TITLE A non-aqueous redox flow battery TECHNICAL FIELD The present invention relates to a battery, specifically to a non- aqueous redox flow battery. The invention further relates to a method of operating such a battery. BACKGROUND In a world where climate neutrality is the goal, green energy sources are essential. More than half of the produced green energy comes from wind and solar power. In order to overcome their intermittency, energy storage applications are inevitable. Redox flow batteries (RFBs) are very promising storage systems in the transition towards renewable energy sources. A redox flow battery generally comprises a negative electrode immersed in a liquid negative electrolyte, a positive electrode immersed in a liquid positive electrolyte, and an ion-permeable separator (e.g., a porous membrane, film, sheet, or panel as well as ion-exchange membranes) between the negative and the positive electrolyte. During charging and discharging, the electrolytes are circulated over their respective electrodes. The electrolytes each comprise either an electrolyte salt (e.g., a lithium, sodium, or organic salt) or a strong acid or base (e.g., NaOH, KOH, HCl or H ₂ SO ₄ ), a redox reactant, and optionally a solvent (e.g. an electrochemically stable organic solvent). They can be broadly classified in aqueous and non-aqueous systems. In aqueous RFBs, the voltage and hence energy density is limited by the water electrolysis potential (usually between 1.15 and 1.55V). Furthermore, the used electrolytes are in most of the cases highly corrosive (operate at pH 0 or 14). Non-aqueous systems operate with organic solvents, which allow for a much broader potential window (up to three times higher) compared to water. Combination of organic solvents with organic redox active materials could pave the way for all carbon-based RFBs with superior energy densities compared to aqueous systems. US10424806B2 discloses a non-aqueous redox flow battery wherein the redox reactant of the positive electrolyte is a dialkoxybenzene compound, and the redox reactant of the negative electrolyte is a viologen compound or a dipyridyl ketone.

US10535891 B2 discloses a redox flow battery comprising a two- electron, redox active, bridged, multi-cyclic compound (“TRBMC”) that comprises a non-aromatic, bridged cyclic portion fused to an aromatic cyclic portion.

US9300000B2 discloses a non-aqueous redox flow battery wherein each redox reactant is selected from an organic compound comprising a conjugated unsaturated moiety, a boron cluster compound, and a combination thereof. The organic redox reactant of the positive electrolyte is selected to have a higher redox potential than the redox reactant of the negative electrolyte.

In non-aqueous flow batteries an organic solvent can be used that can withstand higher potentials, but finding redox active compounds that offer a combination of higher redox potentials, high stability in charged form, and high solubility remains a challenge.

In addition, the currently used organic molecules, however, become more and more complex as they have to fulfill several requirements, e.g. regarding high solubility and stability, as well as high redox potentials and multi-electron redox events to increase capacitance. In general, implementing all these functionalities results in an undesirable higher molecular weight. Moreover, the synthetic complexity of the molecules tends to increase which reduces the feasibility for a large-scale application.

Therefore, there is a need for a redox flow battery that combines high redox potentials, high stability in charged form, and high solubility.

SUMMARY

The present invention therefore relates in a first aspect to a battery comprising a liquid electrolyte, the liquid electrolyte comprising: a compound selected from the group consisting of formula (I), formula (II), formula (III) and formula (IV):

In these formulas, R ¹ and R ² and/or R ³ and R ⁴ together with the aromatic carbon atoms to which they are attached form an optionally substituted cyclic group, or R ¹ , R ² , R ³ , R ⁴ and R ⁵ represent independent from each other H; CN; halogen; NO ₂ ; NR ⁶ 2; N ₂ R ⁶ ; COR ⁶ ; linear, branched or cyclic hydrocarbyl group independently selected from alkyl; linear, branched or cyclic hydrocarbyl group independently selected from alkyl, alkenyl, aryl, aralkyl, alkoxy, alkoxycarbonyl or alkylaryl groups, and one or more combinations thereof; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1-18 carbon atoms. R ⁶ represents H or a linear, branched or cyclic hydrocarbyl group independently selected from alkyl, alkenyl, aryl, aralkyl, alkoxy, alkoxycarbonyl or alkylaryl groups, and one or more combinations thereof; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1- 18 carbon atoms, wherein A- represents an anion. In a second aspect, the present invention relates to a method of manufacturing a battery according to the first aspect, the method comprising: - immersing a negative electrode in a first non-aqueous liquid electrolyte; - immersing a positive electrode in a second non-aqueous liquid electrolyte; - interposing a semi-permeable separator between the negative and positive electrodes; wherein the first non-aqueous liquid electrolyte comprises the liquid electrolyte as described for the first aspect. In a third aspect, the present invention relates to a method of operating a battery according to the first aspect of the present invention. The method comprises: providing the battery according to the first aspect. In a fourth aspect, the present invention relates to a compound according to formula (I). In this compound, R ² and R ⁴ are H. In this compound, R ¹ is CN; COR ⁶ ; linear, branched or cyclic hydrocarbyl group independently selected from alkyl, alkenyl, cyclic alkane, aryl, aralkyl, alkoxy, alkoxycarbonyl or alkylaryl groups; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1-18 carbon atoms. In this compound, R ³ is COOR ⁷ , wherein R ⁷ represents linear, branched or cyclic hydrocarbyl group independently selected from alkyl, alkenyl, cyclic alkane, aryl, aralkyl, alkoxy, alkoxycarbonyl or alkylaryl groups; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1-18 carbon atoms. In a fifth aspect, the present invention relates to a method of manufacturing said compound. Embodiments of the first aspect are applicably correspondingly to the second, third, fourth and fifth aspect according to the present invention. BRIEF DESCRIPTION OF DRAWINGS The present invention is described hereinafter with reference to the accompanying drawings in which embodiments of the present invention are shown and in which like reference numbers indicate the same or similar elements. Figure 1 shows a schematic representation of the working principle of a redox flow battery according to one example according to the present invention. Figure 2A shows the normalized discharge capacity of static, coin cell cycling, up to 3000 cycles. Figure 2B shows the normalized discharge capacity of static, coin cell cycling, up to 10000 cycles. Figure 2C shows the normalized discharge capacity of static, coin cell cycling at higher concentration, up to 1000 cycles. Figure 2D shows the normalized discharge capacity of static, coin cell cycling up to 500 cycles. Figure 2E shows the normalized discharge capacity of static, coin cell cycling at a concentration of 50 mM with various catholytes, up to 1000 cycles. Figure 2F shows the normalized discharge capacity of static, coin cell cycling of the most promising anolyte/catholytes combinations, up to 20000 cycles. Figure 3 shows the volumetric capacity versus the cycle number of flow battery cycling for an example according to the present invention. Figure 4A shows volumetric capacity versus cycle number of flow battery cycling for another example according to the present invention. Figure 4B shows a voltage versus capacity profile for the example of Fig.4A. Figure 5A shows volumetric capacity versus cycle number of flow battery cycling for another example according to the present invention. Figure 5B shows a voltage versus capacity profile for the example of Fig.5A. DESCRIPTION It is an object of the present invention to provide an improved non- aqueous redox flow battery. It is a further object of the present invention to provide a non-aqueous redox flow battery with deep reduction potential anolytes and ultimately a high cell voltage and/or high stability in charged form, and/or high solubility. It is a further object of the present invention to provide a non-aqueous redox flow battery that combines deep reduction potential anolytes and ultimately a high cell voltage, high stability in charged form, and high solubility. It is a further object of the present invention to provide deep reduction potential anolytes for non-aqueous redox flow batteries of which the organic molecules are relatively easy to synthesize. As stated above, the invention relates in a first aspect to a battery comprising a liquid electrolyte comprising a compound selected from the group consisting of formula (I), formula (II), formula (III) and formula (IV). This compound is able to take up an electron. In some types of batteries, such as in redox flow batteries, this compound would be defined as the anolyte. In an embodiment, the liquid electrolyte comprises comprises two or more compounds selected from the group consisting of formula (I), formula (II), formula (III) and formula (IV). When there are two or more of such compounds, preferably said compounds have a similar redox potential. For example, the redox potentials of said two or more compounds differ less than 200 mV. The presence of two or more of such compounds will lead to a similar or higher solubility than when the liquid electrolyte comprises one of such compounds. In an embodiment, the liquid electrolyte comprises a compound selected from the group consisting of formula (I), formula (II) and formula (IV). In an embodiment, R ¹ and R ² and/or R ³ and R ⁴ together with the aromatic carbon atoms to which they are attached form an optionally substituted cyclic group. In other words, the ring depicted in the formula forms a fused ring system with a ring formed by R ¹ and R ² and/or R ³ and R ⁴ . The substituent(s), when present, may be branched or unbranched. In an embodiment, R ¹ and R ² as well as R ³ and R ⁴ together with the aromatic carbon atoms to which they are attached form an optionally substituted cyclic group. Suitable examples of an optionally substituted cyclic group are an optionally substituted cycloalkane group, cycloheteroalkane group, cycloalkene group and cycloheteroalkene group. In an embodiment, the optionally substituted cyclic group is a cycloalkane group. In an embodiment of the first aspect according to the present invention, R ¹ , R ² , R ³ , R ⁴ and R ⁵ represent independent from each other H; CN; halogen; NR ⁶ 2; COR ⁶ ; linear, branched or cyclic hydrocarbyl group independently selected from alkyl, alkenyl, aryl, aralkyl, alkoxy, alkoxycarbonyl or alkylaryl groups, and one or more combinations thereof; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1-10 carbon atoms, preferably from 1-5 carbon atoms; In an embodiment of the first aspect according to the present invention, R ¹ , R ² , R ³ , R ⁴ and R ⁵ represent independent from each other H; CN; COR ⁶ ; linear or branched hydrocarbyl group independently selected from alkyl, alkoxy and alkoxycarbonyl groups, and one or more combinations thereof; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1-10 carbon atoms, preferably from 1-5 carbon atoms; In an embodiment of the first aspect according to the present invention, R ⁶ represents H or a linear, branched or cyclic hydrocarbyl group independently selected from alkyl, alkenyl, aryl, aralkyl, alkoxycarbonyl or alkylaryl groups, and one or more combinations thereof; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1-10 carbon atoms, preferably from 1-5 carbon atoms. In an embodiment of the first aspect according to the present invention, at least one of R ¹ of R ² and at least one of R ³ and R ⁴ is H, preferably wherein R ² = R ⁴ = H. In an embodiment of the first aspect according to the present invention, R ¹ is not H. In an embodiment of this, R ³ is also not H. In an embodiment, at least one of R ¹ and R ³ is branched. In an embodiment, R ¹ and R ³ are independently chosen from CH ₃ , C ₂ H ₅ , C ₃ H ₇ , C ₄ H ₉ , tert-butyl, COOCH ₃ , COOC ₂ H ₅ , COOC ₃ H ₇ , COOC ₄ H ₉ and COOC(CH ₃ ) ₃ . R ³ may be the same or different from R ¹ , preferably R ³ is different from R ¹ . The presence and nature of substituents allow for adjustability of the stability, solubility and redox potential of the compound. When R ³ and R ¹ are different, this asymmetry allows for finetuning of redox potential and a higher solubility. When the compound is according to formula (III), there is an anion present. This anion must be electrochemically inert in the potential window in which the battery is operated. Examples of suitable anions include PF ₆ -, BF ₄ -, N(SO ₂ CF ₃ ) ₂ - (TFSI), CF ₃ SO ₃ - (FSI), CH ₃ SO ₃ -, CH ₃ C ₆ H ₄ SO ₃ - (OTs) or diethyl phosphate. In an embodiment, the compound is selected from formula (I), (II) or (IV). In an embodiment, the liquid electrolyte comprises a compound according to one of the following formulas:

In an embodiment, the liquid electrolyte comprises a compound according to one of the following formulas: In an embodiment, the liquid electrolyte comprises a compound according to one of the following formulas: preferably wherein the compound is according to any one of formulas (1) – (13), more preferably according to any one of formulas (1) – (12). In a preferred embodiment, the liquid electrolyte comprises a compound according to one of the following formulas:

Without wishing to be bound by theory, the inventors believe that, while compounds with formulas (I)-(IV) provide deep reduction potentials, a careful selection of the side groups R ¹ , R ² , R ³ , R ⁴ and R ⁵ will impart stability in the neutral and reduced states on these compounds in combination with high solubility in organic electrolytes via a proper choice of R ¹ , R ² , R ³ , R ⁴ and R ⁵ , via an increase and differentiation in length, branching, and bulkiness of the side groups. In this respect the inventors believe that combinations of branched-linear or branched-branched, and linear side groups, as well as ester groups of different lengths are particular successful for enhancing solubility. In an embodiment of the first aspect according to the present invention, the battery further comprises a solvent that is electrochemically stable in a wide potential range, preferably between -3V to +2V, such as between -2.5V to + 1.5V vs ferrocene/ferrocenium (Fc/Fc ⁺ ), wherein the solvent is acetonitrile; an ether-based solvent, preferably dimethoxyethane; diethyl carbitol, dimethylformamide or a mixture thereof. In an embodiment of the first aspect according to the present invention, the battery further comprises one or more electrolyte salts and a counter ion, wherein the one or more electrolyte salts has a solubility of >0.5 M in the solvent. In a specific embodiment, the one or more electrolyte salt is a quaternary ammonium or imidazolium. In a specific embodiment, the counter ion is chosen from the group of PF ₆ -, BF ₄ -, N(SO ₂ CF ₃ ) ₂ - (TFSI), CF ₃ SO ₃ - (FSI), CH ₃ SO ₃ - or CH ₃ C ₆ H ₄ SO ₃ - (OTs) and diethyl phosphate. In an embodiment of the first aspect according to the present invention, the battery is a redox flow battery. In a specific embodiment, the battery is a non-aqueous redox flow battery. As stated above, the invention relates in a first aspect to a battery comprising a compound according to one of the formulas (I)-(IV). When the battery is a redox flow battery, this compound serves as the anolyte. In other types of batteries, such as lithium ion or lead acid batteries, this compound is often described as material to store electrons, or as material for the anode. In an embodiment, the battery further comprises a material for the cathode (in certain types of batteries such as in redox flow batteries called a catholyte). This catholyte may for instance be chosen from 2,5-di-tert-butyl-1,4-bis(2- methoxyethoxy)benzene (DBBB), 1,4-di-tert-butyl-2-methoxy-5-(2-(2- methoxyethoxy)ethoxy)benzene (DBMMB) as well as 1,4-di-tert-butyl-2,5-bis(2,2,2- trifluoroethoxy)benzene compounds. Other suitable catholytes are 10-[2-(2- methoxyethoxy)ethyl]-10H-phenothiazine (MEEPT) and other phenothiazines, N- (ferrocenylmethyl)-N,N-dimethylethanaminium bis-(trifluoromethanesulfonyl)imide (Fc1N112-TFSI) and other ferrocenium derivatives as well as carbazole derivatives (e.g. 1,3,6,8-tetra-tert-butyl-2-ethylhexyl-carbazole, Cbz1). The catholyte preferably has a high oxidation potential, stability and solubility. The working principle of a redox flow battery 1 is shown in Fig. 1. During discharging of the battery 1, electrons 3 are released via an oxidation reaction from a high chemical potential state on the anode side 5 of the battery 1. The electrons 3 move through the external circuit 7 and provide the electrical energy. Then, the electrons 3 are accepted via a reduction reaction at a lower chemical potential state on the cathode side 9 of the battery 1. To compensate charge neutrality ions move through a membrane 11 inside the cell while on each side 5, 9 the side’s 5,9 respective liquid is circulated in the side’s 5,9 own respective space 13, 15. The voltage, or electromotive force, generated in each cell 17 of the battery 1 is determined by the total difference in chemical potential between the chemical states of the active materials (anolyte and catholyte) on the two sides 5, 9 of the battery 1. The voltage is dependent on the chemical species involved in the reactions as well as the number of cells 17 arranged in series. When charging the battery, this process will take place in a reverse manner. With the present invention, at least one of the objects is achieved. The terms “deep reduction potential”, “high cell voltage”, “high stability in charged form” and “high solubility” may be seen as relative terms, but in the context of this invention they may be defined as follows. A deep reduction potential can be defined as a reduction potential of or deeper than -1.5 V vs Fc/Fc ⁺ , preferably of or deeper than -2 V, more preferably of or deeper than -3 V. High stability may be defined as at least 1000 cycles for cyclability (for instance with at least 50% retention of charge- discharge capacity) and at most 10% per month self-discharge. High solubility determines together with the cell voltage the energy density of the battery. For example, with a cell voltage of 2.75 V (with DBBB or DBMMB catholyte), a solubility of 0.7 M would outperform vanadium RFBs in terms of energy density. The present invention further relates to a method of manufacturing a non-aqueous redox flow battery, the method comprising: - immersing a negative electrode in a first non-aqueous liquid electrolyte; - immersing a positive electrode in a second non-aqueous liquid electrolyte; - interposing a semi-permeable separator between the negative and positive electrodes; wherein the first non-aqueous liquid electrolyte comprises a compound selected from the group consisting of formula (I), formula (II), formula (III) and formula (IV): wherein R ¹ and R ² and/or R ³ and R ⁴ together with the aromatic carbon atoms to which they are attached form an optionally substituted cyclic group, or wherein R ¹ , R ² , R ³ , R ⁴ and R ⁵ represent independent from each other H; CN; halogen; NO ₂ ; NR ⁶ 2; N ₂ R ⁶ ; COR ⁶ ; linear, branched or cyclic hydrocarbyl group independently selected from alkyl, alkenyl, cyclic alkane, aryl, aralkyl, alkoxy, alkoxycarbonyl or alkylaryl groups, and one or more combinations thereof; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1-18 carbon atoms; wherein R ⁶ represents H or a linear, branched or cyclic hydrocarbyl group independently selected from alkyl, alkenyl, aryl, aralkyl, alkoxy, alkoxycarbonyl or alkylaryl groups, and one or more combinations thereof; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1-18 carbon atoms; wherein A- represents an anion. In other terms, the second electrolyte is the liquid electrolyte that is comprised in the battery according to the present invention. The present invention relates further to a method of operating a battery according to the present invention, the method comprising providing the battery according to the first aspect. The present invention also relates to a compound according to formula (I). In this compound, R ² and R ⁴ are H. In this compound, R ¹ is CN; COR ⁶ ; linear, branched or cyclic hydrocarbyl group independently selected from alkyl, alkenyl, cyclic alkane, aryl, aralkyl, alkoxy, alkoxycarbonyl or alkylaryl groups; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1-18 carbon atoms. In this compound, R ³ is COOR ⁷ , wherein R ⁷ represents linear, branched or cyclic hydrocarbyl group independently selected from alkyl, alkenyl, cyclic alkane, aryl, aralkyl, alkoxy, alkoxycarbonyl or alkylaryl groups; wherein said hydrocarbyl group may be substituted or unsubstituted, may contain one or more heteroatoms and has from 1-18 carbon atoms. To clarify, the compound described above is according to the following formula: Both R ¹ and R ⁷ (which is the ester substituent) have as first (viewed from the depicted structure) atom a carbon atom. Further in the chain there may be heteroatoms such as oxygen or nitrogen present (such as when the R-group is CN or COR ⁶ ). In an embodiment, R ¹ is selected from CH ₃ , CH ₂ CH ₃ , C ₃ H ₇ , C ₄ H ₉ , tertbutyl, COOCH ₃ and R ⁷ is selected from CH ₃ , CH ₂ CH ₃ , C ₃ H ₇ , C ₄ H ₉ and C(CH ₃ ) ₃ (thus R ³ is selected from COOCH ₃ , COO CH ₂ CH ₃ , COOC ₃ H ₇ , COOC ₄ H ₉ and COOC(CH ₃ ) ₃ ). In an embodiment, R ¹ is methyl, and R ⁷ is CH ₃ or CH ₂ CH ₃ (thus R ³ is COOCH ₃ or COOCH ₂ CH ₃ ). The compounds of this embodiment are thus methyl 2,5- dicyano-4-methylbenzoate and ethyl 2,5-dicyano-4-methylbenzoate, and have the following formulas: . The present invention relates also to a method of manufacturing the above-described compounds according to formula (I). In an embodiment, this method comprises the provision of a mixture of para-substituted dibromo derivative and copper cyanide. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measured cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope thereof. The scope of the present invention is defined by the appended claims. One or more of the objects of the invention are achieved by the appended claims. EXAMPLES

The present invention is further elucidated based on the Examples below which are illustrative only and not considered limiting to the present invention.

If for a reported value no specific method of measurement is explicitly mentioned, it can be assumed that the same method of measurement is used as described elsewhere in the present description.

Synthesis of dicyanobenzenes

Synthesis routes of several dicyanobenzenes are as follows, and further elaborated below.

Molecules 1 - 3 can all be synthesized with basic chemistry knowledge from cheap benzene derivatives.

2-Ethylterephthalonitrile 3: 1 ,4-Dibromo-2-ethylbenzene (500 mg,

1.89 mmol) was dissolved in anhydrous N-methyl-2-pyrrolidone (NMP) (0.75 M), copper cyanide (339 mg, 3.79 mmol) added, and the mixture stirred under reflux overnight. After cooling down, the mixture was poured into 10% aqueous ammonia and extracted 3 times with dichloromethane (DCM). The combined organic phase was twice washed with demineralized water, dried over MgSO ₄ and concentrated under vacuo. The crude mixture was recrystallized from heptane to yield 163 mg of molecule 3 as white crystals (55%). ¹ H NMR (400 MHz, Chloroform-d) δ 7.73 (d, 1H), 7.64 (s, 1H), 7.59 (dd, 1H), 2.94 (q, 2H), 1.34 (t, 3H). The synthesis of molecule 5 has been described in US009603848B2. 2,5-Diethylterephthalonitrile 6: 1,4-Dibromo-2,5-di-ethylbenzene (300 mg, 1.03 mmol) was dissolved in anhydrous DMF (0.2 M), copper cyanide (202 mg, 2.26 mmol) added, and the mixture stirred under reflux overnight. After cooling down, the mixture was poured into 10% aqueous ammonia and extracted 3 times with DCM. The combined organic phase was twice washed with demineralized water, dried over MgSO ₄ and concentrated under vacuo. The crude mixture was recrystallized from heptane to yield 141 mg of molecule 6 as white crystals (75%). 2,5-Dibromo-1,4-di-tert-butylbenzene: 1,4-Di-tert-butylbenzene (1 g, 5.25 mmol) was dissolved in chloroform (0.5 M) and cooled to 0°C. Bromine (0.54 mL, 10.51 mmol) and iodine (0.133 g, 0.53 mmol) were added, and the mixture stirred for two days. After pouring slowly into 20% aqueous sodium hydroxide the solution discolored and the organic phase was separated. The aqueous phase was extracted 2 times with DCM and the combined organic phase washed with demineralized water, dried over MgSO ₄ and concentrated under vacuo. The crude mixture was recrystallized from heptane to yield 0.82 g of 1,4-dibromo-2,5-di-tert-butylbenzene as white crystals (45%). %). ¹ H NMR (400 MHz, Chloroform-d) δ 7.58 (s, 2H), 1.48 (s, 18H). An analysis of the crude mixture revealed the presence of 48% of 1-bromo-2,5-di-tert- butylbenzene. 2,5-Di-tert-butylterephthalonitrile 7: 1,4-Dibromo-2,5-di-tert- butylbenzene (350 mg, 1.01 mmol) was dissolved in anhydrous DMF (0.2 M), copper cyanide (198 mg, 2.21 mmol) added, and the mixture stirred under reflux for two days. After cooling down, the mixture was poured into 10% aqueous ammonia and extracted 3 times with DCM. The combined organic phase was washed twice with demineralized water, dried over MgSO ₄ and concentrated under vacuo. The crude mixture was recrystallized from heptane to yield 176 mg of molecule 7 as white crystals (73%). ¹ H NMR (400 MHz, Chloroform-d) δ 7.74 (s, 2H), 1.52 (s, 18H). 2,5-Dibromo-4-tert-butyltoluene: 4-tert-Butyl-toluene (1 g, 6.75 mmol) was mixed with chloroform (1 M) and cooled to 0°C. Bromine (0.78 mL, 15.18 mmol) and iodine (0.171 g, 0.67 mmol) were added, and the mixture stirred for two days. After pouring slowly into 20% aqueous sodium hydroxide the solution discolored and the organic phase was separated. The aqueous phase was extracted twice with DCM and the combined organic phase washed with demineralized water, dried over MgSO ₄ and concentrated under vacuo resulting in a colorless oil. The crude mixture was analyzed by NMR and revealed a mixture of 0.99 g (48%) of 2,5-dibromo- 4-tert-butyltoluene and 0.75 g (37%) of unwanted 2,6-dibromo-4-tert-butyltoluene. It was not possible to separate both products by either column chromatography or distillation. ¹ H NMR of 2,5-dibromo-4-tert-butyltoluene (400 MHz, Chloroform-d) δ 7.54 (s, 1H), 7.45 (s, 1H), 2.31 (s, 3H), 1.48 (s, 9H). ¹ H NMR of 2,6-dibromo-4-tert- butyltoluene (400 MHz, Chloroform-d) δ 7.49 (s, 2H), 2.53 (s, 3H), 1.28 (s, 9H). 2-tert-Butyl-5-methylterephthalonitrile 8: A mixture of 2,5- Dibromo-4-tert-butyltoluene (56.5%) and 2,6-Dibromo-4-tert-butyltoluene (43.5%) (500 mg, 1.63 mmol) was dissolved in anhydrous NMP (0.3 M), copper cyanide (293 mg, 3.27 mmol) added, and the mixture heated in the microwave for 20 minutes. After cooling down, the mixture was poured into 10% aqueous ammonia and extracted 3 times with DCM. The combined organic phase was washed twice with demineralized water, dried over MgSO ₄ and concentrated under vacuo to yield a brown oil. Column chromatography in heptane/DCM (starting with 10:1) yielded 103 mg of molecule 8 (31.8%). ¹ H NMR (400 MHz, Chloroform-d) δ 7.68 (s, 1H), 7.62 (s, 1H), 2.54 (s, 3H), 1.51 (s, 9H). As a side product, 69 mg of 2-tert-butyl-5-methyl-1,4-dicyanobenzene (21.2%) could be isolated. ¹ H NMR (400 MHz, Chloroform-d) δ 7.81 (s, 2H), 2.72 (s, 3H), 1.33 (s, 9H). Diethyl 2,5-dicyanoterephthalate 9: A mixture of diethyl 2,5- dibromoterephthalate (2 g, 5.26 mmol) and copper cyanide (1.04 g, 11.58 mmol) was dissolved in DMSO (0.5 M) and the mixture heated for 3 h at 120 °C. After cooling down, the precipitate was filtered off, and the filtrate poured into 10% aqueous ammonia. The solids were filtered off again. Both solids were washed with DCM, the solvents combined and washed twice with demineralized water. After drying over MgSO ₄ and concentration under vacuo beige solids remained. Recrystallization in heptane/ethyl acetate yielded 1.33 g of molecule 9 (93%). ¹ H NMR (400 MHz, Chloroform-d) δ 8.54 (s, 2H), 4.54 (q, J = 7.15 Hz, 4H), 1.49 (t, J = 7.15 Hz, 6H). Dibutyl 2,5-dibromoterephthalate: A mixture of 2,5- dibromoterephthalic acid (2 g, 6.17 mmol) and thionyl chloride (1.5 M) was refluxed for 4 h. After evaporation of the thionyl chloride, the residue was dissolved in anhydrous THF (1.5 M), 1-butanol (3 mL, 32.72 mmol) and the mixture stirred at room temperature overnight.50 mL of DCM were added, and the mixture washed four times with demineralized water. After drying over MgSO ₄ and concentration under vacuo a white gel-like compound remained (52%, 1.4 g). ¹ H NMR (400 MHz, Chloroform-d) δ 8.01 (s, 2H), 4.37 (q, J = 6.62 Hz, 4H), 1.77 (quintet, J = 6.87 Hz, 4H), 1.46 (dt, J = 14.57, 7.39 Hz, 4H), 1.77 (t, J = 7.38 Hz, 6H). Dibutyl 2,5-dicyanoterephthalate 10: A mixture of dibutyl-2,5- dibromoterephthalate (1.4 g, 3.21 mmol) and copper cyanide (0.63 g, 7.06 mmol) was dissolved in DMSO (0.5 M) and the mixture heated for 3 h at 120 °C. After cooling down, the precipitate was filtered off, and the filtrate poured into 10% aqueous ammonia. The solids were filtered off again. Both solids were washed with DCM, the solvents combined and washed twice with demineralized water. After drying over MgSO ₄ and concentration under vacuo a beige solid remained. Recrystallization in heptane/ethyl acetate yielded 0.66 g of molecule 10 (63%). ¹ H NMR (400 MHz, Chloroform-d) δ 8.53 (s, 2H), 4.48 (t, J = 6.65 Hz, 4H), 1.83 (quintet, J = 6.85 Hz, 4H), 1.49 (dt, J = 14.61, 7.33 Hz, 4H), 1.00 (t, J = 7.36 Hz, 6H). 2,5-Bromo-4-methylbenzoic acid: 1,4-Dibromo-2,5- dimethylbenzene (2.61 g, 9.89 mmol) was dissolved in water/nitric acid (55/45%, 0.5 M) and the mixture heated for 6 h at 130 °C in a pressure-tight microwave vial and the pressure released by a needle. The mixture was cooled to 120 °C and stirred for an additional 16 h. After cooling down to 80°C, the pressure was released and the precipitate filtered hot and washed with water. After drying at 70 °C under vacuo, 2.21 g of a white solid of 2,5-dibromo-4-methylbenzoic acid remained (76%). ¹ H NMR (400 MHz, DMSO-d ₆ ) δ 13.56 (br s, 1H), 7.93 (s, 1H), 7.76 (s, 1H), 2.37 (s, 3H). Methyl 2,5-dibromo-4-methylbenzoate: 2,5-Dibromo-4- methylbenzoic acid (1.11 g, 3.78 mmol) was dissolved in thionyl chloride (7.6 mL) and refluxed for 3 h. After evaporation of the thionyl chloride, ethanol (7.6 mL) was added at 0°C and mixture stirred overnight. Concentration under vacuo led to a white powder, 1.14 g of methyl 2,5-dibromo-4-methylbenzoate (98%). ¹ H NMR (400 MHz, Chloroform-d) δ 8.01 (s, 1H), 7.54 (s, 1H), 3.92 (s, 3H), 2.41 (s, 3H). Methyl 2,5-dicyano-4-methylbenzoate 11: A mixture of Ethyl 2,5- dibromo-4-methylbenzoate (1.14 g, 3.70 mmol) and copper cyanide (0.73 g, 8.14 mmol) was dissolved in DMSO (0.5 M) and the mixture heated for 2 h at 130 °C. After cooling down, the mixture was poured into 10% aqueous ammonia and extracted twice with DCM. The combined organic phase was washed twice with demineralized water. After drying over MgSO ₄ and concentration under vacuo a brown solid remained. Recrystallization in heptane/ethyl acetate yielded 0.49 g of beige/yellowish crystals, which turned out to be a mixture between singly and doubly cyanated product. Column chromatography in heptane/DCM (starting with pure heptane, going up to 67% DCM) yielded 0.07 g of molecule 11 (9%). ¹ H NMR (400 MHz, Chloroform-d) δ 8.37 (s, 1H), 7.78 (s, 1H), 4.03 (s, 3H), 2.67 (s, 3H). ¹³ C NMR (100 MHz, Chloroform-d) δ 162.75, 147.05, 136.30, 135.02, 117.18, 116.59, 116.05, 115.61, 53.33, 20.57. Ethyl 2,5-dibromo-4-methylbenzoate: 2,5-Dibromo-4- methylbenzoic acid (0.49 g, 1.65 mmol) was dissolved in thionyl chloride (10 mL) and refluxed for 3 h. After evaporation of the thionyl chloride, ethanol (20 mL) was added at 0°C and mixture stirred overnight. After adding water and DCM, the organic phase was separated and twice washed with demineralized water. Drying over MgSO ₄ and concentration under vacuo led to a dark brown liquid/oil. Column chromatography in heptane/DCM (starting with pure heptane, going up to 33% DCM) yielded 0.174 g of ethyl 2,5-dibromo-4-methylbenzoate (26%). ¹ H NMR (400 MHz, Chloroform-d) δ 7.98 (s, 1H), 7.53 (s, 1H), 4.39 (q, 2H), 1.40 (t, 3H). Ethyl 2,5-dicyano-4-methylbenzoate 12: A mixture of Ethyl 2,5- dibromo-4-methylbenzoate (0.174 g, 0.54 mmol) and copper cyanide (0.11 g, 1.19 mmol) was dissolved in DMSO (0.5 M) and the mixture heated for 1 h at 120 °C and 1 h at 130 °C. After cooling down, the mixture was poured into 10% aqueous ammonia and extracted twice with DCM. The combined organic phase was washed twice with demineralized water. After drying over MgSO ₄ and concentration under vacuo a brown solid remained. Recrystallization in heptane/ethyl acetate yielded 0.11 g of beige/yellowish crystals, which turned out to be a mixture between singly and doubly cyanated product. Column chromatography in heptane/DCM (starting with pure heptane, going up to 67% DCM) yielded 0.024 g of molecule 12 (21%). ¹ H NMR (400 MHz, Chloroform-d) δ 8.36 (s, 1H), 7.77 (s, 1H), 4.48 (q, 2H), 1.45 (t, 3H). Cyclic voltammetry Cyclic voltammetry is a technique used to investigate oxidation and reduction processes of redox-active compounds. By applying a potential sweep in alternating directions, peaks emerge which indicate at which potential oxidation and/or reduction processes occur. In the ideal case their shape follows the Nernst equation and peak separation, and height can be used to determine diffusion coefficients and homogenous electron transfer rates. Cyclic voltammograms of the molecules 1-17 at a concentration of 5 mM in various solvents and electrolyte salts were recorded with a scan rate of 100 mV s ⁻¹ and shown in Table 1. In Table 1, the following abbreviations are used: can = acetonitrile, DMF = dimethylformamide, PhCN = benzonitrile, Pyr = pyridine, DME = dimethoxyethane, TBAPF ₆ = tetrabutylammonium hexafluorophosphate, TEATFSI = tetraethylammonium bis(trifluoromethanesulfonyl)imide, KTFSI = potassium bis(trifluoromethanesulfonyl)imide. Table 1: Reduction potentials vs Fc/Fc ⁺ and solubilities (if determined) of compounds 1-18.

Redox reaction in the battery application with electrolyte salt included The following reactions take place in the battery application with electrolyte salt included: Charging: Reduction: Anolyte + e ⁻ → Anolyte ⁻ Oxidation: Catholyte → Catholyte ⁺ + e ⁻ Discharging: Oxidation: Anolyte ⁻ → Anolyte + e ⁻ Reduction: Catholyte ⁺ + e ⁻ → Catholyte This depicts the redox processes happening while charging and discharging in the battery. While charging, the anolyte takes up an additional electron. On the catholyte side, another molecule gets oxidized, which means an electron gets removed. In order to balance the charges, the positive part of the electrolyte salt, in this case tetrabutylammonium, moves towards the anolyte side and its negative part, the hexafluorophosphate, to the catholyte side. Upon discharging the opposite process takes place. In a nitrogen-filled glove box 4 mL of 5 mM solutions of molecule 3 respectively 7 with 200 mM electrolyte salt (TBAPF ₆ ) in acetonitrile get loaded on each side of an H-cell. With a biologic VSP potentiostat a current is applied and the working side negatively charged. After full charging, the solution is removed and filled into a screw-capped glass vial. Samples for absorption measurements were prepared in a 0.2 mm pathlength quartz cuvette in the glove box but measured outside the glove box at different points in time. While the at the core positions unprotected molecule 3 lost its charge completely after 10 days, the charged solution of molecule 7 still contained 31% of its initial charge after 43 days. CV measurements prior to charging and after storing for 43 days were performed and revealed a similar state-of-charge (SoC) after 43 days of 27%, determined via the open-circuit voltage. A comparison of the peak heights for the reduction implies that 18% of the material has been degraded, which could be another hint towards unwanted reactions of the active material with impurities. Also, the capacity retention of molecule 9 has been evaluated and showed promising results of more than 80% retention after being stored in its reduced state for 20 days. The 50 mM battery cycling of molecule 3 with a commonly used catholyte shows a retention of 73% after 100 cycles. In a real application, the battery might not always get fully charged which is why the charging to various states-of- charge was tested and showed similar retention of 74% after 100 cycles. A 3.22 V battery with molecule 6 and a high oxidation potential catholyte, presented in ACS Appl. Mater. Interfaces 2022, 14, 28834, showed a retention of 61% after 100 cycles. All batteries showed high energy efficiencies of over 70%. Depolarization measurements of a flow batteries of a mixed solution of 50 mM of molecule 3 with 55 mM DBBB and 200 mM TBAPF ₆ in acetonitrile as well 200 mM of molecule 3 with 220 mM DBBB and 400 mM TBAPF ₆ in acetonitrile at 100% SoC have been performed. In both cases a porous Daramic 175 separator has been used. The two measurements revealed almost identical results and higher achievable power densities than a 100 mM MV/TEMPOL and a 1.67 M Vanadium flow battery, shown in J. Noack et al., Energies 2016, 9, 627, due to the more than doubled voltage of the molecule 3/DBBB-system. The maximum power density for both systems could not be determined as the potentiostat limited the current to 0.157 A/cm ² . Extrapolation suggests a very high maximum power density of about 0.39 W/cm ² at a current density of 0.27 A/cm ² . Also, the performance of the molecule 9/MEEPT-system was evaluated and suggested an extrapolated maximum power density of about 0.142 W/cm ² at a current density of 0.176 A/cm ² . Figure 2A shows static, coin cell cycling of 45 or 50 mM of molecule 3 (full squares and diamonds), 7 (half squares and half diamonds), 8 (stars), respectively 12 (circles), and 50 or 55 mM DBBB in 200 mM TBAPF ₆ /acetonitrile if not differently indicated. Only every 50 ^th cycle is shown for better clarity. Figure 2B shows static, coin cell cycling of 50 mM of molecule 7 and 8 with 50 mM DBBB in 200 mM TBAPF ₆ /acetonitrile. Only every 200 ^th cycle is shown for better clarity. Figure 2C shows static, coin cell cycling at high a concentration of 500 mM of molecule 7 and 350 mM of 8 with 550 mM DBMMB in 750 mM TBAPF ₆ /acetonitrile, respectively 385 and 525 mM. Only every 20 ^th cycle is shown for better clarity. Figure 2D shows static, coin cell cycling of molecule 11 at a concentration of 50 mM in 200 mM TBAPF ₆ /acetonitrile. Only every 10 ^th cycle is shown for better clarity. Figure 2E and 2F show static, coin cell cycling of molecule 9 at a concentration of 50 mM with various catholytes at 50 mM in 200 mM TBAPF ₆ /acetonitrile. Only every 20 ^th cycle is shown for better clarity. Even though, coin cells are a static system, the charge-discharge behavior of redox active molecules can be evaluated without wasting large amounts of materials. Thereby, the cycling of the molecules 7 and 8 showed almost an order of magnitude better stability than the one of molecule 3, while molecule 9 exhibited almost two orders of magnitude higher stability than 3. The superior cycling performance of molecule 7, 8 and 9 could be demonstrated over 10,000 cycles. The molecule 9/MEEPT system could even be demonstrated over 20,000 cycles with a loss of only 9% of its initial discharge capacity. Figure 3 shows the volumetric capacity versus cycle number of flow battery cycling of a mixed solution of 50 mM of molecule 7 with 55 mM DBBB and 200 mM TBAPF ₆ in acetonitrile using a porous Daramic 175 separator. To prove the possibility to operate dicyanobenzene flow batteries also at higher concentration, the very soluble molecule 3 was paired with DBMMB. A 0.5 M battery was run for 50 cycles as a demonstration (Figure 4A and 4B). Further optimization has to be done for a better performance. The 50 mM flow battery cycling of molecule 7 shows very promising results. Charging was performed with a constant current of 30 mA/cm ² , with a voltaic cutoff at 3.1 V. Discharging was performed at a constant current of -30 mA/cm ² with a voltaic cutoff at 1.5 V. Very high average coulombic and energy efficiencies of 93.4% and 79% were achieved. The battery’s capacity retention stayed very steady after an initial capacity drop between cycle 50 and 150. The fast capacity drop within the first 150 cycles can be attributed to transport phenomena. After these cycles, about 1.25 mL have been transferred from anolyte to catholyte reservoir including active compounds and electrolyte salt. Analyzing both reservoirs after 500 cycles by cyclic voltammetry revealed that at least 73% of molecule 7 remained in the anolyte reservoir and 80% in the catholyte reservoir, showing an excellent stability of the molecule itself. The decay from cycle 200 to 500 is about -0.06%/cycle which correlates with a decay of about 0.05%/cycle taking the capacity retention of 73% determined by CV. Furthermore, the system of molecule 9/MEEPT was evaluated in a flow battery at a concentration of 50 mM in 200 mM TBAPF ₆ in acetonitrile (Figure 5A and 5B). Thereby, no detectable decay could be determined by the discharge capacity nor cyclic voltammetry with a microelectrode after 100 cycles. Table 2 shows a comparison of molecules 6, 7 and 9 (numbers indicated in brackets in the first column) according to the present invention with anolytes from literature. Table 2: Comparison with literature.

Figure 4A shows volumetric capacity versus cycle number of flow battery cycling of a mixed solution of 500 mM of molecule 3 with 550 mM DBBB and 750 mM TBAPF ₆ in acetonitrile using a porous Daramic 175 separator. Figure 4B shows the charging and discharging voltages in dependance of their SoC (expressed via capacity) for cycle number 1 and 25. The upper lines represent the charging and the lower ones the discharging. The curves reveal relatively high average discharge voltages of 2.43 and 2.33 V for cycle 1, respectively 25, as well as a slight overcharge in the first cycle which can be attributed to active material crossing over through the porous separator. The 500 mM flow battery cycling shows very promising results. Molecule 3 exhibits at least a high solubility of 500 mM in all SoCs in the 0.75 M TBAPF ₆ /acetonitrile mixture. Charging was performed with a constant current of 50 mA/cm ² , followed by a constant voltage at 2.95 V with a cutoff threshold of 15 mA/cm ² . Discharging was performed almost fully at a constant current of -50 mA/cm ² with a very short voltaic hold at 1 V (threshold 6 mA/cm ² ). Energy efficiencies of 55 to 63% were achieved. This could be increased by charging and discharging slower. The battery’s capacity retention might be limited due to unwanted reaction of the anolyte with the catholyte (both are present in both reservoirs in a mixed flow battery) or solvent or electrolyte salt. Sterically more protected derivatives, such as molecules 7 and 8, will mitigate side reaction as well as self-discharge, which could be proven for molecule 7 in UV-vis-NIR as well as coin cell and flow battery cycling experiments. Another very promising possibility to increase the stability is alternating the solvent/electrolyte salt mixture. List of cited documents I ACS Appl. Energy Mat.2019, 2, 2364 II RSC Adv.2019, 9, 13128 III Int. J. Electrochem. Sci.2018, 13, 6676 IV Int. J. Hydrogen Energy 2017, 42, 17488 V J. Power Sources 2020, 445, 227330 VI ChemElectroChem 2022, 9, e202200483 VII ACS Energy Lett.2017, 2, 1156 VIII J. Am. Chem. Soc.2019, 141, 15301 IX Angew. Chem.2019, 131, 7119 X ACS Energy Lett.2016, 1, 705 XI ACS Appl. Energy Mater.2021, 4, 9248 XII Nat. Commun.2020, 11, 3843 XIII Angew. Chem. Int. Ed.2015, 54, 8684