Open Access Article
Qiuli Chen‡
a,
Jinfeng Sun‡b,
Zhengluo Wanga,
Zhiwei Zhaoa,
Yanru Zhanga,
Yang Liub,
Linrui Hou*b and
Changzhou Yuan
*b
aSchool of Materials Science & Engineering, Anhui University of Technology, Ma'anshan, 243002, P. R. China
bSchool of Materials Science & Engineering, University of Jinan, Jinan, 250022, P. R. China. E-mail: mse_houlr@ujn.edu.cn; mse_yuancz@ujn.edu.cn; ayuancz@163.com; Fax: +86-531-82769410; Tel: +86-531-82769410
First published on 2nd March 2018
Cost-efficient carbonaceous materials have been utilized extensively for advanced electrochemical supercapacitors. However, modest gravimetric/volumetric capacitances are the insuperable bottleneck in their practical applications. Herein, we develop a simple yet scalable method to fabricate low-cost micro-/mesoporous N/O-enriched carbon (NOC-K) by using natural rose multiflora as a precursor with KOH activation. The biomass-derived NOC-K is endowed with a large surface area of ∼1646.7 m2 g−1, micro-/mesoporosity with ∼61.3% microporosity, high surface wettability, and a high content of N (∼1.2 at%)/O (∼26.7 at%) species. When evaluated as an electroactive material for supercapacitors, the NOC-K electrode (5 mg cm−2) yields large gravimetric/volumetric specific capacitances of ∼340.0 F g−1 (∼238.0 F cm−3) at 0.5 A g−1, and even ∼200.0 F g−1 (∼140.0 F cm−3) at 5.0 A g−1, a low capacitance decay of ∼4.2% after 8200 consecutive cycles, and a striking specific energy of ∼8.3 W h kg−1 in aqueous KOH electrolyte, benefiting from its intrinsic structural and compositional superiorities. Moreover, a remarkable specific energy of ∼52.6 W h kg−1 and ∼96.6% capacitance retention over 6500 cycles for the NOC-K based symmetric cell are obtained with the organic electrolyte. More promisingly, the competitive NOC-K demonstrates enormous potential towards advanced supercapacitors both with aqueous and organic electrolytes as a sustainable electrode candidate.
In general, the superior specific capacitances (SCs) of carbon materials are mainly determined by their large SSA, hierarchical micro-/mesopores, and high-surface heteroatom content (such as O, N, B, P, etc.).3,4,10–18 To begin with, the high SSA can guarantee massive electroactive sites for fast and efficient charge accommodation.11–20 Besides, bi-modal micro-/mesopores greatly favor for convenient ion diffusion into the electrode bulk by providing ion highways, thus ensuring high-power property, and additionally contribute to large electroactive SSA meanwhile.3,18,20 In addition, heteroatom doping can partially form N- and O-based surface functional groups, which improves the surface wettability of carbon electrodes for effective charge storage.11,21–23 Notedly, the polar functional groups upon the carbon surface exhibit benign affinity to solvents with the same polarity, especially in the ubiquitous aqueous systems.24 Of particular note, the N doping also can change the intrinsic conductivity of carbon materials themselves by adjusting the giving/receiving capacity of the electrons in the C atom, which makes them even better electronic conductivity than their phase-pure counterparts, thanks to the high electronegativity, small atomic size, and additional free electrons of the N itself.3,13,20,21 Therefore, the nitrogen species have attracted the huge interest of many researchers around the word. It is noted that two strategies are generally explored to incorporate the elemental N into the porous carbon materials: post-treatment of porous carbon with extra N-based sources (such as, ammonia, urea, amoine, and so on) and in situ doping by directly pyrolyzing various N-containing precursors. Commonly, the former only introduces N-based functional groups on the surface of the carbon materials, while the latter can easily realize the homogeneous heteroatom doping into the bulk with controlled chemistry, rather than just on the surface of carbon electrodes.10
Recently, biomass materials, different from those artificial materials, have drawn enormous attentions, owing to their connate advantages including richness in nature, low cost, renewability and environmental friendlyness. Carbon materials derived from various biomasses not only possess unique special morphologies, but also can be in situ doped with the heteroatoms originally existing in these biomass precursors. Consequently, this type of carbon materials inherently owns higher market potential due to the lower cost than commercial carbon nanotubes and graphene.25 Nowadays, lots of pioneering works have been reported by using the human hair,14 dead leaves,16 shiitake,18 albumen,20 coconut shell,26 corn stover,27 banana peel,28 cashmere,6 and others as the precursors.
With aforementioned considerations in mind, herein, we proposed that a facile yet efficient synthesis of N/O co-functionalized carbons with bi-modal micro-/mesopores by using rose multiflora as the precursor. After KOH activation process at 800 °C, the as-resulted carbon electrode was obtained with high-content nitrogen (∼1.2 at%) and oxygen (∼26.7 at%) functionally doping, micro-/meso-porous channels with microporosity of ∼61.3%, large SSA (∼1646.7 m2 g−1) and appealing surface wettability. When evaluated as an electrode material for ESs, the KOH-activated nitrogen/oxygen-enriched carbon (denoted as NOC-K) electrode exhibited remarkable electrochemical capacitances both in aqueous KOH solution and organic tetraethylammonium tetrafluoroborate/propylene carbonate (TEABF4/PC) electrolyte, benefiting from its unique structural and compositional superiorities.
:
1 (w/w). Afterwards, the mixture was activated under N2 flow at 800 °C for 2 h in a tubular furnace with a heating rate of 3 °C min−1. After the completion of the activation, the sample was allowed to cool to room temperature. After purified in diluent HCl solution, rinsed with de-ionized water until neutral, and dried at 60 °C, the resulting black NOC-K product was obtained. For comparison, the carbon prepared with the absence of KOH was designed as NOC, while other parameters were kept the same as those for the NOC-K.
:
2
:
1. The mass loading of electroactive material on each electrode is 5 mg for three-electrode evaluation. And two-electrode symmetric cell with two identical electrodes were assembled into coin cells (CR2032) with 6 M KOH as the electrolyte.
In addition, a symmetric cell was also electrochemically characterized in 1 M TEABF4/PC electrolyte. The electrode was prepared by mixing NOC-K or NOC, acetylene black, polyvinylidene difluoride (PVDF) binder (8
:
1
:
1, w/w/w) dispersed in the N-methylpyrrolidinone (NMP). The slurries coated on the Al and steel foils were employed as positive and negative electrodes. After dried in vacuum at 120 °C for 24 h, the electrodes were separated with glass fiber (GF/D, Whatman) in CR2032 coin-type cells for electrochemical measurement in 1 M TEABF4/PC electrolyte.
For the three-electrode systems or symmetric ESs, the mass capacitance (F g−1) and volumetric specific capacitance (VSC) (F cm−3) derived from galvanostatic tests can be calculated by the following equations:
![]() | (1) |
![]() | (2) |
For the symmetric cell, the SE and specific power (SP) were calculated as follows:
![]() | (3) |
![]() | (4) |
![]() | ||
| Fig. 1 Schematic illustration of the fabricating procedure for the NOC-K with the rose multiflora as a starting material. | ||
Fig. 2a shows typical FESEM image of the hierarchical NOC-K product, and the sample presents loose morphology with interconnected hierarchical pores. TEM characterizations (Fig. 2b–d) were also carried out to further examine the specific microstructures of the sample more clearly. As shown in Fig. 2b, the discernable mesoporous structure can be presented for the NOC-K. Further high-magnification TEM (Fig. 2c) and HRTEM (Fig. 2d) observations show staggered and interpenetrated micropores dominating in the NOC-K sample along with several scattered mesopores, which would contribute to even more electroactive sites, and provide convenient and continuous channel meanwhile for the smooth entry and migration of electrolyte ions. As for the KOH activation, the expected products including CO2 and H2O formed in the activation process, and the removal of intercalated metallic K and K-based compounds by following washing both can result in the generation of well-developed pore network in the nano-scale and/or sub-nanometer range observed for the NOC-K specimen. Corresponding energy dispersive spectrometry (EDS) elementary mapping analysis of the NOC-K sample, as evidenced in Fig. 2e–h, visually confirms the homogeneous co-existence and uniform distribution of the C, N and O species in the whole NOC-K sample.
![]() | ||
| Fig. 2 FESEM (a), TEM ((b) and (c)) and HRTEM (d) images of the NOC-K product, corresponding selected TEM area (e) for elemental ((f), C; (g), N and (h), O) mapping images. | ||
To better understand the unique bi-modal porous structure and SSA parameters, we measured N2 sorption isotherms of the NOC-K and NOC samples, as comparatively described in Fig. 3. Obviously, typical adsorption–desorption isotherms of the two, as plotted in Fig. 3a, are close to the combination form of the type I and IV according to the classic International Union of Pure and Applied Chemistry (IUPAC) definition. This distinct feature suggests the co-existence of the micro-/mesopores both in the NOC-K and NOC products, which can be well supported by their mesopore size distribution data (Fig. 3b) derived from the adsorption branches by the Barrett–Joyner–Halenda method, and the micropore size distribution (the inset in Fig. 3b) estimated by the Horvanth–Kawazone method. According to the calculation from the adsorption branch, the BET SSA of the NOC-K is ∼1646.7 m2 g−1, which is approximately two times that of the NOC (∼826.6 m2 g−1). And the total pore volume are estimated as ∼0.39 and ∼0.84 cm3 g−1 for the NOC and NOC-K, respectively, as determined from the N2 adsorption amount at P/P0 = 0.97. According to the ratio of micropore volume to total pore volume, ∼62.4% and ∼61.3% of the porosity for the NOC and NOC-K come from the micropores. As a result, the average pore sizes of the NOC and NOC-K can be derived as ∼2.1 and ∼2.3 nm, respectively. As noted, the NOC-K product obtains high SSA and well-developed porosity with even higher proportion mesoporosity, which are of vital significance to enhance the electrolyte accessibility to the microporous surface for efficient charge storage.
![]() | ||
| Fig. 3 N2 adsorption–desorption isotherms (a), mesopore size distribution (b) and micropore size distribution (the inset in panel b) for the NOC and NOC-K samples. | ||
Fig. 4a demonstrates the Raman spectra of the NOC and NOC-K. Typically, both of the samples mainly show two distinct peaks located at around 1350 cm−1 (D-band) and 1590 cm−1 (G-band), respectively. As is known to all, the D-band corresponds to defect sites or disordered sp2-hybridized carbon atoms of graphite, and the G-band is related to the phonon mode in-plane vibration of sp2-bonded carbon atoms.29 The IG/ID values of the NOC and NOC-K are ∼0.85 and ∼0.65, respectively. The reduced IG/ID value of the NOC-K sample suggests more defective structure and lower graphitization owing to the subsequent high-temperature KOH activation.
Fig. 4b–f collectively summarize the high-resolution elemental XPS spectra of the NOC and NOC-K, and corresponding profiles fitted by using the Gaussian fitting method. Clearly, three elements including elemental carbon, oxygen and nitrogen are all discernable in the two products. Typically, the C 1s spectra for the two, as evidenced in Fig. 4b, can be deconvoluted into following five peaks including C
C (C-I, ∼284.6 eV), C–C (C-II, ∼285.0 eV), C–N/C–O (C-III, ∼285.6 eV), C
O (C-IV, ∼286.6), and O–C
O (C-V, ∼288.4) bonds.30 Fig. 4c shows the deconvolutions of the O 1s spectra to identify the surface oxygen-containing functional groups upon the NOC and NOC-K products. Specifically, the three peaks at binding energies of ∼532.6, ∼533.8 and ∼535.1 eV correspond to the carbonyl (C
O) (O-I), hydroxyl (C–O) (O-II), and carbonxyl (O
C–O) (O-III), respectively.31 The N 1s spectra of the samples (Fig. 4d) can be fitted well into four peaks centered at ∼397.8, ∼399.5, ∼400.6 and ∼402.7 eV, respectively, which are assigned to four types of nitrogen functional group, namely pyridinic (N-I), pyrrolic and/or pyridonic (N-II), quaternary nitrogen (N-III) and oxidized N-oxide species (N-IV).3,8,14,15 Compared with the NOC, total elemental nitrogen and carbon contents of the NOC-K sample are largely increased up to ∼1.2 at% and ∼72.1 at% from ∼0.5 at% and ∼66.2 at%, respectively, while the percentage of elemental oxygen is reduced from ∼33.3 at% to ∼26.7 at%. One especially note that the N-III and N-IV in the NOC-K (Fig. 4f), favoring for the rapid electronic transfer through the carbon electrodes,32,33 are estimated as ∼0.36 and ∼0.26 at%, higher than those for the NOC (N-III, ∼0.21 at%; N-IV, ∼0.12 at%) (Fig. 3e). There is no doubt that the unique feature of the NOC-K is greatly favorable for its enhanced electrochemical energy storage capacity, particularly at high rates, and to some extent, can make up for the lower graphitization of the NOC-K after KOH activation for enhanced rapid charge-storage ability.
Besides, the surface N-/O-containing functional groups play a significant role in enhancing the hydrophilicity and wettability of the carbon electrodes. To further examine the wettability of the NOC and NOC-K electrodes, the contact angle measurement was conducted, as shown in Fig. 5. Visually, the contact angle of the NOC with the 6 M KOH aqueous solution is measured as ∼71.6°, which is greatly larger than that of the NOC-K (∼47°). It confirms the better wettability of the NOC-K sample with the aqueous KOH, which makes the electrolyte easily impregnate and reach the inner pore surface as large as possible for higher electrochemical efficiency of the total surface area towards enhanced electrochemical behaviors.
For further in-depth investigation into the specific electrochemical processes for the NOC and NOC-K electrodes, EIS tests were performed in the frequency from 0.01–105 Hz with an AC signal amplitude of 5 mV at a potential of −0.3 V (vs. SCE). Typical Nyquist plots and corresponding fitted files with the equivalent circuit model (the upper inset) for the two are collected in Fig. 6d, and the fitting parameters are summarized (Table S2, ESI†). The intersections of the plots at the X-axis reasonably represent the solution resistance (Rs), which to a certain extent is related to the intrinsic resistance of electroactive materials itself. As observed from the enlarged high-medium frequency region (the lower inset in Fig. 6d), the Rs of the NOC-K can be estimated as ∼0.4 ohm, somewhat smaller than that of the NOC (∼0.6 ohm). Furthermore, the semicircle, whose diameter stands for the charge transfer resistance (Rct), can be notable in Fig. 6d. The little semicircle of the NOC-K means its small Rct (∼1.6 ohm), and the Rct is approximated to ∼6.4 ohm for the NOC. Evidently, the smaller Rct and Rs values of the NOC-K electrode should well account for its superior power behaviors, as discussed in Fig. 6c, compared with those for the NOC electrode.
In order to investigate the practical application of the NOC-K as potential electrodes for ESs, a symmetric device was assembled by two identical NOC-K electrodes with total mass loading of 6 mg cm−2, along with the 6 M KOH as aqueous electrolyte. Fig. 7a shows the CV curves of the NOC-K based symmetric device with various sweep rates ranged from 50 to 500 mV s−1 in the potential window with a upper voltage of 1.0 V. Obviously, the symmetric device presents representative supercapacitive performance with quasi-rectangular shape voltammetry characteristics with respect to the zero-line and rapid current responses on voltage reversal at each end voltage, even at a high scanning rate of 500 mV s−1, suggesting its outstanding electrochemical properties in KOH. Fig. 7b plots the CP data of the symmetric supercapacitor at different current densities from 0.5 to 10.0 A g−1. It can be seen that all these charge–discharge curves show a standard regular triangle, further supporting its excellent double layer capacitance characteristics. According to the data in Fig. 7b, the SCs of the device can be quantified, and collected in Fig. 7c. Remarkably, the symmetric cell yields high SCs of ∼59.6, ∼56.8, ∼54.1, ∼51.6, ∼50.5, ∼49.1 and ∼44.1 F g−1, at the current densities of 0.5, 1.0, 2.0, 3.0, 4.0, 5.0 and 10.0 A g−1, respectively, that is, just a SC degradation of ∼74.0% is obtained with the current rate increasing from 0.5 to 10 A g−1, strongly suggesting its appealing rate behaviors. As is well known, electrochemical capacitances of any cell are hugely dependent upon the mass loadings of the electroactive material. Fig. 7d contrastively shows the CV curves at a sweep rate of 100 mV s−1 for the three symmetric capacitors with the total mass loadings of 6, 8 and 10 mg, respectively. Evidently, higher electrochemical current responses and larger integrated areas under the I–E curves are observed with the mass loading increasing. As calculated from the CP plots in Fig. 7e, the SCs of the symmetric cells are still maintained as ∼50.8 and ∼50.5 F g−1, which are ∼85.2% and ∼84.7% of the SC for the symmetric cell with the mass loading of 6 mg, when the total mass loading is up to 8 and 10 mg. The excellent SC independence of the NOC-K upon the mass loading fully highlights its remarkable capability for versatile potential applications.
Fig. 7f illustrates the long-span electrochemical stability of the symmetric device with a mass loading of 6 mg. Strikingly, the high SC retention of ∼95.8% over continuous 8200 cycles, i.e., ∼0.05% SC decay per 100 cycles, consistently verifies the excellent long-term cyclic stability of the NOC-K based symmetric cell at a current rate of 0.5 A g−1. Additionally, the CE of the cell is nearly always kept as high as ∼100%, suggesting its highly electrochemical reversibility for efficient charge storage. This fully confirms the promising advantages of the NOC-K electrode for advanced ESs.
Ragone plots of the NOC-K based symmetric devices with the TEABF4/PC and aqueous KOH electrolytes are further depicted in Fig. 9a. As shown in Fig. 9a, the aqueous symmetric cell with 6 M KOH as electrolyte just yields a maximal SE of ∼8.3 W h kg−1 (i.e., ∼2.7 W h L−1) along with a SP of 250 W kg−1, and at the case of 5 kW kg−1, only a SE of ∼6.1 W h kg−1, corresponding to a volumetric SE of ∼2.0 W h L−1, can be rendered. Appealingly, the gravimetric SE for the aqueous cell can be compared to, and/or even higher than others reported before,5,39,40,42,43,48 as plotted in Fig. 9a. As for the organic device, the highest SE of the cell with 1 M TEABF4/PC electrolyte has been obtained as large as ∼52.6 W h kg−1 (∼2.6 W h L−1) at a SP of ∼700 W kg−1. And attractively, the device still can achieve a SE of ∼38.9 W h kg−1 (∼1.9 W h L−1) at a higher SP of ∼14.0 kW kg−1, which is much larger than those for other carbon-based symmetric systems,5,41–49 as summarized in Fig. 9a. Moreover, the symmetric device with the TEABF4/PC electrolyte can efficiently power a light-emitting diode indicator, and maintain it for around 90 s, as visually demonstrated in Fig. 9b and c.
Footnotes |
| † Electronic supplementary information (ESI) available: Electronic supporting information corresponding XRD pattern, charge–discharge plots, electrochemical comparisons with other carbon-based electrodes, EIS simulated results, and cycling performance of the symmetric devices within the potential range from 0.0 to 2.5 or 3.0 V. See DOI: 10.1039/c8ra00858b |
| ‡ These authors contributed equally to this work. |
| This journal is © The Royal Society of Chemistry 2018 |