Effect of boron–nitrogen bonding on oxygen reduction reaction activity of BN Co-doped activated porous carbons

Abstract

The effect of boron and nitrogen (BN) bonding on the activity of the oxygen reduction reaction was investigated when boron and nitrogen atoms are chemically inserted into activated carbons. With the readily available B and N doping precursors of boric acid and urea, heteroatom-doped carbon materials were synthesized under varying pyrolysis temperatures and mixing ratios of the precursor to activated carbon (AC). According to electrochemical and structural analyses, the heteroatom was doped into the carbon lattice at 1000 °C, whereas annealing temperatures below 850 °C did not achieve the doping. It is striking that the BN co-doped AC shows much higher doping contents than the B or N single-doped AC despite the use of the same amounts of doping precursors as in the single doping case. The electrochemical analysis indicates that higher B or N doping contents enhance the level of ORR activity with the four-electron pathway, especially for the BN co-doping case due to the formation of B–N–C bonds. The results of this study demonstrate the potential of heteroatom-doped AC as a cost-effective alternative catalyst for oxygen reduction reaction in fuel cells.

---

1. Introduction

Fuel cells, which convert chemical energy into electrical energy, are currently considered promising candidates for efficient energy conversion technologies.^1,2 In the operation of a fuel cell, the oxygen reduction reaction (ORR) on the cathode limits the overall performance due to its slower reaction rate compared to hydrogen oxidation reaction (HOR) on the anode.^3,4 Therefore, identifying the proper catalyst materials for the cathode has become a major area of research. Currently, Pt-based materials are the most effective ORR catalysts.⁵ However, as the high cost of Pt is the main obstacle for the practical application of fuel cells, alternative catalysts are being widely pursued.^6,7 Abundant and non-precious metals such as iron, cobalt, and metal oxides as well as non-metallic carbon-based materials have been proposed.^8–10

Heteroatom-doped (e.g., B or N) carbon materials have been highlighted as metal-free catalysts feasible for ORR applications. Boron and nitrogen are especially promising candidates for chemical doping, as they act as p-type or n-type dopants, respectively, and have sizes comparable to that of carbon atoms.^11–14 Existing examples of B- or N-doped carbon materials include boron-doped graphene,¹⁵ nitrogen-doped graphene,^16,17 graphene-based carbon nitride nanosheets,¹⁸ nitrogen-doped ordered mesoporous graphitic arrays,¹⁹ and nitrogen-doped carbon nanotube (CNT) arrays.²⁰ Several studies have attempted to synthesize BN co-doped carbon materials. The effect of B–N moieties on electrocatalytic activity was investigated in BN co-doped carbons prepared by the carbonization of polyfurfuryl alcohol²¹ and BN co-doped graphene oxide derived from CO₂.²² Vertically aligned BCN nanotubes²³ and BCN graphene²⁴ were synthesized with BN co-doping, showing positive effects on ORR. However, the mechanism of the enhancement of ORR by BN co-doping has yet to be understood. Specifically, there are limited studies using an identical carbon source to consistently evaluate the effects of B doping, N doping, and BN co-doping on ORR activity. Most prior studies employed graphene or CNT as a carbon source,^{15–21,23,24} but either single doping or BN co-doping was compared to non-doping cases in terms of ORR performance. Thus, with a single carbon source, this work aims to elucidate how BN co-doping demonstrates higher electrocatalytic activity than single B- or N-doping.

Herein, the effects of B-doping, N-doping, and BN co-doping on ORR activity are investigated using a single carbon source of activated carbons (ACs) as a model carbon system. To the best of our knowledge, heteroatom-doping of AC and its potential in fuel cell electrodes have rarely been explored. While there are elegant methods such as chemical vapour deposition²⁵ and arc-discharge²⁶ to incorporate a heteroatom into a carbon network, this study will achieve B- and N-doping into ACs in a cost-effective manner by simple pyrolysis of AC mixed with boric acid and urea as B- and N-doping precursors. AC, boric acid, and urea are inexpensive sources of carbon, boron, and nitrogen, respectively. The pyrolysis temperature and the mixing ratio of AC to boric acid or urea are controlled to maximize the electrocatalytic activity of the AC-based catalysts. The electrodes assembled by the catalysts will be tested using a cyclic voltammetry (CV), rotating disk electrode (RDE) and rotating ring-disk electrode (RRDE) in an effort to evaluate the electrocatalysis performance. The B-doped, N-doped, and BN co-doped ACs will be analysed by scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), Raman spectroscopy and attenuated total reflection Fourier transform infrared (ATR-FTIR) spectroscopy to determine the effects of the microstructure on ORR. Based on these analyses, we will show that the equivalent ORR performance levels exist between the B and N single-doping cases, whereas BN co-doping provides comparable performance to a platinum-based catalyst due to the new formation of B–N bonding in the carbon network.

2. Experiment

2.1. Materials

Commercially available activated carbon (DARCO®, −100 mesh particle size, powder, 600 m² g⁻¹ surface area) and urea (with a purity of >99.5%) were purchased from Sigma-Aldrich. Boric acid (with a purity of >99.0%) was supplied by Fluka. Methanol (with a purity of >99.6%) and sodium hydroxide (with a purity of >95.0%) were purchased from Junsei Chemical Co. Ltd. Ethanol (with a purity of >99.5%) was obtained from Samchun Chemicals. Platinum on graphitized carbon (10 wt% loading) and a Nafion® perfluorinated resin solution (5 wt% in lower aliphatic alcohols and water; contains 15–20% water) were supplied by Sigma-Aldrich. All chemicals were used as received without further purification.

2.2. Synthesis of B-, N-, and BN co-doped carbon materials

For the synthesis of B-doped activated carbon (B-AC), 0.05 g (B-AC 10 : 1) or 0.5 g (B-AC 1 : 1) of boric acid was dissolved in 5 mL of deionized (DI) water and the solution was mixed with 0.5 g of AC for 15 min via ultra-sonication. After the evaporation of the DI water, the mixed powder was loaded onto a 5 cm³ alumina crucible boat. The boat was then put into a horizontal quartz tube and the tube was placed inside a furnace (GSL1100X, MTI Co.). The tube was heated to 700 °C (for B-AC700), 850 °C (for B-AC850), and 1000 °C (for B-AC1000) from room temperature at a heating rate of 5 °C min^-1, held at these temperatures for 2 h, and finally cooled down to room temperature under Ar atmosphere. The solid product was put into a 100 mL glass beaker containing DI water at 85 °C for 4 h with stirring at 500 rpm to remove any unreacted reactants or byproducts. After filtration, the product was washed with DI water at room temperature for 15 min and filtered again. This filtration and washing process was repeated five times. Finally, the product was washed with ethanol for 15 min and evaporated for 6 h at 120 °C. For the synthesis of N-doped activated carbon (N-AC), the same procedure was repeated using 0.05 g (for N-AC 10 : 1) or 0.5 g (for N-AC 1 : 1) of urea instead of boric acid with 0.5 g of AC. For the BN co-doped activated carbon (BN-AC), 0.05 g of boric acid and 0.05 g of urea (for BN-AC 10 : 1 : 1) or 0.5 g of boric acid and 0.5 g of urea (for BN-AC 1 : 1 : 1) were used with 0.5 g of AC while adopting the same procedure used for each single doping case. Pure AC without the mixing of boric acid or urea was also heat-treated with the same procedure as a reference sample.

2.3. Characterization

Scanning electron microscopy (SEM) images were obtained with a field emission SEM (Magellan 400) device by FEI Co. X-ray photoelectron spectroscopy (XPS) data were obtained using the Multi-Purpose XPS (Sigma Probe) device by Thermo VG Scientific equipped with an MXR1 gun (400 μm). The powder samples were put onto carbon tapes for the SEM and XPS analyses. ATR-FTIR spectra were observed by using Alpha FT-IR spectrometer with a diamond ATR module. Raman spectral data were acquired using a high-resolution dispersive Raman microscope (ARAMIS) by Horiba Jobin Yvon. An Ar ion CW laser (514.5 nm) was used for sample excitation.

2.4. Electrochemical analysis

Electrochemical measurements of a cyclic voltammetry (CV), rotating disk electrode (RDE), and rotating ring-disk electrode (RRDE) measurements were performed by a three-electrode system using a RRDE-3A rotating ring disk electrode rotator by ALS. For CV and RDE measurements, Ag/AgCl, platinum wire, and glassy carbon (3 mm diameter) were used as reference, counter, and working electrodes, respectively. Pt ring/GC disk electrode was also used as a working electrode for RRDE measurements. An electrochemical analyzer (CHI 600D) by CH Instruments was used to control the potential. For the measurements, the samples were deposited onto the working electrode. 5 mg of sample was dispersed by ultra-sonication for 15 min in a 1 mL solution of DI water, methanol, and Nafion® perfluorinated resin solution at a volumetric ratio of 15 : 4 : 1. Then, 5 μL of the solution with a dispersed catalyst was loaded onto the polished glassy carbon electrode and dried in air. CV, RDE, and RRDE measurements were performed in 1 M of NaOH electrolyte saturated by oxygen at scan rates of 50 mV s⁻¹, 10 mV s⁻¹, and 10 mV s⁻¹, respectively. All measurements were repeated using at least three electrodes for one sample and 20 cycles for each electrode to insure consistency of the measurements.

3. Results and discussion

3.1. Annealing temperature effect on heteroatom-doping levels and ORR

The effects of different annealing temperatures (700 °C, 850 °C, and 1000 °C) on the doping level were investigated using three different samples of pure AC, B-AC 10 : 1 and N-AC 10 : 1. According to the electrochemical analysis from CV, there was no improvement of ORR activity for either B-AC 10 : 1 or N-AC 10 : 1 compared to AC when they were annealed at 700 °C, as shown in Fig. 1(a). Both the peak potentials and current densities were constant and did not much improve from AC 700.

Fig. 1 CV measurements of samples synthesized at (a) 700 °C and (b) 1000 °C.

The same trend was observed at 850 °C (Fig. S1 in ESI†), but a remarkable positive shift of peak potential was shown at the annealing temperature of 1000 °C (Fig. 1(b)). To understand the reason for this difference, the atomic composition at the surface is investigated by X-ray photoelectron spectroscopy (XPS). B and N atoms or their moieties were not detected by the XPS survey data or in the XPS spectra of the samples annealed at 700 and 850 °C (Table S1 and Fig. S2 in ESI†). On the other hand, B and N were detected on the samples annealed at 1000 °C (Table 1). Here, the results indicate that B- or N-doping on AC is properly achieved at 1000 °C, but not at 700 °C or 850 °C. Therefore, we selected 1000 °C as the preferred annealing temperature for the execution of BN co-doping. We then investigated B-doping, N-doping, and BN co-doping cases using electrochemical measurements and various structural analyses to understand the origin of the enhanced ORR activities of the electrocatalysts.

Table 1 Atomic% of C, O, B, and N as measured by the XPS survey (%) for samples annealed at 1000 °C

Atom	AC	B-AC 10 : 1	B-AC 1 : 1	N-AC 10 : 1	N-AC 1 : 1	BN-AC 10 : 1	BN-AC 1 : 1
C	96.53	94.34	86.21	95.54	91.66	93.08	67.01
O	3.47	5.12	11.93	3.33	3.72	4.63	14.37
B	—	0.54	1.86	—	—	0.87	8.13
N	—		—	1.14	4.6	1.42	10.48

---

For the samples with the 10 : 1 mixing ratio (AC to doping precursor), the doping contents are low, which are 0.54–0.87% for the B atoms (B-AC 10 : 1 & BN-AC 10 : 1) and 1.13–1.42% for the N atoms (N-AC 10 : 1 & BN-AC 10 : 1). For the samples with 1 : 1 mixing ratio, doping contents are higher and the peaks can be easily distinguished from XPS survey spectra (Fig. S3 in ESI†). Due to the higher proportion of N atoms in urea compared to B atoms in boric acid, the N-doping contents are higher when the same precursor ratio is used. For oxygen atoms, the contents increase remarkably in the samples with high mixing ratios of boric acid. This occurs because three oxygen atoms are present for one molecule of boric acid, while only one oxygen atom exists for one molecule of urea.

It is striking that the doping contents of the B and N atoms are significantly increased when they are co-doped, despite the use of the same amount of precursors as employed in the single-doping cases. This is likely due to the binding of B atoms with N atoms, as observed in a prior study,²⁷ which demonstrated that the addition of boron increases the nitrogen doping contents in thin film samples. Our experiment result in Table 1 shows the same trend of an increasing doping amount when B and N are doped simultaneously. The existence of B–N bonding in the carbon network will be proved later with more detailed microscopic and spectroscopic analyses.

3.2. Electrochemical analysis of samples annealed at 1000 °C

To investigate the doping effect, the electrocatalytic activity of the synthesized carbon materials at 1000 °C was evaluated by CV, RDE, and RRDE measurements in 1.0 M NaOH solution saturated with 1 atm of oxygen gas at room temperature.

Table 2 and Fig. 2 show the CV results of B-ACs, N-ACs, and BN-ACs, comparing them with those of the two reference materials of pure AC (non-doping) and Pt on graphitized carbon (Pt/GC 10 wt%). The peak potentials at which the cathodic currents are maximized positively shift as the amount of doping precursors increases, especially for the co-doped ACs. The positive shift of the peak potentials indicates rapid ORR kinetics and higher ORR activity levels. Therefore, higher doping precursors and co-doping have positive effects on ORR. For the BN-AC 1 : 1 case, which was co-doped with the highest levels of precursors, the cathodic current maximizes at −0.197 V; this can be converted to 0.803 V when the value is expressed in terms of standard reversible hydrogen electrode (RHE). If we compare this peak potential with that in the previous study that used the same reference electrode, the BN-AC shows better performance than B-doped carbon nanotube (0.69 V based on the RHE)²⁸ and boron-doped graphene (0.67 V based on the RHE).¹⁵ However, the Pt/GC 10 wt% catalyst still shows much better ORR activity than the B or N-doped AC due to its low peak potential (−0.117 V).

Table 2 CV peak potentials and current densities at −0.8 V for heteroatom-doped ACs annealed at 1000 °C

Weight ratios AC : doping precursor : (co-doping)	Pt/GC 10 wt% (V) (mA cm^−2)	AC (1000 °C) (V) (mA cm^−2)	B-AC (V) (mA cm^−2)	N-AC (V) (mA cm^−2)	BN-AC (V) (mA cm^−2)
10 : 1: (1)	−0.117 ± 0.003 5.5	−0.229 ± 0.005 2.8	−0.213 ± 0.003 2.7	−0.219 ± 0.003 2.5	−0.210 ± 0.002 7.1
1 : 1: (1)			−0.204 ± 0.002 3.1	−0.207 ± 0.003 2.4	−0.197 ± 0.003 6.3

---

Fig. 2 CV of ORR in an O₂-saturated 1.0 M NaOH solution at a scan rate of 50 mV s⁻¹ (a) for B-ACs, (b) for N-ACs, and (c) for BN-ACs synthesized at 1000 °C.

Current density is another indicator of ORR performance in a proton exchange membrane fuel cell (PEMFC).²⁹ BN co-doping has a clear effect on the improved current density (Table 2 & Fig. 2(c)), compared to the single doping of B or N (Fig. 2(a) and (b)). The most striking result is that the two co-doped BN-ACs provide higher current densities (7.1 & 6.3 mA cm⁻² at −0.8 V) than the Pt/GC 10 wt% catalyst (5.5 mA cm⁻² at −0.8 V) at a wide range of potentials, including the potential range for the real fuel cell operation (0.6–0.7 V vs. RHE). This synthesized BN-AC also shows its stability in terms of the consistent ORR activity even after 5000 cycles at a scan rate of 50 mV s⁻¹ (Fig. S4†).

In the rotating disk electrode (RDE) measurements (Fig. 3), results similar to the CV measurements were obtained. Current density increases remarkably for the BN-AC and is comparable to that of the Pt/GC 10 wt% catalyst, consistent with the CV measurements. The onset potential of linear sweep voltammographs (LSV) for the pure AC sample was −0.13 V. The onset potential of the doped materials did not change remarkably but instead slightly shifted in the positive direction for B-AC 10 : 1 and N-AC 10 : 1, at −0.12 V and −0.11 V, respectively. The other samples, i.e., B-AC 1 : 1, N-AC 1 : 1, BN-AC 10 : 1, and BN-AC 1 : 1 samples, show onset potentials almost identical to that of the pure AC sample. Fig. 3 shows the results at a rotating rate of 1600 rpm, but the onset potentials remain at the same level while varying the rotating rates from 400 rpm to 2500 rpm (refer to Fig. S5 in ESI†).

Fig. 3 RDE linear sweep voltammographs for the ORR in 1.0 M of an O₂-saturated NaOH solution with a rotating rate of 1600 rpm at a scan rate of 10 mV s⁻¹ (a) for B-ACs (b) for N-ACs and (c) for BN-ACs prepared at 1000 °C.

Tafel plots were also acquired using RDE data (Fig. S6†). Here, kinetic current densities were calculated from the mass-transport correction of RDE using the following equation:

when I_K is the kinetic current density, I is the measured current density, and I_L is the limiting current density.³⁰ B and/or N-doped materials have smaller Tafel slopes than the slope of pure AC. This indicates they have higher electrocatalytic activity than AC, even though it is inferior to the activity of the Pt/C catalyst.

To understand the ORR mechanism, electron transfer number (n) per oxygen molecule and hydrogen peroxide yield for the ORR are calculated from RRDE measurements, using the equations given below.

where, I_D is the disk current (A), I_R is the ring current (A), and N is the RRDE collection efficiency, which is affected by geometrical features of electrode, and determined as 0.424 herein.

Using eqn (2), electron transfer number (n_e⁻) can be calculated, and the results are shown in Fig. 4(a). The electron transfer number of BN-AC 1 : 1 is close to 4.0 at potentials of −0.3 to −1.0 V, indicating that ORR takes place in the desired four-electron pathway. For the pure AC sample, the two-electron pathway is dominant for the entire range of potentials, which involves the generation of hydroxide ions and results in an inefficient energy conversion (Fig. 4(b)).

Fig. 4 (a) RRDE electron transfer number and (b) hydrogen peroxide yield with a rotating rate of 2500 rpm at a scan rate of 10 mV s⁻¹ in 1 M NaOH.

All of the samples show an identical pattern with varying potentials, that n increases as the potential increases (Fig. 4(a)).

Table 3 shows the n values at −0.6 V. The two-electron pathway dominates the overall ORR reaction in the case of pure AC while co-doped materials provide approximately four electron pathways. The single-doped materials have an intermediate electron transfer number combined between two and four from pure AC and BN co-doped materials. Therefore, for the same source carbon and the same annealing temperature, the enhanced ORR activity can be identified for ACs with B- or N-doping, especially for BN co-doping.

Table 3 Electron transfer number (n) for the ORR at −0.6 V

AC : doping precursor : (co-doping)	AC (1000 °C)	B-AC	N-AC	BN-AC
10 : 1	2.99	3.16	3.47	3.61
1 : 1		3.16	3.38	3.81

---

3.3. Morphological analysis using SEM and XRD

Fig. 5 shows the SEM image of AC, which is the reference sample annealed at 1000 °C, and of BN-AC 1 : 1, which performed best as an electrocatalyst for ORR among all synthesized samples. The porous structure in the annealed AC (Fig. 5(a)) is not different from that in the BN co-doped AC (Fig. 5(b)), indicating that B–N doping does not modify the original structure of AC. Fig. 5(c) shows more magnified SEM image of BN-AC 1 : 1, and Fig. 5(d) presents its elemental mapping results for B, C, N, and O (please refer to Fig. S7† for the mapping of each element). Here, all of the elements are uniformly dispersed in the samples. With elemental mapping analysis, 8%, 70%, 10%, and 13% of B, C, N, and O are detected, respectively.

Fig. 5 SEM images at 1500× for (a) annealed AC and (b) BN-AC 1 : 1, respectively. (c) SEM images for BN-AC 1 : 1 at 20 000×, and (d) its elemental mapping image using EDAX.

X-ray diffraction (XRD), which is an effective technique for investigating the molecular structure of carbon materials, shows identical patterns for all sample types (Fig. S8†). Two broad peaks, which are observed at the 2θ range between 20–30° and 40–50°, are (002) and (100) reflections.³¹ These broad peaks are common for activated carbon³² and indicate its amorphous structure. There are a few sharp peaks, and all of them are assigned to SiO₂, which is likely an impurity in the precursors of the commercially available AC. Any other impurity was not observed in XRD measurements.

3.4. Structural analysis using XPS, FTIR, and Raman spectroscopy

Significant difference was not detected in morphological analyses before and after B- and N-doping through pyrolysis. Therefore, detailed chemical bonding structures should be investigated to understand the reason why the doping affected electrocatalytic activity. XPS deconvolution of B1s and N1s spectra were performed to identify surface compositional information and bonding state.

According to the analysis of the B1s spectra (Fig. 6(a)), the peaks appear at around 189.5 eV, 191 eV and 192.4 eV after deconvolution. These peaks are assigned to BC₃, BC₂O, and BCO₂, respectively.^2,11,15 The proportion of O–B–C moieties increases as the B-doping content increases due to the increased ratio of oxygen atoms in the B-AC 1 : 1 sample from 5.12% to 11.93% compared to the B-AC 10 : 1 sample. Similar to B–C moieties, O–B–C moieties can also act as active sites for ORRs due to the greater electronegativity difference between boron and oxygen atoms.^28,33,34 For the co-doped materials (Fig. 6(b)), another peak is observed at around 190.1 eV. This binding energy is higher than that of B–C bonding (187.5–189.5 eV) and is lower than that of B–O bonding (191–193 eV). It therefore seems to be related to B–N bonding,³⁵ although the exact chemical state may not be assigned. It can be B–N–C moiety³⁶ or B–N–B moiety in hexagonal (h)-BN, for which the peak was reported at 190.2–6 eV.^21,35,37 Alternatively, it may be B–C–N materials, which have an approximate peak range of 188 to 191 eV.^38,39

Fig. 6 (a) XPS B1s scan of B-doped materials and (b) XPS B1s scan of BN co-doped materials.

The N1s spectra are also analysed (Fig. 7(a)), where the peaks are deconvoluted and the centers are observed to exist at around 398.4 eV, 400.1 eV, and 401.3 eV, which are assigned to pyridinic, pyrrolic, and quaternary N atoms.^8,12,16–18 As the N-doping content increases, the proportion of pyridinic-N also increases. It was reported³⁶ that pyridinic N groups in BN co-doped carbon materials act as active sites for ORR and that a higher ratio of the pyridinic N moiety leads to better electrocatalytic activity toward ORRs. Similar to the B1s spectra, the co-doped materials show an additional peak around 397.8–397.9 eV. This additional peak has a lower binding energy than N–C bonding and can be assigned to N–B bonding (Fig. 7(b)).

Fig. 7 (a) XPS N1s scan of N-doped materials (b) XPS N1s scan of BN co-doped materials.

According to the XPS analyses of B1s and N1s, it is certain that there are peaks that are related to B–N bonding. The portion of B–N bonding approximates that of C–N bonding in the N1s spectra (Fig. 7(b)), indicating that those two bonds are combined to each other and the dominant chemical state of B–N related species is not the h-BN form but the B–N–C form.

The attenuated total reflection Fourier transform infrared (ATR-FTIR) spectra also support the existence of B–C, C–N, and B–N bonds instead of h-BN bonds (Fig. 8). Absorption bands at 840 cm⁻¹ (as a shoulder), 1130 cm⁻¹, 2220 cm⁻¹ can be assigned to B–N stretching,⁴⁰ B–C stretching^36,41 and C≡N peak,^42,43 respectively. Those peaks are not observed in absorption spectra of pure AC sample. Peak at 1550 cm⁻¹, which is intensified compared to pure AC, may be the sum of C–C⁴⁴ and C=N bands.⁴⁵ The shoulder at 1420 cm⁻¹ is assigned to C–C ring stretching⁴⁴ or asymmetric B–O stretching.⁴⁶ N–B–H and B–H peaks are also observed at 1030 cm⁻¹ and 2360–2390 cm⁻¹, respectively.⁴⁰ Another absorption bands, which are observed in both spectra of AC and BN-AC 1 : 1, are assigned to carbon related peaks such as C–O, C=O, C–H. Absorption bands related to h-BN, which are observed at around 770 cm⁻¹ and 1360 cm⁻¹, are not observed.^24,36,42,47 The analysis result of ATR-FTIR spectra supports the fact that B–N related moiety does not dominantly exist as h-BN form, but as B–N–C form.

Fig. 8 ATR-FTIR spectra of AC and BN-AC 1 : 1 sample.

The B–N–C moiety appears to have a positive effect on the ORR, as supported by the simulation results.⁴⁸ According to the simulation work, the B–N–C moieties have positive effects on ORRs due to the reduced free energy caused by neighbouring N and B atoms. Thus, our work here confirms the simulation results experimentally.

Another important result is that the sequence of B- and N-doping does not affect the ORR activity for the BN co-doped AC. It was reported³⁶ that BN co-doping has a promoting effect on ORR activity only when doping takes place on graphene separately via a two-step method. It was claimed that h-BN bonding arises during the simultaneous BN co-doping process for graphene and that it is not favourable for ORR activity. However, in the case of AC, the single-step BN co-doped materials show similar or even better ORR activity when compared to the two-step doped materials from the two sequential doping processes of N and B atoms (Fig. S9†). According to the XPS and ATR-FTIR analysis results, B–N–C bonding can arise dominantly for BN co-doped ACs instead of h-BN in graphene,³⁶ contributing to the improved ORR performance. Therefore, the simultaneous heteroatom co-doping method can be used to enhance the ORR electrocatalytic activity without separating the B- and N-doping procedures.

The C1s spectra show similar shapes for all samples types (Fig. S10†). After deconvolution, sp² C–C bonding centered around 284–284.2 eV, sp³ C–C bonding centered around 285.3, and C–O bonding centered around 289 eV are observed.⁴⁹ The existence of about 30% of sp³ C–C bonding indicates that the synthesized materials are not fully crystallized carbon but are amorphous carbon.⁵⁰ As the amount of doping increases for the co-doped materials, the portion of sp³ C–C bonding increases as follows: 27.7% for AC, 28.0% for BN-AC 10 : 1, and 30.5% for BN-AC 1 : 1.

Raman spectroscopy is another useful tool for investigating structural changes in carbon materials.⁵¹ For graphitized carbon materials, two peaks are observed around 1550 cm⁻¹ and 1360 cm⁻¹, which are the G band and the D band, respectively.⁵² The intensity ratio (I_D/I_G) is an important parameter that shows the sp³- to sp²-bonding ratio and the degree of disorder.^53,54 The Raman spectra for pure AC and BN co-doped carbon materials are shown in Fig. 9. Both D band and G band appear around 1340–50 cm⁻¹ and 1600 cm⁻¹, respectively. The broad bands between 2500 and 3300 cm⁻¹ are assigned to 2D, D + G, and 2G bands. The intensity ratios for AC, BN-AC 10 : 1, and BN-AC 1 : 1 are 1.083, 1.058, and 1.003, respectively. The intensity ratio decreases as the doping content increases. This trend differs completely from that associated with graphitized carbon materials such as graphene or CNTs, as heteroatom doping usually creates more defects in the lattice of graphitized carbon materials, and those defects increase the intensity ratio.^55,56 However, even if the doping content and corresponding defects are increased in amorphous carbon materials, the intensity ratio I_D/I_G becomes smaller, which is consistent with previous research on amorphous carbons.^50,53,57 The XRD patterns (Section 3.3) and the C1s spectra analysis by XPS demonstrated that the synthesized materials are amorphous carbon; therefore, the decreased intensity ratio of the co-doped materials may indicate more amorphization of the co-doped activated carbons.

Fig. 9 Raman spectra of AC, BN-AC 10 : 1, and BN-AC 1 : 1.

4. Conclusion

We have introduced a simple and facile method to enhance the electrocatalytic activity of cost-effective activated porous carbon as an ORR catalyst using boron and nitrogen doping. Compared to the B- or N-single doping cases, BN co-doping is more effective way to improve the electrocatalytic activity in terms of peak potential, current density, and favourable pathway of four electron transfer. The co-doped materials show higher doping contents even when they are synthesized under conditions identical to those of single-doped materials, providing positively shifted peak potentials and higher current densities. The main contributor to the enhanced ORR performance is the formation of the B–N–C group in the carbon lattice regardless of the doping sequence of B and N. This work proposes a single-step BN co-doping process to improve the electrocatalytic activity of AC and demonstrates that the BN co-doped AC can be an inexpensive and promising alternative to platinum catalysts for ORR activity.

Acknowledgements

The authors are grateful for the financial support from the Korea CCS R & D Center funded by the Ministry of Science, ICT, and Future Planning (no. NRF-2014M1A8A1049297).

Notes and references

1. Z. Yang, H. Nie, X. A. Chen, X. Chen and S. Huang, J. Power Sources, 2013, 236, 238–249.
2. X. Bo and L. Guo, Phys. Chem. Chem. Phys., 2013, 15, 2459–2465.
3. M. K. Debe, Nature, 2012, 486, 43–51.
4. R. A. Costa and J. R. Camacho, J. Power Sources, 2006, 161, 1176–1182.
5. B. Lim, M. Jiang, P. H. Camargo, E. C. Cho, J. Tao, X. Lu, Y. Zhu and Y. Xia, Science, 2009, 324, 1302–1305.
6. A. Serov and C. Kwak, Appl. Catal., B, 2009, 90, 313–320.
7. M. Lefevre, E. Proietti, F. Jaouen and J. P. Dodelet, Science, 2009, 324, 71–74.
8. Z. S. Wu, S. Yang, Y. Sun, K. Parvez, X. Feng and K. Mullen, J. Am. Chem. Soc., 2012, 134, 9082–9085.
9. G. Wu, K. L. More, C. M. Johnston and P. Zelenay, Science, 2011, 332, 443–447.
10. J. Zhang, A. Byeon and J. W. Lee, Chem. Commun., 2014, 50, 6349–6352.
11. G. Jo and S. Shanmugam, Electrochem. Commun., 2012, 25, 101–104.
12. W. Y. Wong, W. R. W. Daud, A. B. Mohamad, A. A. H. Kadhum, K. S. Loh and E. H. Majlan, Int. J. Hydrogen Energy, 2013, 38, 9370–9386.
13. T. Wu, H. Shen, L. Sun, B. Cheng, B. Liu and J. Shen, New J. Chem., 2012, 36, 1385.
14. K. T. Cho, S. B. Lee and J. W. Lee, J. Phys. Chem. C, 2014, 118, 9357–9367.
15. Z.-H. Sheng, H.-L. Gao, W.-J. Bao, F.-B. Wang and X.-H. Xia, J. Mater. Chem., 2012, 22, 390.
16. Z. Lin, G. Waller, Y. Liu, M. Liu and C.-P. Wong, Adv. Energy Mater., 2012, 2, 884–888.
17. L. Qu, Y. Liu, J.-B. Baek and L. Dai, ACS Nano, 2010, 4, 1321–1326.
18. S. Yang, X. Feng, X. Wang and K. Mullen, Angew. Chem., 2011, 50, 5339–5343.
19. R. Liu, D. Wu, X. Feng and K. Mullen, Angew. Chem., 2010, 49, 2565–2569.
20. K. Gong, F. Du, Z. Xia, M. Durstock and L. Dai, Science, 2009, 323, 760–764.
21. J.-I. Ozaki, N. Kimura, T. Anahara and A. Oya, Carbon, 2007, 45, 1847–1853.
22. A. Byeon and J. W. Lee, J. Phys. Chem. C, 2013, 117, 24167–24173.
23. S. Wang, E. Iyyamperumal, A. Roy, Y. Xue, D. Yu and L. Dai, Angew. Chem., 2011, 50, 11756–11760.
24. S. Wang, L. Zhang, Z. Xia, A. Roy, D. W. Chang, J. B. Baek and L. Dai, Angew. Chem., 2012, 51, 4209–4212.
25. S. Kundu, T. C. Nagaiah, W. Xia, Y. Wang, S. V. Dommele, J. H. Bitter, M. Santa, G. Grundmeier, M. Bron, W. Schuhmann and M. Muhler, J. Mater. Chem. C, 2009, 113, 14302–14310.
26. S. Cui, P. Scharff, C. Siegmund, D. Schneider, K. Risch, S. Klötzer, L. Spiess, H. Romanus and J. Schawohl, Carbon, 2004, 42, 931–939.
27. A. Garsuch, R. Yang, A. Bonakdarpour and J. R. Dahn, Electrochim. Acta, 2008, 53, 2423–2429.
28. L. Yang, S. Jiang, Y. Zhao, L. Zhu, S. Chen, X. Wang, Q. Wu, J. Ma, Y. Ma and Z. Hu, Angew. Chem., 2011, 50, 7132–7135.
29. K. C. Neyerlin, W. Gu, J. Jorne and H. A. Gasteiger, J. Electrochem. Soc., 2006, 153, A1955.
30. Y. Liang, Y. Li, H. Wang, J. Zhou, J. Wang, T. Regier and H. Dai, Nat. Mater., 2011, 10, 780–786.
31. V. Palmre, E. Lust, A. Jänes, M. Koel, A.-L. Peikolainen, J. Torop, U. Johanson and A. Aabloo, J. Mater. Chem., 2011, 21, 2577.
32. C. Pham-Huu, C. Bouchy, T. Dintzer, G. Ehret, C. Estournes and M. J. Ledoux, Appl. Catal., A, 1999, 180, 385–397.
33. J. P. Paraknowitsch and A. Thomas, Energy Environ. Sci., 2013, 6, 2839.
34. A. Byeon, J. Park, S. Baik, Y. Jung and J. W. Lee, J. Mater. Chem. A, 2015, 3, 5843.
35. M. A. Mannan, M. Nagano, K. Shigezumi, T. Kida, N. Hirao and Y. Baba, Am. J. Appl. Sci., 2008, 5, 736.
36. Y. Zheng, Y. Jiao, L. Ge, M. Jaroniec and S. Z. Qiao, Angew. Chem., 2013, 52, 3110–3116.
37. C. I. Chiang, O. Meyer and R. M. C. da Silva, Nucl. Instrum. Methods Phys. Res., Sect. B, 1996, 117, 408–414.
38. A. Perrone, A. P. Caricato, A. Luches, M. Dinescu, C. Ghica, V. Sandu and A. Andrei, Appl. Surf. Sci., 1998, 133, 239–242.
39. M. Dinescu, A. Perrone, A. P. Caricato, L. Mirenghi, C. Gerardi, C. Ghica and L. Frunza, Appl. Surf. Sci., 1998, 127–129, 692–696.
40. J. Zhang, Y. Zhao, D. L. Akins and J. W. Lee, J. Phys. Chem. C, 2011, 115, 8386–8392.
41. D. He, W. Cheng, J. Qin, J. Yue, E. Xie and G. Chen, Appl. Surf. Sci., 2002, 191, 338–343.
42. S. Madakbaş, E. Çakmakçı and M. V. Kahraman, Thermochim. Acta, 2013, 552, 1–4.
43. M. Moreno, M. A. S. Ana, G. Gonzalez and E. Benavente, Electrochim. Acta, 2010, 55, 1323–1327.
44. X.-H. Guan, G.-H. Chen and C. Shang, J. Environ. Sci., 2007, 19, 438–443.
45. S. J. Parikh and J. Chorover, Langmuir, 2006, 22, 8492–8500.
46. D. Peak, G. W. Luther and D. L. Sparks, Geochim. Cosmochim. Acta, 2003, 67, 2551–2560.
47. G. Gao, W. Gao, E. Cannuccia, J. Taha-Tijerina, L. Balicas, A. Mathkar, T. N. Narayanan, Z. Liu, B. K. Gupta, J. Peng, Y. Yin, A. Rubio and P. M. Ajayan, Nano Lett., 2012, 12, 3518–3525.
48. T. Ikeda, M. Boero, S.-F. Huang, K. Terakura, M. Oshima, J.-I. Ozaki and S. Miyata, J. Mater. Chem. C, 2010, 114, 8933–8937.
49. P. Han, Y. Yue, Z. Liu, W. Xu, L. Zhang, H. Xu, S. Dong and G. Cui, Energy Environ. Sci., 2011, 4, 4710–4717.
50. A. C. Ferrari and J. Robertson, Phys. Rev. B: Condens. Matter Mater. Phys., 2000, 61, 14095–14107.
51. F. Tuinstra, J. Chem. Phys., 1970, 53, 1126.
52. Y. Wang, D. C. Alsmeyer and R. L. McCreery, Chem. Mater., 1990, 2, 557–563.
53. J. Schwan, S. Ulrich, V. Batori, H. Ehrhardt and S. R. P. Silva, J. Appl. Phys., 1996, 80, 440–447.
54. J. Filik, P. W. May, S. R. J. Pearce, R. K. Wild and K. R. Hallam, Diamond Relat. Mater., 2003, 12, 974–978.
55. X. Zhou, J. Qiao, L. Yang and J. Zhang, Adv. Energy Mater., 2014, 4, 1301523.
56. C. Xiong, Z. Wei, B. Hu, S. Chen, L. Li, L. Guo, W. Ding, X. Liu, W. Ji and X. Wang, J. Power Sources, 2012, 215, 216–220.
57. J. Zhang, A. Byeon and J. W. Lee, J. Mater. Chem. A, 2013, 1, 8665.

---

Footnote

† Electronic supplementary information (ESI) available: CV & RDE measurement data, XRD, XPS results. See DOI: 10.1039/c5ra00687b

---