Synthesis of 2-X-, 3-X-4,4′-Dicyanobiphenyls (X = CH3, OCH3, F) by Cross-Coupling of the Terephthalonitrile Dianion with Substituted Benzonitriles

Article abstract of DOI:10.1002/ejoc.201500295

A concise noncatalytic synthetic approach to 2- and 3-substituted 4,4′-dicyanobiphenyls by applying terephthalonitrile dianion 1^2- as para-cyanophenylating reagent for neutral benzonitriles is described. Neutral participants are varied to reveal the scope of applicable substrates and to evaluate the electronic and structural factors governing the regioselectivity and efficiency of the cross-coupling. Benzonitriles substituted with Me, OMe and F provide the corresponding dicyanobiphenyls in good yields. The regularities in the reactivities are interpreted in terms of a reaction scheme involving the intermediacy of a charge-transfer complex between 1^2- and benzonitrile, which transforms into a dimeric dianion either through a heterolytic pathway and/or by successive single electron transfer and recombination of the primary generated radical anions. Quantum chemical calculations of the structure and energy of the principal reaction intermediates match the experimental data. CV measurements of the new dicyanobiphenyls are also included. Transition-metal-free biaryl cross-coupling utilises the terephthalonitrile dianion sodium salt for para-cyanophenylation of aromatic nitriles modified with methoxy, methyl, and fluorine substituents, with the formation of 2- and 3-substituted 4,4′-dicyanobiphenyls. The proposed reaction scheme is verified by quantum chemical calculations.

Full text of DOI:10.1002/ejoc.201500295

FULL PAPER
DOI: 10.1002/ejoc.201500295
Synthesis of 2-X-, 3-X-4,4Ј-Dicyanobiphenyls (X = CH3, OCH3, F) by Cross-
Coupling of the Terephthalonitrile Dianion with Substituted Benzonitriles
Roman Yu. Peshkov,[a,b] Elena V. Panteleeva,*[a,b] Lyudmila N. Shchegoleva,[a]
Irina Yu. Bagryanskaya,[a,b] Tatyana V. Rybalova,[a,b] Nadezhda V. Vasilieva,[a] and
Vitalij D. Shteingarts[a]
Keywords: Cross-coupling / Electron transfer / Nucleophilic substitution / Reactive intermediates / Biaryls
A concise noncatalytic synthetic approach to 2- and 3-substi-
tuted 4,4Ј-dicyanobiphenyls by applying terephthalonitrile
dianion 12– as para-cyanophenylating reagent for neutral
benzonitriles is described. Neutral participants are varied to
reveal the scope of applicable substrates and to evaluate the
ies are interpreted in terms of a reaction scheme involving
the intermediacy of a charge-transfer complex between 12–
and benzonitrile, which transforms into a dimeric dianion
either through a heterolytic pathway and/or by successive
single electron transfer and recombination of the primary
electronic and structural factors governing the regioselectiv- generated radical anions. Quantum chemical calculations of
ity and efficiency of the cross-coupling. Benzonitriles substi- the structure and energy of the principal reaction intermedi-
tuted with Me, OMe and F provide the corresponding dicya- ates match the experimental data. CV measurements of the
nobiphenyls in good yields. The regularities in the reactivit- new dicyanobiphenyls are also included.
We previously reported a new biaryl cross-coupling reac-
Introduction
tion utilising terephthalonitrile (1) dianion (12–) alkali metal
Biaryl cross-coupling is a well known and rapidly evolv-
ing class of organic reactions; its development has been
driven largely by intense interest in the biaryl structural mo-
tif, which is widely applicable in pharmaceuticals and mate-
rial science.[1] Currently, the main synthetic approaches to
biaryl scaffolds are based on transition-metal-catalysed
cross-coupling reactions (Suzuki, Negishi etc.).[2] Although
these reactions are well-recognised and convenient methods
in the laboratory and in fine organic synthesis, they still
have some disadvantages, one of which is the use of expens-
ive catalysts that are difficult to recover and that are diffi-
cult to remove from the final product, which is an aspect
that is extremely important in pharmacy.[3] Another prob-
lem is the application of preactivated reagents such as aryl-
boron acids or aryl-metallic compounds, which, in turn, re-
quire additional reagents and special synthetic conditions,
thus being to some extent uneconomical and environmen-
tally harmful. Thus, the development of new synthetic
methodologies for biaryl C–C bond formation by direct C–
H arylation is an urgent problem that deserves a detailed
investigation.[2b]
salt in liquid ammonia as cyanoarylating reagent for neutral
cyanoarenes. Interaction of 12– with benzonitrile, meta-tolu-
nitrile as well as with 2- and 3-cyanobiphenyls has been
shown to provide 4,4Ј-dicyanobiphenyls[4] and dicyanoter-
phenyls[5] in good yields. This approach involves two-elec-
tron reductive activation of dinitrile 1, retaining the initial
structure of the latter while at the same time enhancing its
activity as an electron donor and/or nucleophile towards
another reaction participant, which undergoes direct C–H
arylation without any preactivation. It is noteworthy that
products combining in their structure a biaryl moiety with
the readily and variably modifiable cyano group[6] provide
a wide range of possibilities for their further transforma-
tions into practically important substances and materi-
als.[1,7] The encouraging results of initial trials gave us
reason to study the scope of the reaction in more detail.
With the aim of expanding the synthetic value of the ap-
proach and to reveal the structural and electronic factors
governing this type of interaction, we varied the nature of
the neutral participants. Benzonitriles with different types
and position of substituent (2-, 3-Me; 2-, 3-OMe; 3-NMe2;
2-, 3-, 4-F; 2-Cl; 3-Br) along with other electron-deficient
arenes (nitrobenzene, 3-methylbenzophenone and ethyl
benzoate) were used to assess the interaction with dianion
12–. For our work, it was important that the behaviour of
the selected nitriles had been explored in detail by electro-
chemical reduction.[8,9] To further validation of the reaction
mechanism suggested earlier[4] (see Scheme 1), the obtained
experimental data were compared with the results of quan-
[a] N.N. Vorozhtsov Novosibirsk Institute of Organic Chemistry of
Siberian Branch of Russian Academy of Sciences,
Ac. Lavrentiev Avenue, 9, Novosibirsk, 630090, Russia
E-mail: pantel@nioch.nsc.ru
http://www.nioch.nsc.ru/
[b] Novosibirsk State University,
Pirogova St., 2, Novosibirsk, 630090, Russia
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201500295.
4524
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 4524–4531

Synthesis of 2- and 3-Substituted 4,4Ј-Dicyanobiphenyls
tum-chemical calculations on the electronic structure and Table 1. para-Cyanophenylation of substituted benzonitriles by 12–.
relative energies of the proposed reaction intermediates. The
combined data are particularly valuable for the develop-
ment of a general synthetic approach that involves the ap-
plication of long-living anionic forms of cyanoarene (radi-
cal anions, dianions and cyclohexadienyl anions) as effec-
tive nucleophilic and electron-donating synthons in reac-
tions involving C–C bond formation.[10]
Scheme 1. The proposed[4] scheme for para-cyanophenylation of
substituted benzonitriles by dianion 12–.
Results and Discussion
Terephthalonitrile dianion (12–)[11] was generated by the
addition of two equivalents of metallic sodium[12] to a sus-
pension of dinitrile 1 in liquid ammonia. The salt of 12–
thus obtained was exposed to a twofold excess of function-
alised benzonitrile 2 for 1.5 h under an inert atmosphere of
evaporating ammonia at –33 °C (these experimental condi-
tions previously[4] provided the highest yield of 4,4Ј-di-
cyanobiphenyl in the reaction of 12– with benzonitrile). We
observed that cross-coupling occurred in the reactions of
12– with benzonitrile substituted by methyl (2-Me, 2a; 3-
Me, 2b), methoxy (2-OMe, 2c; 3-OMe, 2d) groups and by
[a] Calculated as mean values of no less than three runs. Deviation
does not exceed 5%.
a fluorine atom (2-F, 2e; 3-F, 2f; 4-F, 2g), providing the
corresponding 3- or 2-substituted 4,4Ј-dicyanobiphenyls
3a–f and 4,4Ј-dicyanobiphenyl (3g) in the case of 2g (see
Table 1). The interaction of 12– with 3-N,N-dimethylamino-
benzonitrile provided the biaryl product in trace amounts
(ca. 0.5% by GC–MS).
nobenzoyl)benzonitrile and their imino forms, correspond-
In addition to target dicyanobiphenyls 3, reaction mix- ing to the coupling of two 2f molecules as well as of 2f
tures contained starting dinitrile 1 (up to 10%), function- and 1. These compounds presumably arise through bonding
alised benzonitriles 2 (10–50%), small amounts of benzoni- between a carbon ring atom of one molecule and the cyano
trile (up to 5%) as the result of 12– protonation[11b,11c] and group of a second in the highly basic reaction medium, due
dicyanobiphenyl 3g (up to 8% as byproduct in cases of 2a– to the comparatively high C2–H acidity of 3-fluorobenzo-
f) as the product of 12– and benzonitrile cross-coupling. The nitrile 2f.
interaction of 12– with nitrile 2f led to the formation of
Individual dicyanobiphenyls 3a–f were isolated from the
some unidentified products (up to 30% in total) besides the reaction mixtures by fractional sublimation. Their structure
target dicyanobiphenyl 3f. According to NMR and GC– was confirmed by NMR, HRMS, IR and elemental analy-
MS data, these byproducts are likely to have a structure sis. Furthermore, X-ray analysis was performed for di-
such as 3-fluoro-2-(3-fluorobenzoyl)benzonitrile, 2-(4-cya- cyanobiphenyls 3a–c and 3f (see Figure 1).
Eur. J. Org. Chem. 2015, 4524–4531
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4525

E. V. Panteleeva et al.
FULL PAPER
(20–30%). Cross-coupling products were also not formed
in the reactions of 12– with nitrobenzene, 3-methylbenzo-
phenone and ethyl benzoate. Reaction mixtures consisted
of starting materials, some reduction products (ca. 10–20%)
and benzonitrile (see the Supporting Information).
Overall, the results show a strong dependence of the yield
of dicyanobiphenyl 3 on the nature of the neutral reaction
participant. We performed an analysis of the regularities of
the reactivities on the basis of the reaction scheme proposed
earlier[4] for the cross-coupling of 12– with benzonitrile
(Scheme 1). The results imply the formation of charge-
transfer complex (CTC) 4 between 12– and nitrile 2, and its
further transformation into dimeric dianion 5 by successive
steps of electron transfer and recombination as the primary
generated radical-anionic pair (RA-pair) 6 (ET-mechanism)
or by synchronous heterolytic C–C bond formation (SNAr-
like mechanism). Subsequent rapid partial aromatisation of
intermediate 5 by cyano group release makes cross-coupling
irreversible and forms monoanion 7, which is stable under
the reaction conditions. The latter, upon exposure to atmo-
spheric oxygen during reaction work up, forms dicyanobi-
phenyl 3. It should be noted that the ET-pathway may not
lead to the desired product when: 1) the primary cage has
time to relax and release the anion radicals; 2) the radical
anion forming from neutral participant is prone to frag-
mentation; 3) the neutral participant is capable of under-
going successive twofold reduction by 12–, thus leading to
neutral 1 and the stable reduced form of this participant.
As an example of a complete sequence of fluorobenzo-
Figure 1. The molecular structure of 3a–c and 3f. In 3a, the CH3
group is disordered over two positions with a C8/H5Ј to C8Ј/H3
occupancy ratio of 0.541:0.459(3). In 3b, only half of the molecule
is situated in the crystalographically independent part, because a
twofold axis crosses the middle bond of C1–C1A; atom C8 of the
methyl group and atom H2 occupies the same position with C8/H2
occupancy ratio of 0.50:0.50. In 3f, only 0.25 part of the molecule is
independent, because one mirror plane crosses the middle bond
C1–C1A; on the other hand, atoms C1, C4, C5, N1 are situated
on another mirror plane, atoms F1 and H1 occupy the same posi-
tion with H1/F1 occupancy ratio of 0.75:0.25.
The highest yield of dicyanobiphenyls 3 was obtained in
the reactions of 12– with tolunitriles 2a and 2b (Table 1,
entries 1 and 2). In general, product yields rise when going
from 3- to 2-substituted benzonitriles, which is clearly due
to decreased steric hindrance when the substituent recedes
further from the position of the newly formed C–C bond.
Furthermore, product yields increased depending on the
nature of the substituent in benzonitriles 2, in the sequence:
dimethylamino ϽϽ methoxy Ͻ fluorine Ͻ methyl; thus re- nitriles 2e–g, we performed quantum-chemical analysis of
flecting the influence of the substituent electronic effects on the structure and energies of the main anionic intermediates
cross-coupling productivity.
of the proposed scheme by MP2/6-31+G(d) method both
In contrast to the results described above, the interaction in vacuo and in polar solvent. Solvent effect was accounted
of 12– with 2-chloro- (2h) and 3-bromobenzonitrile (2i) did for by using the PCM technique with a model solvent with
not provide biphenyl products but resulted in mixtures con- electrostatic parameters of liquid ammonia. The goal was
taining both starting nitriles 1, 2h or 2i and benzonitrile to establish whether the key intermediate – CTC – is stable
Figure 2. Energy profile for main intermediates of reactions 12– + 2e–g according to MP2/6-31+G(d) calculation (GAMESS US[13]) in
the gas phase (A) and in model solvent with electrostatic parameters of liquid ammonia (B).
4526
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 4524–4531

Synthesis of 2- and 3-Substituted 4,4Ј-Dicyanobiphenyls
in polar media as well as in vacuo. In the gas phase (Fig-
In this way, the energy difference between the intermedi-
ure 2, A) CTCs 4e, 4f, and 4g are 45.7, 46.6 and ates in the reactions of 12– with 2e and 2f follows the experi-
41.7 kcalmol–1, respectively, more stable than separated 12– mentally observed decrease in the product yield. The lower
+ 2e, 12– + 2f, and 12– + 2g systems. Plane-to-plane distance energy or even the absence of 5g does not break the rule
in CTCs is 2.82Ϯ0.01 Å, being almost independent of the because incorporation of nucleophile (i.e., supposed rate-
fluorobenzonitrile structure (Figure 3). The degree of limiting step) into position of C4–F of 2g must go slower
charge transfer in all CTCs is high (0.88Ϯ0.02 e– locates on than into C4–H of 2e and 2f, as is usual for σX- and σH
the fluorobenzonitrile molecule, according to the Mulliken complexes in the classic SNAr mechanism.[14]
population analysis). Transformation of CTCs into dimeric
In general, the calculated relative energy changes for the
dianions in the gas phase decreases the system energy for 5e pair 12– + 2 and the CTC 4 resulting in cross-coupling prod-
(6.9 kcalmol–1) and for 5f (1.6 kcalmol–1). For 5g, energy uct 5 are similar in vacuo and in solvent. The difference
reduction is 18.7 kcalmol–1, which is higher than for 5e and concerns the expressed solvent effect on the relative position
5f, clearly due to stabilising electron-withdrawing effect of of the pair of isolated radical anions 1·– + 2·–. This should
the fluorine atom placed in the sp3-C4 geminal position. be due to the fact that the intermediates under consider-
The most stable conformations for 5e–g are “antiperi- ation are charged and isolated. Solvation energy depends
planar”, with the C1–C4–C4Ј–C1Ј dihedral angle close to on the charge value and on the extent of its delocalization
180°, where negatively charged parts of molecule are the in a molecule; that is, on the size of a molecule. Therefore,
most distant (see the Supporting Information). Subsequent the solvation of pair 12– + 2, which are similar in size, with
cyano group release from 5e and 5f, forms 7e and 7f, respec- the double charge fully located at one molecule is much
tively, and lowers the energy by 49.0 and 50.8 kcalmol–1 stronger than that of pair 1·– + 2·–, which are similar in
respectively, clearly due to partial aromatization and charge both size and charges.
separation, and serves as reaction driving force. Monoanion
As we stated earlier,[4] the formation of dimeric dianions
7g was not found to be a stationary point on the reaction 5 is also possible within the intracomplex ET pathway. The
potential energy surface; optimisation of its geometry led very high degree of charge transfer in CTCs 4e–f (see above)
to fluoride abstraction, forming the 1,4-dicyano-1-(4- favours this assumption. Furthermore, the similarity of RA
cyanophenyl)cyclohexa-2,5-dien-4-yl anion, which, in turn SOMO structure has been revealed for all studied nitriles:
(if it exists), is unstable. The tendency of cyclohexadienyl regardless of the nature of the substituent and its position
anions to undergo fast decyanation or defluorination when in the benzonitrile ring, SOMO is localised mainly at the
they contain a cyano group (Yildiz[11b,11c]) or a fluorine ipso- and para-positions of RA (see Figure 4). This result is
atom (Bartak[9b]) in the sp3 geminal position, is well known. also in close agreement with experimental[8,10] data, which
confirms that the spin density localises in the aromatic
nitrile RA predominantly at the para- and ipso-positions.
The structure of 4,4Ј-dicyanobiphenyls 3a–g corresponds to
coupling at the ipso-position of 12– and the para-position of
neutral substrate and is consistent with the participation of
RAs 1·– and 2a–g·– SOMO.
Figure 3. Geometry of CTC 4e in vacuo (A) and in liquid ammonia
(B).
Calculations using the PCM model (Figure 2, B) show
that CTCs 4e, 4f and 4g are more stable in ammonia than
separated 12– + 2e, 12– + 2f and 12– + 2g by 20.1, 21.0 and
17.6 kcalmol–1, respectively. Typical interplane distance for
all CTCs is 2.73Ϯ0.02 Å, and the degree of charge transfer
is 0.92Ϯ0.01 e–. Transformation of 4e into 5e lowers the
energy by 1.3 kcalmol–1; In contrast, transformation of 4f
into 5f increases the energy by 3.5 kcalmol–1. The geome-
tries of 5e and 5f resemble strongly those calculated in the
gas phase. Subsequent cyano group release in both cases
Figure 4. SOMO plots of substituted benzonitrile RAs. DFT UB3-
decreases the energy by 21.2 and 22.3 kcalmol–1, respec-
LYP/6-31+G(d) (US GAMESS).
tively, and allows the reaction to proceed. In ammonia, we
did not find either dianion 5g or monoanion 7g to be sta-
Combining the calculated data with the experimental re-
tionary points on the potential energy surface, possibly due sults described above leads to the conclusion that there is a
to lower C–F bond dissociation barrier compared with that high probability of a competition between SN-like and ET
in the gas phase, caused by unspecific solvation effects.
mechanisms in the formation of new C–C bonds upon the
Eur. J. Org. Chem. 2015, 4524–4531
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4527

E. V. Panteleeva et al.
FULL PAPER
interaction of dianion 12– and neutral aromatic substrates. tuted benzonitrile. This result is clearly due to the realisa-
This competition depends on the variation of neutral par- tion of the ET mechanism under the interaction of both
ticipant structure and influences the cross-coupling produc- these nitriles with dianion 12–. Reported data[9a] illustrating
tivity, changing the yield of the biphenyl product. It is likely the behaviour of 2h and 2i under electrochemical reduction
that the interaction within both competing mechanisms demonstrates the extremely low stability of their RAs; upon
should be slower when going from unsubstituted benzo- formation, they undergo fast halogen anion release, thus
nitrile to more electropositive tolunitriles 2a and 2b and providing cyanophenyl radical 8, the latter further reduces
even slower in the case of methoxybenzonitriles 2c and 2d. into phenyl anion 9, which undergoes protonation by sol-
This statement is consistent with the experimental results. vent to give benzonitrile (Scheme 3). Reported data also
Electron affinity of fluorinated benzonitriles 2e–g is higher support ET realisation in a dissociative manner providing
than that of the other substrates, and we could expect the radical 8 as the product.[9b,15a,15b] Notably, both benzo-
1C
increase of their reactivity by both competing mechanisms. nitriles 2h and 2i have high reductive potential (2hEp
=
However the yields of fluorinated products 3e–g are be- –1.91, 2iEp1C = –1.95 V, Pt electrode, DMF/0.1 m TEAP, vs.
tween those of 3a,b and 3c,d. Earlier, the competition of SCE).[9a] We suppose the reduction of 2h and 2i and oxid-
ET and SN2 was established for alkylation of dianion 12– ation of 12– into 1·– and then to 1 is fast enough to prevent
and the rise of alkyl halide electron affinity increased the interaction of 12– with just formed benzonitrile; thus, we
contribution of ET to a greater extent than for SN2.[10,11a] found no dicyanobiphenyl 3g as product in these cases. It
Probably, biphenyls 3a,b and 3c,d arise predominantly by is apparent that, in reactions with nitrobenzene, 3-methyl-
heterolytic SNAr-like mechanism. Unlike these, the interac- benzophenone and ethyl benzoate provide only reduced
tion of 12– with fluorinated benzonitriles could occur with forms of starting compounds; 12– also works like a re-
a noticeable ET share.[9b,15] We conducted electrochemical duction mediator.
reduction of dinitrile 1 and fluorobenzonitriles 2e–g under
the same conditions to evaluate the possibility of their one-
electron reduction by dianion 12– (see Table 4 in the Sup-
porting Information for cyclic voltammetry (CV) data). The
obtained reduction potentials are in general agreement with
1C
1C
reported values:[9b,11b] 2eEp
= –2.17, 2fEp
= –2.16,
1C
2gEp = –2.37 V (platinum electrode in DMF/0.1 m TEAP,
vs. SCE, v = 0.1 Vs–1). They are higher than the second
Scheme 3. Reaction pathway for halogenated benzonitriles 2h and
2i.
1C
1
2C
potential of dinitrile 1 (1Ep = –1.59, Ep = –2.39 V),
which indicates the possibility of ET and agrees overall with
the observed variation in dicyanobiphenyl 3e–g yields.
The mechanism of product formation in the case of 4-
fluorobenzonitrile 2g is not clear. It may arise by pathways
described in Scheme 1, with the only distinction being aro-
matization of dimeric dianion 5 proceeding through subse-
quent fast defluorination and decyanation (Scheme 2). An-
other probable pathway is dimerization of radical anion 2g·–.
Bartak[9b] detected this species in the case of electrochemi-
cal reduction of 2g when the concentration was more than
0.5 mm, which is about tenfold less than we used. Clearly, a
combination of all these pathways may take place.
To demonstrate the value of the newly synthesized di-
cyanobiphenyls 3a–f for reductive activation and for utili-
sation as electroactive materials, including as semi-
conductors and single-molecule junctions,[7a,7b] their elec-
trochemical behaviour was studied by CV (see Table 2).
Table 2. Reductive potentials of substituted dicyanobiphenyls at a
platinum electrode in DMF/0.1 m TEAP, vs. SCE.
1C
2C
Entry
Substrate
Ep [V]
Ep [V]
1
2
3
4
5
6
7
3a
3b
3c
3d
3e
3f
–1.68
–1.81
–1.65
–1.75
–1.52
–1.59
–1.63
–2.11
–2.12
–2.07
–2.12
–1.97
–2.00
–2.07
3g
All the new dicyanobiphenyls 3a–f resemble unsubsti-
tuted 4,4Ј-dicyanobiphenyl 3g, demonstrating two one-elec-
tron diffusion-controlled peaks, which are reversible even at
low scan rates (0.01 Vs–1). This fact evidences the high sta-
bility of the radical anions and dianions of the dicyanobi-
phenyls in DMF. Typical for all dicyanobiphenyls, the CV
curve of 3a is presented in Figure 5 (for other data see the
Scheme 2. Formation of 3g.
Amongst the halogenated benzonitriles, going from Supporting Information). In comparison to unsubstituted
fluorinated substrates to 2-chloro- (2h) and 3-bromobenzo- dicyanobiphenyl 3g, the influence of substituent on re-
nitrile (2i) resulted in the absence of biphenyl products to- ductive potential in general is not large (ΔE Յ Ϯ0.15 V).
gether with the appearance of large amounts of unsubsti- The presence of a fluorine atom induces a weak electron-
4528
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 4524–4531

Synthesis of 2- and 3-Substituted 4,4Ј-Dicyanobiphenyls
all the structures to correspond to the local minima at the potential
energy surface. Solvent effect was accounted for by the CPCM
technique using the model solvent with electrostatic parameters of
liquid ammonia (solvent radius 1.9 Å, dielectric constant 21). The
SOMO structure of radical anions of 1, 2a–g and benzonitrile was
calculated by using density functional theory in the spin-unrestric-
ted variant, UB3LYP/6-31+G(d). All the calculations were made
with the GAMESS[13] package.
withdrawing effect, whereas methyl and methoxy groups in-
duce an electron-donating effect, which is expressed to a
greater extent in the case of a methyl group.
CV measurements on compounds 1, 2e–g and 3a–g (2 mm solutions
in DMF) were performed with a PG 310 USB potentiostat (HEKA
Elektronik) at 295 K in an argon atmosphere at a stationary Pt
spherical electrode (S = 0.08 cm2) with 0.1 m Et4NClO4 as a sup-
porting electrolyte. The potential sweep rate was 0.1 Vs–1. A stan-
dard electrochemical cell (solution volume was 5 mL) connected
to the potentiostat with a three-electrode scheme was used. Peak
potentials were quoted with reference to a saturated calomel elec-
trode (SCE). All reduction peaks for the compounds 3a–f were dif-
Figure 5. CV curve for 3a.
fusion-controlled, i.e., Ip1C,2Cv–0.5 = const, where Ip is the peak
1C
current.
Conclusions
Liquid ammonia was purified just before use with metallic sodium,
followed by distillation into a reaction vessel, cooled to –80 to
–70 °C. Metallic sodium was freed from oxide film under dry hex-
ane. Commercial terephthalonitrile was purified by sublimation.
Substituted benzonitriles 2a–g were purchased (ABCR, Alfa Aesar
or Sigma-Aldrich) and used without further purification. Melting
points of dicyanobiphenyls 3a–f are uncorrected.
Convenient metal-free biaryl cross-coupling based on the
treatment of a terephthalonitrile dianion with neutral 2-
and 3-substituted benzonitriles is applied for synthesis of
modified 4,4Ј-dicyanobiphenyls. Benzonitriles with electron
affinity close to that of unsubstituted benzonitrile, comply
with such types of cross-coupling. The newly obtained 2-
or 3-methyl, -methoxy and -fluoro substituted 4,4Ј-di-
cyanobiphenyl derivatives are valuable structural blocks for
organic synthesis that are free from transition-metal impuri-
ties. These products are suitable for use in the the design of
new electroactive materials that can also be used in covalent
triazine-based metal-organic frameworks and polymers.[7]
The products have been shown to form stable radical anions
and dianions, thus making them suitable for application in
transformations based on their reductive activation.
General Procedure for the Synthesis of Dicyanobiphenyls 3a–f: Met-
allic sodium (237 mg, 10.3 mmol) was added to a stirred suspension
of dinitrile 1 (640 mg, 5.0 mmol) in liquid ammonia (30–40 mL) at
–33 °C under an atmosphere of evaporating ammonia. The mixture
was kept for 5 min, thus providing a dark-brown suspension of
dianion 12– salt. Substituted benzonitrile 2a–g (10.0 mmol) was
added to a stirred suspension of 12– salt and the reaction mixture
was stirred for 1.5 h at –33 °C (during this time ammonia evapo-
rated by half). Et2O (ca. 20 mL) was added and the reaction mix-
ture was put into contact with air. Stirring was continued until
ammonia evaporated completely and room temperature was
reached. Water (ca. 20 mL) was poured onto the residue to dissolve
inorganic salts. The solid product was filtered off, washed with
Et2O and water, dried in air and analysed by 1H NMR spec-
troscopy. Products from the liquid fraction were extracted with
Et2O (3ϫ 30 mL). The combined organic extract was dried with
MgSO4, filtered and the solvent was removed. The crude residue
Experimental Section
General: 1H, 13C and 19F NMR spectra of all compounds were
acquired with a Bruker Avance-III 500 instrument at 500.13, 125.76
and 470.59 MHz, respectively, in [D6]acetone or [D6]DMSO; chem-
ical shifts (δ) of 1H and 13C are given in ppm relative to TMS using
1
was analysed by H NMR spectroscopy and GC-MS analysis. The
products from the solid fraction were isolated by fractional subli-
mation in vacuo (100–200 °C/2–5 Torr). To isolate product 3a–f
from the liquid fraction, the volatile residues (benzonitrile and 2a–
f) were evaporated under reduced pressure, then the crude product
was sublimated in vacuo.
the solvent signals as the internal standard (δH = 2.05 ppm, δC
=
29.8 and 206.3 ppm for [D6]acetone, δH = 2.50 ppm, δC = 39.5 ppm
for [D6]DMSO), the internal standard for 19F spectra was C6F6 (δ
= –162.9 ppm). Signal assignment and structure justification were
carried out based on the HSQC and HMBC data and on analysis
of the C–F spin–spin coupling constants. IR spectra were recorded
with a Vector-22 instrument for samples pelleted with KBr
(0.25%). GC–MS analysis was performed with a Hewlett–Packard
G1081A instrument consisting of an HP-5890 Series II gas chroma-
tograph and an HP-5971 mass-selective detector (IE, 70 eV) with
an HP5 capillary column. The precise molecular ion weights were
determined with a DFS instrument. Elemental analysis was carried
out with a Carlo Erba automatic C, H, N-analyzer model 1106.
1
4,4Ј-Dicyano-3-methylbiphenyl (3a): M.p. 166–167 °C (ethanol). H
NMR (500.13 MHz, [D6]acetone): δ = 7.94 (d, J = 8.5 Hz, 2 H, 2Ј-
H and 6Ј-H), 7.91 (d, J = 8.5 Hz, 2 H, 3Ј-H and 5Ј-H), 7.84 (br. s,
1 H, 2-H), 7.83 (d, J = 8.2 Hz, 1 H, 5-H), 7.74 (dd, J = 8.0, 1.8 Hz,
1 H, 6-H), 2.61 (s, 3 H, CH3) ppm. 13C NMR (125.76 MHz, [D6]-
acetone): δ = 144.4 (1 C, 1Ј-C), 144.1 (1 C, 1-C), 143.5 (1 C, 3-C),
134.0 (1 C, 5-C), 133.7 (2 C, 3Ј-C and 5Ј-C), 130.0 (1 C, 2-C), 129.0
(2 C, 2Ј-C and 6Ј-C), 126.2 (1 C, 6-C), 119.1 (4Ј-C-CN), 118.3 (4-
The relative energies of the main reaction intermediates were calcu-
lated by using the second order Moller–Plesset perturbation theory,
MP2/6-31+G(d) (UMP2 for the open-shell species). Stationary
structures of the intermediates were located and their type was de-
termined from the analysis of normal vibrations. The latter show
C-CN), 113.5 (4-C), 113.0 (4Ј-C), 20.5 (CH ) ppm. IR (KBr): ν =
˜
3
2225 (CϵN) cm–1. MS (EI): m/z (%) = 218 (100) [M+], 190 (35),
164 (6). HRMS: m/z calcd. for C15H10N2 [M+] 218.0838; found
218.0837. C15H10N2 (218.26): calcd. C 82.55, H 4.62, N 12.84;
found C 81.73, H 5.17, N 12.55.
Eur. J. Org. Chem. 2015, 4524–4531
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4529

E. V. Panteleeva et al.
4,4Ј-Dicyano-2-methylbiphenyl (3b): M.p. 133–134 °C (acetone). 1H C), 121.1 (d, J = 26.4 Hz, 1 C, 3-C), 119.0 (4Ј-C-CN), 117.9 (s, 1
FULL PAPER
NMR (500.13 MHz, [D6]acetone): δ = 7.90 (d, J = 8.4 Hz, 2 H, 3Ј-
H and 5Ј-H), 7.75 (s, 1 H, 3-H), 7.70 (d, J = 8.0 Hz, 1 H, 5-H),
C, 4-C-CN), 114.7 (d, J = 9.9 Hz, 4-C), 113.6 (s, 1 C, 4Ј-C) ppm.
19F NMR (470.59 MHz, [D6]acetone): δ = 48.56 (br. t, J ≈ 9.3 Hz,
7.62 (d, J = 8.4 Hz, 2 H, 2Ј-H and 6Ј-H), 7.46 (d, J = 8.0 Hz, 1 H, 1 F, F2) ppm. IR (KBr): ν = 2237 (CϵN) cm–1. MS (EI): m/z (%)
˜
6-H), 2.32 (s, 3 H, CH3) ppm. 13C NMR (125.76 MHz, [D6]acet-
= 222 (100) [M+], 195 (10). HRMS: m/z calcd. for C15H10N2 [M+]
one): δ = 145.7 (2 C, 1-C and 1Ј-C), 138.0 (1 C, 2-C), 134.9 (1 C,
222.0588; found 222.0589. C14H7FN2 (222.22): calcd. C 75.67, H
3-C), 133.2 (2 C, 3Ј-C and 5Ј-C), 131.3 (1 C, 6-C), 130.8 (1 C, 2Ј- 3.18, F 8.55, N 12.61; found C 75.66, H 3.18, F 8.49, N 12.53.
C and 6Ј-C), 130.6 (1 C, 5-C), 119.2 (4-C-CN), 119.1 (4Ј-C-CN),
Supporting Information (see footnote on the first page of this arti-
112.9 (4-C), 112.6 (4Ј-C), 20.2 (CH ) ppm. IR (KBr): ν = 2226
˜
3
cle): Experimental procedures for reactions of 12– with nitroben-
zene, ethyl benzoate and 3-methylbenzophenone, 1H and 13C NMR
spectra of compounds 3a–f, detailed information about crystal
structure of 3a–c and 3f, cyclic voltammetry of benzonitriles 2e–g
and dicyanobiphenyls 3a–g, optimised geometries of main reaction
intermediates (Figure 2) and radical anions of 1, 2a–g and benzo-
nitrile are presented.
(CϵN) cm–1. MS (EI): m/z (%) = 218 (100) [M+], 203 (9), 190 (58),
178 (10), 140 (8), 95 (12), 81 (14). HRMS: m/z calcd. for C15H10N2
[M+] 218.08449; found 218.08434. C15H10N2 (218.08): calcd. C
82.55, H 4.62, N 12.84; found C 82.23, H 4.70, N 12.77.
4,4Ј-Dicyano-3-methoxybiphenyl (3c): M.p. 198–199 °C (sublima-
tion). 1H NMR (500.13 MHz, [D6]DMSO): δ = 7.99 (d, J = 8.8 Hz,
2 H, 3Ј-H and 5Ј-H), 7.97 (d, J = 8.8 Hz, 2 H, 2Ј-H and 6Ј-H),
7.83 (d, J = 8.0 Hz, 1 H, 5-H), 7.52 (d, J = 1.4 Hz, 1 H, 2-H), 7.44
(dd, J = 8.0, 1.4 Hz, 1 H, 6-H), 4.02 (s, 3 H, OCH3) ppm. 13C
NMR (125.76 MHz, [D6]DMSO): δ = 161.3 (1 C, 3-C), 144.8 (1 C,
1-C), 142.9 (1 C, 1Ј-C), 134.2 (1 C, 5-C), 132.9 (2 C, 3Ј-C and 5Ј-
C), 128.2 (2 C, 2Ј-C and 6Ј-C), 119.8 (1 C, 6-C), 118.6 (1 C, 4Ј-C-
CN), 116.2 (1 C, 4-C-CN), 111.4 (1 C, 4Ј-C), 110.9 (1 C, 2-C), 100.3
Acknowledgments
The authors are grateful to Dr. Ilya V. Eltsov (Novosibirsk State
University) for performing NMR measurements. The centre of col-
lective using (CCU) of the Novosibirsk Institute of Organic Chem-
istry of SB RAS is acknowledged. Financial support of the Chemis-
try and Material Sciences Department of RAS (project “Utilization
of long-living anionic forms of cyanoarenes as versatile synthons in
design of building blocks for new materials and biologically active
compounds”) is acknowledged.
(1 C, 4-C), 56.5 (1 C, OCH3) ppm. IR (KBr):
ν = 2224
˜
(CϵN) cm–1. MS (EI): m/z (%) = 234 (100) [M+], 205 (44), 191
(27), 177 (10), 164 (21), 140 (8). HRMS: m/z calcd. for C15H10N2O
[M+] 234.0788; found 218.0789. C15H10N2O (234.25): calcd. C
76.91, H 4.30, N 11.96; found C 76.02, H 4.35, N 11.84.
4,4Ј-Dicyano-2-methoxybiphenyl (3d): M.p. 187–188 °C (ethanol).
1H NMR (500.13 MHz, [D6]acetone): δ = 7.84 (d, J = 8.5 Hz, 2 H,
3Ј-H and 5Ј-H), 7.77 (d, J = 8.5 Hz, 2 H, 2Ј-H and 6Ј-H), 7.57 (d,
J = 7.8 Hz, 1 H, 6-H), 7.53 (d, J = 1.5 Hz, 1 H, 3-H), 7.47 (dd, J
= 7.8, 1.5 Hz, 1 H, 5-H), 3.95 (s, 3 H, OCH3) ppm. 13C NMR
(125.76 MHz, [D6]acetone): δ = 157.9 (1 C, 2-C), 142.6 (1 C, 1Ј-C),
134.4 (1 C, 1Ј-C), 132.8 (2 C, 3Ј-C and 5Ј-C), 132.4 (1 C, 6-C),
[1] a) J.-P. Corbet, G. Mignani, Chem. Rev. 2006, 106, 2651–2710;
b) K. L. Atkinson, S. M. Morris, M. M. Qasim, F. Castles, D. J.
Gardiner, P. J. W. Hands, S. S. Choi, W.-S. Kim, H. J. Coles,
Phys. Chem. Chem. Phys. 2012, 14, 16377–16385; c) T. R.
Cook, Y.-R. Zheng, P. J. Stang, Chem. Rev. 2013, 113, 734–777;
ˇ
d) T.-H. Chen, I. Popov, O. Zenasni, O. Daugulis, O. S. Mil-
janic´, Chem. Commun. 2013, 49, 6846–6848.
131.3 (2 C, 2Ј-C and 6Ј-C), 119.3 (1 C, 4Ј-C-CN), 119.1 (1 C, 4-C- [2] a) D. Alberico, M. E. Scott, M. Lautens, Chem. Rev. 2007, 107,
174–238; b) A. Lei, H. Zhang, in: C–H and C–X Bond Func-
tionalization (Ed.: X. Ribas), RSC Publishing, Cambridge,
2013, p. 310–327; c) J. Qiu, L. Wang, M. Liu, Q. Shen, J. Tang,
Tetrahedron Lett. 2011, 52, 6489–6491; d) N. Liu, Z.-X. Wang,
J. Org. Chem. 2011, 76, 10031–10038; e) F. M. Piller, P. Ap-
pukkuttan, A. Gavryushin, M. Helm, P. Knochel, Angew.
Chem. Int. Ed. 2008, 47, 6802–6806; Angew. Chem. 2008, 120,
6907; f) H.-Q. Do, O. Daugulis, J. Am. Chem. Soc. 2008, 130,
1128–1129; g) T. Truong, J. Alvarado, L. D. Tran, O. Daugulis,
Org. Lett. 2010, 12, 1200–1203.
CN), 116.0 (1 C, 3-C), 114.2 (1 C, 4-C), 112.5 (1 C, 4Ј-C), 56.7 (1
C, OCH ) ppm. IR (KBr): ν = 2228 (CϵN) cm–1. MS (EI): m/z (%)
˜
3
= 234 (100) [M+], 219 (27), 206 (8), 192 (8), 164 (25), 140 (7).
HRMS: m/z calcd. for C15H10N2O [M+] 234.0788; found 234.0786.
C15H10N2O (234.26): calcd. C 76.91, H 4.30, N 11.96; found C
76.98, H 4.26, N 12.04.
4,4Ј-Dicyano-3-fluorobiphenyl (3e): M.p. 233 °C (ethanol). 1H
NMR (500.13 MHz, [D6]acetone): δ = 8.02–7.93 (m, 5 H), 7.85–
7.81 (m, 2 H) ppm. 13C NMR (125.76 MHz, [D6]acetone): δ =
164.4 (d, JF = 226 Hz, 1 C, 3-C), 147.4 (d, JF = 8.3 Hz, 1 C, 1-C),
143.1 (d, JF = 2.0 Hz, 1 C, 1Ј-C), 135.4 (s, 1 C, 6-C), 133.9 (s, 2 C,
3Ј-C and 5Ј-C), 129.3 (2Ј,6Ј-C), 125.1 (d, JF = 3.2 Hz, 5-C), 119.0
(4Ј-C-CN), 116.1 (d, JF = 21.0 Hz, 1 C, 2-C), 114.4 (s, 1 C, 4-CN),
113.9 (s, 1 C, 4Ј-C), 101.8 (d, JF = 15.7 Hz,1 C, 4-C) ppm. 19F
NMR (470.59 MHz, [D6]acetone): δ = 55.84 (dd, JF,H-ortho = 10.3,
[3] EMEA/CHMP/SWP/4446/2000 (final), Guideline on the speci-
fication limits for residues of metal catalysts or metal reagents.
Committee for Medicinal Products for Human Use (CHMP),
London, 21 February 2008, http://www.ema.europa.eu/docs/
en_GB/document_library/Scientific_guideline/2009/09/
WC500003586.pdf, accessed on September 29, 2011.
[4] E. V. Panteleeva, L. N. Shchegoleva, V. P. Vysotsky, L. M. Pok-
rovsky, V. D. Shteingarts, Eur. J. Org. Chem. 2005, 12, 2558–
2565.
[5] E. V. Panteleeva, I. Y. Bagryanskaya, G. E. Sal’nikov, V. D.
Shteingarts, ARKIVOC (Gainesville, FL, U.S.) 2011, 8, 123–
133.
[6] a) The Chemistry of the Cyano Group (Ed.: Z. Rappoport), In-
terscience Publishers, 1970; b) Comprehensive Organic Trans-
formations (Ed.: R. C. Larock), Wiley-VCH, Weinheim, Ger-
many, 1999.
JF,H-meta = 6.8 Hz, 1 F, 3-F) ppm. IR (KBr): ν = 2228 (CϵN) cm–1.
˜
MS (EI): m/z (%) = 222 (100) [M+], 195 (10). HRMS: m/z calcd.
for C15H10N2 [M+] 222.0588; found 222.0584. C14H7FN2 (222.22):
calcd. C 75.67, H 3.18, F 8.55, N 12.61; found C 74.83, H 3.25, F
8.94, N 12.66.
4,4Ј-Dicyano-2-fluorobiphenyl (3f): M.p. 249–251 °C (sublimation);
1H NMR (500.13 MHz, [D6]acetone): δ = 7.94 (d, J = 8.7 Hz, 2 H,
3Ј-H and 5Ј-H), 7.87–7.76 (m, 5 H, 3,5,6,2Ј,6Ј-H) ppm. 13C NMR
(125.76 MHz, [D6]acetone): δ = 160.1 (d, J = 250 Hz, 1 C, 2-C),
139.5 (d, J = 1.5 Hz, 1 C, 1Ј-C), 133.5 (s, 2 C, 3Ј-C and 5Ј-C), 133.2
(d, J = 13.6 Hz, 1 C, 1-C), 133.1 (d, J = 3.5 Hz, 1 C, 6-C), 131.0
(d, J = 3.2 Hz, 2 C, 2Ј-C and 6Ј-C), 129.9 (d, J = 4.0 Hz, 1 C, 5-
[7] a) D. Vonlanthen, A. Rudnev, A. Mishchenko, A. Käslin, J.
Rotzler, M. Neuburger, T. Wandlowski, M. Mayor, Chem. Eur.
J. 2011, 17, 7236–50; b) A. Mishchenko, L. a. Zotti, D. Von-
lanthen, M. Bürkle, F. Pauly, J. C. Cuevas, M. Mayor, T. Wand-
lowski, J. Am. Chem. Soc. 2011, 133, 184–187; c) M. J. Bojdys,
4530
www.eurjoc.org
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2015, 4524–4531

Synthesis of 2- and 3-Substituted 4,4Ј-Dicyanobiphenyls
S. a. Wohlgemuth, A. Thomas, M. Antonietti, Macromolecules
2010, 43, 6639–6645; d) K. Sumida, M. L. Foo, S. Horike, J. R.
Long, Eur. J. Inorg. Chem. 2010, 3739–3744; e) S. Ren, M. J.
Bojdys, R. Dawson, A. Laybourn, Y. Z. Khimyak, D. J. Ad-
ams, A. I. Cooper, Adv. Mater. 2012, 24, 2357–2361.
[12]
[13]
We found that the alkali metal variation (Li, Na, K) did not
significantly influenced the product yields, probably due to the
same type of ionic association in salts of the reaction interme-
diates. Sodium proved to be the most convenient.
M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert,
M. S. Gordon, J. H. Jensen, S. Koseki, N. Matsunaga, K. A.
Nguyen, S. Su, T. L. Windus, M. Dupuis, J. A. Montgomery,
J. Comput. Chem. 1993, 14, 1347–1363.
[8] a) P. H. Rieger, G. K. Fraenkel, J. Chem. Phys. 1962, 37, 2795;
b) P. H. Rieger, I. Bernal, W. H. Reinmuth, G. K. Fraenkel, J.
Am. Chem. Soc. 1963, 85, 683–693; c) A. R. Buick, T. J. Kemp,
G. T. Neal, T. J. Stone, J. Chem. Soc. A 1969, 666–669.
[9] a) D. E. Bartak, K. J. Houser, B. C. Rudy, M. D. Hawley, J.
Am. Chem. Soc. 1972, 94, 7526–7530; b) K. J. Houser, D. E.
Bartak, M. D. Hawley, J. Am. Chem. Soc. 1973, 95, 6033–6040;
c) M. D. Koppang, N. F. Woolsey, D. E. Bartak, J. Am. Chem.
Soc. 1984, 106, 2799–2805.
[10] a) T. A. Vaganova, E. V. Panteleeva, V. D. Shteingarts, Russ.
Chem. Bull. 2008, 57, 768–779; b) T. A. Vaganova, E. V. Pante-
leeva, V. D. Shteingarts, Russ. Chem. Rev. 2008, 77, 601–619.
[11] a) I. I. Bilkis, E. V. Panteleeva, A. P. Tananakin, V. D. Shtein-
garts, Russ. J. Org. Chem. 1997, 33, 652–659; b) A. Yildiz, M.
Sertel, Electrochim. Acta 1986, 31, 1287–1292; c) A. Yildiz, M.
Sertel, R. Gambert, H. Baumgertel, Electrochim. Acta 1988,
33, 169–170.
[14] a) M. Makosza, Chem. Soc. Rev. 2010, 39, 2855–68; b) F. Ter-
rier, Modern Nucleophilic Aromatic Substitution, Wiley-VCH,
Weinheim, Germany, 2013.
[15] a) M. R. Asirvatham, M. D. Hawley, J. Am. Chem. Soc. 1975,
97, 5024–5026; b) N. V. Efremova, L. V. Podryvanova, N. V.
Vasil’eva, V. F. Starichenko, V. D. Shteingarts, Russ. J. Org.
Chem. 2003, 39, 212–218; c) N. V. Efremova, V. F. Starichenko,
V. D. Shteingarts, Izvestiya Akademii Nauk SSSR Seriya Khim-
icheskaya 1986, 12, 2794–2796; d) V. F. Starichenko, L. N.
Shchegoleva, N. V. Efremova, V. D. Shteingarts, V. O. Saik,
P. V. Schastnev, Chem. Phys. 1985, 100, 79–87.
Received: March 5, 2015
Published Online: June 10, 2015
Eur. J. Org. Chem. 2015, 4524–4531
© 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
4531