Isolation of a hydrogen-bonded complex based on the anthranol/anthroxyl pair: Formation of a hydrogen-atom self-exchange system

Article abstract of DOI:10.1002/anie.201410796

A hydrogen-bonded complex was successfully isolated as crystals from the anthranol/anthroxyl pair in the self-exchange proton-coupled electron transfer (PCET) reaction. The anthroxyl radical was stabilized by the introduction of a 9-anthryl group at the carbon atom at the 10-position. The hydrogen-bonded complex with anthranol self-assembled by π-π stacking to form a one-dimensional chain in the crystal. The conformation around the hydrogen bond was similar to that of the theoretically predicted PCET activated complex of the phenol/phenoxyl pair. X-ray crystal analyses revealed the self-exchange of a hydrogen atom via the hydrogen bond, indicating the activation of the self-exchange PCET reaction between anthranol and anthroxyl. Magnetic measurements revealed that magnetic ordering inside the one-dimensional chain caused the inactivation of the self-exchange reaction.

Full text of DOI:10.1002/anie.201410796

.
Angewandte
Communications
DOI: 10.1002/anie.201410796
Stable Organic Radicals
Isolation of a Hydrogen-Bonded Complex Based on the Anthranol/
Anthroxyl Pair: Formation of a Hydrogen-Atom Self-Exchange
System**
Yasukazu Hirao,* Tohru Saito, Hiroyuki Kurata, and Takashi Kubo*
Abstract: A hydrogen-bonded complex was successfully
isolated as crystals from the anthranol/anthroxyl pair in the
self-exchange proton-coupled electron transfer (PCET) reac-
tion. The anthroxyl radical was stabilized by the introduction
of a 9-anthryl group at the carbon atom at the 10-position. The
hydrogen-bonded complex with anthranol self-assembled by
p–p stacking to form a one-dimensional chain in the crystal.
The conformation around the hydrogen bond was similar to
that of the theoretically predicted PCET activated complex of
the phenol/phenoxyl pair. X-ray crystal analyses revealed the
self-exchange of a hydrogen atom via the hydrogen bond,
indicating the activation of the self-exchange PCET reaction
between anthranol and anthroxyl. Magnetic measurements
revealed that magnetic ordering inside the one-dimensional
chain caused the inactivation of the self-exchange reaction.
Herein, we investigated the self-exchange system of
PCET reactions between phenol and phenoxyl. Theoretical
studies on this self-exchange reaction have been reported.[7,8]
These studies revealed the possibility of the presence of
a concerted reaction of the proton and electron transfer based
on quantum calculations and reaction energy discussions. At
the theoretically predicted transition state, phenol and
phenoxyl form a head-to-head hydrogen bond. The orbital
analyses of this self-exchange system revealed the transfer
pathway of the proton and electron. Proton transfer is
expected to occur through the overlap of the lone pair of
electrons on the oxygen atoms, whereas an electron transfers
through the overlap of p-orbitals distributed on oxygen
atoms. Both overlaps are found in the hydrogen bond. Based
on these studies, we envision that the hydrogen bond in the
phenol/phenoxyl pair can be regarded as a conductive path
rather than a dielectric path. We synthesized and isolated
a PCETactivated complex in the crystalline form to verify the
conductive hydrogen bonds experimentally.
One of the largest barriers to observing the hydrogen-
bonded complex between phenol and phenoxyl is the stability
of phenoxyl. Usually, phenoxyl is stabilized by the introduc-
tion of sterically bulky groups (that is, tert-butyl) at the 2,4,6-
positions having a high spin density. However, the bulky tert-
butyl groups at the 2- and 6-positions can inhibit the
formation of the hydrogen bond. To address this, a new
stable phenoxyl derivative was designed. Anthroxyl is
thought to be a suitable radical for this purpose. Instead of
bulky groups, the extension of the p-conjugated system of the
anthracene ring of anthroxyl is expected to increase the
stability of the radical. Furthermore, the empty space around
the oxygen atom favors the formation of a rigid hydrogen
bond. We decided to introduce a bulky anthryl group at the
10-position of anthroxyl (Scheme 1) to increase the stability,
because of the relatively high spin density at that position.
The anthryl-substituted anthroxyl radical (1), which was
generated by the oxidation of bianthryl, has only been
characterized by electron spin resonance (ESR) spectrosco-
py.[9] We synthesized this anthroxyl derivative 1 by a multi-
E
lectron-transfer reactions involving proton-transfer pro-
cesses are usually discussed in terms of the proton-coupled
electron transfer (PCET) mechanism.[1] Such reactions are
widely found in chemical and biological systems. For example,
the electron transfer in photosystem II (PSII)[2] and ribonu-
cleotide reductase (RNR)[3] were reported to be driven by the
PCET mechanism. In both cases, the oxidation of phenol is
a key step.[4] Phenol releases a proton and an electron and is
converted into phenoxyl. The coupling of proton-transfer
processes to electron-transfer processes reduces the activa-
tion barrier of these reactions. The nature of PCET reactions
has attracted interest of researchers in the field of organic
conductors in the course of exploring superconducting
molecular materials. A quinhydrone complex, consisting of
para-hydroquinone and para-benzoquinone, was one of the
first examples of PCET activated molecular conductors. It
exhibited electron transfer via p–p stacking coupled with
proton transfer through the hydrogen bonds under high-
pressure conditions.[5] Recently, Mori et al. reported a similar
type of the complex based on catechol-fused[6a–c] or pyridyl-
substituted[6d] tetrathiafulvalene (TTF) derivatives. They
observed the proton–electron coupled state.
[*] Dr. Y. Hirao, T. Saito, Prof. Dr. H. Kurata, Prof. Dr. T. Kubo
Department of Chemistry, Graduate School of Science
Osaka University, Toyonaka, Osaka 560-0043 (Japan)
E-mail: y-hirao@chem.sci.osaka-u.ac.jp
[**] This work was supported in part by Scientific Research on
Innovative Areas by MEXT, “Stimuli-responsive Chemical Species
for the Creation of Functional Molecules” (No.2408).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201410796.
Scheme 1. a) The self-exchange PCET reaction between phenol and
phenoxyl pair. b) Anthryl-substituted anthroxyl 1.
2402
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 2402 –2405

Angewandte
Chemie
Figure 1. X-ray structure of the hydrogen-bonded complex between
1 and 2 at 100 K.[12] Ellipsoids are set at 50% probability.
mixing of 1 and 2 in toluene and subsequent slow evaporation
gave brown platelet crystals suitable for X-ray crystallo-
graphic analysis.[12] The shape and color of the crystals were
similar to those of 1. Single-crystal X-ray analysis at 100 K
showed that 1 and 2 were almost coplanar and faced each
other in the form of OC···HO (Figure 1; Supporting Informa-
tion, Table S1). Such a hydrogen-bonded complex in the solid
Scheme 2. Synthetic route to anthryl-substituted anthroxyl 1.
step reaction, and its reduced form anthranol (2) for the
construction of a PCET activated complex, and then isolated
the complex in the crystalline form.
The synthetic route to 1 is shown in Scheme 2. Starting
from anthrone, methylation followed by bromination pro-
vided 9-bromo-10-methoxylanthracene (4) in good yield.
Thereafter, the coupling reaction with anthrone and subse-
quent dehydroxylation afforded methoxybianthryl (5). After
demethylation with BBr3, the resulting anthranol 2 was
oxidized with potassium ferricyanide to form 9-anthryl-
substituted anthroxyl 1. The stability of 1 was found to be
very high; 1 was stable even when it was treated with a polar
protic solvent, such as alcohol or water, in air. Because of the
high stability of 1, pure 1 was obtained as brown platelet
crystals from a toluene solution. X-ray crystallographic
analysis indicated that the carbon atom at the 10-position,
with high spin density, was sandwiched from top and bottom
by the two hydrogen atoms at the peri positions of the
orthogonally substituted anthracene ring (Supporting Infor-
mation, Figure S1). This structural feature is important for
stabilizing the radical. The phenyl-substituted derivative
could not be isolated. The absence of steric repulsion allows
the phenyl ring to rotate and the protection of the spin center
is insufficient. The hyperfine splitting of the solution ESR
spectrum of 1 was mainly derived from the hydrogen atoms at
the 2,4,5,7-positions of anthroxyl (Supporting Information,
Figure S2). We did not find significant leaks of the spin
density into the anthracene substituent. The delocalized spin
distribution of the anthroxyl radical was in agreement with
that estimated from quantum calculations. The redox prop-
erty of 1 was evaluated by cyclic voltammetry in dichloro-
methane solution. The voltammogram showed two reversible
waves at À0.86 V and 0.40 V (vs. Fc/Fc+; Supporting Infor-
mation, Figure S3). The lower redox wave can be assigned to
the electron-reduction process from the radical to the anion,
and the higher wave corresponds to the oxidation process
from the radical to the cation. The detailed description of the
ESR analyses and the temperature dependent superconduct-
ing quantum interference device (SQUID) measurements is
given in the Supporting Information.
state has thus far been rarely observed.[6a,c] One of the C O
À
bond lengths was 1.340(2) ꢀ, which is a typical value for
phenols, suggesting that the hydrogen atom in the hydrogen
bond was completely localized on the alcohol. The intermo-
lecular O···O distance was determined to be 2.767(3) ꢀ, which
is close to the theoretically predicted value (2.799 ꢀ) for the
hydrogen-bonded complex of the phenol/phenoxyl pair.[10]
To observe the self-exchange reaction, the X-ray struc-
tural analysis was also performed at higher temperature
(Supporting Information, Table S2).[12] As summarized in
Table S5, the cell parameters at 200 K are quite different from
those at 100 K. For instance, the cell volume decreased to half
its original value. The X-ray structural analysis at 200 K was
carried out by using this new unit cell. The radical and the
alcohol became indistinguishable from each other in the
À
resulting structure. The C O bond length was observed to be
1.305 ꢀ, which is close to the average bond length of 1 and 2
(Table S3). This averaging phenomenon results in an inver-
sion center between 1 and 2, which subsequently leads to the
volume reduction of the unit cell to half the original volume.
A difference Fourier map generated using the ShelXle
program[11] can be used to visualize the electron distribution
map of the hydrogen atom involved in the hydrogen bond and
provide useful information about the behavior of the hydro-
gen atom. At 100 K, the hydrogen atom was localized on
alcohol 2. However, as shown in Figure 2, the observed
electron density from the hydrogen atom splits into two
centers at 200 K. This distribution map suggested hydrogen
atom-hopping on a double-minimum potential. According to
these analyses, a thermally activated self-exchange reaction,
involving a hydrogen atom exchange between anthranol and
anthroxyl, resulted in the averaged structure observed at
200 K. Compared with the theoretically predicted model of
the phenol/phenoxyl pair, the similar conformation around
the hydrogen bond suggests that the self-exchange reaction
occurring in the anthranol/anthroxyl pair is a concerted PCET
reaction.[7] The intermolecular O···O distance at 200 K was
determined to be 2.758(3) ꢀ, which was longer than the value
(2.4 ꢀ) predicted for the transition state of the phenol/
A hydrogen-bonded complex between anthroxyl 1 and its
reduced form, anthranol 2, was successfully isolated. The
Angew. Chem. Int. Ed. 2015, 54, 2402 –2405
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 2403

.
Angewandte
Communications
The behavior of the molar magnetic susceptibility can be
understood through the intermolecular interactions observed
in the crystal. The crystal packing structure of the hydrogen-
bonded complex contains a one-dimensional chain structure
composed of two intermolecular interactions. One is an
OC···HO hydrogen bonding interaction between anthranol and
the anthroxyl; the other is a p–p stacking interaction between
the planar anthroxyl/anthranol regions. The hydrogen-
bonded dimer was stacked together via the p–p interaction
to form an offset one-dimensional chain as depicted in
Figure 4. At 100 K, the sequence of radical 1 and alcohol 2 in
Figure 2. Difference Fourier map (indicated by arrows) of electron
density around the hydrogen bond in the hydrogen-bonded complex
between 1 and 2 at 200 K. The electron density of split hydrogen
atoms in the O···O bond was observed.
phenoxyl pair.[7] This may be attributed to steric interactions
À
between O and the C H of the neighboring molecule.
The transition observed by X-ray analysis was also
investigated in detail from the perspective of the magnetic
structure. SQUID measurements on the crystalline powder
were expected to be a powerful tool for understanding the
phase transition. The temperature dependence of the molar
magnetic susceptibility (cm) was measured in the temperature
range 2–300 K at a constant field of 1 T (Figure 3). The cm T
Figure 4. Representation of the hydrogen-atom exchange within the
one-dimensional chain of the hydrogen-bonded complex. Anthranol
and anthroxyl are abbreviated as “alc” and “rad”, respectively. Alc and
rad are indistinguishable in the hydrogen-bonded complex at 200 K
and their positions are unclear.
the chain was uniquely determined by X-ray analysis (Fig-
ure 5a). Compounds 1 and 2 were distinguished on the basis
À
of the C O bond length. The p–p separation distance d of the
overlapping anthroxyls was 3.395 ꢀ, whereas anthranol
formed a weak p–p stacking pair (d = 3.544 ꢀ). We could
not determine the sequence at 200 K (Figure 5b). Only the
averaged structure of 1 and 2 was obtained by the X-ray
analysis as mentioned earlier, and the hydrogen-bonded
complex formed a uniform stack with d = 3.497 ꢀ along the
chain. The presence of a hydrogen bond was confirmed by
a broad signal around 3300 cmÀ1 (nO–H) in the infrared (IR)
Figure 3. Temperature dependence of cm T of the crystalline powder
sample of the hydrogen-bonded complex between 1 and 2 under
a constant magnetic field of 1 T.
value of 0.19 emuKmolÀ1 at 300 K was close to the theoretical
value of 0.375/2 emuKmolÀ1 for the isolated S = 1/2 spin of
anthroxyl, which accounts for half of the molecules. Remark-
ably, the cm T versus T plot showed a striking and irregular
decrease of cm T at approximately 125 K. This transition
exhibited a thermal hysteresis of about 15 K between cooling
and heating runs, which is an indication of a first-order
magnetic phase transition. We suspect this transition is related
to the structural change observed in the X-ray analysis. The
two X-ray structures obtained at 100 K and 200 K are
considered to be the structures below and above this phase
transition temperature, respectively. From this measurement,
the inactivity of the self-exchange reaction at 100 K is
ascribed not simply to the height of the self-exchange reaction
barrier, but also to a phase transition.
Figure 5. Crystal packing structure of the hydrogen-bonded complex
between 1 and 2 at a) 100 K and b) 200 K. Ellipsoids are set at 50%
probability; substituted-anthracene and the carbon-bound hydrogen
À
atoms are omitted for clarity. The C O bond length and p–p
separation distances are shown in ꢀ. The hydrogen atom within the
À
O H···O at 200 K is omitted owing to disordering.
2404 www.angewandte.org
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 2402 –2405

Angewandte
Chemie
spectrum measured at 300 K (Supporting Information, Fig-
ure S6).
Keywords: aryloxyl radical · hydrogen bonds ·
organic conductors · proton-coupled electron transfer ·
stable organic radicals
.
Concurrent with these changes in the molecular arrange-
ment, the magnetic phase transition observed at 125 K can be
associated with phase transition from the disordered to the
ordered state. The formation of the strongly stacked radical
pair ceases the self-exchange reaction, and results in the
localization of the hydrogen atom. The decrease of cm Tat the
phase transition can be ascribed to the antiferromagnetic
interaction of the stacked radical pairs, which is thought to be
the driving force for the phase transition (Figure 4). The other
two behaviors of the cm T versus T plot, namely the gradual
decrease of cm T at the high temperature region and the non-
zero value of cm T at a very low temperature, were thought to
stem from a mismatch of the sequence in the chain or from the
radical located on the terminus of the chain. Because the
crystal packing of the hydrogen-bonded complex is very
similar to that of the radical alone (Supporting Information,
Figure S5), an alcohol molecule may be substituted by
a radical. In this mismatched region, there is no hydrogen
atom exchange, and an antiferromagnetic interaction from
the aggregated radicals can occur.
[1] a) J. J. Warren, T. A. Tronic, J. M. Mayer, Chem. Rev. 2010, 110,
6961 – 7001; b) S. Hammes-Schiffer, A. A. Stuchebrukhov,
Chem. Rev. 2010, 110, 6939 – 6960; c) M. H. V. Huynh, T. J.
Meyer, Chem. Rev. 2007, 107, 5004 – 5064; d) C. Costentin, M.
Robert, J.-M. Savꢁant, Chem. Rev. 2010, 110, PR1 – PR40.
[2] a) H. J. Eckert, G. Renger, FEBS Lett. 1988, 236, 425 – 431;
b) C. W. Hoganson, G. T. Babcock, Science 1997, 277, 1953 –
1956; c) H. Matsuoka, J.-R. Shen, A. Kawamori, K. Nishiyama,
Y. Ohba, S. Yamaguchi, J. Am. Chem. Soc. 2011, 133, 4655 – 4660.
[3] a) C. S. Yee, M. C. Y. Chang, J. Ge, D. G. Nocera, J. Stubbe, J.
Am. Chem. Soc. 2003, 125, 10506 – 10507; b) J. Stubbe, D. G.
Nocera, C. S. Yee, M. C. Y. Chang, Chem. Rev. 2003, 103, 2167 –
2201.
[4] C. Costentin, C. Louault, M. Robert, J.-M. Savꢁant, J. Am.
Chem. Soc. 2008, 130, 15817 – 15819.
[5] a) T. Mitani, G. Saito, H. Urayama, Phys. Rev. Lett. 1988, 60,
2299 – 2302; b) K. Nakasuji, K. Sugiura, T. Kitagawa, J. Toyoda,
H. Okamoto, K. Okaniwa, T. Mitani, H. Yamamoto, I. Murata,
A. Kawamoto, J. Tanaka, J. Am. Chem. Soc. 1991, 113, 1862 –
1864.
We conclude that the hydrogen-atom self-exchange
reactions, based on the PCET mechanism between anthranol
and anthroxyl, are inactive during the phase transition
because of the strong antiferromagnetic pairing of the radicals
inside the one-dimensional chain. The temperature-induced
phase transition is reversible and is activated again at high
[6] a) T. Isono, H. Kamo, A. Ueda, K. Takahashi, A. Nakao, R.
Kumai, H. Nakao, K. Kobayashi, Y. Murakami, H. Mori, Nat.
Commun. 2013, 4, 1344; b) H. Kamo, A. Ueda, T. Isono, K.
Takahashi, H. Mori, Tetrahedron Lett. 2012, 53, 4385 – 4388;
c) A. Ueda, S. Yamada, T. Isono, H. Kamo, A. Nakao, R. Kumai,
H. Nakao, Y. Murakami, K. Yamamoto, Y. Nishio, et al., J. Am.
Chem. Soc. 2014, 136, 12184 – 12192; d) S. C. Lee, A. Ueda, H.
Kamo, K. Takahashi, M. Uruichi, K. Yamamoto, K. Yakushi, A.
Nakao, R. Kumai, K. Kobayashi, H. Nakao, Y. Murakami, H.
Mori, Chem. Commun. 2012, 48, 8673 – 8675.
=
temperature. However, the C O IR stretching vibration of
À
anthroxyl and the O H vibration of anthranol at 300 K and
5 K (Supporting Information, Figure S6) provide no direct
evidence for the exchange (such as an anomalous peak shift).
The timescale for the exchange reaction is probably slow.
Further investigations about the dynamics are currently in
progress.
[7] J. M. Mayer, D. A. Hrovat, J. L. Thomas, W. T. Borden, J. Am.
Chem. Soc. 2002, 124, 11142 – 11147.
[8] For other theoretical studies, see; a) A. Sirjoosingh, S. Hammes-
Schiffer, J. Phys. Chem. A 2011, 115, 2367 – 2377; b) M. K.
Ludlow, J. H. Skone, S. Hammes-Schiffer, J. Phys. Chem. B 2008,
112, 336 – 343; c) J. H. Skone, A. V. Soudackov, S. Hammes-
Schiffer, J. Am. Chem. Soc. 2006, 128, 16655 – 16663; d) G. A.
DiLabio, E. R. Johnson, J. Am. Chem. Soc. 2007, 129, 6199 –
6203.
[9] a) T. Yoshida, Y. Ueno, S. Wakabayashi, Kogyo Kagaku Zasshi
1969, 72, 338 – 341; b) L. S. Singer, I. C. Lewis, T. Richerzhagen,
G. Vincow, J. Phys. Chem. 1971, 75, 290 – 291; c) I. C. Lewis, L. S.
Singer, J. Phys. Chem. 1981, 85, 354 – 360; d) P. Devolder, Can. J.
Chem. 1976, 54, 1744 – 1750; e) P. Devolder, P. Goundmand,
Chem. Phys. 1978, 35, 307 – 315.
The high stability of anthroxyl radical 1 allowed us to
isolate a hydrogen-bonded complex with anthranol 2, which
was found to be active in the self-exchange reaction. The
electron distribution map of the hydrogen atom confirmed
the activation of this reaction. Also, the X-ray structure of our
hydrogen-bonded complex was close to that of the predicted
PCETactivated complex of the phenol/phenoxyl pair. Kinetic
studies in the solution phase and theoretical studies examin-
ing PCET reactions[7,8] will be supplemented with the
structural information in this solid-phase study. We envision
shorter hydrogen bond lengths are necessary for activating
the PCET reaction, and we will prepare modified anthroxyls
aimed at generating more efficient PCET systems. Further-
more, we will conduct experimental studies to compare with
the theoretical prediction and relate these findings to the
application of PCET reactions as a new conductive mecha-
nism for organic conductors.
[10] See the Supporting Information of Ref. [7].
[11] “ShelXle: a Qt graphical user interface for SHELXL”: C. B.
Hꢂbschle, G. M. Sheldrick, B. Dittrich, J. Appl. Crystallogr. 2011,
44, 1281 – 1284.
[12] CCDC 1040627 (100K) and CCDC 1040628 (200K) contain the
supplementary crystallographic data for this paper. These data
can be obtained free of charge from The Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
Received: November 5, 2014
Published online: January 7, 2015
Angew. Chem. Int. Ed. 2015, 54, 2402 –2405
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org 2405