Sustainable Energy Storage: Recent Trends and Developments toward Fully Organic Batteries

Abstract In times of spreading mobile devices, organic batteries represent a promising approach to replace the well‐established lithium‐ion technology to fulfill the growing demand for small, flexible, safe, as well as sustainable energy storage solutions. In the last years, large efforts have been made regarding the investigation and development of batteries that use organic active materials since they feature superior properties compared to metal‐based, in particular lithium‐based, energy‐storage systems in terms of flexibility and safety as well as with regard to resource availability and disposal. This Review compiles an overview over the most recent studies on the topic. It focuses on the different types of applied active materials, covering both known systems that are optimized and novel structures that aim at being established.


Introduction
Nowadays, ah igh and steadily increasing demand for technologies and possibilities for the storage of electrical energy exists not only within the industrial world but also in the developingc ountries. Ap articular, ever-growing interest in small, lightweight, mechanicallyflexible and stable, safe, as well as inexpensive energy storage is presentd ue to quickly emerging mobile devices, smart packaging and clothing, as well as the rising Internet of Things. However,t he current leader in mobile energy storage, the lithium-ion battery,e xhibits severald isadvantagesw ith regard to the stated requirements. In particular, safety issues render such batteries unsuitable for applications that include large mechanical stress, potentially even leading to leakage or breaking of the battery,w hich would be disastrous in case of lithium-basedb atteries. [1] Furthermore, the current lithium-ion technology depends on the provision with large amountso fc riticalr esources. Alreadyt oday,h alf of the global productiono fl ithiumi su sed for batteries, which will definitely increasef urther in the future. Another aspectt hat has to be taken into consideration is that most of the known reserves are located in politically or climatically problematic regions of the world. [2] But currentl ithium-ion batteries do not only contain lithium but also large amountso fn ickel, manganese, and cobalt. The latter exhibitsp articularly large restrictions since over 60 %o ft oday's production and of the known reserves are located in politically instable regions and can barely match the predicted demands for the next 30 years ( Figure 1);n ickel does not perform much more promising. [2] Finally,t he disposal of current lithium-ion batteries and the recycling of the contained materials is still an issue under investiga-tion. [3] Consequently,a lternative active materials have to be found to enables ustainable electrochemical energy storage. [4] Organic batteries, whichu tilize organic or polymeric active materials insteado fm etals or metal oxides, represent the most promising approacht oo vercome the technical and economical restrictions of the established metal-based systems. They do not rely on controversial metal deposits, but the activem aterials can prospectively be synthesized from renewable resources in the future. [5] Furthermore, they provide as uperior processability,e nabling the use of printing techniques (e.g., screen printing, inkjet printing) and variouso ther casting methods (e.g.,d octor blading) as wella sr oll-to-roll manufacturing and allow the construction of mechanicallyf lexible devices. [6] Based on the discoveryo ft he conductivity of conjugated polymers in 1977, [7] organic batteries were firstly developed already in the 1980s [8] and commercialized within af ew years by Bridgestone/Seiko and VARTA/BASF. [9] However,t hose systems were based on poly(pyrrole) and poly(aniline), which did not provide as table working voltage and were consequently taken off the market.N ot before 2002, Nakahara et al.p resented the next step in this research field. They reportedaw orking battery that was based on the 2,2,6,6-tetramethyl-4-piperidinyl-Noxyl (TEMPO) radical and startedanew and much larger wave of new materials and conceptst oward the development of organic batteries. [10] Since then,n umerouso rganic active materials intended for the utilization in batteriesw ere investigated. [11] This Reviewg ives ac omprehensive and critical overview over the systems that were developed recently since our last survey in 2016 [12] with af ocus on all-organic approaches.

Working principle
Batteries are based on the concept of an electrochemical cell, that is, two electrodes madeo fr edox-activem aterials that are placed in an electrolyte, with as eparator (e.g.,asalt bridge, a semipermeable membrane) between them,a nd connected electrically.E lectrons move from the material with the lower redox potential (the negative pole) to the material with the higherr edox potential( the positive pole) until the former is completely oxidized and the latter is completely reduced-the cell is discharged. When an externalf orce, namely an electric currenti nt he opposite direction, is applied, the process is reversed,a nd the cell is charged. The active materials are therefore classified into one of three categories:n -type materials In times of spreadingm obile devices, organic batteries represent ap romising approach to replace the well-established lithium-ion technology to fulfill the growing demand for small, flexible, safe, as well as sustainable energy storages olutions. In the last years,l arge efforts have been made regarding the investigation and development of batteries that use organic active materials since they feature superior properties com-pared to metal-based,i np articularl ithium-based, energy-storage systems in terms of flexibility and safety as well as with regard to resource availability and disposal. This Review compiles an overview overt he most recent studies on the topic.I t focuseso nt he differentt ypes of applied active materials, covering both known systemst hat are optimized and novel structures that aim at being established.
("negative"; low redox potential), p-type materials ("positive"; high redox potential), and b-type materials ("bipolar"; medium or high and low redox potential). [11c] During the discharging and charging process, ions move through the electrodes and the electrolyte to allow chargen eutrality (Scheme 1).
In this Review, an organic battery is ab attery that possesses at least one electrode with an organic redox-activem aterial. If both electrodes are based on organic active materials, it is called an all-organic or fully organic battery (although the electrolyte can contain inorganic ions). However,s ince metal electrodes usually provide established and well-known redox processes, most of the novel organic materials are investigated in metal-organic hybrid cells with one electrode based on a metal (e.g.,l ithium, sodium, potassium, magnesium) and an electrolyte containing respective metal cations.T his allows the execution of basic studies in am ore reliable way.U sually the metal constitutes the negative pole since most applied organic materials feature redox potentials that lie above those of the commonly used metals. Nevertheless, an all-organic battery withoutmetallic active materials is eventually desired.

Performance parameters
To assess the suitability of an active material or aw hole cell setup, several performance parameters are theoretically and experimentally accessible: [14] The cell voltage (V [V]) is the working voltage of ac ell. It is determined from the voltage profile of ac harge/discharge experiment, which shows ap lateau in an ideal case, allowing an easy reading of the voltage.I nr eal cases,h owever,t he voltage can change during the course of chargingo rd ischarging (a "sloping" voltage), which necessitates the determination of an arithmetic average. The voltages measured during discharging and charging usually differ;t heir ratio is stated as the voltage efficiency (h V [%]). The maximum voltage of ac ell is called the theoretical voltage( V theo [V]), which can be calculated from the redox potentials (E)oft he employed active materials according to V theo = E cathode ÀE anode .
The capacity (C [Ah]) describes how much chargec an be stored in ac ell. It depends mostly on the number of redox centers,t he number of transferred electrons per redox center, and the accessibilityo ft he redoxc enters. The maximum, resultingf rom the first two values, is represented by the theoretical capacity (C theo [Ah]). Another important parameter is the specific capacity (C spec [mAh g À1 ]),w hich is the amount of storable charger eferred to the mass of applied activem aterial. It is crucial in particular for mobile applicationsa nd can be determined via C spec = nÁF M (n …n umber of transferred electrons per redox reaction, F …F araday constant, M …m olar mass of the redox-active unit). Usually,t he limiting capacity is stated. For the whole cell, C spec is calculated accordingt oC spec, cell À1 = C spec, anode À1 + C spec, cathode À1 .T he capacities measured during discharginga nd charging often differ from each other due to reversible and irreversible losses. The ratio between discharge and charge capacity states an important value, which is called the coulombic efficiency( h c [%]). The energy (E [Wh]) that can be stored in acell can be calculated from the determined cell voltage and capacity (E = V·C). Accordingly,t he maximum theoretical energy (E theo [Wh]) and the specific energy (E spec [mWh g À1 ]) are calculated using the theoretical and specific capacity,r espectively.T he energy density (E dens [Wh L À1 ]) is determined by the storable energy with respectt ot he volumeo ft he material. The ratio between discharge and chargee nergy is the energy efficiency (h W [%]), which is another important parameter for the evaluation of the cell performance.
The performance of ac ell at different charge/discharge currents is expressed by the rate capability,w hich is normally described by the highest current density that gives ac apacity similart ot he capacities at significantly lower current densities. The currentd ensity is usually given in terms of the C-rate, which is derived from the current (i)t hat is required to charge the cell within 1h(C-rate = i applied i 1h ). In Ta ble 1, the latests tate-of-the-art values for lithium-ion [15] and organic [12] batteries are stated for comparison.

Electrodecomponents
Since organic active materials generally do not possess sufficient electrical conductivity,c onductive agentsh avet ob e added during electrode preparation to ensure charget ransport between thec urrent collector andt he redox-actives pecies during the discharging and charging process. Beside ah igh electronic conductivity,t he conductive agent has also to provide al arge surfacea rea to ensure an intimatec ontact to the active material. Furthermore, since ion transport into and out of the electrode is crucial for unhindered discharging and chargingi nt erms of charge neutrality,t he conductivea dditive must offer ak ind of porousa nd flexible network to allow penetration of the electrolyte and ion migration as well as compensation of volumec hanges, which occur due to ion insertion/release. The used conductive additives are usually based on nanostructured carbonmaterials, for example, carbon nanoparticles, [16] mesoporousc arbon, [17] vapor-grown carbon fibers, [18] graphene, [19] and carbon nanotubes. [20] For an optimized interaction, the active materialh as to be thoroughly mixed with the conductive additive to form ac omposite electrode. Alternatively,s everal systemsw ere recently developed that include am ore intimate conjunction betweena dditive and active material, for example, throughc hemical bonds or in situ co-preparation (cf. Section 3).
Beside the active material and the conductive agent, composite electrodes usuallyc ontain binder materials. Binders ensure both the thorough blendingo ft he active material andt he conductive additive as well as as ufficient mechanical stability of the composite electrode. [21] The most common bindersa re fluorinated polymers, such as poly(tetrafluoroethylene) (PTFE) and poly(vinylidene difluoride) (PVdF), which provide high electrochemical stability,b inding capability,a nd electrolyte ab-sorptiona bility. [22] Carboxymethyl cellulose (CMC) is usually used if the composite is processed in an aqueous medium. Furthermore, CMC metal salts, poly(acrylate)s, or poly(3,4-ethylenedioxythiophene)poly(styrenesulfonate) (PEDOT:PSS)o ffer additional ions or semiconductivity,t hus enhancing the performance of the composite electrode. [23] 3. Materials

Quinones
Quinones represent the most popular group of organic active materials for electrochemical energy storage. [24] They offer a stable and reversible redox chemistry,awider ange of electrochemicalp otentials, and af acile synthetic access. [25] The electrochemical charges toragei sb ased on the transition between the reduced hydroquinonea nd the oxidized quinone form (Scheme 2). Usually,q uinones undergo two redox processes that are close in potential, but they can melt into one apparent two-electron process when suitable functional groupsa nd/or electrolyte systems are chosen. Thus, redox systemsa re possible that can store two electrons per molecular unit but possess as table working voltage. During the redox transition, the quinone accepts( or releases) cations-in most cases either protons (in aqueous systems) or alkali metal ions (e.g.,L i + ,N a + , K + ). Thus, quinones are often considered as organic metal-insertionm aterials, as alternatives to inorganic insertion systems, such as LiFePO 4 .
Benzoquinone is the archetypical quinone,o fferingt he simplest possible quinoid structure and, due to its comparably low mass, ah igh theoretical specificc apacity of 496 mAh g À1 (for two electrons). Thus, several approaches have been undertaken to functionalize the basic molecule to optimize its properties with regard to battery application. Yaoa nd co-workers studied ab enzoquinone with four sodium phenoxide groups in ah ybrid sodium cell, which showedasloping voltage (2.4 to 1.8 V) and 150 mAh g À1 but only lost 35 %o fi ts capacity Table 1. Latest benchmark valuesf or lithium-ion and organic batteries. [12,15] Specific capacities and energies ares tated for anodea nd cathode as acombined system. Reviews over 400 cycles. [26] Similarly,ad imethoxybenzoquinone was synthesized and tested in am agnesium-ion cell,s howinga stable voltage and ac apacity of 120 mAh g À1 but losing 80 % of capacity over only 30 cycles due to dissolution of the active material. [27] Furthermore, am ore elaborate 2,3-dichloro-5-hydroxy-6-cyano-1,4-benzoquinonew as tested against lithium but resulted in av oltage that ranged from 2t o0 .5 V. [28] Matsubara et al. investigatedaseries of nine bisbenzoquinones with different halide and alkyl substituents. [29] They achieved stable voltages of ca. 2.8 Vi nh ybrid lithium cells with good capacities duringt he first cycles,b ut they suffered from substantial capacityl osses already during the first 20 cycles. Sodium rhodizonate showed af our-step voltage in as odium-ion cell but a good rate capability with the capacity at 50 Cr esembling7 0% of the 1C-value. [30] However,t he cell lost 20 %o fi ts initial capacity over 100 cycles. The capacity loss could be reduced to 5% by preparing nanorods of the active material. [31] Park and co-workersp resented at riptycene-like trisbenzoquinonew ith a two-stepv oltage and ah igh initial capacity of 400 mAh g À1 in ah ybrid lithium cell. [32] Nevertheless, the cell lost 50 %o ver the first 20 cycles. Besides the small molecules, polymers containing benzoquinone units were investigated, in particular with regard to al ow solubility of the active material in the electrolyte. However, they suffered from sloping voltages as well as low specific capacities. [33] Naphthoquinones are less widespread compared to their smaller counterparts. Park et al. presented ah ybrid lithium cell with 2-hydroxy-1,4-naphthoquinone, called Lawsone, as active material, whichf eatured as loping voltage but ag ood specific capacity, which turned out to be stable over 1000 cycles ( Figure 2). [34] Naphtharazin, 5,8-dihydroxy-1,4-naphthoquinone, and its chlorinated derivatives were likewise tested. [35] Their lithium-ion cellsr evealed high initial capacities of around2 50 to 300 mAh g À1 but suffered from several voltage plateaus during the discharging as well as significantc apacity fading during the first cycles. In contrast, a2 ,3-diamino-functionalized naphthoquinone showedastable discharge voltage as well as ah igh initial capacity ands tabilityo ver 500 cycles. [36] With regard to ad ecreased solubility,ap olymerizable naphthoquinone was preparedv ia coupling to an azide-bearing styrene, forminganaphthotriazolequinone, which wass ubsequently polymerized and used as active materiali nalithium-ion cell. [37] The anthraquinone motif in general offers high stabilities regarding both voltage and capacity and was extensively studied with regard to organic active battery materials in the past. Nevertheless, several derivatives were presented recently.Z hao et al. provedt he applicability of at wofold sulfonated anthraquinone in ah ybrid potassium cell. [38] Emodin, ah ydroxyl-and methyl-functionalized species derived from plant extracts,a llowed the construction of ac ell with as table voltage and a good initial capacity but showedacapacity loss of 70 %o ver 80 cycles. [39] Using quinizarin (1,4-dihydroxyanthraquinone)i n combination with 2-[1H-indol-2-yl(1H-indol-3-yl)methyl]phenol, ac ell comprising two organic active materials (but still a Li + -based electrolyte) was built but revealed av oltage curve that featured no distinct plateaus. [40] Takeda et al. demonstrated an anthraquinone derivative in which carbon atoms in the aromatic rings werer eplaced by nitrogen atoms. [41] Ar espective lithium-ion cell showedas loping voltage and an initial capacity of 250 mAh g À1 ,w hich decreased by 40 %w ithin 30 cycles.T he same contribution presented ac ell based on tetracyano-9,10-anthraquinonedimethane (TCAQ), which was already introduced in 2014 as ar edox-activeg roup in ap olymeric active material. [42] The cell incorporating the small TCAQ molecule featured as table voltage and ag ood initial capacity, which, however,decreasedt o40% over 30 cycles. [41] In another work, by Schubert and co-workers, the polymeric TCAQ was used to build af ully organic cell with at hianthrene-polymerbased counter electrode, which featured ar elativelyh igh voltage of 1.4 V, ag ood capacity,a nd ar easonable capacity fade of 30 %o ver 250 cycles ( Figure 3). [43] Furthermore, ap olymer based on 9,10-di(1,3-dithiol-2-ylidene)-9,10-dihydroanthracene (exTTF) was used in ah ybrid zinc cell, resulting in av oltage of ca. 1.2 Vand acapacity loss of only 5% over 1000 cycles. [44] Further quinoid moieties are less common, but af ew examples do exist. Diamino-and dinitrophenanthrenequinones were used in hybridl ithium cells, which revealed unstabled ischargev oltages and fast capacity fading. [45] Likewise, al ithiumion cell comprising5 ,7,12,14-tetraaza-6,13-pentacenequinone as active material showed several discharge voltages and ac apacity drop of 80 %o ver 250 cycles. [46] Liang et al. presented a cell that used pyrene-4,5,9,10-tetraone and PbO 2 as redoxactive electrode materials, whichf eatured ah igh specific capacity of 350 mAh g À1 ,acapacity loss of only 5% over 1500 cycles, and an excellentr ate capability ( Figure 4). [47] While the abovementioned polymers contain the active quinone moietya ss ide groups attached to the backbone,L iao and co-worker presented polymers with ab ackbonef ormed by anthraquinone units, which are linked either via the 1-and 4-or 2-and 6-carbon atoms and were utilized as active material in hybrid magnesium cells. [48] Althoughb oth cells are comparable with regard to capacitya nd cycling stability, the former shows as table discharge voltage, in contrast to as loping voltage in the latter system. Ap olymer containing pyrene-4,5,9,10-tetraonei ni ts backbone resulted in as loping cell voltage andacapacity loss of 30 %over 100 cycles. Notably,the introduction of an ethynyl moiety into the backbone led to asignificantly improved rate capability. [49] Two-dimensional covalent organic frameworks (COFs) containing redox-active quinone moieties were presented by Wu et al., [50b] AbruÇae tal., [50c] and Wang et al. [50a] Applyinge xfoliated few-layer nanosheets of the COFs in combination with al ithium counter electrode, ac ell featuring ad ischarge voltage of 2.8 Va nd ac apacity that is stable over 1000 cycles was demonstrated ( Figure 5). [50a] 3.1.1. Quinone-sulfurp olymers Poly(quinonyl sulfide)s represent another populargroup of quinone-based redox-activep olymers. As for most polymerics ys-tems used as activeb attery materials, they combine al ow solubility with almostu naltered electrochemical characteristics. In the case of sulfide polymers,t he sulfur linkersp rovide, depending on their length, potentially additional redox activity. While this can increaset he provided capacity,s uch systems often suffer from sloping voltages or decreased stabilities. [51] Yaoa nd co-workers presented ac ell based on poly(anthraquinonyl sulfide) (PAQS) and Ni(OH) 2 as active electrode materials and achieved as table dischargev oltage of 1.0 Vand ac apacity of 190 mAh g À1 ,w hichl ost only 10 %o ver 1300 cycles. Furthermore,t he obtained cell showedan oticeable temperature stability ( Figure 6). [47] As ulfide polymer derived from sodiumphenoxide-functionalized benzoquinoner evealed ag ood performance in ah ybrid sodiumc ell losing only 15 %c apacity over 500 cycles. [52] 3.1.2. Quinones linked to conductive substrates As already stated, electrodes made of organic active materials usuallyc ontain significant amountso fc onductive additive, mostlyc arbon materials, to ensure charge transport between the active material and the current collector.C onsequently, the interaction betweent he organic compound and the carbon additive is crucial for the performance of the cell. Thus, several  groups aim at the improvement of those interactions, for example,through the introductiono fcovalentb onds or exploitation of p-p interactions. Notably,t he former approach is rather rare compared to the latter.N evertheless, Campidelli et al. prepared multi-walled carbon nanotubes (MWCNTs) with anthraquinone moieties that were covalently attached via reductive diazonium coupling (Scheme 3). Organic electrodes without conductive additives or bindersw ere prepared and investigated in ah ybrid lithium cell, resulting in as loping voltage (2.2 to 1.7 V) and ac apacity of 100 mAh g À1 that was stable at least for 50 cycles. [53] The application of ar educed graphene oxide (RGO)w ith anthraquinone attached via azide coupling in al ithium-ion cell resulted in ad ischarge curve without ad istinct voltage plateau. [54] Non-covalentc oupling of quinones and conductive carbon materials relies on the p-p interactions between the aromatic systemso ft he quinones and the graphene or carbon nanotubes. Those systemsare usually prepared either by in situ syn-  thesis of the redox-active compounds in ac arbon-containing suspension or by co-dispersion of both materials. Several demonstrated cells suffered from aw orking voltage that changed by more than 1V during discharging, [55] most likely due to a significant capacitive contribution of the large-surface graphene materials in those composites or because of ap oor cycling stability. [56] In other cases, the intimatel inkage of active materiala nd conductive carbon led to an increased stability [57] or rate capability. [58] Nevertheless,f or all of the abovementioned cells, additional conductive carbon additive wasu sed for the preparation of the electrodes. In contrast, Wei et al. prepared poly(2,5-dihydroxyl-1,4-benzoquinonyl sulfide) on singlewalled carbon nanotubes (SWCNTs), which wass ubsequently used as electrode materialw ithout furtherc onductive additive or binder.T he resulting hybrid lithium cell showed ac apacity of 120 mAh g À1 with only 15 %l oss over 500 cycles. [59] Beside carbon, other conductive materials were linked to the quinone active materials as well. Chen and co-workers encapsulatedb enzoquinone into TiO 2 spherest od ecrease the solubility while retaining electrical conductivity. [60] The resulting lithium-ion cell showed ag ood initial capacity of 300 mAh g À1 but lost around2 0% duringt he first 100 cycles. Furthermore, severals ystems were proposed based on ac onductive polymer with quinone side groups. However,m ost reports did not present aw orking cell. [61] In contrast, Sjçdin et al. showed an all-organic cell, which used poly(EDOT-benzoquinone) and poly(EDOT-anthraquinone) as active electrode materials as well as am etal-free, pyridine-based electrolyte (Figure7). However, the cell featured an unstable voltagea nd capacity. [62] In summary,q uinones still represent the mostp opulara nd promisingc lass of organica ctive electrode materials in terms of specific capacity and stability,i np articular when they are integrated in polymeric compounds. Thus, several all-organic systems have been presented that are based partly or solely on quinone derivatives.

Diimidesand dianhydrides
Aromatic diimides offer ab road range of redoxp otentials, stable electrochemical processes, and at wo-electron-storage capability per molecule. [63] In addition, they tend to stack, leading to al ow solubility in many electrolytes, thus improving the long-term stability.S imilart oq uinones,t he redox reactions are based on the transformation between two carbonyl and two hydroxyl groups, with the radical transition state being stabilized through the aromatic system, and the diimides can be appliedaso rganic metal-insertion materials (Scheme 4).
Phenylenediimide is the simplest aromatic diimide,a nd, althoughi ti sl ess electrochemically stable compared to its polycyclic counterparts according to theory, [64] several derivatives were prepared for an application in organic batteries. Kothandaraman and co-workers tested an acetate-functionalized phenylenediimide in al ithium-as well as as odium-ion cell. [64,65] Althoughb oth cells showed as table voltage plateau and comparable capacities, the latter revealed as ignificantly lower capacity retention. Rings composed of phenylenediimidea nd cyclohexylm oieties werel ikewise studied as active materials.
Here, the performance of the cell depends strongly on the extento ft he formed rings. [66] In comparison to phenylenediimide, naphthalenediimide (NDI) possessesalarger aromatic system and al ower ring strain within the imide moieties. Thus, NDIs are usually more stable and more often used as active battery material. For example, cells with the NDI analogs of the abovementioned acetate-containing diimides featured significantly improved long-term stability. [64,65] In contrast, al ikewise simple NDI, N,N'-diphenyl-NDI, showed as table discharge voltage in ah ybrid lithium-ion cell but ac apacity loss of 25 %o ver only 100 cycles. [67] Perylene diimidefeatures an even larger aromatic system, offering ap otentially higher stabilization of charged states and was used by Cao and co-workers in a hybrid lithium-ion cell,w hich showedastable voltage and a specific capacity of 110mAh g À1 (80 %o ft he theoretical value) that remained stable over 200 cycles. [68] An acetate-functionalized derivativel ed to al ower capacity (which is expectedd ue to the higherm ass), which was stable over 100 cycles. [69] To enhance the long-term stability,a lso polymers containing NDI redox-activem oieties were thoroughly investigated. Dominko and co-workerst ested as imple polymer of nitrogen-linked NDI units in ah ybrid magnesium-ionc ell. Notably,t he rather low initial capacity of 30 mAh g À1 increased over 100 cycles to 75 mAh g À1 ,w hichw as explained by as welling of the electrode. [70] Ap hthalein and at ri(ethylene glycol)b ridge were introducedi nto an NDI polymer by Mecerreyes and co-workers. [71] The resulting polymers were appliedi nl ithium-ion cells and showed capacities of 120 and 80 mAh g À1 ,r espectively (corresponding to 82 and 75 %o ft he theoretical capacities), which remained stable over at least 100 cycles. Xu et al. incorporated as ulfonyl moiety as linking unit and, after the electrolyte was optimized, achieved af lat voltage plateau as wella sa specific capacity of 120 mAh g À1 that remained stable over 400 cycles after aperiod of equilibration. [72] Due to an additional redox-active unit, the applicationo facarbonyl linker led to an even higherc apacity of 160 mAh g À1 ,w hich was stable over at least 50 cycles. [73] Dong et al. used an ethylene-linkedN DI polymer to build an all-organic cell with ap olytriphenylamine counter electrode and ah ighly concentrated (21 m)a queous lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) electro-lyte. [74] The resulting workingv oltage strongly decreased during the discharge, but the capacity loss was only 15 %o ver 700 cycles (Figure8,r ight).L ikewise, Lu et al. built af ully or-ganic cell from an ethylene-linked perylened iimide polymer and polytriphenylamineu sing Mg(ClO 4 ) 2 in the electrolyte. [75] Again, the discharge voltage was slightly sloping with an aver-  [62] Scheme4.To p: Molecular structures of commonaromatic diimides with theoretical specific capacities. Bottom:General redox mechanism of aromatic diimides. age of 1.0 V, but the capacity of 80 mAh g À1 remained stable over remarkable 2000 cycles. In addition, the cell showed a good rate capability (Figure 9). Another all-organic cell was presentedb yP icard and co-workers, based on hexyl-linked perylened iimidea nd perylene tetracarboxylate. [76] The cell featured as lightly slopingv oltage with an average of 1.0 Va nd an initial capacityo f80mAh g À1 ,w hich decreased only by 20 % over 200 cycles.
Aiming at an improved lithium intercalation,t wo-dimensional polymericn etworks were prepared from perylene diimide and triptycene or tri-b-ketoenamine moieties. Due to the increased mass per redox-activeu nit, the specific capacities are low,b ut, in particular for the former system, the capacities are rather stable. [77] Besides nitrogen-connectedp olymer chains, also alternative configurations were studied. On the one hand, conjugatedp olymers from aromatic diimides linked via the carbocycles through phenylene or ethylene bridges were investigated but suffered from decreasing specific capacities. [78] On the other hand, NDIs were introduced as side groupsi np olynorbornenes by Nishidea nd co-workers. [79] The resulting hybrid lithium-ion cells showeds table voltages and capacities of 100 mAh g À1 ,w hich even remained stable over 500 cycles for ap henyl-functionalized NDI ( Figure 10).

Mixed diimide-quinone polymers
With the goal of achieving ah ighers pecific capacity, several polymers containing an anthraquinone bridge linking the aromatic diimide moieties werep repared and tested in hybrid lithium-ion cells. [80] Sincet he resulting polymers are conjugated,t hey all showed aw orkingv oltage that decreased by ca. 0.5 Vd uring discharge. According to the work of Xu et al.,n os ignificant differences between the 1,4and 1,5-linking anthraquinone are observable. [80b] For an NDIanthraquinone co-polymer,ah igh specific capacity of up to 190 mAh g À1 was observed, which corresponds to 80 %o ft he theoretical value assuming four transferred electrons. [

Aromatic diimides linked to conductive carbons pecies
Despite their conjugatedn ature,d iimide-based active materials, as most of the other organic active materials, are usually combinedw ith electrically conductive materials, namely carbon materials, in organic battery electrodes to ensure sufficient charget ransport. [12] Aiming at af urther enhancement of the charget ransfer between the active material and the conductive carbon, p-p or ionic bonds between the diimides and graphene or carbon nanotubes (CNTs) are used to form composite compounds. Thus, alkylammonium-functionalized naphthalenea nd perylened iimides were combined with RGO via co-dispersion,e xploiting ionic interactionsb etween the cationic alkylammonium and anionic graphene oxide groups. Although the former showed as trongly decreasing voltage over the course of the discharge in ah ybrid lithium-ion cell, [81] the perylenespeciesresulted in astable voltage and acapacity of 75 mAh g À1 ,w hichd ecreased by only 25 %o ver 500 cycles. [82] Furthermore, composites of polymerica romatic diimides and carbon materials were prepared, usually via in situ polymerization in ac arbon-containing dispersion. Similarly,p oly(aromatic diimide)s on graphene and CNTsw ere synthesized. [80d, 83] While mosto ft he obtained hybrid cells showed sloping workingv oltages, high stabilities [83b,d] and improved rate capabilities [80d, 83d] were achievedd ue to the linkage to the conductive carbon compounds (Figure 11). Besides diimides, perylened ianhydrides were likewise studied in composites with conductive carbonm aterials for hybrid lithium-ion cells. [84] Here, the combination with RGO revealed the best performance with as  Thus, aromatic imides are still among the most popularm aterials for organic battery electrodes. They allow the construction of cells that feature very stable capacities over several hundredso re ven thousands of cycles. [75, 79a, 83d] Furthermore, they were successfully appliedi na ll-organic cells, in particular in combination with polytriphenylamines. [74][75][76] The interesting approachofcombining diimides with quinone units in co-polymers showed, however,n os ignificant improvement. In contrast, intimate linkaget oc onductive carbon species (mostly graphene and CNTs) enhanced both the long-term cycling stability as well as the rate capabilityo ft he applied materials. [80d, 83d] 3.3. Other carbonyl compounds

Terephthalatesa nd other aromatic carboxylates
The charge-storage capability of terephthalates ando ther aromatic carboxylates is based on the reversible reduction of the carbonyl moieties of the carboxylate groups (Scheme 5). Like other carbonyl compounds, the terephthalates act as metal-insertion compounds when they are used with am etal-based electrolyte. [63] Several functionalized terephthalates had been synthesized and wereu sed in hybrid lithium-ion cells. Most of them revealed significant capacity drops or showed no stable voltage at all. [85] Nevertheless, a2 ,5-dimethyl-terephthalate enabled a stable discharge voltage of 0.8 Va gainstl ithium and as table specific capacity of 160 mAh g À1 over at least 50 cycles, [86] whereas the application of a2 ,5-bis(phenylamine) terephthalate led to av oltage of 3.2 Vb ut only am oderate capacity of 70 mAh g À1 . [87] Instead of lithium or other alkaline-metal counter cations, Wang et al. used terephthalates with alkaline-earthm etal ions. [88] Since they form different crystal structures with different ion pathways than their alkaline metal analogs,a better lithium-ion transport through the electrode was anticipated. The studied hybrid lithium-ion cells showed very high initial capacities of over 300 mAh g À1 ,w hich dropped to 170, 140, and 100 mAh g À1 during the first 10 cycles for calcium, barium, and strontiumt erephthalate, respectively (corresponding to 65, 80, and 50 %o ft he theoretical capacity), re-maining constanto ver the next 40 cycles. Az inc terephthalate showedasimilarb ehavior and reached as table capacity of 180 mAh g À1 (75 %o ft heory) over 80 cycles after an equilibrationo fc a. 20 cycles in ah ybrid lithium-ion cell. [89] Notably, the amorphous form revealed am uch better performance compared to the crystalline ones. Poizota nd co-workersu tilized a2 ,5-dihydroxyterephthalate with one magnesium and two lithium cations per molecule in al ithium-organic cell, which featured am oderate capacity of 80 mAh g À1 ,r emaining stable over at least 80 cycles. [90] Subsequently,t he authors applied the material in an all-organic, Li + -based, and symmetric cell, exploiting the carboxylate-as well as the quinone-based redox processes of the compound (Figure 12). The observed capacityw as 80 mAh g À1 andd ecreased by only 20 %o ver 300 cycles.
Polymers containing redox-active terephthalate moieties were also tested as electrode materials, notably also poly(ethylene terephthalate) from recycled bottles, [91] but the resulting cells mostly suffered from vastly decreasing discharge voltages. [33b, 91, 92]

Terephthalatesl inked to conductive substrates
As most organic materials that are used for electrochemical energy storage, terephthalates do not provide electrical conductivity that is high enought oe nsure sufficient charge transport during charginga nd discharging. Thus, ac onductive additive has to be added. Aiminga ta no ptimum interaction, several composites in which the redox-activea nd conductive material are closely connected were studied. The most common systemsa re terephthalates combined with graphene. Ac omposite of sodium terephthalate and graphene, achieved via co-dispersion, resulted in ah ybrid sodiumc ell with aw orking voltage of only 0.3 Vb ut ag ood initial capacity of 220 mAh g À1 ,w hich decreased by only 20 %o ver 500 cycles. [93] In contrast, ac omparable system based on potassium resulted in as loping voltage of 1.2 to 0.8 Va nd an initial capacity of only 120 mAh g À1 . [94] In am ixed approach, which used potassium terephthalate on graphene in as odium-ion cell, also ad ecreasingd ischarge voltage andaspecific capacity of 150 mAh g À1 but an improved rate capability compared to the graphene-free cell were achieved. [95] An alternative to the utilization of carbon as conductor is represented by the application of silver particles. Li and co-worker prepared ac omposite of calcium terephthalate ands ilver via dispersion of the organic compound in as ilver nitrate solution with subsequent drying and reduction of the silver ions. [96] The built hybrid lithium-ion test cell revealed as table voltage of 1.0 Vand ac apacity of 90 mAh g À1 .N otably,t he silver content determined the rate capability,w ith an optimum at 5wt% (Figure 13). The same group demonstrated ac ell where the silver is formed in situ through the utilization of silver terephthalate. Within in the first cell cycle, the silver cations werei rreversibly reduced and formed uniformly dispersed silver nanoparticles. [97] With its low molar mass accompanied by ap otential twoelectron storage ability,t erephthalates offer the possibility of high specific capacities. Indeed, many systems with high initial capacities were demonstrated, but most of them suffered from significant losses during the first 100 cycles. Furthermore, compared to quinones and diimides, relatively lowv oltages were observedf or respective hybrid metal cells. As for quinones and aromatic diimides, composites that possess ac lose connection between active and conductive materials werep resented, which showed good stabilities andr ate capabilities. However, they are clearly inferiortothe bestcomparable quinone-or diimide-based systems.

Stable organic radicals
Stable organic radicals are characterized by the presenceo fa n unpaired electron within the molecule. They are stabilized through the introduction of sterically demanding substituents and electron resonance, thus impeding problemst hat are related to radical compounds, such as dimerizationt hrough the formation of bonds between radicals. [11d, 107] In charge-storage applications, organic radicals usually store one chargep er redox-active unit, leadingt oanon-radical, thus more stable speciesi nt he charged state. Additionally,t he redox processes of organic radicals are based only on the transfer of as ingle  [90] ChemSusChem 2019, 12,4093 -4115 www.chemsuschem.org 2019 The Authors. Publishedb yWiley-VCH Verlag GmbH &Co. KGaA, Weinheim electron withoutl arger structural changes of the molecule, which resultsi ni mproved electron-transfer rates. [13] On the other hand, counterions have to migrate during redox processes to maintain electroneutrality,whichaffects the related redox kinetics.T his can occur either via ion expulsion or uptake, depending on the ion types present and ion concentrations as well as the appliedc harge-transfer rate, as recently shown for polymers containing TEMPO radicals. [108] Nevertheless, high charginga nd discharging currents are possible leading to high power densities. The most common organic radical motifs are the nitroxyl and the phenoxyl radicals( Scheme 6).

TEMPO-based materials
The TEMPOr adical is the by far most popularo rganic radical and has been used in batteries since 2002. [10] Thus, it is well known and most of the studies that were recently conducted focusedo nt he variation and optimizationo fe stablished systems. To enhance charge-transfer rates through enhanced electronic conductivities, TEMPO-containing polymers were combined with conductive carbon materials (graphite,g raphene, CNTs, etc.) either by in situ polymerization on the material, [109] by co-dispersion, [110] by encapsulation, [111] or by grafting-on approaches. [112] The resulting materials, on the one hand, often showed as loping or multistep voltage, [109][110][111] most likely due to an additional, capacitive contribution by the high-surfacearea carbonm aterial. On the other hand, as ignificantly improvedr ate capability comparedt os imple composites of active materiala nd conductive additive wasachieved [111,112] and also aw orkingc ell without additives was demonstrated. [109] Nishidea nd co-workers used SWCNTst hat were prepared by enhanced direct-injection pyrolytic synthesis (eDIPS), which showed ah igh crystallinity and enabledt he preparation of an electrode based on poly(2,2,6,6-tetramethylpiperidinyloxy-4-yl Figure 13. Top: SEM images of calciumt erephthalate silverc omposites, showing adecreasing particle size and increasingp article uniformity with increasing silver precursor content. Bottom:Performance characteristics of respective hybrid lithium cells, in particular the charge/discharge capacity over 130 consecutive cycles and voltage profiles at different currents and silver precursor contents, demonstrating an optimum of the latter at 5wt%.( Reprinted with permission from the Royal SocietyofC hemistry,Copyright 2016). [96] ChemSusChem 2019, 12,4093 -4115 www.chemsuschem.org 2019 The Authors. Publishedb yWiley-VCH Verlag GmbH &Co. KGaA, Weinheim acrylamide) (PTAm) with less than 5% CNTs. [113] The resulting electrode was appliedi na na ll-organic cell with poly(anthraquinone-substituted ethylene imine) (PAQE) as counter electrode, using an aqueous sodiumc hloride electrolyte. The cell delivered as table discharge voltage of 1.0 V, ar emarkable rate capability,a nd an initial specific capacity of 80 mAh g À1 ,w hich retained7 0% after 1000 cycles( Figure 15). Alternatively, TEMPO units were coupled to ac onjugated polymer,n amely poly(dithieno[3,2-b:2',3'-d]pyrrole, for enhanced charge transfer. [114] Electrodes without additives were prepared via electropolymerization, leadingo nly to as mallm ass loading of 0.1 mg cm À2 .T he resulting hybrid lithium-ion cell revealed a decreasing dischargev oltage around3 .3 Va nd an initial capacity of 60 mAhg À1 . Alternatively, systemsw ith TEMPO units linked to conjugated polymer chains were studied in lithium-ion cells, but while ap olyaniline-based materialr evealed as trongly sloping dischargev oltage (between 3.7 and 2.8 V) and unstablec apacities, [115] aPEDOT-containing polymer led to avery low initial capacity of 35 mAh g À1 . [116] With regard to an improved ionic conductivity,a ni midazolium-containing polymerici onic liquid bearing TEMPO groups was synthesized, but at est cell did not  [76] Scheme6.To p: Molecular structures of organic radical moieties and their preferential redox reactions.Bottom:Organic radicals recently used in organic cells. Reviews show any distinct voltage plateau. [117] For the same purpose, a COF with TEMPO units was prepared. [50a] However,t he resulting lithium-ion cell revealed at wo-step discharge voltage and, comparedt oi ts quinone counterparts( cf. Section 3.1), al ow specific capacity. [50a] In contrast, Iwasa et al. presented ab endable hybrid lithium-ion cell using ag el of ac ross-linked TEMPO polymer. [118] The utilization of the gel electrode led to as table voltage of 3.5 V, ac apacity of 105 mAh g À1 ,w hich remained nearly stable over 500 cycles, as well as an excellent rate capability ( Figure 16).
Beside the conductive additive, organic electrodes usually requireapolymeric binder form echanical stability andt horough contacta mong the components.Z hang and co-workers developed aT EMPO-containing polymer based on poly(ethylene maleic anhydride), which they applied in ab inder-free electrode. [119] The relatedh ybrid lithium-ion cell showed a Reviews stable voltage of 3.6 Vand ac apacity of 80 mAh g À1 ,w hich decreasedb yonly 10 %o ver remarkable 2000 cycles.

Other stable organic radicals
Most other organic radicals that weres tudied for battery applications are, like the TEMPO itself, nitroxyl radicals. Ad idehydro-PROXYL( 2,2,5,5-tetramethyl-1-pyrrolidinyloxy) radicalw as introduced in at riphenylamine polymer. [120] Theo btained hybrid lithium-ion cell revealed ap rominent voltage plateau at 3.7 Vb ut also as econd one at 2.7V ,w hich was assigned to the polytriphenylamine. The overall capacity was 120 mAh g À1 , losing 10 %o ver 100 cycles. Subsequently,t he same group attempted to enhancet he electrochemical performance by the introduction of MWCNTs,b ut no significant improvementw as achieved. [121] Ap olymer of the 1,1,3,3-tetramethylisoindolin-2yloxyl radical showedacomparablep erformance in al ithiumion cell, [122] whereas ap olyphosphazene bearing an N-tertbutyl-N-oxylamino phenylr adicalu nit resulted in ah igher initial capacity of 145 mAh g À1 but ac apacity drop of 35 %d uring the first 50 cycles. [123] Aphenoxyl radicalwas used in apolynorbornenei nf orm of tetramethylphenoxyl units, but ab uilt hybrid lithium-ion cell showed ac apacity of only 60 mAh g À1 , losing 20 %c apacity over the first 100 cycles. [124] In conclusion, no new organic radicals that represent promising candidates for organic batteries were revealed recently. Nevertheless, systemsb ased on already established motifs, namely on TEMPOa nd didehydro-PROXYL, were further optimized,i np articular with regard to higher electrical and ionic conductivities.

Conjugated polymers
The first organic compounds that were used for electrochemical energy storageb elonged to the class of conjugated polymers. [8] However,s ince those systems weren ot able to provide stable voltages and capacities, the first approaches were quickly discarded. Indeed, am ain drawbacko fc onjugated polymers is the factt hat their redox potentials usually dependo nt heir state of charge, leadingt ot he described sloping workingv oltage. [125] This is caused by the semiconductor-like electronic band structure of conjugatedp olymers, which is formed by the overlap of the p orbitals of the single "monomer" units of the polymer. [126] Nevertheless, the very same band structure providesc onjugated polymers with electronic and ionic conductivities that are superior to most otherr edox-active organic compounds when they are partly oxidized (or reduced), that is, doped. Thus, ad ecreased need for conductive additives can be achieved, making conjugated polymers interesting candidates for organic batteries( Figure 17). Furthermore, mostc onjugated polymers can be obtained via electrochemical polymerization,enabling an in situ formation in electrode setups.

Polytriphenylamine
Polytriphenylamines differ from the abovementioned polymers in their ability to form two-dimensional structures. Jiang and co-workersp reparedp olytriphenylamine networks that form microporous structures, which give rise to an enhanced ion transport( Figure 18). [131] The resulting hybrid lithium-ion cell revealed as table discharge voltage, ac apacity of 100 mAh g À1 , which remained stable over 200 cycles, and ag ood rate capability.
Through replacing someo ft he triphenylamine units by triphenylbenzene moieties, Zhang and co-workers created at wodimensional network with different pore sizes. The resulting hybridl ithium-ion cell showed ap erformance that resembled mostly that of ap olytriphenylamine-based analog, except for a slightly improvedr ate capability. [132] Ac ell based on ap olytriphenylamine and ag raphitee lectrode and ap otassium hexafluorophosphate-based electrolyte was demonstrated, displaying as table voltage but al ow capacity of 60 mAh g À1 and a coulombic efficiency of only 80 %. [133] Dong and co-workers prepared ap olytriphenylamine on MWCNTsvia in situ polymerizationo ft riphenylamine in adispersion of MWCNTs. [121] Subsequently,t he authors built ah ybrid lithium-ion cell without additional conductive additives, which revealed as table voltage of 3.7 Va nd ac apacity of 100 mAh g À1 ,r etaining 85 %a fter 100 cycles.
Further electrode materials based on functionalized polytriphenylamines were prepared. Ap olymer from triphenylamine and aniline, at riphenylamine-phenylene-imine polymer,a nd a methoxy-bearing polytriphenylamine suffered from strongly changing discharge voltages. [134] In contrast,l ithium-ion cells containing polytriphenylamines with cyano and methyl groups revealed stable voltages and capacities as well as good rate capabilities. [

Reviews
Conjugated polymers are still populara ctivem aterials foro rganic batteries. However,m ost of the studied compounds re-vealed unstable dischargev oltages, as long knownf or such systems, as well as low and decreasing capacities. As an excep- Figure 18. Two-dimensional polytriphenylamine networks in hybrid lithium cells. To pl eft:Molecular structure. To pr ight:C apacity development at different charge/discharge currents and over 1600 cycles for bothsystems (black and red), demonstrating both good rate capabilities and long-termstabilities. Middle: SEM images of the obtained microporous structures with different specific surface areas,showing aggregates of 60-80 nm in both cases. Bottom:C harge/discharge curves of the respective hybrid lithium cells at different currents. (Reprinted with permission from Elsevier,C opyright 2016). [131] ChemSusChem 2019, 12,4093 -4115 www.chemsuschem.org 2019 The Authors. Publishedb yWiley-VCH Verlag GmbH &Co. KGaA, Weinheim tion, polytriphenylamine proved to be ap romising basis for organic electrodes, with several examples that provided stable voltages and capacities as well as excellent rate capabilities. Thus, polytriphenylamines were even already used in all-organic cells combined with aromatic diimides (cf. Section 3.2). [74,75] 3.6. Aromatic compounds 3.6.1. Heterocycles Severalh eterocycles have been tested as organic charge-storage materials in batteries (Figure 19). Their aromatic nature enables them, in principle, to reversibly accept or release electrons. Cai et al. built ah ybrid sodium-ionc ell using poly(5-cyanoindole). The cell showed ad ecreasingd ischarge voltage, but its capacity,w hich reached 100 mAh g À1 after as hort equilibration, decreased only by 15 %o ver 1000 cycles. [135] Higashibayashi and co-workersp repared electrodes based on dicarbazole and di(dihydroacridine) for lithium-ion cells. [136] Notably,t he applicability of the dimers depended strongly on the linkage. While an only CÀC-coupled carbazole dimerr evealed no usable workingv oltage, its counterpart that was CÀC-and NÀ N-coupled revealed as lightly slopingv oltage around3 .5 Vand ac apacity of 50 mAh g À1 ,w hich remained stable over 20 cycles.Alikewise assembled dihydroacridine dimerr esulted in as table voltage and as lightly higher initial capacityo f 80 mAh g À1 but lost 35 %o vero nly 20 cycles.F urther promising heterocycles werep resented, in particular phenazine, [137] phenothiazine, [138] and thianthrene, [43] with the latter being even appliedi naworking, fully organic cell (Figure 3).

Other aromatic systems
Some larger polycyclic aromatic compounds were also investigated with regardt oa na pplicationi no rganic batteries ( Figure 19). 5,12-Diaminorubicene was used in ah ybrid lithium-ion cell, resultingi nasloping voltage and ac apacity loss of 35 %o ver the first 60 cycles. [139] Coronene,o nt he other hand, delivered as table voltage of 4.0 Va nd ac apacity that was rather low (35 mAh g À1 )b ut retained 85 % after 950 cycles. [139] Furthermore, Wang and co-workersp repared a series of azobenzenes andt ested them as active materials in hybrids odium-ion cells. [140] Notably,t he cells weret he more stable the more carboxylate groups were attached to the azobenzene, resulting in ac apacity of 170 mAh g À1 ,r etaining9 5% after 100 cycles.

Concluding Remarks
Duringt he recent years, many efforts were made with respect to the development of active organic electrode materials for electrochemical energy storage. Several new structuralm otifs were studied regardingt heir electrochemical behavior and their applicability in batteries. However,u pt on ow,t he most successful systemsa re the already established ones, in particular simple aromatic quinones (benzo-, naphtho-, and anthraquinone),a romatic diimides (naphthalene and perylene diimide), and the stable organic radicalT EMPO (cf. Sections 3.1, 3.2, and 3.4.1, respectively). Al arge amount of research was dedicated to the optimization of these compounds by the introduction of functional groups,integrationinto polymers, or combination of different redox-active moieties. In particular, an increasing interesti nt he development of mixed materials containing a redox-active organic species and an electrically conductive compound (e.g.,g raphene, carbon nanotubes (CNTs), as well as silver particles) is apparent. The components are either connectedd irectly to each other via chemical bonds (e.g.,u sing clickingt echniques) and p-p interactions (in particular for graphene and CNTs), or conductive nanoparticles are formed in situ within the active-material network. The driving force for this research is the desire for making the use of conductive additives as well as bindersr edundanti nf uture organic electrodes, thus increasing the overall specific capacity of the battery. Beside the established redox units, other aromatic materials, namely terephthalates, poly(triphenylamine)s, and some heterocyclic compounds (cf. Sections 3.3.1, 3.5, and 3.6.1, respectively) proved to be promising candidates. Consequently,t he number of fully organic approaches is increasing, ands everal workinga ll-organic cells were presented (Table 2) during the last years, revealing relatively high voltages, [43] capacities, [90] and stabilities. [75] There is still necessity for improvement and optimization but also many possibilities to do so due to the large toolbox that is provided by organic chemistry.H owever,n ot only the active materials offer many possibilitieso fd evelopment, but also the accompanying components, namely conductivea dditives as well as electrolyte salts and solvents, allow ac omprehensive optimization with regard to enhanced interactions and Figure 19. Heterocyclic moieties and other aromatic compounds that were tested as active materials for organic batteries.  poly(ethylenei mine-anthraquinone) 1.0 80 (À30 %/ 1000 cycles)a q. NaCl electrolyte [113]