies [30,57]. That is to say, above 76% of the total N in N/Cs-900 electrocatalyst is ORR active (Table 1). Of particular interest is that there exist only three main types of N-doping structures (pyridinic-N 37.23%, Fe-N 11.15%, graphitic-N 47.61%) in Fe/N-Cs-900 electrocatalyst (Fig. 4b, Table 1), in which almost all the N species are ORR active due to that Fe-N species is also regarded as the ORR active sites as proved by plenty of studies [9,15,69]. At present, there is still debate on pyridinic-N and graphitic-N in terms of their contribution in promoting ORR activity. For example, Yasuda et al. [70] have found that graphitic-N can catalyze ORR predominately through a 2e pathway, while pyridinic-N catalyze ORR via a combined 4e and 2e pathway after investigating two kinds of N-doped graphenes, one with 90% of 108

Applied Catalysis A, General 559 (2018) 102–111

T. Sun et al.

Fig. 7. (a) ORR polarization plots of N/Cs-900, Fe/N-Cs-900 and Pt/C in O2-saturated 0.5 M H2SO4. (b) H2O2 yields and n values of Fe/N-Cs-900 and Pt/C in O2saturated 0.5 M H2SO4. (c, d) Endurance tests of Fe/N-Cs-900 (c) and Pt/C (d) for 10 000 cycles in O2-saturated 0.5 M H2SO4.

Fig. 8. Bar plots showing ARE1/2 of the RDE measurements, comparing Pt/C with Fe/N-Cs-900 in O2-saturated 0.5 M H2SO4 at 1600 rpm, 30 s hold, before and after adding corresponding concentration of Cl− ion (a), urea (b), methanol (c), and CO (d).

shows ORR activity primarily in alkaline electrolyte [21,75–77]. As expected, N/Cs-900 electrocatalyst really showed a significant lower ORR activity (Eo = 0.772 V, E1/2 = 0.565 V, Jl = 4.70 mA cm−2) in comparison with Pt/C reference electrocatalyst (Eo = 0.891 V, E1/ −2 ) in 0.5 M H2SO4 solution (Fig. 7a). 2 = 0.738 V, Jl = 4.74 mA cm Conversely, Fe/N-Cs-900 electrocatalyst exhibited an ORR activity (Eo = 0.845 V, E1/2 = 0.717 V, Jl = 5.89 mA cm−2) nearly comparable to Pt/C reference electrocatalyst in acidic media. The Tafel slopes were

further facilitates exposure of ORR catalytic sites as well as mass transport, which are also favorable for improving ORR activity. The outstanding ORR activity of N/Cs-900 and Fe/N-Cs-900 electrocatalysts in alkaline media prompted us to investigate their ORR activities in acidic media. In the present study, however, we focused on Fe/N-Cs-900 electrocatalyst because previous researches have proved that Fe/N-C electrocatalyst can display promising ORR activity in both alkaline and acidic electrolytes, whereas metal-free N/C electrocatalyst 109

Applied Catalysis A, General 559 (2018) 102–111

T. Sun et al.

found to be 85 and 70 mV dec–1, respectively, for Pt/C and Fe/N-Cs-900 electrocatalysts in acidic media (Fig. S17b, see SI), also demonstrating a fast ORR kinetic process on Fe/N-Cs-900 electrocatalyst. In addition, RRDE experiments confirmed that the H2O2 yield (below 1.4% at all potentials) and n value (above 3.96 at all potentials) measured for Fe/ N-Cs-900 electrocatalyst in 0.5 M H2SO4 solution are very close to Pt/C reference electrocatalyst (Fig. 7b), confirming a high 4e selectivity for the ORR. Moreover, Fe/N-Cs-900 electrocatalyst also exhibited better long-term stability in acidic media in comparison with Pt/C reference electrocatalyst (Fig. 7c). Considering that Pt/C electrocatalyst is easy to poisoning and meanwhile poison influences can severely damage its ORR activity, the poison tolerance of Fe/N-Cs-900 and Pt/C electrocatalysts were finally assessed by RDE measurements. According to Kucernak et al. [51], the loss in ORR activity can be quantitatively estimated from the J1/2 values at E1/2 in the presence and absence of contaminants. The activity retained at E1/2 (ARE1/2) can be calculated by the following equation:

Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.apcata.2018.04.028. References [1] B.C. Steele, A. Heinzel, Nature 414 (2001) 345–352. [2] M.K. Debe, Nature 486 (2012) 43–51. [3] A. Forner-Cuenca, J. Biesdorf, L. Gubler, P.M. Kristiansen, T.J. Schmidt, P. Boillat, Adv. Mater. 27 (2015) 6317–6322. [4] L. Lin, Z.K. Yang, Y.F. Jiang, A.W. Xu, ACS Catal. 6 (2016) 4449–4454. [5] H.A. Gasteiger, S.S. Kocha, B. Sompalli, F.T. Wagner, Appl. Catal. B: Environ. 56 (2005) 9–35. [6] K. Gong, F. Du, Z. Xia, M. Durstock, L. Dai, Science 323 (2009) 760–764. [7] D.S. He, D. He, J. Wang, Y. Lin, P. Yin, X. Hong, Y. Wu, Y. Li, J. Am. Chem. Soc. 138 (2016) 1494–1497. [8] X. Wang, M. Vara, M. Luo, H. Huang, A. Ruditskiy, J. Park, S. Bao, J. Liu, J. Howe, M. Chi, Z. Xie, Y. Xia, J. Am. Chem. Soc. 137 (2015) 15036–15042. [9] S.H. Ahn, X. Yu, A. Manthiram, Adv. Mater. 29 (2017) 1606534. [10] M. Qiao, C. Tang, G. He, K. Qiu, R. Binions, I.P. Parkin, Q. Zhang, Z. Guo, M.M. Titirici, J. Mater. Chem. A 4 (2016) 12658–12666. [11] S. Yang, X. Feng, X. Wang, K. Müllen, Angew. Chem., Int. Ed. 50 (2011) 5339–5343. [12] H.J. Choi, N.A. Kumar, J.B. Baek, Nanoscale 7 (2015) 6991–6998. [13] J. Zhu, D. Xu, C. Wang, W. Qian, J. Guo, F. Yan, Carbon 115 (2017) 1–10. [14] T.Y. Ma, J. Ran, S. Dai, M. Jaroniec, S.Z. Qiao, Angew. Chem., Int. Ed. 54 (2015) 4646–4650. [15] P. Chen, L.K. Wang, G. Wang, M.R. Gao, J. Ge, W.J. Yuan, Y.H. Shen, A.J. Xie, S.H. Yu, Energy Environ. Sci. 7 (2014) 4095–4103. [16] G. Lu, Y. Zhu, L. Lu, K. Xu, H. Wang, Y. Jin, Z.J. Ren, Z. Liu, W. Zhang, J. Power Sources 315 (2016) 302–307. [17] S. Ratso, I. Kruusenberg, M. Käärik, M. Kook, R. Saar, P. Kanninen, T. Kallio, J. Leis, K. Tammeveski, Appl. Catal. B: Environ. 219 (2017) 276–286. [18] K.A. Kuttiyiel, Y. Choi, K. Sasaki, D. Su, S.M. Hwang, S.D. Yim, T.H. Yang, G.G. Park, R.R. Adzica, Nano Energy 29 (2016) 261–267. [19] Y. Hou, Z. Wen, S. Cui, S. Ci, S. Mao, J. Chen, Adv. Funct. Mater. 25 (2015) 872–882. [20] H. Wang, W. Wang, M. Gui, M. Asif, Z. Wang, Y. Yu, J. Xiao, H. Liu, ACS Appl. Mater. Interfaces 9 (2016) 335–344. [21] X. Wang, H. Zhang, H. Lin, S. Gupta, C. Wang, Z. Tao, H. Fu, T. Wang, J. Zheng, G. Wu, X. Lia, Nano Energy 25 (2016) 110–119. [22] G. Wu, K.L. More, C.M. Johnston, P. Zelenay, Science 332 (2011) 443–447. [23] H. Yu, A. Fisher, D. Cheng, D. Cao, ACS Appl. Mater. Interfaces 8 (2016) 21431–21439. [24] N.P. Subramanian, S.P. Kumaraguru, H. Colon-Mercado, H. Kim, B.N. Popov, T. Black, D.A. Chen, J. Power Sources 157 (2006) 56–63. [25] M. Yang, D. Yang, H. Chen, Y. Gao, H. Li, J. Power Sources 279 (2015) 28–35. [26] M. Favaro, L. Ferrighi, G. Fazio, L. Colazzo, C.D. Valentin, C. Durante, F. Sedona, A. Gennaro, S. Agnoli, G. Granozzi, ACS Catal. 5 (2014) 129–144. [27] C. Zhang, J. Sha, H. Fei, M. Liu, S. Yazdi, J. Zhang, Q. Zhong, X. Zou, N. Zhao, H. Yu, Z. Jiang, E. Ringe, B.I. Yakobson, J. Dong, D. Chen, J.M. Tour, ACS Nano 11 (2017) 6930–6941. [28] J. Zhang, L. Dai, ACS Catal. 5 (2015) 7244–7253. [29] C. Zhu, S. Fu, J. Song, Q. Shi, D. Su, M.H. Engelhard, X. Li, D. Xiao, D. Li, L. Estevez, D. Du, Y. Lin, Small 13 (2017) 1603407. [30] L. Dai, Y. Xue, L. Qu, H.J. Choi, J.B. Baek, Chem. Rev. 115 (2015) 4823–4892. [31] S. Ratso, I. Kruusenberg, M. Käärik, M. Kook, R. Saar, M. Pärs, J. Leis, K. Tammeveski, Carbon 113 (2017) 159–169. [32] Y. Ye, F. Cai, H. Li, H. Wu, G. Wang, Y. Li, S. Miao, S. Xie, R. Si, J. Wang, X. Bao, Nano Energy 38 (2017) 281–289. [33] G. Wang, K. Jiang, M. Xu, C. Min, B. Ma, X. Yang, J. Power Sources 266 (2014) 222–225. [34] J. Yang, X. Wang, B. Li, L. Ma, L. Shi, Y. Xiong, H. Xu, Adv. Funct. Mater. 27 (2017) 1606497. [35] J. Zhang, Y. Song, M. Kopeć, J. Lee, Z. Wang, S. Liu, J. Yan, R. Yuan, T. Kowalewski, M.R. Bockstaller, K. Matyjaszewski, J. Am. Chem. Soc. 139 (2017) 12931–12934. [36] X. Wang, L. Zou, H. Fu, Y. Xiong, Z. Tao, J. Zheng, X. Li, ACS Appl. Mater. Interfaces 8 (2016) 8436–8444. [37] W. Ding, L. Li, K. Xiong, Y. Wang, W. Li, Y. Nie, S. Chen, X. Qi, Z. Wei, J. Am. Chem. Soc. 137 (2015) 5414–5420. [38] Y. Hou, T. Huang, Z. Wen, S. Mao, S. Cui, J. Chen, Adv. Energy Mater. 4 (2014) 1400337. [39] Z. Xiang, Y. Xue, D. Cao, L. Huang, J.F. Chen, L. Dai, Angew. Chem., Int. Ed. 53 (2014) 2433–2437. [40] S. Lu, Y. Jin, H. Gu, W. Zhang, Sci. China Chem. 60 (2017) 999–1006. [41] L. Hao, S. Zhang, R. Liu, J. Ning, G. Zhang, L. Zhi, Adv. Mater. 27 (2015) 3190–3195. [42] G. Lu, Y. Zhu, K. Xu, Y. Jin, Z.J. Ren, Z. Liu, W. Zhang, Nanoscale 7 (2015) 18271–18277. [43] G. Lu, H. Yang, Y. Zhu, T. Huggins, Z.J. Ren, Z. Liu, W. Zhang, J. Mater. Chem. A 3 (2015) 4954–4959. [44] P. Pachfule, V.M. Dhavale, S. Kandambeth, S. Kurungot, R. Banerjee, Chem. Eur. J. 19 (2013) 974–980. [45] H.T. Chung, J.H. Won, P. Zelenay, Nat. Commun. 4 (2013) 1922.

ARE1/2 = (J1/2c/J1/2) × 100 where J1/2 is the ORR current density of a given electrocatalyst at E1/2 in the absence of contaminants, J1/2c is the current densities at the same E1/2 (determined for such electrocatalyst in the absence of contaminants) in the presence of different concentrations of contaminants. As depicted in Fig. 8, the ARE1/2 values of Fe/N-Cs-900 electrocatalyst were only slightly decreased after the introduction of different concentrations of urea, Cl−, methanol, and CO (Fig. S19–20, see SI). However, the ARE1/2 values of Pt/C reference electrocatalyst were found to be zero or drastically decreased in the presence of these contaminants (Fig. S19 − 20, see SI). These results confirm that Fe/N-Cs900 electrocatalyst has superior poison tolerance in comparison with Pt/C electrocatalyst. The good methanol and CO tolerance for Fe/N-Cs900 electrocatalyst in acidic electrolyte can be further confirmed by its chronoamperometric responses (Fig. S21, see SI).

4. Conclusions In summary, a novel kind of tetrazine-based polypyrrole spheres (PTPys) was prepared by Friedel-Crafts polymerization of bis(N-pyrrolyl)tetrazine (TPy) with dimethoxymethane (DMM) in dichloroethane using methanesulfonic acid as the catalyst. It was found that the dropwise addition of methanesulfonic acid to the TPy/DMM solution was the key process in synthesis of polymer spheres. The as-synthesized PTPys possessed a well-defined spherical architecture with diameter ranging from 100–300 nm. The N/Cs-900 electrocatalyst was prepared by directly pyrolyzing PTPys at 900 °C, and the Fe/N-Cs-900 electrocatalyst was synthesized by pyrolyzing PTPys/Fe(OAc)2 mixture at the same temperature. The N/Cs-900 and Fe/N-Cs-900 electrocatalysts both have well-defined spherical architectures, relatively high surface areas, and relatively high N contents. Owing to the synergistic influences, the Fe/N-Cs-900 electrocatalyst exhibits excellent ORR activities in both alkaline and acidic electrolytes together with superior tolerance toward urea, Cl−, MeOH, and CO in acidic media, whereas the N/Cs900 displays ORR activity primarily in alkaline electrolyte. This approach can fabricate heteroatom-doped porous carbon spheres dispensing with the need for any templates, which is of great explored potentiality and utilized value.

Acknowledgments Financial support from Program for NSFC (51674219), China Postdoctoral Science Foundation (2017M610502), Foundation of Hunan Educational Committee (17C1522) and the Construct Program of the Key Discipline in Hunan Province is greatly acknowledged.

110

Applied Catalysis A, General 559 (2018) 102–111

T. Sun et al.

[62] G. Panomsuwan, N. Saito, T. Ishizaki, RSC Adv. 6 (2016) 114553–114559. [63] W.J. Jiang, L. Gu, L. Li, Y. Zhang, X. Zhang, L.J. Zhang, J.Q. Wang, J.S. Hu, Z. Wei, L.J. Wan, J. Am. Chem. Soc. 138 (2016) 3570–3578.. [64] G. Wu, P. Zelenay, Acc. Chem. Res. 46 (2013) 1878–1889. [65] J. Kim, S. Lim, S.K. Kim, D.H. Peck, B. Lee, S.H. Yoon, J. Doohwan, J. Nanosci. Nanotechnol. 11 (2011) 6350–6358. [66] Z. Chen, D. Higgins, H. Tao, R.S. Hsu, Z. Chen, J. Phys. Chem. C 113 (2009) 21008–21013. [67] J.D. Wiggins-Camacho, K.J. Stevenson, J. Phys. Chem. C 113 (2009) 19082–19090. [68] J.D. Wiggins-Camacho, K.J. Stevenson, J. Phys. Chem. C 115 (2011) 20002–20010. [69] Z.Y. Wu, X.X. Xu, B.C. Hu, H.W. Liang, Y. Lin, L.F. Chen, S.H. Yu, Angew. Chem., Int. Ed. 54 (2015) 8179–8183. [70] S. Yasuda, L. Yu, J. Kim, K. Murakoshi, Chem. Commun. 49 (2013) 9627–9629. [71] C.V. Rao, Y. Ishikawa, J. Phys. Chem. C 116 (2012) 4340–4346. [72] K. Parvez, S. Yang, Y. Hernandez, A. Winter, A. Turchanin, X. Feng, K. Müllen, ACS Nano 6 (2012) 9541–9550. [73] T. Ishizaki, S. Chiba, Y. Kaneko, G. Panomsuwan, J. Mater. Chem. A 2 (2014) 10589–10598. [74] R.S. Zhong, Y.H. Qin, D.F. Niu, J.W. Tian, X.S. Zhang, X.G. Zhou, S.G. Sun, W.K. Yuan, J. Power Sources 225 (2013) 192–199. [75] T.N. Tran, M.Y. Song, K.P. Singh, D.S. Yang, J.S. Yu, J. Mater. Chem. A 4 (2016) 8645–8657. [76] W. Niu, L. Li, X. Liu, N. Wang, J. Liu, W. Zhou, Z. Tang, S. Chen, J. Am. Chem. Soc. 137 (2015) 5555–5562. [77] Y.C. Wang, Y.J. Lai, L. Song, Z.Y. Zhou, J.G. Liu, Q. Wang, X.D. Yang, C. Chen, W. Shi, Y.P. Zheng, M. Rauf, S.G. Sun, Angew. Chem., Int. Ed. 54 (2015) 9907–9910.

[46] G.C.K. Liu, J.R. Dahn, Appl. Catal. A: Gen. 347 (2008) 43–49. [47] D. Sebastián, A.G. Ruíz, I. Suelves, R. Moliner, M.J. Lázaro, V. Baglio, A. Stassi, A.S. Aricò, Appl. Catal. B: Environ. 115−116 (2012) 269–275. [48] R. Ma, B.Y. Xia, Y. Zhou, P. Li, Y. Chen, Q. Liu, J. Wang, Carbon 102 (2016) 58–65. [49] J. Tang, J. Liu, C. Li, Y. Li, M.O. Tade, S. Dai, Y. Yamauchi, Angew. Chem., Int. Ed. 54 (2015) 588–593. [50] T. Zhou, Y. Zhou, R. Ma, Z. Zhou, G. Liu, Q. Liu, Y. Zhu, J. Wang, Carbon 114 (2017) 177–186. [51] D. Malko, T. Lopes, E. Symianakis, A.R. Kucernak, J. Mater. Chem. A 4 (2016) 142–152. [52] S. Cai, Z. Meng, H. Tang, Y. Wang, P. Tsiakaras, Appl. Catal. B: Environ. 217 (2017) 477–484. [53] X. Zhu, S.M. Mahurin, S.H. An, C.L. Do-Thanh, C. Tian, Y. Li, L.W. Gill, E.W. Hagaman, Z. Bian, J.H. Zhou, J. Hu, H. Liu, S. Dai, Chem. Commun. 50 (2014) 7933–7936. [54] Y. Luo, B. Li, W. Wang, K. Wu, B. Tan, Adv. Mater. 24 (2012) 5703–5707. [55] X. Zhu, S. Ding, C.W. Abney, K.L. Browning, R.L. Sacci, G.M. Veith, C. Tian, S. Dai, Chem. Commun. 53 (2017) 7645–7648. [56] J.H. Zhu, Q. Chen, Z.Y. Sui, L. Pan, J. Yu, B.H. Han, J. Mater. Chem. A 2 (2014) 16181–16189. [57] L. Lai, J.R. Potts, D. Zhan, L. Wang, C.K. Poh, C. Tang, H. Gong, Z. Shen, J. Lin, R.S. Ruoff, Energy Environ. Sci. 5 (2012) 7936–7942. [58] Z.S. Wu, S. Yang, Y. Sun, K. Parvez, X. Feng, K. Müllen, J. Am. Chem. Soc. 134 (2012) 9082–9085. [59] J. Yang, D.J. Liu, N.N. Kariuki, L.X. Chen, Chem. Commun. 3 (2008) 329–331. [60] S. Ratso, I. Kruusenberg, M. Vikkisk, U. Joost, E. Shulga, I. Kink, T. Kallio, K. Tammeveski, Carbon 73 (2014) 361–370. [61] J. Wu, W. Li, D. Higgins, Z. Chen, J. Phys. Chem. C 115 (2011) 18856–18862.

111