- Email: [email protected]
pitch oxide composite for supercapacitors
Journal of Colloid and Interface Science 461 (2016) 96–103
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science journal homepage: www.elsevier.com/locate/jcis
Nitrogen-doped hierarchical porous carbon with high surface area derived from graphene oxide/pitch oxide composite for supercapacitors Yuan Ma, Chang Ma ⇑, Jie Sheng, Haixia Zhang, Ranran Wang, Zhenyu Xie, Jingli Shi ⇑ State Key Laboratory of Separation Membranes and Membrane Processes, School of Materials Science and Engineering, Tianjin Polytechnic University, Tianjin 300387, China
g r a p h i c a l a b s t r a c t
a r t i c l e
i n f o
Article history: Received 27 May 2015 Revised 21 August 2015 Accepted 27 August 2015 Available online 28 August 2015 Keywords: Nitrogen-doped Hierarchical pore Graphene oxide Pitch oxide Supercapacitors
a b s t r a c t A nitrogen-doped hierarchical porous carbon has been prepared through one-step KOH activation of pitch oxide/graphene oxide composite. At a low weight ratio of KOH/composite (1:1), the as-prepared carbon possesses high specific surface area, rich nitrogen and oxygen, appropriate mesopore/micropore ratio and considerable small-sized mesopores. The addition of graphene oxide plays a key role in forming 4 nm mesopores. The sample PO–GO-16 presents the characteristics of large surface area (2196 m2 g 1), high mesoporosity (47.6%), as well as rich nitrogen (1.52 at.%) and oxygen (6.9 at.%). As a result, PO–GO-16 electrode shows an outstanding capacitive behavior: high capacitance (296 F g 1) and ultrahigh-rate performance (192 F g 1 at 10 A g 1) in 6 M KOH aqueous electrolyte. The balanced structure characteristic, low-cost and high performance, make the porous carbon a promising electrode material for supercapacitors. Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction Electrochemical capacitors (EDLC, known as ultracapacitors or supercapacitors) hold promise for a wide range of applications, including portable electric equipments, uninterruptable power sources, medical devices, load leveling and hybrid electric vehicles, ⇑ Corresponding authors. E-mail addresses: [email protected] (C. Ma), [email protected] (J. Shi). http://dx.doi.org/10.1016/j.jcis.2015.08.065 0021-9797/Ó 2015 Elsevier Inc. All rights reserved.
and so forth [1]. In recent years, great efforts have been made in developing better supercapacitor electrode materials. Porous carbon materials are widely used and studied as supercapacitor electrodes because of their relatively low cost, high specific surface area, good electric conductivity, rich source and excellent chemical stability [2]. In EDLCs, energies are stored by the electrostatic charge uptake at the electrolyte/electrode interfacial regions of porous carbons. The electrochemical performance of supercapacitors is greatly
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
dependent on high specific surface area (SSA) and a rational pore structure. Generally, carbon materials with higher SSA have a better capability for charge accumulation at the electrode/electrolyte interface. Recently, numerous researches have held that specific capacitance is not always linearly proportional to the SSA, especially for carbons with ultra-high SSA. This is because not all micropores in electrodes are necessarily accessible to the electrolyte ions, due to that closed pores or narrow bottle-necks may dramatically prevent or slow ion transport and impair the capacitive performance of supercapacitors. Hence, the micropore amount should be controlled at a particular level. The pores with bigger size have been reckoned more beneficial for convenient ion transportation and higher accessible surface. Hierarchical porous carbons, containing considerable mesopores or macropores, have inspired high hopes and been proved possess distinctive potential for high-performanced supercapacitors [3,4]. However, too many bigger pores or pores with too big diameter always lead to moderate even low surface area, which are obviously not conducive to high-performanced supercapacitors [5]. So, mesopores with relative small size should be introduced with an appropriate proportion. So far, great efforts have been made to gain welldesigned hierarchical porous carbon materials for supercapacitors. Template method is most commonly used to create mesopores. A variety of templates, such as SiO2 colloid, magnesium, nickel oxide, have been employed [6–8]. However, it always means that timeconsuming washing process, introduction of impurity are unavoidable. These deficiencies hinder their practical applications. Hence, a simplified and convenient strategy to hierarchical porous carbons is still needed. Besides pore structure, the surface functionalization of electrode materials is another crucial factor for the electrochemical performance [9]. In general, carbon materials are of poor surface wettability for electrolyte solution owing to the high temperature carbonization, which impedes the penetration of the electrolyte ions into the pores of carbons. Introduction of surface functionalization, especially nitrogen functionalizations, has been proven to be a promising approach to improve their surface wettability, electric conductivity and capacitance properties [10,11]. However, development of nitrogen functionalized electrode materials by a simple method is still required for further advancement in supercapacitors. Herein, we demonstrate a facile strategy to nitrogen-containing hierarchical porous carbons through one-step KOH activation of pitch oxide/graphene oxide composite. With less KOH consumption and simpler process, the as-prepared carbon possessed high specific surface area (2196 m2 g 1), rich nitrogen (1.52 at.%) and oxygen (6.9 at.%), appropriate mesopore/micropore ratio (mesoporosity of 47.6%) and considerable small-sized mesopores (4 nm). As a result, an outstanding capacitive performance, including high capacitance (296 F g 1) and ultrahigh-rate performance (192 F g 1 at 10 A g 1) in 6 M KOH aqueous electrolyte was achieved.
2. Experimental
97
dissolved in 500 ml deionized water by stirring 2 h at 75 °C water bath. The pitch oxide aqueous solution was obtained. Subsequently, solid content of the pitch oxide aqueous solution was measured. Graphene oxide (GO) was prepared by oxidizing the natural graphite powder and subsequent exfoliation by ultrasonication by a modified Hummers’ method [12,13]. Firstly, 1 g graphite powder (Qingdao Black Dragon Graphite Co. Ltd.) and 0.5 g sodium nitrate (Tianjin Kemiou Chemical Reagent Co., Ltd.) were mixed with 30 ml sulfuric acid (98%, Shanghai Macklin Biochemical Co., Ltd) in an ice bath under vigorous stirring. Then, 3 g potassium permanganate (Tianjin Kemiou Chemical Reagent Co., Ltd.) was slowly added into the system in 30 min while the temperature was kept from exceeding 20 °C. After 2 h, the mixture was moved to 35 °C oil bath and maintained for 12 h. Subsequently, 230 ml deionized water was slowly added, and then the mixture was heated to 98 °C for 60 min. 150 ml deionized water and 20 ml hydrogen peroxide (30 wt.%, Tianjin Kemiou Chemical Reagent Co., Ltd.) were added into the mixture. After centrifugation and washing to remove residual salts, the graphite oxide aqueous solution was prepared. Then, the graphite oxide aqueous solution was ultrasonicated for 2 h and centrifuged at 8000 r min 1. The supernatant colloidal substance was collected. Subsequently, solid content of GO aqueous solution was measured. PO/GO compound was prepared by mixing pitch oxide aqueous solution and graphene oxide aqueous solution with different ratios under ultrasonic and stirring, followed by desiccation at 100 °C. Then, the PO/GO compound was mixed with KOH at PO/GO: KOH = 1:1. Carbonization was conducted at 800 °C for 1 h with a heating-rate of 3 °C min 1 in high purity nitrogen. The product was washed using dilute HCl and deionized water repeatedly. The obtained solid was dried at 80 °C to obtain porous carbons PO–GO-X (X means the ratio of PO/GO = 4, 8, 16, 24). The preparation process was displayed in Scheme 1. For comparison, PO and GO were activated by the same process, from which the resultant carbons were named PO and GO, respectively. 2.2. Materials characterizations The morphology and structure of samples were examined by field-emission scanning electron microscopy (FESEM, Hitachi S-4800). The crystallographic structures of the materials were investigated by a Rigaku D/MAX-ga diffractometer with filtered Cu Ka radiation (k = 0.15406 nm). X-ray photoelectron spectroscopy (XPS) was measured by an X-ray photoelectron spectrometer (ESCALAB 250) to analyze the surface characteristics of the samples. Nitrogen adsorption and desorption isotherms was conducted at 77 K to investigate the porous texture of the samples on a Quantachrome ASMS analyzer. All samples were degassed at 200 °C for 12 h before the measurement. The Brunauer–Emmett– Teller (BET) method was utilized to calculate the specific surface area. Micropore size distribution was estimated by density functional theory (DFT), while mesopore size distribution by Barrett–J oyner–Halenda (BJH) model from adsorption branch isotherms. Pore volume was calculated from the adsorbed amount at a relative pressure P/P0 = 0.994.
2.1. Material preparation 2.3. Electrochemical measurement Pitch oxide (PO) was prepared by treatment of pitch (Isotropic pitch, Shanghai Dongdao Carbon Chemical Industry Co., Ltd.) using mixed acid. Firstly, 20 ml nitric acid and 80 ml sulfuric acid were blended and stirred in 40 °C water bath for 10 min. The pitch was slowly added into the system in 30 min. After 2 min, 1000 ml deionized water was added into the mixture. Solid precipitate of the mixture was collected by repeated vacuum filtration. Subsequently, the collected solid precipitate and 1 g NaOH were
The working electrode was fabricated by mixing the active material, carbon black and polytetrafluoroethylene (PTFE) with the weight ratio of 80:10:10 and then dried in a vacuum oven at 100 °C for 24 h. Finally, the electrode was pressed onto the nickel foam current collectors at 8 MPa for test. The weight of the electrode is 1.5–2.0 mg. The electrochemical performance of the electrodes was measured in three-electrode cells in 6 M KOH
98
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
Scheme 1. Illustration of the preparation of PO–GO-X.
aqueous electrolyte with platinum sheet and Hg/HgO electrodes as counter and reference electrodes, respectively. All the electrochemical investigations were carried out on a CHI 660C electrochemical workstation using cyclic voltammetry (CV), galvanostatic charge/discharge and electrochemical impedance spectroscopy (EIS) techniques. CV tests were investigated between 0.9 and 0 V (vs. Hg/HgO). Galvanostatic charge/discharge was performed in the same potential range at the current densities ranging from 0.1 to 10 A g 1. EIS test was performed with alternate current amplitude of 0.005 V in the frequency range of 0.01 to 100 kV.
3. Results and discussion 3.1. Morphology and textural properties Fig. 1 presents the surface morphology of the obtained porous carbons derived from PO/GO composite precursor with different ratios. From the four SEM images, one can see clearly their remarkable differences. It can be confirmed that the content of the GO is responsible for the changes. When the GO/PO ratio is set to 1:24, a very low proportion of GO in the composite precursor, the carbon displays no clear distinction between other activated carbons, for example, glucosamine based activated carbons [14]. With the ratio rising to 1:16, some bigger holes, being held up by nanoparticles
with diameter of about 100 nm, can be observed. It can be referred that the graphene plays a role of framework support, which can make sure a homogeneous activation. Increasing the percentage of GO further, as shown in Fig. 1(b), the structural guiding effect of graphene becomes more apparent. The clearly-seen carbon sheets are considered as a composite consisting of graphene and the activated carbon particle with a smaller size. With a highest content of GO, the PO–GO-4 shows many obvious folds. The folds should be from the aggregation of graphene. Fig. 2 shows the N2 sorption isotherms of all PO–GO-Xs, GO and PO. The isotherms for all PO–GO-Xs possess distinct adsorption and desorption branches, as shown in Fig. 2(a), which belong to type-IV profile in virtue of the International Union of Pure and Applied Chemistry (IUPAC) classification. In details, the adsorbed quantity increases steeply to a high value at relative low pressure (close to 0), demonstrating abundant micropores in carbons; the hysteresis loop at medium pressure (0.45–1.0) means the existence of considerable mesopores; at high pressure close to 1.0, the adsorbed curves seem to be relatively smooth, suggesting the negligible quantity of macropores. GO presents a negligible quantity at low relative pressure and a apparent hysteresis loop at medium pressure, meaning typical mesoporous characteristic. PO displays a high low-pressure adsorption quantity and an insignificant hysteresis loop at medium pressure, suggesting a typical microporous characteristic. Their difference demonstrates that graphene has higher skeleton rigidity and is possible to form mesopore wall.
Fig. 1. SEM images of the obtained porous carbons. (a) PO–GO-4, (b) PO–GO-8, (c) PO–GO-16, (d) PO–GO-24.
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
99
Fig. 2. (a) N2 sorption isotherms, (b) pore size distribution of the obtained porous carbons, in which the micropore distribution is calculated from DFT method, while mesopore region from BJH method.
However, because of existence of aggregation, GO has a small adsorption quantity. With a small amount of GO being added, PO–GO-16 and PO–GO-24 show a slight enhancement in low-pressure adsorbed quantity, meaning increased micropores. It may be because that some closed micropores are opened due to existing of some graphene-directing activation channels. However, the adsorbed quantity at low pressure decreases remarkably when more GO is added. That is because that PO is easier to be activated and to form micropores than GO due to numerous structural defect in aromatic-hydrocarbons of pitch in compared with relatively complete hexagonally bonded carbon of graphene [15,16]. With a relative small KOH content (1:1), a moderate activation degree is predictable, which also suggests that mesopores are usually not formed [17]. For PO and PO–GO-24, without GO or with low enough GO content, the hysteresis loop is much smaller, which implies that both PO and PO–GO-24 present principally microporous characteristic and too little GO does not bring about a noticeable change in mesopores. In contrast, those samples with relatively high GO content show more evident hysteresis loops. Therefore, there are reasons for believing that graphene makes a positive role in producing mesopores. However, too much GO does not bring about more mesopores. The sample PO–GO-4 exhibits a less adsorbed quantity at high pressure. It can be referred that the aggregation of graphene impedes the formation of more mesopore walls. Fig. 2(b) presents the pore size distribution for all samples. Based on the sorption isotherms analysis, a significant change with different additions of GO is plainly seen at mesopore region. It’s
well known that BJH method is most suitable to esteem mesopores. So, the mesopore distribution was calculated according to BJH method. For micropores, DFT model is distinctly better than BJH method. Therefore, DFT method was selected to estimated the micropore distribution. It can be seen that PO present both narrow pore distribution at about 0.5 nm and a wide pore distribution at 1–4 nm. The wide micropore distribution is much common for chemical-activated carbons. GO has no micropore distribution, suggesting it is more difficult to be activated than PO. A weak mesopore peak at about 4 nm can be observed. It can be deduced that mesopores are principally due to stacking of graphene. By comparison with PO, all PO–GO-X samples show relatively distinct micropore and mesopore distributions in spite of weak mesopore peak for PO–GO-24 and PO–GO-4. PO–GO-4, PO–GO-8 and PO–GO-16 all show relatively wider micropores than PO–GO-24, which may be attributed to framework effect of homogeneously dispersed graphene in carbon matrix. All PO–GO-Xs display a centered mesopore distribution at 4 nm, relatively small mesopores. In conformity with the sorption isotherms, PO–GO-8 and PO–GO-16 possess much more 4 nm mesopores and PO–GO-24 and PO–GO-4 show a faint 4 nm mesopore peak. Such specific mesopores may be determined by molecular structure and size of PO and GO. Generally, the small mesopores still greatly contribute to total specific surface area. Table 1 lists the pore structure parameters of all samples. Although very limited KOH is used, KOH/composite ratio is 1:1, PO–GO-16 still shows a fairly large specific surface area of 2196 m2 g 1 and pore volume of 1.55 cm3 g 1. PO–GO-24 presents
100
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
a microporosity of 91.5%, even higher than PO, and lower pore volume of 0.98 cm3 g 1 than PO. It is in accordance with the pore size distribution result that PO has a wide pore distribution at 1–4 nm. With the GO/PO ratio increased from 1/24 to 1/16, not only mesoporosity and total volume are notably raised, but also the BET surface area is increased from 2019 to 2196 m2 g 1. It should be due to slightly-increased micropores and a relatively small mesopores. Both PO–GO-8 and PO–GO-16 have a high mesoporosity, close to 50%. The appropriate percentage of mesopores and relative small pore size give a balance between high surface area and enlarged pore diameter. The PO–GO-Xs are further analyzed by XRD patterns. As shown in Fig. 3, one broad and low intensity diffraction peak centered at about 20.8° (2h), which can be approximately indexed as (0 0 2) plane of standard graphite, is observed in the XRD patterns, revealing the characteristics of amorphous carbon usually given by KOH activation procedure [18]. The interlayer spacing, d0 0 2 of the porous carbons was calculated by Bragg’s equation to be 0.427 nm, much higher than 0.365 nm of graphite. Maybe the high content of acidic groups, including carboxyl, hydroxyl or epoxy, on PO molecules and carbon plane of GO should be the reason for the incompletely developed graphite microcrystallite [19]. Another weak peak at 43.3° (2h), corresponding to (1 0 0) diffractions of graphitic layers, was hardly distinguishable, further verifying the incompletely developed graphite microcrystallite. Fig. 4 shows the XPS spectrum of PO–GO-Xs. From the wide XPS spectrum, apart from the strongest C1 peak (284 eV) and apparent O1 peak (533 eV), a weak N1s peak at 402 eV also can be observed clearly, demonstrating that the obtained porous carbons possess a significant amount of O and a small amount of N [20]. Among them, the O could stem from the rich hydroxyl, hydroxyl and sulfo groups of both PO and GO, while N only has one source, namely nitro group of PO treated by mixed acid. Based on the systematic survey of nitrogen- or oxygencontaining carbons, the peaks of the O1s and N1s spectras of various samples are assigned (as shown in Fig. 5). For oxygen, there are three main peaks centered at 530.3 ± 0.2, 531.7 ± 0.2, 533.1 ± 0.2 eV, which represent highly conjugated forms of carbonyl oxygen such as quinine or pyridone groups (OI), AC@O groups (OII), CAOH groups and/or CAOAC groups (OIII), respectively [21–24]. In the same way, the chemical state of nitrogen atoms could be assigned to two types: pyrrolic/pyridone nitrogen (N5, 400.0 ± 0.2 eV) and pyridine-N-oxide (N-X, 402.2 ± 0.2 eV), respectively [23,25,26]. Table 2 lists the relative surface concentrations of oxygen and nitrogen species obtained by fitting the N1s and O1s spectra. With different PO/GO ratios, the obtained porous carbons present different content of nitrogen and similar content of oxygen. The more the GO, the higher the residual nitrogen. It seems farfetched that the carbon from precursor with more PO has less nitrogen. Actually, the difference in content of PO is not very remarkable, just from 80 wt.% of PO–GO-4 to 96 wt.% of PO–GO-24. The evolution and migration of heteroatom in carbon is complex, however, it is
Fig. 3. XRD patterns of PO–GO-Xs.
Fig. 4. The XPS survey spectra.
believed that number of graphene, acting as obstacle for diffusion of heteroatoms, and the porosity, serving as pathway for escaping of N or O-containing gas, exert a crucial influence on resulting N or O content. Because of high graphene content and low porosity, PO– GO-4 shows a relatively high N content. Recent surveys have revealed that N5, OI and OII have electrochemically activity, meaning ability of contributing the pseudocapacitance [27,28]. All PO– GO-Xs own high proportion of active N and O, which would produce considerable pseudocapacitance. In addition to the pseudocapacitance contribution, both the oxygen and nitrogen groups, especially their basic ones, significantly improve the wettability of the surface and enhance the access of the electrolyte ions. 3.2. Electrochemical performance The electrochemical performances of PO–GO-Xs were investigated by cyclic voltammograms (CV) tests and galvanostatic charge/discharge measurements in 6 M KOH solution. Fig. 6 illustrates a set of CV curves of PO–GO-16 at various scan rates from
Table 1 Summary of specific surface area (SBET), total pore volume, average pore diameter, and pore-size distribution data for all samples. Vmicro (%) and Vmeso (%) are volume percentages of micropores and mesopores, respectively. Samples
GO PO PO–GO-4 PO–GO-8 PO–GO-16 PO–GO-24
SBET (m2 g
158 1824 1172 1493 2196 2019
1
)
Total pore volume (cm3 g
0.18 1.18 0.73 1.10 1.55 0.98
1
)
Average pore diameter (nm)
2.31 1.95 2.50 2.94 2.82 1.93
Pore size distribution Vmicro (%)
Vmeso (%)
25.3 85.4 81.2 53.3 52.4 91.5
74.7 14.6 18.8 46.7 47.6 8.5
101
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
Fig. 5. N1s spectra and O1s spectra.
0.005 to 0.1 V s 1 in the potential range between 0.9 and 0 V (vs. Hg/HgO). The electrode presents a typical capacitive behavior with quasi-rectangular shaped voltammetry characteristics as well as broad humps in the CVs, indicating a combination of electrical double-layer capacitance and pseudocapacitance related to redox reactions of the heteroatoms in the electrode materials [29]. With the scan rate increased from 0.005 to 0.1 V s 1, the specific capacitance of vertical ordinate, calculated from current density, displays very insignificant decay, revealing an outstanding rate performance. Even at high rate of 0.1 V s 1, the CV curve still retains near-rectangular shape and no dramatic distortion of the curve could be observed, indicating a highly reversible system and a swift current response. Fig. 6(b) shows CV curves of PO–GO-Xs at scan rates of 0.01 V s 1. All electrodes display a
similar CV curves, including a quasi-rectangular shape, broad humps and rapid current response. Typical galvanostatic charge/discharge curves of PO–GO-16 at different current densities from 0.1 to 10 A g 1 are displayed in Fig. 6(c). It can be seen that all the galvanostatic charge/discharge curves exhibit the shape of isosceles triangular and the charge/discharge curves are highly symmetric. These characteristics also demonstrate that the electrode possesses an ideal capacitive performance and excellent electrochemical reversibility. Fig. 6(d) exhibits galvanostatic charge/discharge curves of all PO–GO-Xs at current density of 1 A g 1. One can clearly seen that PO–GO-16 has the best charge/discharge performance. It can be also found that the ohmic drop of all electrodes is very small, proving a high electrical conductivity of electrodes. The specific capacitances of PO–GO-Xs at different current densities are calculated according to charge/discharge curves, as shown in Fig. 6(e). It can be found that the bigger the specific surface area, the higher the specific capacitance. PO–GO-16 possesses a specific capacitance of 296 F g 1 at 0.1 A g 1 in the 6 M KOH aqueous solution, much higher than that of PO–GO-4 (175 F g 1). With much the same specific area, PO–GO-24 has a relatively low specific capacitance of 256 F g 1. It may be due to the higher content of mesopores, which can provide more accessible surface area. It’s important to note that PO–GO-16, which was prepared with a lower KOH/precursor ratio (1:1), not only presents considerable specific surface area, but also a high mesoporosity, combination of which leads to excellent capacitive performance. Table 3 lists SBET, mesoporosity and capacitance of the reported pitchbased porous carbons, mainly activated ones. Comparison reveals that the PO/GO-based activated carbon possesses a competitive capacitive performance over other pitch-based carbons even if limited KOH is used. With the current density continuously increasing, all PO–GO-Xs present a capacitance drop, but still hold high capacitance retentions. Owing to high surface area and abundant micropores, both the PO–GO-16 and PO–GO-24 show more significant capacitance loss, but still keep high capacitance retentions of 66% and 71%. Fig. 6(f) shows the electrochemical impedance spectroscopy measured in the range from 0.01 to 100 kHz in the threeelectrode cell. The semicircle of the Nyquist plot in the highfrequency range is related to the porous structure of the carbon electrode and corresponds to the charge transfer resistance. The smaller diameter of semicircle implies the lower charge transfer resistance. PO–GO-24 has the biggest diameter of semicircle, which is bigger than that of PO–GO-16. Considering much the same specific surface area, mesopores are thought to provide a more convenient route for ion transferring. PO–GO-4 and PO–GO-8 display the smaller semicircle. It should be mainly attributed to graphene, which brings about an enhanced electron conductivity. The line of the Nyquist plot in the low-frequency range is related to the capacitive behavior of the electrode. All electrodes display nearly vertical lines, which reflect an idea supercapacitive behavior.
Table 2 XPS analysis and XPS peak positions and relative surface concentrations of oxygen and nitrogen species obtained by fitting the N 1s and O 1s spectra.
a b
Sample
N (at.%)
PO–GO-4 PO–GO-8 PO–GO-16 PO–GO-24
2.77 1.68 1.52 0.92
N1s (%)a N-5 (400.0 eV)
N-X (402.2 eV)
94.2 80.5 86.7 76.2
5.8 19.5 13.3 23.8
Relative contents of functional groups in N1s. Relative contents of functional groups in O1s.
O (at.%)
O1s (%)b O-I (530.3 eV)
O-II (531.7 eV)
O-III (533.1 eV)
6.6 7.1 7.2 5.5
11.7 8.2 6.9 8.7
62.1 48.0 36.5 46.2
26.2 43.8 56.6 45.1
102
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
Fig. 6. (a) CV curves of PO–GO-16 at different scan rates from 0.005 to 0.1 V s 1. (b) CV curves of PO–GO-Xs at scan rates of 0.01 V s 1. (c) Charge/discharge curves of PO–GO-16 at different current densities from 0.1 to 10 A g 1. (d) Charge/discharge curves of PO–GO-Xs at current densities from 1 A g 1. (e) Variation of the specific capacitance of PO–GO-Xs as a function of current density. (f) Nyquist plots of PO–GO-Xs measured in the three-electrode cell.
Table 3 Comparison of specific capacitances and specific surface area with other pitch-based carbons. Precursor
Methods
SBET (m2 g
Heat-treated pitch Sulfonated pitch Pitch-based carbon fibers Pitch Coal tar pitch/melamine Coal tar pitch based ACMs PO/GO
KOH(1:4) KOH(1:3) Ammonia KOH(1:3) KOH(1:3) P123 KOH(1:1)
2003 3548 1454 983 2573 860 2196
3E: three-electrode. 2E: two-electrode.
1
)
Vmeso/Vtotal (%)
C (F g
31.7 59.3 – 11.4 56.0 – 47.6
170 277 146 276 228 172 296
1
)
(0.5 A g 1) (2 mV s 1) (1 A g 1) (5 mV s 1) (1 A g 1) (10 mV s 1) (0.1 A g 1)
Cell
Electrolytes
Ref.
3E 3E 2E 3E 3E 3E 3E
6M 6M 6M 1M 6M 1M 6M
[30] [31] [32] [33] [34] [35] This work
KOH KOH KOH H2SO4 KOH H2SO4 KOH
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
4. Conclusion In summary, we demonstrate the facile synthesis of a novel hierarchical porous carbon with high specific surface area, rich nitrogen and oxygen, appropriate mesopore/micropore ratio and considerable small-sized mesopores through one-step activation of pitch oxide/graphene oxide precursor. The addition of graphene plays a key role in forming 4 nm mesopores. As-prepared sample PO–GO16 presents the characteristics of large surface area (2196 m2 g 1), high mesoporosity (47.6%), as well as rich nitrogen (1.52 at.%) and oxygen (6.9 at.%). As a result, PO–GO-16 electrode shows outstanding capacitive behavior: high capacitance (296 F g 1) and ultrahighrate performance (192 F g 1 at 10 A g 1) in 6 M KOH aqueous electrolyte. Therefore, taking the low-cost, less KOH consumption and simplified process into account, the well-designed and high-performanced PO–GO-Xs provide new opportunities for widespread application of supercapacitors. References [1] [2] [3] [4] [5] [6]
P. Simon, Y. Gogotsi, B. Dunn, Science 343 (2014) 1210. L. Hao, X. Li, L. Zhi, Adv. Mater. 25 (2013) 3899. Y. Li, Z. Li, P.K. Shen, Adv. Mater. 25 (2013) 2474. Z. Chen, D. Weng, H. Sohn, M. Cai, Y. Lu, RSC Adv. 2 (2012) 1755. F. Béguin, V. Presser, A. Balducci, E. Frackowiak, Adv. Mater. 14 (2014) 2219. Q. Wang, J. Yan, Y. Wang, T. Wei, M. Zhang, X. Jing, Z. Fan, Carbon 67 (2014) 119. [7] D.W. Wang, F. Li, M. Liu, G.Q. Lu, H.M. Cheng, Angew. Chem. Int. Ed. 120 (2008) 379. [8] C. Ma, J. Shi, Y. Song, D. Zhang, X. Zhai, M. Zhong, Q. Guo, L. Liu, Int. J. Electrochem. Sci. 7 (2012) 7587. [9] X.Y. Chen, C. Chen, Z.J. Zhang, D.H. Xie, X. Deng, Ind. Eng. Chem. Res. 52 (2013) 10181.
103
[10] X. Zhai, Y. Song, J. Liu, P. Li, M. Zhong, C. Ma, W. Huiqi, J. Electrochem. Soc. 159 (2012) E177. [11] S. Zhong, C. Zhan, D. Cao, Carbon 85 (2015) 51. [12] Y. Bai, R.B. Rakhi, W. Chen, H.N. Alshareef, J. Power Sources 233 (2013) 313. [13] W.S. Hummers, R.E. Offeman, JACS 80 (1958) 1339. [14] L. Zhao, L.Z. Fan, M.Q. Zhou, H. Guan, S.Y. Qiao, M. Antonietti, M.-M. Titirici, Adv. Mater. 22 (2010) 5202. [15] S. Lee, Y. Eom, B.-J. Kim, I. Mochida, S.-H. Yoon, B.C. Kim, Carbon 81 (2015) 694. [16] H. Bai, C. Li, G. Shi, Adv. Mater. 23 (2011) 1089. [17] L. Zhu, Q. Gao, Y. Tan, W. Tian, J. Xu, K. Yang, C. Yang, Micropor. Mesopor. Mater. 210 (2015) 1. [18] J. Jiang, H. Chen, Z. Wang, L. Bao, Y. Qiang, S. Guan, J. Chen, J. Colloid Interface Sci. 452 (2015) 54. [19] W. Gao, L.B. Alemany, L. Ci, P.M. Ajayan, Nat. Chem. 1 (2009) 403. [20] K.-S. Kim, S.-J. Park, Electrochim. Acta 56 (2011) 10130. [21] H. Li, Z. Song, X. Zhang, Y. Huang, S. Li, Y. Mao, H.J. Ploehn, Y. Bao, M. Yu, Science 342 (2013) 95. [22] Y. Zhao, C. Hu, Y. Hu, H. Cheng, G. Shi, L. Qu, Angew. Chem. Int. Ed. 45 (2012) 11371. [23] M. Seredych, D. Hulicova-Jurcakova, G.Q. Lu, T.J. Bandosz, Carbon 46 (2008) 1475. [24] J.H. Zhou, Z.J. Sui, J. Zhu, P. Li, D. Chen, Y.C. Dai, W.K. Yuan, Carbon 45 (2007) 785. [25] Y.J. Kim, Y. Abe, T. Yanagiura, K.C. Park, M. Shimizu, T. Iwazaki, S. Nakagawa, M. Endo, M.S. Dresselhaus, Carbon 45 (2007) 2116. [26] E. Raymundo-Piñero, D. Cazorla-Amorós, A. Linares-Solano, J. Find, U. Wild, R. Schlögl, Carbon 40 (2002) 597. [27] X. Yang, D. Wu, X. Chen, R. Fu, J. Phys. Chem. C 114 (2010) 8581. [28] C. Ma, Y. Song, J. Shi, D. Zhang, Q. Guo, L. Liu, J. Colloid Interface Sci. 395 (2013) 217. [29] X. Fan, C. Yu, J. Yang, Z. Ling, J. Qiu, Carbon 70 (2014) 130. [30] G.-Y. Heo, S.-J. Park, Powder Technol. 239 (2013) 94. [31] Y. Guo, Z.-Q. Shi, M.-M. Chen, C.-Y. Wang, J. Power Sources 252 (2014) 235. [32] K. Torchała, K. Kierzek, G. Gryglewicz, J. Machnikowski, Electrochim. Acta 167 (2015) 348. [33] M.-S. Park, S. Cho, E. Jeong, Y.-S. Lee, J. Ind. Eng. Chem. 23 (2015) 27. [34] C. Zhong, S. Gong, L.E. Jin, P. Li, Q. Cao, Mater. Lett. 156 (2015) 1. [35] Y. Cheng, T. Li, C. Fang, M. Zhang, X. Liu, R. Yu, J. Hu, Appl. Surf. Sci. 282 (2013) 862.
Contents lists available at ScienceDirect
Journal of Colloid and Interface Science journal homepage: www.elsevier.com/locate/jcis
Nitrogen-doped hierarchical porous carbon with high surface area derived from graphene oxide/pitch oxide composite for supercapacitors Yuan Ma, Chang Ma ⇑, Jie Sheng, Haixia Zhang, Ranran Wang, Zhenyu Xie, Jingli Shi ⇑ State Key Laboratory of Separation Membranes and Membrane Processes, School of Materials Science and Engineering, Tianjin Polytechnic University, Tianjin 300387, China
g r a p h i c a l a b s t r a c t
a r t i c l e
i n f o
Article history: Received 27 May 2015 Revised 21 August 2015 Accepted 27 August 2015 Available online 28 August 2015 Keywords: Nitrogen-doped Hierarchical pore Graphene oxide Pitch oxide Supercapacitors
a b s t r a c t A nitrogen-doped hierarchical porous carbon has been prepared through one-step KOH activation of pitch oxide/graphene oxide composite. At a low weight ratio of KOH/composite (1:1), the as-prepared carbon possesses high specific surface area, rich nitrogen and oxygen, appropriate mesopore/micropore ratio and considerable small-sized mesopores. The addition of graphene oxide plays a key role in forming 4 nm mesopores. The sample PO–GO-16 presents the characteristics of large surface area (2196 m2 g 1), high mesoporosity (47.6%), as well as rich nitrogen (1.52 at.%) and oxygen (6.9 at.%). As a result, PO–GO-16 electrode shows an outstanding capacitive behavior: high capacitance (296 F g 1) and ultrahigh-rate performance (192 F g 1 at 10 A g 1) in 6 M KOH aqueous electrolyte. The balanced structure characteristic, low-cost and high performance, make the porous carbon a promising electrode material for supercapacitors. Ó 2015 Elsevier Inc. All rights reserved.
1. Introduction Electrochemical capacitors (EDLC, known as ultracapacitors or supercapacitors) hold promise for a wide range of applications, including portable electric equipments, uninterruptable power sources, medical devices, load leveling and hybrid electric vehicles, ⇑ Corresponding authors. E-mail addresses: [email protected] (C. Ma), [email protected] (J. Shi). http://dx.doi.org/10.1016/j.jcis.2015.08.065 0021-9797/Ó 2015 Elsevier Inc. All rights reserved.
and so forth [1]. In recent years, great efforts have been made in developing better supercapacitor electrode materials. Porous carbon materials are widely used and studied as supercapacitor electrodes because of their relatively low cost, high specific surface area, good electric conductivity, rich source and excellent chemical stability [2]. In EDLCs, energies are stored by the electrostatic charge uptake at the electrolyte/electrode interfacial regions of porous carbons. The electrochemical performance of supercapacitors is greatly
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
dependent on high specific surface area (SSA) and a rational pore structure. Generally, carbon materials with higher SSA have a better capability for charge accumulation at the electrode/electrolyte interface. Recently, numerous researches have held that specific capacitance is not always linearly proportional to the SSA, especially for carbons with ultra-high SSA. This is because not all micropores in electrodes are necessarily accessible to the electrolyte ions, due to that closed pores or narrow bottle-necks may dramatically prevent or slow ion transport and impair the capacitive performance of supercapacitors. Hence, the micropore amount should be controlled at a particular level. The pores with bigger size have been reckoned more beneficial for convenient ion transportation and higher accessible surface. Hierarchical porous carbons, containing considerable mesopores or macropores, have inspired high hopes and been proved possess distinctive potential for high-performanced supercapacitors [3,4]. However, too many bigger pores or pores with too big diameter always lead to moderate even low surface area, which are obviously not conducive to high-performanced supercapacitors [5]. So, mesopores with relative small size should be introduced with an appropriate proportion. So far, great efforts have been made to gain welldesigned hierarchical porous carbon materials for supercapacitors. Template method is most commonly used to create mesopores. A variety of templates, such as SiO2 colloid, magnesium, nickel oxide, have been employed [6–8]. However, it always means that timeconsuming washing process, introduction of impurity are unavoidable. These deficiencies hinder their practical applications. Hence, a simplified and convenient strategy to hierarchical porous carbons is still needed. Besides pore structure, the surface functionalization of electrode materials is another crucial factor for the electrochemical performance [9]. In general, carbon materials are of poor surface wettability for electrolyte solution owing to the high temperature carbonization, which impedes the penetration of the electrolyte ions into the pores of carbons. Introduction of surface functionalization, especially nitrogen functionalizations, has been proven to be a promising approach to improve their surface wettability, electric conductivity and capacitance properties [10,11]. However, development of nitrogen functionalized electrode materials by a simple method is still required for further advancement in supercapacitors. Herein, we demonstrate a facile strategy to nitrogen-containing hierarchical porous carbons through one-step KOH activation of pitch oxide/graphene oxide composite. With less KOH consumption and simpler process, the as-prepared carbon possessed high specific surface area (2196 m2 g 1), rich nitrogen (1.52 at.%) and oxygen (6.9 at.%), appropriate mesopore/micropore ratio (mesoporosity of 47.6%) and considerable small-sized mesopores (4 nm). As a result, an outstanding capacitive performance, including high capacitance (296 F g 1) and ultrahigh-rate performance (192 F g 1 at 10 A g 1) in 6 M KOH aqueous electrolyte was achieved.
2. Experimental
97
dissolved in 500 ml deionized water by stirring 2 h at 75 °C water bath. The pitch oxide aqueous solution was obtained. Subsequently, solid content of the pitch oxide aqueous solution was measured. Graphene oxide (GO) was prepared by oxidizing the natural graphite powder and subsequent exfoliation by ultrasonication by a modified Hummers’ method [12,13]. Firstly, 1 g graphite powder (Qingdao Black Dragon Graphite Co. Ltd.) and 0.5 g sodium nitrate (Tianjin Kemiou Chemical Reagent Co., Ltd.) were mixed with 30 ml sulfuric acid (98%, Shanghai Macklin Biochemical Co., Ltd) in an ice bath under vigorous stirring. Then, 3 g potassium permanganate (Tianjin Kemiou Chemical Reagent Co., Ltd.) was slowly added into the system in 30 min while the temperature was kept from exceeding 20 °C. After 2 h, the mixture was moved to 35 °C oil bath and maintained for 12 h. Subsequently, 230 ml deionized water was slowly added, and then the mixture was heated to 98 °C for 60 min. 150 ml deionized water and 20 ml hydrogen peroxide (30 wt.%, Tianjin Kemiou Chemical Reagent Co., Ltd.) were added into the mixture. After centrifugation and washing to remove residual salts, the graphite oxide aqueous solution was prepared. Then, the graphite oxide aqueous solution was ultrasonicated for 2 h and centrifuged at 8000 r min 1. The supernatant colloidal substance was collected. Subsequently, solid content of GO aqueous solution was measured. PO/GO compound was prepared by mixing pitch oxide aqueous solution and graphene oxide aqueous solution with different ratios under ultrasonic and stirring, followed by desiccation at 100 °C. Then, the PO/GO compound was mixed with KOH at PO/GO: KOH = 1:1. Carbonization was conducted at 800 °C for 1 h with a heating-rate of 3 °C min 1 in high purity nitrogen. The product was washed using dilute HCl and deionized water repeatedly. The obtained solid was dried at 80 °C to obtain porous carbons PO–GO-X (X means the ratio of PO/GO = 4, 8, 16, 24). The preparation process was displayed in Scheme 1. For comparison, PO and GO were activated by the same process, from which the resultant carbons were named PO and GO, respectively. 2.2. Materials characterizations The morphology and structure of samples were examined by field-emission scanning electron microscopy (FESEM, Hitachi S-4800). The crystallographic structures of the materials were investigated by a Rigaku D/MAX-ga diffractometer with filtered Cu Ka radiation (k = 0.15406 nm). X-ray photoelectron spectroscopy (XPS) was measured by an X-ray photoelectron spectrometer (ESCALAB 250) to analyze the surface characteristics of the samples. Nitrogen adsorption and desorption isotherms was conducted at 77 K to investigate the porous texture of the samples on a Quantachrome ASMS analyzer. All samples were degassed at 200 °C for 12 h before the measurement. The Brunauer–Emmett– Teller (BET) method was utilized to calculate the specific surface area. Micropore size distribution was estimated by density functional theory (DFT), while mesopore size distribution by Barrett–J oyner–Halenda (BJH) model from adsorption branch isotherms. Pore volume was calculated from the adsorbed amount at a relative pressure P/P0 = 0.994.
2.1. Material preparation 2.3. Electrochemical measurement Pitch oxide (PO) was prepared by treatment of pitch (Isotropic pitch, Shanghai Dongdao Carbon Chemical Industry Co., Ltd.) using mixed acid. Firstly, 20 ml nitric acid and 80 ml sulfuric acid were blended and stirred in 40 °C water bath for 10 min. The pitch was slowly added into the system in 30 min. After 2 min, 1000 ml deionized water was added into the mixture. Solid precipitate of the mixture was collected by repeated vacuum filtration. Subsequently, the collected solid precipitate and 1 g NaOH were
The working electrode was fabricated by mixing the active material, carbon black and polytetrafluoroethylene (PTFE) with the weight ratio of 80:10:10 and then dried in a vacuum oven at 100 °C for 24 h. Finally, the electrode was pressed onto the nickel foam current collectors at 8 MPa for test. The weight of the electrode is 1.5–2.0 mg. The electrochemical performance of the electrodes was measured in three-electrode cells in 6 M KOH
98
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
Scheme 1. Illustration of the preparation of PO–GO-X.
aqueous electrolyte with platinum sheet and Hg/HgO electrodes as counter and reference electrodes, respectively. All the electrochemical investigations were carried out on a CHI 660C electrochemical workstation using cyclic voltammetry (CV), galvanostatic charge/discharge and electrochemical impedance spectroscopy (EIS) techniques. CV tests were investigated between 0.9 and 0 V (vs. Hg/HgO). Galvanostatic charge/discharge was performed in the same potential range at the current densities ranging from 0.1 to 10 A g 1. EIS test was performed with alternate current amplitude of 0.005 V in the frequency range of 0.01 to 100 kV.
3. Results and discussion 3.1. Morphology and textural properties Fig. 1 presents the surface morphology of the obtained porous carbons derived from PO/GO composite precursor with different ratios. From the four SEM images, one can see clearly their remarkable differences. It can be confirmed that the content of the GO is responsible for the changes. When the GO/PO ratio is set to 1:24, a very low proportion of GO in the composite precursor, the carbon displays no clear distinction between other activated carbons, for example, glucosamine based activated carbons [14]. With the ratio rising to 1:16, some bigger holes, being held up by nanoparticles
with diameter of about 100 nm, can be observed. It can be referred that the graphene plays a role of framework support, which can make sure a homogeneous activation. Increasing the percentage of GO further, as shown in Fig. 1(b), the structural guiding effect of graphene becomes more apparent. The clearly-seen carbon sheets are considered as a composite consisting of graphene and the activated carbon particle with a smaller size. With a highest content of GO, the PO–GO-4 shows many obvious folds. The folds should be from the aggregation of graphene. Fig. 2 shows the N2 sorption isotherms of all PO–GO-Xs, GO and PO. The isotherms for all PO–GO-Xs possess distinct adsorption and desorption branches, as shown in Fig. 2(a), which belong to type-IV profile in virtue of the International Union of Pure and Applied Chemistry (IUPAC) classification. In details, the adsorbed quantity increases steeply to a high value at relative low pressure (close to 0), demonstrating abundant micropores in carbons; the hysteresis loop at medium pressure (0.45–1.0) means the existence of considerable mesopores; at high pressure close to 1.0, the adsorbed curves seem to be relatively smooth, suggesting the negligible quantity of macropores. GO presents a negligible quantity at low relative pressure and a apparent hysteresis loop at medium pressure, meaning typical mesoporous characteristic. PO displays a high low-pressure adsorption quantity and an insignificant hysteresis loop at medium pressure, suggesting a typical microporous characteristic. Their difference demonstrates that graphene has higher skeleton rigidity and is possible to form mesopore wall.
Fig. 1. SEM images of the obtained porous carbons. (a) PO–GO-4, (b) PO–GO-8, (c) PO–GO-16, (d) PO–GO-24.
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
99
Fig. 2. (a) N2 sorption isotherms, (b) pore size distribution of the obtained porous carbons, in which the micropore distribution is calculated from DFT method, while mesopore region from BJH method.
However, because of existence of aggregation, GO has a small adsorption quantity. With a small amount of GO being added, PO–GO-16 and PO–GO-24 show a slight enhancement in low-pressure adsorbed quantity, meaning increased micropores. It may be because that some closed micropores are opened due to existing of some graphene-directing activation channels. However, the adsorbed quantity at low pressure decreases remarkably when more GO is added. That is because that PO is easier to be activated and to form micropores than GO due to numerous structural defect in aromatic-hydrocarbons of pitch in compared with relatively complete hexagonally bonded carbon of graphene [15,16]. With a relative small KOH content (1:1), a moderate activation degree is predictable, which also suggests that mesopores are usually not formed [17]. For PO and PO–GO-24, without GO or with low enough GO content, the hysteresis loop is much smaller, which implies that both PO and PO–GO-24 present principally microporous characteristic and too little GO does not bring about a noticeable change in mesopores. In contrast, those samples with relatively high GO content show more evident hysteresis loops. Therefore, there are reasons for believing that graphene makes a positive role in producing mesopores. However, too much GO does not bring about more mesopores. The sample PO–GO-4 exhibits a less adsorbed quantity at high pressure. It can be referred that the aggregation of graphene impedes the formation of more mesopore walls. Fig. 2(b) presents the pore size distribution for all samples. Based on the sorption isotherms analysis, a significant change with different additions of GO is plainly seen at mesopore region. It’s
well known that BJH method is most suitable to esteem mesopores. So, the mesopore distribution was calculated according to BJH method. For micropores, DFT model is distinctly better than BJH method. Therefore, DFT method was selected to estimated the micropore distribution. It can be seen that PO present both narrow pore distribution at about 0.5 nm and a wide pore distribution at 1–4 nm. The wide micropore distribution is much common for chemical-activated carbons. GO has no micropore distribution, suggesting it is more difficult to be activated than PO. A weak mesopore peak at about 4 nm can be observed. It can be deduced that mesopores are principally due to stacking of graphene. By comparison with PO, all PO–GO-X samples show relatively distinct micropore and mesopore distributions in spite of weak mesopore peak for PO–GO-24 and PO–GO-4. PO–GO-4, PO–GO-8 and PO–GO-16 all show relatively wider micropores than PO–GO-24, which may be attributed to framework effect of homogeneously dispersed graphene in carbon matrix. All PO–GO-Xs display a centered mesopore distribution at 4 nm, relatively small mesopores. In conformity with the sorption isotherms, PO–GO-8 and PO–GO-16 possess much more 4 nm mesopores and PO–GO-24 and PO–GO-4 show a faint 4 nm mesopore peak. Such specific mesopores may be determined by molecular structure and size of PO and GO. Generally, the small mesopores still greatly contribute to total specific surface area. Table 1 lists the pore structure parameters of all samples. Although very limited KOH is used, KOH/composite ratio is 1:1, PO–GO-16 still shows a fairly large specific surface area of 2196 m2 g 1 and pore volume of 1.55 cm3 g 1. PO–GO-24 presents
100
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
a microporosity of 91.5%, even higher than PO, and lower pore volume of 0.98 cm3 g 1 than PO. It is in accordance with the pore size distribution result that PO has a wide pore distribution at 1–4 nm. With the GO/PO ratio increased from 1/24 to 1/16, not only mesoporosity and total volume are notably raised, but also the BET surface area is increased from 2019 to 2196 m2 g 1. It should be due to slightly-increased micropores and a relatively small mesopores. Both PO–GO-8 and PO–GO-16 have a high mesoporosity, close to 50%. The appropriate percentage of mesopores and relative small pore size give a balance between high surface area and enlarged pore diameter. The PO–GO-Xs are further analyzed by XRD patterns. As shown in Fig. 3, one broad and low intensity diffraction peak centered at about 20.8° (2h), which can be approximately indexed as (0 0 2) plane of standard graphite, is observed in the XRD patterns, revealing the characteristics of amorphous carbon usually given by KOH activation procedure [18]. The interlayer spacing, d0 0 2 of the porous carbons was calculated by Bragg’s equation to be 0.427 nm, much higher than 0.365 nm of graphite. Maybe the high content of acidic groups, including carboxyl, hydroxyl or epoxy, on PO molecules and carbon plane of GO should be the reason for the incompletely developed graphite microcrystallite [19]. Another weak peak at 43.3° (2h), corresponding to (1 0 0) diffractions of graphitic layers, was hardly distinguishable, further verifying the incompletely developed graphite microcrystallite. Fig. 4 shows the XPS spectrum of PO–GO-Xs. From the wide XPS spectrum, apart from the strongest C1 peak (284 eV) and apparent O1 peak (533 eV), a weak N1s peak at 402 eV also can be observed clearly, demonstrating that the obtained porous carbons possess a significant amount of O and a small amount of N [20]. Among them, the O could stem from the rich hydroxyl, hydroxyl and sulfo groups of both PO and GO, while N only has one source, namely nitro group of PO treated by mixed acid. Based on the systematic survey of nitrogen- or oxygencontaining carbons, the peaks of the O1s and N1s spectras of various samples are assigned (as shown in Fig. 5). For oxygen, there are three main peaks centered at 530.3 ± 0.2, 531.7 ± 0.2, 533.1 ± 0.2 eV, which represent highly conjugated forms of carbonyl oxygen such as quinine or pyridone groups (OI), AC@O groups (OII), CAOH groups and/or CAOAC groups (OIII), respectively [21–24]. In the same way, the chemical state of nitrogen atoms could be assigned to two types: pyrrolic/pyridone nitrogen (N5, 400.0 ± 0.2 eV) and pyridine-N-oxide (N-X, 402.2 ± 0.2 eV), respectively [23,25,26]. Table 2 lists the relative surface concentrations of oxygen and nitrogen species obtained by fitting the N1s and O1s spectra. With different PO/GO ratios, the obtained porous carbons present different content of nitrogen and similar content of oxygen. The more the GO, the higher the residual nitrogen. It seems farfetched that the carbon from precursor with more PO has less nitrogen. Actually, the difference in content of PO is not very remarkable, just from 80 wt.% of PO–GO-4 to 96 wt.% of PO–GO-24. The evolution and migration of heteroatom in carbon is complex, however, it is
Fig. 3. XRD patterns of PO–GO-Xs.
Fig. 4. The XPS survey spectra.
believed that number of graphene, acting as obstacle for diffusion of heteroatoms, and the porosity, serving as pathway for escaping of N or O-containing gas, exert a crucial influence on resulting N or O content. Because of high graphene content and low porosity, PO– GO-4 shows a relatively high N content. Recent surveys have revealed that N5, OI and OII have electrochemically activity, meaning ability of contributing the pseudocapacitance [27,28]. All PO– GO-Xs own high proportion of active N and O, which would produce considerable pseudocapacitance. In addition to the pseudocapacitance contribution, both the oxygen and nitrogen groups, especially their basic ones, significantly improve the wettability of the surface and enhance the access of the electrolyte ions. 3.2. Electrochemical performance The electrochemical performances of PO–GO-Xs were investigated by cyclic voltammograms (CV) tests and galvanostatic charge/discharge measurements in 6 M KOH solution. Fig. 6 illustrates a set of CV curves of PO–GO-16 at various scan rates from
Table 1 Summary of specific surface area (SBET), total pore volume, average pore diameter, and pore-size distribution data for all samples. Vmicro (%) and Vmeso (%) are volume percentages of micropores and mesopores, respectively. Samples
GO PO PO–GO-4 PO–GO-8 PO–GO-16 PO–GO-24
SBET (m2 g
158 1824 1172 1493 2196 2019
1
)
Total pore volume (cm3 g
0.18 1.18 0.73 1.10 1.55 0.98
1
)
Average pore diameter (nm)
2.31 1.95 2.50 2.94 2.82 1.93
Pore size distribution Vmicro (%)
Vmeso (%)
25.3 85.4 81.2 53.3 52.4 91.5
74.7 14.6 18.8 46.7 47.6 8.5
101
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
Fig. 5. N1s spectra and O1s spectra.
0.005 to 0.1 V s 1 in the potential range between 0.9 and 0 V (vs. Hg/HgO). The electrode presents a typical capacitive behavior with quasi-rectangular shaped voltammetry characteristics as well as broad humps in the CVs, indicating a combination of electrical double-layer capacitance and pseudocapacitance related to redox reactions of the heteroatoms in the electrode materials [29]. With the scan rate increased from 0.005 to 0.1 V s 1, the specific capacitance of vertical ordinate, calculated from current density, displays very insignificant decay, revealing an outstanding rate performance. Even at high rate of 0.1 V s 1, the CV curve still retains near-rectangular shape and no dramatic distortion of the curve could be observed, indicating a highly reversible system and a swift current response. Fig. 6(b) shows CV curves of PO–GO-Xs at scan rates of 0.01 V s 1. All electrodes display a
similar CV curves, including a quasi-rectangular shape, broad humps and rapid current response. Typical galvanostatic charge/discharge curves of PO–GO-16 at different current densities from 0.1 to 10 A g 1 are displayed in Fig. 6(c). It can be seen that all the galvanostatic charge/discharge curves exhibit the shape of isosceles triangular and the charge/discharge curves are highly symmetric. These characteristics also demonstrate that the electrode possesses an ideal capacitive performance and excellent electrochemical reversibility. Fig. 6(d) exhibits galvanostatic charge/discharge curves of all PO–GO-Xs at current density of 1 A g 1. One can clearly seen that PO–GO-16 has the best charge/discharge performance. It can be also found that the ohmic drop of all electrodes is very small, proving a high electrical conductivity of electrodes. The specific capacitances of PO–GO-Xs at different current densities are calculated according to charge/discharge curves, as shown in Fig. 6(e). It can be found that the bigger the specific surface area, the higher the specific capacitance. PO–GO-16 possesses a specific capacitance of 296 F g 1 at 0.1 A g 1 in the 6 M KOH aqueous solution, much higher than that of PO–GO-4 (175 F g 1). With much the same specific area, PO–GO-24 has a relatively low specific capacitance of 256 F g 1. It may be due to the higher content of mesopores, which can provide more accessible surface area. It’s important to note that PO–GO-16, which was prepared with a lower KOH/precursor ratio (1:1), not only presents considerable specific surface area, but also a high mesoporosity, combination of which leads to excellent capacitive performance. Table 3 lists SBET, mesoporosity and capacitance of the reported pitchbased porous carbons, mainly activated ones. Comparison reveals that the PO/GO-based activated carbon possesses a competitive capacitive performance over other pitch-based carbons even if limited KOH is used. With the current density continuously increasing, all PO–GO-Xs present a capacitance drop, but still hold high capacitance retentions. Owing to high surface area and abundant micropores, both the PO–GO-16 and PO–GO-24 show more significant capacitance loss, but still keep high capacitance retentions of 66% and 71%. Fig. 6(f) shows the electrochemical impedance spectroscopy measured in the range from 0.01 to 100 kHz in the threeelectrode cell. The semicircle of the Nyquist plot in the highfrequency range is related to the porous structure of the carbon electrode and corresponds to the charge transfer resistance. The smaller diameter of semicircle implies the lower charge transfer resistance. PO–GO-24 has the biggest diameter of semicircle, which is bigger than that of PO–GO-16. Considering much the same specific surface area, mesopores are thought to provide a more convenient route for ion transferring. PO–GO-4 and PO–GO-8 display the smaller semicircle. It should be mainly attributed to graphene, which brings about an enhanced electron conductivity. The line of the Nyquist plot in the low-frequency range is related to the capacitive behavior of the electrode. All electrodes display nearly vertical lines, which reflect an idea supercapacitive behavior.
Table 2 XPS analysis and XPS peak positions and relative surface concentrations of oxygen and nitrogen species obtained by fitting the N 1s and O 1s spectra.
a b
Sample
N (at.%)
PO–GO-4 PO–GO-8 PO–GO-16 PO–GO-24
2.77 1.68 1.52 0.92
N1s (%)a N-5 (400.0 eV)
N-X (402.2 eV)
94.2 80.5 86.7 76.2
5.8 19.5 13.3 23.8
Relative contents of functional groups in N1s. Relative contents of functional groups in O1s.
O (at.%)
O1s (%)b O-I (530.3 eV)
O-II (531.7 eV)
O-III (533.1 eV)
6.6 7.1 7.2 5.5
11.7 8.2 6.9 8.7
62.1 48.0 36.5 46.2
26.2 43.8 56.6 45.1
102
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
Fig. 6. (a) CV curves of PO–GO-16 at different scan rates from 0.005 to 0.1 V s 1. (b) CV curves of PO–GO-Xs at scan rates of 0.01 V s 1. (c) Charge/discharge curves of PO–GO-16 at different current densities from 0.1 to 10 A g 1. (d) Charge/discharge curves of PO–GO-Xs at current densities from 1 A g 1. (e) Variation of the specific capacitance of PO–GO-Xs as a function of current density. (f) Nyquist plots of PO–GO-Xs measured in the three-electrode cell.
Table 3 Comparison of specific capacitances and specific surface area with other pitch-based carbons. Precursor
Methods
SBET (m2 g
Heat-treated pitch Sulfonated pitch Pitch-based carbon fibers Pitch Coal tar pitch/melamine Coal tar pitch based ACMs PO/GO
KOH(1:4) KOH(1:3) Ammonia KOH(1:3) KOH(1:3) P123 KOH(1:1)
2003 3548 1454 983 2573 860 2196
3E: three-electrode. 2E: two-electrode.
1
)
Vmeso/Vtotal (%)
C (F g
31.7 59.3 – 11.4 56.0 – 47.6
170 277 146 276 228 172 296
1
)
(0.5 A g 1) (2 mV s 1) (1 A g 1) (5 mV s 1) (1 A g 1) (10 mV s 1) (0.1 A g 1)
Cell
Electrolytes
Ref.
3E 3E 2E 3E 3E 3E 3E
6M 6M 6M 1M 6M 1M 6M
[30] [31] [32] [33] [34] [35] This work
KOH KOH KOH H2SO4 KOH H2SO4 KOH
Y. Ma et al. / Journal of Colloid and Interface Science 461 (2016) 96–103
4. Conclusion In summary, we demonstrate the facile synthesis of a novel hierarchical porous carbon with high specific surface area, rich nitrogen and oxygen, appropriate mesopore/micropore ratio and considerable small-sized mesopores through one-step activation of pitch oxide/graphene oxide precursor. The addition of graphene plays a key role in forming 4 nm mesopores. As-prepared sample PO–GO16 presents the characteristics of large surface area (2196 m2 g 1), high mesoporosity (47.6%), as well as rich nitrogen (1.52 at.%) and oxygen (6.9 at.%). As a result, PO–GO-16 electrode shows outstanding capacitive behavior: high capacitance (296 F g 1) and ultrahighrate performance (192 F g 1 at 10 A g 1) in 6 M KOH aqueous electrolyte. Therefore, taking the low-cost, less KOH consumption and simplified process into account, the well-designed and high-performanced PO–GO-Xs provide new opportunities for widespread application of supercapacitors. References [1] [2] [3] [4] [5] [6]
P. Simon, Y. Gogotsi, B. Dunn, Science 343 (2014) 1210. L. Hao, X. Li, L. Zhi, Adv. Mater. 25 (2013) 3899. Y. Li, Z. Li, P.K. Shen, Adv. Mater. 25 (2013) 2474. Z. Chen, D. Weng, H. Sohn, M. Cai, Y. Lu, RSC Adv. 2 (2012) 1755. F. Béguin, V. Presser, A. Balducci, E. Frackowiak, Adv. Mater. 14 (2014) 2219. Q. Wang, J. Yan, Y. Wang, T. Wei, M. Zhang, X. Jing, Z. Fan, Carbon 67 (2014) 119. [7] D.W. Wang, F. Li, M. Liu, G.Q. Lu, H.M. Cheng, Angew. Chem. Int. Ed. 120 (2008) 379. [8] C. Ma, J. Shi, Y. Song, D. Zhang, X. Zhai, M. Zhong, Q. Guo, L. Liu, Int. J. Electrochem. Sci. 7 (2012) 7587. [9] X.Y. Chen, C. Chen, Z.J. Zhang, D.H. Xie, X. Deng, Ind. Eng. Chem. Res. 52 (2013) 10181.
103
[10] X. Zhai, Y. Song, J. Liu, P. Li, M. Zhong, C. Ma, W. Huiqi, J. Electrochem. Soc. 159 (2012) E177. [11] S. Zhong, C. Zhan, D. Cao, Carbon 85 (2015) 51. [12] Y. Bai, R.B. Rakhi, W. Chen, H.N. Alshareef, J. Power Sources 233 (2013) 313. [13] W.S. Hummers, R.E. Offeman, JACS 80 (1958) 1339. [14] L. Zhao, L.Z. Fan, M.Q. Zhou, H. Guan, S.Y. Qiao, M. Antonietti, M.-M. Titirici, Adv. Mater. 22 (2010) 5202. [15] S. Lee, Y. Eom, B.-J. Kim, I. Mochida, S.-H. Yoon, B.C. Kim, Carbon 81 (2015) 694. [16] H. Bai, C. Li, G. Shi, Adv. Mater. 23 (2011) 1089. [17] L. Zhu, Q. Gao, Y. Tan, W. Tian, J. Xu, K. Yang, C. Yang, Micropor. Mesopor. Mater. 210 (2015) 1. [18] J. Jiang, H. Chen, Z. Wang, L. Bao, Y. Qiang, S. Guan, J. Chen, J. Colloid Interface Sci. 452 (2015) 54. [19] W. Gao, L.B. Alemany, L. Ci, P.M. Ajayan, Nat. Chem. 1 (2009) 403. [20] K.-S. Kim, S.-J. Park, Electrochim. Acta 56 (2011) 10130. [21] H. Li, Z. Song, X. Zhang, Y. Huang, S. Li, Y. Mao, H.J. Ploehn, Y. Bao, M. Yu, Science 342 (2013) 95. [22] Y. Zhao, C. Hu, Y. Hu, H. Cheng, G. Shi, L. Qu, Angew. Chem. Int. Ed. 45 (2012) 11371. [23] M. Seredych, D. Hulicova-Jurcakova, G.Q. Lu, T.J. Bandosz, Carbon 46 (2008) 1475. [24] J.H. Zhou, Z.J. Sui, J. Zhu, P. Li, D. Chen, Y.C. Dai, W.K. Yuan, Carbon 45 (2007) 785. [25] Y.J. Kim, Y. Abe, T. Yanagiura, K.C. Park, M. Shimizu, T. Iwazaki, S. Nakagawa, M. Endo, M.S. Dresselhaus, Carbon 45 (2007) 2116. [26] E. Raymundo-Piñero, D. Cazorla-Amorós, A. Linares-Solano, J. Find, U. Wild, R. Schlögl, Carbon 40 (2002) 597. [27] X. Yang, D. Wu, X. Chen, R. Fu, J. Phys. Chem. C 114 (2010) 8581. [28] C. Ma, Y. Song, J. Shi, D. Zhang, Q. Guo, L. Liu, J. Colloid Interface Sci. 395 (2013) 217. [29] X. Fan, C. Yu, J. Yang, Z. Ling, J. Qiu, Carbon 70 (2014) 130. [30] G.-Y. Heo, S.-J. Park, Powder Technol. 239 (2013) 94. [31] Y. Guo, Z.-Q. Shi, M.-M. Chen, C.-Y. Wang, J. Power Sources 252 (2014) 235. [32] K. Torchała, K. Kierzek, G. Gryglewicz, J. Machnikowski, Electrochim. Acta 167 (2015) 348. [33] M.-S. Park, S. Cho, E. Jeong, Y.-S. Lee, J. Ind. Eng. Chem. 23 (2015) 27. [34] C. Zhong, S. Gong, L.E. Jin, P. Li, Q. Cao, Mater. Lett. 156 (2015) 1. [35] Y. Cheng, T. Li, C. Fang, M. Zhang, X. Liu, R. Yu, J. Hu, Appl. Surf. Sci. 282 (2013) 862.