hierarchical porous carbon aerogel derived from bagasse ... · metal oxides and conducting polymers...
TRANSCRIPT
Nanoscale
PAPER
Publ
ishe
d on
29
July
201
4. D
ownl
oade
d by
Uni
vers
ity o
f H
oust
on o
n 07
/03/
2016
17:
28:3
0.
View Article OnlineView Journal | View Issue
Hierarchical poro
aState Key Laboratory of Crystal Materials,
China. E-mail: [email protected] of Chemical Materials, China Aca
621900, P. R. ChinacSchool of Materials Science and Engineerin
E-mail: [email protected] of Chemistry, Faculty of Scien
11001, Kingdom of Saudi Arabia. E-mail: ahePromising Centre for Sensors and Electro
Najran-11001, Kingdom of Saudi Arabia
† Electronic supplementary informa10.1039/c4nr03574g
‡ Pin Hao and Zhenhuan Zhao contribute
Cite this: Nanoscale, 2014, 6, 12120
Received 26th June 2014Accepted 16th July 2014
DOI: 10.1039/c4nr03574g
www.rsc.org/nanoscale
12120 | Nanoscale, 2014, 6, 12120–1212
us carbon aerogel derived frombagasse for high performance supercapacitorelectrode†
Pin Hao,‡a Zhenhuan Zhao,‡a Jian Tian,a Haidong Li,a Yuanhua Sang,a Guangwei Yu,a
Huaqiang Cai,b Hong Liu,*a C. P. Wong*c and Ahmad Umar*de
Renewable, cost-effective and eco-friendly electrodematerials have attractedmuch attention in the energy
conversion and storage fields. Bagasse, the waste product from sugarcane that mainly contains cellulose
derivatives, can be a promising candidate to manufacture supercapacitor electrode materials. This study
demonstrates the fabrication and characterization of highly porous carbon aerogels by using bagasse as
a raw material. Macro and mesoporous carbon was first prepared by carbonizing the freeze-dried
bagasse aerogel; consequently, microporous structure was created on the walls of the mesoporous
carbon by chemical activation. Interestingly, it was observed that the specific surface area, the pore size
and distribution of the hierarchical porous carbon were affected by the activation temperature. In order
to evaluate the ability of the hierarchical porous carbon towards the supercapacitor electrode
performance, solid state symmetric supercapacitors were assembled, and a comparable high specific
capacitance of 142.1 F g�1 at a discharge current density of 0.5 A g�1 was demonstrated. The fabricated
solid state supercapacitor displayed excellent capacitance retention of 93.9% over 5000 cycles. The high
energy storage ability of the hierarchical porous carbon was attributed to the specially designed pore
structures, i.e., co-existence of the micropores and mesopores. This research has demonstrated that
utilization of sustainable biopolymers as the raw materials for high performance supercapacitor
electrode materials is an effective way to fabricate low-cost energy storage devices.
1. Introduction
Recently, supercapacitors, known as electrochemical capacitors(ECs), with high power density, high specic capacitance andexcellent cycling stability characteristics have received consid-erable attention due to increased demand of energy storagedevices. Typically, supercapacitors deliver a power density thatis an order of magnitude higher (10 000 W kg�1) than that oflithium ion batteries, and an energy density that is two orders ofmagnitude higher (10 W h kg�1) than that of electrolyticcapacitors.1,2 Supercapacitors can be quickly charged and have
Shandong University, Jinan 250100, P. R.
demy of Physical Engineering, Mianyang,
g, Georgia Tech, Atlanta, GA 30032, USA.
ce and Arts, Najran University, Najran-
nic Devices (PCSED), Najran University,
tion (ESI) available. See DOI:
d equally to this work.
9
ultralong lifetimes even when discharged at a high currentdensity. Thus, supercapacitors have specic applications, suchas backup energy, electric vehicles, wind power and even toreplace batteries.3 Commercial supercapacitors mainly useporous carbon as the active electrode materials but they exhibitrelatively low energy density.4,5 This is the reason why consid-erable research effort has mainly focused on enhancing theenergy storage abilities of supercapacitors. For this, transitionmetal oxides and conducting polymers are widely used as activeelectrode materials that store electrical energy based on theredox reactions.6–9 However, supercapacitors using transitionmetal oxides and conducting polymers possess poor cyclabilityand rate capability despite the high energy density and speciccapacitance. Research has thus been focused on increasing theenergy density without sacricing the cycle life or high powerdensity.
The energy storage ability of a supercapacitor is determinedby the choice and structure of the active electrode materials.10
Carbon materials, including activated carbon, carbon bers,carbon nanotubes, carbon aerogels, graphene and carbide-derived carbons, are widely used as active electrode mate-rials.11–19 These carbon materials have very good supercapacitorbehavior. However, most of them are prepared from mineralmaterials and petroleum, which are expected to be depleted in
This journal is © The Royal Society of Chemistry 2014
Paper Nanoscale
Publ
ishe
d on
29
July
201
4. D
ownl
oade
d by
Uni
vers
ity o
f H
oust
on o
n 07
/03/
2016
17:
28:3
0.
View Article Online
the near future. The high cost of raw materials, environmentaldestructiveness of preparation and complicated manufacturinglargely limit the further applications of most supercapacitors.Thus, research has focused on the utilization of green repro-ducible biomass or their derivatives to produce porous carbonmaterials, which are critical to the sustainable development andenvironmental protection.20,21 Accordingly, considerable effortshave been made to prepare carbon based on natural sourcessuch as carbonization of chicken feathers and banana peels forenergy storage devices.22–29 Although various carbon basedmaterials were prepared by a variety of natural ways and usedfor energy storage devices, the electrochemical performances ofthe fabricated devices were unsatisfactory.
Different from other reported literature, the present researchdemonstrates the fabrication of hierarchical porous carbonmaterials using bagasse as a raw material. Bagasse is theindustrial waste of sugar rening industries, which is producedaer sucrose extraction from the sugarcane plant. Bagasse has ahigh proportion of cellulose; hence, it presents itself as apotential and interesting candidate for the production ofporous carbon electrode materials.30 Bagasse originates fromabundant and essentially free renewable biomasses, and withthe growing demand for sugar, much more bagasse will beproduced as industrial waste.
Unlike the proposed idea presented in this paper, in previousstudies, cellulose was utilized as the supporting materials andsubstrates, which mainly has the advantages of mechanicalproperties and exibility of cellulose.31–33 However, these ideasstill could not make full use of cellulose to prepare the all-carbon materials, such as carbon nanotube or graphene, whichare expensive. Therefore, converting bagasse to an electrodematerial is supposed to be an economical and environmentallyfriendly route.
For the structure of carbon materials, the electrochemicalperformance of ECs is dependent on several parameters, suchas the electric conductivity of the electrode materials, pore sizeand distribution, and specic surface area. The electricconductivity affects the power density that the ECs can deliver.The specic surface area and pore size affects the speciccapacitance and energy density. For example, KOH-activatedporous graphene exhibits a very high specic surface area(3100 m2 g�1).34 However, the increase in specic surface areacannot always positively enhance the specic capacitance. Thisis because with increasing specic surface area, the porevolume of micropores increases and the size of the microporesdecreases to below the size of hydrated electrolyte ions, whichdecreases the accessibility of pores.3 The ideal ECs electrodematerials should have a hierarchical porous structure contain-ing macropores (larger than 50 nm) for the ion-bufferingreservoir, mesopores (2–50 nm) for ion transportation andmicropores (less than 2 nm) for the enhancement of chargestorage. Moreover, it is also desirable that the porous carbonstructures should have multi-scale pores.35
In this study, hierarchical porous carbon aerogels wereprepared from freeze drying cellulose aerogels using bagasse asthe raw material. The direct carbonization of cellulose aerogelcreates numerous macro- and meso-pores. A post KOH
This journal is © The Royal Society of Chemistry 2014
activation process, which is widely used for manufacturingporous carbon was employed to produce micropores on thewalls of the mesoporous carbon aerogels. Therefore, the presentwork provides a general hint for the utilization of sustainablematerials for hierarchical porous carbon used for super-capacitors, which is effective, easily to operate, low cost, andenvironmental friendly. The special hierarchical porous struc-tures demonstrate high electrochemical performance andvarious fascinating advantages. Bagasse, the raw material, isindustrial waste that can be efficiently utilized for the large-scale production of hierarchical porous carbon materials.
Furthermore, the prepared hierarchical carbon aerogelspossess high specic surface area with reasonable pore sizedistributions and hence exhibited good energy storage perfor-mance and high cycling stability. Moreover, the effects of theactivation temperatures on the pore size distribution were alsoinvestigated and their corresponding electrochemical perfor-mance was compared. Finally, the real supercapacitive perfor-mance of hierarchical carbon aerogel with optimizedconditions were studied using an all-solid-state supercapacitor.
2. Experimental2.1 Materials
The bagasse was obtained from the sugarcane (purchased froma supermarket), which were dried under laboratory conditionsand then dried again at 100 �C. The cellulose used in this studywas extracted from the above bagasse. All other chemicals wereof reagent grade and used without further purication. Deion-ized water was used throughout the experiments.
2.2 Preparation of cellulose aerogel
The cellulose was puried from the bagasse by an alkalinehydrolysis, according to the reported method.36 Initially, inorder to remove the lignin and hemicellulose, the dried bagassewas kept at 80 �C in a 4% sodium hydroxide solution for 4 h.Following the persistent discoloration of the product bleachedwith a sodium chlorite/glacial acetic acid mixture, the residuallignin and hemicellulose were removed. The bleached cellulosebers were washed repeatedly to attain a neutral pH. Thecellulose aerogel was then prepared as follows. The rst stepwas the dissolution of cellulose, in which the solvent mixture ofNaOH/urea/H2O (7.5 : 11.5 : 81 w/w) was precooled to �12 �Cand the desired amount (7 wt%) of the cellulose was dispersedinto the solvent system under vigorous stirring for 4 h at �6 �Cto obtain a transparent cellulose sol. The mixture was then keptfor 12 hours at 50 �C to ensure complete gelation. Secondly, thegels were regenerated in water at room temperature, which werethen frozen to �80 �C for 12 h and dried using a freeze-dryer.
2.3 Preparation and activation of carbon aerogel
The resultant cellulose aerogels were pyrolyzed at 800 �C for 3 hunder owing N2. The obtained carbon aerogels were thenmixed with a KOH solution at KOH/carbon aerogel weight ratioof 3 : 1 under vigorous stirring. Aer stirring for 2 h, the mixturewas then dried at 110 �C to prepare the impregnated sample,
Nanoscale, 2014, 6, 12120–12129 | 12121
Nanoscale Paper
Publ
ishe
d on
29
July
201
4. D
ownl
oade
d by
Uni
vers
ity o
f H
oust
on o
n 07
/03/
2016
17:
28:3
0.
View Article Online
which was then heated to various temperatures under a N2 owfor 2 h. The activation temperatures were varied from 700 �C to900 �C (samples are denoted as K700, K800 and K900). Aerheating to the particular temperatures, the samples were cooledto room-temperature under a continuous N2 ow. The obtainedsamples were washed thoroughly with deionized water in orderto remove the residual chemicals. Finally, the activated carbonaerogels were obtained aer drying the prepared samples at50 �C for 12 h.
2.4 Characterizations of hierarchical porous carbon aerogels
The morphologies and structural properties of the preparedmaterials were characterized by eld emission scanning elec-tron microscopy (FESEM; HITACHI S-4800) and transmissionelectron microscopy (TEM; JEOL JEM2100F) equipped withhigh-resolution imaging features. The scattering and furtherstructural properties were investigated using a Raman-scat-tering (Jobin–Yvon HR 800) spectrometer.
The porous characteristics of the prepared materials wereexamined by N2 adsorption/desorption experiments at 77 Kusing ASAP 2020 V3.02 H. The specic surface area wasmeasured according to the Brunauer–Emmett–Teller (BET)method, and the BJHmethod was used to calculate the pore sizedistribution and pore volumes.
2.5 Electrochemical measurements
To obtain the electrochemical properties of the preparedsamples, both three-electrode congurations and two-electrodecongurations were used. To prepare the electrode used in thethree electrode set-up, according to the conventional optimi-zation methods, the supercapacitor electrodes were made ofdifferent activated carbon aerogels, acetylene black and poly-
Scheme 1 Schematic of the fabrication of highly porous bagasse-derive
12122 | Nanoscale, 2014, 6, 12120–12129
vinylideneuoride (PVDF).37 Prior to preparing the electrode,PVDF was added to N-methyl pyrrolidone (NMP) solution (0.01 gmL�1), and then the carbon aerogel and acetylene black wereadded to the above solution with a carbon aerogel/acetyleneblack/PVDF weight ratio of 8 : 1 : 1. Aer stirring for 12 h, themixture was pressed onto a nickel foil with a size of 1 cm � 1cm. The prepared electrodes were then dried at 60 �C in air for12 h to remove the NMP solution. In a conventional three-electrode cell, a Pt wire and a Ag/AgCl electrode (saturated withKCl (aq)) were used as the counter and reference electrodes,respectively. An aqueous solution of 6 M KOH was used as theelectrolyte. To evaluate the performance of the activated carbonaerogel as two symmetrical electrodes, a two-electrode cong-uration was used.
To prepare the KOH/PVA gel electrolyte, 4.5 g of KOH wasadded to 60 mL of deionized water and 6 g of PVA was thenadded. The mixture was then heated to 85 �C under continuousstirring until the solution became clear. Two electrodes, madeby the above mentioned method were immersed into the KOH/PVA solution for 5 min, keeping the foil without carbon mate-rials above the solution, and then overlaid the two electrodeshead to head until solidied the gel at room temperature.
Cyclic voltammetry (CV) measurements, galvanostaticcharging–discharging (GCD) measurements and electro-chemical impedance spectroscopy (EIS) were performed byusing a CHI 660C Electrochemical Workstation (CH Instru-ments, China).
3. Results and discussion
The main process of preparing hierarchical porous carbonaerogels derived from bagasse is schematically illustrated inScheme 1. The rst step is the dissolution of cellulose bers,
d carbon aerogels with hierarchical pore structure.
This journal is © The Royal Society of Chemistry 2014
Paper Nanoscale
Publ
ishe
d on
29
July
201
4. D
ownl
oade
d by
Uni
vers
ity o
f H
oust
on o
n 07
/03/
2016
17:
28:3
0.
View Article Online
which involved the purication of cellulose from bagasse. Thissolution was then placed for a few hours to form the gel, andkept in water for regeneration. Aer freeze-drying and carbon-ization process, carbon aerogels can be obtained. To furtheroptimize the pore size distribution and prepare hierarchicalporous structures, the obtained carbon aerogels were thenmixed with a KOH solution, aer drying at 110 �C, the mixturewas activated at different temperatures under a N2 ow for 2 h.The resulting mixture was then washed and dried, yieldinghierarchical nanoporous structures. Fig. S1† demonstrates thesystematic and detailed processes for the fabrication of bagasse-derived carbon aerogels.
To examine the general morphologies, all the preparedcarbon aerogel samples were examined by FESEM and resultsare demonstrated in Fig. 1, which depicts the generalmorphologies of the as-synthesized carbon aerogels with andwithout post activation treatments. The observed FESEMimages shown in Fig. 1(a) and (b) clearly reveal that the carbonaerogels possess porous structures, which exhibit large chan-nels with the average diameters in the range of 50–100 mm. It isalso clear from the observed FESEM images that the channelsare well connected with each other, which is promising for theelectrolyte ions diffusion. The morphology of the as-obtainedcellulose aerogels (Fig. S2†) is different from that aer carbon-ization. Before carbonization, cellulose aerogels possess well-shaped macroporous structures, with channel diameters are inthe range of 450 � 50 nm. Interestingly, the channels exhibituniform diameter distribution, which is the typical character-istic of aerogels. Aer carbonization, the uniform porous
Fig. 1 Representative SEM images of the carbon aerogels activated atdifferent temperatures: as-carbonized sample (a and b); and activatedsamples (c and d) K700; (e) K800 and (f) K900.
This journal is © The Royal Society of Chemistry 2014
structures of cellulose aerogels were broken and the macro-pores reassemble together to form large particles and le withvery large cracks, and transferred to porous structure with poresizes in the range of 80 � 20 mm. The high resolution SEMimage of carbonized aerogel conrmed that there were nosecondary pores in the aerogels (Fig. 1(b)). The transformationto such a porous structure with 80 � 20 mm diameters of thepores was caused by the local melting of cellulose and thencarbonization of the cellulose aerogel during the heating andcarbonization process. It is well known that the glass transitiontemperature of dried cellulose is �200 �C,38 and when thesample was heated to �200 �C, the melting process of thecellulose causes the connection of the cellulose molecules, andform complete surface in some part of the aerogels due to thehigh surface tension. Because of the lack of enough mass ofcellulose of the porous aerogels, the aerogels cannot connect tobe an entire bulk structures, and the network is broken andforms porous structure with very large pores. Fig. 1(c)–(f)demonstrate the typical FESEM images of the carbon aerogelsaer activation. As shown in the images, all three samples showthe presence of open and interconnected pores, forming athree-dimensional macroporous framework. When the samplewas activated at 700 �C, the carbon aerogels displayed unin-terrupted 3D networks constructed with carbon walls from thecarbonization of biopolymers of cellulose, and the pore sizeswere less than 2 mm, which are much smaller than those in theas-carbonized sample (Fig. 1(c) and (d)). However, when theactivation temperatures were increased to 800 �C (Fig. 1(e)) and900 �C (Fig. 1(f)), the three-dimensional macroporous frame-works were unchanged.
The detailed morphological and structural properties of theporous structure of carbon aerogels were further observed usingTEM. Fig. 2 exhibits the typical TEM and HRTEM images ofcarbon aerogels activated at 700 �C. The obtained TEM imagesare fully consistent with the FESEM observations (Fig. 2(a)),which demonstrates that the channel wall of porous carbonaerogels are composed of small carbon particles, about 20–40nm in diameter, which connect together to form a mesoporousstructure (Fig. 1(c) and (d)). The high resolution TEM image ofthe selected area shown by square bracket in Fig. 2(a) exhibitsthat there are large number of micropores in the carbon parti-cles (Fig. 2(b)). The diameter of the pores are uniform, and thepore sizes are in the range of 2 � 1 nm (Fig. 2(b)). High-reso-lution image of the carbon aerogels also reveals that the aero-gels have a highly disordered slot hole-like porous carbonstructure with a large fraction of micro- and meso-pores
Fig. 2 TEM (a) and (b) and HRTEM (c) images of the K700 sample.
Nanoscale, 2014, 6, 12120–12129 | 12123
Fig. 3 (a–d) CV curves of the porous carbon aerogels electrodes at varied potential scan rates in 6 M KOH aqueous solution ((a) K700; (b) K800;(c) K900; (d) without activation); (e) CV curves of the four electrodes at the potential scan rates of 2 mV s�1; (f) specific capacitances of the fourelectrodes as a function of scan rate derived from (a–d).
Nanoscale Paper
Publ
ishe
d on
29
July
201
4. D
ownl
oade
d by
Uni
vers
ity o
f H
oust
on o
n 07
/03/
2016
17:
28:3
0.
View Article Online
(Fig. 2(c)), which facilitates ion transport during the charging–discharging process. Fig. 2(c) also shows that there are someordered lattice fringes in the particles, which is from thegraphitization of the carbon aerogels to some extent. TheseTEM images substantially indicate that the activation processetched the carbon aerogels and has signicantly tuned porousstructure. Thus, from all the morphological characterizations,one can gure out the microstructures of KOH activated carbonaerogels. The KOH-activated carbon aerogels have a typicalhierarchical porous structure. The rst level is very large mac-ropores, which are larger than 50 nm in diameter, the secondlevel is the mesopores in the channel wall of the large pores,which is composed of several connections of carbon nano-particles and is about 2–50 nm in diameter, and the third levelis micropores in the carbon nanoparticles, which composed ofgraphite-like structures and is smaller than 2 nm.
To evaluate the electrochemical performance of the porouscarbon aerogels, cyclic voltammetry (CV) tests were carried outin a three-electrode conguration in an aqueous solution of 6 MKOH, and the results are depicted in Fig. 3. Fig. 3(a–d) presentthe CV curves of the four samples at various scanning rates inthe potential ranges from�1 to 0 V. As shown in the gures, theCV curves of three-electrode system exhibit an approximatelyrectangular shape, even at the scan rate of 200 mV s�1, indi-cating a near-ideal capacitive behavior with good rate capability.The slight deviation from the ideal rectangular shape may bedue to the ESR (Equivalent Series Resistance) and the pore sizedistribution of samples. Fig. 3(e) is the CV curve of the fourelectrodes at the potential scan rates of 2 mV s�1. From Fig. 3(e),one can see that the current density of K700 is the highestcompared with the other samples, and all samples displayed amore rectangular shape at a low scan rate. Fig. 3(f) summarizes
12124 | Nanoscale, 2014, 6, 12120–12129
the gravimetric specic capacitance of the samples according tothe CV testing. The prepared carbon aerogels electrode exhibi-ted the lowest specic capacitance and only 88 F g�1 wasobtained even at the lowest scan rate of 2 mV s�1. The samplesof K800 and K900 displayed an increased specic capacitance of118.8 and 133.6 F g�1 at 2 mV s�1, respectively. Impressively, thespecic capacitance of the K700 electrode reached 268.4 F g�1 ata scan rate of 2 mV s�1. These values are much higher thanthose of the graphene-based macro-mesoporous carbon (226 Fg�1 at 1 mV s�1), and two-dimensional graphene-based meso-porous carbon sheets (148 F g�1).39,40 Although capacitance losscan be observed when increasing the scan rate from 2 mV s�1 to200 mV s�1 for all the carbon samples, the K700 electrode stilldemonstrates a specic capacitance of as high as 202.1 F g�1.
The Nyquist plot of impedance behavior as a function offrequency is oen used to evaluate the frequency response ofsupercapacitors. Fig. S3† shows the impedance curves obtainedfor different aerogel samples at different frequencies rangingfrom 0.01 Hz to 100 kHz, and a magnied view of the high-frequency region is shown in the inset. Interestingly, it wasobserved that the K700 sample displayed nearly ideal capacitivebehavior with a vertical slope at low frequencies. The ESR(Equivalent Series Resistance) of the samples, i.e., K700, K800,K900, and without activation were 0.67 U, 0.74 U, 0.75 U, and0.93 U, respectively. Generally, ESR contains three differentcomponents, the contact resistance, solution resistance and thebulk resistance of the electrode. The K700 sample has the lowestESR value, indicating that the contact resistance and bulkresistance is very low. This may be the reason why the currentdensity of the K700 sample shown in the CV proles is higherthan that of K800 and K900. ESR data determines the rate thatthe supercapacitor can be charged and discharged, and it is a
This journal is © The Royal Society of Chemistry 2014
Paper Nanoscale
Publ
ishe
d on
29
July
201
4. D
ownl
oade
d by
Uni
vers
ity o
f H
oust
on o
n 07
/03/
2016
17:
28:3
0.
View Article Online
very important factor to determine the power density of asupercapacitor.41 Hence, it is believed that the electrochemicalperformance of as-prepared carbon aerogels can be improveddramatically by an activation process.
The enhancement of the specic capacitance of the activatedsamples is suggested to be related to the degree of graphitiza-tion and the increase in surface area and the manipulation ofthe pore size and distribution, which is further demonstrated byRaman scattering measurements and N2 adsorption/desorptionexperiments.
Raman scattering measurements were carried out tounderstand the structural evolution of the modied carbonaerogels activated at different temperatures. Fig. 4(a) showsthe Raman spectra of all the samples prepared at variousactivation temperatures. For comparison, the Raman scat-tering result for carbon without activation is also displayed inthe spectra. All the samples exhibited two distinct peakslocated at 1350 cm�1 and 1582 cm�1, which were indexed tothe D band and G band vibrations, respectively. The D bandreects the disordered carbon, while the G band is caused bythe vibration of ordered carbon.42,43 It was found that boththe D and G bands of carbon without activation displayed abroad shape and low intensity, indicating the possible highcontent of amorphous carbon. Fig. 4(b) shows the full widthat the half maximum (FWHM) values of the D-band and G-band of the samples, which reveals that the FWHM values ofthe D- and G-bands of carbon aer activation are lower thanthose without activation, indicating a sharper feature ofpeaks. The D- and G-bands of activated carbon aerogels showsharper and high intensity peaks, which indicate theimproved graphite degree aer activation in the samples.This conclusion is further demonstrated by the intensity
Fig. 4 (a) Raman spectra of the pristine and activated samples; (b) full wintegrated intensities of the D- and G-bands (IG/ID); (d) nitrogen adsorptio(calculated by using the BJH model); (f) detail view of (e).
This journal is © The Royal Society of Chemistry 2014
ratio (ID/IG) of the D- and G-bands, which reects the ratio ofdisordered carbon and ordered carbon in the carbon aerogels(Fig. 4(c)). The values of ID/IG for activated carbon aerogelsdecreased to less than 0.8, indicating a very high content ofordered carbon in the activated carbon aerogel samples. Theincreased content of ordered carbon indicates that the postactivation process can not only be expected to create pores,but also enhance the graphite degree. The slight increase inID/IG aer activation at 800 and 900 �C indicates that thedegree of graphitization in the sample activated at 700 �C isthe highest while graphitization decreases with increase inthe activation temperature. The improved graphite degree ofactivated carbon aerogels can signicantly increase theelectric conductivity, which should decrease the equivalentseries resistance of supercapacitors. This is consistent withthe results of the impedance behaviour. The FTIR spectra ofcarbon aerogels activated at different temperatures are alsoshown in Fig. S4.‡ There is no obvious peak located at thespectra, indicating that the prepared carbon aerogels werevery clean and there are no other elements or residualorganic compounds in the samples except for C. Themechanical property of the carbon aerogels was also perfect.Fig. S5(a)‡ shows a photograph of a carbon aerogel, sup-porting a mass of 50 g with no deformation. In addition, asshown in Fig. S5(b),‡ we placed a carbon aerogel on a spiderweb to illustrate the light weight originating from highporosity of the sample.
Fig. 4(d) exhibits the typical N2 adsorption/desorptionisotherms of all samples. As illustrated in Fig. 4(d), the type-IV nitrogen sorption isotherms for the samples suggest theexistence of different pore sizes from micro- to macro-pores.Major adsorption of the samples occurs at a low relative
idth at half maximum (FWHM) of the D- and G-band peaks; (c) ratio ofn/desorption isotherms of the samples; (e) pore size distribution for N2
Nanoscale, 2014, 6, 12120–12129 | 12125
Nanoscale Paper
Publ
ishe
d on
29
July
201
4. D
ownl
oade
d by
Uni
vers
ity o
f H
oust
on o
n 07
/03/
2016
17:
28:3
0.
View Article Online
pressure of less than 0.1, giving rise to an almost horizontalplateau at higher relative pressures. These curves indicatethat the samples possess high microporosity.28 Furthermore,the hysteresis between the adsorption and desorptionbranches (at 0.5–1.0 P/P0) of the samples also suggests theexistence of different types of pores with no pore-blockingeffect during desorption. This result can be demonstrated bythe pore size distribution calculated from the BJH model, asshown in Fig. 4(e) and (f). Fig. 4(e) and (f) demonstrate thatall the three samples have both mesopores (2–50 nm) andmicropores (<2 nm). The mesopores have a pore size of �2nm for the three samples, which is fully consistent with theobserved HRTEM image, and the sizes of the micropores aretypically centered at 0.6 nm. Table 1 summarizes the specicsurface area and pore volumes of samples. Aer activationwith KOH etching, the specic surface area is amplied to2064.79 m2 g�1. The specic surface area and total porevolume of K700 is lower than that of K900, while the micro-pore volume of the K700 sample is typically the highest withthe proper mesopore volume. The micropore volume of K900sample is the lowest value with the highest surface area andmesopore volume, indicating that the increase in specicsurface area cannot always positively enhance the speciccapacitance.
To further investigate the relationship between porousstructure and electrochemical performance of electrode mate-rials, correlation analysis is conducted based on the results ofN2 adsorption/desorption and cyclic voltammetry (CV) tests.The correlation, indicating the related degree of variable, can becalculated using the following equation:
Correlation ðx; yÞ ¼nX
ðxyÞ �X
xX
yffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi��nX
X 2 ��X
X�2���nX
Y 2 ��X
Y�2��s (1)
where x and y are variables, and n is the number of x or y. Thecalculated Correlation (X5, X3) and Correlation (X5, X4) were85% and 61%, respectively, suggesting that the speciccapacitance is highly relevant with a micropore volume andonly moderately associated with the mesopore volume. Theincrease in micropore volume can enhance the speciccapacitance of the electrode materials effectively, whileincreasing the mesopore volume had little effect. TheCorrelation (X6, X3) and Correlation (X6, X4) were calculated
Table 1 Physical properties and specific capacitance obtained from CV scapacitance obtained at scan rate of 2mV s�1. The rate capability (R) is defiat 2 mV s�1
Sample IDSpecic surfacearea cm2 g�1
Total pore volumecm3 g�1
Micropore vocm3 g�1
Variable X1 X2 X3
K700 1892.4 0.86 0.58K800 1641.6 0.80 0.46K900 2064.8 1.41 0.28
12126 | Nanoscale, 2014, 6, 12120–12129
to be 92% and 99%, respectively, indicating that rate capa-bility is mainly associated with the mesoporous volume of thematerials. The correlation analysis results showed that themicropores and mesopores play different roles on the elec-trochemical properties of the electrode materials and theenergy storage capacity is determined by the microporevolume, while the rate capability is mainly decided by mes-opore volume. When electrolyte ions immerse into the pores,the distance between the electrolyte ions and micropore wallson both sides are the shortest, indicating the strongestelectrostatic adsorption, which leads to the best energystorage capacity. However, mesopores mainly store electro-lyte to shorten the electrolyte ions diffusion distance, and atthe same time, can supply electrolyte ions in the process ofcharging and discharging quickly. For ideal ECs electrodematerials, it should have a hierarchical porous structure. TheK700 sample typically has three types of pores, i.e., micropore(<2 nm), mesopores (2–50 nm) and macropores (>50 nm),indicating a hierarchical porous structure, which is consis-tent with the observed FESEM and TEM results.
Moreover, electrochemical properties of a porous carbonelectrode material are not always directly proportional to itsspecic surface area. Aside from the specic surface area,effective pore size distribution, good electrical conductivity (toimprove the power density) also affects the performance ofsupercapacitors. Combining all these factors, one can under-stand why the K700 sample shows the best electrochemicalperformance.
The special hierarchical porous structure is attributed to thereaction between KOH and carbon in the aerogels during theactivation process.44,45 As mentioned in the experimentalsection, the activation process occurred at a high temperaturetreatment of as-synthesized carbon aerogels and KOH pene-trated in the pores of the carbon aerogels. At high temperatures,the reaction between carbon on the surface of the wall of thechannels of the aerogels and KOH can occur according toformula (2).
4 KOH + C / K2CO3 + K2O + 2H2 (2)
This reaction preferentially occurs on the surface of thewall of the channel in as-synthesized carbon aerogels, andthe surface can be corroded to form nanopores. From theRaman scattering results, it is clear that the as-carbonizedaerogels possess a certain extent of graphitization. Normally,
canning at 2 mV s�1 in the three electrode testing. Cs is defined as thened as the ratio of the capacitance value obtained at 100mV s�1 to that
lume Mesopore volumecm3 g�1
Specic capacitanceCs F g�1
Ratecapability R
X4 X5 X60.23 268.4 82.5%0.27 133.6 82.3%1.0 118.8 57.4%
This journal is © The Royal Society of Chemistry 2014
Paper Nanoscale
Publ
ishe
d on
29
July
201
4. D
ownl
oade
d by
Uni
vers
ity o
f H
oust
on o
n 07
/03/
2016
17:
28:3
0.
View Article Online
the low graphitized part can be corroded rst, and leavehigher graphitized nanoparticles. This is why the hierar-chical porous structure can be formed aer the activationprocess, and why the graphite peak of the activated sample ismuch higher than the unactivated carbon aerogels. Byincreasing the activation temperature, the reaction betweencarbon and KOH becomes more violent and KOH can reactwith the more graphitized area in the carbon particle, whichcauses a decrease in the peaks of graphite in the Ramanspectra. It is known that the conductivity of the carbon aer-ogel is related to the degree of graphitization of the sample.Therefore, the conductivity of the K700 sample should be thehighest because of the highest graphitization degree in thesample activated by KOH at 700 �C, which is fully consistentwith the ESR results.
As is well known, the electrolyte plays an important role insupercapacitors, the potential range and ionic conductivityof the electrolyte can inuence supercapacitor behaviorsignicantly. Many reports are based on liquid electrolytesbecause of their high ionic conductivity. However, there arealso many existing disadvantageous in the liquid electrolytessuch as low potential range, leakage, corrosion and explo-sions. Therefore, replacing liquid electrolytes with solidelectrolytes in the supercapacitor has been explored by manyresearchers. The solid electrolytes have several advantagesover liquid ones, which includes easy handling, increasedsafety, exibility in packing because of the solid electrolytesetc.46 In order to assess the reliable electrochemical perfor-mance of the carbon aerogels, a two-electrode solid statesymmetrical supercapacitor device was assembled to char-acterize the electrochemical storage capability of the K700sample in the KOH/PVA solid electrolyte. Fig. 5(a) shows a
Fig. 5 (a) Schematic of the fabricated hierarchical porous carbon supercasample as a function of the scan rate derived from (b), (d) charge–dischspecific capacitance of the K700 sample as a function of the current de
This journal is © The Royal Society of Chemistry 2014
schematic of an all-solid-state symmetric supercapacitorusing hierarchical porous carbon supercapacitor. The porouscarbon based cell was constructed by sandwiching a KOH/PVA gel electrolyte with two pieces of as prepared carbonelectrodes via a facile and effective process, as discussed inthe experimental section. The CV curves at various potentialscan rates ranging from 2 to 200 mV s�1 and galvanostaticcharging–discharging curves as a function of the currentdensity varied from 0.5 to 10 A g�1 were investigated,respectively. As shown in Fig. 5(b), the CV curves at variousscan rates are rectangular-like in shape and symmetric from0 to 1 V, indicating a near-ideal capacitive behavior with goodrate performance. The highest specic capacitance reached189 F g�1 at 2 mV s�1 and decrease to 99.8 F g�1 at 200 mV s�1
(Fig. 5(c)). The galvanostatic charge–discharge curves(Fig. 5(d)) at various current densities showed goodsymmetry, nearly linear slopes and quick current–voltageresponse, which conrms the good supercapacitor charac-teristics of the K700 sample. The calculated specic capaci-tance from the discharge curves are shown in Fig. 5(e). Thespecic capacitance at 0.5 A g�1 is about 142.1 F g�1, which ismuch higher than the electrode prepared from bananasusing KOH activation (65 F g�1).47 An IR drop of 0.02 V can beobtained at 0.5 A g�1, indicating a negligible leakage currentof the sample. In addition, the CV curves of a two-electrodesupercapacitor using a liquid electrolyte at various potentialscan rates ranging from 2 to 200 mV s�1 were also investi-gated. As shown in Fig. S6,† the CV curves at various scanrates are rectangular-like in shape, indicating near-idealcapacitive behavior with good rate capability.
The electrochemical impedance spectra were also con-ducted from 0.01 Hz to 100 kHz for the solid and liquid
pacitor; (b) CV curves of K700 sample; (c) specific capacitance of K700arge curves of the K700 sample at different current densities, and (e)nsity derived from (d).
Nanoscale, 2014, 6, 12120–12129 | 12127
Nanoscale Paper
Publ
ishe
d on
29
July
201
4. D
ownl
oade
d by
Uni
vers
ity o
f H
oust
on o
n 07
/03/
2016
17:
28:3
0.
View Article Online
electrolyte supercapacitors, as shown in Fig. S7.† It was foundthat the ESR of K700 using solid electrolyte was about 0.76 U,which is similar to that of the aqueous electrolyte (�0.72 U).While the charge transfer resistance showed signicantdifference and this value for solid electrolyte was about 5.4 U,which is much higher than that for an aqueous electrolyte(�2.6 U). At the low frequency range, a nearly vertical line canbe observed, indicating the ideal supercapacitve features ofthe symmetrical supercapacitor. The results of EIS of thesymmetrical supercapacitor indicated that K700 sample hadthe best supercapacitor performance.
To assess the cycling stability of K700 for supercapacitorapplications, the galvanostatic charge–discharge experimentwas performed for 5000 cycles. Fig. 6(a) shows the cyclingstability of the symmetric electrodes at a constant charge anddischarge current density of 1 A g�1 for 5000 cycles. Thecoulombic efficiency retained about 93.9% with a smalldecrease from 132 to 123.9 F g�1, indicating excellent cyclabilityof the symmetric cell based on the hierarchical porous carbonaerogels. The power density and energy density of thesymmetric supercapacitor is shown in a Ragone plot (Fig. 6(b)).Benetting from a high specic capacitance of up to 142.1 F g�1
(obtained at a current density of 0.5 A g�1) and an operatingvoltage of 1 V, the cell displayed an energy density of�19.74Whkg�1 and a power density of 0.5 kW kg�1. Although the elec-trochemical performance of the K700 sample was notoutstanding compared with a few individual cases preparedfrom the carbonization of mixtures of organic reagents, therequirement of highly toxic organic reagents in these othercases can discourage their use in electrode preparation. Thus,the presented work is highly promising if one considers theavailability of bagasse as a biomass waste and the ease of thepreparation method.
4. Conclusions
In summary, a supercapacitor electrode material with excellentperformance has been achieved by the carbonization of cellu-lose aerogels using bagasse, which is an industrial waste, andactivation with KOH. The resulting carbon aerogels have hier-archical macro-, meso-, and micro- porous structures andpossess extremely high specic surface area of 1892.4 m2 g�1.The specic capacitance calculated from the charge–discharge
Fig. 6 (a) Cycle performance and coulombic efficiency at 1 A g�1 over5000 cycles and (b) Ragone plots of K700 sample.
12128 | Nanoscale, 2014, 6, 12120–12129
experiments using a symmetric supercapacitor was about 142.1F g�1 at 0.5 A g�1 with excellent capacitance retention at�93.86% over 5000 cycles, which translates to the highestenergy density of �19.74 W h kg�1 and a power density of �0.5kW kg�1. The obtained results demonstrate the possibility forthe large scale production of hierarchical porous carbons bychoosing low-cost biomass waste, i.e. bagasse, as a rawmaterial.The presented synthesis method also opens a new approach forthe large-scale production of carbon electrodes that are espe-cially suitable for supercapacitors.
Acknowledgements
The authors are thankful for fundings from the NationalNatural Science Foundation of China (Grant no. 51372142), theInnovation Research Group (IRG: 51321091) and the “100Talents Program” of the Chinese Academy of Sciences. Thanksfor the support from the “thousands talents” program forpioneer researcher and his innovation team, China.
References
1 Z. Wu, K. Parvez, X. Feng and K. Mullen, Nat. Commun., 2013,4, 2487.
2 P. Simon and Y. Gogotsi, Nat. Mater., 2008, 7, 845–854.3 P. Sharma and T. S. Bhatti, Energy Convers. Manage., 2010,51, 2901–2912.
4 A. Stein, Z. Wang andM. A. Fierke, Adv. Mater., 2009, 21, 265–293.
5 Y. Gogotsi and P. Simon, Sci. Magazine, 2011, 334, 917–918.6 L. Mao, K. Zhang, H. S. O. Chan and J. Wu, J. Mater. Chem.,2012, 22, 1845–1851.
7 J. Sun, W. Li, B. Zhang, G. Li, L. Jiang, Z. Chen, R. Zou andJ. Hu, Nano Energy, 2014, 4, 56–64.
8 R. Ramya, R. Sivasubramanian and M. V. Sangaranarayanan,Electrochim. Acta, 2013, 101, 109–129.
9 K. Wang, H. Wu, Y. Meng and Z. Wei, Small, 2014, 10, 14–31.10 P. Sharma and T. S. Bhatti, Energy Convers. Manage., 2010,
51, 2901–2912.11 Z. Niu, H. Dong, B. Zhu, J. Li, H. H. Hng, W. Zhou, X. Chen
and S. Xie, Adv. Mater., 2013, 25, 1058–1064.12 J. Li, X. Wang, Y. Wang, Q. Huang, C. Dai, S. Gamboa and
P. J. Sebastian, J. Non-Cryst. Solids, 2008, 354, 19–24.13 Z. Niu, L. Zhang, L. Liu, B. Zhu, H. Dong and X. Chen, Adv.
Mater., 2013, 25, 4035–4042.14 Q. Cheng, J. Tang, J. Ma, H. Zhang, N. Shinya and L. C. Qin, J.
Phys. Chem. C, 2011, 115, 23584–23590.15 J. Yan, T. Wei, Z. Fan, W. Qian, M. Zhang, X. Shen and
F. Wei, J. Power Sources, 2010, 195, 3041–3045.16 Z. Niu, P. Luan, Q. Shao, H. Dong, J. Li, J. Chen, D. Zhao,
L. Cai, W. Zhou, X. Chen and S. Xie, Energy Environ. Sci.,2012, 5, 8726–8733.
17 C. Liu, Z. Yu, D. Neff, A. Zhamu and B. Z. Jang, Nano Lett.,2010, 10, 4863–4868.
18 C. X. Guo and C. M. Li, Energy Environ. Sci., 2011, 4, 4504–4507.
This journal is © The Royal Society of Chemistry 2014
Paper Nanoscale
Publ
ishe
d on
29
July
201
4. D
ownl
oade
d by
Uni
vers
ity o
f H
oust
on o
n 07
/03/
2016
17:
28:3
0.
View Article Online
19 Z. Niu, J. Chen, H. H. Hng, J. Ma and X. Chen, Adv. Mater.,2012, 24, 4144–4150.
20 B. Subhadra, Energy Sci. Technol., 2013, 5, 36–49.21 A. A. White, M. S. Platz, D. M. Aruguete, S. L. Jones,
L. D. Madsen and R. D. Wesson, ACS Sustainable Chem.Eng., 2013, 1, 871–877.
22 H. Wang, E. Zhu, J. Yang, P. Zhou, D. Sun and W. Tang, J.Phys. Chem. C, 2012, 116, 13013–13019.
23 L. Wei, M. Sevilla, A. B. Fuertes, R. Mokaya and G. Yushin,Adv. Energy Mater., 2011, 1, 356–361.
24 H. Sun, L. Cao and L. Lu, Energy Environ. Sci., 2012, 5, 6206–6213.
25 W. Sia, J. Zhou, S. Zhang, S. Li, W. Xing and S. Zhuo,Electrochim. Acta, 2013, 107, 397–405.
26 Y. Liang, D. Wu and R. Fu, Sci. Rep., 2013, 3, 1119.27 O. Inganas and S. Admassie, Adv. Mater., 2013, 26, 830–848.28 Q. Wang, Q. Cao, X. Wang, B. Jing, H. Kuang and L. Zhou, J.
Power Sources, 2013, 225, 101–107.29 Y. Lv, L. Gan, M. Liu, W. Xiong, Z. Xu, D. Zhu and
D. S. Wright, J. Power Sources, 2012, 209, 152–157.30 E. Raymundo-Pinero, F. Leroux and F. Beguin, Adv. Mater.,
2006, 18, 1877–1882.31 Z. Gui, H. Zhu, E. Gillette, X. Han, G. W. Rubloff, L. Hu and
S. B. Lee, ACS Nano, 2013, 7, 6037–6046.32 Y. R. Kang, Y. L. Li, F. Hou, Y. Y. Wen and D. Su, Nanoscale,
2012, 4, 3248–3253.33 H. Zhu, Z. Jia, Y. Chen, N. Weadock, J. Wan, O. Vaaland,
X. Han, T. Li and L. Hu, Nano Lett., 2013, 13, 3093–3100.34 Y. Zhu, S. Murali, M. D. Stoller, K. J. Ganesh, W. Cai,
P. J. Ferreira, A. Pirkle, R. M. Wallace, K. A. Cychosz,
This journal is © The Royal Society of Chemistry 2014
M. Thommes, D. Su, E. A. Stach and R. S. Ruoff, Science,2011, 332, 1537–1541.
35 M. Zhi, C. Xiang, J. Li, M. Li and N. Wu, Nanoscale, 2013, 5,72–88.
36 D. Bhattacharya, L. T. Germinario and W. T. Winter,Carbohydr. Polym., 2008, 73, 371–377.
37 M. Yang, Y. Zhong, L. Su, J. Wei and Z. Zhou, Chem.–Eur. J.,2014, 20, 1–9.
38 L. Szczesniak, A. Rachocki and J. Tritt-Goc, Cellulose, 2008,15, 445–451.
39 Z. S. Wu, Y. Sun, Y. Z. Tan, S. Yang, X. Feng and K. Mullen, J.Am. Oil Chem. Soc., 2012, 134, 19532–19535.
40 S. Yang, X. Feng, L. Wang, K. Tang, J. Maier and K. Mullen,Angew. Chem., Int. Ed., 2010, 49, 4795–4799.
41 Q. Cheng, J. Tang, J. Ma, H. Zhang, N. Shinya and L. C. Qin,Carbon, 2011, 49, 2917–2925.
42 A. Janes, H. Kurig and E. Lust, Carbon, 2007, 45, 1226–1233.43 F. Bonhomme, J. C. Lasseguesz and L. Servant, J.
Electrochem. Soc., 2001, 148, 450–458.44 G. G. Stavropoulos, Fuel Process. Technol., 2005, 86, 1165–
1173.45 H. Wang, Z. Li, Y. Huang, Q. Li and X. Wang, J. Mater. Chem.,
2010, 20, 3883–3889.46 G. Ma, J. Li, K. Sun, H. Peng, J. Mu and Z. Lei, J. Power
Sources, 2014, 256, 281–287.47 V. Subramanian, C. Luo, A. M. Stephan, K. S. Nahm,
S. Thomas and B. Wei, J. Phys. Chem. C, 2007, 111, 7527–7531.
Nanoscale, 2014, 6, 12120–12129 | 12129