micromachines

Review Recent Studies on with Next-Generation Structures

Juho Sung and Changhwan Shin * Department of Electrical and Computer Engineering, Sungkyunkwan University, Suwon 16419, Korea; [email protected] * Correspondence: [email protected]; Tel.: +82-31-290-7694

 Received: 27 November 2020; Accepted: 16 December 2020; Published: 18 December 2020 

Abstract: Supercapacitors have shown great potential as a possible solution to the increasing global demand for next-generation energy storage systems. Charge repositioning is based on physical or chemical mechanisms. There are three types of supercapacitors—the electrochemical double layer, the pseudocapacitor, and a hybrid of both. Each type is further subdivided according to the material used. Herein, a detailed overview of the working mechanism as well as a new method for capacitance enhancement are presented.

Keywords: ; electrical double layer ; EDLC; negative capacitance; energy storage system; pseudocapacitor; micropores; nanopores; ultracapacitor; Helmholtz model

1. Introduction The Fourth Industrial Revolution, which facilitates the emergence of a hyper-connected society, is the basis for data processing between people and things as well as things and things. The essence anchored in a host of electronic devices highlights the energy consumption problem to address the ever-increasing energy demand. However, the challenge confronted (based on expenditure) is proving hard to overcome. Self-powered devices are mostly based on lithium- batteries, which were proposed three decades ago. They demonstrate superior properties as compared to energy storage systems (ESSs) based on other materials with the highest energy storage capacity; however, the world no longer refers to them as “perfect batteries”. Moreover, the progress of battery technology would be considered as a stumbling block to the development of electrically powered industry in some parts of the world. Insufficient cycles to deliver energy owing to membrane shrinkage by shutdown, power density, and thermal stability are now more noteworthy aspects in particular fields. Most importantly, the energy storage capacity, which is considered a strength, is no longer increasing. The major frameworks to solve these issues are classified as follows: (1) ultra-low-power devices based on energy efficiency with various techniques to overcome the theoretical limit of energy consumption efficiency with constant energy capability [1–4] and (2) focusing on developing the next generation of ESSs to complement lithium-ion batteries that have reached their innovation limit in a specific area [5–7]. There are various requirements to implement these next-generation ESSs, but to date some systems to satisfy certain conditions remain nonexistent [8]. Of these, using the extreme surface area and distance from the to the charged layer on the have particularly attracted attention owing to their superior specific power. Capacitors with superior characteristics (called supercapacitors), which are unavailable in conventional batteries, exhibit excellent functionality in many areas, including power density, charge/discharge cycles, operation over a wide temperature range, and reliability, which have been noted as limitations of batteries [8–13]. However, based on energy density, it is evident that supercapacitors must be studied intrinsically.

Micromachines 2020, 11, 1125; doi:10.3390/mi11121125 www.mdpi.com/journal/micromachines Micromachines 2020Micromachines, 11, 1125 2020, 11, x 22 of of 25 25

MicromachinesMicromachines[8,9,10,11,12,13]. 2020, 11 2020, x , 11, x However, based on energy density, it is evident that supercapacitors must2 of 25 be 2 of studied 25 Among aintrinsically. few solutions for overcoming the limitations of low energy density, the main area of [8,9,10,11,12,13].[8,9,10,11,12,13].Among However, However, a few based solutions based on energy on for energy overcoming density, density, it is theevident it islimitations evident that supercapacitors that of lowsupercapacitors energy must density, mustbe studied the be mainstudied area of interestMicromachines involves 2020, 11, x the development of new materials or structures for the of2 of supercapacitors. 25 intrinsically.intrinsically.interest involves the development of new materials or structures for the electrodes of supercapacitors. The electrode materials in laboratories and industries are typically reliant on carbon, metal oxides, [8,9,10,11,12,13].AmongTheAmong However, a electrodefew a solutions few based materialssolutions on for energy overcoming forin laboratories overcomingdensity, the it is limitationsand evidentthe industrilimitations that of es supercapacitorslow are of energy lowtypically energy density, reliant must density, the be on studiedmain carbon,the mainarea metal ofarea oxides,of mixed metal oxides, and conductive polymers. Since the onset of supercapacitor manufacturing, carbon intrinsically.interestinterest involvesmixed involves themetal development the oxides, development and of newconductive of materials new materials polymers. or structures or structuresSince for the the foronset electrodes the ofelectrodes supercapacitor of supercapacitors. of supercapacitors. manufacturing, materialsAmongThe electrodeThe have a fewelectrodecarbon been solutionsmaterials usedmaterials materials becausefor in havelaboratoriesovercoming in beenlaboratories of their used andthe broad because limitations industriand surface industri ofes their areof areaeslow brtypically areoad energy [7 typically, surface14 ,15reliant density,]. area Metalreliant on [7,14,15]. thecarbon, oxideson main carbon, Metalmetal area are mainlymetal oxidesofoxides, oxides, are used mainly for pseudocapacitors,interestmixed involvesmixed metalused the metal oxides, development andfor oxides,pseudocapacitors, due and to conductiveand theirof newconductive high materials andpolymers. capacity due polymers. or to structures Sincetheir and Since highlowthe for onset resistancecapathe the city onsetelectrodesof supercapacitorand of they lowsupercapacitor ofcan resistancesupercapacitors. be manufacturing, attractive they manufacturing, can as be electrode attractive materialsThe electrodecarboncarbon which materials materialsas materialselectrode have have in a laboratories higherhavebeenmaterials usedbeen energy usedbecause whichand because industri and ofhave their power ofesa brtheir higherareoad than typicallybr surface oaden conventionalergy surface areareliant and [7,14,15]. area poweron [7,14,15]. carbon, carbon-based Metalthan metal Metal oxidesconventional oxides,oxides are supercapacitors. mainly are carbon-based mainly Formixed conductiveused metal usedfor oxides, pseudocapacitors,supercapacitors. for polymers, pseudocapacitors, and conductive a reduction-oxidationFor and conductive dueandpolymers. to due their topolymers, Since theirhigh processcapahighthe a onset cityreduction-oxidationcapa and is cityof used supercapacitorlow and toresistance low store resistance process and theymanufacturing, release iscanthey used be can charges.attractiveto bestore attractive and Various release as electrodecharges. materials Various which conductive have a higherpolymers en ergyhave andbeen power widely than studied conventional because they carbon-based are easy to produce conductivecarbon materialsas polymerselectrode have been materials have used been because which widely ofhave their studied a br higheroad becausesurface energy area they and [7,14,15]. arepower easy Metal than to produceoxidesconventional are and mainly arecarbon-based inexpensive supercapacitors.supercapacitors.and are For inexpensive conductive For conductive compared polymers, polymers, to a other reduction-oxidation a materialsreduction-oxidation fabricated process forprocess issupercapacitor used is usedto store to storeandelectrodes release and release [11]. comparedused for pseudocapacitors, to other materials and due fabricated to their high for supercapacitor capacity and low electrodes resistance they [11]. can be attractive as electrodecharges.charges. materials Various VariousRecently, conductivewhich conductive ahave new polymers atheory higher polymers havewas energy proposedhavebeen and beenwidely power withwidely studied athan layerstudied becauseconventional combination because they are they carbon-basedof easy aare ferroelectric toeasy produce to produce exhibiting Recently, a new theory was proposed with a layer combination of a ferroelectric exhibiting negative supercapacitors.and areand inexpensive arenegative For inexpensive conductive capacitance compared compared polymers, toand other todielectric, a other reduction-oxidationmaterials materials which fabricated couldfabricated processfor enhance supercapacitor for is supercapacitor usedthe capacitance to store electrodes and electrodes th releaseat [11]. had [11]. been hitherto capacitancecharges. VariousRecently, andconsideredRecently, conductive dielectric, a new a astheorynew polymersan which theoryinherent was could haveproposedwas limitation. beenproposed enhance widelywith In withadditiona the layerstudied capacitancea layercombination to because previously combination they that of exhibiting area had ferroelectricof easy a been ferroelectric to the produce hitherto ability exhibiting exhibitingto considered handle low- asand an arenegative inherent inexpensivenegative capacitancepower limitation. capacitancecompared technologies and to dielectric, Inand other addition dielectric,[1,2,3,4], materials which to ferroelectricwhich facould previouslybricated could enhance for mateenhance supercapacitor exhibiting therials capacitance the with capacitance the electrodesnegative abilitythat hadth capacitanceat[11]. to beenhad handle beenhitherto low-powereffectshitherto have technologiesRecently,consideredconsidered a [demonstrated 1 newas–4 an], theory ferroelectricasinherent an inherent wasthe limitation. abilityproposed materialslimitation. of InESSs with addition In within a additioncombination layer negativeto previouslycombination to previously with capacitance exhibiting dielectric of exhibitinga ferroelectric elayers.theffects ability the The have ability exhibiting to theory handle demonstrated to handleis basedlow- low- on ionic the polarization, which leads to ferroelectric effects, as well as dipolar polarization. It has evinced an abilitynegativepower of capacitance ESSspower technologies in technologies combination and dielectric,[1,2,3,4], [1,2,3,4], with whichferroelectric dielectric ferroelectric could enhance layers.mate rialsmate Thethe rialswith capacitance theory withnegative isnegative based th atcapacitance had oncapacitance ionicbeen hithertoeffects polarization, effects have have which demonstrateddemonstratedimproved the ability the capacitance, ability of ESSs of ESSswhichin combination in resulted combination inwith a theoretical withdielectric dielectric layers.increase layers. The in storagetheory The theory is energy based is based ofon up ionic onto twoionic times leadsconsidered to ferroelectric as an inherent eff limitation.ects, as well In addition as dipolar to previously polarization. exhibiting It has the evinced ability to an handle improved low- capacitance, polarization,polarization,and which also whichexperimentally leads leadsto ferroelectric to ferroelectricachieved effects, an increaseeffects, as well as in as wellstorage dipolar as dipolarenergy polarization. ofpolarization. up to 1.5It has times It evinced has [16]. evinced This an presents an whichpower resultedtechnologies in a theoretical[1,2,3,4], ferroelectric increase inmate storagerials energywith negative of up tocapacitance two times effects and also have experimentally demonstratedimprovedimproved the acapacitance, new ability capacitance, possibility of ESSs which to inwhich increaseresultedcombination resulted capacitancein a theoretical within a theoreticaldielectric which increase can layers. increase overcome in storageThe in theorystorage the energy limitations is energy based of up onof to upsupercapacitors. ionictwo to times two times The achieved an increase in storage energy of up to 1.5 times [16]. This presents a new possibility to increase polarization,and alsoand which experimentally alsoresults experimentally leads were to experimentally ferroelectric achieved achieved an effects,increase confirmed, an increase as in well storage as in ascomprehensively storage dipolar energy energy polarization.of up of todesc up 1.5 ribedto times It 1.5 has intimes [16]. Sectionevinced This[16]. 4. presentsThisan presents capacitanceimproveda new capacitance,a possibilitynew which possibilityThis can towhich articleincrease overcome to resultedincrease not capacitance only the incapacitance describes a limitations theoretical which the which can increaseanalysis of overcome can supercapacitors. overcome ofin supercapacitorsstorage the limitations the energy limitations The of basedof results up supercapacitors. toof on supercapacitors.two werevarious times experimentally electrochemical The The confirmed,and alsoresults experimentallyresults were asdouble-layer comprehensively wereexperimentally experimentally achieved models confirmed,an describedbut increase confirmed, also introducesas in comprehensivelyinstorage as Section comprehensively newenergy4. ways of desc up to improvetoribed desc 1.5 ribed timesin Sectioncapacitance in [16]. Section 4.This beyond 4.presents the fatal limit by a newThis possibilityThis article articleutilizingThis to not increasearticle not only negative only not capacitance describes describesonly capacitance describes thewhichthe analysis materials.the can analysis overcome of of supercapacitors supercapacitorsof supercapacitors the limitations based basedof based on supercapacitors. various on various various electrochemical electrochemical The electrochemical The characteristics of capacitors, supercapacitors, and batteries, which are typical ESSs, are double-layerresultsdouble-layer weredouble-layer experimentally models models models but but confirmed, also also but introduces introduces also asintroduces comprehensively new new waysnew ways ways to toimprove desc improveto ribedimprove capacitance in capacitanceSection capacitance beyond4. beyond beyond the fatal the the limitfatal fatal limitby limit by by utilizingutilizing negativedescribed negative capacitance in Table capacitance 1. materials.The indicatorsmaterials. for specific energy density, specific power density, charge storage utilizingThis article negative not only capacitance describes materials. the analysis of supercapacitors based on various electrochemical double-layerThe models characteristicsmechanism,The characteristics but also cycling introducesof capacitors, of performance capacitors, new supercapacitors, ways stsupercapacitors, toorage improve region, and capacitance charge batteries,and batteries, temperature, beyond which which theare fataldischarge typical are limit typical ESSs, bytemperature, ESSs, are are and The characteristics of capacitors, supercapacitors, and batteries, which are typical ESSs, are utilizingdescribed negativedescribed galvanostaticin Tablecapacitance in Table 1. The 1.discharge materials.indicators The indicators curves for specific for are specific presented.energy en density,ergy In density, the specific case specific powerof galvanostatic power density, density, charge discharge charge storage storagecurves, this describedThemechanism, characteristicsmechanism, inallows Table cycling1 us .cycling of to Theperformance figurecapacitors, indicatorsperformance out that supercapacitors,storage the forst capacitororage region, specific region, and chargeand energy su chargebatteries,percapacitor temperature, density, temperature, which could discharge specific are use dischargetypical the powertemperature, total ESSs, temperature, stored density,are energy and chargeand rapidly storagedescribedgalvanostatic mechanism, ingalvanostatic Tablethrough 1. discharge The cycling the dischargeindicators slope curves performance of forcurves curves. are specific presented. are In encontrast,presented. storageergy In density, th region,the eIn battery case the specific chargeofcase shows galvanostatic powerof temperature, galvanostatica relatively density, discharge chargeuniform discharge discharge storagecurves, state. curves, temperature, this this andmechanism, galvanostaticallowsallows uscycling to usfigure toperformance discharge figure out that out the that curvesst capacitororage the capacitor areregion, and presented. su chargeandpercapacitor su percapacitortemperature, In the could case could usedischarge ofthe use galvanostatic total the temperature, storedtotal stored energy discharge energyand rapidly rapidly curves, Table 1. Description of the properties of energy storage systems. thisgalvanostatic allowsthroughthrough us thedischarge to slope the figure slope of curves curves. out of curves. thatare In presented.contrast, the In capacitorcontrast, th eIn battery ththee and batterycase shows supercapacitor of showsgalvanostatic a relatively a relatively uniform could discharge uniform use state. curves, the state. total this stored energy rapidlyallows us through to figureEnergy out the thatStorage slope the System of capacitor curves. and In su contrast,percapacitor Capacitors the batterycould use shows the total Supercapacitors a relativelystored energy uniform rapidly state. Batteries through the slopeSpecific of curves. energyTable In density contrast,Table 1. Description (Wh 1. Description thkge−1 )battery of the ofproperties shows the0.01 propertiesto 0.1a relativelyof energy of energy storage uniform storage systems. 0.1state. tosystems. 50 10 to 200 Specific power density (W kg−1) 103 to ~107 1 to ~106 10 to 100 EnergyEnergy Storage Storage System System Capacitors Capacitors Supercapacitors Supercapacitors Batteries Batteries Table 1. Description of the properties of energy storageCharge Separation systems. Faradaic Specific energy density (Wh kg−1) −1 0.01 to 0.1 0.1 to 50 10 to 200 SpecificCharge energyTable storage density 1. Description mechanism (Wh kg ) of the properties0.01Charge to 0.1 of Separation energy storage Charge systems. 0.1 adsorption to 50 (desorption) 10 toIntercalation 200 Specific power density (W kg−1) −1 3 37 7 6 6 Specific power density (W kg ) 10 to ~1010 to ~10 1 toIntercalation ~101 to ~10 (deintercalation) 10 to 100 10(deintercalation) to 100 EnergyEnergy Storage Storage System System Capacitors Capacitors Supercapacitors Supercapacitors Batteries Batteries Storage region Surface ChargeCharge Separation SeparationSurface FaradaicFaradaic Surface to bulk SpecificCharge energy storage density mechanism (Wh kg−1) 1 0.01 to 0.1 0.1 to 50 10 to 200 Specific energyChargeCycling density storage performance mechanism (Wh kg− ) Charge0.01Charge Separation to 0.1 SeparationInfinite Charge Charge adsorption adsorption 0.1 to (desorption) 50>500,000 (desorption) IntercalationIntercalation 10 to 200500~2000 Specific power density (W kg−1) 1 103 to ~1037 7 Intercalation1 to ~106 (deintercalation)6 10 to(deintercalation) 100 Specific powerCharge density temperature (W kg− ) (°C) 10 to ~10 −20 to 100 Intercalation1 to ~10 (deintercalation)−40 to 65 (deintercalation)10 to 100 0 to 45 Storage region Surface Charge SeparationSurface FaradaicSurface to bulk StorageDischarge region temperature (°C) Surface−20 to 100 Charge SeparationSurface−40 to 65 SurfaceFaradaic to− 20bulk to 60 ChargeCycling storage performancemechanism Charge SeparationInfinite Charge adsorption>500,000 (desorption) Intercalation500~2000 Charge storageCyclingCharging mechanism performance time (sec) Charge SeparationInfinite10− 6 to 10−3Charge adsorption>500,000 (desorption) 1 to 10 Intercalation500~2000103 to 105 ChargeCharge temperature temperature (°C) (°C) −20 to 100−20 to 100Intercalation Intercalation (deintercalation)−40 to (deintercalation) 65−40 to 65 (deintercalation) 0(deintercalation) to 45 0 to 45 StorageStorageDischarge region regionDischarge temperature temperature (°C) (°C) Surface−20Surface to 100−20 to 100 Surface−40 Surfaceto 65−40 to 65 Surface to −bulk20 Surface to 60−20 to to 60 bulk CyclingCyclingCharging performance performanceCharging time (sec) time (sec) Infinite10Infinite −6 to 10−63 to 10−3 >500,000 1 >to500,000 10 1 to 10 500~200010 3 to 500~2000101053 to 105 Charge temperature (°C) −20 to 100 −40 to 65 0 to 45 Charge temperature (◦C) 20 to 100 40 to 65 0 to 45 Discharge temperature (°C) −20 to −100 −40 to 65 − −20 to 60 Discharge temperature (◦C) 20 to 100 40 to 65 20 to 60 Charging time (sec) 10−6 to 10− −63 3 1 to 10 − 10 3 to 105 − 3 5 Charging timeGalvanostatic (sec) discharge curves10 − to 10− 1 to 10 10 to 10

GalvanostaticGalvanostaticGalvanostatic discharge discharge discharge curves curves curves

Galvanostatic discharge curves Note: Galvanostatic discharge curves of supercapacitors were classified into three types: (1) EDLC, (2) Pseudocapacitors (), and (3) Pseudocapacitors (intercalation).

2. Classification of Supercapacitors

Supercapacitors, which are also referred to as double-layer capacitors, gold capacitors, power caches, power capacitors, ultracapacitors, or super condensers, are classified into three types, depending on the charge storage mechanism: (1) the electrical double-layer capacitor (EDLC), which is based on a Micromachines 2020, 11, 1125 3 of 25 charge storage mechanism that physically stores charges on the electrode surface with an electrical double layer without causing irreversible chemical reactions [8–11,17]; (2) a pseudocapacitor, which is operated by a that relies on a mechanism that stores charge only in a limited area based on surface, unlike a battery, wherein the reaction occurs on the spacious electrode region from the surface to the bulk region; and (3) hybrid capacitors that adopt both mechanisms, composed of EDLC electrodes and electrodes based on chemical mechanisms, such as pseudocapacitors or batteries, which can exhibit the characteristics of both systems and provide intermediate performances. Here again, they are subdivided according to the material of the electrode or electrolyte (see Figure1)[18]. Notably, EDLCs can withstand millions of cycles owing to the mechanism that stores charge on the electrode surface with an electrical double layer based on a constantly reversible capability. This property allows the device to not only exceed the life of the system but also the quality of the device compared to batteries. The reversible capability does not stir deviation up in the electrode, thereby eliminating the swelling that occurs in typical redox reactions found in spacious amounts of battery active materials during charge and discharge cycles. Supercapacitor electrodes undergo electrochemical dynamics through polarization resistance, and no rate limiting occurs within the bulk range as in redox battery electrodes. Based on the EDLC structure, many research teams have spent their efforts on improving the charge storage quality regarding the high specific surface area of porous materials, which are mainly used for electrodes. Until pseudocapacitors based on intercalation were embraced, pseudocapacitors have also been synonymously used with “surface redox reactions”. Redox-active materials, such as metal oxides, mixed metal oxides, sulfides, selenides, nitrides, conductive polymers, carbon materials, and composite with carbon materials, have been used on the active electrode. The energy based on the chemical reaction of a solid substance is mainly stored in chemical bonds—that is, ionic bonds, covalent bonds, and hydrogen bonds. When exploited in electrochemical reactions, solid materials with high bond energies are converted into low bond energy materials to release electricity. Electrical and chemical potentials in the electrolyte can be reversibly converted by an interaction between charged/discharged electrode materials and chemical reactions. Pseudocapacitors do not depend on metal chemistry in the bulk area; thus, there is less risk of metal plating, which poses significant performance degradation, failure mechanism, and safety issues that can lead to short-circuiting and uncontrollable energy chemical reactions [6,18–20]. The material conventionally used in the active region of the electrode is described in Table2. Active carbon, carbon nanotube(CNT)/graphene, metal oxide pseudocapacitance, metal oxide faradaic reactions, and conducting polymers are described according to their benefits, issues, and electrode descriptions [21,22].

Table 2. Properties of various electrode materials [21–23].

Electrode Materials Advantages Issues and Challenges Electrode Description high surface area (over 1000 m2 g 1) low specific capacitance, Activated carbons − AC cost, chemical stability, and availability conductivity 2 1 high surface area (over 2600 m g− ), volumetric capacitance, thermal conductivity, flexibility, agglomeration, GO CNT/Graphene chemical, thermal, mechanical stability nano–micro MWCNT corrosion resistance transformation, ion accessibility, Metal oxides- high specific capacitance, intrinsic stability, MnO , RuO pseudocapacitance porosity, adhesion, conductivity 2 2 relatively high cost Co O NiO, CuO, Fe O high specific capacitance, long cycles life, stability, 3 4, 2 3, Metal TiO , etc. high conductivity, good electrochemical barren reveres, 2 oxides/hydroxides-faradaic Metal hydroxides reversibility, and high rate capability relatively high cost (Ni(OH)2, Co(OH)2), relatively high storage capacity, low cost, conductivity, Conducting polymers low environmental impact, porosity PANI, PPy, PTh stability high voltage window, Micromachines 2020, 11, x 4 of 24

Co3O4, NiO, CuO, Metal high specific capacitance, long cycles life, stability, Fe2O3, TiO2, etc. oxides/hydroxides- high conductivity, good electrochemical barren reveres, Metal hydroxides faradaic reversibility, and high rate capability relatively high cost (Ni(OH)2, Co(OH)2), relatively high storage capacity, low cost, conductivity, MicromachinesConducting2020 polymers, 11, 1125 low environmental impact, porosity PANI, PPy, PTh4 of 25 stability high voltage window,

Pseudocapacitors based on redox reactions as well asas otherother chemicalschemicals exhibitexhibit higherhigher specificspecific capacitancecapacitance propertiesproperties thanthan EDLCsEDLCs basedbased on physicalphysical charge storage operations. However, there are advantages andand disadvantagesdisadvantages ofof energyenergy storagestorage viavia physicalphysical andand bulkbulk chemicalchemical reactions.reactions. The conceptconcept ofof exploitingexploiting bothboth physicalphysical andand chemicalchemical mechanismsmechanisms is presentedpresented asas hybridhybrid supercapacitors.supercapacitors. This This is is a a combination combination of of materials commonly usedused inin pseudocapacitor-basedpseudocapacitor-based electrodes– with EDLC-based electrodes electrodes and and lithium lithium electrode–electrolytes electrode–electrolytes with with EDLC EDLC - -basedbased electrodes. The The second second approach approach is is considered considered as a a lithium-ion lithium-ion capacitor, which mainly entails thethe useuse ofof metal metal oxides oxides or or mixed mixed metal metal oxides oxides and and carbon carbon electrodes electrodes to construct to construct the cell.the cell. Owing Owing to this to architecturethis architecture of arrangement, of arrangement, the storage the storage cell functions cell functions in a wider in a workingwider working potential potential window, window, thereby exhibitingthereby exhibiting a largercapacitance a larger capacitance value. This value. approach Thisprovides approach a 2–3-timesprovides highera 2–3-times energy higher density energy than thatdensity of traditional than that of EDLCs traditional [21]. EDLCs [21].

Figure 1. ClassificationClassification of supercapacitors [[24].24].

3. Electrical Double-Layer Capacitors (EDLCs)

3.1. Theory of the Electrochemical Double Layer As thethe namename itselfitself indicates,indicates, supercapacitorssupercapacitors are devicesdevices thatthat maximizemaximize thethe capacitancecapacitance ofof conventionalconventional capacitors. These These devices based on physicalphysical charge storage withwith electricalelectrical operationoperation exhibit aa chargecharge outputoutput like like capacitors; capacitors; thus, thus, the the capacitance capacitance “C “”C depends” depends on on the the dielectric dielectric constant constantεr d ofεr of the the electrolyte, electrolyte, the the eff effectiveective thickness thickness “ “”d” of of the the charge charge separation separation layer layer fromfrom electrodes,electrodes, andand thethe accessible surfacesurface “ “AA”” through through the the distance distance between between the the electric electr layersic layers based based on Helmholtzon Helmholtz theory theory and theand porous the porous structure structure [21]. [21]. InIn conventionalconventional capacitors,capacitors, surfacesurface areaarea AA isis aa two-dimensionaltwo-dimensional single-planarsingle-planar electrode,electrode, whilewhile thethe supercapacitor supercapacitor is is a athree-dimensional three-dimensional porous porous electrode—i.e., electrode—i.e., the the surface surface area area is maximized. is maximized. The Thedielectric dielectric layer layer of a of supercapacitor a supercapacitor comprises comprises a single a single or or several several molecular molecular layers, layers, so so “ “dd”” is is only a smallsmall molecularmolecular distance,distance, ranging ranging from from a fewa few Å Å to to several several Å. Å. This This is a is huge a huge diff erence,difference, considering considering that anthat electrostatic an electrostatic capacitor capacitor has an has insulator an insulator layer layer that reachesthat reaches up to up several to several microns. microns. The charge separation layer is formed with ions when a charged object is placed in an electrolyte. The mutual interfacial structure between the electrodes creates a certain structure. The balancing counter charge on the charged surface of this object is formed correspondingly in the electrode and concentrated near the surface. Various theories have been proposed to explain the structure wherein charges are formed, and a complete theory explaining all the mechanical contents remains Micromachines 2020, 11, x 5 of 24

The charge separation layer is formed with ions when a charged object is placed in an electrolyte. The mutual interfacial structure between the electrodes creates a certain structure. The balancing Micromachinescounter charge2020, on11, 1125the charged surface of this object is formed correspondingly in the electrode5 ofand 25 concentrated near the surface. Various theories have been proposed to explain the structure wherein charges are formed, and a complete theory explaining all the mechanical contents remains undetermined. There are theories for understanding the charge separation layer, as represented in undetermined. There are theories for understanding the charge separation layer, as represented in Figure2. Figure 2.

FigureFigure 2.2. ModelsModels ofof thethe electrochemicalelectrochemical doubledouble layer.layer. ((aa)) HelmholtzHelmholtz model;model; ((bb)) Gouy–ChapmanGouy–Chapman oror didiffuseffuse model; model; ( c(c)) Stern Stern model; model; ((dd)) GrahameGrahame model;model; ((ee)) Bockris,Bockris, Devanathan, and and Muller Muller model model [25–27] [25–27]..

TheThe HelmholtzHelmholtz model model is theis the first first theory theory to provide to provide an approximation an approximation for the arrangementfor the arrangement between thebetween electrode the electrode and layer and of ions layer called of ions the called Helmholtz the Helmholtz layer or plane layer with or plane a finite with distance. a finite However, distance. thereHowever, are some there problems are some with problems interactions with thatinteractions occur further that occur away fromfurther the away electrode from which the electrode include thewhich electrolyte include concentrationthe electrolyte andconcentration all the given and environments, all the given environments, besides the double besides layer the arisingdouble fromlayer thearising interface from (Figurethe interface2a) [ 25 (Figure,28]. Chapman 2a) [25,28]. did Chap not takeman thisdid intonot take account this andinto proposedaccount and an electricalproposed double-layeran electrical theorydouble-layer and considered theory and the considered applied potential the applied and electrolyte potential concentration. and electrolyte Gouy concentration. suggested thatGouy a layersuggested of positive that a ionic layer charge of positive that counteractsionic charge the that charge counteracts affecting the the charge charged affecting electrode the appearscharged inelectrode the electrolyte appears surrounding in the electrolyte the charged surrounding solid, but the that charged ions didsolid, not but appear that ions only did on thenot surfaceappear withonly aon di fftheusion surface layer with based a ondiffusion the Boltzmann layer based distribution on the thatBoltzmann demonstrated distribution the concentration that demonstrated of charged the ionsconcentration with the counter of charged ions ions near with the surface. the counter This ions is called near thethe disurface.ffuse layer. This However,is called the this di modelffuse layer. also hasHowever, problems. this Becausemodel also ions has are problems. expressed Because as point ions charges, are expressed the specific as size point of thecharges, ion is the not specific considered. size Theof the theory ion is does not notconsidered. recognize The that theory the finite does space not re ofcognize ions cannot that the exist finite near space the electrode.of ions cannot Secondly, exist itnear fails the when electrode. applied Secondly, to highly chargedit fails when double applied layers to (Figure highly2b) charged [ 25,29– double34]. Stern layers modified (Figure the 2b) di ff[25,29–usion double34]. Stern layer modified to integrate the diffusion the theories double of Helmholtz layer to andintegrate Gouy–Chapman the theories and of Helmholtz proposed the and compact Gouy– layerChapman presented and byproposed the improved the compact understanding layer presented of the architectures by the improved of the ions understanding with a specific of sizethe andarchitectures limiting space of the between ions with the a specific electrode size and and electrolyte, limiting space but also between extended the electrode the diffusion andlayer electrolyte, to the bulkbut also layer. extended This distance the diffusion is generally layer considered to the bulk as la theyer. radius This ofdistance the ions. is generally Consequently, considered the potential as the chargesradius of and the concentration ions. Consequently, of the di theffusion potential portion char ofges the and layer concentration are sufficiently of the low diffusion to justify portion treating of ionsthe layer with pointare sufficiently charges (Figure low 2toc) justify [ 25,35 –treating37]. Grahame ions with developed point charges the theory (Figure of electrical 2c) [25,35–37]. double layersGrahame based developed on four regions—the the theory inner of electrical Helmholtz do planeuble inlayers front ofbased the electrode on four with regions—the small molecular inner distances;Helmholtz the plane outer in Helmholtz front of the plane, electrode which with consists small of molecular Helmholtz distances; layers; the the di outerffusion Helmholtz region; and plane, the bulkwhich region consists suggested of Helmholtz by Stern. layers; The the theory diffusion of specific region; adsorptionand the bulk was region applied. suggested This by bond Stern. occurs The ontheory the surfaceof specific ofthe adsorption metal electrode. was applied. The negativeThis bond ions occurs are adsorbedon the surface on the of surfacethe metal of electrode. the electrode The regardlessnegative ions of its are charge, adsorbed not byon the surface electrical of force, the electrode but via chemicalregardless interaction. of its charge, The not ions by referred the electrical to as ghost ions act as pseudocapacitive particles. Cations rarely stick to the electrode. The adsorption of ghost ions would be reduced while keeping the solution and electrodes clean. The capacitor centered on the reaction by specifically adsorbed ions is called a pseudocapacitor, which is found at the interface Micromachines 2020, 11, 1125 6 of 25 between the electrolyte and electrode (Figure2d) [ 25–27]. Bockris, Devanathan, and Muller’s model shows a phenomenon for the preponderance of solvent molecules around the interface. Solvated molecules, anions, cations, and electrodes were suggested, and the dipole moment was considered for both the charge separation and diffusion layers. The model is acceptable for the values applied to understand high-current-power EDLCs. However, the concept cannot perfectly explain carbon-based models with porous structures (Figure2e) [25,38].

3.2. Variation of Capacitance According to Pore Size and Structure Although the theory of electric double layers on plane surfaces has been clarified, problems arise when applying electric double layers to pores in actual spaces. The characteristics of ion electrical adsorption in porous structures make the process of charge storage difficult to deal with. Understanding the principle of charge storage through the formation of layers by cations and anions remains insufficient to explain experimental results when applied in nanopores. Generally, the inverse effect of pore size on capacitance is also known. Ions have different movement patterns with pores compared to the surface or bulk region. The specific surface area, which exhibits an evincive connection between the pore sizes, is important in order to understand its impact on specific capacitance. Additionally, it has been an attractive subject of study to numerous research groups in the past two decades. The capacitance generally increases proportionally to the surface area, but in the micropore region a difference in increasing value occurs. As the movement of ions has architectures with a finite size, the potential charges are greatly affected by the size of the pores. Each ion that plays a role in the storage of charge has a specific size, thereby allowing access to structures above a certain scale. Even if the surface area is increased, in regions where the pore size is too small for the solvated ion to approach it does not fully contribute to the layer, which should interact with electrodes. The specific and volumetric capacitance values indicate significant increases and are inversely commensurate to the nonlinear proportion depending on the chlorination temperature, which is largely involved in pore size control [39,40]. It is important to adjust the temperature, density, and intensity values for each material. The targets that constitute the electric double layer with the ions in the electrode layer are activated carbon and graphene, which have two-dimensional structures, and carbon nanotubes (CNTs), which have three-dimensional structures. Activated carbon changes the pore size depending on the temperature. Because the size of the pores determines the number of ions that can pass through the bilayer, if the conductivity is greatest at one point in the nano-meter range, then it decreases rapidly at higher and lower temperatures based on that state [39–44]. The reasons for this are as follows. First, when they are smaller than the largest point, solvated ions do not have the minimum length to form a double layer. That is, the optimal pore size may vary depending on the scale of the electrolyte ions (Figure3)[ 42–44]. According to their characteristics, there is a change in the interaction mechanism between the electrodes and ions. To understand this phenomenon, micropores are analyzed from various perspectives. The first aspect concerns layer formation as a cylinder (Figure3). There are many inquiries regarding the interaction between the electrolyte and the electrode fabricated with carbon material to adequately describe the electrostatic capacity depending on the pore shape and size. The analysis of capacitance for a cylindrical model is as follows, and it is analyzed by Gauss’s law (Figure3c): Z b Q Vba = ( )dr (1) − a 2πεrε0rL Z Q b dr Vba = (2) −2πεrε0L a r Q b Vba = ln . (3) −2πεrε0L a Micromachines 2020, 11, x 7 of 24

𝑄 𝑑𝑟 (2) Micromachines 2020, 11, 1125 =− 7 of 25 2𝜋ε𝜀𝐿 𝑟

=− ln . (3) In the equation, “εr” represents the dielectric constant of the electrolyte, the permittivity of a vacuum is expressed as “ε0”, the distance from the center of the pore to the cation is expressed “a”, In the equation, “𝜀” represents the dielectric constant of the electrolyte, the permittivity of a and the radius of the pore is expressed as “b”. vacuum is expressed as “𝜀”, the distance from the center of the pore to the cation is expressed “a”, and the radius of the pore is expressed as “b”. Q = CVba, (4) 𝑄=𝐶𝑉, (4) ! 1 b − C = 2πεrε0L ln , (5) 𝐶=2𝜋𝜀𝜀 𝐿(lna ), (5) wherewhere the the total total charge charge on on the the cylinder cylinder is is expressed expressed as as “Q “”,Q”, and andC Crepresents represents the the capacitance capacitance of of the the supercapacitor.supercapacitor. This This is appliedis applied at aat distance a distance of ~0of nm~0 nm or lessor less from from the electrodethe electrode to ions. to ions. Depending Depending on theon size the ofsize the of pores, the pores, the number the number of ions of that ions can that contribute can contribute to the capacitanceto the capacitance and the and size the of thesize layer of the oflayer ions changeof ions change (Figure 3(Figureb). 3b).

Figure 3. (a) Brief structure of a supercapacitor with a view of cylindrical theory (endohedral Figure 3. (a) Brief structure of a supercapacitor with a view of cylindrical theory (endohedral capacitors/electric double-cylinder capacitor), (b) shape of the electrolyte ion placed and the cylindrical capacitors/electric double-cylinder capacitor), (b) shape of the electrolyte ion placed and the electrode according to the pore size. The first picture shows that the ions are not very involved in charge cylindrical electrode according to the pore size. The first picture shows that the ions are not very storage because they have a very small pore structure. The second figure shows one ion interacting involved in charge storage because they have a very small pore structure. The second figure shows with an electrode to store charge. The third figure shows more ions forming the same layer and acting one ion interacting with an electrode to store charge. The third figure shows more ions forming the on the electrode. (c) The shape of pores observed in a three-dimensional shape; “a” is the distance from same layer and acting on the electrode. (c) The shape of pores observed in a three-dimensional shape; the center to a layer of several ions, the radius of the pores is expressed as “b”, and the cylinder length “a” is the distance from the center to a layer of several ions, the radius of the pores is expressed as “b”, is expressed as “L”[21,45,46]. and the cylinder length is expressed as “L” [21,45,46]. The second model tries to understand the interaction between ions and electrodes based on slits. DependingThe second on the intervalmodel tries between to understand the slits, the the types interactio whereinn between the ions ions of the and electrolyte electrodes with based electrode on slits. contributeDepending are on diff theerent interval which isbetween described the on slits, Figure the4 b–ftypes [45 ,wherein46]. the ions of the electrolyte with electrodeIf it is contribute larger than are point differenta in Figure which4 g,is described the distance on betweenFigure 4b–f the [45,46]. electrode and ions increases becauseIf oneit is electrodelarger than and point several a in solvation Figure 4g, ions the constitute distance a between double layer. the electrode Here, solvation and ions ions increases act as anbecause insulator, one and electrode the distance and several is applied solvation to the ions capacitor’s constitute width a double of the dielectriclayer. Here, layer. solvation When ions thepore act as b size is greater than a certain spot in Figure4g, the electrode pores and the layer of ions that can be formed start to increase; thus, the capacitance starts to increase again [39]. Nanopores sometimes have Micromachines 2020, 11, x 8 of 24

an insulator, and the distance is applied to the capacitor’s width of the dielectric layer. When the pore Micromachines 2020, 11, 1125 8 of 25 size is greater than a certain spot b in Figure 4g, the electrode pores and the layer of ions that can be formed start to increase; thus, the capacitance starts to increase again [39]. Nanopores sometimes minimalhave minimal spaces spaces to form to electrical form electrical layers withlayers solvated with solvated ions (Figure ions (Figure4b) as well 4b) as well electrical as electrical layers with layers solvatedwith solvated ions (Figure ions (Figure4c). If 4c). the poreIf the sizepore is size less is than less than or equal or equal to a to, (b) a, (b) (c)→ (c) applies. applies. (d) (d) represents represents → thethe case case of of solvated solvated ions ions with with nanopores nanopores having having extra extra space space to to form form an an electrical electrical layer, layer, and and (e) (e) shows shows thatthat the the nanopore nanopore has has su sufficientfficient solvated solvated ions ions to to form form an an electrical electrical layer layer with with each each electrode electrode and and that that therethere is is adequate adequate space space in in each each electrode. electrode. If If the the pore pore size size exceeds exceedsa anda andb bor or less, less, (c) (c) →(d) (d) is is applicable. applicable. → IfIf the the pore pore size size is is greater greater than thanb, b (e), (e) →(f) (f) is is applied applied (Figure (Figure4)[ 4)39 [39,40,42].,40,42]. →

Figure 4. (a) Brief structure of activated carbon particles with various pore sizes, not based on a circle, butFigure based 4. on (a) the Brief electrode structure theory of activated of walls carbon and walls. particles (b) Nanopores with various do pore not havesizes, the not minimum based on a space circle, requiredbut based to formon the an electrode electrical th layereory withof walls solvated and walls. ions, ( bc)) nanoporesNanopores have do not minimal have the space minimum to form space an electricalrequired layer to form with an solvated electrical ions, layer (d with) nanopores solvatedhave ions, extra(c) nanopores space to have form minimal an electrical space layer to form with an solvatedelectrical ions, layer (e) with nanopores solvated have ions, suffi (dcient) nanopores space for have each extra electrode space such to form that solvated an electrical ions canlayer form with ansolvated electrical ions, layer (e) with nanopores each electrode, have sufficient (f) electric space layer for each for carbon electrode grain such at thethat surface, solvated (g ions) schematic can form diagraman electrical of normalized layer with capacitance each electrode, according (f) electric to pore layer size. for Whencarbon the grain pore at size the issurface, less than (g) orschematic equal todiagrama, which of is normalized indicated by capacitance a red-colored according line, (b ,toc) pore is applicable, size. When and the when pore the size pore is less size than is more or equal than to a anda, which less thanis indicated or equal by to a bred-colored, which is indicated line, (b,c) byis applicable, a yellow-colored and when line, the (c, dpore) is applicable.size is more If than the a poreand size less is than larger or thanequal b, to which b, which is indicated is indicated by aby green-colored a yellow-colored line withline, a(c conventional,d) is applicable. view, If the (e,f )pore is appliedsize is [25larger,39,40 than,42]. b, which is indicated by a green-colored line with a conventional view, (e,f) is applied [25,39,40,42]. Depending on how the capacitor interacts with the carbon surface of the counter ion, these EDLCs can also be divided into endoderm and ectoderm capacitors, distinguished by the way the carbon surface interacts with the charged molecule layer. In the endoderm, counter ions enter the pores and

MicromachinesMicromachines 2020 2020, 11, 11, x, x 9 9of of 24 24

Micromachines 2020 11 DependingDepending, on on, 1125 how how the the capacitor capacitor interacts interacts with with the the carbon carbon surface surface of of the the counter counter ion, ion, these9 these of 25 EDLCsEDLCs can can also also be be divided divided into into endoderm endoderm and and ecto ectodermderm capacitors, capacitors, distinguished distinguished by by the the way way the the carbon surface interacts with the charged molecule layer. In the endoderm, counter ions enter the carbonform EDLs surface with interacts the external with electrodes. the charged This molecule type of chargelayer. In accumulation the endoderm, is observed counter for ions nanoporous enter the pores and form EDLs with the external electrodes. This type of charge accumulation is observed for porescarbons and with form negative EDLs with surface the curvatures. external electrodes. These structures This type are of conventionally charge accumulation shown onis observed the activated for nanoporous carbons with negative surface curvatures. These structures are conventionally shown on nanoporouscarbon, carbide-derived carbons with carbon, negative and surface template curvatur carbon.es. InThese the casestructures of an outerare conventionally surface capacitor, shown it was on the activated carbon, carbide-derived carbon, and template carbon. In the case of an outer surface theobserved activated that carbon, ions existed carbide-derived on the outer carbon, surface and of the template carbon particlescarbon. In [21 the]. case of an outer surface capacitor, it was observed that ions existed on the outer surface of the carbon particles [21]. capacitor,The third it was model observed tries that to interpret ions existed both on theories the outer with surface a cylindrical of the carbon and slit particles structure [21]. (Figure 5). The third model tries to interpret both theories with a cylindrical and slit structure (Figure 5). ThisThe theory third suggests model thattries thereto interpret are significant both theories cylinders with that a cylindrical affect the normalizedand slit structure capacitance (Figure and 5). This theory suggests that there are significant cylinders that affect the normalized capacitance and Thisslits whichtheory aresuggests based onthat micropores there are significant with inaccessible cylinders architectures that affect accordingthe normalized to the sizecapacitance of the pores. and slits which are based on micropores with inaccessible architectures according to the size of the pores. Asslits this which theory are combinedbased on micropores conventional with theories inaccessible about microarchitectures and macropores, according itto is the more size compatible of the pores. to As this theory combined conventional theories about micro and macropores, it is more compatible to Asinterpret this theory lots of combined phenomena. conventional The cylinder theories radius abou sizet micro plays and a role macropores, of consolidation it is more with compatible capacitance, to interpret lots of phenomena. The cylinder radius size plays a role of consolidation with capacitance, interpretsince the lots pore of size phenomena. reaches certain The cylinder values. radius Additionally, size plays the a role size of of consolidation the slits affects with the capacitance, normalized since the pore size reaches certain values. Additionally, the size of the slits affects the normalized sincecapacitance, the pore since size the reaches size is certain changed values. with variousAddition rangesally, the [47, 48size]. Bothof the the slit macropores affects the and normalized micropore capacitance, since the size is changed with various ranges [47,48]. Both the macropore and micropore capacitance,structure is captured since the by size scanning is changed with microscopy various ranges (SEM) [47,48]. images Both of highlythe macropore pores activated and micropore carbons structure is captured by scanning electron microscopy (SEM) images of highly pores activated (HAPC).structure Theis captured images by by SEM scanning are 1 µ melectron to 10 µ mmicroscopy with different (SEM) elements images (Figure of highly6)[47 ].pores activated carbonscarbons (HAPC). (HAPC). The The images images by by SEM SEM are are 1 1 μ μmm to to 10 10 μ μmm with with different different elements elements (Figure (Figure 6) 6) [47]. [47].

FigureFigure 5. 5. An An analysis analysis model model forfor for thethe the theorytheo theoryryof of of nanopores, nanopores, nanopores, showing showing showing a a model a model model that that that combines combines combines the the the theory theory theory of ofaof structurea a structure structure made made made of cylindersof of cylinders cylinders and and and micropores micropores micropores [25 [25,39,40,42,45,48]., 39[25,39,40,42,45,48].,40,42,45,48].

FigureFigure 6. 6. ( A(A−−CC) )Macro/micropole-dominated Macro/micropole-dominated carbon carbon on on SE SEMM (scanning (scanning electron electron microscope) microscope) images images Figure 6. (A–C) Macro/micropole-dominated carbon on SEM (scanning electron microscope) images withwith different different scale scale (10 (10 μ μmm and and 1 1 μ μm)m) and and ( D(D) )element element mapping mapping of of highly highly porous porous activated activated carbon carbon with different scale (10 µm and 1 µm) and (D) element mapping of highly porous activated carbon [47]. [47]. Reprinted with permission from Ref. [47]. Copyright 2017 American Chemical Society. [47]Reprinted. Reprinted with permissionwith permission from from Ref. [ 47Ref.]. Copyright[47]. Copyright 2017 American2017 American Chemical Chemical Society. Society.

CarbonCarbon materials materials with with positively positively curved curved surfaces, surfaces, such such as as nanofibersnano nanofibersfibers and and carbon carbon ions, ions, exhibit exhibit certaincertain types types of of behaviors. behaviors. Since SinceSince the thethe curvature curvaturecurvature of ofof grap graphenegraphenehene is is zero, zero, it it does does not not belong belong to to the the first first first two two categories,categories, and and it it is is classifiedclassified classified as as a a graphene graphene capacitor capacitor itself. itself. The capacitance in the plane is shown in series with the electrical double layer, wherein “a0” is The capacitance inin thethe plane plane is is shown shown in in series series with with the the electrical electrical double double layer, layer, wherein wherein “a0” “ isa0 few” is fewmolecularfew molecularmolecular distances distancesdistances which whichwhich are radii areare to radiiradii solvated toto solvatedsolvated cation, thecation,cation, half the ofthe distance halfhalf ofof fromdistdistanceance the slitfromfrom is expressedthethe slitslit isis

Micromachines 2020, 11, x 10 of 24 expressed as “b”, and “deff” is the effective separation distance between the electrode surface and the oppositely charged ions, which do not simply point charges, and the ionic radii place a role of the location of the charge densities. In the recently discovered carbon, a realistic approximation of the pore shape is a slit rather than a cylinder. Accordingly, a polarity located in-between two electrodes separated by a single layer of charge has been proposed. The capacitive formula for the sandwich capacitor, which is located in between two electrodes of the same polarity and is formed with a layer formed with a counter ion separated with 2b, which is indicated in the pore width (Figure 7). Because the counter electrode shares the total net charge of the opposite charged ions, slit capacitors can be considered as two capacitors in parallel, represented by the left and right-hand electrodes and the charged molecule layer in the middle, as shown in Figure 7. Therefore, the total capacitance Ctot is calculated as follows [44,49]:

Micromachines 2020, 11, 1125 𝐶 = . 10(6) of 25

In the equation, 𝜀 represents the dielectric constant of electrolyte, 𝜀 is the permittivity of a vacuum,as “b”, and and “ dtheeff” surface is the e ffareaective of separationthe electrode distance is represented between as the A electrode. surface and the oppositely charged ions, which do not simply point charges, and the ionic radii place a role of the location of the 𝑑 =(𝑏−𝑎 ) charge densities. , (7) whereIn deff the is recentlythe effective discovered separation carbon, distance a realistic between approximation the electrode of surface the pore and shape the isopposite a slit rather charged than ions,a cylinder. and the Accordingly, total capacitance a polarity of the located capacitor in-between is obtained two as electrodesfollows: separated by a single layer of charge has been proposed. The capacitive formula for the sandwich capacitor, which is located in 𝐶 = . (8) between two electrodes of the same polarity and is formed with a layer formed with a counter ion separated with 2b, which is indicated in the pore width (Figure7). Because the counter electrode shares All the analyses were performed in one region, similar to the interaction method between the the total net charge of the opposite charged ions, slit capacitors can be considered as two capacitors dielectric layer and the electrode, which was performed in a typical capacitator. The theory notified that in parallel, represented by the left and right-hand electrodes and the charged molecule layer in the the corrections of charge separation by the locations of charge densities are extremely important in middle, as shown in Figure7. Therefore, the total capacitance C is calculated as follows [44,49]: producing a reliable capacitance model, especially for micropores.tot However, as the present model is derived for parallel plates, we should notice that a precise structure of the electrode was not considered 2εrε0A for the reason of feasibility. The areal total capaCtotcitance= of the. system should be obtained [44]. (6) de f f

FigureFigure 7. 7. SchematicSchematic diagram ofof aa capacitorcapacitor formed formed by by a a solvated solvated cation cation with with a negative a negative polarity polarity located locatedin the middle in the of middle two electrodes of two electrodes separated by separated a single layer by a of single charge layer [39,44 of,50 charge]. [39,44,50].

4. NegativeIn the Capacitance equation, εr forrepresents Capacity the Enhancement dielectric constant of electrolyte, ε0 is the permittivity of a vacuum, and the surface area of the electrode is represented as A. A polarized molar solvent is generated in the compact double layer, which acts as an insulator, and becomes polarized to form the electrode surface and maintains dipole planes. Recently, theories de f f = (b a0), (7) and experiments have been proposed involving a combination− of negative capacitance and dielectric material,where de ffwhichis the can effective enhanc separatione the capacitance distance [16]. between the electrode surface and the opposite charged ions,The and electrical the total properties capacitance of of a dielectric the capacitor capacitor is obtained could asbe follows:expressed with Fd, which indicates free energy; Dd, which means displacement field; and Cd−1 inverse capacitance without external bias. A

2εrε0A f f−1 capacitor with a ferroelectric layer is also exCpressedtot = with. its corresponding parameters with F , C (8) b a and Pf mean spontaneous polarization. To fabricate− supercapacitors0 using negative capacitance, supercapacitorsAll the analyses should were be consolidated performed in with one region,a ferroelectric similar layer to the on interaction the dielectric method stack. between Secondly, the sincedielectric the depolarization layer and the field electrode, on the which ferroelectric was performed layer could in affect a typical the capacitator.spontaneous The polarization theory notified state, that the corrections of charge separation by the locations of charge densities are extremely important in producing a reliable capacitance model, especially for micropores. However, as the present model is derived for parallel plates, we should notice that a precise structure of the electrode was not considered for the reason of feasibility. The areal total capacitance of the system should be obtained [44].

4. Negative Capacitance for Capacity Enhancement A polarized molar solvent is generated in the compact double layer, which acts as an insulator, and becomes polarized to form the electrode surface and maintains dipole planes. Recently, theories and experiments have been proposed involving a combination of negative capacitance and dielectric material, which can enhance the capacitance [16]. The electrical properties of a dielectric capacitor could be expressed with Fd, which indicates 1 free energy; Dd, which means displacement field; and Cd− inverse capacitance without external bias. A capacitor with a ferroelectric layer is also expressed with its corresponding parameters with Ff, Micromachines 2020, 11, 1125 11 of 25

1 Cf− and Pf mean spontaneous polarization. To fabricate supercapacitors using negative capacitance, supercapacitors should be consolidated with a ferroelectric layer on the dielectric stack. Secondly, Micromachinessince the depolarization 2020, 11, x field on the ferroelectric layer could affect the spontaneous polarization11 state,of 24 interfacial charge, which is expressed with σIF, should be fabricated between ferroelectric and dielectric interfacial charge, which is expressed with σIF, should be fabricated between ferroelectric and layers to adjust Pf at zero voltage. Ferroelectric-dielectric capacitor could enhance its capacity with dielectric layers to adjust Pf at zero voltage. Ferroelectric-dielectric capacitor could enhance its the integration of electrical energy by polarization on the ferroelectric layer. As the ferroelectric layer capacity with the integration of electrical energy by polarization on the ferroelectric layer. As the based on negative capacitance could demonstrate a significantly high ratio, the capacity is abruptly ferroelectric layer based on negative capacitance could demonstrate a significantly high ratio, the enhanced at Va. These fabrication processes are interpreted in Figure8. capacity is abruptly enhanced at Va. These fabrication processes are interpreted in Figure 8.

Figure 8. IllustrationsIllustrations and and plots plots to to express express the the integration integration of capacitance caused by both the dielectric layer and the ferroelectric ferroelectric layer to enhance enhance capacity. capacity. ( a) Schematic of the the dielectric dielectric capacitor capacitor including including Dd-Ed curve, Cd−11Dd c-urve, and Fd-Dd curve without external bias. (b) Schematic of ferroelectric Dd-Ed curve, Cd− Dd c-urve, and Fd-Dd curve without external bias. (b) Schematic of ferroelectric capacitor capacitor including Pf-Ef1 curve, Cf−1-Pf curve shown as “s” and Ff-Pdf curve with polarization on 2Q. including Pf-Ef curve, Cf− -Pf curve shown as “s” and Ff-Pdf curve with polarization on 2Q. (c–e) Schematic (ofc– capacitore) Schematic compounded of capacitor with compounded both ferroelectric with both and ferroelectric dielectric layer and showsdielectric significant layer shows charge significant storage whichcharge isstorage expressed which as is the expressed area above as thethe Q-Varea curve.above (thec) shows Q-V curve. a state (c at) shows zero bias, a state (d) demonstratesat zero bias, (d at) demonstrates at Va, and (e) expresses at Vmax. Reprinted with permission from Ref. [16]. Copyright 2019. The Va, and (e) expresses at Vmax. Reprinted with permission from Ref. [16]. Copyright 2019. The authors authorsare under are Creative under Creative Commons Commons License License 4.0 (https: 4.0 //(https://creativecommons.org/licenses/by/4.0/).creativecommons.org/licenses/by/4.0/).

The improvement in capacitance resulted resulted in in an an incr increaseease in in the the storage storage energy energy up up to to 1.5 1.5 times. times. Ionic polarization inducesinduces ferroelectricferroelectric e effectsffects as as well well as as dipolar dipolar polarization. polarization. Ferroelectrics Ferroelectrics that that lead lead to tonegative negative capacitance capacitance may bemay applied be applied in the electrolyte,in the electrolyte, and ferroelectric and ferroelectric materials may materials be deposited may onbe depositedthe active carbonon the active layer tocarbon establish layer the to eestaffectblish of negative the effect capacitance. of negative capacitance. The insulatorinsulator layer layer formed formed through through oxidation oxidation on theon generalthe general carbon carbon layer played layer aplayed role in a inducing role in inducingcapacitance capacitance reduction. reduction. This is theThis reason is the reason why the why distance the distance increased increased as the as insulator the insulator layer layer was wasformed formed in the in molecularthe molecular dielectric dielectric layer, layer, and theand insulatorthe insulator layer layer formed formed through through oxidation oxidation has ahas low a lowdielectric dielectric constant. constant. However, However, ferroelectrics ferroelectric exhibiteds exhibited increased capacitancesincreased capacitances when they encounteredwhen they encounteredthe dielectric the layer dielectric and had layer a high and dielectric had a high constant. dielectric The constant. theory for The the theory enhancement for the ofenhancement capacitance withof capacitance a combination with a of combination the dielectric of and the ferroelectricdielectric and layers ferroelectric is given by:layers is given by: 𝑉 = Q. (9) V = . (9) d c Vd is the voltage applied to the dielectric capacitor. Vd is the voltage applied to the dielectric capacitor. 𝑉 =𝑎(𝑄−Δ𝑄) b(Q−Δ𝑄) . (10) 3 Vf is the voltage applied to theV ferroelectric= a(Q ∆ capacitor.Q) + b(Q ∆Q) . (10) f − − 𝑉 = 𝑎 𝑄−𝑎Δ𝑄𝑏(𝑄−Δ𝑄). (11)

Vtot is the voltage applied to the dielectric–ferroelectric capacitor. The total energy value for a capacitor combined with a combination layer is summarized as follows: ( ) 𝑊 = 𝑉 𝑄 𝑑𝑄. (12)

Micromachines 2020, 11, 1125 12 of 25

Vf is the voltage applied to the ferroelectric capacitor.   1 3 Vtot = + a Q a∆Q + b(Q ∆Q) . (11) c − −

Vtot is the voltage applied to the dielectric–ferroelectric capacitor. The total energy value for a capacitor combined with a combination layer is summarized as follows:

Z Q Wtot = Vtot(Q)dQ. (12) 0

The equation for energy is given by:   1 2 3 b 4 Wtot = a Q b∆QQ + Q . (13) 2c − − 4

Wtot is the total energy stored on the dielectric–ferroelectric capacitor. After normalizing W, the difference with the existing capacitance should be summarized to the whole storage system at the same voltage to use the steep slope region of the ferroelectric capacitor. Using the first-order 2 approximation on the w0 = CV /2 by Taylor’s expansion terms yields the following:

W  1 aC. (14)

w0 v=vmax −

The equation of normalizing W could explain the independence of b and rely only on the product “aC”. When the product is equal to the value of “ 1” for optimizing the capacitance matching, − the enhancement of W becomes approximately twice as great, meaning that the energy stored at the identical voltage is doubled [16,51]. The experimental results according to the theory did not show a significant difference and indicated an increase in energy storage of up to 1.5 times. However, the experimental result was only applicable to solid dielectric layers, and not to electrolytes; thus, it can be expected to have different effects when applied to supercapacitors using molecular solvent on electrolytes as the dielectric layer. Although theoretical efficiency is expressed as 2 times that of the conventional device, the capacitor with HZO (ferroelectric) layer and Ta2O5 (dielectric) layer (to match aC) has exhibited the enhancement (W/W ) about 1.1 due to the value of aC which is exhibited about 0.16. The thickness of both HZO 0 − and the Al2O3 layer is investigated about the effect on the energy storage properties. Figure9 depicts efficiency (%) versus discharged energy density W (J/cm 3) curves for 7.7 nm HZO thickness with h di − various Al2O3 thicknesses. Increasement of Al2O3 thickness which means volume fraction of the dielectric material affects not only the increase in the maximum obtainable W but also efficiency. h di The maximum obtainable W is achieved for 7.7 nm HZO and 4 nm Al O with 121 J cm 3. The ratio of h di 2 3 − dielectric/ferroelectric thickness would be an effective method to deal with W , while the ferroelectric h di characteristics remain. The frequency, including the small-signal capacitance per area and loss factor/pulse with W , h di temperature, and the cycling stability of the capacitor composed with ferroelectric/dielectric stack 2 are checked in Figure 10. The Tan (δ) measured with 0.6% during C is about 1.3 µF cm− . While the 3 cycles are up to 1,000,000 times, the discharged energy density stays at about 38 J cm− below the voltage of 10.6 V. While the efficiency is maintained at about 99%, W decreases to about 31 J cm 3 h di − on 100,000,000 times of cycling by the pinning of ferroelectric domains which would not affect the capacitance enhancement. The capacitor composed with HZO layer exhibited high both discharged energy density and efficiency during an increase in temperature of up to 150 ◦C. These experimental results show that a supercapacitor based on negative capacitance effect operation could be feasible based on the described designs. Micromachines 2020, 11, x 12 of 24

The equation for energy is given by:

𝑊 = −𝑎 𝑄 −𝑏Δ𝑄𝑄 𝑄. (13)

Wtot is the total energy stored on the dielectric–ferroelectric capacitor. After normalizing W, the difference with the existing capacitance should be summarized to the whole storage system at the same voltage to use the steep slope region of the ferroelectric capacitor. Using the first-order approximation on the 𝑤 =𝐶𝑉 /2 by Taylor’s expansion terms yields the following: ≅1−𝑎𝐶. (14) The equation of normalizing W could explain the independence of b and rely only on the product “aC”. When the product is equal to the value of “−1” for optimizing the capacitance matching, the enhancement of W becomes approximately twice as great, meaning that the energy stored at the identical voltage is doubled [16,51]. The experimental results according to the theory did not show a significant difference and indicated an increase in energy storage of up to 1.5 times. However, the experimental result was only applicable to solid dielectric layers, and not to electrolytes; thus, it can be expected to have different effects when applied to supercapacitors using molecular solvent on electrolytes as the dielectric layer. Although theoretical efficiency is expressed as 2 times that of the conventional device, the capacitor with HZO (ferroelectric) layer and Ta2O5 (dielectric) layer (to match aC) has exhibited the enhancement (W/W0) about 1.1 due to the value of aC which is exhibited about −0.16. The thickness of both HZO and the Al2O3 layer is investigated about the effect on the energy storage properties. Figure 9 depicts efficiency (%) versus discharged energy density 〈Wd〉 (J/cm−3) curves for 7.7 nm HZO thickness with various Al2O3 thicknesses. Increasement of Al2O3 thickness which means volume fraction of the dielectric material affects not only the increase in the maximum obtainable 〈Wd〉 but also efficiency. The maximum obtainable 〈Wd〉 is achieved for 7.7 nm HZO and 4 nm Al2O3 with 121 Micromachines 2020, 11, 1125 13 of 25 J cm−3. The ratio of dielectric/ferroelectric thickness would be an effective method to deal with 〈Wd〉, while the ferroelectric characteristics remain.

Micromachines 2020, 11, x 13 of 24

The frequency, including the small-signal capacitance per area and loss factor/pulse with 〈Wd〉, FigureFigure 9. 9.Plots Plots of of capacitor capacitor compoundedcompounded with both a a ferroelectric ferroelectric (HZO) (HZO) and and dielectric dielectric layer layer (Al (Al2O3)O ) temperature, and the cycling stability of the capacitor composed with ferroelectric/dielectric stack2 3 are withwith various various layer layer thicknesses. thicknesses. ((aa)E) Efficiencyfficiency (%)(%) versusversus di dischargedscharged energy energy density density 〈WWd〉(J/cm(J/cm−3) for3) for checked in Figure 10. The Tan (δ) measured with 0.6% during C is about 1.3 μF cm−h2. Whiledi the− cycles 7.77.7 nm nm HZO HZO compounded compounded withwith 1.51.5 nmnm toto 44 nmnm AlAl2O3 layerlayer thicknesses. thicknesses. (b ()b Efficiency)Efficiency (%) (%) versus versus are up to 1,000,000 times, the discharged energy density2 3 stays at about 38 J cm−3 below the voltage of discharged energy density 〈Wd〉(J/cm−33) for 4 nm Al2O3 compounded with 7.7 nm and 11.3 nm HZO discharged energy density Wd (J/cm− ) for 4 nm Al2O3 compounded with 7.7 nm and 11.3 nm HZO 10.6 V. While the efficiencyh isi maintained at about 99%, 〈Wd〉 decreases to about 31 J cm−3 on thicknesses.thicknesses. (c ()c Charging) Charging ( Q(Q),c), Discharging Discharging ((QQd), and and Charge Charge loss loss ( (QQloss), ),which which means means QcQ-Q-dQ, according, according 100,000,000 times of cyclingc by the pinningd of ferroelectric domainsloss which wouldc dnot affect the to measured charges versus applied voltage at capacitor fabricated with the 4 nm Al2O3/7.7 nm HZO. to measured charges versus applied voltage at capacitor fabricated with the 4 nm Al2O3/7.7 nm HZO. capacitance(d) Charges enhancement. versus applied The voltage capacitor with composedan experimental with andHZO theoretical layer exhibited capacity high increase both (W discharged/W0). (d) Charges versus applied voltage with an experimental and theoretical capacity increase (W/W0). energyReprinted density with and permission efficiency from during Ref. an[16]. increase Copyright in temperature2019. The authors of up are to under 150 °C.Creative These Commons experimental Reprinted with permission from Ref. [16]. Copyright 2019. The authors are under Creative Commons resultsLicense show 4.0 that (https://creativec a supercapacitorommons.org/licenses/by/4.0/). based on negative capacitance effect operation could be feasible License 4.0 (https://creativecommons.org/licenses/by/4.0/). based on the described designs.

Figure 10. Plots of major factors with the stability of the capacitor composed with Figure 10. Plots of major factors with the stability of the capacitor composed with ferroelectric/dielectric ferroelectric/dielectric stack. (a) Both capacitance density C (left) and loss factor tan (δ) (right) versus stack. (a) Both capacitance density C (left) and loss factor tan (δ) (right) versus frequency without DC frequency without DC bias for frequency stability. (b) Discharged energy density 〈Wd〉 (left) and bias for frequency stability. (b) Discharged energy density Wd (left) and efficiency (right) versus efficiency (right) versus cycling times under 10.6 V for cyclingh stability.i (c) 〈Wd〉 versus frequency by cycling times under 10.6 V for cycling stability. (c) Wd versus frequency by pulse under 12.7 V for pulse under 12.7 V for frequency stability. (d) 〈Wd〉 h(left)i and efficiency (right) versus temperature at frequency stability. (d) W (left) and efficiency (right) versus temperature at 14.4 V for temperature 14.4 V for temperatureh dstability.i Reprinted with permission from Ref. [16]. Copyright 2019. The stability.authors Reprinted are under withCreative permission Commons from License Ref. [4.016 ].(https://creativecommons.org/licenses/by/4.0/). Copyright 2019. The authors are under Creative Commons License 4.0 (https://creativecommons.org/licenses/by/4.0/). Additionally, the fact-confirming evidence and theoretical perspective for negative capacitance and inductance induced by polarization switching is also discussed in Figure 11. The analysis is described with the plot of the real and imaginary Nyquist impedances of ferroelectric capacitors based on nanoscale zirconium oxide (nano-f-ZrO2). Measurements were conducted with an impedance analyzer in the frequency range from 1 MHz to 10 mHz at 1 V direct current (DC) voltage and 10 mV alternating current (AC) perturbation. You can see that the RC circuit (shown in the inset of Figure 11) gives a semi-circle of negative imaginary impedance in the Nyquist plot [52].

Micromachines 2020, 11, 1125 14 of 25

Additionally, the fact-confirming evidence and theoretical perspective for negative capacitance and inductance induced by polarization switching is also discussed in Figure 11. The analysis is described with the plot of the real and imaginary Nyquist impedances of ferroelectric capacitors based on nanoscale zirconium oxide (nano-f-ZrO2). Measurements were conducted with an impedance analyzer in the frequency range from 1 MHz to 10 mHz at 1 V direct current (DC) voltage and 10 mV alternating current (AC) perturbation. You can see that the RC circuit (shown in the inset of Figure 11) givesMicromachines a semi-circle 2020, 11 of, x negative imaginary impedance in the Nyquist plot [52]. 14 of 24

FigureFigure 11. 11. Nyquist plot plot based based on on the the real real and and imaginary imaginary parts parts at the at complex the complex impedance impedance (Zr + (ZrjZi) +of jZi)MFM of MFMcapacitor capacitor stacked stacked with Pt–ZrO2–Pt. with Pt–ZrO2–Pt. Resistor-c Resistor-capacitorapacitor (RC) dummy (RC) dummycell provided cell provided by Metrohm by MetrohmAUTOLAB, AUTOLAB, KM Utrecht, KM Netherlands Utrecht, Netherlands is exhibited is exhibitedin an insert in with an insert RC component with RC component 100 Ω and 100 an ΩRC andelement an RC (1 element μF) capacitance (1 µF) capacitance with 1 MΩ with resistance 1 MΩ resistanceintegrated integrated in parallel. in Reprinted parallel. Reprintedwith permission with permissionfrom Ref. from[52]. Ref. Copyright [52]. Copyright 2019. 2019.The author The authorss are areunder under Creative Creative Commons Commons LicenseLicense 4.04.0 (https:(https://creativecommons.org/licenses/by/4.0/).//creativecommons.org/licenses/by/4.0/).

5.5. Theories Theories of of Pseudocapacitors Pseudocapacitors InIn the the mid-1920s, mid-1920s, scientists scientists discovered discovered the the presence presence of of pseudocapacitance pseudocapacitance inin some some electrode electrode materialsmaterials of of metal-ion metal-ion batteries, batteries, and and they they called called this this phenomenon phenomenon specific specific adsorption, adsorption, which which is is based based onon chemical chemical processes processes [ 53[53].]. TheThe termterm “pseudocapacitance” “pseudocapacitance” isis synonymoussynonymous withwith thethe surface surface redox redox reaction,reaction, which which is is based based on on chemical chemical reactions, reactions, such such as as batteries. batteries. When When electrical electrical energy energy is storedis stored in ain solida solid material material by by chemical chemical reaction, reaction, a material a material with with high high electrochemical electrochemical potential potential is formed. is formed. AsAs the the high high electrochemical electrochemical potential potential of material of material based onbased the chemicalon the chemical reaction ofreaction a solid substanceof a solid issubstance mainly stored is mainly in chemical stored bonds in chemical which includesbonds which ionic bonds,includes covalent ionic bonds, bonds, andcovalent hydrogen bonds, bond, and thehydrogen material bond, is disconnected the material or changed is disconnected with chemical or bondschanged to thewith low chemical electrochemical bonds potentialto the low of theelectrochemical material, the electrochemical potential of the reactions material, emitting the electrochemical chemical energy reactions that canemitting be transferred chemical as energy electrical that energycan be istransferred generated. as The electrical difference energy in the is reactiongenerate betweend. The difference the state ofin chargethe reaction and state between of discharge the state canof charge be explained and state by the of followingdischarge equationcan be explaine with thed Gibbs by the free following energy: equation with the Gibbs free energy: ∆G = ∆U + P∆V V∆P T∆S S∆T. (15) Δ𝐺 = Δ𝑈 𝑃Δ𝑉− − − 𝑉Δ𝑃− − 𝑇Δ𝑆−𝑆Δ𝑇. (15) TheThe energy energy transformation transformation of aof given a given chemical chemical reaction reaction can be can explained be explai byned the Gibbsby the free Gibbs energy. free Inenergy. the equation, In the equation,G is the G Gibbs is the free Gibbs energy, free energy, the internal the internal energy energy of the system of the system is represented is represented as U, theas pressureU, the pressure is expressed is expressed as P, the temperatureas P, the temperature is expressed is asexpressedT, and V asis theT, and volume. V is the The volume. capacitance The ofcapacitance the pseudocapacitor of the pseudocapacitor can be 10–100 can times be higher10–100 thantimes that higher of the than EDLCs that of [20 the,54 ,EDLCs55]. [20,54,55]. BothBoth pseudocapacitors pseudocapacitors and and batteries batteries are based are based on chemical on chemical reactions, reactions, but there arebut major there di arefferences major indifferences the charge in andthe dischargecharge and behaviors discharge of behaviors pseudocapacitive of pseudocapacitive materials in materials different in time different domains. time domains. Pseudocapacitance, a faradaic activation involving surface or near-surface redox phenomenon, offers a means of achieving high energy density at high charge-discharge rates. Pseudocapacitors are not capacitors with a conventional mechanism perspective of using only electrical energy. However, output properties are shown as capacitors rather than batteries. In rare cases, pseudocapacitive materials (lithium in Nb2O5) have exhibited interlayer insertion processes which are essential phenomena of intercalation pseudocapacitance rather than rapid reactions through conventional surface redox reactions. Since the concept of “intercalation pseudocapacitance” was proposed, it was no longer the synonymous mean of the surface redox reaction. The intercalation pseudocapacitance shows a faradaic charge storage mechanism, which is

Micromachines 2020, 11, 1125 15 of 25

Pseudocapacitance, a faradaic activation involving surface or near-surface redox phenomenon, offers a means of achieving high energy density at high charge-discharge rates. Pseudocapacitors are not capacitors with a conventional mechanism perspective of using only electrical energy. However, output properties are shown as capacitors rather than batteries. In rare cases, pseudocapacitive materials (lithium in Nb2O5) have exhibited interlayer insertion processes which are essential phenomena of intercalation pseudocapacitance rather than rapid reactions through conventional surface redox reactions. Since the concept of “intercalation pseudocapacitance” was proposed, it was no longer the synonymous mean of the surface redox reaction. The intercalation pseudocapacitanceMicromachines 2020, 11, x shows a faradaic charge storage mechanism, which is more similar to batteries15 of 24 than the redox pseudocapacitance. The cyclic voltammetry results of intercalation pseudocapacitance more similar to batteries than the redox pseudocapacitance. The cyclic voltammetry results of appear closer to that of a battery than a capacitor. Supercapacitors can satisfy their purpose by intercalation pseudocapacitance appear closer to that of a battery than a capacitor. Supercapacitors design, from EDLCs showing typical capacitor characteristics to pseudocapacitors showing a medium can satisfy their purpose by design, from EDLCs showing typical capacitor characteristics to performance as capacitors and batteries (Figure 12)[10,56,57]. pseudocapacitors showing a medium performance as capacitors and batteries (Figure 12) [10,56,57].

Figure 12. Schematic illustration of a supercapacitor classified by capacitive type. (a) Schematic Figure 12. Schematic illustration of a supercapacitor classified by capacitive type. (a) Schematic illustration of EDLC, which has cations and anions that are physically layered from electrodes. The illustration of EDLC, which has cations and anions that are physically layered from electrodes. Theschematic schematic cyclic cyclic voltmeter voltmeter has has the the same same shape shape as as a acapacitor. capacitor. (b (b) )Schematic Schematic illustration illustration of redoxredox pseudocapacitor,pseudocapacitor, which which stores stores and and releases releases energy ener throughgy through chemical chemical adsorption adsorption reactions reactions at the surface at the ofsurface the electrode. of the electrode. The schematic The schematic cyclic voltmeter cyclic voltmeter is not the is not same the as same a capacitor as a capacitor but has but a similar has a similar shape. (shape.c) Schematic (c) Schematic illustration illustration of intercalation of intercalation pseudocapacitor, pseudocapacitor, which stores which and stores releases and energyreleases through energy chemicallythrough chemically transform reactionstransform based reactions on the based faradaic on chargethe faradaic storage charge mechanism. storage The mechanism. schematic cyclic The voltmeterschematic shows cyclic thevoltmeter characteristics shows between the characteri the capacitorstics between and the battery.the capacitor (d) Schematic and the illustration battery. ( ofd) hybridSchematic capacitor illustration integrated of hybrid both electrodecapacitor types,integrated which both is EDLC electrode and redoxtypes, pseudocapacitivewhich is EDLC and material. redox (pseudocapacitivee) Schematic illustration material. of hybrid (e) Schematic capacitor integratedillustration with of bothhybrid electrode , integrated which are EDLCwith both and intercalationelectrode types, pseudocapacitive which are EDLC material and intercalat [5,6,10,21ion,55 pseudocapacitive]. material [5,6,10,21,55].

PseudocapacitivePseudocapacitive types types are are usually usua basedlly based on electrode on electrode materials materials regarding regarding their specific their capacitance. specific Therecapacitance. are various There materials are various with theirmaterials theoretical with capacitancetheir theoretical listed incapacitance Table3. The listed composites in Table exhibited 3. The theircomposites characteristics exhibited according their characteristics to the properties according of materials to the that properties have identical of materials advances that and have challenges. identical advances and challenges. CNT and graphene, which demonstrate stability in various regions including chemical, thermal, and mechanical consolidations, could demonstrate a composite showing the intermediate properties of the other materials. Table 4 shows various data, including graphene/MnO2/CNTs, GR/MCNTs/MnO2, CNTs/PANI/GR, CNTs/GO/PANI, and CNTs/PANI, which exhibited specific capacitances of 372, 355, 477, 569, 589, and 838, which are significantly enhanced compared to a device manufactured with only graphene and CNTs. Additionally, these composite exhibited life cycling times with significant enhancement, which is considered as the fatal limitation of conducting polymers. 3D Co3O4/MnO2, 3D Co3O4/MnO2, Co3O4/MnO2@GO, Co3O4/MnO2@GO, NiCo2O4–MnO2/GF, Co3O4, and CuCo2O4/NiO are demonstrated with a significant capacitance compared to the other composite due to the characteristics of both Co3O4 and NiO, which have theoretical specific capacitances of 3560 and 2573. However, these composites have fatal drawbacks in both materials, since the composites have not only advantages but also drawbacks.

Micromachines 2020, 11, 1125 16 of 25

CNT and graphene, which demonstrate stability in various regions including chemical, thermal, and mechanical consolidations, could demonstrate a composite showing the intermediate properties of the other materials. Table4 shows various data, including graphene /MnO2/CNTs, GR/MCNTs/MnO2, CNTs/PANI/GR, CNTs/GO/PANI, and CNTs/PANI, which exhibited specific capacitances of 372, 355, 477, 569, 589, and 838, which are significantly enhanced compared to a device manufactured with only graphene and CNTs. Additionally, these composite exhibited life cycling times with significant enhancement, which is considered as the fatal limitation of conducting polymers. 3D Co3O4/MnO2, 3D Co3O4/MnO2, Co3O4/MnO2@GO, Co3O4/MnO2@GO, NiCo2O4–MnO2/GF, Co3O4, and CuCo2O4/NiO are demonstrated with a significant capacitance compared to the other composite due to the characteristics of both Co3O4 and NiO, which have theoretical specific capacitances of 3560 and 2573. However, these composites have fatal drawbacks in both materials, since the composites have not only advantages but also drawbacks. These supercapacitors exhibited fatal limitations on cycling times or chemical, physical, or thermal stabilities.

Table 3. The theoretical capacitance of various composite with issues.

1 Composite Specific Capacitance (F g− ) Issue NiO 2573 Cycling stability Co3O4 3560 Toxicity, Cycling stability CuCo2O4 984 Cycling stability V2O5 2120 Low conductivity, Cycling stability MnO2 1380 Limited thickness layer RuO2, xH2O 1200–2200 High cost, blockage of accessible surface area

Table 4. Specific capacitance of various composites.

Specific Electrolyte Current Density Composite Capacitance Year Ref. 1 Solution (ic)/Scan Rate (v) (F g− ) 1 MnO2/porous carbon 459 KOH 1.0 A g− 2014 [58] 1 GO/MnO2 315 1 m Na2SO4 0.5 A g− 2015 [59] 1 3D GR/MnO2 267 1.5 M Li2SO4 200 mV s− 2018 [60] 2 3D Co3O4/MnO2 1397 2.0 M KOH 1 mA cm− 2016 [61] 1 3D Co3O4/MnO2 1184 1 M LiPF6 1.0 A g− 2014 [62] 1 MnO2@GO 1518 1.0 M Na2SO4 1.0 A g− 2020 [63] 1 Co3O4/MnO2@GO 1358 1.0 M Na2SO4 1.0 A g− 2020 [63] 1 Co3O4/MnO2@GO 1718 1.0 M Na2SO4 1.0 A g− 2020 [63] 1 graphene/MnO2/CNTs 372 1 m Na2SO4 0.5 A g− 2012 [64] 1 GR/MCNTs/MnO2 355 1 m Na2SO4 0.3 A g− 2013 [65] 2 CNT/TiO2/PANI 477 0.5 M H2SO4 0.4 µA mm− 2013 [66] 1 CNTs/PANI/GR 569 1.0 M HCl 0.1 A g− 2011 [67] 1 CNTs/GO/PANI 589 1.0 M H2SO4 0.2 A g− 2013 [68] 1 CNTs/PANI 838 1.0 M H2SO4 1 mV s− 2011 [69] 1 MWCNTs/PANI 233 H3PO4-PVA gel 1 A g− 2013 [70] 1 MWCNTs/PPy 427 1.0 M Na2SO4 5 mV s− 2010 [71] 1 MWCNTs/PPy/MnO2 365 0.5 M Na2SO4 5 mV s− 2014 [72] 1 MWCNTs/PANI 560 0.1 M H2SO4 1 mV s− 2010 [73] 1 MWCNTs/Pd/PANI 920 1.0 M H2SO4 2 mV s− 2013 [74] 1 d-CNTs/PPy 587 0.1 M NaClO4 3 A g− 2012 [75] 2 C60-PANI-EB 776 1.0 M H2SO4 1 mA cm− 2012 [76] 1 MnO2/AC 324 Na2SO4 0.1 A g− 2008 [77] 2 MnO2/AC 345 KOH 10 mA cm− 2015 [78] 1 δ-MnO2/AC 360.5 1M NaNO3 4 A g− 2020 [79] Micromachines 2020, 11, 1125 17 of 25

Table 4. Cont.

Specific Electrolyte Current Density Composite Capacitance Year Ref. 1 Solution (ic)/Scan Rate (v) (F g− ) 1 MnO2/CNT 199 1.0 M Na2SO4 0.1 A g− 2008 [80] 1 MnO2/CNT 325.5 1 m Na2SO4 0.3 A g− 2009 [81] 1 NiCo2O4–MnO2/GF 2577 1 m Na2SO4 1 A g− 2017 [82] 1 PPy-graphene 165 1.0 M NaCl 1 A g− 2010 [83] 1 PAA@MnO2/PPy 564 0.2 M FeCl3 10 mV s− 2018 [84] PAA@MnO2/PPy 692 0.2 M FeCl3 0.5 A/g 2018 [84] 1 PAA@ MnO2 288 0.2 M FeCl3 10 mV s− 2018 [84] 2 PANI-Si 409 0.5 M H2SO4 40 mA cm− 2010 [85] 2 PANI/MnO2 1292 1.0 M LiClO4 4.0 mA cm− 2010 [86] 1 MWCNTs/MnO2/PPy 806 1 M Na2SO4 1 A g− 2019 [87] 1 NiO 1700 6 M Hg/HgO KOH 2 A g− 2012 [88] 1 Co3O4 2735 to 1471 2 M KOH 2 to 10 A g− 2012 [89] 1 CuCo2O4/NiO 2219 1.0 M NaOH 1 A g− 2017 [90] 1 CuCo2O4 743 1.0 M NaOH 1 A g− 2017 [90] 1 NiO 1296 1.0 M NaOH 1 A g− 2017 [90] Note: Polypyrrole (PPy), Polyaniline (PANI), Polyacrylic acid (PAA), GR (graphene), GO (graphene oxide), MCNTs (carbon nanotubes), MWCNT (Multi-walled carbon nanotube), and CNT (carbon nanotube).

Additionally, there are devices taking advantage of properties with specific architectures such as micro flowers, nanosheets, nanobelts, etc. ZnCo2O4 with Ni foam fabricated in a microflower structure exhibited a superior specific capacitance with a reasonable cycling stability. While CuCo2O4 with 1 Ni foam consolidated with nanosheets types shows a specific capacitance of 1330 F g− with 70% capacitance retention after 5000 charge/discharge cycles, CuCo2O4 fabricated in a nanobelt structure 2 exhibited a 809 specific capacitance with a 127% capacitance retention after 1800 cycles at 2 mA cm− in Table5.

Table 5. Composite with specific architectures with various factors.

Specific Electrolyte Current Density Composite Capacitance Cycling Stability Year Ref. 1 Solution (ic)/Scan Rate (v) (F g− ) ZnCo O -Ni foam 2 4 2256 1 M KOH 2 mA cm 2 2017 [91] (microflowers) 90% capacitance − retention after 2000 cycles at ZnCo O -Ni foam 2 4 170010 mA cm 2 1 M KOH 30 mA cm 2 2017 [91] (microflowers) − − ZnCo O -Ni foam 2 4 2037 1 M KOH 2 mA cm 2 2017 [91] (nanosheets) 80% capacitance − retention after 2000 cycles at ZnCo O -Ni foam 2 4 71910 mA cm 2 1 M KOH 30 mA cm 2 2017 [91] (nanosheets) − − 127% capacitance CuCo2O4 1 809 retention after 1800 cycles at 2.0 M KOH 0.667A g− 2015 [92] (nanobelts) 2 2 mA cm− CuCo O -Ni foam 2 4 1330 3 M KOH 2 A g 1 2018 [93] (nanosheets) 70% capacitance − retention after 5000 cycles at CuCo2O4-Ni foam 1 9382 A g− 3 M KOH 60 A g 1 2018 [93] (nanosheets) −

Intercalation pseudocapacitance is based on the faradaic charge storage mechanism conventionally used for batteries with high energy densities rather than reversible or near-surface redox reactions. The material with pseudocapacitive characteristics exhibited the electrical response of the EDLC identically. As the charge state changes continuously with the potential, it results in a proportional constant, which is the reason why these devices could be considered as a capacitance. Micromachines 2020, 11, 1125 18 of 25

In pseudocapacitive electrodes, different charge storage mechanisms can be distinguished potential difference deposition, redox reactions of transition metal oxides, intercalation pseudocapacitance, reversible electrochemical doping, and dedoping in conductive polymers. The materials used to fabricate these electrodes are usually metal oxides, mixed metal oxides, sulfides, selenides, nitrides, conductive polymers, carbon materials, and composites with carbon materials. Since materials with various elements and structures may exhibit different electrochemical potentials, it is necessary to carefully design the chemical bonding of solid materials that can be controlled by the surface and interfacial chemical reactions [6,94]. The limitations of the pseudocapacitor are similar to that of the battery, and the degree is also located between that of the EDLC and the battery. As the chemical mechanism of the slow parody process occurs, the performance of specific power by the pseudocapacitor is generally lower than that of the EDLCs. Electrodes exhibiting pseudocapacitance are susceptible to expansion and contraction during charge and discharge cycles, resulting in a poor mechanical stability and shortened cycle life; they are very trivial compared to batteries, which are based on chemical energy with the bulk electrode region and not only the surface layer. However, they show a higher energy storage capacity than that of EDLCs and have both an electrochemically reversible structure to achieve a high cycle compared to a battery and a high power density.

6. Other Next-Generation Supercapacitors

6.1. Redox Electrolyte Enhanced Supercapacitors As there are confinement strategies with enhancement on the energy storage capacity with supercapacitors, utilizing various materials could be attractive as electrode materials and structures, which have a higher energy and power than conventional supercapacitors or electrolyte materials. There are different suggestions that are not a conventional method or based on negative capacitance effect but that rely on the redox additive/active electrolyte. Alternative approaches can demonstrate enhancement in various electrode materials; for instance, active carbon electrodes could demonstrate a 1 1 1 2 1 specific capacitance with 220 F g− , 630.6 F g− , and 901 F g− at 2.65 mA cm− and 1 A g− . Active carbon electrolyte devices have redox additive electrolytes such as VOSO4 or hydroquinone [95–97]. Electrolytes could be divided into three groups. The first group is redox additive-liquid electrolytes. Secondly, there are the redox-active liquid electrolytes. The last group is redox additive-polymer gel electrolytes. Supercapacitors based on redox electrolyte-enhanced supercapacitors could be comparable to batteries. However, there are issues to solve regarding redox electrolyte-enhanced supercapacitors compared to the liquid electrolytes with a low ion mobility and ion accessibility with the electrode, causing a poor contact area [23].

6.2. Piezoelectric Supercapacitors Piezoelectric materials have attracted particular attention to energy harvesting technology by means of nanogenerators. As devices fabricated with zinc oxide nanowire arrays could be bent, strain field and charge separation are formed by the coupling of a piezoelectric layer with zinc oxides which exhibit semiconducting properties. The separator utilized on a supercapacitor could be replaced by a piezoelectric film which transforms physical stress with piezoelectric potential [98]. Conventional piezoelectric films that have issues of stability, flexibility, and cost had been proposed with promising alternatives such as polyamide (PA) or polyvinylidene fluoride (PVDF) with an enhanced β crystalline phase content [99]. The piezoelectric layer stressed by the environment generates the potential to drive anions or cations with electrochemical potential. The polarization of ions generates a potential difference with the thickness of the separator. The piezoelectric field induces the movement of H+ to screen piezoelectric field. To obtain chemical equilibrium, oxidation and redox reactions should occur on the positive electrode and negative electrode. Devices operating based on mechanical potential without external electric energy have been demonstrated [100]. A flexible Micromachines 2020, 11, 1125 19 of 25 piezoelectric device has been proposed with carbon clothes with a maximum specific capacitance of 2 2 357.6 F m− at 8 Am− [101]. A PVDF-ZnO film supercapacitor applied a mechanical force charged from 35 to 145 mV with 300 s [102]. Additionally, other devices generate electrochemical energy with a different method to that of piezoelectric supercapacitors; they are called thermal self-charging supercapacitors. Thermal self-charging with an electrochemical double layer capacitor in an open circuit with a bias of 80–300 mV is achieved with a thermal energy of 65 ◦C. The charge generation could be achieved at room temperature after the device is heated at a high temperature.

6.3. Electrochromic Supercapacitors Electrochromic supercapacitors quantifiably determine their electrical capacity by an optical method based on some electrode materials which had demonstrated their color transformation during electrochemical operation [103]. The visual change is the most noticeable thing to figure out. Normalized optical density could be associated with energy density variation, since color according to remaining electrical capacity needs a significant definition. Calibration curves are dependent on various hybrid supercapacitors, which could be a method to measure energy capacity. A tungsten electrode which changes from transparent to blue is considered the most promising device due to compatibility with both supercapacitors and electrochromic materials [104]. Electrochromic supercapacitors are a method to develop devices with imaginative applications.

6.4. Fiber Shaped Supercapacitors Flexible and portable electronic devices have potential to become mainstream in the Fourth Industrial Revolution, which facilitates the emergence of a hyper-connected society, and it is the basis for data processing between people and things as well as between things and things. The emerging fiber-shaped supercapacitors have helped to create ESSs or portable electronic systems. However, conventional devices have a specific structure based on a planar format. These formats had been changed by fabrication with carbon nanotubes (CNTs), which demonstrate high flexibility, tensile strength, electrical conductivity, and stability. However, CNT materials 1 have significantly low specific capacitances, with 4.5 to 5 F g− [105]. Additionally, there are challenges of strategies with electrochemical performance and physical stability, including flexibilities. 2 1 The stretchable asymmetric FSS demonstrated a specific capacitance of 60.435 mF cm− at 10 mV s− . To enhance the energy density, ppy, which demonstrates a broad range of negative potential and MnO2, would be electrodeposited onto CNT film (CNT@PPy, CNT@MnO2) due to the inherently high energy density of composites. The FSS demonstrated significant stabilities with an 88% capacitance after 200 stretching cycles with physically a 20% strain [106].

6.5. Shape Memory Supercapacitors Shape memory supercapacitors are usually based on shape memory polymer fibers with recoverable materials such as CNT and graphene. These supplement materials reinforce shape memory polymer fibers with electrical conductivity, recovering stability, or reduction in stress. Representative materials of shape memory supercapacitors are usually MWCNT, PANI, and polymer electrolyte, which could be recovered through thermal energy, triggering shape memory effects [107]. MWCNT has unique properties based on a large aspect ratio that could endure huge mechanical stress without noticeable degradation [108]. CNT-coated polyurethane–poly(ε-caprolactone) PU-PCL electrode demonstrated a superior cycle stability, with 96.5% capacitance retention after 10,000 cycles due to the recovering stability of the substrate with poly(acrylic acid)–poly(ethylene oxide), which is used as electrolyte. Additionally, the PU-PCL supercapacitor exhibits a specific capacitance with 1 1 37 F g− at 0.5 A g− [109]. MWCNT/MnO2/SMPU hybrid film not only could be changed with various structures with mechanical forces but also could reconsolidate its initial formation within 1 sec when the thermal environment is adjusted above a glass transition temperature. The device demonstrates Micromachines 2020, 11, 1125 20 of 25

2 a specific capacitance of 69.1 mF cm− and a cycle stability with a 95.3% capacitance retention after 3000 cycles [110]. These recoverable and structure adjustable supercapacitors are considered as the most prospective devices for applications such as electronic textiles, flexible electronic systems, and wearable devices [109,110].

7. Conclusions As the world which facilitates the emergence of a hyper-connected society requires next-generation ESS, supercapacitors draw attention as one of the candidates with their superior properties in many key indicators, such as power density, cycle stability, temperature range, and safety, except for energy storage capacity, which could be considered as the most important factor in the prevalent view. The major frameworks to solve these issues conventionally focus on utilizing different materials and facial structures. We discuss various electrodes—not only materials such as ACs, MWCNT, metal oxides (MnO2, RuO2, etc), metal hydroxides (Ni(OH)2, Co(OH)2, etc.), and conducting polymers (PANI, PPy, etc), but also redox electrolyte. However, it is not enough to satisfy the demand for energy capacity with only these strategies. In this work, we synthesize not only the charge repositioning principle according to supercapacitor types, but also various interaction theories of electrodes with electrolytes which consist of both a dielectric layer and a charge separation layer. We suggest that a novel theory reported recently by a research group focusing on negative capacitance demonstrated the energy storage capacity enhancement through the combination of ferroelectric and dielectric layers could be one of the solutions to the noticeable hurdle that supercapacitors should overcome. Additionally, next-generation supercapacitors, including redox electrolyte-enhanced capacitors, which rely on the enhancement on energy storage capacity with the redox electrolyte; piezoelectric supercapacitors, which could generate energy by means of nanogenerators, electrochromic supercapacitors, which could quantifiably determine their electrical capacity via appearance, and fiber-shaped supercapacitors and shape memory supercapacitors, which are presented for wearable or stretchable energy storage systems, were described regarding the promising potential of supercapacitors.

Author Contributions: Writing—original draft preparation, J.S.; writing—review and editing J.S. and C.S.; supervision, C.S. All authors have read and agreed to the published version of the manuscript. Funding: This work was supported by the National Research Foundation of Korea (NRF) through a grant funded by the Korean Government (MSIP) (No. 2020R1A2C1009063). Conflicts of Interest: The authors declare no conflict of interest.

References

1. Jo, J.; Choi, W.Y.; Park, J.-D.; Shim, J.W.; Yu, H.-Y.; Shin, C. Negative capacitance in organic/ferroelectric capacitor to implement steep switching MOS devices. Nano Lett. 2015, 15, 4553–4556. [CrossRef] 2. Khan, A.I.; Chatterjee, K.; Wang, B.; Drapcho, S.; You, L.; Serrao, C.; Bakaul, S.R.; Ramesh, R.; Salahuddin, S. Negative capacitance in a ferroelectric capacitor. Nat. Mater. 2015, 14, 182–186. [CrossRef][PubMed] 3. Ko, E.; Lee, J.W.; Shin, C. Negative Capacitance FinFET With Sub-20-mV/decade Subthreshold Slope and Minimal Hysteresis of 0.48 V. IEEE Electron Device Lett. 2017, 38, 418–421. [CrossRef] 4. Ko, E.; Shin, J.; Shin, C. Steep switching devices for low power applications: Negative differential capacitance/resistance field effect transistors. Nano Converg. 2018, 5, 2. [CrossRef][PubMed] 5. Simon, P.; Gogotsi, Y.; Dunn, B. Where do batteries end and supercapacitors begin? Science 2014, 343, 1210–1211. [CrossRef][PubMed] 6. Augustyn, V.; Simon, P.; Dunn, B. Pseudocapacitive oxide materials for high-rate electrochemical energy storage. Energy Environ. Sci. 2014, 7, 1597–1614. [CrossRef] 7. Conway, B.E. Electrochemical Supercapacitors: Scientific Fundamentals and Technological Applications; Springer Science & Business Media: Berlin/Heidelberg, Germany, 2013. 8. Miller, J.R.; Simon, P. Electrochemical capacitors for energy management. Science 2008, 321, 651–652. [CrossRef] Micromachines 2020, 11, 1125 21 of 25

9. Kötz, R.; Carlen, M. Principles and applications of electrochemical capacitors. Electrochim. Acta 2000, 45, 2483–2498. [CrossRef] 10. Liu, J.; Wang, J.; Xu, C.; Jiang, H.; Li, C.; Zhang, L.; Lin, J.; Shen, Z.X. Advanced Energy Storage Devices: Basic Principles, Analytical Methods, and Rational Materials Design. Adv. Sci. (Weinh) 2018, 5, 1700322. [CrossRef] 11. Iro, Z.S. A Brief Review on Electrode Materials for Supercapacitor. Int. J. Electrochem. Sci. 2016, 11, 10628–10643. [CrossRef] 12. Laidler, K.J. The development of the Arrhenius equation. J. Chem. Educ. 1984, 61, 494. [CrossRef] 13. Ma, S.; Jiang, M.; Tao, P.; Song, C.; Wu, J.; Wang, J.; Deng, T.; Shang, W. Temperature effect and thermal impact in lithium-ion batteries: A review. Prog. Nat. Sci. Mater. Int. 2018, 28, 653–666. [CrossRef] 14. Stoller, M.D.; Park, S.; Zhu, Y.; An, J.; Ruoff, R.S. Graphene-based ultracapacitors. Nano Lett. 2008, 8, 3498–3502. [CrossRef] 15. Wu, H.; He, D.; Wang, Y.; Fu, M.; Liu, Z.; Wang, J.; Wang, H. Graphene as the electrode material in supercapacitors. In Proceedings of the 2010 8th International Vacuum Electron Sources Conference and Nanocarbon, Nanjing, China, 14–16 October 2010; pp. 465–466. 16. Hoffmann, M.; Fengler, F.P.G.; Max, B.; Schroeder, U.; Slesazeck, S.; Mikolajick, T. Negative Capacitance for Electrostatic Supercapacitors. Adv. Energy Mater. 2019, 9, 1901154. [CrossRef] 17. Pandolfo, A.G.; Hollenkamp, A.F. Carbon properties and their role in supercapacitors. J. Power Sources 2006, 157, 11–27. [CrossRef] 18. Sharma, P.; Bhatti, T.S. A review on electrochemical double-layer capacitors. Energy Convers. Manag. 2010, 51, 2901–2912. [CrossRef] 19. Christen, T.; Carlen, M.W. Theory of Ragone plots. J. Power Sources 2000, 91, 210–216. [CrossRef] 20. Wang, G.; Zhang, L.; Zhang, J. A review of electrode materials for electrochemical supercapacitors. Chem. Soc. Rev. 2012, 41, 797–828. [CrossRef][PubMed] 21. Ratha, S.; Samantara, A.K. Supercapacitor: Instrumentation, Measurement and Performance Evaluation Techniques; Springer: Berlin/Heidelberg, Germany, 2018. 22. Borenstein, A.; Hanna, O.; Attias, R.; Luski, S.; Brousse, T.; Aurbach, D. Carbon-based composite materials for supercapacitor electrodes: A review. J. Mater. Chem. A 2017, 5, 12653–12672. [CrossRef] 23. Senthilkumar, S.; Selvan, R.K.; Melo, J. Redox additive/active electrolytes: A novel approach to enhance the performance of supercapacitors. J. Mater. Chem. A 2013, 1, 12386–12394. [CrossRef] 24. Miller, E.E.; Hua, Y.; Tezel, F.H. Materials for energy storage: Review of electrode materials and methods of increasing capacitance for supercapacitors. J. Energy Storage 2018, 20, 30–40. [CrossRef] 25. Endo, M.; Takeda, T.; Kim, Y.; Koshiba, K.; Ishii, K. High power electric double layer capacitor (EDLC’s); from operating principle to pore size control in advanced activated carbons. Carbon Lett. 2001, 1, 117–128. 26. Grahame, D.C. The electrical double layer and the theory of electrocapillarity. Chem. Rev. 1947, 41, 441–501. [CrossRef][PubMed] 27. Grahame, D.C. Thermodynamic Properties of the Electrical Double Layer; Amherst College: Amherst, MA, USA, 1954. 28. Helmholtz, H. About electrical interfaces (translated title). Ann. Phys. Chem 1879, 7, 337. [CrossRef] 29. Gouy, M. Sur la constitution de la charge électrique à la surface d’un électrolyte. J. Phys. Theor. Appl. 1910, 9, 457–468. [CrossRef] 30. Gupta, A.; Stone, H.A. Electrical double layers: Effects of asymmetry in electrolyte valence on steric effects, dielectric decrement, and ion–ion correlations. Langmuir 2018, 34, 11971–11985. [CrossRef] 31. Chapman, D.L. LI. A contribution to the theory of electrocapillarity. Lond. Edinb. Dublin Philos. Mag. J. Sci. 1913, 25, 475–481. [CrossRef] 32. Bolt, G. Analysis of the validity of the Gouy-Chapman theory of the electric double layer. J. Colloid Sci. 1955, 10, 206–218. [CrossRef] 33. Outhwaite, C.W.; Bhuiyan, L.B.; Levine, S. Theory of the electric double layer using a modified poisson–boltzman equation. J. Chem. Soc. Faraday Trans. 2 Mol. Chem. Phys. 1980, 76, 1388–1408. [CrossRef] 34. Valleau, J.; Torrie, G. The electrical double layer. III. Modified Gouy Chapman theory with unequal ion − sizes. J. Chem. Phys. 1982, 76, 4623–4630. [CrossRef] 35. Stern, O. The theory of the electrolytic double-layer. Z. Elektrochem. 1924, 30, 1014–1020. Micromachines 2020, 11, 1125 22 of 25

36. Chan, D.Y.; Healy, T.W.; Supasiti, T.; Usui, S. Electrical double layer interactions between dissimilar oxide surfaces with charge regulation and Stern–Grahame layers. J. Colloid Interface Sci. 2006, 296, 150–158. [CrossRef][PubMed] 37. Uvarov, N. Estimation of the surface potential in superionic oxide conductors using the Stern model. Solid State Ion. 2008, 179, 783–787. [CrossRef] 38. Bockris, J.M.; Devanathan, M.; Müller, K. On the structure of charged interfaces. In ; Elsevier: Amsterdam, The Netherlands, 1965; pp. 832–863. 39. Chmiola, J.; Yushin, G.; Gogotsi, Y.; Portet, C.; Simon, P.; Taberna, P.-L. Anomalous increase in carbon capacitance at pore sizes less than 1 nanometer. Science 2006, 313, 1760–1763. [CrossRef] 40. Simon, P.; Burke, A. Nanostructured carbons: Double-layer capacitance and more. Electrochem. Soc. Interface 2008, 17, 38. 41. Barbieri, O.; Hahn, M.; Herzog, A.; Kötz, R. Capacitance limits of high surface area activated carbons for double layer capacitors. Carbon 2005, 43, 1303–1310. [CrossRef] 42. Largeot, C.; Portet, C.; Chmiola, J.; Taberna, P.-L.; Gogotsi, Y.; Simon, P. Relation between the ion size and pore size for an electric double-layer capacitor. J. Am. Chem. Soc. 2008, 130, 2730–2731. [CrossRef][PubMed] 43. Chmiola, J.; Largeot, C.; Taberna, P.L.; Simon, P.; Gogotsi, Y. Desolvation of ions in subnanometer pores and its effect on capacitance and double-layer theory. Angew. Chem. Int. Ed. 2008, 47, 3392–3395. [CrossRef] 44. Feng, G.; Qiao, R.; Huang, J.; Sumpter, B.G.; Meunier, V. Ion distribution in electrified micropores and its role in the anomalous enhancement of capacitance. ACS Nano 2010, 4, 2382–2390. [CrossRef] 45. Huang, J.; Sumpter, B.G.; Meunier, V. A universal model for nanoporous carbon supercapacitors applicable to diverse pore regimes, carbon materials, and electrolytes. Chem. A Eur. J. 2008, 14, 6614–6626. [CrossRef] 46. Huang, J.; Sumpter, B.G.; Meunier, V.Theoretical model for nanoporous carbon supercapacitors. Angew. Chem. Int. Ed. 2008, 47, 520–524. [CrossRef][PubMed] 47. Tang, K.; Chang, J.; Cao, H.; Su, C.; Li, Y.; Zhang, Z.; Zhang, Y. Macropore-and micropore-dominated carbon derived from poly (vinyl alcohol) and polyvinylpyrrolidone for supercapacitor and capacitive deionization. ACS Sustain. Chem. Eng. 2017, 5, 11324–11333. [CrossRef] 48. Liu, H.J.; Wang, J.; Wang, C.X.; Xia, Y.Y. Ordered Hierarchical Mesoporous/Microporous Carbon Derived from Mesoporous Titanium-Carbide/Carbon Composites and its Electrochemical Performance in Supercapacitor. Adv. Energy Mater. 2011, 1, 1101–1108. [CrossRef] 49. Zhang, S.; Pan, N. Supercapacitors performance evaluation. Adv. Energy Mater. 2015, 5, 1401401. [CrossRef] 50. Heimböckel, R.; Hoffmann, F.; Fröba, M. Insights into the influence of the pore size and surface area of activated carbons on the energy storage of electric double layer capacitors with a new potentially universally applicable capacitor model. Phys. Chem. Chem. Phys. 2019, 21, 3122–3133. [CrossRef][PubMed] 51. Shoron, O.F.; Raghavan, S.; Freeze, C.R.; Stemmer, S. BaTiO3/SrTiO3 heterostructures for ferroelectric field effect transistors. Appl. Phys. Lett. 2017, 110, 232902. [CrossRef] 52. Cheng, P.-H.; Yin, Y.-T.; Tsai, I.-N.; Lu, C.-H.; Li, L.-J.; Pan, S.C.; Shieh, J.; Shiojiri, M.; Chen, M.-J. Negative capacitance from the inductance of ferroelectric switching. Commun. Phys. 2019, 2, 32. [CrossRef] 53. Naoi, K.; Simon, P. New materials and new configurations for advanced electrochemical capacitors. J. Electrochem. Soc. (JES) 2008, 17, 34–37. 54. González, A.; Goikolea, E.; Barrena, J.A.; Mysyk, R. Review on supercapacitors: Technologies and materials. Renew. Sustain. Energy Rev. 2016, 58, 1189–1206. [CrossRef] 55. Chen, K.; Xue, D. Materials chemistry toward electrochemical energy storage. J. Mater. Chem. A 2016, 4, 7522–7537. [CrossRef] 56. Brousse, T.; Bélanger, D.; Long, J.W. To Be or Not To Be Pseudocapacitive? J. Electrochem. Soc. 2015, 162, A5185–A5189. [CrossRef] 57. Chang, J.-K.; Tsai, W.-T. Microstructure and pseudocapacitive performance of anodically deposited manganese oxide with various heat-treatments. J. Electrochem. Soc. 2005, 152, A2063–A2068. [CrossRef] 58. Liu, M.; Gan, L.; Xiong, W.; Xu, Z.; Zhu, D.; Chen, L. Development of MnO 2/porous carbon microspheres with a partially graphitic structure for high performance supercapacitor electrodes. J. Mater. Chem. A 2014, 2, 2555–2562. [CrossRef] 59. Wu, Z.-S.; Ren, W.; Wang, D.-W.; Li, F.; Liu, B.; Cheng, H.-M. High-energy MnO2 nanowire/graphene and graphene asymmetric electrochemical capacitors. ACS Nano 2010, 4, 5835–5842. [CrossRef][PubMed] Micromachines 2020, 11, 1125 23 of 25

60. Xiong, C.; Li, T.; Zhao, T.; Dang, A.; Ji, X.; Li, H.; Etesami, M. Three-dimensional graphene/MnO2 nanowalls hybrid for high-efficiency electrochemical supercapacitors. Nano 2018, 13, 1850013. [CrossRef]

61. Zheng, X.; Han, Z.; Yang, W.; Qu, F.; Liu, B.; Wu, X. 3D Co3O4@ MnO2 heterostructures grown on a flexible substrate and their applications in supercapacitor electrodes and photocatalysts. Dalton Trans. 2016, 45, 16850–16858. [CrossRef][PubMed] 62. Qiu, K.; Yan, H.; Zhang, D.; Lu, Y.; Cheng, J.; Lu, M.; Wang, C.; Zhang, Y.; Liu, X.; Luo, Y. Hierarchical 3D

Co3O4@ MnO2 core/shell nanoconch arrays on Ni foam for enhanced electrochemical performance. J. Solid State Electrochem. 2015, 19, 391–401. [CrossRef] 63. Obodo, R.M.; Onah, E.O.; Nsude, H.E.; Agbogu, A.; Nwanya, A.C.; Ahmad, I.; Zhao, T.; Ejikeme, P.M.;

Maaza, M.; Ezema, F.I. Performance Evaluation of Graphene Oxide Based Co3O4@GO, MnO2@ GO and Co3O4@ MnO2@ GO Electrodes for Supercapacitors. Electroanalysis 2020, 32, 2786–2794. [CrossRef] 64. Cheng, Y.; Lu, S.; Zhang, H.; Varanasi, C.V.; Liu, J. Synergistic effects from graphene and carbon nanotubes enable flexible and robust electrodes for high-performance supercapacitors. Nano Lett. 2012, 12, 4206–4211. [CrossRef] 65. Deng, L.; Hao, Z.; Wang, J.; Zhu, G.; Kang, L.; Liu, Z.-H.; Yang, Z.; Wang, Z. Preparation and capacitance

of graphene/multiwall carbon nanotubes/MnO2 hybrid material for high-performance asymmetrical electrochemical capacitor. Electrochim. Acta 2013, 89, 191–198. [CrossRef] 66. Liu, Q.; Nayfeh, O.; Nayfeh, M.H.; Yau, S.-T. Flexible supercapacitor sheets based on hybrid nanocomposite materials. Nano Energy 2013, 2, 133–137. [CrossRef] 67. Lu, X.; Dou, H.; Yang, S.; Hao, L.; Zhang, L.; Shen, L.; Zhang, F.; Zhang, X. Fabrication and electrochemical capacitance of hierarchical graphene/polyaniline/carbon nanotube ternary composite film. Electrochim. Acta 2011, 56, 9224–9232. [CrossRef] 68. Ning, G.; Li, T.; Yan, J.; Xu, C.; Wei, T.; Fan, Z. Three-dimensional hybrid materials of fish scale-like polyaniline nanosheet arrays on graphene oxide and carbon nanotube for high-performance ultracapacitors. Carbon 2013, 54, 241–248. [CrossRef] 69. Zheng, L.; Wang, X.; An, H.; Wang, X.; Yi, L.; Bai, L. The preparation and performance of flocculent polyaniline/carbon nanotubes composite electrode material for supercapacitors. J. Solid State Electrochem. 2011, 15, 675–681. [CrossRef] 70. Lin, H.; Li, L.; Ren, J.; Cai, Z.; Qiu, L.; Yang, Z.; Peng, H. Conducting polymer composite film incorporated with aligned carbon nanotubes for transparent, flexible and efficient supercapacitor. Sci. Rep. 2013, 3, 1353. [CrossRef] 71. Fang, Y.; Liu, J.; Yu, D.J.; Wicksted, J.P.; Kalkan, K.; Topal, C.O.; Flanders, B.N.; Wu, J.; Li, J. Self-supported supercapacitor membranes: Polypyrrole-coated carbon nanotube networks enabled by pulsed electrodeposition. J. Power Sources 2010, 195, 674–679. [CrossRef] 72. Grover, S.; Shekhar, S.; Sharma, R.K.; Singh, G. Multiwalled carbon nanotube supported polypyrrole manganese oxide composite supercapacitor electrode: Role of manganese oxide dispersion in performance evolution. Electrochim. Acta 2014, 116, 137–145. [CrossRef] 73. Zhou, Y.; Qin, Z.-Y.; Li, L.; Zhang, Y.; Wei, Y.-L.; Wang, L.-F.; Zhu, M.-F. Polyaniline/multi-walled carbon nanotube composites with core–shell structures as supercapacitor electrode materials. Electrochim. Acta 2010, 55, 3904–3908. [CrossRef] 74. Giri, S.; Ghosh, D.; Malas, A.; Das, C.K. A facile synthesis of a palladium-doped polyaniline-modified carbon nanotube composites for supercapacitors. J. Electron. Mater. 2013, 42, 2595–2605. [CrossRef] 75. Hu, Y.; Zhao, Y.; Li, Y.; Li, H.; Shao, H.; Qu, L. Defective super-long carbon nanotubes and polypyrrole composite for high-performance supercapacitor electrodes. Electrochim. Acta 2012, 66, 279–286. [CrossRef] 76. Xiong, S.; Yang, F.; Jiang, H.; Ma, J.; Lu, X. Covalently bonded polyaniline/fullerene hybrids with coral-like morphology for high-performance supercapacitor. Electrochim. Acta 2012, 85, 235–242. [CrossRef] 77. Zhang, F.; Ma, H.; Chen, J.; Li, G.-D.; Zhang, Y.; Chen, J.-S. Preparation and gas storage of high surface area microporous carbon derived from biomass source cornstalks. Bioresour. Technol. 2008, 99, 4803–4808. [CrossRef]

78. Deng, D.; Kim, B.-S.; Gopiraman, M.; Kim, I.S. Needle-like MnO2/activated carbon nanocomposites derived from human hair as versatile electrode materials for supercapacitors. RSC Adv. 2015, 5, 81492–81498. [CrossRef] Micromachines 2020, 11, 1125 24 of 25

79. Long, X.; Tian, L.; Wang, J.; Zhang, L.; Chen, Y.; Emin, A.; Wang, X.; Xie, W.; Liu, D.; Fu, Y. Interconnected δ-MnO2 nanosheets anchored on activated carbon cloth as flexible electrode for high-performance aqueous asymmetric supercapacitors. J. Electroanal. Chem. 2020, 877, 114656. [CrossRef] 80. Zhang, H.; Cao, G.; Wang, Z.; Yang, Y.; Shi, Z.; Gu, Z. Growth of manganese oxide nanoflowers on vertically-aligned carbon nanotube arrays for high-rate electrochemical capacitive energy storage. Nano Lett. 2008, 8, 2664–2668. [CrossRef] 81. Huang, M.; Mi, R.; Liu, H.; Li, F.; Zhao, X.L.; Zhang, W.; He, S.X.; Zhang, Y.X. Layered manganese oxides-decorated and nickel foam-supported carbon nanotubes as advanced binder-free supercapacitor electrodes. J. Power Sources 2014, 269, 760–767. [CrossRef] 82. Garakani, M.A.; Abouali, S.; Xu, Z.-L.; Huang, J.; Huang, J.-Q.; Kim, J.-K. Heterogeneous, mesoporous NiCo 2 O 4–MnO 2/graphene foam for asymmetric supercapacitors with ultrahigh specific energies. J. Mater. Chem. A 2017, 5, 3547–3557. [CrossRef] 83. Biswas, S.; Drzal, L.T. Multilayered nanoarchitecture of graphene nanosheets and polypyrrole nanowires for high performance supercapacitor electrodes. Chem. Mater. 2010, 22, 5667–5671. [CrossRef] 84. Lu, X.; Shen, C.; Zhang, Z.; Barrios, E.; Zhai, L. Core–Shell Composite Fibers for High-Performance Flexible Supercapacitor Electrodes. ACS Appl. Mater. Interfaces 2018, 10, 4041–4049. [CrossRef] 85. Liu, Q.; Nayfeh, M.H.; Yau, S.-T. Supercapacitor electrodes based on polyaniline–silicon nanoparticle composite. J. Power Sources 2010, 195, 3956–3959. [CrossRef] 86. Zou, W.-Y.; Wang, W.; He, B.-L.; Sun, M.-L.; Yin, Y.-S. Supercapacitive properties of hybrid films of manganese dioxide and polyaniline based on active carbon in organic electrolyte. J. Power Sources 2010, 195, 7489–7493. [CrossRef] 87. Chen, Y.; Zhang, X.; Xu, C.; Xu, H. The fabrication of asymmetry supercapacitor based on MWCNTs/MnO2/PPy composites. Electrochim. Acta 2019, 309, 424–431. [CrossRef] 88. Yang, M.; Li, J.; Li, H.; Su, L.; Wei, J.; Zhou, Z. Mesoporous slit-structured NiO for high-performance pseudocapacitors. Phys. Chem. Chem. Phys. 2012, 14, 11048–11052. [CrossRef][PubMed]

89. Yuan, C.; Yang, L.; Hou, L.; Shen, L.; Zhang, X.; Lou, X.W.D. Growth of ultrathin mesoporous Co3O4 nanosheet arrays on Ni foam for high-performance electrochemical capacitors. Energy Environ. Sci. 2012, 5, 7883–7887. [CrossRef]

90. Qiu, K.; Lu, M.; Luo, Y.; Du, X. Engineering hierarchical nanotrees with CuCo2O4 trunks and NiO branches for high-performance supercapacitors. J. Mater. Chem. A 2017, 5, 5820–5828. [CrossRef]

91. Pan, Y.; Gao, H.; Zhang, M.; Li, L.; Wang, Z. Facile synthesis of ZnCo2O4 micro-flowers and micro-sheets on Ni foam for pseudocapacitor electrodes. J. Alloys Compd. 2017, 702, 381–387. [CrossRef]

92. Vijayakumar, S.; Lee, S.-H.; Ryu, K.-S. Hierarchical CuCo2O4 nanobelts as a supercapacitor electrode with high areal and specific capacitance. Electrochim. Acta 2015, 182, 979–986. [CrossRef]

93. Abbasi, L.; Arvand, M. Engineering hierarchical ultrathin CuCo2O4 nanosheets array on Ni foam by rapid electrodeposition method toward high-performance binder-free supercapacitors. Appl. Surf. Sci. 2018, 445, 272–280. [CrossRef] 94. Ho, M.; Khiew, P.; Isa, D.; Tan, T.; Chiu, W.; Chia, C.H. A review of metal oxide composite electrode materials for electrochemical capacitors. Nano 2014, 9, 1430002. [CrossRef] 95. Roldan, S.; Granda, M.; Menendez, R.; Santamaría, R.; Blanco, C. Mechanisms of energy storage in carbon-based supercapacitors modified with a quinoid redox-active electrolyte. J. Phys. Chem. C 2011, 115, 17606–17611. [CrossRef] 96. Senthilkumar, S.; Selvan, R.K.; Ponpandian, N.; Melo, J.; Lee, Y. Improved performance of electric double layer capacitor using redox additive (VO 2+/VO 2+) aqueous electrolyte. J. Mater. Chem. A 2013, 1, 7913–7919. [CrossRef] 97. Roldán, S.; Blanco, C.; Granda, M.; Menéndez, R.; Santamaría, R. Towards a further generation of high-energy carbon-based capacitors by using redox-active electrolytes. Angew. Chem. Int. Ed. 2011, 50, 1699–1701. [CrossRef][PubMed] 98. Wang, Z.L.; Song, J. Piezoelectric nanogenerators based on zinc oxide nanowire arrays. Science 2006, 312, 242–246. [CrossRef][PubMed] 99. Li, B.; Xu, C.; Zhang, F.; Zheng, J.; Xu, C. Self-polarized piezoelectric thin films: Preparation, formation mechanism and application. J. Mater. Chem. C 2015, 3, 8926–8931. [CrossRef] Micromachines 2020, 11, 1125 25 of 25

100. Kim, Y.-S.; Xie, Y.; Wen, X.; Wang, S.; Kim, S.J.; Song, H.-K.; Wang, Z.L. Highly porous piezoelectric PVDF membrane as effective lithium ion transfer channels for enhanced self-charging power cell. Nano Energy 2015, 14, 77–86. [CrossRef] 101. Song, R.; Jin, H.; Li, X.; Fei, L.; Zhao, Y.; Huang, H.; Chan, H.L.-W.; Wang, Y.; Chai, Y. A rectification-free piezo-supercapacitor with a polyvinylidene fluoride separator and functionalized carbon cloth electrodes. J. Mater. Chem. A 2015, 3, 14963–14970. [CrossRef] 102. Ramadoss, A.; Saravanakumar, B.; Lee, S.W.; Kim, Y.-S.; Kim, S.J.; Wang, Z.L. Piezoelectric-driven self-charging supercapacitor power cell. ACS Nano 2015, 9, 4337–4345. [CrossRef] 103. Yang, P.; Sun, P.; Mai, W. Electrochromic energy storage devices. Mater. Today 2016, 19, 394–402. [CrossRef] 104. Yang, P.; Sun, P.; Chai, Z.; Huang, L.; Cai, X.; Tan, S.; Song, J.; Mai, W. Large-scale fabrication of pseudocapacitive glass windows that combine electrochromism and energy storage. Angew. Chem. 2014, 126, 12129–12133. [CrossRef] 105. Cai, Z.; Li, L.; Ren, J.; Qiu, L.; Lin, H.; Peng, H. Flexible, weavable and efficient microsupercapacitor wires based on polyaniline composite fibers incorporated with aligned carbon nanotubes. J. Mater. Chem. A 2013, 1, 258–261. [CrossRef] 106. Yu, J.; Lu, W.; Smith, J.P.; Booksh, K.S.; Meng, L.; Huang, Y.; Li, Q.; Byun, J.H.; Oh, Y.; Yan, Y. A High Performance Stretchable Asymmetric Fiber-Shaped Supercapacitor with a Core-Sheath Helical Structure. Adv. Energy Mater. 2017, 7, 1600976. [CrossRef] 107. Voit, W.; Ware, T.; Dasari, R.R.; Smith, P.; Danz, L.; Simon, D.; Barlow, S.; Marder, S.R.; Gall, K. High-strain shape-memory polymers. Adv. Funct. Mater. 2010, 20, 162–171. [CrossRef] 108. Sekitani, T.; Noguchi, Y.; Hata, K.; Fukushima, T.; Aida, T.; Someya, T. A rubberlike stretchable active matrix using elastic conductors. Science 2008, 321, 1468–1472. [CrossRef][PubMed] 109. Li, T.; Fang, X.; Pang, Q.; Huang, W.; Sun, J. Healable and shape editable supercapacitors based on shape memory polyurethanes. J. Mater. Chem. A 2019, 7, 17456–17465. [CrossRef] 110. Ma, Y.-Y.; Yi, G.-B.; Wang, J.-C.; Wang, H.; Luo, H.-S.; Zu, X.-H. Shape-controllable and-tailorable multi-walled carbon nanotube/MnO2/shape-memory polyurethane composite film for supercapacitor. Synth. Met. 2017, 223, 67–72. [CrossRef]

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).