Redox-Active Guanidines in Proton-Coupled Electron-Transfer Reactions: Real Alternatives to Benzoquinones?

Guanidino-functionalized aromatics (GFAs) are readily available, stable organic redox-active compounds. In this work we apply one particular GFA compound, 1,2,4,5-tetrakis(tetramethylguanidino)benzene, in its oxidized form in a variety of oxidation/oxidative coupling reactions to demonstrate the scope of its proton-coupled electron transfer (PCET) reactivity. Addition of an excess of acid boosts its oxidation power, enabling the oxidative coupling of substrates with redox potentials of at least +0.77 V vs. Fc+ /Fc. The green recyclability by catalytic re-oxidation with dioxygen is also shown. Finally, a direct comparison indicates that GFAs are real alternatives to toxic halo- or cyano-substituted benzoquinones.

Proton‐coupled electron transfer (PCET) is important for biological and bioinspired (photosynthetic) processes as well as synthetic chemistry,1, 2, 3 and has been studied intensively mechanistically.4, 5 Quinones are especially versatile organic PCET reagents. Their redox‐properties and the pK a values of their corresponding hydroquinones can be varied by the introduction of substituents,6, 7, 8 and also by electronic excitation.9, 10 Figure 1 shows as examples the three benzoquinones BQ, CA and DDQ. The 1 e− redox potentials (E red vs. Fc+/Fc) of DDQ (+0.14 V) and CA (−0.35 V) are significantly higher than that of BQ (−0.88 V).8 However, the reduction of the pK a value of the reduced form that accompanies the increase of E red of the oxidized form leads to a certain “leveling” effect on the PCET reactivity.7, 8 Benzoquinones with a relatively high oxidation potential, for example, CA and DDQ, are used in a number of PCET reactions as stoichiometric oxidation reagents, often in combination with a strong acid,11, 12, 13 and in some reactions also catalytically, for example, DDQ together with nitrite.14 Low‐potential p‐benzoquinone derivatives were used as redox‐mediators in biomimetic catalysis and as redox catalysts (often together with transition metal complexes).15, 16, 17 Moreover, low‐potential o‐quinone‐type catalysts were recently shown to enable manifold (bioinspired) aerobic oxidations.18, 19, 20, 21, 22, 23, 24, 25, 26

Figure 1

Lewis structures of p‐benzoquinone (BQ), chloranil (CA), and 2,3‐dichloro‐5,6‐dicyano‐benzoquinone (DDQ).

Despite of the outstanding success story of quinones, some drawbacks oppose their large‐scale applications. Hence cyano‐ or halo‐substituted benzoquinones like CA are highly toxic, as they induce reactive oxygen species and oxidative stress, showing an inflammatory response both in vivo and in vitro.27 Moreover, the recycling of the quinones is sometimes problematic due to side reactions.11 Strong oxidizing reagents are required for high‐potential quinones. For low‐potential quinones catalytic oxidation of the hydroquinone with dioxygen is possible,28 but could be hampered by the formation of quinhydrones (the 1:1 complex between benzoquinone and hydroquinone exhibits a binding energy of more than 20 kJ mol−1)29 at high concentrations.28

We recently developed a new class of PCET reagents, namely redox‐active guanidines, that do not have these disadvantages.30, 31, 32, 33 One example is 1,2,4,5‐tetrakis‐(tetramethylguanidino)benzene (1), which could be readily oxidized to the dication 1 2+ (Scheme 1). The loss of aromaticity upon oxidation leads to significantly different C−C bond distances in the C6 ring and a distinct colour change from pale yellow for neutral 1 to intense green for the dication 1 2+. Proton‐coupled electron transfer (PCET) reactions of oxidized 1,2,4,5‐tetrakis(tetramethylguanidino)benzene (1 2+, Scheme 1)34 with some inorganic (thiol to disulfides, phosphines to diphosphines)31 and organic substrates with relatively low redox‐potential (e.g., phenols to biphenols, catechols to benzoquinones) were already reported.31, 32 The π‐system of 1 2+ accepts the electrons and the nitrogen lone pairs accept the protons. Using a copper co‐catalyst, 1 2+ could be used as an organocatalyst with dioxygen as the terminal oxidant.32 The pK a value of ca. 25.3 for (1+H)+ in CH3CN sharply decreases upon oxidation. Interestingly, green 1 2+ still is a Lewis35 and Brønsted base, and is protonated with HBF4 ⋅Et2O to blue (1+H)3+ and orange (1+2 H)4+ (with a pK a value of ca. 13 in CH3CN close to CF3COOH, Scheme 1).36 The reduction potential increases from 1 2+ (E 1/2=−0.73 V vs. Fc+/Fc in CH3CN) to (1+2H)4+ by ca. 0.7 V.36

Scheme 1

Lewis structures of characterized, stable states, starting with neutral 1,2,4,5‐tetrakis(tetramethylguanidino)‐benzene (1), relevant for the PCET reactivity. The colors characteristic for the oxidized states are highlighted.

Herein, we demonstrate the wide scope of its PCET reactivity, especially in combination with strong acids. Nine oxidative coupling/ oxidation reactions were studied with substrates that differ largely in their redox potentials. In addition, we show that efficient regeneration of the PCET reagent 1(BF4)2 is possible by catalytic oxidation with dioxygen. Finally, we compare its PCET properties with benzoquinones.

The salts 1(BF4)2 (oxidized GFA) and (1+2 H)(BF4)4 (oxidized and protonated GFA)36 as well as 1(BF4)2 in combination with an excess of the strong acid HBF4 ⋅OEt2 were applied. The substrates are grouped in low‐potential [Eqs. (1–4)] and high potential substrates [Eqs. (5–9)], for example, +0.52 V for NPh3 (S),37 +0.66 V for 4,4′‐dibromo‐triphenylamine (S), +0.74 V for 3,3′′,4,4′′‐tetramethoxy‐o‐terphenyl (S),13b and +0.77 V vs. Fc+/Fc for 4‐nitro‐triphenylamine (S).38 To allow for a direct comparison, all reactions were carried out in CH3CN solution. The yields were estimated from NMR signal integration (see the Supporting Information).

Oxidative coupling of 2,6‐di‐tert‐butyl‐phenol (S) to the diketone (P) gives best results (82 % yield) with 1(BF4)2 [Eq. (1)] rather than (1+2 H)(BF4)4. Oxidation of 3,5‐di‐tert‐butyl‐catechol (S) to the o‐ benzoquinone (P) is fast with 1(BF4)2, but gives slightly better yields with (1+2 H)(BF4)4 [Eq. (2)]. Both reactions are presumably initiated by deprotonation. Catechol deprotonation is favored by the intramolecular O−H⋅⋅⋅O bridge in the resulting monoanion.7, 39

The oxidative coupling of benzylamine (S) to give N‐(phenylmethylene)benzenemethanamine (P) and the oxidation of o‐phenylene‐diamine (S) to give 2,3‐diaminophenazine (P) give best results with (1+2 H)(BF4)4 (79 % respectively 96 % yield, see Eqs. (3) and (4)]. With 1(BF4)2, these reactions are much slower, giving less than 10 % yield after 25 h at 60 °C (see the Supporting Information for details). UV/Vis and 1H NMR spectra (see the Supporting Information) indicate that (1+2 H)4+ first protonates the amine, in line with the acidity of (1+2 H)4+.36

When benzylamine oxidation was repeated with 0.5 equivalents of 1(BF4)2 and slightly more than 1 equivalent of NH4PF6, the reaction proceeded with a similar rate and slightly better yield (84 %).

Next we inspected the reactivity toward substrates with higher redox potentials, requiring the addition of excess acid. Happily, oxidative coupling of triphenylamine and derivatives with electron‐withdrawing and ‐donating groups [S–S, Eq. (5)–(8)] is accomplished in less than 2.3 h with excellent yields with a combination of 1(BF4)2 (equimolar amount) and excess HBF4⋅OEt2 (see the Supporting Information). The results demonstrate the superior functional‐group tolerance of such coupling reactions. UV/Vis experiments showed the presence of reaction intermediates, arising from substrate oxidation (see the Supporting Information).12h, 40, 41 In principle, the triphenylamine derivatives could be protonated by the strong acid. However, in all cases fast oxidation was observed, indicating that protonation plays no significant role.

Finally, we tested an intramolecular oxidative coupling reaction of 3,3′′,4,4′′‐tetramethoxy‐o‐terphenyl (S) [Eq. (9)], a substrate with a high oxidation potential of 0.74 V vs. Fc+/Fc. Application of 1 equivalent of 1(BF4) with an excess of HBF4⋅OEt2 leads to 99 % triphenylene coupling product.

Obviously, (1+2 H)(BF4)4 forms from 1(BF4)2 upon acid addition, but the oxidation potential of these organic substrates is still higher than the reduction potential of (1+2 H)(BF4)4.36 Hence the addition of excess acid boosts the oxidation power, as also found for benzoquinones in aqueous42, 43 and aprotic solutions.12b, 12h, 12k, 44, 45

The guanidinium salt could easily be separated from the reaction mixture. We already showed that (1+2 H)2+ can be quantitatively reconverted to 1 2+ by catalytic oxidation with dioxygen.32 In new experiments we tested the recyclability of 1(BF4)2 from the reduced tetraprotonated compound (1+4 H)(BF4)4, that is formed in the experiments with (1+2 H)(BF4)4 or an excess of acid. Indeed, quantitative formation of 1(BF4)2 (NMR studies, see the Supporting Information) within 30 min at 60 °C was achieved by catalytic oxidation with dioxygen in the presence of 2 equivalents of NEt3 (Scheme 2) with a simple, commercially available catalyst (3 mol % of a 1:1 mixture of CuCl2/[Cu(H2O)6](BF4)2), independent of the concentration of (1+4 H)(BF4)4 (62, 17, and 8 mmol L−1). A complex formation between the product (1 2+) and the reactant [(1+2 H)2+ or even (1+4 H)4+], as observed in the case of benzoquinone (quinhydrone complex), is prohibited by strong electrostatic repulsion.

Scheme 2

Regeneration of 1(BF4)2 from the reduced and two‐ or fourfold protonated forms (cat.=[CuCl2/Cu(H2O)6(BF4)2]).

The reaction between dihydro‐benzoquinone and 1 2+ in CH3CN leads quantitatively in 35 min at r.t. to BQ,31 showing that 1 2+ is a stronger PCET reagent than BQ. To gain more information, we calculated the energetics of the reactions in Table 1 by using the B3LYP functional in combination with a def2‐SV(P) or def2‐TZVP basis set. The solvent effect was estimated with the conductor‐like screening model (COSMO) at a relative permittivity ϵ r of 40. Calculations with and without BF4 − counter‐ions gave similar results (see the Supporting Information); we here present results with BF4 −. According to these calculations, 1(BF4)2 is similar to BQ with respect to the thermodynamics of its PCET reactions, and slightly weaker than CA. On the other hand, (1+2 H)(BF4)4 is a significantly stronger PCET reagent than all three quinones BQ, CA and DDQ.

Table 1

Reaction energies, enthalpies (at 0 K) and Gibbs free energies (at 298 K) for the reaction between the benzoquinones BQ, CA or DDQ and 1(BF4)2 respectively (1+2 H)(BF4)4 from B3LYP+COSMO/def2‐TZVP calculations at ϵ r=1 and 40.

	X	X′	ΔE [kJ mol−1]		ΔH (0 K) [kJ mol−1]	ΔG (298 K) [kJ mol−1]
			ϵ r=1	ϵ r=40	ϵ r=1	ϵ r=1
BQ	H	H	−1.8	9.9	1.5	6.0
CA	Cl	Cl	9.9	12.9	12.7	15.2
DDQ	Cl	CN	45.2	59.8	47.3	48.2

−82.3

−67.4

−76.2

−71.0

−70.6

−64.3

−65.0

−61.8

−35.3

−17.5

−30.4

−28.8

The effect of hydrogen‐bonding and protonation on the redox‐potential of quinones in aqueous and aprotic solutions41, 46 was already studied. Moreover, estimates for the pK a value of protonated BQ were reported (e.g., from Pourbaix diagrams).47, 48 On the other hand, monoprotonation of BQ in significant amounts requires the use of superacidic HF/AsF5 and low temperature, since the salt (BQ+H)AsF6 decomposes already above −60 °C.49 By contrast, (1+2 H)(BF4)4 is a storable compound, being stable in the solid state and in solution at ambient conditions.36 Consequently, the double‐proton transfer from (1+2 H)(BF4)4 to BQ to give 1(BF4)2 and (BQ+2 H)(BF4)2 (exhibiting almost symmetric F⋅⋅⋅H−O bonds between cation and anion, with F−H: 1.360 Å and O−H: 1.059 Å) was calculated (B3LYP+COSMO/def2‐TZVP) to be associated with a high positive reaction energy of +251 kJ mol−1 at ϵ r=40. Accordingly, no reaction was observed when (1+2 H)(BF4)4 was dissolved together with BQ in CH3CN. Interestingly, (BQ+2 H)(BF4)2 decomposes in the calculations for ϵ r=1 by fluoride abstraction from the anion to a complex BQ(HF)2(BF3)2 (see the Supporting Information). Moreover, (CA+2 H)(BF4)2 defines no minimum structure at both ϵ r=1 and 40, but converges again to the product of fluoride abstraction, CA(HF)2(BF3)2 (Figure 2). The reaction between (1+2 H)(BF4)4 and CA to give, instead of protonated CA, the favoured complex CA(HF)2(BF3)2 exhibits a reaction energy of +317 kJ mol−1 at ϵ r=40.

Figure 2

a) Comparison between the experimentally derived structure of the stable compound (1+2 H)(BF4)4 in the solid state (a) and the structure of the CA(HF)2(BF3)2 complex obtained in the attempt to calculate the analogue two‐fold‐protonated CA with two BF4 − counter‐ions (b).

In summary we demonstrated the preeminent PCET reactivity and efficient recyclability (by green oxidation of (1+2 H)(BF4)2 or (1+4 H)(BF4)4 with dioxygen) of the tetrakisguanidine 1(BF4)2. This PCET reagent is readily synthesized (in two steps starting from commercially available 1,2,4,5‐tetraaminobenzene‐tetrahydrochloride), easy to handle, and thermally stable.30, 34, 36, 50 The results show that the combination of 1(BF4)2 with a strong acid allows the fast and near quantitative oxidative coupling of substrates with high redox potentials (at least +0.77 V vs. Fc+/Fc) at mild conditions, making the compound a real alternative to traditionally applied toxic benzoquinone derivatives.

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

Click here for additional data file.