a review of graphene-based electrochemical microsupercapacitors

22
A Review of Graphene-Based Electrochemical Microsupercapacitors Guoping Xiong, a, b Chuizhou Meng, c, d Ronald G. Reifenberger, a, e Pedro P. Irazoqui, c, d Timothy S. Fisher* a, b a Birck Nanotechnology Center, Purdue University, West Lafayette, IN 47907, USA b School of Mechanical Engineering, Purdue University, West Lafayette, IN 47907, USA c Center for Implantable Devices, Purdue University, West Lafayette, IN 47907, USA d Weldon School of Biomedical Engineering, Purdue University, West Lafayette, IN 47907, USA e Department of Physics , Purdue University, West Lafayette, IN 47907, USA *e-mail: [email protected] Received: May 20, 2013 Accepted: August 20, 2013 Published online: && &&, 2013 Abstract The rapid development of miniaturized electronic devices has led to a growing need for rechargeable micropower sources with high performance. Among different sources, electrochemical microcapacitors or microsupercapacitors provide higher power density than their counterparts and are gaining increased interest from the research and engi- neering communities. To date, little work has appeared on the integration of microsupercapacitors onto a chip or flexible substrates. This review provides an overview of research on microsupercapacitors, with particular emphasis on state-of-the-art graphene-based electrodes and solid-state devices on both flexible and rigid substrates. The ad- vantages, disadvantages, and performance of graphene-based microsupercapacitors are summarized and new trends in materials, fabrication and packaging are identified. Keywords: Microsupercapacitors, Supercapacitors, Energy storage, Graphene, Nanomaterials, Electrochemistry DOI: 10.1002/elan.201300238 1 Introduction of Supercapacitors The need to store and use energy on diverse scales in a modern technological society necessitates the design of large and small energy systems, among which electrical energy storage systems such as batteries and supercapaci- tors have attracted much interest in the past several de- cades [1]. Supercapacitors, also known as ultracapacitors, or electrochemical capacitors, with fast power delivery and long cycle life, are playing an important role in com- plementing or even replacing batteries in many applica- tions [1, 2]. The first patent on supercapacitors was grant- ed to Becker at General Electric Corp. in 1957 [3], in which he proposed a capacitor based on porous carbon material with high surface area. Later in 1969, first at- tempts to market such devices were undertaken by SOHIO [4]. Between the late 1970s and the 1980s, Conway successfully fabricated supercapacitors with high specific capacitance and low internal resistance using RuO 2 as an active material [5]. In the 1990s, supercapaci- tors began to attract attention because of the emergence of hybrid electric vehicles [6]. Supercapacitors offered the promise to supplement batteries and fuel cells in hybrid electric vehicles in providing the necessary power for ac- celeration, and additionally to allow for the recuperation of brake energy. These promising studies prompted the U.S. Department of Energy to initiate supercapacitor de- velopment programs. A comprehensive review of the his- torical background, properties, and principles of superca- pacitors has been provided by Conway [1]. Supercapacitors effectively fill the gap between batter- ies and conventional capacitors (e.g., electrolytic capaci- tors or metalized film capacitors) [6]. They provide higher power density than batteries and fuel cells and higher energy density than conventional capacitors, while offering long lifetimes. In recent years, much progress has been achieved in both theoretical understanding [7–12] and experimental design [2, 13, 14] of high-performance supercapacitors. Meanwhile, their low energy density and high production costs have emerged as major challenges for the future development [1, 2, 6, 13]. As the size of portable electronic devices becomes smaller, low-power integrated circuits in devices such as sensors, microprocessors and wireless communication chips will make increasing use of miniature embedded microelectromechanical systems (MEMS) that operate in controlled/uncontrolled environments to gather, process, store and communicate information. Thus, there is a demand for integrated power sources to meet biological [15], medical [16] and environmental [17] applications. Micropower systems can be fabricated with length scales in the micrometer range and with improved performance by decreasing diffusion lengths to achieve this purpose. Among these micropower systems, microbatteries [18– Special Issue GRAPHENE Electroanalysis 2013, 25, No. &, 1 – 22 # 2013 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim &1& These are not the final page numbers! ÞÞ Review

Upload: pratik-sarkar

Post on 28-Nov-2015

58 views

Category:

Documents


2 download

DESCRIPTION

A Review of Graphene-Based ElectrochemicalMicrosupercapacitors

TRANSCRIPT

Page 1: A Review of Graphene-Based Electrochemical  Microsupercapacitors

A Review of Graphene-Based ElectrochemicalMicrosupercapacitors

Guoping Xiong,a, b Chuizhou Meng,c, d Ronald G. Reifenberger,a, e Pedro P. Irazoqui,c, d Timothy S. Fisher*a, b

a Birck Nanotechnology Center, Purdue University, West Lafayette, IN 47907, USAb School of Mechanical Engineering, Purdue University, West Lafayette, IN 47907, USAc Center for Implantable Devices, Purdue University, West Lafayette, IN 47907, USAd Weldon School of Biomedical Engineering, Purdue University, West Lafayette, IN 47907, USAe Department of Physics , Purdue University, West Lafayette, IN 47907, USA*e-mail: [email protected]

Received: May 20, 2013Accepted: August 20, 2013Published online: && &&, 2013

AbstractThe rapid development of miniaturized electronic devices has led to a growing need for rechargeable micropowersources with high performance. Among different sources, electrochemical microcapacitors or microsupercapacitorsprovide higher power density than their counterparts and are gaining increased interest from the research and engi-neering communities. To date, little work has appeared on the integration of microsupercapacitors onto a chip orflexible substrates. This review provides an overview of research on microsupercapacitors, with particular emphasison state-of-the-art graphene-based electrodes and solid-state devices on both flexible and rigid substrates. The ad-vantages, disadvantages, and performance of graphene-based microsupercapacitors are summarized and new trendsin materials, fabrication and packaging are identified.

Keywords: Microsupercapacitors, Supercapacitors, Energy storage, Graphene, Nanomaterials, Electrochemistry

DOI: 10.1002/elan.201300238

1 Introduction of Supercapacitors

The need to store and use energy on diverse scales ina modern technological society necessitates the design oflarge and small energy systems, among which electricalenergy storage systems such as batteries and supercapaci-tors have attracted much interest in the past several de-cades [1]. Supercapacitors, also known as ultracapacitors,or electrochemical capacitors, with fast power deliveryand long cycle life, are playing an important role in com-plementing or even replacing batteries in many applica-tions [1,2]. The first patent on supercapacitors was grant-ed to Becker at General Electric Corp. in 1957 [3], inwhich he proposed a capacitor based on porous carbonmaterial with high surface area. Later in 1969, first at-tempts to market such devices were undertaken bySOHIO [4]. Between the late 1970s and the 1980s,Conway successfully fabricated supercapacitors with highspecific capacitance and low internal resistance usingRuO2 as an active material [5]. In the 1990s, supercapaci-tors began to attract attention because of the emergenceof hybrid electric vehicles [6]. Supercapacitors offered thepromise to supplement batteries and fuel cells in hybridelectric vehicles in providing the necessary power for ac-celeration, and additionally to allow for the recuperationof brake energy. These promising studies prompted theU.S. Department of Energy to initiate supercapacitor de-

velopment programs. A comprehensive review of the his-torical background, properties, and principles of superca-pacitors has been provided by Conway [1].

Supercapacitors effectively fill the gap between batter-ies and conventional capacitors (e.g., electrolytic capaci-tors or metalized film capacitors) [6]. They providehigher power density than batteries and fuel cells andhigher energy density than conventional capacitors, whileoffering long lifetimes. In recent years, much progress hasbeen achieved in both theoretical understanding [7–12]and experimental design [2, 13, 14] of high-performancesupercapacitors. Meanwhile, their low energy density andhigh production costs have emerged as major challengesfor the future development [1,2, 6,13].

As the size of portable electronic devices becomessmaller, low-power integrated circuits in devices such assensors, microprocessors and wireless communicationchips will make increasing use of miniature embeddedmicroelectromechanical systems (MEMS) that operate incontrolled/uncontrolled environments to gather, process,store and communicate information. Thus, there isa demand for integrated power sources to meet biological[15], medical [16] and environmental [17] applications.Micropower systems can be fabricated with length scalesin the micrometer range and with improved performanceby decreasing diffusion lengths to achieve this purpose.Among these micropower systems, microbatteries [18–

Specia

lIssu

eG

RA

PH

EN

E

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim &1&

These are not the final page numbers! ��

Review

Page 2: A Review of Graphene-Based Electrochemical  Microsupercapacitors

20], microsupercapacitors [19,21], microfuel cells [22]and piezoelectric energy harvesters [23] have been ex-plored in the recent years. Microbatteries based on thinfilm solid-state Li/TiS were introduced to provide powerfor microsystems in the late 1990s [20], and since thenthe use of enhanced materials has further improved theirperformance [24, 25]. Despite these improvements, inher-ent problems associated with microbatteries remain,making them unable to satisfy various requirements ofautonomous microsystems (e.g., stability over long-termcycling, high power-high charge/discharge rates, and im-

Specia

lIs

sue

GR

AP

HE

NE

Guoping Xiong received his B. S.degree at Huazhong University ofScience and Technology in 2005 andM. S. degree at Tsinghua University,Beijing, China in 2008. He is nowworking on a PhD degree under theguidance of Prof. Timothy S. Fisherand Prof. Ronald G. Reifenberger.His research focuses on carbon nano-structure synthesis and applications inelectrochemical energy storage (e.g.,supercapacitors, microsupercapacitorsand lithium ion batteries).

Chuizhou Meng received the B. S.degree and Ph.D. degree in physicsfrom Tsinghua University, Beijing,China, in 2006 and 2011, respectively.During his graduate research, he fo-cused on the fabrication of advancedCNT/polymer nano-composites andtheir application in energy harvesting,conversion and storage. In August2011, he became a post-doctoral re-search associate in the Center for Im-plantable Devices and WeldonSchool of Biomedical Engineering,Purdue University, West Lafayette, IN, USA. His current re-search interest focuses on the fabrication of micro-supercapaci-tors and their integration with antenna, MEMs, ASICs into mini-ature RF-wireless implantable medical devices for clinical appli-cations such as targeted reinnervation, epilepsy, glaucoma, andcardiology.

Ron Reifenberger is a professor ofPhysics at Purdue University anda member of Purdue�s Center forSensing Science and Technology. Hereceived his undergraduate degree inPhysics from John Carroll Universityin 1970 and his PhD in Physics fromthe University of Chicago in 1976. Hejoined the Physics faculty at Purduein 1978 following a two-year post-doctoral appointment in the PhysicsDepartment at the University of Tor-onto. Upon joining the faculty atPurdue, Reifenberger initiated a program to measure photo-in-duced field emitted electrons from a variety of metals. Since1986, Reifenberger�s scanning probe group has been active infurthering inter-disciplinary nanoscale research at Purdue by es-tablishing collaborations with faculty throughout campus. Hisgroup has focused on research problems that emphasize the roleof scanning probe microscopy (SPM) as one of the key enablersof nanotechnology. His current research is focused on non-lineardynamics of SPM cantilevers, micro patterning of substrates forthe rapid detection of targeted bacteria, and fundamental meas-urements related to current flow in molecules, carbon nanotubesand Au nanocluster networks.

Pedro Irazoqui received his B.Sc. andM.Sc. degrees in Electrical Engineer-ing from the University of NewHampshire, Durham in 1997 and1999 respectively, and the Ph.D. inNeuroengineering from the Universi-ty of California at Los Angeles in2003 for work on the design, manu-facture, and packaging, of implanta-ble integrated-circuits for wirelessneural recording. Currently he is anassociate professor in the WeldonSchool of Biomedical Engineeringand the School of Electrical and Computer Engineering atPurdue University, where his lab is pursuing research into a mod-ular approach to the design of biological implants. He is the Di-rector of the Center for Implantable Devices working towardsclinical treatment of physiological disorders, using miniature,wireless, implantable systems. Specific research and clinical appli-cations explored in the center include: epilepsy, addiction, glau-coma, heart failure, and neural control of prostheses.

Timothy S. Fisher (PhD in Mechani-cal Engineering, 1998, Cornell)joined Purdue�s School of MechanicalEngineering and Birck Nanotechnolo-gy Center in 2002 after several yearsat Vanderbilt University. He is anAdjunct Professor in the Internation-al Centre for Materials Science at theJawaharlal Nehru Centre for Ad-vanced Scientific Research(JNCASR) and co-directs theJNCASR-Purdue Joint NetworkedCentre on Nanomaterials for Energy.From 2009 to 2012, he served as a Research Scientist at the AirForce Research Laboratory�s newly formed Thermal Sciencesand Materials Branch of the Materials and Manufacturing Direc-torate. In 2013 he became the James G. Dwyer Professor in Me-chanical Engineering at Purdue. Prior to his graduate studies, hewas employed from 1991 to 1993 as a design engineer in Motoro-la’s Automotive and Industrial Electronics Group. His researchhas included studies of nanoscale heat transfer, carbon nanoma-terial synthesis, coupled electro-thermal effects in semiconductorand electron emission devices, energy conversion and storagematerials and devices, microfluidic devices, biosensing, and relat-ed computational methods ranging from atomistic to continuumscales.

&2& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.

Page 3: A Review of Graphene-Based Electrochemical  Microsupercapacitors

munity to severe environments, particularly low tempera-tures). Microsupercapacitors, on the other hand, cancouple with microbatteries and energy harvesting micro-systems to provide high peak power, long cycle life, andhigh charge/discharge rates, while maintaining reasonableenergy densities for practical applications in microsys-tems. Moreover, today�s technological advancements inmicro and nano-scale fabrication provide a solid founda-tion for fabrication of microsupercapacitors. Thus it isreasonable to project that microsupercapacitors will satis-fy a variety of micropower demands and will complementor even replace microbatteries in electrochemical energystorage systems where high power delivery is required inshort times.

Figure 1 contains a customary Ragone plot that com-pares the volumetric specific energy and power densitiesof a Li thin film battery, a commercial 3.5 V/25 mF super-capacitor, a 63 V/220 mF electrolytic capacitor, and a mi-crosupercapacitor based on activated carbon (AC), a typi-cal electrode material, cited from [26]. Electrolytic capac-itors have higher charge/discharge rates and thus higherpower densities than AC microsupercapacitors, but theirspecific energy is more than two orders of magnitudelower. Meanwhile, lithium batteries and conventional su-percapacitors do not provide the ultrafast charge/dis-charge rates demonstrated by AC microsupercapacitors.Moreover, the specific energy of AC microsupercapaci-tors is roughly two orders of magnitude higher than com-mercial conventional supercapacitors and even higherthan Li thin film batteries.

In this review, we highlight recent developments in on-chip microsupercapacitor research. A particular focus ison electrochemical performance of graphene-based elec-trode materials. Section 1 contains an introductory back-ground on supercapacitors and particularly microsuperca-pacitors. Section 2 briefly discusses the fundamentals of

supercapacitors, while Section 3 highlights structural dif-ferences between conventional (or macro-) and microsu-percapacitors. Section 4 reviews graphene-based electrodematerials and their electrochemical performance. Sec-tion 5 reports developments in solid-state graphene-basedmicrosupercapacitors on both flexible and rigid substrateswhile Section 6 summarizes fabrication techniques of gra-phene-based electrodes. Finally, Section 7 contains con-clusions and a forward-looking outlook.

2 Fundamentals of Supercapacitors

2.1 Basic Structures and Performance Evaluation

A conventional supercapacitor consists of two electrodes(symmetric or asymmetric) and a separator sandwichedbetween them that are sealed in organic or aqueous elec-trolyte liquid. Recently, all-solid-sate supercapacitorshave also been designed using solid-state gel or polymerelectrolytes. The configuration of a conventional superca-pacitor is shown in Figure 2. Unlike electrolytic capaci-tors, where charge accumulates on two conductors sepa-rated by a dielectric, supercapacitors store charge at theinterface between an electrode and an electrolyte solu-tion. When charged, the negative ions in the electrolytesdiffuse to the positive electrode, while the positive ionsdiffuse to the negative electrode to create two separatelayers of capacitive storage. Consequently, in such a two-terminal configuration each electrode-electrolyte inter-face represents a capacitor so that the complete cell canbe considered as two capacitors in series, as shown inFigure 2. The cell capacitance for the supercapacitor cellcan be calculated from:

ðCcellÞ�1 ¼ ðC1Þ�1 þ ðC2Þ�1 ð1Þ

Specia

lIssu

eG

RA

PH

EN

E

Fig. 1. Comparative Ragone plot of a Li thin film battery, commercial supercapacitor, electrolytic capacitor and AC microsupercapa-citor, reprinted with permission from [26].

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de &3&

These are not the final page numbers! ��

Graphene-Based Electrochemical Microsupercapacitors

Page 4: A Review of Graphene-Based Electrochemical  Microsupercapacitors

where Ccell is the capacitance of the two-terminal deviceand C1 and C2 represent the capacitances of the two elec-trodes, respectively [13,14, 27].

Supercapacitors can be classified into two main typesin terms of working mechanism: (a) electric double-layercapacitors and (b) pseudocapacitors. Electric double-layercapacitors (EDLCs) store energy through ion adsorption;namely, the charge accumulations are achieved with elec-trostatically positive and negative charges separately re-siding on interfaces between electrolyte and electrodes.The charge transfer process in EDLCs is non-faradic, i.e. ,ideally no electron transfer takes places across the elec-trode interface. Pseudocapacitors store energy throughfast redox reactions between the electrolyte and electro-active materials on the electrode surface. Electron trans-fer causes charge accumulation, and the charge transferprocess is faradic in nature.

For electric double-layer supercapacitors, the specificcapacitance C (in F/g) of each electrode is approximatedby that of a parallel-plate capacitor [28,29]:

C ¼ ere0A=dm ð2Þ

where er is the relative permittivity, eo is the permittivityof vacuum (8.85� 10�12 F/m), A is the surface area of theelectrode accessible to the electrolyte ions, m is the massof active materials in grams; and d, the Debye length, isused to estimate the charge separation distance [2, 29].The Debye length is usually approximated as:

d ¼ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

eoerkBT2e2z2NAc1

r

ð3Þ

where er is the relative dielectric constant, kB is the Boltz-mann constant (1.38 �10�23 J/K), T is the temperature inKelvin, e is the elementary electronic charge (1.60 �10�19 C), z is the (integer) valence of the ionic species, NA

is the Avagadro constant (6.02 �1023 mol�1), and c1 is thebulk molar concentration (in moles/m3) of the ionic spe-

cies. Typically, d=0.3 nm for a 1 M concentration ofa monovalent ionic species in water (er =80) at 25 8C. Inelectric double-layer capacitors, according to Equation 2,high specific surface area (typically>1500 m2/g) of theactive electrode materials and charge separations d closeto atomic dimensions are the most important factors con-tributing to extremely high capacitance [14,30].

For pseudocapacitors, faradaic capacitance (CF) is cal-culated from the charge stored (Dq) and the change inpotential (DV) by [1, 5]:

CF ¼@ Dqð Þ@ DVð Þ ð4Þ

Experimentally, specific capacitances can be calculatedfrom both cyclic voltammetry (CV) and galvanostaticcharge/discharge. Capacitances derived from CV tests arecalculated from [31,32]:

C ¼ 12sm Vh � Vlð Þ

I

Vl!Vh!Vl

I Vð ÞdV ð5Þ

where s is the scan rate in V/s, Vh and Vl are high and lowpotential limits of the CV tests in V, I is the instantaneouscurrent in CV curves, and V is the applied voltage in V.Capacitances derived from galvanostatic charge/dischargetests are calculated from [33]:

C ¼ Id=Mv ð6Þ

where Id is the discharge current in A, and v is the slopeof the discharge curve after the initial voltage drop (IRdrop, VIR) of the discharge curves. The internal resistanceR (in W) is determined from the voltage drop at the be-ginning of a discharge curve by [34,35]:

R ¼ DVIR=2Id ð7Þ

where DVIR is the voltage dropped across the internal re-sistance in V. The factor of “2” is associated with the in-

Specia

lIs

sue

GR

AP

HE

NE

Fig. 2. Schematic representation of a supercapacitor cell.

&4& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.

Page 5: A Review of Graphene-Based Electrochemical  Microsupercapacitors

stant switch of current direction at the transition fromcharging and discharging.

The energy density (E in Wh/kg) and power density (Pin kW/kg) including maximum power density (Pmax inkW/kg) delivered for a single cell supercapacitor aregiven by:

E ¼ Ccell V2=2m ð8Þ

P ¼ E=Dt ð9Þ

Pmax ¼ V2=4mRs ð10Þ

where V is the cell voltage, Ccell is the total capacitance ofthe cell, P is the average power, ~t is the discharge timeand Rs is the equivalent series resistance (ESR in W).Maximum power is achieved when the internal resistanceof an energy storage device equals the load resistance.Energy and power densities strongly depend on the volu-metric capacitance of the device, cell voltage and ESR.The cell voltage is limited by the thermodynamic stabilityof the electrolyte solution. ESR derives from varioustypes of resistance associated with the intrinsic electronicproperties of the electrode matrix and electrolyte solu-tion, mass transfer resistance of the ions in matrix pores,contact resistance between the current collector and theelectrode, the ionic resistance of ions moving through theseparator and the electrolyte resistance [14].

The cyclic stability of a supercapacitor depends onmany factors such as mechanical properties and chemistryof the electrode materials. For instance, oxygen-rich func-tional groups on treated carbon electrodes usually pro-duce higher capacitance than untreated samples in organ-ic electrolyte; however, the functional groups can be det-rimental to cyclic stability of electrodes, resulting in in-creased series resistance and deterioration of capacitance[2, 14].

Leakage current and self-discharge rates are also im-portant factors to evaluate supercapacitor performance.Charged supercapacitors are in a state of higher freeenergy than the discharged state, and will undergo self-discharge because of this thermodynamic driving force.During self-discharge processes, a small amount of leak-age current will cause supercapacitors lose voltage(charge) as they idle in a charged state. This loss of volt-age may limit supercapacitor usage in some commercialapplications if the self-discharge rate is too high. Conse-quently, this issue is beginning to attract more attentionin supercapacitor fabrication [36,37], but systematic un-derstanding regarding the underlying decay mechanismsremains an area of active study. Several mathematicalmodels to understand and predict self-discharge profileshapes for single electrodes have been reported [1,38].

A wide variety of performance metrics have been usedto characterize supercapacitors, and unfortunately therelacks widespread agreement on which quantities to mea-sure and what units to use. This lack of standardized ap-proaches in measuring and reporting results creates diffi-

culties when attempting to compare the performance ofdifferent devices. A number of performance metrics canbe quoted, and most automated measurement systems arecapable of providing these metrics. A confounding issueis the conversion of measured values into normalizedunits. The measured capacitance (energy or power) isoften normalized by either the mass of an active ingredi-ent, the projected (Euclidean) area, or the device volume,although the volumetric and areal values are believed tobe more practically useful as compared to gravimetricones [39, 40]. Because the deciding factor(s) are often notspecified, uncertainties associated with the final quotedvalues are rarely known. One way to remedy this issue isto adopt standard radar plots that summarize the perfor-mance of a particular device.

The overall performance of supercapacitors should beoptimized in order to meet the following goals: (i) thepower density should be greater than that of batterieswhile maintaining an acceptable energy density; (ii) theelectrochemical cyclic stability should be high; (iii) thefinal device should exhibit fast charge/discharge rates,and (iv) a device should have reasonably low self-dis-charge rates. The example radar plots shown in Figure 3compare the major metrics of performance of two micro-supercapacitors reported in [41]. In Figure 3 a, red andblue curves are generated by connecting the electricalperformance data points from laser-scribed graphene(LSG)-based microsupercapacitors using a polyvinyl alco-hol (PVA)-H2SO4 gelled electrolyte and an ionogel elec-trolyte (a mixture of an ionic liquid with fumed silicananopowder), respectively. Figure 3 b compares durabili-ty-related metrics. In general, a larger area encompassedwithin a radar plot indicates better overall performance.As is apparent in Figure 3, the microsupercapacitor withPVA-H2SO4 electrolyte exhibits a slightly higher area-normalized capacitance, inferior cyclic stabilities, anda lower operating voltage, resulting in lower energy andpower densities than the microsupercapacitors based onthe ionogel electrolyte.

2.2 Electrode Materials and Electrolytes

Three main types of materials are frequently used as con-ventional supercapacitor active electrode materials: (i)carbon materials, e.g., carbon aerogel [27], activatedcarbon [42], carbon nanotubes [43,44] and graphene[29,45]; (ii) electroactive oxide or hydrous oxide films oftransition metals, e.g., MnO2 [33, 46,47], RuO2 [48], NiO[49], Co3O4 [49], MoO3 [50]; (iii) conducting polymers,e.g., polypyrrole [51], polyaniline [52] and polythiophene[53]. Apart from symmetric electrodes (anode and cath-ode using the same electrode materials), some supercapa-citors are designed based on asymmetrical electrode con-figurations (i.e. , one electrode consists of electrostaticcarbon material while the other consists of faradaic ca-pacitance material). One obvious advantage of suchasymmetric supercapacitors is that both electric double-layer capacitance and faradaic capacitance mechanisms

Specia

lIssu

eG

RA

PH

EN

E

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de &5&

These are not the final page numbers! ��

Graphene-Based Electrochemical Microsupercapacitors

Page 6: A Review of Graphene-Based Electrochemical  Microsupercapacitors

occur simultaneously, rendering a higher working voltagewindow and higher energy and power densities in super-capacitors than with symmetric electrodes (see Equa-tions 8–10). In recent years, extensive research has beencarried out on this topic [54–59].

The electrolyte is also a critical factor that influencessupercapacitor performance. The requirements for a goodelectrolyte include a wide voltage window, high electro-chemical stability, high ionic concentration and low sol-vated ionic radius, low resistivity, low viscosity, low vola-tility, low toxicity, low cost, and availability at high purity[13]. The electrolytes used in supercapacitors can be clas-sified into two main types: conventional liquid electro-lytes and solid-state electrolytes.

Conventional liquid electrolytes include: (i) aqueouselectrolytes, (ii) organic liquid electrolytes and (iii) ionicliquid electrolytes. The three types of electrolytes havebeen addressed in prior reviews [6,13]. Aqueous and or-ganic choices for liquid electrolytes each have their ownadvantages. Aqueous electrolytes such as acids (e.g.,

H2SO4) and alkalis (e.g., KOH) tend to have higher ionicconductivity (up to 1 S/cm) and higher dielectric constant[60], and usually give a higher specific capacitance foractive materials than organic electrolytes. Combined withlower operation cost, aqueous electrolytes have been ex-tensively used in supercapacitor design. Despite these ad-vantages, the cell voltage of aqueous electrolyte-based su-percapacitors is usually restricted to ca. 1 V, which is setby the decomposition voltage of water at 1.23 V [14,27].This value is generally lower than that of the organic elec-trolytes, which tends to be greater than 2.5 V [61]. Thisvoltage window of supercapacitors is closely related toenergy and power densities (see Equations 8–10).

Solid-state electrolytes (also known as fast ion conduc-tors or superioinc conductors) conduct electricity due tothe movement of ions through voids or defects in theircrystal lattice. These materials are relatively new andhave been applied to the design and fabrication of bothmacro- and microsupercapacitors. Solid electrolytes offermany advantages over their liquid counterparts[34,35, 62, 63]. Solid electrolytes do not exhibit leakagesince they are well dispersed and bound into a polymermatrix. Conversely, liquid electrolytes require robust en-capsulation to prevent leakage. Electrolyte leakage isa severe safety issue when the sealed electrolytes are en-vironmentally hazardous. In addition, solid-state gel orpolymer electrolytes offer dual functionality as they com-bine the separator and the electrolyte into a single layer.This situation should be contrasted to liquid electrolytes,where a separator to avoid electrical contact between theelectrodes is needed. Encouragingly, there are reportsthat the performance of solid-state gel electrolytes iscomparable to liquid counterparts [34,64, 65].

Because of the foregoing advantages, solid-state elec-trolytes are becoming more common in the design of con-ventional supercapacitors that are small, thin, lightweight,and flexible, in order to meet the requirements of rapidlygrowing modern markets, such as multifunctional porta-ble electronic devices. Solid-state electrolytes, such asPVA/H2SO4 [34,35], PVA/H3PO4 [44, 66] and 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide([EMIM][NTf2], C-TRI)/fumed silica nanopowder gelelectrolyte [67] and PVDF/ BMIM+BF4� ionic liquidpolymer electrolyte [68] have been successfully imple-mented in the design of both flexible and nonflexible con-ventional supercapacitors and show excellent electro-chemical performance.

The advantages offered by solid-state electrolytes areparticularly beneficial in microsupercapacitor applicationssuch as power sources for MEMS applications wherewafer-level packaging, cost, yield, and reliability are allequally important. Packaging is typically performed usinginterfacial bonding or deposition sealing techniques, asdescribed in [69]. It appears that the sealing and packag-ing processes for microsupercapacitors are not yet welldeveloped since the preparation of such microdevices isconfined to scientific research, i.e. , characterization ofmaterials and structures. Except for one recent study

Specia

lIs

sue

GR

AP

HE

NE

Fig. 3. Example radar plots to compare two microsupercapaci-tors cited from [41]. Red and blue curves are generated by con-necting data points from LSG-based microsupercapacitors usinga PVA-H2SO4 gelled electrolyte and an ionogel electrolyte, re-spectively. The electrical performance is compared in (a), whiledurability metrics are presented in (b). If a data point is not plot-ted, no information about that category is available in the refer-ence.

&6& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.

Page 7: A Review of Graphene-Based Electrochemical  Microsupercapacitors

[70], most studies have considered non-encapsulated mi-crodevices, with performance tested either in open air orcontrolled glove-box environments. Solid-state electro-lytes, however, provide a fast and facile way to encapsu-late either inorganic acid or ionic liquid electrolyte intoa matrix (e.g., PVA, PVDF, fumed silica nanopowder)while avoiding electrolyte leakage. In contrast, proce-dures to seal organic electrolytes reliably for microsuper-capacitor applications are complicated. Moreover, solid-state electrolytes also allow microsupercapacitors to befabricated with more functionality (e.g., flexibility andtransparency). The development of solid-state, graphene-based microsupercapacitors is discussed further in Sec-tion 5.

3 Structural Differences Between Macro- andMicrosupercapacitors

The various principles mentioned above apply to bothmacro- and microsupercapacitors, which differ both intheir structural design and ionic diffusion behavior asschematically illustrated in Figure 4. Macro-supercapaci-tors (conventional supercapacitors) employ vertical sand-wich structures, consisting of two electrodes and a separa-tor that are soaked with an electrolyte (Figure 4 a). Mi-crosupercapacitors predominantly use an interdigitatedstructure for on-chip design (Figure 4b) [26,71–74], al-though electrodes can also be designed in sandwich [75]and roll-like [76] structures. Microsupercapacitors withelectrodes having a characteristic dimension that ap-proaches 25 mm will further exploit the properties of ul-tramicroelectrodes (UMEs) [77], which have been inten-sively studied because they enable unprecedented spatialand temporal resolution in electrochemical measure-ments. UMEs have combined high current densities with

low measuring currents to open up new areas in sensorsand scanning electrochemical microscopy. New effectsarise because of a time-dependent change in mass trans-port, in which traditional, one-dimensional diffusionfields are replaced by rapidly varying fields that are spa-tially inhomogeneous. A further advantage offered by aninterdigitated architecture becomes noteworthy when lim-ited area considerations become important, as is the casefor microsupercapacitor integration onto integrated cir-cuit chips.

For macrosupercapacitors, major factors affecting thecharge/discharge rates are the thicknesses of the electro-des and separator. The thickness of separators (ts) is usu-ally 20 to 30 mm, and that of electrodes (t1) in the rangeof tens to hundreds of microns. Assuming fixed electrodearea, the only way to improve charge storage is to in-crease the thickness of the electrodes, which leads toa higher ion diffusion length and thus lower charge/dis-charge rates and power densities. On the other hand, inan interdigitated finger structure, electrodes of width we

are separated on an insulating plane by a gap wg. The ca-pacitance for such a structure is proportional to the ratiowe/wg [78]. Due to resistance of the electrolyte itself [6],as wg increases the longer ion diffusion path leads toa higher ESR, and thus a lower maximum power. Be-cause both we and wg can be adjusted by fabricationmethods (Section 6 below), the average migration dis-tance of ions can be controlled. In order to increase thecapacitance of the microdevice without compromising theion diffusion length and power, it is also possible to in-crease the thickness of the electrodes (t2) to fabricate 3Dinterdigitated structures, in which case the migration dis-tance of ions is not directly increased. Note that typicalvalues of t2, we and wg are in the range of tens to hun-dreds of microns. Taken together, 2D supercapacitor ar-chitectures are desirable for cost-effective mass produc-

Specia

lIssu

eG

RA

PH

EN

E

Fig. 4. Difference between (a) a sandwich structure, commonly used in conventional supercapacitors and (b) an interdigitated fingerstructure, used in microsupercapacitors.

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de &7&

These are not the final page numbers! ��

Graphene-Based Electrochemical Microsupercapacitors

Page 8: A Review of Graphene-Based Electrochemical  Microsupercapacitors

tion when layers of functionality are built up [79]. Themerits of 3D electrodes with a small area footprint andshort transport lengths become more obvious when on-board sources of power become increasingly smaller [79].Microbatteries based on 3D structures show better prop-erties than planar thin film batteries for autonomousMEMS [25, 80]. Thus, it is reasonable to conclude that de-signing 3D electrode structures for microsupercapacitorswill be of high interest in future development of the tech-nology. For this class of devices, most of the reported spe-cific capacitances in literature have been normalized byEuclidean area or volume of electrodes instead of themass of active materials.

In summary, a crucial method to increase the specificcapacitance and energy density of a microsupercapacitoris to increase the thickness of the active materials (i.e. ,3D structures) in a given footprint area. A key factor toincrease power density is to decrease the spacing betweenadjacent electrodes. High specific surface areas and highelectronic conductivity of the active materials are alsoprerequisites for the fabrication of efficient 3D electro-des.

4 State-of-the-Art Graphene-BasedMicrosupercapacitors

The goal of microsupercapacitor devices is to improvetheir performance in a limited footprint area by usinghigh-capacitance active materials and well-designed 3Dstructures. Efforts to increase the energy and power den-sities by designing novel 3D structures for the electrodeshave appeared in recent years [19,21], and various nano-structured materials have been used in the microsuperca-pacitor electrodes. Like their more conventional counter-parts, microsupercapacitors are usually fabricated fromthree main types of materials: (i) carbon materials withhigh specific surface area (ii) conducting polymers and(iii) metal oxides with high pseudocapacitance. We sum-marize and analyze the recent development in electrodematerials and structures (particularly based on graphene),as well as their electrochemical performance for microsu-percapacitor applications.

4.1 Carbon Materials as non-Faradaic Electrodes

Because of the advantages of low cost, easy processing,non-toxicity, high specific surface area, good electronicconductivity, high chemical stability, and wide operatingtemperature range, carbon materials are promising forlarge-scale fabrication. To enable their use as supercapa-citor electrode materials, they must have [81]: (i) highspecific surface areas, of the order of 1000 m2/g, (ii) goodintra- and inter-particle conductivity in porous matrices,and (iii) good electrolyte accessibility to intra-pore re-gions. Prior work [2, 13] indicates that carbon-based elec-trochemical capacitors function similarly to electrochemi-cal double-layer capacitors, which rely on high specific

area to accumulate non-faradaic charges at the boundarybetween an electrode and an electrolyte. Thus, unlikepseudocapacitive materials, carbon-based active materialsexhibit true capacitive behavior and excellent chemicalstability upon cycling [14]. To date, carbon materials withhigh specific areas such as carbon nanotubes [82], gra-phene [83], activated carbon [26,78, 84], carbide-derivedcarbon [74], and carbon onions [26] have been reportedas active electrode materials in microsupercapacitors. Atable summarizing the reported electrochemical perfor-mance of the microsupercapacitors based on these mate-rials is provided at the end of this section.

Activated carbons, produced by either thermal activa-tion or chemical activation, are the most widely used elec-trode materials because they have a high specific surfacearea (approx. 1200 m2/g [14]), good electrochemical sta-bility, and a relatively high electronic conductivity inaqueous and organic electrolytes [85–90]. The carbonsgenerally contain planar networks of hexagonal carbonrings with a size and stacking determined by the particu-lar carbon preparation method employed. Generally,there is little order between the sheets and no long-range3D order [85]. Conventional supercapacitors based on ac-tivated carbons, with different pore sizes (ranging from0.9 nm to 1.5 nm), are reported with specific capacitancesof 27.9 to 400 F/g in aqueous potassium hydroxide elec-trolyte, equivalent to area-normalized capacitances of 11to 46 mF/cm2 [86, 87]. To date, only a few studies havebeen reported [26,78, 84] on microsupercapacitors basedon activated carbons.

Pech et al. fabricated microsupercapacitors with acti-vated carbons by electrophoretic deposition [26] andinkjet methods [84]. However, the microelectrodes witha few micrometer thicknesses showed relatively low area-normalized capacitances (<5 mF/cm2 in 1 M Et4NBF4

propylene carbonate electrolyte). Shen et al. [78] filled 50to70-mm-thick activated carbons into etched channels onsubstrates as microelectrodes, exhibiting much higherarea-normalized capacitance (90.7 mF/cm2) and powerdensity (51.5 mW/cm2). Durou and co-workers [70] fabri-cated microsupercapacitors by depositing activated car-bons combined with 15wt% PVDF as active materialsinto KOH-etched silicon cavities. The device exhibited anarea-normalized capacitance of 81.0 mF/cm2 (correspond-ing to a volumetric capacitance of 6.8 F/cm3) at 5 mV/s scan rate and a specific energy density of 257 mJ/cm2,5.7 times higher than the reported performance forcarbon-based microsupercapacitors [78]. The highest re-ported energy density was realized when the depositedactive materials reached a thickness of several hundredmicrons. However, this device exhibited a high equivalentseries resistance, approx. 200 W (45.4 Wcm2), leading toa relatively low specific power of 34.4 mW/cm2. The highinternal resistance was attributed to the electrolyte resist-ance and particularly the poor contact between the cur-rent collector and the electrode material (in this case, ac-tivated carbon and binder). A conductive agent needs tobe added to reduce the series resistance of the electrodes.

Specia

lIs

sue

GR

AP

HE

NE

&8& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.

Page 9: A Review of Graphene-Based Electrochemical  Microsupercapacitors

In addition to the foregoing factors, one that may havebeen ignored is that the gaps between two adjacent elec-trodes are not filled with electrolyte instead with solid sil-icon. Because the active materials are confined to theetched silicon cavities, one might expect longer diffusionlengths and thus higher resistance during ion transport.Although activated carbons are good candidates for on-chip microsupercapacitor electrode materials, their specif-ic capacitance might be further improved by adding pseu-docapacitive materials like their more conventional coun-terparts [91].

Recently, electrochemically activated carbon microelec-trode arrays derived from the pyrolysis of patterned pho-toresist were reported [19,21, 92–94]. The microelectrodearrays were fabricated through the following steps [92]:(i) 2D interdigitated patterns were created by photoli-thography using SU-8 25 photoresist; (ii) a second photo-lithography step was employed using SU-8 100 photore-sist to create cylindrical posts on patterned fingers; (iii)SU-8 structures were pyrolyzed at 1000 8C for 1 h in form-ing gas (95 % N2 and 5 % H2); (iv) electrochemical activa-tion was performed on carbon microelectrode arraysbefore CV and galvanostatic charge/discharge characteri-zation. The microsupercapacitor device exhibited a specif-ic geometric capacitance of 75 mF/cm2 at a scan rate of5 mV/s and a fairly good cyclic stability (capacitance lossof approx. 13% after 1000 CV cycles). After carboniza-tion, the measured post diameters of the pyrolyzedcarbon arrays ranged from 53 to 68 mm, which limited theaccessible specific surface area. The overall electrochemi-cal performance could be further improved with (i) de-creased diameter of the pyrolyzed carbon posts; (ii) in-creased density of the posts; and (iii) decreased internalresistance by depositing metal current collectors.

CNTs have attracted interest as electrode materials forconventional supercapacitors [34,35, 95–97] because oftheir unique structure, high surface area, low mass densi-ty, outstanding chemical stability and excellent electronicconductivity [98–103]. Compared with activated carbons,CNTs have several advantages: (i) the electrical conduc-tivity of CNTs is greater than 100 S/cm, higher than acti-vated carbon (2.5 S/cm); (ii) CNT electrodes are binder-free and each tube is connected directly to the substrate,assuming that CNT arrays are grown by CVD and notdispersed onto the substrate from a liquid suspension,while activated carbon electrodes contain binder that in-creases the contact resistance between particles; (iii) mostof the open space in CNT electrodes consists of meso-pores that contribute to double-layer capacitance and fastion transport rates [104], whereas the pore distribution ofactivated carbons contains a mixture of micropores(<2 nm), mesopores (2 to 50 nm), and macropores(>50 nm). Micropores can significantly increase surfacearea but fail to produce the effect of double-layer capaci-tance due to the impedance of ion diffusion and ion-siev-ing effects, particularly when larger organic electrolytesare used [1, 14,105]. This realization seems to be contra-dicted by recent studies suggesting an anomalous increase

in carbon capacitance at pore sizes less than 1 nanometer[7, 106,107].

To date, few studies of CNT-based microsupercapaci-tors have been reported [82,108]. Microelectrodes basedon vertically aligned CNT arrays seem better than ran-domly aligned CNTs because random tubes might extendinto the gap region between two adjacent electrodes toproduce a short circuit. However, because of weak vander Waals force between the tubes in the array and thepoor mechanical bonding of CNTs to the supportingmetal films, the internal resistance tends to be large. Thereported capacitance of as-prepared CNT microsuperca-pacitors was 36.5 F/g, with a calculated energy density ofapprox. 0.4 Wh/kg and a power density of approx. 1 kW/kg.

A common problem with CNT arrays as electrodes isassociated with poor substrate bonding that is exposedwhen they are wetted by an aqueous electrolyte to causenot only detachment of CNTs from the substrate but alsodegradation of the vertical orientation resulting in poorcyclic stability. Consequently, highly ordered CNT arrayelectrodes with high quality electronic properties and me-chanical robustness are required. To date, balancing theseproperties remains a challenge, although several attemptshave been made. A typical method is to fill the array withadditional conducting polymer and binder to achieveboth good electronic conductivity and also mechanical ro-bustness [109]. Chen et al. [110] used a conducting poly-mer composite layer to achieve good electrical connectiv-ity, but only 90 % of the CNT length (approximately 2 mmthick) was exposed and part of the CNT array was dam-aged. New fabrication techniques are needed to achievehighly ordered CNT array electrodes that are optimizedfor CNT microsupercapacitor applications.

Since a mechanically exfoliated graphene monolayerwas first observed and characterized in 2004 [111], muchresearch in both scientific and engineering applications ofgraphene has been carried out worldwide, including ex-tensive attempts to use graphene in conventional superca-pacitors [14,29, 45,112]. Among the graphene materials,reduced graphene oxide (rGO) is most frequently used asan active material in conventional supercapacitors be-cause of its low-cost, scalability, wet-chemical propertiesand the high density of chemically active defect sites[113–115]. rGO is also of high interest in the fabricationof microsupercapacitor electrodes. Interestingly, GO canbe used as a solid electrolyte [116]. When a substantialamount of water is entrapped in the layered GO, it be-comes a strongly anisotropic ion conductor as well as anelectrical insulator, making it both a viable electrolyteand an electrode separator.

Gao et al. [83] used a laser technique to write rGO pat-terns directly on free-standing hydrated GO films for mi-crosupercapacitors. Three different geometric patternswere studied (see Figure 5), and a concentric circular pat-tern exhibited the highest capacitance density (0.51 mF/cm2) and volumetric capacitance (3.1 F/cm3), consideringonly the active thickness of the electrodes. These values

Specia

lIssu

eG

RA

PH

EN

E

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de &9&

These are not the final page numbers! ��

Graphene-Based Electrochemical Microsupercapacitors

Page 10: A Review of Graphene-Based Electrochemical  Microsupercapacitors

are nearly twice those of a comparable sandwich struc-ture. The energy density for this device is calculated to be4.3 �10�4 Wh/cm3, with a power density of 1.7 W/cm3. As-prepared sandwich and concentric circular devicesshowed fairly good cyclic stabilities with less than 35 %loss of capacitance after 10000 cycles.

Recently, rGO electrodes have been patterned bydirect laser irradiation of graphite oxide films under am-bient conditions [117, 118]. The electrical properties of as-prepared laser-scribed graphene can be tuned over 5orders of magnitude of conductivity by varying the laserintensity and laser irradiation treatments. Scalable fabri-cation of rGO-based solid-state microsupercapacitorsover large areas was demonstrated by this laser scribingmethod on graphite oxide films using a standard Light-Scribe DVD burner (see Figure 6) [41]. Microsupercapa-citors with an ionogel electrolyte, fumed silica nanopow-der with the IL 1-butyl-3-methylimidazolium bis(trifluor-omethylsulfonyl)imide (FS-IL), exhibit a power density of200 W/cm3, a frequency response with an RC time con-stant of 19 ms and a low leakage current (<150 nA after12 h). Microsupercapacitors based on GO seem to bepromising for large-scale fabrication; however, as-pre-pared GO-based supercapacitors may face temperaturerestrictions that require use near or below room tempera-ture to avoid serious stability problems; temperature pro-grammed desorption experiments indicate the decomposi-tion of graphene oxide begins at a relatively low tempera-ture of 70 8C [119].

Ultrathin supercapacitors based on an open “planar”architecture between opposing graphene thin films have

been reported [66]. The structure, comprised of grapheneflakes that are well connected along the flake edges, isnotable because it has sufficient porosity to maximize thecoverage of the electrode surface by electrolyte ions.These “planar” supercapacitors are based either on few-layer graphene or multilayer rGO thin films with PVA/H3PO4 polymer gel as an electrolyte. The effect of gra-phene edges reportedly leads to specific capacitances of80 mF/cm2 for few-layer graphene electrodes, while muchhigher specific capacitances (0.394 mF/cm2) were ob-served for multilayer rGO electrodes. Multi-layer gra-phene electrodes with the same area and similar amountsof electrolyte were compared in order to demonstrate theadvantages offered by the open “planar” structure overa more conventional stacked structure. The area-normal-ized capacitance derived from the “planar” structure(0.394 mF/cm2) was almost 3 times higher than ofa stacked layer-by-layer structure (0.14 mF/cm2), indicat-ing the “planar” geometry�s superior charge mobility andeffective utilization of the electrochemical surface area.

Microsupercapacitors based on rGO/CNT compositeelectrodes have been fabricated by combining electrostat-ic spray deposition (ESD) and photolithography lift-offmethods [120]. The fabrication process can be brieflysummarized as (see Figure 7): (i) creation of interdigitat-ed metal current collectors by conventional photolithog-raphy and wet etching; (ii) preparation of a removablemask by spin coating Omnicoat and SU-8, followed byphotolithography using a semi-automated, four-camera,optical front/backside mask aligner; (iii) oxygen plasmaetching to remove the excess Omnicoat; (iv) ESD of the

Specia

lIs

sue

GR

AP

HE

NE

Fig. 5. Schematics of CO2 laser-patterning of free-standing hydrated GO films to fabricate rGO�GO�rGO devices with in-plane andsandwich geometries and a digital image of the result, reprinted with permission from [83].

&10& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.

Page 11: A Review of Graphene-Based Electrochemical  Microsupercapacitors

hybrid rGO/CNT materials on the substrates; and (v) re-moval of the mask. The ESD was demonstrated to bothdeposit and reduce GO to rGO. The microsupercapaci-tors based on the hybrid electrodes exhibit a capacitanceof 6.1 mF/cm2 at a scan rate of 0.01 V/s and a resistive-ca-pacitive time constant of 4.8 ms in 3 M KCl electrolyte.

Carpet-based microsupercapacitors have been fabricat-ed from 3D graphene/CNTs grown in situ on Ni electro-des [121]. The interdigitated graphene/CNT structure forthe microdevice is shown in Figure 8. The microsuperca-pacitor shows an impedance phase angle of �81.58 at

a frequency of 120 Hz, comparable to commercial alumi-num electrolytic capacitors for alternating current line fil-tering applications. The microdevice also delivers a volu-metric energy density of 2.42 mWh/cm3 in ionic liquidand a maximum power density of 115 W/cm3 in aqueouselectrolyte at a rate of 400 V/s. The remarkable perfor-mance of the device was attributed to seamless nanotube/graphene junctions at the interface of the differingcarbon allotropic forms.

Carbide-derived carbon (CDC) is a promising elec-trode material for microsupercapacitors with high volu-

Specia

lIssu

eG

RA

PH

EN

E

Fig. 6. (a–c) Schematic diagram of fabrication process for microsupercapacitors by laser scribing method. (d, e) Flexible microsuper-capacitors with high areal density, reprinted with permission from [41].

Fig. 7. (a) Schematic drawing of fabrication procedures of microsupercapacitors .Inset: Digital photograph of a fabricated device. (b,c) Top view SEM images of rGO/CNT-based interdigital microelectrode arrays, reprinted with permission from [120].

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de &11&

These are not the final page numbers! ��

Graphene-Based Electrochemical Microsupercapacitors

Page 12: A Review of Graphene-Based Electrochemical  Microsupercapacitors

metric capacitance [74]. CDC can be produced by selec-tively etching metals from metal carbides using chlorineat elevated temperatures in a process that is similar todry-etching techniques employed for MEMS and micro-chip fabrication. CDC is reported to have the followingadvantages [74]: (i) microstructures can be preciselytuned by tailoring the synthesis conditions; (ii) the precur-sor TiC is conductive and can be deposited in a uniformfilm by chemical or physical vapor deposition technique;and (iii) CDC coatings are strongly adherent with anatomically perfect interface leading to low impedance. Aschematic of the fabrication process is found in Figure 9.Electrochemical results in both TEABF4 and H2SO4 elec-trolyte indicate that volumetric capacitance decreaseswith increasing coating thickness. For a CDC film ofapprox. 50 mm in thickness, volumetric capacitances ofapprox. 60 F/cm3 and 90 F/cm3 were measured inTEABF4 and H2SO4 electrolyte, respectively. As the coat-ing thickness decreased to approx. 2 mm, the volumetriccapacitance increased to nearly 180 F/cm3 in TEABF4

electrolyte and 160 F/cm3 in 1 M H2SO4. The decrease in

volumetric capacitance with thicker films was most likelydue to microstructural rearrangement from surface stressrelaxation, which resulted in porosity collapse and pertur-bation of the interconnected structure that facilitates elec-tron conduction.

Porous and highly conducting nanocrystalline graphiticcarbon film, derived from the graphitization of NiTi alloyand n-type polycrystalline SiC at temperatures less than1050 8C, has also been used to fabricate on-chip microsu-percapacitors [122]. Electrochemical characterization re-veals that incorporated nitrogen in the carbon electrodemight induce pseudo-capacitance. The capacitance wascalculated to be 743 mF/cm2, comparable to the values re-ported for CNT electrodes [108].

High-power microsupercapacitors with an interdigitat-ed structure have been reported using nanometer-size(approx. 7 nm) carbon onions as active materials [26].These carbon onions, with a specific surface area of 500m2/g, were first produced by annealing nanodiamondpowder at 1800 8C (Figure 10 b) and then deposited fromcolloidal suspensions using an electrophoretic deposition

Specia

lIs

sue

GR

AP

HE

NE

Fig. 8. (a) Schematic of the structure of G/CNTCs-MCs. Inset: enlarged scheme of Ni-G-CNTCs pillar structure that does not showthe Al2O3 atop the CNTCs; (b) SEM image of a fabricated G/CNTCs-MC, reprinted with permission from [121].

Fig. 9. Schematic of the fabrication of a microsupercapacitor integrated onto a silicon chip based on the bulk CDC film process.Standard photolithography techniques can be used for fabricating CDC capacitor electrodes (oxidative etching in oxygen plasma) anddeposition of gold current collectors, reprinted with permission from [74].

&12& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.

Page 13: A Review of Graphene-Based Electrochemical  Microsupercapacitors

technique onto interdigitated gold current collectors pat-terned on silicon wafers (see Figure 10c). Cyclic voltam-mograms were recorded with a scan rate as high as200 V/s. The microsupercapacitor maintained a specificcapacitance of 0.9 mF/cm2 at a scan rate of 100 V/s andvolumetric power densities that were comparable to elec-trolytic capacitors. Such high rate capabilities can be at-tributed to the unique zero-dimensional carbon onionstructure, which made the surface fully accessible to ionadsorption (see Figure 10 a). The large pore sizes (averagesize of 10 mm as indicated by SEM characterization inFigure 10e) after particle aggregation gives a lower inter-nal resistance for ion transport during high-rate charge/discharge processes. Chemical activation of these onion-like carbon materials using KOH or NaOH [42,123, 124]may further increase the specific surface area and thusthe volumetric capacitances of the microdevice.

In summary, carbon materials with high surface areaand large pore sizes have been extensively investigated aselectrode materials for microsupercapacitor applications.A summary of electrochemical performance of carbon-based microsupercapacitors is provided in Table 1. Be-cause device data in the literature have been recordedusing different setups (e.g., two- or three-electrode testingsystems) and calculations of capacitances and capacitiesoften are performed with respect to an active electrodemass, a full prototype device, or without describing thenormalization procedure, it is generally difficult to obtainan unambiguous comparison of all parameters of an elec-trode material or a device. In the table, we therefore sum-marize the data presented in the original reports in orderto make a comparison. As evident from Table 1, 3D elec-trodes (thickness greater than several tens of micron-meters) in general tend to provide better electrochemicalperformance in ion transport between electrodes and pro-duce a relatively higher energy density.

Future research in carbon-based microsupercapacitorswill likely emphasize electrodes designed to have a higher

specific surface area with a pore-size distribution compat-ible with the need for moderate surface modification.Above all, thicker electrodes within a fixed footprint area(i.e., 3D structures) are required to further optimize over-all capacitance, energy and power densities without com-promising stability.

4.2 Carbon and Pseudocapacitive Materials as FaradicElectrodes

Electric double-layer capacitors that rely on physical ionadsorption at the boundary between electrode and elec-trolyte will only give limited capacitance, typically in therange of 10–50 mF/cm2 [13,14]. Pseudocapacitance, how-ever, may be 10–100 times larger because of faradaiccharge transfer. Consequently, supercapacitors based onpseudocapacitive materials further increase energy andpower densities. If they additionally maintain a goodcyclic stability, they are highly desirable. The chargestored in such supercapacitors includes both non-faradaiccharge in the double-layer and faradaic charge, as activepseudocapacitive materials undergo fast and reversiblesurface redox reactions. To date, considerable effort hasbeen devoted to developing electrode materials for con-ventional supercapacitors that exhibit pseudocapacitance.Among these pseudocapacitive materials, conductingpolymers [126–130] and metal oxides [131–134] are themost frequently used. Recently, researchers have begunto incorporate these pseudocapacitive materials in micro-supercapacitor electrodes to further increase area-normal-ized capacitance as well as energy and power densities[135]. Among the metal oxides, RuO2 in its amorphoushydrous form (RuO2·xH2O) has been found to be an ex-cellent material for supercapacitor applications. However,faradaic reactions are confined to the outermost layersuch that a large portion of underlying RuO2·xH2O re-mains unreacted. Moreover, ruthenium-based electrodes

Specia

lIssu

eG

RA

PH

EN

E

Fig. 10. Design of the interdigitated microsupercapacitors with carbon onion electrodes, reprinted with permission from [26].

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de &13&

These are not the final page numbers! ��

Graphene-Based Electrochemical Microsupercapacitors

Page 14: A Review of Graphene-Based Electrochemical  Microsupercapacitors

are expensive and suffer from a diminished high-rate ca-pability [48].

Hydrous ruthenium dioxide (RuOxHy or RuO2·0.5H2O)on-chip microsupercapacitors were fabricated by a laserengineering approach under ambient temperature and at-mospheric conditions [136, 137]. Microsupercapacitorsbased on mixtures of sulfuric acid with the RuO2·0.5H2O

electrode material exhibits a specific capacitance of160 F/g, an energy density of 22 mWh/g and a power den-sity of 96.5 mW/g. Microsupercapacitors based on hy-drous RuO2 (hRuO2) and RuO2 nanorods shows a dis-charge capacitance of 40.7 mF/cm2 at a current of 5 mA,decreasing to 11.9 mF/cm2 when the current is at 75 mA[135].

Specia

lIs

sue

GR

AP

HE

NE

Table 1. Summary of carbon-based microsupercapacitors reported in contemporary literature.

Electrode(thickness)

Electrolyte Capacitance(mF/cm2)

Capacitance cyclestability

Energydensity

Powerdensity

Ref.

Activated carbon(5 mm) [a]

1 M Et4NBF4 inPC

11.6 – 10 mWh/cm3 ~40 W/cm3

[d][26]

Activated carbon(1–2 mm) [a]

1 M Et4NBF4 inPC

2.1 – 1.8 mWh/cm2 44.9 mW/cm2 [84]

Activated carbon(50–70 mm) [a]

1 M NaNO3 90.7 – – 51.5 mW/cm2 [78]

Activated carbon(336 mm) [a]

1 M Et4NBF4 inPC

81 – 71.4 mWh/cm2 34.4 mW/cm2 [70]

Photoresist-derivedcarbon(115–140 mm) [a]

0.5 M H2SO4 75 12.3% loss after 1000cycles

– – [92]

CNT arrays(80 mm) [a]

Ionic liquidBMIM-PF6

0.428 No capacitance lossafter 10 cycles

– – [108]

CNT array(34 mm) [a]

0.1 M NaSO4 36.5 F/g – ~0.4 Wh/kg [d] ~1 kW/kg[d]

[82]

Reduced GO(22 mm) [a]

Water-entrappedGO

0.51 30% loss after 10 000cycles

0.43 mWh/cm3 1.7 W/cm3 [83]

Reduced GO(22 mm) [b]

Water-entrappedGO

~0.25 35% loss after 10 000cycles

0.19 mWh/cm3 9.4 W/cm3 [83]

1–2 layer graphene [a] PVA/H3PO4 poly-mer gel

0.08 – 2.8 nWh/cm2 2 mW/cm2 [66]

Reduced GO film(10 nm) [a]

PVA/H3PO4 poly-mer gel

0.394 No capacitance lossafter 1500 cycles

14 nWh/cm2 9 mW/cm2 [66]

Reduced GO(7.6 mm) [a]

PVA/H2SO4 poly-mer gel

2.3 3% loss after 1000cycles

0.3 mWh/cm3

[d]70 W/cm3

[d][41]

Reduced GO(7.6 mm) [a]

FS-IL Ionogels 1.79 [c] No capacitance lossafter 30 000 cycles

1 mWh/cm3 [d] 200 W/cm3 [41]

Reduced GO/CNTs(6 mm) [a]

3 M KCl 6.1 little capacitance lossafter 1000 cycles

0.68 mWh/cm3 ~2.5 W/cm3

[d][120]

Graphene/CNTs(20 mm) [a]

1 M Na2SO4 2.16 – 0.16 mWh/cm3 115 Wh/cm3 [121]

Graphene/CNTs(20 mm) [a]

1-Butyl-3-methyli-midazolium tetra-fluoroborate

3.93 1.6 % capacitance lossafter 8000 cycles

2.42 mWh/cm3 135 Wh/cm3 [121]

Carbide-derived carbon(50 mm) [b]

1 M H2SO4 ~450 [c] – – – [74]

Carbide-derived carbon(50 mm) [b]

1 M TEABF4 ~300 [c] – – – [74]

Onion-like carbon(7 mm) [a]

1 M Et4NBF4 inPC

1.7 almost no loss after10 000 cycles

~2 mWh/cm3[d]

1000 W/cm3 [26]

Mesocarbon microbead(100 mm) [b]

Solid-state[BMIM][BF4]

100 47% loss after 8000cycles

10 mWh/cm2 0.575 mW/cm2 [125]

Super P carbon black(unknown thickness) [b]

1.5 M H2SO4 ~0.8 – – – [75]

[a] In-plane type. [b] Sandwich type. [c] Estimated from the given information in the literature. [d] Estimated from Ragone plots inthe literature.

&14& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.

Page 15: A Review of Graphene-Based Electrochemical  Microsupercapacitors

Unlike RuO2·xH2O, MnO2 has attracted much atten-tion as a potentially commercial pseudocapacitive elec-trode material because of its low cost, low toxicity andmost importantly, high theoretical specific capacitance(1370 F/g) [46]. Thick MnO2 layers were electrodepositedon CNT patterns on two different metal stack layers: Fe�Al/SiO2 and Fe�Al/Au/Ti/SiO2 [82]. The 20-nm-thick Aufilm was intended to reduce internal resistance, but it wasalso found to impede CNT growth since the CNTs werefound to grow randomly with an average length of only1.4 mm. Non-rectangular CV curves at a low scan rate(10 mV/s) showed a large internal resistance for bothcases (with and without a Au metal layer). Adding anoth-er Au layer did not alleviate the internal resistance prob-lem but definitely hindered the CNT growth and conse-quently limited the area-normalized capacitance, energydensity and power density. The high internal resistancemay result from excessive MnO2 deposition, blocking thepores for fast ion transport. In another study, MnO2/nano-fiber patterning was achieved by microfluidic etching formicrosupercapacitor electrodes [138]. The specific capaci-tance was approx. 25 mF/cm2 at a current density of0.5 mA/cm2. A large internal resistance was observed,and the device showed a cyclic stability characterized bya 4 % loss in capacitance after 500 cycles at 2 mA/cm2.

Conducting polymers, such as polypyrrole (PPy), poly-aniline (PANI) and polythiophene (PTP), have been fre-quently used as pseudocapacitive electrode materials inconventional supercapacitors [13,31, 139,140]. To date,the design and fabrication of microsupercapacitors basedon conducting polymers are in the early stages. The mostrecent reports of microsupercapacitors based on a con-ducting polymer is confined to PANI and PPy, which aredirectly coated on current collectors with limited specificsurface areas [71,72, 93, 141,142]. Conducting polymer-based microsupercapacitors predominately focus on elec-trochemically coating conducting polymers on metal cur-rent collectors, pre-patterned by conventional lithographytechniques. The advantage of this method is the ease infabrication, since PPy and polyaniline are electropolymer-izable and electroactive in aqueous media. However, par-allel fabrication issues require special consideration. Forinstance, when electrodepositing PANI on interdigitatedcurrent collectors, the aniline monomer exhibits a high af-finity to the underlying SiO2 substrate [141, 143], leadingto preferential PANI growth in a lateral direction. Thislateral growth bridges the gap between two adjacent elec-trodes causing an electrical short circuit. Thus post-treat-ment (e.g., oxygen plasma cleaning) may be needed toremove PANI to avoid the shorting problems.

The main limitation of conducting polymer-based su-percapacitors seem to be their poor cyclic stabilities[144–146], high self-discharge rates [34, 147], low capacitiesdue to the suboptimal doping [139, 140], and limited masstransport within thick polymer layers [139, 148, 149]. Onepossible solution to these issues is to coat a thin layer ofconducting polymer on a conducting template witha large specific area [34, 35,150–152]. Consequently, coat-

ing of conducting polymers on templates (e.g., graphene)might be an effective way to enhance electrochemicalperformance of microsupercapacitors.

A newly published study combined rGO and PANI asactive materials for electrodes [153]. The microelectrodeswere prepared by in situ electrodeposition of PANI nano-rods on the surface of rGO patterns fabricated by micro-molding in capillaries. The rGO patterns were found tobe relatively uniform, although aggregation of rGO in thepatterns along with defects caused by the closed channelat the end of each finger electrode was observed. After4500 s of PANI electrodeposition, dense PANI nanorods,with diameters of 20 nm and heights of 100 nm to 200 nm,covered the surface of rGO. This solution-based methodallows uniform and controllable in situ electrochemicalgrowth of conducting PANI nanorod arrays on patternedrGO thin films over large areas. The as-prepared microsu-percapacitors exhibited a specific capacitance of 970 F/gat a discharge current density of 2.5 A/g, as well as goodstability, retaining 90% of initial capacitance after 1700consecutive cycles.

Graphitic nanosheets (nanowalls) [154], or graphiticpetals (GPs), containing a few layers of graphitic carbonand growing roughly perpendicularly to a substrate overa large surface area offer many advantages as active elec-trode materials because of their high specific area andhigh electrical conductivity. The formation of petals re-quires a plasma environment, different from non-plasmaCVD of planar graphene. The petals can be directlygrown on different substrates (e.g., Ni foil, carbon cloth,carbon nanotubes) without any binder for conventionalsupercapacitor applications [155–158]. Recently, con-trolled growth of these graphitic petals on insulating sub-strates was demonstrated by a simple scribing method[159], providing an efficient means to pattern GP interdi-gitated electrodes for on-chip planar microsupercapacitorapplications. The interdigitated GP nanostructures canserve either as independent electrodes or templates fordeposition of pseudocapacitive materials in microsuperca-pacitor applications.

Table 2 provides a summary of microsupercapacitorperformance based on pseudocapacitive materials report-ed in the literature so far. Similar to the practice followedin Table 1 for carbon-based microsupercapacitors, wechoose to summarize the data presented in the originalreports rather than translate the results into a commonset of units.

5 Solid-State and Flexible Graphene-BasedMicrosupercapacitors

Conventional supercapacitors often use electrolytes in theform of liquid (either aqueous or organic-based solution).A consequence of this choice is the risk of electrolyteleakage during use. Consequently, high-level safety pack-aging techniques have been developed in which the elec-trolyte seal is robust and leak-proof but the encapsulation

Specia

lIssu

eG

RA

PH

EN

E

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de &15&

These are not the final page numbers! ��

Graphene-Based Electrochemical Microsupercapacitors

Page 16: A Review of Graphene-Based Electrochemical  Microsupercapacitors

is bulky. This is not an important issue in conventionalmacroscale applications, where there is no overriding sizeconstraint. The situation is different in microscale systemswhere the space constraint for energy storage devices canbe severe.

To date, electrode patterns are typically confined toseveral square millimeters with a micrometer-size spacingbetween electrodes that can be readily achieved usingwell-developed microfabrication technique like photoli-thography. However, the development of advanced pack-aging techniques has not kept pace with the size reduc-tion of the micropatterned electrodes. As a result, mostexisting literature on microsupercapacitors reports theenergy performance in liquid electrolyte without encapsu-lation. To enable microsupercapacitor technology, a solid-state device configuration is necessary for microscaledenergy storage components to be directly integrated ontochips. The key task to achieving a solid-state microsuper-capacitor lies in the development of a solid-state electro-lyte. Several pioneering studies have appeared to addressthis need.

An early work on the development of a sandwich-typesolid-state microsupercapacitor was carried out by Hoand co-workers [163]. A porous carbon material and anionic liquid-based solid-state electrolyte were adopted.They demonstrated a direct-write, printing method to ad-ditively fabricate solid-state, thick-film microsupercapaci-tors directly on a substrate at room temperature underambient conditions. First, electrode and electrolyte pastesolutions were prepared in N-methyl pyrrolidone (NMP)according to the following component proportions:50wt% mesophase microbead (MCMB) carbon material(size range 6–28 mm), 2 wt% acetylene black (AB) as con-ductive additive, 24wt% polyvinylidene difluoride(PVDF) as polymer binder, and 24wt% 1-butyl-3-methyl-imidazolium tetrafluoroborate (BMIM+BF4

�) ionic liquidelectrolyte for electrode pastes; 50wt% PVDF and50wt.% BMIM+BF4

� ionic liquid for the electrolytepastes. A custom-built printer consisting of a pressureregulator and a 3-axis stage with micron resolution wasdeveloped for the direct writing of microsupercapacitors.A typical microsupercapacitor consists of a bottom

Specia

lIs

sue

GR

AP

HE

NE

Table 2. Summary of pseudocapacitive microsupercapacitor performance.

Electrode(thickness)

Electrolyte Capacitance(mF/cm2)

Capacitance cyclestability

Energydensity

Powerdensity

Ref.

Hydrous RuO2

(10 mm) [a]0.5 M H2SO4 174 F/g – 23 Wh/kg 96.5 W/kg [160]

Hydrous RuO2

(1.9 mm) [a]0.5 M H2SO4 40.7 – ~1.1 mWh/cm2

[e]~0.4 mW/cm2

[e][135]

W-RuO2

(1 mm) [b]LiPON 54.2 – – – [161]

MnO2/CNT array(34 mm) [a]

0.1 M NaSO4 176 F/g – 10.3 Wh/kg 960 W/kg [82]

MnO2 nanofiber(1 mm) [a]

Solid-state H3PO4-PVA

341.4 F/g 4% loss after 500cycles

– – [138]

PPy vs. PPy [a] 0.1 M H3PO4 ~1.6–14 [c] – – – [141]

PPy vs. PPy [a] 0.5 M Et4NBF4 inACN

~3.9 [c] – – – [141]

PPy vs. PPT [a] 0.5 M Et4NBF4 inACN

~5.2 [c] – – – [141]

PPy on 3D Si(150 mm) [a]

1 M KCl 30 almost no loss after800 cycles

– – [162]

PPy on 3D Si(150 mm) [a]

Solid-state LiClO4-PVA

29 – – 2.2 mW/cm2 [71]

PPy on 3D Si beams(unknown) [a]

0.5 M NaCl electro-lytes

56 – – 0.56 mW/cm2 [72]

PPy/PR-derived carbon(140 mm) [a]

1 M KCl 78.35 44% loss after 1000cycles

– 0.63 mW/cm2 [93]

PANI nanowire(400 nm) [a]

Solid-state H2SO4-PVA

~23.5 [d] 4% loss after 1000cycles

2.9mWh/cm2 [d] 50 mW/cm2

[d][142]

PANI nanorod/rGO(100–200 nm) [a]

Solid-state H3PO4-PVA

970 F/g 10% loss after 1700cycles

– – [153]

[a] In-plane type. [b] Sandwich type. [c] Estimated from the capacitance and effective device area in the literature. [d] Estimated fromthe specific volumetric capacitance and device dimension in the literature. [e] Estimated from the value and device dimension in theliterature.

&16& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.

Page 17: A Review of Graphene-Based Electrochemical  Microsupercapacitors

carbon electrode film (thickness approx. 30 mm), a gelelectrolyte separator (thickness 15–30 mm) and a topcarbon electrode film (thickness approx. 30 mm) ona stainless steel substrate within a 1 cm � 1 cm footprint.This ionic liquid-based microsupercapacitor exhibiteda high breakdown voltage of 1.7 V. At a current densityof 0.1 mA/cm2, the specific capacitance of the microsuper-capacitor increased substantially within the first 30 000cycles, and then reached a stable value of 0.55 mF/cm2 forthe next 70000 cycles. Further work on integration ofsuch a microsupercapacitor with a MEMS piezoelectricvibration energy harvester on the same chip was alsodemonstrated [125]. An improved specific capacitance of53 mF/cm2 was achieved using thicker electrodes. It showsan energy density and power density of 10 mWh/cm2 and0.575 mW/cm2, respectively. Although the performance ofthe MEMS and microsupercapacitors were tested sepa-rately, without any electrical integration of the energyharvesting and energy storage components, this workdemonstrates the feasibility of integration of a microsuper-capacitor with a MEMS on the same chip.

To increase the device specific capacitance, Sun andChen [71] designed and fabricated microsupercapacitorsbased on PPy films polymerized on 3D interdigital Si/SiO2/Ni electrode [162]. LiClO4-PVA solid-state electro-lyte was coated on the electrode substrate to forma solid-state microsupercapacitor. At a low current loadof 0.5 mA/cm2, the device shows an area-normalized ca-pacitance of 125 mF/cm2. As the current load increases to5 mA/cm2, the device can deliver a power density of2.2 mW/cm2 with a lower specific capacitance of 29 mF/cm2.

Microenergy storage with the added functionality ofmechanical flexibility creates a new requirement toenergy storage that extends power sources beyond tradi-tional use, e.g., to enable a wearable electronic device.Not surprisingly, a significant effort has arisen to develop

a solid-state, mechanically flexible microsupercapacitor.In prior work, flexible solid-state microsupercapacitorswere fabricated with conducting polymers (e.g., PANI[142], PPy [73,141,164], and PTP [141]) or metal oxides(e.g., MnO2 [165]) as electrodes because such electrodematerials could be deposited on patterned current collec-tors electroplated on flexible polymer substrates. Thengel polymer electrolyte is coated on the patterned elec-trodes to form a flexible solid-state microsupercapacitor.However, conducting polymer and metal oxide-basedelectrodes have the inevitable physical limitation of poorcharge/discharge cycling performance and inferior electri-cal conductivity (see Section 4.2). Future research shouldfocus on the addition of porous graphene-based materialsas templates to further exploit the electrochemical prop-erties of the pseudocapacitive materials in flexible micro-supercapacitors.

Xue and co-workers [153] fabricated flexible solid-statemicrosupercapacitors by in situ electrodeposition ofPANI nanorods on the surface of rGO patterns that areprepared by micromolding in capillaries (MIMIC). Thefabrication procedure is summarized in Figure 11. First,a flat PDMS substrate and a patterned PDMS stampwere pressed together to achieve conformal contact. Adrop of well-dispersed GO aqueous solution was placedat each end of the PDMS stamp. After heating undervacuum with hydrazine solution and then degassing for30 min at room temperature, the system was thenwarmed to 70 8C for 12 h to form the patterned rGO elec-trodes. PANI was electrodeposited on the patterned GOinterdigital electrodes using a three-electrode electro-chemical system at a constant potential of 0.75 V vs. Ag/AgCl in a 0.05 M aniline solution with 0.5 M H2SO4 asthe supporting solution. H3PO4-PVA gel electrolyte wasused to coat the microsupercapacitor device to achievea solid-state flexible microsupercapacitor. Compared withmicrosupercapacitors that employ PANI deposited on

Specia

lIssu

eG

RA

PH

EN

E

Fig. 11. Schematic diagram of fabrication process for rGO/PANI microelectrodes, reprinted with permission from [153].

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de &17&

These are not the final page numbers! ��

Graphene-Based Electrochemical Microsupercapacitors

Page 18: A Review of Graphene-Based Electrochemical  Microsupercapacitors

bare gold microelectrodes, the PANI/rGO-based microde-vice showed slightly higher specific capacitance (970 F/gat a current density of 2.5 A/g, as compared with approx.940 F/g for Au/PANI), better rate capability, and en-hanced cycle stability (10 % capacitance decay after 1700charge/discharge cycles, as compared with 20 % after lessthan 500 cycles for Au/PANI), which was ascribed to thesynergistic effect of rGO and PANI nanorod arrays.

Instead of using conventional microfabrication tech-niques (e.g., photolithography in clean room) that requirehigh standard fabrication environment and complicatedmultiple fabrication processes, Kaner�s group [41] demon-strated a scalable fabrication of flexible all-solid-state mi-crosupercapacitors by a simple direct laser writing on gra-phene oxide films using a standard LightScribe DVDburner, with a spatial resolution of 20 mm. Briefly, 16 mLof aqueous GO dispersion (2.7 mg/mL) was drop-cast onPET bonded on the surface of a media disc. After beingdried overnight under ambient conditions, a uniform GOlayer was formed on PET surface. The GO-coated discwas subsequently inserted into a DVD drive for laserscribing to obtain designed interdigitated micropatterns.A flexible all-solid-state microsupercapacitor was ach-ieved after dropping a solid-state electrolyte overcoat(e.g., PVA-H2SO4 or FS-IL Ionogels). Herein, the laser-scribed graphene patterns serve as both an active materialand a current collector while the remaining GO serves asspacing between two electrodes. As-prepared flexible mi-crosupercapacitors exhibit good energy storage perfor-mance (see Table 1).

In summary, the development of solid-state graphene-based microsupercapacitors is still in early stages. Design-ing a composite electrode composed of highly porous gra-phene-based materials plus a pseudocapacitive additiveremains a goal to improve the electrochemical propertiesof the microdevice. In order to solve electrolyte leakageproblems associated with packaging, researchers haveused different solid-state electrolytes to fabricate gra-phene-based solid-state microsupercapacitors with littleor no encapsulation. By integrating devices on flexiblesubstrates (e.g., polymer-based films), researchers haveachieved new types of solid-state microsupercapacitorswith a mechanical flexibility which may find use in flexi-ble thin film microenergy storage devices. Meanwhile,predictable issues surrounding the use of solid-state elec-trolytes, such as larger internal resistance and inferiorrate capability compared with their liquid counterparts,still require attention. Future research on flexible micro-supercapacitors should focus on the design of new, high-performance electrode materials and better solid-stateelectrolytes.

6 Fabrication Techniques for Graphene-BasedElectrodes

This section summarizes various techniques that havebeen used in the fabrication of microsupercapacitors.

First and foremost, progress in micro and nanofabricationtechniques provides a scalable basis for fabricating micro-supercapacitor electrodes using conventional lithography-based techniques. These techniques are often selectedwhen the patterned graphene-based materials are thinand require a precise separation between two adjacentelectrodes. In order to pattern thick active materials asmicroelectrodes, selective etching of active materials isperformed using metal masks patterned by conventionaloptical lithography techniques. In this case, the gap widthbetween two adjacent electrodes can be easily adjustedby controlling the optical lithography parameters. As anexample, 200-mm-thick electrodes of carbide-derivedcarbon films have been fabricated for microsupercapaci-tors using patterned metal masks on top of the active ma-terials [74]. However, this technique might be not appli-cable when the substrate surface is rough or porous (e.g.,free-standing porous CNT network). Effective etchingmethods to pattern active materials should also be takeninto consideration. For instance, an impractically longtime of oxygen plasma etching might be needed to pat-tern a 200-mm-thick graphite foil to fabricate microelectr-odes.

Inkjet printing technology has proven efficient whenpatterning liquid precursor materials such as structuralpolymers, conducting polymers, sol-gel materials, ceram-ics, nanoparticles, nucleic acid and protein arrays for thefabrication of electronic devices, sensors, and the func-tionalization of biomedical surfaces [166–168]. Inkjetprinting offers the following advantages: (i) short process-ing time, (ii) low capital and production costs, (iii) applic-ability to non-planar substrates, (iv) ease in processing,particularly when compared to photolithographic tech-niques, and (v) an easy path to meet industrial scale-upneeds. For these reasons, inkjet printing is considered tobe simpler, more environmentally friendly and cost effec-tive than vacuum-based methods [166, 169]. Inkjet print-ing has been successfully used to print conducting metalpatterns from Ag [170, 171], Pd [172], Au [173], Pt [174],Cu [175], and conducting polymers [176]. As an example,to investigate the potential of inkjet printing for nanoe-lectronic applications, Bhuvana and co-workers [172] suc-cessfully used inkjet printing methods to pattern a net-work of SWCNTs by printing two layers of an aqueoussolution of SWCNTs wrapped with single-stranded DNA.Pech et al. [84] prepared an ink of activated carbonpowder with a PTFE polymer binder in ethylene glycolstabilized with a surfactant. It was then possible to depos-it this material on patterned gold current collectors withthe substrate heated at 140 8C in order to assure a goodhomogeneity. Various microsupercapacitors were de-signed with this technique with interdigitated fingersranging from 40 to 100 mm width. However, commonproblems associated with inkjet methods include largedrops of ink that preclude the ability to print fine interdi-gitated fingers, limited precision for narrow electrodesand gaps, and coalescence of drops affecting the printquality.

Specia

lIs

sue

GR

AP

HE

NE

&18& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.

Page 19: A Review of Graphene-Based Electrochemical  Microsupercapacitors

Femtosecond lasers have been widely used for produc-ing micron features and 3D microdevices in many fieldsdue to their advantages of nanometer spatial resolutionand 3D prototyping capability [177–179]. The resolutionrequired to create precise microdevices has continuallyimproved [180,181]. A laser “forward transfer” techniquehas been used to fabricate conformal, mesoscale passiveelectronic components, including metal interconnects,multilayer capacitors, inductors, and resistors. Using thistransfer technique, Pique et al. [182] demonstrated theability to rapidly prototype temperature, biological andchemical sensor devices. This matrix assisted pulsed laserevaporation direct-write process is compatible witha broad class of materials such as metals, polymers, bio-materials or composites as multilayers or discrete struc-tures on a single substrate. A direct femtosecond laser re-duction process to make graphene-based electronic mi-crocircuits on graphene oxide films has been demonstrat-ed [183]. Free-standing and flexible microsupercapacitorson a GO film were fabricated based on the laser writingtechniques [83]. Laser writing has also been used to fabri-cate microsupercapacitors at a large scale by a standardLightScribe DVD burner [41]. A laser drive can write de-sired graphene circuits onto a GO film following comput-er-designed patterns. Various microdevices with differentsizes and shapes can be produced on a single run. Thismethod would be promising for large-scale fabrication ofmicrosupercapacitors if the cost of femtosecond laserequipment can be reduced.

Micromolding in capillary (MIMIC) has been used tofabricate microstructures of organic polymers, inorganicand organic salts, ceramics, metals, and crystalline micro-particles in many different kinds patterns [184, 185]. Theprotocols underlying this method are discussed elsewhere[184, 186]. The technique relies on the spontaneous fillingof channels with a fluid by capillary action, in which therate and the extent of filling are determined by the bal-ance between interfacial thermodynamics and viscositydrag [185]. The merits of this method in fabricating mi-cropattern electrodes are: (i) the fabrication of a mold inMIMIC is simple; it requires only the conformal contactof a substrate with an elastomeric mold; (ii) only limited(and in some cases no) access to facilities for lithography;and (iii) the production of multiple copies of an elasto-meric component from a single lithographic master. How-ever, this method is limited to low-viscosity liquids. Thisliquid-based process can be an effective alternative tofabricate microsupercapacitor electrodes. Large-scale mi-cropatterns of continuously conductive rGO films thatare centimeters in length and micrometers in width onvarious substrates were fabricated using the micromold-ing and their capabilities in a sensing application weredemonstrate in a recent study [186]. Microsupercapaci-tors based on rGO micropatterns fabricated by thismethod were reported recently [153].

7 Conclusion and Outlook

Fabrication of 3D electrochemical microsupercapacitorsis relatively new and rapidly growing endeavor whencompared to more conventional supercapacitors. For thisreason, it is worthwhile to review the primary scientificliterature to learn trends and identify existing bench-marks. Particular emphasis was placed here on the latestdevelopments of carbon-based materials used to fabricatesolid-state (both flexible and rigid) microsupercapacitors.A major challenge remains to increase the thickness ofthe active materials (i.e. , 3D structures) in order to in-crease the specific capacitance and energy density of a mi-crosupercapacitor, without sacrificing the cyclic stabilityand power densities in a given footprint for future micro-supercapacitor design. Material constraints are requiredbecause the long-term goal is to produce on-chip devices.A number of directions can be identified that require fur-ther improvement.

(i) Electrode materials. Integrating new carbon nanoma-terials such as CNTs and graphene into microsupercapaci-tors would be a good choice to improve specific surfacearea, capacitance and energy storage. These nanomateri-als not only can be directly used as electrodes but alsocould be used as nanotemplates for pseudocapacitive ma-terials to further increase their utilization efficiency andmore importantly to solve the long-term cyclic problemassociated with volume change and swelling during theion doping/undoping process.

(ii) Fabrication and integration techniques. New tech-niques to fabricate and integrate 3D microelectrodesneed further development (e.g., how to transfer and inte-grate active materials onto temperature sensitive sub-strates). A cost analysis of the wide variety of fabricationtechniques already in use is required before large-scaleapplications are possible.

(iii) Electrolyte development and packaging issues.Solid-state electrolytes define a new trend for the fabrica-tion and packing of multifunctional (e.g., flexible andtransparent) microsupercapacitors. More choices forsolid-state electrolytes that are fast ion conductors are re-quired so that the cell voltage of a microdevice can behigher than 4 V to further increase energy density. Solid-state electrolytes also hold promise to solve packagingissues confronting microscale supercapacitors. Lastly, newpackaging techniques are needed in order for microsuper-capacitors to be useful in practical applications.

Acknowledgements

The authors thank the U.S. Air Force Research Laborato-ry (AFRL), and its Office of Scientific Research (AFOSR)under the MURI Program on Nanofabrication of Tunable3D Nanotube Architectures (PM: Dr. Joycelyn Harrison),for financial support in this work.

Specia

lIssu

eG

RA

PH

EN

E

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de &19&

These are not the final page numbers! ��

Graphene-Based Electrochemical Microsupercapacitors

Page 20: A Review of Graphene-Based Electrochemical  Microsupercapacitors

References

[1] B. E. Conway, Electrochemical Supercapacitors: ScientificFundamentals and Technological Applications, KluwerAcademic/Plenum Publishers, New York 1999.

[2] P. Simon, Y. Gogotsi, Nat. Mater. 2008, 7, 845.[3] H. E. Becker, U.S. Patent, 1957, 2 800 616 (to General

Electric).[4] D. I. Boos, U.S. Patent, 1970, 3 536 963 (TO Standard Oil,

SOHIO).[5] B. E. Conway, J. Electrochem. Soc. 1991, 138, 1539.[6] R. Kotz, M. Carlen, Electrochim. Acta 2000, 45, 2483.[7] J. S. Huang, B. G. Sumpter, V. Meunier, Chem. Eur. J.

2008, 14, 6614.[8] P. Wu, J. Huang, V. Meunier, B. G. Sumpter, R. Qiao, ACS

Nano 2011, 5, 9044.[9] G. A. Feng, R. Qiao, J. S. Huang, B. G. Sumpter, V. Meuni-

er, J. Phys. Chem. C 2010, 114, 18012.[10] L. Yang, B. H. Fishbine, A. Migliori, L. R. Pratt, J. Am.

Chem. Soc. 2009, 131, 12373.[11] S. Wang, S. Li, Z. Cao, T. Y. Yan, J. Phys. Chem. C 2010,

114, 990.[12] Y. Shim, H. J. Kim, ACS Nano 2010, 4, 2345.[13] G. Wang, L. Zhang, J. Zhang, Chem. Soc. Rev. 2012, 41,

797.[14] A. G. Pandolfo, A. F. Hollenkamp, J. Power Sources 2006,

157, 11.[15] P. Mohseni, K. Najafi, S. J. Eliades, X. Q. Wang, IEEE T.

Neur. Sys. Reh. 2005, 13, 263.[16] A. Nieder, J. Neurosci. Meth. 2000, 101, 157.[17] M. Hautefeuille, C. O’Mahony, B. O’Flynn, K. Khalfi, F.

Peters, Microelectron. Reliab. 2008, 48, 906.[18] F. Albano, Y. S. Lin, D. Blaauw, D. M. Sylvester, K. D.

Wise, A. M. Sastry, J. Power Sources 2008, 185, 1524.[19] M. Beidaghi, C. L. Wang, P. Soc. Photo-Opt. Ins. 2010,

7679, 76791G1.[20] S. D. Jones, J. R. Akridge, Solid State Ionics 1996, 86–8,

1291.[21] M. Beidaghi, W. Chen, C. L. Wang, Proc. SPIE. 2011,

8035, 80350J1.[22] J. D. Morse, Int. J. Energ. Res. 2007, 31, 576.[23] K. A. Cook-Chennault, N. Thambi, A. M. Sastry, Smart

Mater. Struct. 2008, 17, 043001.[24] J. Hassoun, S. Panero, P. Reale, B. Scrosati, Adv. Mater.

2009, 21, 4807.[25] J. W. Long, B. Dunn, D. R. Rolison, H. S. White, Chem.

Rev. 2004, 104, 4463.[26] D. Pech, M. Brunet, H. Durou, P. H. Huang, V. Mochalin,

Y. Gogotsi, P. L. Taberna, P. Simon, Nat. Nanotechnol.2010, 5, 651.

[27] E. Frackowiak, F. Beguin, Carbon 2001, 39, 937.[28] E. Frackowiak, Phys. Chem. Chem. Phys. 2007, 9, 1774.[29] L. L. Zhang, R. Zhou, X. S. Zhao, J. Mater. Chem. 2010,

20, 5983.[30] A. K. Shukla, S. Sampath, K. Vijayamohanan, Curr. Sci.

India 2000, 79, 1656.[31] H. Li, J. Wang, Q. Chu, Z. Wang, F. Zhang, S. Wang, J.

Power Sources 2009, 190, 578.[32] R. Y. Song, J. H. Park, S. R. Sivakkumar, S. H. Kim, J. M.

Ko, D. Y. Park, S. M. Jo, D. Y. Kim, J. Power Sources 2007,166, 297.

[33] G. H. Yu, L. B. Hu, M. Vosgueritchian, H. L. Wang, X.Xie, J. R. McDonough, X. Cui, Y. Cui, Z. N. Bao, NanoLett. 2011, 11, 2905.

[34] C. Meng, C. Liu, L. Chen, C. Hu, S. Fan, Nano Lett. 2010,10, 4025.

[35] C. Meng, C. Liu, S. Fan, Electrochem. Commun. 2009, 11,186.

[36] H. A. Andreas, K. Lussier, A. M. Oickle, J. Power Sources2009, 187, 275.

[37] J. J. Niu, W. G. Pell, B. E. Conway, J. Power Sources 2006,156, 725.

[38] B. Pillay, J. Newman, J. Electrochem. Soc. 1996, 143, 1806.[39] J. R. Miller, Science 2012, 335, 1312.[40] Y. Gogotsi, P. Simon, Science 2011, 334, 917.[41] M. F. El-Kady, R. B. Kaner, Nat. Commun. 2013, 4, 1475,

doi:10.1038/ncomms2446.[42] K. Babel, K. Jurewicz, J. Phys. Chem. Solids 2004, 65, 275.[43] A. Izadi-Najafabadi, S. Yasuda, K. Kobashi, T. Yamada,

D. N. Futaba, H. Hatori, M. Yumura, S. Iijima, K. Hata,Adv. Mater. 2010, 22, E235.

[44] M. Kaempgen, C. K. Chan, J. Ma, Y. Cui, G. Gruner, NanoLett. 2009, 9, 1872.

[45] M. D. Stoller, S. J. Park, Y. W. Zhu, J. H. An, R. S. Ruoff,Nano Lett. 2008, 8, 3498.

[46] M. Toupin, T. Brousse, D. Belanger, Chem. Mater. 2004,16, 3184.

[47] A. E. Fischer, K. A. Pettigrew, D. R. Rolison, R. M.Stroud, J. W. Long, Nano Lett. 2007, 7, 281.

[48] B. J. Lee, S. R. Sivakkumar, J. M. Ko, J. H. Kim, S. M. Jo,D. Y. Kim, J. Power Sources 2007, 168, 546.

[49] R. F. Zhou, C. Z. Meng, F. Zhu, Q. Q. Li, C. H. Liu, S. S.Fan, K. L. Jiang, Nanotechnology 2010, 21, 345701–1.

[50] W. Tang, L. L. Liu, S. Tian, L. Li, Y. B. Yue, Y. P. Wu, K.Zhu, Chem. Commun. 2011, 47, 10058.

[51] H. H. Chang, C. K. Chang, Y. C. Tsai, C. S. Liao, Carbon2012, 50, 2331.

[52] Y.-Y. Horng, Y.-C. Lu, Y.-K. Hsu, C.-C. Chen, L.-C. Chen,K.-H. Chen, J. Power Sources 2010, 195, 4418.

[53] M. Mastragostino, C. Arbizzani, F. Soavi, J. Power Sources2001, 97–8, 812.

[54] A. Burke, J. Power Sources 2000, 91, 37.[55] S. B. Ma, K. W. Nam, W. S. Yoon, X. Q. Yang, K. Y. Ahn,

K. H. Oh, K. B. Kim, Electrochem. Commun. 2007, 9, 2807.[56] P. C. Chen, G. Z. Shen, Y. Shi, H. T. Chen, C. W. Zhou,

ACS Nano 2010, 4, 4403.[57] V. Ganesh, S. Pitchumani, V. Lakshminarayanan, J. Power

Sources 2006, 158, 1523.[58] V. Khomenko, E. Raymundo-Pinero, E. Frackowiak, F.

Beguin, Appl. Phys. A 2006, 82, 567.[59] J. H. Chae, K. C. Ng, G. Z. Chen, P. I. Mech. Eng. A, J.

Pow. 2010, 224, 479.[60] B. E. Conway, W. G. Pell, J. Solid State Electron. 2003, 7,

637.[61] M. F. Rose, C. Johnson, T. Owens, B. Stephens, J. Power

Sources 1994, 47, 303.[62] A. M. Stephan, Eur. Polym. J. 2006, 42, 21.[63] C. A. Vincent, Prog. Solid. State. Chem. 1987, 17, 145.[64] S. Nohara, H. Wada, N. Furukawa, H. Inoue, M. Morita, C.

Iwakura, Electrochim. Acta 2003, 48, 749.[65] M. Morita, J. L. Qiao, N. Yoshimoto, M. Ishikawa, Electro-

chim. Acta 2004, 50, 837.[66] J. J. Yoo, K. Balakrishnan, J. S. Huang, V. Meunier, B. G.

Sumpter, A. Srivastava, M. Conway, A. L. M. Reddy, J. Yu,R. Vajtai, P. M. Ajayan, Nano Lett. 2011, 11, 1423.

[67] Y. J. Kang, H. Chung, C. H. Han, W. Kim, Nanotechnology2012, 23, 065401.

[68] C. C. Ho, D. A. Steingart, ECS Trans. 2008, 16, 35.[69] M. Esashi, J. Micromech. Microeng. 2008, 18[70] H. Durou, D. Pech, D. Colin, P. Simon, P. L. Taberna, M.

Brunet, Microsyst. Technol. 2012, 18, 467.[71] W. Sun, X. Y. Chen, Microelectron. Eng. 2009, 86, 1307.

Specia

lIs

sue

GR

AP

HE

NE

&20& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.

Page 21: A Review of Graphene-Based Electrochemical  Microsupercapacitors

[72] W. Sun, R. L. Zheng, X. Y. Chen, J. Power Sources 2010,195, 7120.

[73] J. H. Sung, S. Kim, K. H. Lee, J. Power Sources 2004, 133,312.

[74] J. Chmiola, C. Largeot, P. L. Taberna, P. Simon, Y. Gogotsi,Science 2010, 328, 480.

[75] H. J. In, S. Kumar, Y. Shao-Horn, G. Barbastathis, Appl.Phys. Lett. 2006, 88, 083104.

[76] H. X. Ji, Y. F. Mei, O. G. Schmidt, Chem. Commun. 2010,46, 3881.

[77] C. G. Zoski, Electroanalysis 2002, 14, 1041.[78] C. W. Shen, X. H. Wang, W. F. Zhang, F. Y. Kang, J. Power

Sources 2011, 196, 10465.[79] D. R. Rolison, R. W. Long, J. C. Lytle, A. E. Fischer, C. P.

Rhodes, T. M. McEvoy, M. E. Bourga, A. M. Lubers,Chem. Soc. Rev. 2009, 38, 226.

[80] D. Golodnitsky, V. Yufit, M. Nathan, I. Shechtman, T. Ri-penbein, E. Strauss, S. Menkin, E. Peled, J. Power Sources2006, 153, 281.

[81] B. E. Conway, W. A. Adams, J. de Oliveira, W. G. Pell, J.Power Sources 1999, 80, 134.

[82] C. C. Liu, D. S. Tsai, W. H. Chung, K. W. Li, K. Y. Lee,Y. S. Huang, J. Power Sources 2011, 196, 5761.

[83] W. Gao, N. Singh, L. Song, Z. Liu, A. L. M. Reddy, L. J.Ci, R. Vajtai, Q. Zhang, B. Q. Wei, P. M. Ajayan, Nat.Nanotechnol. 2011, 6, 496.

[84] D. Pech, M. Brunet, P. L. Taberna, P. Simon, N. Fabre, F.Mesnilgrente, V. Conedera, H. Durou, J. Power Sources2010, 195, 1266.

[85] D. Y. Qu, J. Power Sources 2002, 109, 403.[86] D. Y. Qu, H. Shi, J. Power Sources 1998, 74, 99.[87] H. Shi, Electrochim. Acta 1996, 41, 1633.[88] C. Emmenegger, P. Mauron, P. Sudan, P. Wenger, V. Her-

mann, R. Gallay, A. Zuttel, J. Power Sources 2003, 124,321.

[89] A. Alonso, V. Ruiz, C. Blanco, R. Santamaria, M. Granda,R. Menedez, S. G. E. de Jager, Carbon 2006, 44, 441.

[90] M. Nakamura, M. Nakanishi, K. Yamamoto, J. PowerSources 1996, 60, 225.

[91] T. Brousse, M. Toupin, D. Belanger, J. Electrochem Soc2004, 151, A614.

[92] M. Beidaghi, W. Chen, C. L. Wang, J. Power Sources 2011,196, 2403.

[93] M. Beidaghi, C. L. Wang, Electrochim. Acta 2011, 56, 9508.[94] W. Chen, M. Beidaghi, V. Penmatsa, K. Bechtold, L.

Kumari, W. Z. Li, C. L. Wang, IEEE T. Nanotechnol. 2010,9, 734.

[95] E. Frackowiak, K. Jurewicz, S. Delpeux, F. Beguin, J.Power Sources 2001, 97–8, 822.

[96] H. Zhang, G. P. Cao, Z. Y. Wang, Y. S. Yang, Z. J. Shi,Z. N. Gu, Nano Lett. 2008, 8, 2664.

[97] H. Zhang, G. Cao, Z. Wang, Y. Yang, Z. Shi, Z. Gu, Elec-trochem. Commun. 2008, 10, 1056.

[98] S. Iijima, Nature 1991, 354, 56.[99] R. Saito, G. Dresselhaus, M. S. Dresselhaus, J. Appl. Phys.

1993, 73, 494.[100] M. S. Dresselhaus, G. Dresselhaus, R. Saito, Solid State

Commun. 1992, 84, 201.[101] T. W. Ebbesen, H. J. Lezec, H. Hiura, J. W. Bennett, H. F.

Ghaemi, T. Thio, Nature 1996, 382, 54.[102] M. M. J. Treacy, T. W. Ebbesen, J. M. Gibson, Nature 1996,

381, 678.[103] J. P. Issi, L. Langer, J. Heremans, C. H. Olk, Carbon 1995,

33, 941.[104] H. Zhang, G. P. Cao, Y. S. Yang, J. Power Sources 2007,

172, 476.

[105] G. Salitra, A. Soffer, L. Eliad, Y. Cohen, D. Aurbach, J.Electrochem. Soc. 2000, 147, 2486.

[106] J. Chmiola, G. Yushin, Y. Gogotsi, C. Portet, P. Simon,P. L. Taberna, Science 2006, 313, 1760.

[107] J. Chmiola, C. Largeot, P. L. Taberna, P. Simon, Y. Gogotsi,Angew. Chem. Int. Ed. 2008, 47, 3392.

[108] Y. Q. Jiang, Q. Zhou, and L. Lin, IEEE 22nd Int. Conf. onMEMS 2009, 587.

[109] A. K. Shukla, T. P. Kumar, Curr. Sci. India 2008, 94, 314.[110] J. Chen, Y. Liu, A. I. Minett, C. Lynam, J. Z. Wang, G. G.

Wallace, Chem. Mater. 2007, 19, 3595.[111] K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y.

Zhang, S. V. Dubonos, I. V. Grigorieva, A. A. Firsov, Sci-ence 2004, 306, 666.

[112] X. Huang, X. Y. Qi, F. Boey, H. Zhang, Chem. Soc. Rev.2012, 41, 666.

[113] H. R. Byon, S. W. Lee, S. Chen, P. T. Hammond, Y. Shao-Horn, Carbon 2011, 49, 457.

[114] J. Yan, Z. J. Fan, T. Wei, W. Z. Qian, M. L. Zhang, F. Wei,Carbon 2010, 48, 3825.

[115] C. G. Liu, Z. N. Yu, D. Neff, A. Zhamu, B. Z. Jang, NanoLett. 2010, 10, 4863.

[116] S. Cerveny, F. Barroso-Bujans, A. Alegria, J. Colmenero, J.Phys. Chem. C 2010, 114, 2604.

[117] V. Strong, S. Dubin, M. F. El-Kady, A. Lech, Y. Wang,B. H. Weiller, R. B. Kaner, ACS Nano 2012, 6, 1395.

[118] M. F. El-Kady, V. Strong, S. Dubin, R. B. Kaner, Science2012, 335, 1326.

[119] I. Jung, D. A. Field, N. J. Clark, Y. W. Zhu, D. X. Yang,R. D. Piner, S. Stankovich, D. A. Dikin, H. Geisler, C. A.Ventrice, R. S. Ruoff, J. Phys. Chem. C 2009, 113, 18480.

[120] M. Beidaghi, C. L. Wang, Adv. Funct. Mater. 2012, 22,4501.

[121] J. Lin, C. Zhang, Z. Yan, Y. Zhu, Z. Peng, R. H. Hauge, D.Natelson, J. M. Tour, Nano Lett. 2013, 13, 72.

[122] F. Liu, A. Gutes, I. Laboriante, C. Carraro, R. Maboudian,Appl. Phys. Lett. 2011, 99, 112104.

[123] Y. W. Zhu, S. Murali, M. D. Stoller, K. J. Ganesh, W. W.Cai, P. J. Ferreira, A. Pirkle, R. M. Wallace, K. A. Cychosz,M. Thommes, D. Su, E. A. Stach, R. S. Ruoff, Science 2011,332, 1537.

[124] M. A. Lillo-Rodenas, D. Cazorla-Amoros, A. Linares-Solano, Carbon 2003, 41, 267.

[125] L. M. Miller, C. C. Ho, P. C. Shafer, P. K. Wright, J. W.Evans, R. Ramesh, Proc. IEEE ECCE 2009, 30, 2627.

[126] Q. Wu, Y. X. Xu, Z. Y. Yao, A. R. Liu, G. Q. Shi, ACSNano 2010, 4, 1963.

[127] C. Peng, D. Hu, G. Z. Chen, Chem. Commun. 2011, 47,4105.

[128] D. W. Wang, F. Li, J. P. Zhao, W. C. Ren, Z. G. Chen, J.Tan, Z. S. Wu, I. Gentle, G. Q. Lu, H. M. Cheng, ACSNano 2009, 3, 1745.

[129] A. Clemente, S. Panero, E. Spila, B. Scrosati, Solid StateIonics 1996, 85, 273.

[130] A. Laforgue, P. Simon, C. Sarrazin, J. F. Fauvarque, J.Power Sources 1999, 80, 142.

[131] L. H. Bao, J. F. Zang, X. D. Li, Nano Lett. 2011, 11, 1215.[132] J. A. Yan, E. Khoo, A. Sumboja, P. S. Lee, ACS Nano

2010, 4, 4247.[133] X. Y. Lang, A. Hirata, T. Fujita, M. W. Chen, Nat. Nano-

technol. 2011, 6, 232.[134] C. C. Hu, K. H. Chang, M. C. Lin, Y. T. Wu, Nano Lett.

2006, 6, 2690.[135] C. C. Liu, D. S. Tsai, D. Susanti, W. C. Yeh, Y. S. Huang,

F. J. Liu, Electrochim. Acta 2010, 55, 5768.[136] C. B. Arnold, R. C. Wartena, B. Pratap, K. E. Swider-

Lyons, A. Pique, Mater. Res. Soc. Symp. P. 2002, 698, 275.

Specia

lIssu

eG

RA

PH

EN

E

Electroanalysis 2013, 25, No. & , 1 – 22 � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.electroanalysis.wiley-vch.de &21&

These are not the final page numbers! ��

Graphene-Based Electrochemical Microsupercapacitors

Page 22: A Review of Graphene-Based Electrochemical  Microsupercapacitors

[137] C. B. Arnold, R. C. Wartena, B. Pratap, K. E. Swider-Lyons, A. Pique, Proc. SPIE 2002, 4637, 353.

[138] M. Q. Xue, Z. Xie, L. S. Zhang, X. L. Ma, X. L. Wu, Y. G.Guo, W. G. Song, Z. B. Li, T. B. Cao, Nanoscale 2011, 3,2703.

[139] L. Nyholm, G. Nystrçm, A. Mihranyan, M. Strømme, Adv.Mater. 2011, 23, 3751.

[140] G. A. Snook, P. Kao, A. S. Best, J. Power Sources 2011,196, 1.

[141] J. H. Sung, S. J. Kim, K. H. Lee, J. Power Sources 2003,124, 343.

[142] K. Wang, W. J. Zou, B. G. Quan, A. F. Yu, H. P. Wu, P.Jiang, Z. X. Wei, Adv. Energy Mater. 2011, 1, 1068.

[143] M. Nishizawa, M. Shibuya, T. Sawaguchi, T. Matsue, I.Uchida, J. Phys. Chem. 1991, 95, 9042.

[144] S. R. Sivakkumar, W. J. Kim, J.-A. Choi, D. R. MacFarlane,M. Forsyth, D.-W. Kim, J. Power Sources 2007, 171, 1062.

[145] S. Mondal, K. Barai, N. Munichandraiah, Electrochim.Acta 2007, 52, 3258.

[146] P. Novak, K. Muller, K. S. V. Santhanam, O. Haas, Chem.Rev. 1997, 97, 207.

[147] H. H. Rehan, J. Power Sources 2003, 113, 57.[148] R. Liu, S. Il Cho, S. B. Lee, Nanotechnology 2008, 19,

215710–1.[149] J. Z. Wang, S. L. Chou, J. Chen, S. Y. Chew, G. X. Wang, K.

Konstantinov, J. Wu, S. X. Dou, H. K. Liu, Electrochem.Commun. 2008, 10, 1781.

[150] G. Nystrom, A. Razaq, M. Stromme, L. Nyholm, A. Mih-ranyan, Nano Lett. 2009, 9, 3635.

[151] L. Z. Fan, Y. S. Hu, J. Maier, P. Adelhelm, B. Smarsly, M.Antonietti, Adv. Funct. Mater. 2007, 17, 3083.

[152] J. Yan, T. Wei, B. Shao, Z. Fan, W. Qian, M. Zhang, F.Wei, Carbon 2010, 48, 487.

[153] M. A. Q. Xue, F. W. Li, J. Zhu, H. Song, M. N. Zhang, T. B.Cao, Adv. Funct. Mater. 2012, 22, 1284.

[154] Y. H. Wu, B. J. Yang, G. C. Han, B. Y. Zong, H. Q. Ni, P.Luo, T. C. Chong, T. S. Low, Z. X. Shen, Adv. Funct. Mater.2002, 12, 489.

[155] J. R. Miller, R. A. Outlaw, B. C. Holloway, Science 2010,329, 1637.

[156] X. Zhao, H. Tian, M. Y. Zhu, K. Tian, J. J. Wang, F. Y.Kang, R. A. Outlaw, J. Power Sources 2009, 194, 1208.

[157] T. C. Hung, C. F. Chen, W. T. Whang, Electrochem. Solid.St. 2009, 12, K41.

[158] G. P. Xiong, K. P. S. S. Hembram, R. G. Reifenberger, T. S.Fisher, J. Power Sources 2012, 227, 254.

[159] G. P. Xiong, K. P. S. S. Hembram, D. N. Zakharov, R. G.Reifenberger, T. S. Fisher, Diamond Relat. Mater. 2012,27–28, 1.

[160] C. B. Arnold, R. C. Wartena, B. Pratap, K. E. Swider-Lyons, A. Piqu�, Mater. Res. Soc. Symp. Proc. 2002, 698,Q3.2.1.

[161] H. K. Kim, S. H. Cho, Y. W. Ok, T. Y. Seong, Y. S. Yoon, J.Vac. Sci. Technol. B 2003, 21, 949.

[162] W. Sun, X. Y. Chen, J. Power Sources 2009, 193, 924.

[163] C. C. Ho, D. A. Steingart, J. W. Evans, P. K. Wright, ECSTrans. 2008, 16, 35.

[164] J. H. Sung, S. J. Kim, S. H. Jeong, E. H. Kim, K. H. Lee, J.Power Sources 2006, 162, 1467.

[165] M. Q. Xue, Z. Xie, L. S. Zhang, X. L. Ma, X. L. Wu, Y. G.Guo, W. G. Song, Z. B. Li, T. B. Cao, Nanoscale 2011, 3,2703.

[166] P. Calvert, Chem. Mater. 2001, 13, 3299.[167] H. P. Le, J. Imaging Sci. Techn. 1998, 42, 49.[168] V. Kumar, J. W. Boley, Y. Yang, H. Ekowaluyo, J. K.

Miller, G. T. C. Chiu, J. F. Rhoads, Appl. Phys. Lett. 2011,98, 153510.

[169] K. Woo, D. Kim, J. S. Kim, S. Lim, J. Moon, Langmuir2009, 25, 429.

[170] M. L. Allen, M. Aronniemi, T. Mattila, A. Alastalo, K.Ojanpera, M. Suhonen, H. Seppa, Nanotechnology 2008,19, 175201.

[171] S. M. Bidoki, D. M. Lewis, M. Clark, A. Vakorov, P. A.Millner, D. McGorman, J. Micromech. Microeng. 2007, 17,967.

[172] T. Bhuvana, W. Boley, B. Radha, B. D. Dolash, G. Chiu, D.Bergstrom, R. Reifenberger, T. S. Fisher, G. U. Kulkarni,Micro Nano Lett. 2010, 5, 296.

[173] E. Chow, J. Herrmann, C. S. Barton, B. Raguse, L. Wiec-zorek, Anal. Chim. Acta 2009, 632, 135.

[174] P. Shah, Y. Kevrekidis, J. Benziger, Langmuir 1999, 15,1584.

[175] D. P. Li, D. Sutton, A. Burgess, D. Graham, P. D. Calvert,J. Mater. Chem. 2009, 19, 3719.

[176] D. Pede, G. Serra, D. De Rossi, Mater. Sci. Eng. C,Biomim. 1998, 5, 289.

[177] D. Du, X. Liu, G. Korn, J. Squier, G. Mourou, Appl. Phys.Lett. 1994, 64, 3071.

[178] I. Barbulovic-Nad, M. Lucente, Y. Sun, M. J. Zhang, A. R.Wheeler, M. Bussmann, Crit. Rev. Biotechnol. 2006, 26,237.

[179] R. R. Gattass, E. Mazur, Nat. Photonics 2008, 2, 219.[180] S. Kawata, H. B. Sun, T. Tanaka, K. Takada, Nature 2001,

412, 697.[181] A. Chimmalgi, T. Y. Choi, C. P. Grigoropoulos, K. Komvo-

poulos, Appl. Phys. Lett. 2003, 82, 1146.[182] A. Pique, D. W. Weir, P. K. Wu, B. Pratap, C. B. Arnold,

B. R. Ringeisen, R. A. McGill, R. C. Y. Auyeung, R. A.Kant, D. B. Chrisey, Proc. SPIE 2002, 4637, 361.

[183] Y. L. Zhang, L. Guo, S. Wei, Y. Y. He, H. Xia, Q. D. Chen,H. B. Sun, F. S. Xiao, Nano Today 2010, 5, 15.

[184] E. Kim, Y. N. Xia, G. M. Whitesides, J. Am. Chem. Soc.1996, 118, 5722.

[185] Y. N. Xia, E. Kim, G. M. Whitesides, Chem. Mater. 1996, 8,1558.

[186] Q. Y. He, H. G. Sudibya, Z. Y. Yin, S. X. Wu, H. Li, F.Boey, W. Huang, P. Chen, H. Zhang, ACS Nano 2010, 4,3201.

Specia

lIs

sue

GR

AP

HE

NE

&22& www.electroanalysis.wiley-vch.de � 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Electroanalysis 2013, 25, No. & , 1 – 22

�� These are not the final page numbers!

Review G. Xiong et al.