Yi
Zhang‡
ab,
Xiaodong
Zhuang‡
b,
Yuezeng
Su
*a,
Fan
Zhang
*b and
Xinliang
Feng
bc
aSchool of Aeronautics and Astronautics, Shanghai Jiao Tong University, Shanghai 200240, China. E-mail: yzsu@sjtu.edu.cn
bSchool of Chemistry and Chemical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China. E-mail: fan-zhang@sjtu.edu.cn
cMax Planck Institute for Polymer Research, Ackermannweg 10, D-55128 Mainz, Germany
First published on 27th March 2014
A novel B/N co-doped porous carbon nanosheet with a high heteroatom doping content has been successfully prepared. Using amino-functionalized graphene oxide (GO) as the template, oxidation polymerization of aniline, 3-aminophenylboronic acid, and m-phenylenediamine generates GO-based polyaniline nanosheets functionalized with boronic acid (GO-CBP). After high-temperature treatment, graphene-based B/N co-doped carbon nanosheets (G-CBP) are obtained, which show a typical 2D morphology with a thickness of ∼20 nm. After CO2 activation at 1000 °C, the obtained porous carbon nanosheets (G-CBP-a) have a thickness of ∼17 nm and a high specific surface area of 363 m2 g−1. Benefiting from its high surface area, unique 2D sheet nanostructure, and high heteroatom-doping contents (5.4% B and 5.3% N), G-CBP-a exhibits excellent electrochemical performance for the oxygen reduction reaction under alkaline conditions (0.1 M KOH), with a low half-wave potential (−0.27 V for G-CBP-a versus −0.18 V for Pt/C), a dominant four-electron transfer mechanism (n = 3.78 at −0.45 V), and excellent methanol tolerance and durability (10% current decrease after 20000 s operation), as well as a high diffusion-limiting current density (JL = −4.5 mA cm−2).
Graphene, a two-dimensional aromatic monolayer of carbon atoms, has attracted great attention owing to its exceptional physical, chemical, and mechanical properties, as well as its potential applications in electronics, sensors, supercapacitors, and batteries.22–24 Undoubtedly, heteroatom-doped graphene also holds great promise for catalysis.25 In addition to the fabrication of doped graphene by chemical vapor deposition26 involving a vacuum-based system, which is generally complicated and expensive for large-scale production, there is another strategy for the preparation of doped graphene: the direct pyrolysis of heteroatom-containing precursors with graphene or graphene oxide (GO).27
In this work, we present a novel strategy to synthesize B/N co-doped porous carbon nanosheets with high heteroatom doping contents (5.4% B and 5.3% N). Using amino-functionalized GO as the template, oxidation polymerization of aniline, 3-aminophenylboronic acid, and the cross-linker m-phenylenediamine28 generates GO-based polyaniline nanosheets functionalized with boronic acid (GO-CBP). After high-temperature treatment at 1000 °C, the as-produced carbon nanosheets (G-CBP) show a typical 2D morphology with a thickness of about 20 nm. Further activation of G-CBP with CO2 at 1000 °C yields porous carbon nanosheets (G-CBP-a) with a thickness of ∼17 nm and a high specific surface area of 363 m2 g−1. Benefiting from the high surface area, the unique 2D sheet nanostructure, and the high heteroatom-doping contents (5.4% B and 5.3% N), G-CBP-a exhibits excellent electrochemical performance with respect to the oxygen reduction reaction (ORR) under alkaline conditions (0.1 M KOH), with a low half-wave potential (−0.27 V for G-CBP-a versus −0.18 V for Pt/C), a dominant four-electron transfer mechanism (n = 3.78 at −0.45 V), and excellent methanol tolerance and durability (10% current decrease after 20000 s operation), as well as a high diffusion-limiting current density (JL = −4.5 mA cm−2).
The morphologies of AGO, GO-CBP, G-CBP, and G-CBP-a were investigated by scanning electron microscopy (SEM), transmission electron microscopy (TEM), and atomic force microscopy (AFM). The SEM image (Fig. S1a†) shows the typical curved, layer-like feature of AGO, which is similar to that of GO. GO-CBP maintained this 2D morphology, ranging in the sheet size from 200 nm to several micrometers. Based on the SEM and TEM images in Fig. 1a and b, no free porous polymer particles or naked graphene sheets were observed, indicating that most of the monomers were polymerized on the surface of AGO. The AFM image further revealed the same 2D morphology observed in the SEM and TEM images (Fig. 1c). After pyrolysis, G-CBP exhibited a typical 2D sheet nanostructure (Fig. 1d and e), similar to that of GO-CBP. After CO2 activation at 1000 °C, G-CBP-a also exhibited a flexible and crumpled morphology similar to that of GO-CBP and G-CBP (TEM images in Fig. 1b, e, and h). Further, GO-CBP, G-CBP, and G-CBP-a had average thicknesses of ∼30, ∼20, and ∼17 nm, respectively. The decreased thickness of G-CBP can be attributed to the thermal treatment, which caused decomposition and recombination of the carbon framework, while the thickness decrease caused by CO2 activation should be ascribed to the reduction reaction between CO2 and carbon at high temperature.
![]() | ||
Fig. 1 SEM (a, d, and g), TEM (b, e, and h), and AFM (d, f, and i) images of GO-CBP (a, b, and c), G-CBP (d, e, and f), and G-CBP-a (g, h, and i). |
Fourier transform infrared (FT-IR) spectra of CBP, AGO, and GO-CBP are shown in Fig. S2.† The characteristic peaks from the FT-IR spectrum of GO-CBP at 1612 and 1507 cm−1 correspond to the C–C stretching vibration of the quinoid ring and the benzenoid ring, respectively.30 Asymmetric B–O stretching was observed at 1412 cm−1.31 The thermal stability of CBP and GO-CBP was further studied by thermogravimetric analysis (TGA). As shown in Fig. S3,† the weight loss of GO-CBP (21%) is less than that of CBP (24%) at 400 °C. The residue from GO-CBP had the highest weight (50%) compared with that from CBP (47%) and AGO (42%) at 800 °C. This result suggests that incorporation of the graphene layer within polyaniline enhances the thermal stability of the composite.
The porous nature of GO-CBP, G-CBP, and G-CBP-a was evaluated by nitrogen physisorption measurements. The pore size distribution and specific surface area were calculated from the adsorption isotherms based on the nonlocal density functional theory (NL-DFT) and Brunauer–Emmett–Teller (BET) methods. It was found that the isotherms of GO-CBP, G-CBP, and G-CBP-a were type III according to the IUPAC classification.32 A significant increase in the adsorption isotherm when the relative pressure was beyond 0.95 implied the presence of meso/macropores in the samples. A small hysteresis loop at a relative pressure of 0.4–0.95 was observed, indicating the presence of silt pores.32 The pore size distributions of GO-CBP, G-CBP, and G-CBP-a are presented in Fig. 3b. G-CBP and G-CBP-a show peaks at around 1.5, 3.0, and 4.1 nm, suggesting the presence of numerous micro/mesopores. The detailed results are summarized in Table 1. The BET surface areas of GO-CBP, G-CBP, and G-CBP-a were calculated to be 39, 110, and 363 m2 g−1, respectively. The total pore volume of G-CBP-a reached 1.58 cm3 g−1, which is more than twice that of G-CBP (0.664 cm3 g−1). The high specific surface area and pore volume of G-CBP-a with respect to G-CBP can be explained by the drastic reaction between carbon and CO2 at high temperature which produces CO and creates rich micro- and mesopores (1.5, 2.5, and 3.4 nm in Fig. 2b).33
![]() | ||
Fig. 2 (a) Nitrogen adsorption/desorption isotherms of GO-CBP, G-CBP, and G-CBP-a. (b) Pore size distributions calculated by the NL-DFT method. |
GO-CBP | G-CBP | G-CBP-a | |
---|---|---|---|
a Surface area calculated from the N2 adsorption isotherm using the BET method. b Surface area calculated from the N2 adsorption isotherm using the Langmuir method. c Total pore volume at p/p0 = 0.99. d Average pore size for p/p0 between 0.05 and 0.99. e N1, N2, and N3 correspond to quaternary N, pyridinic N and the N in the N–B–C moiety, respectively. f B1 and B2 correspond to the B in the N–B–C moiety and in BC2O, respectively. | |||
S BET (m2 g−1) | 39.1 | 110.4 | 362.9 |
S Langmuir (m2 g−1) | 59.1 | 131.9 | 531.2 |
V tot (cm3 g−1) | 0.456 | 0.664 | 1.58 |
d av (nm) | 46.7 | 24.0 | 17.1 |
N% (wt%) | 9.71 | 10.4 | 5.3 |
N1/N2/N3e | — | 65/19/16 | 17.7/37/45.3 |
B (wt%) | 1.12 | 2.1 | 5.4 |
B1/B2f | — | 60/40 | 43/57 |
X-ray photoelectron spectroscopy (XPS) was further used to investigate the chemical nature of GO-CBP, G-CBP, and G-CBP-a (Fig. S4† and 3). The boron and nitrogen components that were calculated from the XPS results are listed in Table 1. G-CBP has a higher nitrogen content (10.4%) than G-CBP-a (5.3%), while the boron content in G-CBP (2.1%) is lower than that in G-CBP-a (5.4%). This result indicates that boron is more stable than nitrogen in the carbon framework under high temperature and a CO2 atmosphere. As shown in Fig. 3, both the N 1s and B 1s spectra of G-CBP and G-CBP-a show similar chemical compositions. The N in G-CBP and G-CBP-a can be fitted by three peaks corresponding to quaternary N (N1, 401 ± 0.1 eV), pyridinic N (N2, 399.2 ± 0.1 eV), and N in the N–B–C configuration (N3, 398.4 ± 0.1 eV).10 It can be assumed that the proportion of these nitrogen components exhibited a great change before and after activation. The B in G-CBP and G-CBP-a can be fitted by the B–CO10 (B1, 191.9 ± 0.1 eV) and N–B–C (B2, 190.8 ± 0.1 eV) moieties.10 The high percentage of N–B bonds (3.1 wt% based on B2) indicates that nitrogen and boron can be easily coordinated under high-temperature treatment.34 The relative content of B1 significantly increased after the activation process, possibly due to the higher reactivity of carbon with CO2 than boron with CO2.
To gain insight into the electrocatalytic behavior of G-CBP-a, we examined the electrocatalytic activities of the B/N co-doped porous carbon nanosheets toward the ORR under alkaline conditions (0.1 M KOH). The ORR catalytic activity of G-CBP and G-CBP-a was first evaluated by cyclic voltammetry (CV) (Fig. 4a) in N2- and O2-saturated 0.1 M KOH as well as in O2-saturated 0.1 M KOH solution with 1.0 M methanol. The oxygen reduction peak for G-CBP-a was observed at −0.24 V, whereas the signal vanished in N2-saturated 0.1 M KOH. G-CBP exhibited the same phenomenon, while the oxygen reduction peak for G-CBP (−0.32 V) was more negative than that of G-CBP-a, which indicates higher energy consumption for oxygen reduction. No noticeable change was observed in the CV results of G-CBP and G-CBP-a in O2-saturated 0.1 M KOH solution with 1.0 M methanol. This result indicates good methanol tolerance for both G-CBP and G-CBP-a.
To further explore the ORR performance of G-CBP-a, rotating-ring disk electrode (RRDE) voltammetry and rotating-disk electrode (RDE) voltammetry studies were performed. The onset potential for G-CBP-a in RRDE voltammograms was at approximately −0.23 V (Fig. 4b), which is close to that identified from CV measurements (−0.24 V, Fig. 4a). The onset potentials for G-CBP and Pt/C are −0.25 and −0.07 V, respectively. The RDE voltammetric profiles in O2-saturated 0.1 M KOH solution showed that the current density was enhanced upon increasing the rotation rate (from 225 to 1600 rpm, Fig. 4c). Koutecky–Levich (K–L) plots (the inset in Fig. 4c) with a good linear relationship for the G-CBP-a electrode were drawn from the linear sweep voltammetry (LSV) curves (Fig. 4c) at various rotation rates. Linearity and parallelism of the plots are usually taken as an indication of first-order reaction kinetics with respect to the concentration of dissolved O2. According to eqn (S2)–(S4),†1,35 the number of transferred electrons (n) and the kinetic current density (JK, Fig. 4e) can be calculated from the slope and intercept of the K–L plots, respectively. Notably, the calculated number of transferred electrons (n) per O2 molecule involved in the ORR was 3.78 at −0.45 V. A similar value of n (∼3.85 at −0.45 V) can be calculated from the RRDE curve (Fig. 4b and eqn (S1)†).36 In contrast, the calculated n for G-CBP was only 3.25 at −0.45 V. These results highlight that, for G-CBP-a, the ORR proceeds via a primary four-electron pathway.
The corresponding LSV curves for G-CBP, G-CBP-a, and Pt/C in an O2-saturated 0.1 M KOH solution at a rotation rate of 1600 rpm are compared in Fig. 4d. G-CBP-a exhibited a high diffusion-limiting current, reaching −4.5 mA cm−2, which was close to that of Pt/C (−5.1 mA cm−2). G-CBP showed a relatively low diffusion-limiting current of about −3.8 mA cm−2. Remarkably, G-CBP-a displayed a low half-wave potential at −0.27 V, which was only 90 mV higher than that of Pt/C (−0.18 V) (Fig. 4d) and 50 mV lower than that of G-CBP. Additionally, the kinetic current density of G-CBP-a (21.8 mA cm−2) at −0.45 V was slightly higher than that of Pt/C (20.4 mA cm−2), and significantly greater than that of G-CBP (8.43 mA cm−2). These results clearly reveal the enhanced effect on ORR activity by CO2-activated B/N co-doped porous carbon nanosheets. Mostly, the results could be explained by integration of the large specific surface area, large pore volume37 and high N/B-doping contents of as-prepared samples.38 Actually, nitrogen as one of the most effective doping elements39 confidently enables improvement of the activity of metal-free carbon-based materials.40 Our work demonstrated that doping with different contents of N (10.4% to 5.3%) and B (2.1% to 5.4%) essentially offered quite different catalytic activities of the resulting materials either before or after activation with CO2; thus it is quite desirable for further exploring the contribution of B/N in such a kind of catalytic system.
Since durability is one of the major concerns in current fuel cell technology, the stability of G-CBP-a over time was tested at a constant voltage of −0.36 V for 20000 s in a 0.1 M KOH solution saturated with O2 at a rotation rate of 1600 rpm (Fig. 4f). Remarkably, the corresponding current–time (i–t) chronoamperometric response of G-CBP-a exhibited a fast decrease of 8% within the first 500 s, and then a slow attenuation with a high current retention of 89% after 20
000 s. In contrast, Pt/C showed a gradual decrease, with a current loss of approximately 76% measured after 20
000 s. This suggests that the durability of G-CBP-a is superior to the commercially available Pt/C catalyst.
Footnotes |
† Electronic supplementary information (ESI) available: Experimental details; SEM and TEM of AGO; FTIR and TGA; XPS spectra of GO-CBP; some electrochemical data. See DOI: 10.1039/c4ta00814f |
‡ These two authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2014 |