A Stable Hexakis(guanidino)benzene: Realization of the Strongest Neutral Organic Four-Electron Donor

Article abstract of DOI:10.1002/anie.201611189

The growing demand for efficient batteries has stimulated the search for redox-active organic compounds with multistage redox behavior, as materials with large charge capacity. Herein we report the synthesis and properties of the first hexakis(guanidino)benzene derivative: a strong neutral organic electron donor with reversible multistage redox behavior and a record low redox potential for donation of four electrons. Detailed structural and spectroscopic characterization of three redox states (0, +2, and +4) reveal its unique electronic features. Despite its nitrogen richness, the compound is thermally robust and can be readily purified by sublimation.

Full text of DOI:10.1002/anie.201611189

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201611189
German Edition: DOI: 10.1002/ange.201611189
Electron Donors
A Stable Hexakis(guanidino)benzene: Realization of the Strongest
Neutral Organic Four-Electron Donor
Benjamin Eberle, Elisabeth Kaifer, and Hans-Jçrg Himmel*
Abstract: The growing demand for efficient batteries has
stimulated the search for redox-active organic compounds with
multistage redox behavior, as materials with large charge
capacity. Herein we report the synthesis and properties of the
first hexakis(guanidino)benzene derivative: a strong neutral
organic electron donor with reversible multistage redox
behavior and a record low redox potential for donation of
four electrons. Detailed structural and spectroscopic character-
ization of three redox states (0, + 2, and + 4) reveal its unique
electronic features. Despite its nitrogen richness, the compound
is thermally robust and can be readily purified by sublimation.
T
here is an increasing demand for new batteries that are
superior to Li-ion batteries in terms of safety and cost.
Organic materials are very attractive for such purposes but
generally suffer from low charge capacity.[1–3] Hence, strat-
egies for the synthesis of organic compounds that display
multistage redox behavior need to be developed. In partic-
ular, open-shell molecules with degenerate frontier orbitals
have been synthesized in recent years as multistage redox-
active compounds.[4–7] Such charge-storage materials are often
electron acceptors and therefore generally exhibit negative
charge regimes. In this work, we follow the opposite strategy:
we exploit the strong capability of guanidino groups to
stabilize positive charges (as demonstrated by guanidino-
functionalized aromatic compounds (GFAs), a class of strong
organic electron donors[8,9]). To achieve multistage redox
behavior, we attempted to attach the maximum number of
guanidino groups to an aromatic compound, and conse-
quently achieved the synthesis of the first hexakis(guanidi-
no)benzene derivative. Hexaaminobenzene[10] was chosen as
direct precursor, since guanidinylation of amino-substituted
aromatic compounds via a formamidinium or imidazolium
salt (“Vilsmeier salt”) is a generally reliable route to GFAs.
Nevertheless, reaction of hexaaminobenzene with chloro-
N,N,N’,N’-tetramethylformamidinium chloride leads to
a product mixture, from which 2 was isolated as the major
product (Scheme 1 and Figure S1). It is formed in a double
cyclization reaction between a guanidino group and an ortho
amino group.[11] Apparently such reactions are favored with
an increasing number of amino groups, explaining why even
a large excess of the Vilsmeier salt does not suppress this
Scheme 1. Reactions leading to 2 and to the first hexakis(guanidino)-
benzene 1b.
pathway. Fortunately, the problem could be circumvented by
the use of imidazolium salts, since the embedding of the two
dialkylamino groups in a cyclic system hampers their elimi-
nation. Indeed, reaction of hexaaminobenzene with 2-chloro-
1,3-dimethyl-4,5-dihydro-1H-imidazolium chloride led to the
first hexakis(guanidino)benzene, 1b (Scheme 1).
Compound 1b is thermally stable and can be sublimed
without decomposition under vacuum at 2508C, yielding the
analytically pure substance. The colorless powder is
extremely oxygen-sensitive; solutions of 1b rapidly turn
yellow under air. On the other hand, it can be stored for
several months in solid form under an inert-gas atmosphere.
In the IR spectrum, an intense band at 1664 cmÀ1 is assigned
1
=
to the n(C N) stretching modes (Figure S2). In the H NMR
spectrum, two signals appear at d = 2.61 and 3.06 ppm due to
the methyl and methylene protons, respectively (Figure S3).
The new hexakis(guanidino)benzene crystallizes from CH2Cl2
solution, and Figure 1 illustrates its solid-state structure. As
already observed for other GFAs, the planes of the NCN2
groups of each guanidino group are highly tilted relative to
the plane of the aromatic C6 ring (see explanation in the
literature[12]). The guanidino N C bonds measure 1.268(2)/
=
1.277(2)/1.274(2) ꢀ.
The cyclic voltammetry (CV) curve (Figure 2 and Fig-
ure S4) of 1b in CH3CN solution shows two fully reversible
two-electron redox waves at E1/2 = À0.96 V (Eox = À0.93 V)
and À0.43 V (Eox = À0.39 V) vs. Fc/Fc+, and a one-electron
redox wave at E1/2 =+ 0.82 V (Eox =+ 0.88 V). Apparently 1b
is a strong electron donor, although the potential for the first
two-electron oxidation (1b/1b2+) is not exceptional for
organic electron donors.[13–18] In contrast, the second wave at
[*] B. Eberle, Dr. E. Kaifer, Prof. Dr. H.-J. Himmel
Anorganisch-Chemisches Institut
Ruprecht-Karls-Universitꢀt Heidelberg
Im Neuenheimer Feld 270, 69120 Heidelberg (Germany)
E-mail: hans-jorg.himmel@aci.uni-heidelberg.de
Supporting information for this article can be found under:
http://dx.doi.org/10.1002/anie.201611189.
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
Figure 1. Structure of 1b. Vibrational ellipsoids drawn at the 50%
probability level. Hydrogens omitted for clarity. Selected bond lengths
[ꢁ]: N1–C1 1.421(2), N1–C4 1.268(2), N2–C4 1.403(2), N3–C4 1.386-
(2), N4–C2 1.419(2), N4–C9 1.277(2), N5–C9 1.390(2), N6–C9 1.392-
(2), N7–C3 1.415(2), N7–C14 1.274(2), N8–C14 1.398(2), N9–C14
1.399(2), C1–C2 1.403(3), C1–C3 1.402(3), C2–C3’ 1.406(3).
Figure 3. a) Comparison between the experimental (black curve)
UV/Vis spectra recorded at À1.35 V, À0.75 V and +1.05 V vs. Fc/Fc+
and the simulated spectra (based on TD-DFT (B3LYP-D3/TZVP)
computations, red curve). b) Photo of CD3CN solutions of 1b after
addition of 2, 3, and 4 equiv of NO(SbF6).
Figure 3a displays the UV/Vis spectra obtained in spec-
troelectrochemical studies and recorded at potentials that
favor the neutral, di- and tetracationic species (Figure S5 for
all spectra). With increasing charge, the lowest-energy
electronic transition experiences a strong bathochromic
shift. The spectrum of 1b2+ exhibits a band at 472 nm, and
that of 1b4+ a strong band at 604 nm, with a shoulder at
553 nm. At very high potentials (1.35 V), the intense band at
604 nm of 1b4+ decreases, and two weak features show at 407
and 771 nm, that are tentatively assigned to the pentacation
1bC5+ (Figures S5 and S10).
Figure 2. CV curve of 1b in CH3CN (nBu4N(PF6) as supporting electro-
lyte, 100 mVsÀ1, ferrocene (Fc) as external standard). The curves for
4b in CH3CN and 3a in CH2Cl2 are also shown.
À0.43 V for the redox step 1b2+/1b4+ is indeed a record low
value for a neutral molecular organic compound, proving the
unique four-electron donor ability of 1b. From the potential
difference between the 1b/1b2+ and 1b2+/1b4+ redox steps,
a DG value of 102 kJmolÀ1 is estimated for the disproportio-
nation of 1b2+ into 1b and 1b4+. The reversible one-electron
step at E1/2 =+ 0.82 V is tentatively assigned to the 1b4+/1bC5+
redox step, thus indicating that 1b is a molecular organic five-
electron donor.[19]
Figure 2 shows that the redox potential decreases with an
increasing number of guanidino groups (from E1/2 = À0.22 V
for 3a/3a2+[12] to À0.76 V for 4b/4b2+[20] and finally to À0.96 V
for 1b/1b2+, Table S1). Other GFA compounds, for which
reversible four-electron oxidation is feasible, are 1,4,5,8-
tetrakis(tetramethylguanidino)naphthalene (two two-elec-
tron waves were found at E1/2 = À0.65 V and + 0.17 V in
CH2Cl2 vs. Fc/Fc+) and two different hexakis(guanidine)tri-
phenylenes.[21] However, 1b is a much stronger four-electron
donor.
Addition of NO(SbF6) to CH3CN solutions of 1b leads to
its chemical oxidation (see Figure S6, for experiments with
Fc(PF6)). The photo of NMR tubes containing 1b2+ (2 equiv
of NO+ added), 1b2+/1b4+ (3 equiv of NO+ added), and 1b4+
(4 equiv of NO+ added) is displayed in Figure 3b. In line with
the UV/Vis spectra in Figure 3a, solutions of 1b2+ are yellow
and solutions of 1b4+ are intensely blue colored. The green
color of solutions with 3 equiv of NO+ is due to the
superposition of the colors of 1b2+ and 1b4+. Oxidation to
1
1b2+ leads to downfield shifts of the H NMR signals. For
1b4+, two signals at d = 2.74 (methyl protons) and 3.76 ppm
1
(methylene protons) appear in the H NMR spectrum (Fig-
ure S7). In the 13C NMR spectrum, also two signals at d =
33.24 and 49.33 ppm are detected. After addition of 8 equiv of
the oxidant, the NMR spectrum shows weak and broad
signals, indicating the formation of a paramagnetic species
(Figure S8a). The EPR spectrum of this solution contains
2
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
a single signal at g = 2.003 (Figure S9), which prevails in the
solid material after solvent removal, and is assigned to 1bC5+.
The dication 1b2+ and the tetracation 1b4+ were structur-
ally characterized (Figure 4). Crystals of (1b)Cl2 were
obtained during the synthesis of 1b and presumably arise
from traces of O2; crystalline 1b(SbF6)4 precipitated from
a CD3CN solution of 1b and 4 equiv of NO(SbF6). The
dication 1b2+ exhibits a p-benzoquinone-type structure in the
solid state. Two guanidino groups participate directly in
charge delocalization, consequently showing a shortening of
À
the bond to the C6 ring (N4 C2) to 1.297(2) ꢀ and an
À
elongation of the former imino bond (N4 C9) to 1.366(2) ꢀ.
The bond lengths in the other four guanidino groups are
À
closer to the values in neutral 1b. In the C6 ring, two C C
bonds are shortened (1.374(2) ꢀ), indicating loss of aroma-
ticity. The tetracation 1b4+ has a more delocalized bonding
situation, and all six guanidino groups participate in the
À
charge dispersion (see SI, Table S2). With 1.475(10) (C1 C2),
Figure 4. a) Structure of (1b)Cl2. Vibrational ellipsoids drawn at the
À
À
À
1.447(9) (C1 C3), and 1.463(10) ꢀ (C2 C3’), all C C bonds
in the central C6 ring are similar and elongated with respect to
neutral 1b. The relative orientation of the guanidino groups in
1b4+ leads to a paddle-wheel-type conformer that differs from
the structures of 1b and 1b2+.
50% probability level. Hydrogens omitted for clarity. Selected bond
lengths [ꢁ]: N1–C1 1.382(2), N1–C4 1.302(2), N2–C4 1.362(2), N3–C4
1.373(2), N4–C2 1.297(2), N4–C9 1.366(2), N5–C9 1.326(2), N6–C9
1.332(2), N7–C3 1.386(2), N7–C14 1.310(2), N8–C14 1.357(2), N9–
C14 1.365(2), C1–C2 1.470(2), C1–C3 1.374(2), C2–C3’ 1.470(2).
b) Structure of 1b(SbF6)4. In this case, a ball-and-stick representation
is chosen for clarity; again, all hydrogens are omitted. Selected bond
lengths [ꢁ]: N1–C1 1.300(8), N1–C4 1.383(9), N2–C4 1.327(9), N3–C4
1.325(8), N4–C2 1.298(8), N4–C9 1.384(9), N5–C9 1.328(8), N6–C9
1.338(9), N7–C3 1.304(9), N7–C14 1.384(9), N8–C14 1.334(8), N9–
C14 1.323(9), C1–C2 1.475(10), C1–C3 1.447(9), C2–C3’ 1.463(10). The
corresponding Lewis representations of 1b2+ and 1b4+ are shown on
the right.
To complete our analysis, we carried out quantum
chemical computations (B3LYP-D3/TZVP). To restrict the
computational effort, only two conformers are considered
herein, namely one with the guanidino groupsꢁ orientation
akin to that in the solid-state structures of 1b and (1b)Cl2, and
a paddle-wheel-type conformer as found in solid 1b(SbF6)4. In
line with the solid-state structures, the paddle-wheel confor-
mer is more stable for 1b4+, while the other conformer is
favored for 1b and 1b2+ (see Tables S3 and S4, and Scheme S2
for a natural population analysis). In Figure 3a, the simulated
electronic spectra (based on TD-DFT computations) are
compared to the experimental spectra in solution (see also
Figure S10). The computations confirm the bathochromic
trend with increasing charge (309 nm for 1b, 526 nm for 1b2+,
and 548 nm for 1b4+).
The new compound 1b exhibits a high (theoretical)
charge capacity of 144 AhkgÀ1 (storage of four electrons)
Scheme 2. Comparison between charge regimes (electron acceptor vs. electron donor) and theoretical charge capacities of R3TOT and 1b. Relative
orbital energies are DFT estimates (B3LYP-D3/TZVP computations for 1b4+; ROB3LYP/UB3LYP/6-31G(d,p)[4] for R3TOT).
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
and even 180 AhkÀ1 (for release of five electrons), making it
attractive as a possible alternative for Li-ion battery technol-
ogy (ca. 150–170AhkÀ1).[4] Importantly, its charge regime is
opposite to that of other prominent molecules with high
charge capacity (Scheme 2), in particular the radicals 6-
oxophenalenoxyl (6OPO), tribrominated trioxotriangulene
Br3TOT (192 AhkÀ1), and tri-tert-butylated trioxotriangulene
(tBu)3TOT (219 AhkÀ1).[22] Scheme 2 illustrates the similarity
of the frontier molecular orbitals (in the case of 1bn+, these
are composed of benzene p-orbitals with some guanidino
contributions, Figure S11). For the radical pentacation 1bC5+
as well as the neutral radicals R3TOT, these three orbitals are
occupied by a single electron. Although the relative orbital
energies are only DFT estimates, 1b-based materials are
expected to give a much greater output voltage than the
R3TOT-based materials, which might result in a lower self-
discharge rate and more power or less current, respectively.
Because of its high thermal stability and its ready sublimation
at a convenient temperature, we expect compound 1b to show
promising processability. The application of 1b as a battery
material is currently under further evaluation in our labo-
ratory.
[4] Y. Morita, S. Nishida, T. Murata, M. Moriguchi, A. Ueda, M.
Satoh, K. Arifuka, K. Sato, T. Takui, Nat. Mater. 2011, 10, 947 –
951.
[5] S. Nishida, Y. Yamamoto, T. Takui, Y. Morita, ChemSusChem
2013, 6, 794 – 797.
[6] Y. Morita, S. Nishida, J. Kawai, K. Fukui, S. Nakazawa, K. Sato,
D. Shiomi, T. Takui, K. Nakasuji, Org. Lett. 2002, 4, 1985 – 1988.
[7] S. Nishida, Y. Morita in Organic Redox Systems: Synthesis
Properties, and Applications, (Ed.: T. Nishinaga), Wiley, Hobo-
ken, 2015, pp. 177 – 244.
[8] B. Eberle, O. Hꢂbner, A. Ziesak, E. Kaifer, H.-J. Himmel, Chem.
Eur. J. 2015, 21, 8578 – 8590.
[9] H.-J. Himmel, Z. Anorg. Allg. Chem. 2013, 639, 1940 – 1952.
[10] a) B. Flꢂrscheim, E. L. Holmes, J. Chem. Soc. 1929, 330 – 337;
b) I. E. Knaggs, Proc. R. Soc. London Ser. A 1931, 131, 612 – 620;
c) B. Kohne, K. Praefcke, Liebigs Ann. Chem. 1987, 265 – 265;
d) D. A. Dixon, J. C. Calabrese, J. S. Miller, Angew. Chem. Int.
Ed. Engl. 1989, 28, 90 – 92; Angew. Chem. 1989, 101, 79 – 81;
e) G. Khisamutdinov, V. Korolev, I. Kondyukov, I. Abdrakhma-
nov, S. Smirnov, A. Fainzilberg, V. Dorokhov, Russ. Chem. Bull.
1993, 42, 136 – 138.
[11] A. Peters, H. Herrmann, M. Magg, E. Kaifer, H.-J. Himmel, Eur.
J. Inorg. Chem. 2012, 1620 – 1631.
[12] J. Hornung, O. Hꢂbner, H.-J. Himmel, RSC Adv. 2016, 6, 39323 –
39329.
[13] J. Broggi, T. Terme, P. Vanelle, Angew. Chem. Int. Ed. 2014, 53,
384 – 413; Angew. Chem. 2014, 126, 392 – 423.
[14] E. Doni, J. A. Murphy, Chem. Commun. 2014, 50, 6073 – 6087.
[15] J. L. Segura, N. Martin, Angew. Chem. Int. Ed. 2001, 40, 1372 –
1409; Angew. Chem. 2001, 113, 1416 – 1455.
Acknowledgements
Financial support by the Deutsche Forschungsgemeinschaft
(DFG) and the allocation of computational resources of the
bwForCluster JUSTUS (State of Baden-Wꢂrttemberg, frame-
work program bwHPC-C5) is gratefully acknowledged. We
thank Dr. Olaf Hꢂbner, Dr. Lutz Greb, and Prof. Dr. Markus
Enders for fruitful discussions.
[16] N. Martꢃn, Chem. Commun. 2013, 49, 7025 – 7027.
[17] H. S. Farwaha, G. Bucher, J. A. Murphy, Org. Biomol. Chem.
2013, 11, 8073 – 8081.
[18] S. S. Hanson, E. Doni, K. T. Traboulsee, G. Coulthard, J. A.
Murphy, C. A. Dyker, Angew. Chem. Int. Ed. 2015, 54, 11236 –
11239; Angew. Chem. 2015, 127, 11388 – 11391.
[19] a) E. M. Pꢄrez, M. Sierra, L. Sꢅnchez, M. R. Torres, R. Viruela,
P. M. Viruela, E. Ortꢃ, N. Martꢃna, Angew. Chem. Int. Ed. 2007,
46, 1847 – 1851; Angew. Chem. 2007, 119, 1879 – 1883; b) M.
Gallego, J. Calbo, J. Aragꢆ, R. M. K. Calderon, F. H. Liquido, T.
Iwamoto, A. K. Greene, E. A. Jackson, E. M. Pꢄrez, E. Ortꢃ,
D. M. Guldi, L. T. Scott, N. Martꢃn, Angew. Chem. Int. Ed. 2014,
53, 2170 – 2175; Angew. Chem. 2014, 126, 2202 – 2207.
[20] D. Emeljanenko, A. Peters, V. Vitske, E. Kaifer, H.-J. Himmel,
Eur. J. Inorg. Chem. 2010, 4783 – 4789.
Conflict of interest
The authors declare no conflict of interest.
Keywords: batteries · guanidine · organic electron donors ·
oxidation · redox-active molecules
[21] a) V. Vitske, C. Kçnig, E. Kaifer, O. Hꢂbner, H.-J. Himmel, Eur.
J. Inorg. Chem. 2010, 115 – 126; b) V. Vitske, P. Roquette, S.
Leingang, C. Adam, E. Kaifer, H. Wadepohl, H.-J. Himmel, Eur.
J. Inorg. Chem. 2011, 1593 – 1604; c) A. Lebkꢂcher, C. Wagner,
O. Hꢂbner, E. Kaifer, H.-J. Himmel, Inorg. Chem. 2014, 53,
9876 – 9896.
[22] Y. Morita, T. Ohba, N. Haneda, S. Maki, J. Kawai, K. Hatanaka,
K. Sato, D. Shiomi, T. Takui, K. Nakasuji, J. Am. Chem. Soc.
2000, 122, 4825 – 4826.
[1] M. Miroshnikov, K. P. Divya, G. Babu, A. Meiyazhagan,
L. M. R. Arava, P. M. Ajayan, G. John, J. Mater. Chem. A
2016, 4, 12370 – 12386.
[2] T. Janoschka, N. Martin, M. D. Hager, U. S. Schubert, Angew.
Chem. Int. Ed. 2017, 56, 686 – 711; Angew. Chem. 2017, 129, 702 –
729.
[3] S. Muench, A. Wild, C. Friebe, B. Haeupler, T. Janoschka, U. S.
Schubert, Chem. Rev. 2016, 116, 9438 – 9484.
Manuscript received: November 17, 2016
Final Article published: && &&, &&&&
4
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Electron Donors
Multistage electron donor: The first
hexakis(guanidino)benzene derivative
donates four electrons at a record low
potential.
B. Eberle, E. Kaifer,
H.-J. Himmel*
&&&& — &&&&
A Stable Hexakis(guanidino)benzene:
Realization of the Strongest Neutral
Organic Four-Electron Donor
Angew. Chem. Int. Ed. 2017, 56, 1 – 5
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!