Hexakis(4-(N-butylpyridylium))benzene: A six-electron organic redox system

Article abstract of DOI:10.1021/jo701944c

(Chemical Equation Presented) The cobalt-catalyzed cyclotrimerization of bis(4-pyridyl)acetylene affords hexakis(4-pyridyl)benzene in moderate yield. Alkylation with n-butyltriflate gives hexakis(4-(N-butylpyridylium))benzene triflate (1⁶⁺), wh

Full text of DOI:10.1021/jo701944c

Hexakis(4-(N-butylpyridylium))benzene: A Six-Electron Organic
Redox System
Zhenfu Han,‡ Thomas P. Vaid,*,‡ and Arnold L. Rheingold§
Department of Chemistry and Center for Materials InnoVation, Washington UniVersity, St. Louis, Missouri
63130, and Department of Chemistry and Biochemistry, UniVersity of California, San Diego,
La Jolla, California 92093
Vaid@wustl.edu
ReceiVed September 4, 2007
The cobalt-catalyzed cyclotrimerization of bis(4-pyridyl)acetylene affords hexakis(4-pyridyl)benzene in
moderate yield. Alkylation with n-butyltriflate gives hexakis(4-(N-butylpyridylium))benzene triflate (16+),
which can be reduced with Na/Hg in DMF to neutral 10. A single-crystal X-ray diffraction structure
reveals that 10 has a chair-cyclohexane-like core and a [6]radialene structure. Cyclic voltammetry shows
that 16+ is reversibly reduced to 12+ in one four-electron step and 12+ is reversibly reduced to 10 in one
two-electron step. A reduction by four electrons at one potential is unprecedented for a molecule in
which the electrochemically active centers are in electronic communication. The large structural
transformation from 16+ to 10 is responsible for the “potential inversion” in the cyclic voltammetry, and
DFT calculations suggest a possible structure for the stable intermediate 12+. A comparison is made to
the electrochemistry and structural transformations in a previously prepared [4]radialene analogue of 10.
Introduction
Recently, we have synthesized a number of highly reducing,
neutrally charged molecules.1-5 Those molecules have included
the neutral form of phenyl viologen2 and the neutral form of an
“extended viologen”3 (see Figure 1). A neutral viologen is very
easily oxidized to its dication because two quinoid rings become
aromatic when two electrons are removed. In our extended
viologen in Figure 1, four quinoid rings become aromatic when
the neutral molecule is oxidized to the dication. We therefore
FIGURE 1. The dicationic and neutral forms of an extended viologen.
expected that the neutral extended viologen would be a stronger
reducing agent than a regular neutral viologen. Judged by its
(2+/0) redox potential, the neutral extended viologen is indeed
more reducing than a regular neutral viologen and is in fact the
most reducing neutral organic molecule known.3
‡ Washington University, St. Louis.
§ University of California, San Diego.
(1) Cissell, J. A.; Vaid, T. P.; Rheingold, A. L. J. Am. Chem. Soc. 2005,
127, 12212-12213.
(2) Porter, W. W., III; Vaid, T. P. J. Org. Chem. 2005, 70, 5028-5035.
(3) Porter, W. W., III; Vaid, T. P.; Rheingold, A. L. J. Am. Chem. Soc.
2005, 127, 16559-16566.
After our work on the extended viologen, we were interested
in investigating the chemistry and electrochemistry of other
viologen-like molecules. One intriguing possibility was the
hexacation hexakis(4-(N-alkylpyridylium))benzene (see Scheme
(4) Cissell, J. A.; Vaid, T. P.; Rheingold, A. L. Inorg. Chem. 2006, 45,
2367-2369.
(5) Cissell, J. A.; Vaid, T. P.; Yap, G. P. A. J. Am. Chem. Soc. 2007,
129, 7841-7847.
10.1021/jo701944c CCC: $40.75 © 2008 American Chemical Society
Published on Web 12/18/2007
J. Org. Chem. 2008, 73, 445-450
445

Han et al.
SCHEME 1
1 for the structure). It seemed it might be possible to reduce
the molecule all the way to its neutral form. The hexacation
(with the alkyl group ) methyl) had been synthesized previ-
ously,6,7 but its electrochemistry was not examined, and it
seemed that a more straightforward synthetic route might be
possible. The cobalt-catalyzed cyclotrimerization of alkynes to
benzene derivatives is a thoroughly studied and utilized reac-
tion,8 and recently, Mu¨llen and co-workers have made extensive
use of the reaction to create hexaarylbenzenes (and then the
oxidative cyclodehydrogenation of hexaarylbenzenes to create
hexa-peri-hexabenzocoronenes).9 It seemed that the cobalt-
catalyzed trimerization of bis(4-pyridyl)acetylene might provide
a convenient route to hexakis(4-pyridyl)benzene, and alkylation
would yield the hexacation of interest. Herein we report the
synthesis, isolation, and characterization of hexakis(4-(N-
butylpyridylium))benzene as both a hexacation and neutral
molecule.
Crystal Structure of 10. Yellow crystals of 10 were grown
by slowly cooling a hot, concentrated solution in toluene to room
temperature. The single-crystal X-ray structure of 10 is shown
in Figure 2. The C6 core of 10 has a chair-cyclohexane-like
structure, and the bond lengths of 10 (Table 1) match those
expected from the pattern of double and single bonds shown
for 10 in Scheme 1. The structure of 10 classifies it as a [6]-
radialene.10,11 Clearly a large structural rearrangement has
occurred in the conversion of 16+ to 10, which has implications
for the electrochemistry of 1, as discussed below.
Electrochemistry. The cyclic voltammogram of 16+ in DMF
is shown in Figure 3. The redox wave at 0 V is due to a
ferrocene internal standard. There are two reversible reductions
of 16+ at -1.14 and -1.33 V (vs Fc0/+). Because 10 was
synthesized by the reduction of 16+ with Na/Hg in DMF, and
the potential of Na/Hg in nonaqueous solvents is about -2.36
V,12 and no further reductions are observed in the cyclic
voltammogram down to -2.50 V, it is very likely that the two
observed reduction waves together represent the reduction of
16+ to 10. Because the two redox waves are of unequal
magnitude, they cannot be due to two three-electron reductions
(16+f13+f10) and must be due to either a four-electron and
subsequent two-electron reduction (16+f12+f10) or a five-
electron and subsequent one-electron reduction (16+f11+f10).
When the potential is swept at a constant rate (as in this cyclic
voltammetry experiment), the peak current, ip, is expected to
be proportional to n3/2, where n is the number of electrons
transferred.13 For four-electron and two-electron waves, the ratio
of the two ip should be 23/2 ) 2.83, and for five-electron and
Results and Discussion
Synthesis. Hexakis(4-pyridyl)benzene and its methylated
hexacation hexakis(4-(N-methylpyridylium))benzene have been
prepared previously.6,7 However, that work was focused mainly
on the synthesis of tris(pyridylium)propenyl and tris(pyridyli-
um)cyclopropenyl compounds, and it seemed that a more
efficient route to hexakis(4-pyridyl)benzene might be possible.
We found that the cobalt-catalyzed cyclotrimerization of bis-
(4-pyridyl)acetylene (available by the bromination and double
dehydrobromination of bis(4-pyridyl)ethylene) provides hexakis-
(4-pyridyl)benzene in 25% yield (Scheme 1). Alkylation of
hexakis(4-pyridyl)benzene with n-BuOTf in CH2Cl2 afforded
the hexacation hexakis(4-(N-butylpyridylium))benzene triflate
(16+ triflate) in 94% yield as a colorless, crystalline solid. Reduc-
tion of 16+ with sodium amalgam in DMF gave the neutral radia-
lene 10 in 66% yield as a bright yellow, air-sensitive powder.
one-electron waves, the ratio of the two ip should be 53/2
)
11.2. The experimental ratio of the two ip is 3.16, indicating
that it is very likely that the redox processes are 16+ + 4 e- T
12+ and 12+ + 2 e- T 10.
(10) Gholami, M.; Tykwinski, R. R. Chem. ReV. 2006, 106, 4997-
5027.
(6) Breslow, R.; Crispino, G. A. Tetrahedron Lett. 1991, 32, 601-604.
(7) Crispino, G. A.; Breslow, R. J. Org. Chem. 1992, 57, 1849-1855.
(8) Agenet, N.; Gandon, V.; Vollhardt, K. P. C.; Malacria, M.; Aubert,
C. J. Am. Chem. Soc. 2007, 129, 8860-8871 and references therein.
(9) Wu, J.; Pisula, W.; Mu¨llen, K. Chem. ReV. 2007, 107, 718-747.
(11) Hopf, H.; Maas, G. Angew. Chem., Int. Ed. Engl. 1992, 31, 931-
954.
(12) Connelly, N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877-910.
(13) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Fundamen-
tals and Applications; John Wiley & Sons: New York, 1984.
446 J. Org. Chem., Vol. 73, No. 2, 2008

Hexakis(4-(N-butylpyridylium))benzene
FIGURE 2. Two views of the solid-state structure of the [6]radialene 10 (only the first carbon of each n-butyl group is shown).
TABLE 1. Selected Bond Lengths in the Structure of 10
bond
length (Å)
C(1)-C(2)
1.489(2)
1.360(2)
1.452(2)
1.336(2)
1.373(2)
1.459(2)
C(1)-C(11)
C(11)-C(15)
C(15)-C(14)
C(14)-N(1)
N(1)-C(16)
Ideally, the separation between the potentials of the peak
anodic and peak cathodic currents, ∆Ep, should be 59 mV for
the one-electron ferrocene wave, 30 mV for the two-electron
1
2+ + 2 e- T 10 wave, and 15 mV for the four-electron 16+
+
FIGURE 3. Cyclic voltammogram of 16+ in DMF with 0.10 M [NBu4]-
4 e- T 12+ wave.13 However, we observe a ∆Ep of 74 mV for
ferrocene, 76 mV for the 12+ + 2 e- T 10 wave, and 46 mV
for the 16+ + 4 e- T 12+ wave (at 50 mV/s sweep rate). The
∆Ep values were similar at sweep rates of 25 and 100 mV/s.
These ∆Ep would seem to indicate that the two redox waves
observed for 1 correspond to two-electron and one-electron
processes, but that would mean that 16+ has been reduced only
to 13+ in the cyclic voltammetry experiments, and that seems
highly unlikely. The reason for the unexpected ∆Ep values is
not clear.
[PF6], with ferrocene0/+ internal standard at 0 V, scan rate 50 mV/s.
Sharp peaks in cyclic voltammograms such as the ones
observed for 16+ + 4 e- T 12+ are sometimes caused by
adsorption of an analyte to the electrode. For a strictly
diffusional process, with no adsorption to the electrode, the peak
current should be proportional to the square root of the potential
sweep rate. We obtained cyclic voltammograms of 1 at 25, 50,
100, 150, and 200 mV/s, starting at -0.50 V and sweeping in
the negative direction. A plot of the peak reductive current, ipc,
for the 16+ + 4 e- f 12+ process versus the square root of the
potential sweep rate is linear (R2 ) 0.997) with a near-zero
intercept (see Figure S1 in Supporting Information). Therefore,
it appears that the 16+ + 4 e- T 12+ process is diffusion-
controlled, and the sharp peaks in the cyclic voltammogram are
not due to adsorption.
The reversible oxidation/reduction of a molecule by four
electrons at one potential is without precedent, at least for
molecules in which the redox-active centers are in electronic
communication. There are many examples of molecules with
several identical redox-active units that are not in electronic
communication and are therefore oxidized or reduced at the
same potential. For example, a dendrimer peripherally func-
tionalized with n ferrocene units will generally undergo an
oxidation by n electrons at one potential.16-18 In contrast, it is
likely that in the reduction of 16+ at least some of the reduction
potentials are inverted;19 in other words, 15+ + e- f 14+ may
Attempts to isolate 12+ by the conproportionation of a 2:1
ratio of 10 and 16+ were unsuccessful. In polar solvents such as
DMF, a 2:1 ratio of 10 and 16+ yielded dark red solutions,
presumably due to the presence of 12+. However, removal of
the solvent or addition of a less polar solvent such as ether led
to disproportionation of the dark red species to 10 and 16+. A
solution containing 12+ was created by combining 10 and 16+
in a 2:1 ratio in CH3CN; its UV-vis spectrum had absorbances
at 232, 259, and 503 nm, with absorptivity ratios of 1.8:1.5:1.
The absorbance at 259 nm may be due to 16+
.
The 10/2+ potential of -1.33 V shows that 10 is a moderately
strong reducing agent, with an oxidation potential similar to
that of cobaltocene.12 The (0/+) potential of methyl viologen14
is -1.26 V (vs ferrocene0/+), and the (0/2+) potential of our
previously reported four-ring extended viologen (see
Figure 1) is -1.48 V.3 The (0/2+) potential of a previously
reported three-ring extended viologen15 is -1.32 V, very similar
to that of 1.
(14) Baglio, J. A.; Calabrese, G. S.; Harrison, D. J.; Kamieniecki, E.;
Ricco, A. J.; Wrighton, M. S.; Zoski, G. D. J. Am. Chem. Soc. 1983, 105,
2246-2256.
(15) Takahashi, K.; Nihira, T.; Akiyama, K.; Ikegami, Y.; Fukuyo, E.
Chem. Commun. 1992, 620-622.
(16) Alvarez, J.; Ren, T.; Kaifer, A. E. Organometallics 2001, 20, 3543-
3549.
(17) Sengupta, S. Tetrahedron Lett. 2003, 44, 7281-7284.
(18) Suzuki, M.; Nakajima, R.; Tsuruta, M.; Higuchi, M.; Einaga, Y.;
Yamamoto, K. Macromolecules 2006, 39, 64-69.
J. Org. Chem, Vol. 73, No. 2, 2008 447

Han et al.
FIGURE 4. Plausible structures for 14+ and 12+
.
have a reduction potential that is more positive than that of 16+
+ e- f 15+, so both reductions happen at the same potential.
That type of potential inversion generally occurs when there is
a large structural rearrangement upon electron transfer,19 and
the crystal structure of 10 shows that at some point between
16+ and 10 a very large structural rearrangement does occur.
Density functional theory (DFT) calculations, described below,
clarify the structural rearrangements that occur in the reduction
steps 16+ f 14+ f 12+ f 10.
FIGURE 5. View along a C2-axis of the calculated (B3LYP/6-31G*)
structure of 12+ in the “twisted” geometry, with n-butyl groups
substituted by methyl groups, and hydrogens not shown.
10) starting in all four of the structures outlined above, for 16
total minimizations. For all four oxidation states, the expected
geometry (16+ benzene, 14+ 1,3-cyclohexadiene, 12+ cyclohex-
ene, 10 chair cyclohexane) was the lowest energy configuration,
with the other configurations either higher in energy, or
converting to the expected configuration during the minimiza-
tion, or converting to another geometry altogether that was
higher in energy than the global minimum.
DFT Calculations. It has not been possible to isolate 1 in
the intermediate oxidation states 12+ or 14+ (or any other
intermediate oxidation state), so we undertook DFT calculations
on 16+, 14+, 12+, and 10 to better understand the intermediate
geometries in the large structural rearrangement that occurs in
the reduction of 16+ to 10. Simply drawing two valence-bond
structures for 14+ and 12+ (Figure 4) suggests one set of plausible
structures for 14+ and 12+. The hexacation 16+ has a benzene-
like core, and neutral 10 has a chair-cyclohexane-like core, and
possible structures for 14+ and 12+ have the 1,3-cyclohexadiene
and cyclohexene-like cores, respectively, shown in Figure 4.
To convert from the structure of 16+ to the suggested 1,3-
cyclohexadiene structure of 14+, one pair of neighboring
peripheral pyridyl rings must twist such that two CH groups
move past each other, from “above” the central C6 ring to
“below” the central C6 ring and vise versa. Similarly, a
conversion from the 1,3-cyclohexadiene core to the cyclohexene
core requires another set of CH groups to move past each other,
as does a conversion from the cyclohexene-like core to the chair-
cyclohexane-like core. Those sterically difficult conversions
separate the four structures (with benzene, 1,3-cyclohexadiene,
cyclohexene, and chair cyclohexane cores) into local energetic
minima.
In one of the geometry optimizations, when 10 was started
in the 1,3-cyclohexadiene geometry, the final structure had a
C6 core with a “twisted” geometry. When the structures of 14+
and 12+ were optimized starting from that twisted geometry,
they had lower energies than any of the other optimized
structures for their respective oxidation states. The optimized
structure of 12+ in the twisted geometry is shown in Figure 5.
It has overall molecular D2 symmetry, with three mutually
perpendicular C2 axes. The structure is very similar to that of
the dication of hexakis(dimethylamino)benzene, which is dis-
cussed further below. It is likely that the twisted D2 structure is
the true global minimum for 12+, as the calculated energy of
the twisted structure is 10 kcal/mol lower in energy than the
structure with the next lowest energy, the cyclohexene core
structure. The situation is somewhat less clear for 14+, where
the twisted structure has an energy only 2.4 kcal/mol lower than
the nearest in energy, the 1,3-cyclohexadiene-like structure. In
addition, all the calculations were done without a solvent model,
and adding a solvent might change the energy ordering. Neither
12+ nor 14+ has an electric dipole in the twisted structure. In
contrast, the cyclohexene-like structure for 12+ has a large
calculated dipole moment of 10.4 D, pointing in the direction
expected from the valence-bond picture of Figure 4. Similarly,
the 1,3-cylohexadiene structure of 14+ has a calculated dipole
moment of 7.1 D, pointing in the direction expected from
Figure 4. The presence of a polar solvent may lower the energies
of the dipolar structures relative to the twisted structures.
However, it is unlikely that a polar solvent would change the
relative energetics of the two structures of 12+ by more than 10
kcal/mol, so the twisted structure is likely to be the true global
minimum for 12+. The atomic coordinates for the optimized
We calculated energy-minimized structures (Gaussian 03,20
B3LYP/6-31G*, with the n-butyl groups substituted by methyl
groups) for all four oxidation states of concern (16+, 14+, 12+
,
(19) Evans, D. H.; Hu, K. J. Chem. Soc., Faraday Trans. 1996, 92,
3983-3990.
(20) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A.
D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A.
G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian
03; Gaussian, Inc.: Wallingford, CT, 2004.
448 J. Org. Chem., Vol. 73, No. 2, 2008

Hexakis(4-(N-butylpyridylium))benzene
structures of 16+, 14+ (in both the 1,3-cyclohexadiene and
twisted geometries), 12+ (in both the cyclohexene and twisted
geometries), and 10 are included in the Supporting Information.
Cyclic voltammetry shows that the disproportionation of 14+
to 16+ and 12+ in DMF (eq 1) is thermodynamically favored or
has a free energy change very close to zero, while the
disproportionation of 12+ to 16+ and 10 in DMF (eq 2) has a
positive ∆G. The DFT-calculated energies of the relevant species
do not shed any light on the energetics of eqs 1 and 2, as they
indicate that both disproportionation reactions are disfavored
by over 200 kcal/mol. However, the DFT calculations do show
that there is one particularly stable structure for 12+, the
twisted structure in Figure 5. The electrochemical behavior
of 1 shows that all oxidation states other than 10, 12+, and 16+
are unstable toward disproportionation. The absence of one
particularly favorable structure in the various DFT-optimized
geometries of 14+ is at least consistent with that observation.
FIGURE 6. The tetracationic and neutral forms of 2 (R ) CH3 or
CO2CH2CH3).
one-electron steps to 23+ and 24+. Similarly, the R ) CH3
version of 20 can be reversibly oxidized in a single two-electron
step to 22+, but its oxidation to 23+ is irreversible.
Neither form of 2 (R ) CH3 or CO2CH2CH3) in any of its
oxidation states has been structurally characterized. We calcu-
lated energy-minimized structures of 24+, 22+, and 20 (Gaussian
03,20 B3LYP/6-31G*, R ) CH3). The calculated structure of
2 14+ h 16+ + 12+ ∆G e 0
3 12+ h 16+ + 2 10 ∆G > 0
(1)
(2)
2
4+ has the expected cyclobutadiene core and aromatic pyridy-
lium substituents twisted from the plane of the cyclobutadiene
core by 38° in a propeller-like arrangement. The neutral 20 is a
[4]radialene with alternating up and down quinoidal pyridyl
groups in a butterfly-like structure. In 22+, all of the C-C bonds
in the cyclobutane core are equivalent and 1.460 Å in length.
The pyridyl groups of 22+ are all twisted from the plane of the
cyclobutane core by 28° in a propeller arrangement, giving 22+
overall D4 symmetry. The structural transformations from 20 to
22+ to 24+ explain its electrochemical behavior. The structural
change from 20 to 22+ requires at least one pair of CH groups
on neighboring pyridyl groups to twist past each other, a
sterically difficult process. There are only two important
energetic minima in the structural space of 2, the propeller-like
and butterfly-like geometries, that are separated by a substantial
energy barrier. The change between those structures occurs when
20 is oxidized to 22+ or 22+ is reduced to 20. The structural
rearrangement from 22+ to 24+ has no major energy barrier.
Those findings are consistent with the observed inverted
Comparison to Hexakis(dimethylamino)benzene. The cal-
culated optimal structure of 12+ is similar to that of the dication
of hexakis(dimethylamino)benzene. Hexakis(dimethylamino)-
benzene has the expected planar benzene core, while its dication,
which has been characterized crystallographically, has a twisted
core like that in Figure 5 and an overall approximate D2 mole-
cular symmetry.21,22 Because of the structural rearrangement,
the first two oxidation potentials of hexakis(dimethylamino)-
benzene are inverted, and it is electrochemically oxidized by
two electrons in one step.22 It appears that it is possible to elec-
trochemically oxidize the molecule to the 6+ oxidation state,23
but no oxidation states higher than 2+ have been isolated.
Another related set of molecules, hexakis(4-(N,N-diarylamino)-
phenyl)benzenes24-27 and hexakis(4-(N,N-dialkylamino)phenyl)-
benzenes,26,28 has been synthesized by cobalt-catalyzed alkyne
cyclotrimerization. Those compounds and their (poly)cations
were studied electrochemically and spectroscopically. Both the
hexakis(4-(N,N-diarylamino)phenyl)benzenes24,25 and the hexakis-
(4-(N,N-dialkylamino)phenyl)benzenes26,28 can be oxidized to
the 6+ oxidation state, but only the monocations of hexakis-
(4-(N,N-diarylamino)phenyl)benzenes have been isolated, and
none have been structurally characterized.
potentials for the processes 20 - e- f 2+ and 2+ - e- f 22+
,
f
whereas the potentials for 22+ - e- f 23+ and 23+ - e-
24+ are in the normal order (at least for R ) CO2CH2CH3, for
which those oxidations are observable and reversible).
Conclusions
The reduction of 16+ to 10 entails a large structural rear-
rangement, from the benzene-like core of 16+ to the chair-
cyclohexane-like core of the [6]radialene 10. That structural
rearrangement explains the potential inversions observed in the
cyclic voltammetry of 1, and DFT calculations suggest a
Comparison to the [4]Radialene Analogue. Two [4]-
radialene analogues of 10 have been previously synthesized and
studied electrochemically; they are pictured as 20 in Figure 6,
with R ) CH3 or CO2CH2CH3.29-31 Cyclic voltammetry showed
that 20 (R ) CO2CH2CH3) can be oxidized in one reversible,
two-electron step to the dication 22+ and then in two reversible,
structure for the one stable intermediate oxidation state, 12+
.
The four-electron reduction of 16+ to 12+ at one potential is
unprecedented for a molecule with electronic communication
between the electrochemically active centers.
(21) Chance, J. M.; Kahr, B.; Buda, A. B.; Toscano, J. P.; Mislow, K.
J. Org. Chem. 1988, 53, 3226-3232.
(22) Speiser, B.; Wu¨rde, M.; Maichle-Mossmer, C. Chem.sEur. J. 1998,
4, 222-233.
(23) Speiser, B.; Wu¨rde, M.; Quintanilla, M. G. Electrochem. Commun.
2000, 2, 65-68.
Experimental Section
(24) Lambert, C.; Noll, G. Angew. Chem., Int. Ed. 1998, 37, 2107-
2110.
Published procedures were employed for the synthesis of bis-
(4-pyridyl)acetylene32 and butyl triflate.33
(25) Lambert, C.; Noll, G. Chem.sEur. J. 2002, 8, 3467-3477.
(26) Rosokha, S. V.; Neretin, I. S.; Sun, D.; Kochi, J. K. J. Am. Chem.
Soc. 2006, 128, 9394-9407.
(29) Horner, M.; Hu¨nig, S. Angew. Chem., Int. Ed. Engl. 1977, 16, 410-
411.
(30) Horner, M.; Hu¨nig, S. Liebigs Ann. Chem. 1983, 642-657.
(31) Horner, M.; Hu¨nig, S.; Reissig, H. U. Liebigs Ann. Chem. 1983,
658-667.
(27) Hirao, Y.; Ishizaki, H.; Ito, A.; Kato, T.; Tanaka, K. Eur. J. Org.
Chem. 2007, 186-190.
(28) Sun, D.; Rosokha, S. V.; Kochi, J. K. Angew. Chem., Int. Ed. 2005,
44, 5133-5136.
J. Org. Chem, Vol. 73, No. 2, 2008 449

Han et al.
Cyclic voltammetry was performed using a potentiostat with
electrical connection to the inside of a nitrogen-filled drybox, where
all materials and solutions were maintained during the measure-
ments. Cyclic voltammetry was performed in anhydrous DMF with
0.10 M [Bu4N][PF6] supporting electrolyte and a ferrocene internal
standard. The working electrode and pseudoreference electrode were
0.50 mm diameter platinum disks, and the counter electrode was a
2.5 mm diameter platinum disk.
Neutral Radialene (10). Manipulations were carried out on a
Schlenk line or in a glovebox under N2 or vacuum. A solution of
hexakis(4-(N-butylpyridylium))benzene triflate (1.067 g, 0.600
mmol) in 20 mL of DMF was added to sodium amalgam (Na,
0.0911 g, 3.96 mmol in 10 g Hg). The solution turned violet
immediately, then turned dark red after a few minutes. Gradually,
the color of the solution became yellow-brown. The mixture was
stirred at 22 °C for 12 h. The DMF was removed under vacuum.
Benzene (250 mL) was added to the residue, and the suspension
was filtered. The volatiles were removed from the filtrate under
vacuum to leave a red-brown solid. Ether (30 mL) was added to
the residue, forming a red-brown solution with undissolved yellow
solid. The solid was collected by filtration and washed with ether
to afford the neutral radialene (0.3501 g) in 66% yield as bright
yellow powder: 1H NMR (300 Hz, C6D6) δ 5.90 (d, J ) 8.1 Hz,
12H), 5.78 (d, J ) 8.1 Hz, 12H), 2.52 (t, J ) 6.6 Hz, 12H), 1.07
(m, 12H), 0.93 (m, 12H), 0.66 (t, J ) 7.2 Hz, 18H); 13C NMR (75
Hz, C6D6) δ 129.8, 121.3, 119.9, 113.0, 53.7, 32.9, 20.1, 14.2; UV-
vis (C6H6) 321 nm, log ꢀ ) 4.83; IR (Nujol, cm-1) 1652, 1407,
1218, 1190, 1048, 1006, 928, 794, 769; HRMS (MALDI) calcd
for C60H78N6 (M+) 882.6288, found 882.6285.
Hexakis(4-pyridyl)benzene. A solution of bis(4-pyridyl)acety-
lene (4.680 g, 26.0 mmol) and CpCo(CO)2 (0.374 g, 2.08 mmol)
in 60 mL of dioxane was heated to reflux under N2 for 3 days. The
volatiles were removed by rotary evaporation. The residue was
purified by silica gel chromatography with CH2Cl2/MeOH (5:1) as
the eluent, affording the white product (1.205 g) in 25% yield. The
identity and purity of the product were confirmed by comparison
1
of its H NMR spectrum with the reported spectrum.6,7
Hexakis(4-(N-butylpyridylium))benzene Triflate (16+ Tri-
flate). Butyl triflate (1.591 g, 7.72 mmol) was added to a solution/
suspension of hexakis(4-pyridyl)benzene (0.6301 g, 1.17 mmol)
in 75 mL of dry CH2Cl2, and the mixture was stirred at 22 °C under
N2 for 24 h. The precipitate was collected by filtration and washed
with CH2Cl2 to furnish hexakis(4-(N-butylpyridylium))benzene
triflate (1.960 g) in 94% yield as white powder: 1H NMR (300
Hz, CD3CN) δ 8.42 (d, J ) 5.7 Hz, 12H), 8.00 (d, J ) 5.7 Hz,
12H), 4.31 (t, J ) 7.5 Hz, 12H), 1.73 (m, 12H), 1.16 (m, 12H),
0.87 (t, J ) 7.2 Hz, 18H); 13C NMR (75 Hz, CD3CN) δ 154.2,
146.2, 136.8, 131.2, 122.3 (q, J ) 319 Hz, -O3SCF3), 62.7, 33.8,
20.1, 14.0; UV-vis (CH3CN) 262 nm, log ꢀ ) 4.83; IR (Nujol,
cm-1) 1642, 1265, 1225, 1158, 1031, 638. Anal. Calcd (found)
for C66H78F18N6O18S6: C, 44.59 (44.88); H, 4.42 (4.17); N, 4.73
(4.68).
Acknowledgment. We thank Prof. Dennis Evans for a
helpful discussion, and Dayna Turner for some experimental
assistance. Funding was provided by NSF Grant CHE-0133068,
and the computational chemistry resource was supported by NSF
Grant CHE-0443511.
Supporting Information Available: Crystal structure data in
CIF format, crystallographic details, plot of ipc vs sweep rate, and
coordinate files for the calculated structures of 16+, 14+, 12+, and
10. This material is available free of charge via the Internet at
http://pubs.acs.org.
(32) Coe, B. J.; Harries, J. L.; Harris, J. A.; Brunschwig, B. S.; Coles,
S. J.; Light, M. E.; Hursthouse, M. B. Dalton Trans. 2004, 2935-2942.
(33) Hong, Y.-R.; Gorman, C. B. J. Org. Chem. 2003, 68, 9019-9025.
JO701944C
450 J. Org. Chem., Vol. 73, No. 2, 2008