Dicyanoaromatic radical anions as mixed valence species

Article abstract of DOI:10.1002/poc.2905

The reduction of nine symmetric dicyanoaromatic radical anions by sodium amalgam in the presence of cryptand[2.2.2] was studied using cyclic voltammetry and using optical and electron paramagnetic resonance (EPR) spectroscopies in two aprotic solvents. Al

Full text of DOI:10.1002/poc.2905

Research Article
Received: 18 October 2011,
Revised: 29 November 2011,
Accepted: 2 December 2011,
Published online in Wiley Online Library: 2012
(wileyonlinelibrary.com) DOI: 10.1002/poc.2905
Dicyanoaromatic radical anions as mixed
valence species
Álvaro Moneoa, M. Fernanda N. N. Carvalhoa and João P. Teloa*
The reduction of nine symmetric dicyanoaromatic radical anions by sodium amalgam in the presence of cryptand[2.2.2]
was studied using cyclic voltammetry and using optical and electron paramagnetic resonance (EPR) spectroscopies in
two aprotic solvents. All radicals are charge-delocalized (Class III) mixed valence species, as shown by the weak
solvatochromism of their low-energy optical bands and by the vibrational structure exhibited by most of the bands.
The maximum of the low-energy optical transition decreases logarithmically with the number of bonds (n) between
cyano groups in a series of p-phenylene-bridged radicals, from p-phenyl (n = 5) to p-quaterphenyl (n = 17), although
the energy of the latter lies higher than the value predicted by the linear regression. The energy of this band decreases
linearly with the cos2 value of the torsion angle between phenyl rings in the spectra of biphenyl-4,4′-dicarbonitrile
radical anion and their methyl-substituted derivatives. The fact that charge delocalization is maintained in radicals with
non-Kekulé bridges, with unusually large bridges and with bridges with highly twisted biphenyl systems suggests that
cyano charge-bearing units have small reorganization energies and induce high electronic couplings through the
bridges. Copyright © 2012 John Wiley & Sons, Ltd.
Keywords: charge-transfer optical bands; conjugation; dicyanoaromatic radical anions; Marcus–Hush theory; organic mixed
valence
range of properties can be achieved by changing the nature of
the aromatic bridge on these radicals. Radical anions with small
INTRODUCTION
bridges, where the two nitro groups are conjugated in what is
often called a Kekulé substitution pattern, like p-dinitrobenzene
Mixed valence (MV) compounds are characterized by having two
charge-bearing units (CBU) symmetrically attached to a bridge,
or 2,6-dinitronaphthalene radical anions, are delocalized Class III
and being at an oxidation level for which the charges at the CBU
[6,7]
[1–4]
MV species in all aprotic solvents studied.
Increasing the
could be different. The Marcus–Hush two-state model
for
distance between nitro groups increases the l/2Hab ratio and
may result in localization of charge. 4,4′-dinitrobiphenyl and 4,
electron transfer employs parabolas representing diabatic states
with the charge on the left and on the right (traced curves on
Fig. 1). These diabatic surfaces are allowed to interact through an
electronic coupling term Hab, producing the adiabatic energy
surfaces shown as solid lines in Fig. 1. When the vertical reorgani-
zation energy (l), which is the energy gap between the diabatic
parabolas at their energy minima, exceeds 2Hab, the charge is
instantaneously localized in one of the CBU, and the compound
4′-dinitrotolane are examples of Class II MV species in solvents
[8–11]
that induce a high reorganization energy value.
Decreasing
the conjugation (and hence Hab) by the twisting around the
central bond on substituted biphenyl bridges, as in 2,2′-dimethyl-
4,4′-dinitrobiphenyl radical anion, also leads to localization of
charge.[9,10] Dinitroaromatic radical anions with non-Kekulé substi-
tution patterns like 1,3-dinitrobenzene, 2,7-dinitronaphthalene, or
2,7-dinitrodibenzodioxin radical anions have small enough Hab
values to make them localized Class II radicals in all solvents
studied. [12–15]
Although there are several studies that explore the effects of
the nature of the bridge on the properties of organic MV
species, the same amount of effort have not been applied to
the systematic study of the nature of the charge-bearing unit.
Changing the bridge affects mostly the electronic coupling
Hab, but MV compounds with the same bridge and different
CBU will differ both in the electronic coupling between the
[5]
is charge-localized, or Class II in the Robin-Day classification.
As shown in Fig. 1, the lower energy surface for a Class II
compound has double minima, and interconvertion between the
two minima corresponds to the thermally-activated electron
transfer between CBU. When l < 2Hab, there is only a single
minimum in the ground state adiabatic energy surface, and the
compound is delocalized (or Class III).
Class III and Class II compounds can usually be distinguished by
their low-energy optical absorption band, often referred to as the
intervalence band. In Class II systems, the vertical excitation from
the ground state minimum is to a steeply sloping region of the
upper energy surface, producing the typical wide and Gaussian
shaped intervalence bands with no vibrational fine structure,
typical of localized MV compounds. In delocalized Class III MV
systems, the excited state has a minimum vertical from the
ground state minimum, so the optical bands are narrow and often
show vibrational structure.
* Correspondence to: J. P. Telo, Centro de Química Estrutural, Instituto
Superior Técnico, Technical University of Lisbon, Av. Rovisco Pais, 1049-001
Lisboa, Portugal. E-mail: jptelo@ist.utl.pt
a Á. Moneo, M. Fernanda N. N. Carvalho, J. P. Telo
Centro de Química Estrutural, Instituto Superior Técnico, Technical University
of Lisbon, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
Dinitroaromatic radical anions have been the subject of our
studies on MV chemistry for the last several years.[6–15] A wide
J. Phys. Org. Chem. (2012)
Copyright © 2012 John Wiley & Sons, Ltd.

A. MONEO ET AL.
Class III
Class II
λ/2Hab=4
λ/2Hab=0.8
λ
2Hab
λ
2Hab
Figure 2. Optical spectra of 18NDN, 26NDN, and 18ADN radical anions
in MeCN at room temperature
0.0
0.5
1.0
0.0
0.5
1.0
ET coordinate
ET coordinate
Figure 1. Marcus–Hush diagrams for localized (Class II, left-hand side)
and delocalized (Class III, right-hand side) MV compounds using the
frequently applied two-state model
before, the two-state model does not apply unaltered to Class III
MV compounds.[7,16] Nevertheless, although not being correct to
assign it to 2Hab, Eop often decreases exponentially with the
distance between the charge-bearing units, as expected for an
electronic coupling, and in most cases, its change with the CBU
substitution pattern somewhat accompany the predictable
change in Hab. For example, the energy of the (0,0) maximum
occurs at 14,010 cmÀ1 for the conjugated 26NDN radical anion,
but at much lower energy (9500 cmÀ1) for the non-conjugated
18NDN (Fig. 2 and Table 1). Although with a shorter distance
between CBU, theÀlack of direct resonance between the cyano
CBU and the bridge and in the reorganization energy. In this
work, we explore the MV chemistry of the dicyanoaromatic radi-
cal anions of Chart 1 and compare their properties with other
MV compounds with similar bridges.
RESULTS AND DISCUSSION
•
groups in 18NDN causes a much smaller electronic coupling
than in 26NDN•À, and this effect is reflected in the relative values
of Eop. The fact that the non-Kekulé 18NDN radical anion is a delo-
calized Class III radical anion led us to prepare the also non-Kekulé
18ADN. Increasing the distance between CBU should decrease Hab
and possibly induce charge localization. Surprisingly, not only the
narrow optical band of 18ADN radical anion shows it is still Class
III but it also occurs at a higher energy than the one from the
shorter 18NDN. The difference between the first and the second
reduction potentials (Table 1) is a measure of the electronic inter-
action between CBU. This difference is similar for both radicals,
although higher for the 1,8-naphthalene-bridged radical, which is
The optical spectra of 18NDN, 26NDN, and 18ADN radical anions
are shown in Fig. 2. The low-energy absorption bands are narrow
and clearly have vibrational fine structure, showing that these
radicals are delocalized Class III MV compounds. Although the term
"intervalence band" is also used for Class III systems, in this case the
Frank–Condon vertical transition does not involve a net charge
transfer between CBU. Application of Marcus–Hush two-state
model to Class III compounds implies that the energy of the
vertical transition equals 2Hab, as shown in Fig. 1. A large number
of studies use the Eop value of the lowest-energy (0,0) band of Class
III MV compounds as an accurate measure of 2Hab. However,
delocalized MV systems are symmetric, and the electronic
transition shown in the right-hand side of Fig. 1 occurs between
symmetric electronic states. This symmetry-forbidden transition
should produce an optical band of very low intensity, which is
clearly not the case of the bands shown in Fig. 2. As pointed out
not consistent with 18NDNÀ having a smaller Eop. This shows that
•
the correlation of Eop with 2Hab should be taken cautiously.
The low-energy optical bands of the oligo-(p-phenylene)
dicarbonitrile radical anions on Fig. 3 show a typical decrease of
Eop, with the distance between CBU. Terephthalonitrile (Ph1DN)
CN CN
CN
CN
NC
CN
18NDN
26NDN
18ADN
NC
CN
NC
CN
NC
CN
n
n=1,2,3,4
PhnDN
M2Ph2DN
M4Ph2DN
Chart 1. Dicyanoaromatic compounds studied in this work
wileyonlinelibrary.com/journal/poc
Copyright © 2012 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. (2012)

DICYANOAROMATIC RADICAL ANIONS
Table 1. Cyclic voltammetry data (in V, vs Fc/Fc+, Pt electrode) of dicyanoaromatic compounds, optical spectra data of their
radical anions and calculated dihedral angle between phenyl rings
Reduction
Potential
Optical
Spectrum
I
II
Compound
Eo
II Eo
ΔEo = IEo1/2 À Eo1/2
Eop (cmÀ1 a
)
e (MÀ1 cmÀ1 a
)
θb
1/2
1/2
18NDN
26NDN
18ADN
À1.92c
À1.93c
À1.67
À1.59e
À2.00
À2.08
À2.17
À2.27
À2.42
À2.71
À2.75d
À2.70
À2.37
À2.37e
À2.79d
À2.54
À2.36
–
0.83
0.77
9500
14,010
10,160
1670
8130
5620
–
–
–
0.70
0.78e
0.79
Ph1DN
Ph2DN
Ph3DN
Ph4DN
M2Ph2DN
M4Ph2DN
23,200
13,230
7680
6820
10,090
6600
3427
21,300
13,850
10,070
7760
–
0.46
0.19
<0.04
0.30
0.28
10ꢀ
16ꢀ
20ꢀ
43ꢀ
69ꢀ
À2.72
À2.99d
5230
aMaximum of the low-energy (0,0) band of the radical anion.
bCalculated dihedral angle between phenyl rings of the radical anions.
cIn good agreement with the values in reference
[38]
dNot reversible, value of Ep/2.
eMeasured with a glassy carbon working electrode.
Figure 3. Optical spectra of radical anions from the oligo-(p-pheny-
Figure 4. Comparison of the optical spectra of Class III Ph4DN radical
anion (red lines) with the Class II quaterphenylquinone radical anion
(green lines) in dimethylformamide (full lines) and MeCN (traced lines).[20]
lene)dicarbonitriles Ph1DN to Ph4DN in DMF. The spectrum of
Ph2DN (right-hand side axis) is plotted with 1/3 the intensity of the
other spectra (left-hand side axis). The band at 11300cmÀ1 of Ph4DN•À
(red line) belongs to the dianion formed by disproportionation
The bands at 10,000–12,000 cmÀ1 on the Ph4DN À spectra belong to the
•
dianion formed by disproportionation
radical anion absorbs mainly in the ultraviolet (UV), but the
intervalence band shifts to lower energies with the increase of
the bridge size. Surprisingly to us, all radicals are delocalÀized Class
The decay of Eop with distance on Class III MV systems is
normally exponential, and this has been one of the arguments to
ascribe the energy of the optical transition to 2Hab. The data are
fitted with the equation ln(Eop/2) = constant – bNÁN/2, where bN is
a dimensionless distance decay constant and N the number of
bonds connecting the CBU. [21] A plot of ln (Eop/2) for radical anions
Ph1DN to Ph4DN versus N is shown in Fig. 5. A linear fit with
a decay constant of bN = 0.28 is obtained for the first three N
values (N = 5, 9, and 13), but the point for the 17-bonded Ph4DN
compound lies somewhat higher than what is predicted by the
linear fit. The value of the decay constant is in good agreement
•
III systems, even the long quaterphenyl-bridged Ph4DN , with 17
bonds connecting the CBU. Most examples of compounds
with bridges of this size found on the literature are Class II MV
systems. [17–19] Figure 4 compares the spectra of Ph4DN•À in two
solvents with the ones of the t-butyl-protected quaterphenyl
quinone radical anion, which show a wide Gaussian-shaped band
having a band maximum that is strongly dependent on solvent,
[20]
as is typical of Class II mixed-valence compounds.
The
maximum of the quinone band moves from 5890 cmÀ1 in
dimethylformamide (DMF) to 7970 cmÀ1 in MeCN, a blue-shift of
2080 cmÀ1 compared with the small 100cmÀ1 red-shift of the
Class III Ph4DN radical anion band maximum between the same
two solvents. Localization of charge occurs because the reorgani-
zation energy increases (through its solvent component lS) and
the electronic coupling decrease with the increase of distance be-
tween the CBU, eventually making the ratio l/2Hab smaller than 1.
with the bN = 0.30 value found both for a series of Class III violene
[22]
radical cations
and for a series of Class III quinone radical
anions.[20]
The calculated (UB3LYP/6-31G*) twiÀst angle at the central
bond of the biphenyl bridge of Ph2DN is 10ꢀ. The twist angle
•
between phenyl rings increases to 16 for Ph3DN À and to an av-
ꢀ
•
erage of 20ꢀ for the longer Ph4DN•À, so there is no reason to sus-
pect that the attenuation of the decay of Eop with distance for
J. Phys. Org. Chem. (2012)
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

A. MONEO ET AL.
10
9.5
9
βN=0.28
8.5
8
NC
5
CN
7.5
7
n
3
7
9
11 13 15 17 19
number of bonds between CN
Figure 6. Comparison of optical spectra of Ph2DN, M2Ph2DN, and
M4Ph2DN radical anions in dimethylformamide (full lines) and of
M4Ph2DN radical anion in MeCN (broken line). The full spectra of
Ph2DN•À is shown in Fig. 3
Figure 5. Plot of ln(Eop/2) versus the number of bonds N connecting
cyano groups for the Class III radical anions Ph1DN to Ph4DN. The line
shown is a fit for the first three points
the 17-bonded Ph4DN À is due to some structural effect. Accord-
•
structure, interestingly more severe in the second band than in
the low-energy one. The absence of vibrational structure could
suggest that the methylated radicals are delocalized Class II
species. However, the widths at half height of the low-energy
ing to superexchange theory, electron tunneling through the
bridge is facilitated by low-energy gaps between the singly occu-
pied molecular orbital (SOMO) of the donor (the reduced CN
group in this case) and the lowest unoccupied molecular
orbital of the bridge. In that case, the scheme of Fig. 1 is no
longer valid because bridge-centered electronic surfaces have to
be taken into account in a 3-stage model. In long-range electron
transfer, as it occurs in DNA or in molecular wires, reduced bridge
states (or oxidized states, in the case of hole-transfer) are effective
intermediates, and an electron hopping process is achieved.
•À
•À
bands of M2Ph2DN and M4Ph2DN , Δn1/2 = 3570 cmÀ1 and
ꢀ
Δn1/2 = 3230 cmÀ1, respectively, are smaller than the minimum
ꢀ
width that Hush predicted for Class II bands, ΔnH1/u2sh = [16RT(ln2)
ꢀ
1/2
•À
. This equation yields ΔnH1/u2sh = 481À2 cmÀ1 for M2Ph2DN
ꢀ
Eop
]
and Δ n H1/u2sh = 3891 cmÀ1 for M4Ph2DN . Moreover, changing
the solvent from DMF to MeCN produces only minor changes
in band maxima, as shown in Fig. 6 for the tetramethylated
radical, indicating that both radicals are in fact delocalized
Class III MV species. The two-state model assumes that
Eop = 2Hab, and the electronic coupling is predicted to depend
on cos2 (θ), where θ is the twist angle between any of the p
orbitals along the conjugation path. Figure 7 shows a linear plot
of Eop versus cos2 (θ) for the radicals of Fig. 6, where θ is the
calculated dihedral angle between the aromatic rings, confirm-
ing the theoretical prediction.
•
ꢀ
Electron hopping is characterized by a very shallow distance
[23,24]
dependence of the electron transfer rate.
The quaterphenyl-
4,4′′′-dicarbonitrile radical anion is Class III and effective electron
transfer never occurs, but we suggest that, in this case, the bridge-
donor energy gap is small enough to produce an electronic
coupling that is higher than what the distance dependence plot
would predict. In fact, it is 120 mV easier to reduce p-quaterphenyl
(E01/2 =À2.28 V vs Ag/AgCl) than p-terphenyl (E01/2 = À2.40 V). [25] A
similar effect was found in the Class III optical spectra of 4,
4′-dinitroazobenzene and 4,4′-dinitrostilbene radical anions.
Although the distance between nitro groups is similar, the Eop
value of the azobenzene-bridged radical is 1.48 times that for the
stilbene-bridge radical. Introduction of the nitrogens in the bridge
lowers the reduction potential for the bridge relative to that of the
hydrocarbon bridge.[11]
Cyclic voltammetry data (Table 1) show that the difference
between the two reduction potentials, ΔEo, varies according to
the predictable change in Hab. All the planar dicyano radical anions
with 1,4-phenylene, 1,8- or 2,6-naphtylene or 1,8-anthracenylene
bridges have large ΔEo values. ΔEo decreases with the distance
between CBU in the oligo-(phenylene)dinitrile series. The longer
In an attempt to induce charge localization, we prepared
the dimethyl-substituted and the tetramethyl-substituted biphe-
nyldicarbonitriles M2Ph2DN and M4Ph2DN. Introduction of the
ortho-methyl groups in the biphenyl bridge will force some twist-
ing around the central bond of biphenyl, which should decrease
the electronic coupling. Theoretical calculations (UB3LYP/6-31G*)
predict a dihedral angle between rings of 10ꢀ for the unsubstituted
15
12
9
Ph2DN radical anion, of 43 for the dimethylated M2Ph2DNÀ and
ꢀ
•
6
of 69ꢀ in the tetramethylated M4Ph2DN radical anion. The neutral
compounds are more twisted than the parent radical anions, with
a dihedral angle of 37ꢀ in Ph2DN and both methylated neutral
compounds M2Ph2DN and M4Ph2DN optimizing with orthogonal
phenyl rings. Clearly, the radical anions allow a higher degree of
steric hindrance in order to increase the resonance between rings.
The optical spectra of these radicals (Fig. 6) show that the
maximum of the low energy band decreases with the degree
of methyl substitution, as predicted by the two-state model, for
which Eop = 2Hab. This is accompanied by a loss of the vibrational
θ
CN
NC
3
0
0
0.2
0.4
0.6
cos2(θ)
0.8
1
Figure 7. Dependence of the energy of the band maximum on the twist
angle θ of the central bond of the biphenyl bridge for Ph2DN (structure
show in the inset), M2Ph2DN and M4Ph2DN radical anions
wileyonlinelibrary.com/journal/poc
Copyright © 2012 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. (2012)

DICYANOAROMATIC RADICAL ANIONS
Ph4DN shows an unresolved wave that probably correspond to
a two-electron reduction. Unfortunately, controlled-potential
electrolysis of this compound was accompanied by the precipita-
tion of a highly insoluble product, which prevented electron
counting. ΔEo values in Table 1 are higher than in dinitro radical
anions with the same bridges. For example, ΔEo = 0.067 V for 4,
4′-dinitrobiphenyl, [26] a much smaller value than the ΔEo = 0.46 V
found for Ph2DN. This shows that the electronic couplings in
dicyano compounds are higher than in dinitro compounds with
similar bridges. The value of ΔEo also decreases with the twist angle
between phenyl rings in the three compounds with biphenyl
bridges, following the decrease in conjugation. A similar effect
was recently found in a series of dicyanobiphenylcyclophanes. [27]
The fact that all the dicyanoaromatic radical anions presented
in this work are charge-delocalized, even the ones with bridges
that are supposed to induce a small electronic coupling is
unexpected. Other organic MV radicals with similar bridges, like
the quaterphenylquinone radical anion of Fig. 4, are charge-
localized. The radical anion of 2,2′-dimethyl-4,4′-dinitrobiphenyl,
which is the dinitro equivalent of M2Ph2DN, is Class II in DMF
The internal component of the reorganization energy in electron
transfer between benzonitrile units is almost half the value
calculated for the transfer between nitrobenzene units. This
difference was already predicted long ago by Formosinho using
the intersecting-state model. [29]
The EPR spectra of the dicyano radical anions studied show
small nitrogen coupling constants and large ring hydrogen cou-
pling constants (see experimental part). The hydrogen atoms of
the biphenyl-bridged Ph2DN•À have hyperfine coupling
constants of 1.79 G and 1.03 G in DMF. Because of the higher
electron-withdrawing power of the nitro group, the hydrogen
coupling constants are 1.18 G and 0.22 G in the parent 4,
4′-dinitrobiphenyl radical anion spectrum in the same solvent.
The higher spin density on the bridges of the dicyano radical
anions, as compared with the parent dinitro radical anions,
forces the polyphenyl-bridges to have smaller twist angles to
accommodate the extra resonance, increasing the electronic
coupling. On the other hand, less spin density (and, therefore, less
charge) on the CBU will make the solvent component of the
reorganization energy smaller, because ls depends strongly on
the change of the dipole moment upon electron transfer.
and MeCN with high l values of 11,000 cmÀ1and 12,800 cmÀ1
,
respectively.[9] Charge-delocalization occurs when l/2Hab < 1,
so these dicyano radicals, as argued below, must have reorgani-
zation energies smaller than the parent dinitroaromatic radicals
and higher electronic couplings. Theoretical calculations show
that the phenyl rings of the dicyano radical anions are less
twisted than in the corresponding dinitro radical anions with
the same bridges, which necessarily implies increased electronic
couplings. At the UB3LYP/6-31G* level, the 4,4′-dinitrobiphenyl
radical anion is calculated to have aÀtwist angle of 19ꢀ, almost
twist angles of 22ꢀ, bigger than the 16ꢀ value found for Ph3DN•À.
The same effect is found for radicals with the methylated
biphenyl bridges, with 2,2′,6,6′-tetramethyl-4,4′-dinitrobiphenyl
radical anion optimizing with a dihedral angle between rings
of 77ꢀ, also bigger than the value of 69ꢀ found for the parent
M2Ph2DN•À. Although these differences seem small, its effect
in the electronic coupling is enhanced because of the depen-
dence of Hab on the cos2 (θ).
CONCLUSIONS
Marcus–Hush analysis of the optical spectra of nine dicyanoaro-
matic radical anions has shown that these mixed-valence species
are charge-delocalized. Theoretical calculations show that the
oligo-(p-phenylene)dicarbonitrile radical anions are less twisted
than the dinitro radical anions with the same bridges, suggesting
that the cyano charge-bearing unit induces higher electronic
coupling energies throughout the bridges. Apparently, the lower
electron-withdrawing power of the cyano group allows a higher
charge density on the bridges, which results in higher electronic
couplings. Estimates show that the vibrational component of
the reorganization energy in the electron transfer between
benzonitrile units is almost half the value calculated for the transfer
between nitrobenzene units. The congruence of these factors
explains the preservation of charge delocalization even in radicals
with bridges that normally induce charge localization.
•
twice the one calculated for Ph2DN . 4,4′′-dinitroterphenyl has
The reorganization energy is normally considered to have a
vibrational component lv, which depends upon the change of
internal coordinates of the atoms, and a solvent component ls,
which accounts for the change in solvation upon electron transfer.
The internal component of the reorganization energy of both
types of compounds can be compared using the four-point
method of Nelsen. [28] l is the energy of the vertical transition from
the ground-state minimum shown in the left side of Fig. 1, and
corresponds to an electron transfer between CBU with no change
in geometry. Estimates of lv are obtained by computing the
energy change between the radical anion calculated with the
geometry of the neutral compound (nÀ) and the relaxed neutral
(n0), and between the neutral calculated at the geometry of the
radical anion (a0) and the relaxed radical anion (aÀ), that is,
lv′ = a0 À aÀ + nÀ À n0. Using benzonitrile and nitrobenzene as
model, CBU yields the values shown below.
EXPERIMENTAL
Terephthalonitrile (Ph1DN) and 4,4′-biphenyldicarbonitrile (Ph2DN) were
purchased from Aldrich and used as received.
Naphthalene-1,8-dicarbonitrile (18NDN)
This compound was prepared from acenaphthenequinone in two steps
using published procedures. [30,31] mp= 239–240 ꢀC (lit.[30] 230–231ꢀC). 1H
NMR (300 MHz; DMSO) d 8.47 (d, 2H, J = 7.6 Hz), 8.35 (d, 2H, J = 6.3 Hz),
7.82 (m, 2H). 13C NMR d 138.5, 135.3, 132.9, 127.8, 127.1, 116.7, 107.0; ultra-
violet–visible (UV–VIS) (MeCN) 323.5 nm (log e = 3.78), 305.5 nm (log = 3.98);
EPR (radical anion, MeCN) a2N = 0.53G, a2H = 6.05G, a2H = 4.35 G,
a2H = 0.14G.
Naphthalene-2,6-dicarbonitrile (26NDN)
2,6-dibromonaphthalene (2.2g, 7.1 mmol) and cuprous cyanide (1.5g,
16.7mmol) were refluxed in dry N-methylpyrrolidinone under nitrogen
atmosphere for 3h. The mixture was cooled, poured into ice-water,
and filtered. The solids were purified by elution with toluene from a column
packed with silica to yield 0.96 g (76%) of a white solid. mp 306–307 ꢀC; 1H
NMR (300 MHz; CDCl3) d 8.30 (d, 2H, J= 1.2Hz), 8.02 (d, 2H, J = 8.7Hz), 7.76
NO2
CN
v' = 4850 cm-1
v' = 2460 cm-1
J. Phys. Org. Chem. (2012)
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

A. MONEO ET AL.
(dd, 2H, J = 8.7 and J= 1.2Hz); 13C NMR d 134.0, 133.5, 129.8, 128.2, 118.2,
112.7; UV–VIS (DMF) 322 nm (log e = 3.43), 306.5 nm (log e = 3.27),
272.5nm (log e = 3.92), 261.5nm (log e = 4.00); EPR (radical anion, MeCN)
a2N = 1.18 G, a2H = 3.37G, a2H = 3.35 G, a2H = 0.87G.
Cyclic voltammetry (CV) and controlled potential electrolysis (CPE)
were performed in Bu4NBF4/DMF (0.1 M) using a Pt wire (CV) or a
Pt-gauze (CPE) as working electrodes. Potentials were measured versus
Fc/Fc+ (internal reference) at 200 mV/s. A potentiostat/galvanostat Radi-
ometer Analytical Voltammetry PST050 VoltaLabW (72 rue d’Alsace
69627 Villeurbanne Cedex Lyon France) equipment was used for VC
and CPE studies.
Anthracene-1,8-dicarbonitril (18ADN)
[32]
Theoretical calculations were performed using Gaussian 03. [37]
1,8-diiodoanthracene
(1g, 2.3mmol) was reacted with CuCN (0.5g,
5.58mmol) as described above to yield 0,45 g (60%) of yellow crystals; mp
[33]
300ꢀC (Lit
300,5ꢀC); 1H NMR (300 MHz; CDCl3), d 9.16 (s, 1H), d 8.64
Acknowledgements
(s, 1H), d 8.30 (dd, 2H, J= 8.7 and J = 1.2Hz), d 8.07 (dd, 2H, J= 6.9 and
J = 1.2Hz), d 7.62 (dd, 2H, J= 8.7 and J = 6.9Hz); UV–VIS (DMF) 406 nm (log
e = 3.72), 385 nm (log e = 3.84), 364 nm (log e = 3.78); EPR (radical anion,
DMF) a2N = 0.28G, aH = 5.33 G, a2H = 4.08 G, a2H = 3.34 G, aH = 3.23G,
a2H = 0.28 G.
We thank Dr. Alexandra Antunes for the use of the 500MHz NMR
facility, Professor Dennis Evans (Purdue University, IN, USA) for
providing a sample of 2,6-dibromonaphthalene, and Professor
José Paulo Farinha for the use of a Shimadzu UV/Vis/NIR
spectrometer. Support by Fundação Para a Ciência e Tecnologia
through its Centro de Química Estrutural and Projects PEst-
OE/QUI/UI0100/2011 and PTDC/QUI-QUI/101433/2008 is gratefully
acknowledged.
Biphenyl-4,4′-dicarbonitrile (Ph2DN)
UV–VIS (DMF) 280 nm (log e = 4.49); EPR (radical anion, DMF) a2N = 0.28 G,
a4H = 1.79 G, a4H = 1.03 G.
Terphenyl-4,4′′-dicarbonitrile (Ph3DN)
REFERENCES
[34]
This compound was prepared using published procedures.
mp >
303–306 ꢀC (Lit[35] 299–300 ꢀC); 1H NMR (300 MHz; CDCl3) 7.75 (AB, 8H),
7.72 (s, 4H); 13C NMR d 144.80, 139.57, 136.81, 132.89, 131.55, 128.10,
127.82, 118.93, 111.55; UV–VIS (DMF) 339.5 nm (log e = 3.76). The complex
EPR spectrum was not interpreted.
[1] R. A. Marcus, J. Chem. Phys. 1956, 24, 966–978.
[2] R. A. Marcus, N. Sutin, Biochim. Biophys. Acta 1985, 811, 265–322.
[3] N. S. Hush, Prog. Inorg. Chem. 1967, 8, 391–444.
[4] N. S. Hush, Coord. Chem. Rev. 1985, 64, 135–157.
[5] M. B. Robin, P. Day, Adv. Inorg. Radiochem. 1967, 10, 247–422.
[6] S. F. Nelsen, A. E. Konradsson, M. N. Weaver, J. P. Telo, J. Am. Chem.
Soc. 2003, 125, 12493–12501.
Quaterphenyl-4,4′′′-dicarbonitrile (Ph4DN)
[7] S. F. Nelsen, M. N. Weaver, J. I. Zink, J. P. Telo, J. Am. Chem. Soc. 2005,
To a stirred solution of 4,4′-dibromobiphenyl (0.42g, 3,2mmol) and tetrakis
(triphenylphosphine)palladium(0) (0.1g) in 10ml of toluene was added
under nitrogen atmosphere 5ml of a 2 M aqueous solution of Na2CO3,
followed by 4-cyanophenylboronic acid (0.5g, 6.8mmol) in 5ml of
methanol. The mixture was vigorously stirred at 80ꢀC for 24 h, filtered
and the powder extracted with chloroform to yield 0.38 g of crude Ph4DN
(78%). The crude solid was purified by elution with toluene from a column
packed with silica to yield 0.24 g (50%) of a white solid. mp 285–287 ꢀC; 1H
NMR (500MHz; Acetone-d6) 7.99 (d, 4H, J= 8.4), 7.93 (d, 4H, J= 8.4Hz), 7.92
(s, 8H); 13C NMR d 144.98, 140.50, 138.44, 132.71, 127.77, 127.75, 127.60,
118.91, 111.08; UV–VIS (DMF) 306 nm (log e = 4.23). The complex EPR
spectrum was not interpreted.
127, 10611–10622.
[8] R. M. Hoekstra, J. P. Telo, Q. Wu, R. M. Stephenson, S. F. Nelsen, J. I.
Zink, J. Am. Chem. Soc. 2010, 132, 8825–8827.
[9] J. P. Telo, S. F. Nelsen, Y. Zhao, J. Phys. Chem. A 2009, 113, 7730–7736.
[10] S. F. Nelsen, K. P. Schultz, J. P. Telo, J. Phys. Chem. A 2008, 112,
12622–12628.
[11] S. F. Nelsen, M. N. Weaver, J. P. Telo, J. Am. Chem. Soc. 2007, 129,
7036–7043.
[12] J. P. Telo, A. Moneo, M. F. N. N. Carvalho, S. F. Nelsen, J. Phys. Chem. A
2011, 115, 10738–10743.
[13] J. P. Telo, A. S. Jalilov, S. F. Nelsen, J. Phys. Chem. A 2011, 115,
3016–3021.
[14] S. F. Nelsen, M. N. Weaver, A. E. Konradsson, J. P. Telo, T. Clark, J. Am.
Chem. Soc. 2004, 126, 15431–15438.
[15] J. P. Telo, M. C. B. L. Shohoji, B. J. Herold, G. Grampp, J. Chem.
Soc., Faraday Trans. 1992, 88, 47–51.
2,2′-dimethylbiphenyl-4,4′-dicarbonitrile (M2Ph2DN)
[16] S. F. Nelsen, M. N. Weaver, Y. Luo, J. V. Lockard, J. I. Zink, Chem.
Phys. 2006, 324, 195–201.
[17] C. Lambert, G. Nöll, J. Am. Chem. Soc. 1999, 121, 8434–8442.
[18] S. V. Rosokha, D.-L. Sun, J. K. Kochi J. Phys. Chem. A 2002, 106,
2283–2292.
[19] S. Barlow, C. Risko, S.-J. Chung, N. M. Tucker, V. Coropceanu, S. C.
Jones, Z. Levi, J.-L. Brédas, S. R. Marder, J. Am. Chem. Soc. 2005,
127, 16900–16911.
2,2′-dimethyl-4,4′-dibromobiphenyl [35] (1 g, 0.29 mmol) was reacted with
CuCN (0.47 g, 5.3 mmol) and the product purified as described previously
[36]
to yield 0.35 g (65%) of a white solid. mp 112–113 ꢀC (lit
113 ꢀC); 1H
NMR (300 MHz; CDCl3) d 7.6 (d, 2H, J = 0.6 Hz), 7.56 (dd, 2H, J = 7.8 Hz,
J = 0.6), 7.17 (d, 2H, J = 8.7), 2.08 (s, 6H); 13 C NMR d 144.63, 137.15,
133.86, 129.84, 129.68, 118.73, 112.27, 19.68. The complex EPR spectrum
was not interpreted.
[20] S. F. Nelsen, M. N. Weaver, J. P. Telo, J. Phys. Chem. A 2007, 111,
10993–10997.
2,2′,6,6′-tetramethylbiphenyl-4,4′-dicarbonitrile (M4Ph2DN)
[21] G. L. Closs, J. R. Miller, Science 1988, 240, 440–447.
[22] S. F. Nelsen, H. Q. Tran, M. A. Nagy, J. Am. Chem. Soc. 1998, 120,
298–304.
[23] O. S. Wenger, Inorg. Chim. Acta 2009, 374, 3–9.
[24] M. Cordes, B. Giese, Chem. Soc. Rev. 2009, 38, 892–901.
[25] K. Meerholzt, J. Heinze, Electrochim. Acta 1996, 41, 1839–1854.
[26] N. A. Macías-Ruvalcaba, J. P. Telo, D. H. Evans, J. Electroanal. Chem.
2007, 600, 294–302.
4,4′-Dibromo-2,2′,6,6′-tetramethylbiphenyl[36] (0.71 g, 1.92 mmol) was
reacted with CuCN (0.41 g, 4.6mmol) and purified as described previously
1
to yield 0.12 g (24%) of a white solid. mp 252–253ꢀC; H NMR (300 MHz;
CDCl3) 7.46 (s, 4H), 2.09 (s, 12H); 13C NMR d 134.41, 136.77, 131.53, 118.95,
111.88, 19.66; UV–VIS (DMF) 283 nm (log e = 3.44), 274 nm (log e = 3.47).
EPR (radical anion, DMF) a2N = 0.34 G, a4H = 1.52G, a12H = 0.69 G.
Radical anions were prepared in vacuum sealed glass cells equipped
with an EPR tube and a quartz optical cell. The concentration of the samples
was determined spectrophotometrically before reduction using the UV
data earlier. Reduction was achieved by contact with 0.2% Na–Hg amalgam.
A large excess of Cryptand [2.2.2] was used to sequester the sodium
counter-ion.
[27] D. Vonlanthen, A. Rudnev, A. Mishchenko, A. Käslin, J. Rotzler,
M. Neuburger, T. Wandlowski, M. Mayor, Chem. Eur. J. 2011, 17,
7236–7250.
[28] S. F. Nelsen, S. C. Blackstock, Y. Kim, J. Am. Chem. Soc. 1987, 109,
677–682.
[29] S. J. Formosinho, J. Chem. Soc. Perkin Trans 2 1988, 1209–1212.
[30] G. Ege, E. Beisiege, Synthesis 1974, 22–23.
wileyonlinelibrary.com/journal/poc
Copyright © 2012 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. (2012)

DICYANOAROMATIC RADICAL ANIONS
[31] J. B. Wright, J. Org. Chem. 1964, 29, 1905–1909.
[32] M. Goichi, K. Segawa, S. Suzuki, S. Toyota, Synthesis 2005,
2116–2118.
[33] H. Woldmann, R. Stengl, Chem. Ber. 1950, 83, 167
[34] M. A. Ismail, R. K. Arafa, R. Brun, T. Wenzler, Y. Mao, W. D. Wilson,
C. Generaux, A. Bridges, J. E. Hall, D. W. Boykin, J. Med. Chem.
2006, 49, 5324–5332.
[35] A. Helms, D. Heiler, G. McLendon, J. Am. Chem. Soc. 1992, 114,
6227–6238.
[36] W. Theilacker, W. Ozegowski, Chem. Ber. 1940, 73, 33–40.
[37] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J. C.
Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci,
M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada,
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.
Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian,
J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E.
Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. Ochterski,
P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg,
V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas,
D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz,
Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M.
W. Gill, B. G. Johnson, W. Chen, M. W. Wong, C. Gonzalez, J. A. Pople,
Gaussian 03, (Revision D01), Gaussian, Inc., Wallingford CT, 2005.
[38] P. D. Astudillo Sánchez, D. H. Evans, J. Electroanal. Chem. 2011, 660,
91–96.
J. Phys. Org. Chem. (2012)
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc