Hard@Soft Integrated Morning Glory Like Porous Carbon as aCathode for a High-Energy Lithium Ion CapacitorDong Yan, Shu-Hua Li, Li-Ping Guo, Xiao-Ling Dong, Zhi-Yuan Chen, and Wen-Cui Li*
State Key Laboratory of Fine Chemicals, School of Chemical Engineering, Dalian University of Technology, Dalian 116024, People’sRepublic of China
*S Supporting Information
ABSTRACT: A lithium ion capacitor (LIC) is a hybrid energystorage device that combines the energy storage mechanism oflithium ion batteries and supercapacitors and presents theircomplementary features. However, imbalances of the capacity andkinetics between cathode and anode still remain challenges. Herein,to address the issue of low capacity in the cathode, we constructed amorning glory like porous material crimped from an openingsandwich structure consisting of a hard carbon inner layer and softcarbon outer layer, by coating graphene oxide containing flakyphenolic resin with coal tar pitch followed by carbonization andactivation. Both the hard@soft carbon integrated design and thereduced graphene oxide network cocontributed to a favorableelectrical conductivity and a developed microporosity. To deal withthe sluggish kinetics limitation of the anode, a structure-optimized MnO@C electrode prepared by pore creation of CO2 froman MnCO3@polydopamine precursor was chosen due to its excellent rate performance. The LIC with this configuration candeliver a maximum energy density of 117.6 Wh kg−1 and an favorable capacity retention of 76% after 3000 cycles at 2050 Wkg−1. Furthermore, the energy density of 27.8 Wh kg−1 can still be delivered even at a high power density of 10250 W kg−1. Ingeneral, this well-designed hybrid capacitor has a potential to make up the gap between lithium ion batteries andsupercapacitors.
KEYWORDS: lithium ion capacitor, coal tar pitch, porous carbon, MnO, three-electrode test
■ INTRODUCTIONA lithium ion capacitor (LIC) is a kind of hybrid energystorage device with a relatively high energy density, whichcombines the energy storage mechanisms of lithium ionbatteries and supercapacitors and has been brought into focusrecently.1 Its structure usually consists of a capacitor-typecathode, an insertion-type anode, and commercial lithium ionbattery electrolyte. During the charging process, lithium ionsare inserted into the anode material; meanwhile the anions areadsorbed on the surface of the cathode. On the basis of eq 1,this asymmetrical design gains a higher voltage window (V1and V2) and specific capacitance (CT), which are the sources ofhigh energy density.2
E C V12 T
2=(1)
C C C1 1 1
T a c= +
(2)
VV V
21 2=
+(3)
E, CT, Ca, Cc, V, V1, and V2 are the energy density, totalcapacitance, anode and cathode capacitance, operating voltage,
and lower and upper limits of the voltage range, respectively.Due to the differences in both capacity and kinetics, thecathode and anode influence the performance of the full celldifferently. On the basis of the bucket effect of the wholedevice, the lower capacitance of the cathode (Cc ≪ Ca) is theprimary limiting factor for energy density. Similarly, the farlower rate of the anode limits the power density of the fullcell.3 Therefore, reasonable designing and matching for thecathode and anode materials are the main routes to realize agreat enhancement in such an asymmetric capacitor.4
The configuration of a high-surface-area porous carboncathode5 and the a conversion-type transition-metal oxideanode such as MnO,6 MnFe2O4,
7 Fe3O4@NC,8 Fe3O4−graphene hybrid,9,10 etc. has been regarded as a promisingmatch for an LIC. However, challenges still remain in thesesystems for both the cathode and anode. For a porous carboncathode, in order to improve the energy density of the LIC,high surface area is the main requirement for the capacitance.However, the surface area of porous carbon usually comesfrom the contribution of micropores, which are not fully
Received: October 5, 2018Accepted: November 26, 2018Published: November 26, 2018
favorable for the rapid diffusion of electrolyte ions. In this case,a good rate capability cannot be assured, especially in theorganic electrolyte of an LIC. Thus, a hierarchical porestructure (including macro-, meso-, and micropores) has greatpotential to afford a high capacitance by expanding the surfacearea, as well as possesses developed diffusion pathways for agood rate capability.11 Nevertheless, a highly developed porousstructure always means low electrical conductivity in porouscarbon, which also needs to be considered in a capacitivematerial. In terms of a transition metal oxide anode, to endowa perfect rate performance, effort should be made to design arational nanostructure with highly conductive matrix and richlithiation active sites.7
On the basis of the carbon chemistry, soft carbon with well-graphitized characteristics generally presents good conductiv-ity; hard carbon containing disordered graphitic microcrystalseasily forms developed porosity. Coal tar pitch, a flexiblebyproduct of the coal chemical industry, is an ideal soft carbonprecursor due to its high degree of aromatic ring structure andhigh carbon residue.12 In this paper, by keeping in mind thecontradiction between high porosity and electrical conductivityof the carbon electrode material, we constructed a morningglory like porous carbon by coating graphene oxide containingflaky phenolic resin with coal tar pitch followed by carbon-ization and activation. The porous carbon with a surface areaof higher than 2000 m2/g exhibits an excellent specificcapacitance and rate performance. Moreover, for the sake ofenergy density in the full cell, a porous MnO@C was chosen asthe anode due to its high capacity and low redox potential(<0.5 V vs Li/Li+). By pore creation of CO2 from theMnCO3@polydopamine precursor, the porous MnO@Celectrode with a high surface charge storage of ∼34% canalleviate the kinetics imbalance between the cathode andanode. This hybrid lithium ion capacitor can deliver a highenergy density up to 117.6 Wh kg−1 at 410 W kg−1 and asuperior cycling life.
■ EXPERIMENTAL SECTIONSynthesis of Morning Glory Like Porous Carbon. Herein, coal
tar pitch and phenolic resin were used as soft and hard carbonsources, respectively. First, resorcinol was polymerized with form-aldehyde around graphene oxide (GO) nanosheets catalyzed byasparagine in aqueous solution at 90 °C for 4 h, according to ourprevious work.13 Afterward, the coal tar pitch (softening point 50 °C)was preliminarily mixed with the obtained phenolic resin, and thenthe same quantity of K2CO3 was added to the carbon precursor as anactivating agent with a subequent 2 h of ball-milling. The collectedmixture was subject to prepolycondensation in a muffle oven at 350°C for 1 h, and then the same quantity of K2CO3 as above was addedagain for an extra 2 h of ball-milling (the total mass ratio of K2CO3 tocarbon source is 2:1). Finally, the mixture was transferred to a tubefurnace for the carbonization and activation process at 800 °C for 2 hunder an Ar atmosphere. After repetitive washing with diluted HCland deionized water and drying at 80 °C, the porous carbon wascollected and correspondingly named PC-x, where x represents themass percentage of coal tar pitch in the carbon source and was variedas 85, 75, and 50 wt %. For control samples, PC-T and PC-R wereprepared only from the coal tar pitch and flaky phenolic resin,respectively. PC-75* was a reference prepared with 75 wt % coal tarpitch but without the prepolycondensation step.
Synthesis of Porous MnO@C Composite. MnO@C wassynthesized via a hydrothermal process as reported before.14 Briefly,KMnO4 and glucose were dissolved in 80 mL of deionized water andhydrothermally treated at 160 °C for 4 h. After centrifugation andwashing with deionized water, the obtained product MnCO3 wascoated by polydopamine in solution at 30 °C and further calcinated at500 °C under an argon flow to form MnO@C composite.
Materials Characterization. Scanning electron microscopy(SEM) and transmission electron microscopy (TEM) investigationswere respectively carried out with an FEI Nova NanoSEM 450instrument and an FEI Tecnai F30 instrument. Nitrogen adsorptionisotherms were measured at 77.4 K with an ASAP 2020 sorptionanalyzer (Micromeritics Instruments, USA). The Brunauer−Em-mett−Teller (BET) method was used to calculate the specific surfacearea (SBET). Pore size distributions (PSDs) were calculated from theslit pore geometry. X-ray diffraction (XRD) measurements were takenon a PANalytical X’Pert3 powder diffractometer. Raman spectra were
Figure 1. (a, b) SEM images of PC-75. (c, d) TEM image and HRTEM image of PC-75, respectively.
ACS Applied Materials & Interfaces Research Article
taken with a Raman spectrometer with 532 nm laser excitation (DXRSmart Raman).Electrochemical Characterization. To prepare the cathode, the
as-synthesized porous carbon was mixed with polytetrafluoroethylene(PTFE) and acetylene black in a ratio of 8:1:1 in ethanol to form auniform mixture. The mixture was rolled into a film and punched into12 mm diameter electrodes (2−3 mg). The electrode was pressedover a stainless-steel mesh (15 mm diameter) and dried at 100 °Cunder vacuum overnight. For the anode side, MnO@C was mixedwith conductive carbon black (Super-P) and binder (LA133 andCMC) in a ratio of 8:1:1 with DI water. The slurry was coated ontocopper foil by a doctor blade technique followed by heating at 100 °Cunder vacuum overnight. The sheet was punched into 12 mmdiameter electrodes with a mass loading of 1−2 mg. An electrolyte of1 M LiPF6 in 1:1:1 (volume ratio) ethylene carbonate (EC), dimethylcarbonate (DMC), and ethyl methyl carbonate (EMC) and apolyethene-based separator were employed. Galvanostatic charging−discharging studies were conducted by a Land CT2001A instrumentat 28 °C in a cell incubator. The full capacitors were also assembled incoin cells with a prelithiated anode (charge−discharge for five cyclesand charge to 0.4 V) and cathode in the same electrolyte. Theoptimized mass ratio of cathode and anode was about 2:1. Cyclicvoltammetry (CV) measurements and electrochemical impedancespectroscopy (EIS) were carried out on an electrochemical work-station (CH Instruments Inc., Shanghai, People’s Republic of China,CHI660E).The energy density (E, Wh kg−1) and power density (P, W kg−1)
were calculated on the basis of the equations15,16
E IV t C V V V V VIm
td12
( )( )t
t
1 2 1 21
2∫= = + − = Δ(4)
PEt
VmI= = Δ
(5)
VV V
21 2Δ =
+(6)
where V1 and V2 are the lower and upper limits of the voltage window(V), I is the discharge current (A), t is the discharge time (h), m is thetotal active material mass of both electrode (kg), and ΔV is theaverage working voltage ((0.1 + 4)/2 = 2.05 V).
■ RESULTS AND DISCUSSIONAs the cathode of an LIC, high surface area and good electricalconductivity are always two contradictory but expectedproperties for porous carbon. For this reason, we designed aporous carbon (PC-75) with an open-sandwich structureconsisting of a soft carbon outer layer and hard carbon innerlayer. From the SEM image (Figure 1a), PC-75 exhibits aunique morning glory like morphology. The diameter of theopening is roughly 100 nm (Figure 1b), and the thickness ofthe flower petal is about 35 nm. On consideration of the ∼10nm thickness of the phenolic resin based carbon flakes (FigureS2b), the single coating layer of the coal tar pitch isapproximately 12 nm in thickness. The almost transparentappearance of PC-75 in a transmission electron microscope(TEM) confirms its ultrathin nature and visible openingsapproximately 100 nm in size (Figure 1c). As is more clearlyseen in the HRTEM image (Figure 1d), the microcrystalcarbon sheet was coated by a highly graphitized carbon layerwith a distinct boundary.Scheme 1 depicts the overall formation process of this
morning glory like porous carbon (PC-75). First, the flakyphenolic resin (thickness ∼80 nm) was ball-milled with coaltar pitch and K2CO3 to give a homogeneous disperson (FigureS1a). After a prepolycondensation at 350 °C, the pitch was firstmelted and spread out on the surface of the flaky resin to form
a sandwich structure (Figure S1b). Meanwhile, K2CO3 buriedin the particles during the prepolycondensation process cancontribue to the hierarchical pore structure of PC-75. Thisprocess also improved the cross-linking of the polymer,reduced the thermoplasticity of the coal tar pitch, and avoidedthe blockage of the pores during the subsequent high-temperature activation process. In the subsequent high-temperature activation step, these flakes were converted fromstraight to curved by the tension stress derived fromhomogeneous thermal shrinkage to form a morning glorylike morphology that was further mutually cross-linkedtogether to create numbers of holes dispersed over thewhole carbon particle, which can work as the electrolytereservoirs for the fast diffusion of electrolyte ions.17 For thehigh surface area, K2CO3 activation started at a temperaturehigher than 600 °C and created abundant micropores.18 At thesame time of pyrolysis, the GO embedded in phenolic resinwas converted to reduced graphene oxide (RGO), whichcontributed to the conductivity as well.To trace the origin of this flowerlike morphology, we
prepared a series with varying mass percentages of coal tarpitch in the carbon source. The sample PC-T derived fromthermoplastic pure coal tar pitch exhibits a cellular structurewith a smooth surface (Figure S2a), whereas the sample PC-Rprepared solely from the phenolic resin shows a sheet structure(Figure S2b). When the content of coal tar pitch is lowered,the flaky phenolic resin in sample PC-50 is incompletelycoated by coal tar pitch (Figure S2c). A higher content of coaltar pitch results in the PC-85 sample having some blockedchannels due to too great an amount of coating (Figure S2d).Graphitization degrees of the carbons respectively derivedfrom two precursors were investigated by HRTEM, where PC-T shows a typical soft carbon characteristic with relativelyordered fringes (Figure S2e) and PC-R shows a typical hardcarbon feature with turbostratic amorphous microcrystals(Figure S2f). In addition, prepolycondensation is crucial forthe formation of a flowerlike morphology. The sample (PC-75*) prepared without the prepolycondensation step lost theprevious opened structure (Figure S3a), possibly due to theflowing of pitch during the pyrolysis, which was confirmed bythe N2 adsorption−desorption isotherm (Figure S3b) as well.Therefore, the flowerlike morphology and opening structureare mainly dependent on the rational mixture ratio and thenecessary prepolycondensation step.The N2 adsorption−desorption isotherms and correspond-
ing PSDs are shown in Figure 2a,b. The specific surface areaand the other pore structure parameters of all samples are alsosummarized in Table 1. The sample PC-T shows a type Iisotherm, indicating rich amounts of micropores. A specific
Scheme 1. Synthesis of Morning Glory Like Porous CarbonUsing Coal Tar Pitch and Flaky Phenolic Resin asPrecursors
ACS Applied Materials & Interfaces Research Article
surface area of 1826 m2 g−1 and 0.5−4 nm pores alsoconfirmed the microporous features (Table 1), which resultedfrom activation of K2CO3 toward the thermoplastic pitch. ThePC-75 features a type IV isotherm, reflecting the existence ofmesopores, and the SBET value increases to 2163 m2 g−1. Acontinuous increase in the resin content to 50 wt % resulted inthe sample PC-50 showing a more prominent type IV isothermwith a pronounced hysteresis loop, implying many moremacropores and mesopores. In contrast, PC-85 shows a nearlytype I isotherm, due to the excess amount of coal tar pitchused. PC-R has a smaller Vads increase (P/P0 > 0.9) andhysteresis loop in comparison to PC-50, due to the absence ofthe morning glory like morphology.The graphitization degree for porous carbons was measured
by XRD and Raman spectra. In the Figure 2c, all XRD patternsshow two broad peaks at a 2θ value between 20 and 30° and a2θ value of 43°, assigned to the typical (002) and (100)reflections, respectively. Their sharpness indicates the crystal-lite sizes of Lc(002) and La(110) in the amorphous carbon.17
After a background removal, the La(110) values of PC-T, PC-R, PC-85, PC-75, and PC-50 are respectively calculated to be2.35, 2.11, 2.33, 2.26, and 2.16 on the basis of the Scherrerformula, which explains the better graphitization of pitch-derived carbon which is beneficial to conductivity and electrontransfer. Meanwhile, all Raman spectra (Figure 2d) show twowell-defined characteristic bands at 1350 and 1580 cm−1,
which belong to the D and G modes of carbon, respectively.The G band (1580 cm−1) is the vibration of sp2-bondedcarbon atoms in a cycle or chain structure, while the D band(1350 cm−1) is ascribed to the breathing mode of sp3-bondedcarbon atoms in the cyclic structure: for example, edges, otherdefects, and disordered carbon.5,19,20 Their intensity ratiosbetween D and G bands (ID/IG) after fitting have beencalculated to be 1.11, 1.17, 1.11, 1.09, and 1.12 for PC-T, PC-R, PC-85, PC-75, and PC-50, respectively, indicating that theintroduction of the coal tar pitch is able to increase thegraphitization degree of porous carbon.
Electrochemical Performance of a Single Electrode.Before the contstruction of LICs, the electrochemical perform-ances of PC-T, PC-R, PC-75, and PC-50 were investigated inhalf-cells with lithium ion battery electrolyte and metalliclithium. PC-85 was eliminated for the electrochemical tests dueto the relatively low micropore surface area and volume. Thecyclic voltammetry (CV) profiles ranging from 3 to 4.5 V at 5mV s−1 of the four discussed samples (Figure 3a) exhibit asimilar quasi-rectangular shape, suggesting that the capacity ismainly contributed by the electrochemical double layer. Theirgalvanostatic charge/discharge curves (GCs) at a currentdensity of 1 A g−1 (Figure 3b) all have nearly ideal triangularshapes without any deviation and prove the above conclusiononce again. Furthermore, their superior electrochemicalreversibility is also revealed by the good symmetries betweencharging and discharging curves of the GCs.The EIS of PCs can demonstrate the conductivity
contribution from the coal tar pitch precursor. In Figure 3c,each Nyquist plot has a semicircle in the high-frequency regionfollowed by a sloping line in the low-frequency region. Thediameter of the semicircle defined as charge transfer resistance(Rct) is always related to the conductivity of the capacitivematerial.21 Therefore, a lower Rct value of PC-T and PC-75 incomparison to that of PC-R indicates a favorable electricalconductivity contributed by the high degree of aromaticfeatures of the pitch precursor. However, PC-50 has a muchlarger Rct in comparison to the others, which mainly suffersfrom its overlarge Vtol. The electrical conductivities of thesesamples were also investigated by a four-probe method (TableS1), which corroborated the EIS results and also confirmed theideal of hard@soft integrated structures. In the Bode plots(Figure 3d), in comparison to PC-T, the higher phase angle atlow frequency (0.01 Hz) of PC-75 denotes its bettercapacitance behavior due to the introduction of a hierarchicalpore structure, which can behave like ion-buffering reservoirsto accelerate electrolyte ion transfer into and out of the internalchannels of carbon.22 Furthermore, the plots of capacitanceretention vs applied current density (Figure 3e) also confirm
Figure 2. (a) N2 adsorption−desorption isotherms and (b) pore sizedistributions of PC-T, PC-R, PC-85, PC-75, and PC-50. Theisotherms of the last four samples are respectively vertically offsetby 100, 200, 300, and 400 cm3 g−1 (STP). (c) XRD patterns and (d)Raman spectra of these samples.
Table 1. Structural Parameters of Six Different Porous Carbons
aMass ratio in carbon source. bBET specific surface area (m2 g−1). cMicropore surface area (m2 g−1). dSingle point desorption total pore volume(cm3 g−1). eMicropore volume (cm3 g−1), based on t-plot method. fSpecific capacitance (F g−1) at 1 A g−1.
ACS Applied Materials & Interfaces Research Article
this conclusion, where the hierarchical porous carbon of PC-75has a better rate performance.From the aspect of capacity (Figure 3f), PC-75 exhibits a
higher specific capacity (∼50 mAh g−1) in comparison to PC-T (∼42 mAh g−1) at 1 A g−1, which is attributed to itsincreased specific surface area and superior rate performance(Figure S4a). Under the joint action of favorable conductivityand rational hierarchical pore structure, the smallest iR dropfor PC-75 can be observed in Figure S4b, particularly in highcurrent density operations. In addition, cycle stability is also animportant criterion to evaluate the behavior of cathodematerials in LIC. Table S2 gives the capacity retention ratiosof the discussed samples after 5000 cycles at 1 A/g−1, wherePC-75 has a good cycle stability of 83.8%, much better thanthat of PC-R (68.8%). The better stability may come from thelower oxygen content of PC-75 (Table S3).23 Overall, PC-75offers the best performance in comparison to the others in half-cell investigations and has been selected as the cathode in thefull cell of the LIC for further study.To meet the high energy target of an LIC, MnO with low
intercalation potential was chosen as the anode material tosatisfy a high voltage for the full cell. However, thecharacteristics of low structural stability and sluggishintercalation kinetics limit the cycle stability and rateperformance of the full cell. Thus, the MnCO3@polydop-amine-derived porous MnO@C electrode we reported earlier14
has been employed in this work. Its outstanding electro-chemical performance is believed to come from the following.
First, by the pore creation of pyrolytically produced CO2, theMnCO3 precursor can increase the intercalation sites of theMnO electrode and lead to a higher surface charge storage.24
On the basis of calculations,25 the outer charge (nondiffusioncontrol capacity) accounts for 32% of the total charge (FigureS5). Second, the MnCO3 precursor has a much larger volumethan MnO, which can reserve space to partially buffer thedramatic volume change upon lithiation/delithiation. Third,the pyrolysis from polydopamine to nitrogen-doped carboncan construct a conductive network on the surface of MnO forthe high rate demand.
Electrochemical Performance of Li Ion Capacitors. Ahigh energy density LIC was constructed by an anode ofporous MnO@C nanocubes and a cathode of PC-75 in lithiumion battery electrolyte for a detailed electrochemicalinvestigation (Figure 4a). As demonstrated in Figure 4b,
after consideration of the CV behavior of the MnO@C anodeand PC-75 cathode, an operating voltage of 0.1−4 V for thefull cell configuration has been determined by the potentialrange of both sides. To increase the full cell voltage, the anodewas cycled and prelithiated to 0.4 V in the assembled half-celland then paired with PC-75 in the coin cell to make up a wholeLIC (MnO@C//PC-75). After the mass loading was adjusted(Figure S6a), the mass ratio of the cathode and anode wasoptimized to 2:1 and the corresponding cyclic voltammetrycurve is shown in Figure S6b. From the GC curves in Figure4c, the specific capacitance values of MnO@C//PC-75 are
Figure 3. Electrochemical performance characteristics of samples in ahalf-cell configuration: (a) CV curves of PCs at a scan rate of 5 mVs−1; (b) galvanostatic charge−discharge curves of PCs at 1 A g−1; (c)Nyquist plots and (d) Bode plots of PCs in the frequency region of100 kHz to 0.01 Hz; (e) capacity retain ratios of PCs at currentdensities from 0.2 to 10 A g−1; (f) long-term cycling performances ofPCs at 1 A g−1.
Figure 4. (a) Illustration of the LIC, in which PC-75 and MnO@Care used as the cathode and anode, respectively. (b) Diagram of theoperating potential range for the hybrid configuration. (c) GC curvesof MnO@C//PC-75 at 0.2, 0.5, 1, 2, and 5 A g−1. (d) Ragone plot ofMnO@C//PC-75. Values reported for other lithium ion capacitorsare added for comparison. (e) Cycle performance of MnO@C//PC-75 with a mass ratio of 1:2 at 1 A g−1.
ACS Applied Materials & Interfaces Research Article
respectively 54.2, 44.8, 36.7, 28.4, and 15.6 F g−1 at currentdensities of 0.2, 0.5, 1, 2, and 5 A g−1 (all of the calculations inthe full cell discussion are based on the weight of both thecathode and anode). Meanwhile, all galvanostatic charge/discharge curves exhibit only a slight deviation from the linearslope of an ideal supercapacitor, indicating a favorablecapacitive character of the as assembled hybrid capacitor.The energy density and power density of MnO@C//PC-75are also displayed in the Ragone plot (Figure 4d), where themaximum energy density of 117.6 Wh kg−1 can be achieved ata power density of 410 W kg−1 and 27.8 Wh kg−1 can bedelivered even at a high power density of 10250 W kg−1. Oncomparison with reports of LICs in a very similarconfiguration, such as MnO-C//CNS,26 MnO-C//AC,27
SnO2−C//C,28 etc.,29,30 the LIC reported here has a verypromising energy density due to the novel synthesis of hard@soft integrated morning glory like porous carbon as well as therational match between the cathode and anode. Meanwhile, afurther comparison with other configurations is also given inTable S4. After 3000 cycles at 2050 W kg−1 (1 A g−1), areasonable capacity retention of 76% was achieved for this LIC(Figure 4e).Finally, a three-electrode test, containing an MnO@C
anode, PC-75 cathode, and Li reference electrode, was appliedto detect the performance evolutions of the anode and cathodeduring a long-term cycling and explore the real reason for thecapacity decay.31 The potentials of the anode and cathode canbe obtained by referencing them to a Li metal electrode (0 Vvs Li/Li+). In a typical cycle of the three-electrode test (Figure5a), the plots of cell voltage and cathode and anode potentials
revealed that both the cathode and anode operate in a safepotential range (0−4.5 V, without lithium deposition andelectrolyte decomposition), Thus, 0.4 V is a suitableprelithiated potential for MnO@C. A potential of 0.4 V isthe main discharge platform region of MnO, and the relativelylow potential increases the voltage of the full cell, which canalso provide suitable active sites to avoid the lithium depositionreaction during the first charge of the full cell. By observing thefinal potential of the anode and cathode in charge/discharge(Figure 5b), we find that the cyclic potential window of theanode keeps expanding along the cycle, which means a capacitydecay of the anode. In detail, the conversion-type MnO anodecan change its potential following the charge and dischargeprocess. When a fresh MnO anode was used, a potential rangeof 0.2−1.7 V was enough to meet the capacity requirement ofthe full cell. With the long-term cycle runs, the capacity of the
anode will be reduced more or less and a wider potential rangeis required to satisfy the capacity of the full cell.Correspondingly, the cyclic potential window of the anodewas expanded. This expanding also encompassed the potentialwindow (U = Umax − Umin) of the cathode, reduced the playcapacity (Q = CU) of the cathode, and eventually led tocapacity loss of the full cell. Oppositely, the capacitance (C) ofPC-75 was calculated to be nearly unchanged after 1500 cycles(from 96.9 to 96.7 F g−1, close to the value from the half-celltest), confirming the excellent cycling stability of PC-75 in thefull cell of an LIC.
■ CONCLUSION
In summary, we developed a novel method to prepare a uniquemorning glory like porous carbon by rationally compositingcoal tar pitch and flaky phenolic resin. The opened porousstructure produced by the flaky phenolic resin gives anenhancement of surface area and ion diffusion efficiency toporous carbon. Meanwhile, the coating feature of coal tar pitchwas applied to ensure a good electrical conductivity of porouscarbon. By varying the amount of pitch in the carbon source,the porous structure and specific surface area have beenoptimized to improve both capacity and rate performance.After assembly with a structure-optimized MnO@C anode, thereported LIC finally delivers an excellent maximum energydensity of 117.6 Wh kg−1 and a favorable cyclability of 76%capacity retain after 3000 cycles at 2050 W kg−1. Finally, aninvestigation of capacity loss in the full cell was also analyzedby a three-electrode test with a lithium reference electrode,which reveals that the anode stability is the main reason forcapacity loss during the cycling.
■ ASSOCIATED CONTENT
*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acsami.8b17340.
SEM images, HRTEM images, and N2 adsorption−desorption isotherm of a reference sample, specificcapacitance and iR drop at different current densities ofPCs, calculation of outer charge storage of the MnO@Canode at different scan rates in CV tests, charge−discharge cycling performance in different mass ratios ofanode to cathode and a typical CV curve of MnO@C//PC-75 at a scan rate of 10 mV s−1, capacity retentionratio after 5000 cycles of the discussed samples, andelectrochemical properties of other LICs in organicelectrolyte (PDF)
ORCIDWen-Cui Li: 0000-0001-8066-7144NotesThe authors declare no competing financial interest.
■ ACKNOWLEDGMENTS
This project was financially supported by National NaturalScience Funds of China (No. 21776041).
Figure 5. (a) Cell voltage and electrode potential in a typical cycle ofMnO@C//PC-75. (b) Plots of the final potential of the cathode andanode during cycling. These four scatter plots are respectively themaximum and minimum of the cathode potential and the maximumand minimum of the anode potential.
ACS Applied Materials & Interfaces Research Article
■ REFERENCES(1) Shen, L.; Lv, H.; Chen, S.; Kopold, P.; van Aken, P. A.; Wu, X.;Maier, J.; Yu, Y. Peapod-like Li3VO4/N-Doped Carbon Nanowireswith Pseudocapacitive Properties as Advanced Materials for High-Energy Lithium-Ion Capacitors. Adv. Mater. 2017, 29, 1700142−1700149.(2) Yang, M.; Zhong, Y.; Ren, J.; Zhou, X.; Wei, J.; Zhou, Z.Fabrication of High-Power Li-Ion Hybrid Supercapacitors byEnhancing the Exterior Surface Charge Storage. Adv. Energy Mater.2015, 5, 1500550−1500556.(3) Xu, D.; Chao, D.; Wang, H.; Gong, Y.; Wang, R.; He, B.; Hu, X.;Fan, H. J. Flexible Quasi-Solid-State Sodium-Ion CapacitorsDeveloped Using 2D Metal-Organic-Framework Array as Reactor.Adv. Energy Mater. 2018, 8, 1702769−1702779.(4) Wang, H.; Zhu, C.; Chao, D.; Yan, Q.; Fan, H. J. NonaqueousHybrid Lithium-Ion and Sodium-Ion Capacitors. Adv. Mater. 2017,29, 1702093−1702110.(5) Jain, A.; Jayaraman, S.; Ulaganathan, M.; Balasubramanian, R.;Aravindan, V.; Srinivasan, M. P.; Madhavi, S. Highly MesoporousCarbon from Teak Wood Sawdust as Prospective Electrode for theConstruction of High Energy Li-ion capacitors. Electrochim. Acta2017, 228, 131−138.(6) Wang, H.; Xu, Z.; Li, Z.; Cui, K.; Ding, J.; Kohandehghan, A.;Tan, X.; Zahiri, B.; Olsen, B. C.; Holt, C. M.; Mitlin, D. HybridDevice Employing Three-dimensional Arrays of MnO in CarbonNanosheets Bridges Battery-Supercapacitor Divide. Nano Lett. 2014,14, 1987−1994.(7) Lee, W. S.V.; Peng, E.; Li, M.; Huang, X.; Xue, J. M. RationalDesign of Stable 4V Lithium Ion Capacitor. Nano Energy 2016, 27,202−212.(8) An, C.; Liu, X.; Gao, Z.; Ding, Y. Filling and Unfilling CarbonCapsules with Transition Metal Oxide Nanoparticles for Li-ionHybrid Supercapacitors: Towards Hundred Grade Energy Density.Sci. China Mater. 2017, 60, 217−227.(9) Zhang, S.; Li, C.; Zhang, X.; Sun, X.; Wang, K.; Ma, Y. HighPerformance Lithium-Ion Hybrid Capacitors Employing Fe3O4-Graphene Composite Anode and Activated Carbon Cathode. ACSAppl. Mater. Interfaces 2017, 9, 17136−17144.(10) Liang, T.; Wang, H.; Xu, D.; Liao, K.; Wang, R.; He, B.; Gong,Y.; Yan, C. High-Energy Flexible Quasi-Solid-State Lithium-IonCapacitors Enabled by a Freestanding rGO-Encapsulated Fe3O4Nanocube Anode and a Holey rGO Film Cathode. Nanoscale 2018,10, 17814−17823.(11) Li, B.; Xiao, Z. J.; Chen, M.; Huang, Z. Y.; Tie, X. Y.; Zai, J. T.;Qian, X. F. Rice Husk-Derived Hybrid Lithium-Ion Capacitors withUltra-high Energy. J. Mater. Chem. A 2017, 5, 24502−24507.(12) Guan, T. T.; Li, K. X.; Zhao, J. H.; Zhao, R. J.; Zhang, G. L.;Zhang, D. D.; Wang, J. L. Template-Free Preparation of Layer-Stacked Hierarchical Porous Carbons from Coal Tar Pitch for HighPerformance All-Solid-State Supercapacitors. J. Mater. Chem. A 2017,5, 15869−15878.(13) Hao, G. P.; Lu, A. H.; Dong, W.; Jin, Z. Y.; Zhang, X.-Q.;Zhang, J. T.; Li, W. C. Sandwich-Type Microporous CarbonNanosheets for Enhanced Supercapacitor Performance. Adv. EnergyMater. 2013, 3, 1421−1427.(14) Cheng, F.; Li, W. C.; Lu, A. H. Using Confined CarbonateCrystals for the Fabrication of Nanosized Metal Oxide@Carbon withSuperior Lithium Storage Capacity. J. Mater. Chem. A 2016, 4,15030−15035.(15) Xia, Q. Y.; Yang, H.; Wang, M.; Yang, M.; Guo, Q. B.; Wan, L.M.; Xia, H.; Yu, Y. High Energy and High Power Lithium-IonCapacitors Based on Boron and Nitrogen Dual-Doped 3D CarbonNanofibers as Both Cathode and Anode. Adv. Energy Mater. 2017, 7,1701336−1701344.(16) Li, Y.; Wang, H.; Huang, B.; Wang, L.; Wang, R.; He, B.; Gong,Y.; Hu, X. Mo2C-Induced Solid-Phase Synthesis of Ultrathin MoS2Nanosheet Arrays on Bagasse-Derived Porous Carbon Frameworksfor High-Energy Hybrid Sodium-Ion Capacitors. J. Mater. Chem. A2018, 6, 14742−14751.
(17) Thangavel, R.; Kaliyappan, K.; Kang, K.; Sun, X.; Lee, Y.-S.Going Beyond Lithium Hybrid Capacitors: Proposing a New High-Performing Sodium Hybrid Capacitor System for Next-GenerationHybrid Vehicles Made with Bio-Inspired Activated Carbon. Adv.Energy Mater. 2016, 6, 1502199−1502207.(18) Sun, Y.; Wei, J.; Wang, Y. S.; Yang, G.; Zhang, J. P. Productionof Activated Carbon by K2CO3 Activation Treatment of CornstalkLignin and its Performance in Removing Phenol and SubsequentBioregeneration. Environ. Technol. 2010, 31, 53−61.(19) Wang, C.; Huang, J.; Qi, H.; Cao, L.; Xu, Z.; Cheng, Y.; Zhao,X.; Li, J. Controlling Pseudographtic Domain Dimension ofDandelion Derived Biomass Carbon for Excellent Sodium-ionStorage. J. Power Sources 2017, 358, 85−92.(20) Niu, J.; Shao, R.; Liang, J.; Dou, M.; Li, Z.; Huang, Y.; Wang, F.Biomass-Derived Mesopore-Dominant Porous Carbons with LargeSpecific Surface Area and High Defect Density as High PerformanceElectrode Materials for Li-Ion Batteries and Supercapacitors. NanoEnergy 2017, 36, 322−330.(21) Li, B.; Dai, F.; Xiao, Q.; Yang, L.; Shen, J.; Zhang, C.; Cai, M.Nitrogen-Doped Activated Carbon for a High Energy HybridSupercapacitor. Energy Environ. Sci. 2016, 9, 102−106.(22) Gokhale, R.; Aravindan, V.; Yadav, P.; Jain, S.; Phase, D.;Madhavi, S.; Ogale, S. Oligomer-Salt Derived 3D, Heavily NitrogenDoped, Porous Carbon for Li-Ion Hybrid Electrochemical CapacitorsApplication. Carbon 2014, 80, 462−471.(23) Naoi, K.; Ishimoto, S.; Miyamoto, J. i.; Naoi, W. SecondGeneration ‘Nanohybrid Supercapacitor’: Evolution of CapacitiveEnergy Storage Devices. Energy Environ. Sci. 2012, 5, 9363−9373.(24) Wang, P. Y.; Wang, R. T.; Lang, J. W.; Zhang, X.; Chen, Z. K.;Yan, X. B. Porous Niobium Nitride as a Capacitive Anode Material forAdvanced Li-Ion Hybrid Capacitors with Superior Cycling Stability. J.Mater. Chem. A 2016, 4, 9760−9766.(25) Li, S.; Chen, J.; Cui, M.; Cai, G.; Wang, J.; Cui, P.; Gong, X.;Lee, P. S. A High-Performance Lithium-Ion Capacitor Based on 2DNanosheet Materials. Small 2017, 13, 1602893−1602901.(26) Zhao, Y.; Cui, Y.; Shi, J.; Liu, W.; Shi, Z.; Chen, S.; Wang, X.;Wang, H. Two-Dimensional Biomass-Derived Carbon Nanosheetsand MnO/Carbon Electrodes for High-Performance Li-Ion Capaci-tors. J. Mater. Chem. A 2017, 5, 15243−15252.(27) Liu, C.; Zhang, C.; Song, H.; Zhang, C.; Liu, Y.; Nan, X.; Cao,G. Mesocrystal MnO Cubes as Anode for Li-Ion Capacitors. NanoEnergy 2016, 22, 290−300.(28) Qu, W.-H.; Han, F.; Lu, A.-H.; Xing, C.; Qiao, M.; Li, W.-C.Combination of a SnO2-C Hybrid Anode and a Tubular MesoporousCarbon Cathode in a High Energy Density Non-aqueous Lithium IonCapacitor: Preparation and Characterization. J. Mater. Chem. A 2014,2, 6549−6557.(29) Liu, C.; Zhang, C.; Song, H.; Nan, X.; Fu, H.; Cao, G. MnONanoparticles with Cationic Vacancies and Discrepant CrystallinityDispersed into Porous Carbon for Li-Ion Capacitors. J. Mater. Chem.A 2016, 4, 3362−3370.(30) Ulaganathan, M.; Aravindan, V.; Ling, W. C.; Yan, Q.; Madhavi,S. High Energy Li-ion Capacitors with Conversion Type Mn3O4Particulates Anchored to Few Layer Graphene as the NegativeElectrode. J. Mater. Chem. A 2016, 4, 15134−15139.(31) Shan, X.-Y.; Wang, Y.; Wang, D.-W.; Weng, Z.; Li, F.; Cheng,H.-M. A Smart Self-Regenerative Lithium Ion Supercapacitor with aReal-Time Safety Monitor. Energy Storage Mater. 2015, 1, 146−151.
ACS Applied Materials & Interfaces Research Article
