6,6-Dicyanopentafulvenes: Teaching an old dog new tricks

Personal Account

DOI: 10.1002/tcr.201402060

T H E

C H E M I C A L

R E C O R D

6,6-Dicyanopentafulvenes: Teaching an

Old Dog New Tricks

Aaron D. Finke^[a]and François Diederich*^[a]

^[a]Laboratory of Organic Chemistry, ETH Zurich, Vladimir-Prelog-Weg 3, 8093 Zurich

(Switzerland), E-mail: diederich@org.chem.ethz.ch, Fax: (+41) 44-632-1109

Received: July 8, 2014

Published online: ■■

Chem. Rec. 2014, ••, ••–••

Wiley Online Library

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

ABSTRACT: 6,6-Dicyanopentafulvene (DCF) is a fascinating molecular entity that consists of a

cyclopentadiene ring conjugated to an exocyclic double bond bearing two cyano groups on its

periphery. Herein, we give a brief history of the chemistry of DCFs prior to our arrival to the ﬁeld

in 2011, followed by a summary of our work. We show how substitution on the ring and the

exocyclic bond affects the HOMO and LUMO energies of pentafulvenes and how the design of

DCFs was exploited computationally for the ﬁrst time. Shortly after the report of the ﬁrst rational

synthesis of DCFs, we discovered that DCFs had a vast and astonishing array of reactivities to form

new molecular entities. Simple, catalyst-free reactions between DCF acceptors and electron-rich

donors led to the formation of scaffolds of exceptional complexity. Furthermore, our discovery that

DCFs are capable of undergoing mild pentafulvene-to-benzene rearrangements challenges previous

conventions of fulvene chemistry.

Keywords: chromophores, dicyanopentafulvenes, donor–acceptor systems, radical ions,

rearrangement

1. Introduction

Pentafulvene (C₆H₆) is a nonvalence isomer of benzene that

displays a rich and fascinating diversity of chemistry. However,

after the heydays of pentafulvene chemistry between 1960 and

1980, the scaffold was largely ignored.^[1]Recently, there has

been a small but vigorous resurgence in interest in pentafulvene

chemistry, as the application of sophisticated computational

techniques unraveled the electronic properties of pentafulvenes

and their potential applicability as functional molecular mate-

rials in devices. Particular to their use in electronic devices, it

was found that the HOMO–LUMO gap of pentafulvenes was

much more susceptible to tuning by chemical substitution than

the gap in aromatic compounds.

Fig. 1. a) The structure of DCF with the pentafulvene numbering scheme. b)

HOMO and c) LUMO of DCF (B3LYP/TZVP).^[3]

In recent years, we have embarked on a research program

to develop new organic donor–acceptor chromophores and

explore their optoelectronic properties that result from intra-

molecular charge-transfer (CT) interactions.^[2]We found it

appealing to electronically tune pentafulvenes towards strong

electron-accepting properties because there is a dearth of

organic acceptors relative to the number of their donor coun-

terparts. At the same time, we were looking for new modes of

reactivity and ultimately access to new materials unobtainable

by other methods. Pentafulvene-based compounds are one of

the most surprising and richly varied scaffolds we have studied.

Substitution with strong electron acceptors at the exocy-

clic 6-position with, for example, cyano groups, leads to a

dramatic lowering of the LUMO energy, while having less of an

effect on the HOMO.^[3]Such 6,6-dicyanopentafulvenes

(DCFs) have had a recent resurgence in interest for that reason:

this property makes them excellent electron acceptors. The

frontier orbitals of pentafulvene are at the origin of its remark-

able properties. A depiction of the orbitals of unsubstituted

DCF (dipole moment 0.60 D) is shown in Figure 1.^[3]

is dispersed throughout the cyclopentadiene ring. It is clear

from the orbital depiction where this dipole moment origi-

nates; there is no HOMO isosurface on C(6), while the

LUMO is spread throughout. We have calculated the effects of

substitution at C(6). Substitution with cyano groups leads to a

dramatic lowering of the LUMO energy by 2 eV, with a low-

ering of the HOMO by only about 1 eV. Unsurprisingly, sub-

stitution on the cyclopentadiene ring primarily affects the

HOMO. Therefore, we can tune the HOMO–LUMO gap of

DCFs by substitution at the ring.^[3,4]

1.1. The History of DCFs Before 2011

The ﬁrst preparation of a DCF was reported in 1974, wherein a

dicyanovinyl–molybdenum complex reacted with two equiva-

lents of diphenylacetylene to give a green solid, which turned

out to be tetraphenyl-DCF 1 (Figure 2).^[5]However, this is the

only example, to date, for the preparation of a DCF starting

from a metal complex, and the synthetic potential of such

methods remains unexplored, despite the well-known analo-

gous preparation of cyclopentadienones from alkynes and

metal carbonyls (e.g., in the formation of the

tetraphenylcyclopentadienone–Ru dimer by Shvo et al.).^[6]

The unsubstituted pentafulvene moiety displays a dipole

moment of 0.44 D due to its nonbenzenoid connectivity.^[1c]

The positive charge resides on C(6), while the negative charge

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

Fig. 2. The ﬁrst reported preparation of stable DCF 1.^[5]

Fig. 3. The preparation of 1 by Lehnert’s TiCl₄/pyridine Knoevenagel condensation conditions. Given are the

electrochemical data, recorded in CH₂Cl₂containing 0.1 m nBu₄PF₆versus the ferrocene couple (Fc/Fc⁺), and the

maximum of the longest-wavelength band, λ_max, in CH₂Cl₂.^[12]

In the mid-1980s, Junek et al. reported that, surprisingly,

1,2,3,4-tetrachlorocyclopentadiene reacted with tetracyano-

ethylene (TCNE) through a metathesis-type transformation to

henylcyclopentadienone and malononitrile failed.^[9]This

appeared to be the death knell for DCF chemistry,

and for nearly 20 years no reports on DCFs were published.

After synthetic chemists abandoned DCFs, computational

chemists picked up interest in them in the early 2000s. In

2006, Michl and co-workers proposed that certain benzo-fused

DCF derivatives might have acted as efﬁcient singlet ﬁssion

materials in organic solar cells, leading to charge-separated

states; however, the proposed compounds have yet to be pre-

pared.^[10]Ottosson et al. have led the ﬁeld in the computational

form 1,2,3,4-tetrachloro-DCF

2

(not shown) and

malononitrile.^[7]However, in our hands this reaction was sadly

found not to be general for cyclopentadienes. Later, they inves-

tigated the reactivity of 2 with anilines to generate push–pull

dyes.^[8]Katritzky et al. reported the preparation of DCF-type

dyes through various chemistries, but stressed that the

preparation of 1 by Knoevenagel condensation of tetrap-

Aaron Finke (born 1984) grew up in Las

Vegas, NV, and studied chemistry at the

University of Arizona (2002–2006). He

received his PhD in chemistry (2006–

2011) from the University of Illinois,

Urbana-Champaign, working in the

laboratory of Prof. Jeffrey S. Moore.

Afterwards, he moved to the laboratory

of Prof. François Diederich as a NSF-

IRFP Postdoctoral Fellow (2011–2014).

He is currently a Postdoctoral Fellow at the Swiss Light

Source, Paul Scherrer Institute. His research interests include

the physical organic chemistry of conjugated molecules and

X-ray crystallography.

François Diederich was born in the

Grand-Duchy of Luxemburg (1952)

and studied chemistry at the University

of Heidelberg (1971–1977). He joined

Prof. H. A. Staab at the Max-Planck-

Institut für Medizinische Forschung in

Heidelberg for his doctoral dissertation

(1977–1979). After postdoctoral studies

with Prof. O. L. Chapman at UCLA (1979–1981), he

returned to Heidelberg for his Habilitation (1981–1985).

Subsequently, he joined the Faculty in the Department of

Chemistry and Biochemistry at UCLA, where he became Full

Professor in 1989. Since 1992, he has been Professor in the

Laboratory of Organic Chemistry in the Department of

Chemistry and Applied Biosciences at ETH Zurich.

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

Fig. 4. Preparation of push–pull DCF 4 by oxidative ring-closure of a bis-donor-substituted pentacyanohexatriene

3E/Z with ceric ammonium nitrate^[15]or wet silica gel.^[16]Given are the electrochemical data, recorded in CH₂Cl₂

containing 0.1 m nBu₄PF₆versus Fc/Fc⁺, and the maximum of the longest-wavelength band, λ_max, in CH₂Cl₂.^[16]

Fig. 5. Preparation of tetraalkynyl-DCF 5 via the corresponding cyclopentadienone.^[18,19]Given are the electrochemi-

cal data, recorded in CH₂Cl₂containing 0.1 m nBu₄PF₆versus Fc/Fc⁺, and the maximum of the longest-wavelength

band, λ_max, in CH₂Cl₂.

synthesis of DCFs that overcame the restrictions reported by

Katritzky et al. They found that, even though

tetraphenylcyclopentadienone was unreactive under classical

Knoevenagel conditions, the use of Lehnert’s TiCl₄/pyridine

conditions^[13]led to the formation of tetraphenyl-DCF 1 in

65% yield (Figure 3). We found these conditions were general

for the preparation of a number of DCFs. The Swager group

also demonstrated that, even if the parent cyclopentadienone

was susceptible to dimerization, the dimer could be “cracked”

at elevated temperatures after Knoevenagel condensation with

malononitrile. Thus, DCFs tend to be more resistant to

dimerization than their cyclopentadienone counterparts. In

addition, they showed that DCFs displayed two one-electron

reductions in CH₂Cl₂(E_red,1= −0.94 V and E_red,2= −1.40 V vs.

Fc/Fc⁺) at potentials more positive than that of the parent

cyclopentadienones in MeCN (E_{red,tetraphenylcyclopentadienone}

=

−1.52 V vs. Fc/Fc⁺).^[12,14]

Fig. 6. EPR spectra of the radical anions of a) DCF 1 and b) DCF 5 in THF

(296 K).^[3]

chemistry of fulvenes and published extensively on the excited-

and triplet-state aromaticity of certain fulvene derivatives.^[11]In

2010, they proposed that suitably substituted DCFs could

exhibit electron afﬁnities greater than that of the linchpin

acceptor 7,7,8,8-tetracyano-p-quinodimethane (TCNQ).^[4]

Clearly, pentafulvenes and DCFs, in particular, still had much

to offer to those who were taking on their synthesis and study.

That same year, synthetic chemists returned to DCFs.

Swager and co-workers,^[12]at last, reported a general

2. Push–Pull Substituted and

Perethynylated DCFs

At this point, we entered the fray and discovered unexpectedly

that a bis-(N,N-dimethylanilino)-substituted pentacyano-

hexatriene 3E/Z, obtained by the [2+2] cycloaddition–

retroelectrocyclization reaction (CA-RE; see below)^[2]between

4-ethynyl-N,N-dimethylaniline and N,N-dimethylanilino

(DMA)-substituted

1,1,2,4,4-pentacyanobuta-1,3-diene,

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

Fig. 7. The formal [2+2] CA-RE of electron-rich alkynes and the electron-poor oleﬁns TCNE (top) and TCNQ

(bottom).^[2]

Fig. 8. Divergence in regioselectivity in the CA-RE reaction with DCFs.^[3]Given are the electrochemical data,

recorded in CH₂Cl₂containing 0.1 m nBu₄PF₆versus Fc/Fc⁺, and the maximum of the longest-wavelength band, λ_max

,

in CH₂Cl₂.

underwent oxidative cyclization with (NH₄)₂Ce(NO₃)₆to

form push–pull-substituted DCF 4, liberating HCN in the

process (Figure 4).^[15]Later, we optimized the conditions for

the formation of 4 using silica gel in CH₂Cl₂/H₂O.^[16]DCF 4

octacyano[4]dendralene through a CA-RE reaction of a

bis(azulenyl)diyne with TCNE.^[17]

While the optoelectronic properties of 4 were outstanding,

the mechanism of its formation did not appear to be general and

we wished to pursue a more rational design of a class of strongly

accepting DCFs. Inspired by the work of the groups of

Rubin^[18]and Tobe^[19]in the mid-1990s, we knew that

tetraalkynylcyclopentadienones were stable and easy to prepare

and handle, providing that the alkynes were terminally substi-

tuted with a bulky group, such as tert-butyl or triisopropylsilyl

(TIPS). We prepared silyl-protected tetraalkynyl-DCF 5 in the

hopes that it would also behave as a strong acceptor, and it

has some remarkable properties, including

low-energy CT band

a

strong,

(1.59 eV;

(λ_max= 782 nm

ε = 27500 m⁻¹cm⁻¹) that extends well into the near-IR

(1100 nm), and electrochemical reductions (E_red= −0.45 V

and −0.96 V vs. Fc/Fc⁺in CH₂Cl₂) at potentials more

positive than those of tetraphenyl-DCF 1.^[15]Shoji et al.

prepared

a

push–pull

azulenyl

DCF

similar

to 4, but as a byproduct of the formation of an

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

Fig. 9. Two CA-RE reaction mechanisms leading to the divergence in regioselectivity of DCFs 9 and 1. The RDS was

assumed to be the initial attack of 6 on the DCF.^[3]

does: 5 displays two one-electron reductions at potentials

intermediate to those of 1 and 4 (E_red= −0.57 V and −1.13 V vs,

Fc/Fc⁺, CH₂Cl₂; Figure 5).^[3]It was at this time that we decided

to look deeper into the reactivity of DCFs.

Spin localization, and thus, charge stabilization, in 1 and 4 are

quite different, and this not only explains the approximate

500 mV difference in their reduction potentials, but also their

divergence in reactivity, as described below.

3. Radical Anions of DCFs

4. CA-RE Reactivity of DCFs

Because DCFs are good electron acceptors that have stable and

reversible reductions, we decided to study the electron para-

magnetic resonance (EPR) properties of the radical anions of 1

and 5.^[3]Treatment of a solution of the DCF with sodium

mirror in a closed system (THF, 294 K) generated the radical

anion, the EPR spectrum of which could be measured. The

EPR spectra of 1^•–and 5^•–are indeed staggeringly different

(Figure 6). Tetraalkynyl-DCF 5^•–has an EPR spectrum with

distinct hyperﬁne coupling, arising from the cyano nitrogen

atoms, and hyperﬁne coupling of weaker intensity from each

silicon atom; thus, the spin of the radical anions is located in

these areas of the molecule.

Since 2005, we focused much of our research on the synthesis,

reactivity, and optoelectronic properties of nonplanar push–

pull chromophores.^[2]Key to this program has been the devel-

opment of the formal [2+2] CA-RE reaction between electron-

rich alkynes and electron-poor oleﬁns (Figure 7).^[17,20]Most of

our studies have focused on the reactions of the electron-poor

oleﬁns TCNE and TCNQ, but we have demonstrated that less

electron-deﬁcient oleﬁns, such as dicyanovinylarenes^[21]and

cyanoimines,^[22]are also competent in this reaction. We pre-

supposed that DCFs would also participate as acceptors in the

CA-RE reaction, and they do, but the reality was far from what

we expected.

The EPR spectrum of tetraphenyl-DCF 1^•–is not

resolved, but application of the electron nuclear double reso-

nance (ENDOR) technique revealed hyperﬁne coupling

arising from the protons of the phenyl groups.^[3]However,

unlike in the case of 5^•–, the four phenyl groups do not localize

spin equally. We know the orientations of the phenyl groups

from X-ray crystallographic studies: the phenyl groups attached

to C(2) and C(5) are oriented almost perpendicularly to the

plane of the DCF, while the phenyl groups at C(3) and C(4) of

the DCF are oriented about 50° relative to the DCF;^[12]there-

fore, the last two phenyl groups display more spin localization.

Tetralkynyl-DCF 5 was reacted with electron-rich alkyne

6 to give CA-RE adduct 7 in good yield.^[3]Owing to the steric

bulk and the more negative reduction potential of 5 relative to

TCNE and TCNQ, elevated temperatures were required for

the transformation to take place. Tetraphenyl-DCF 1 also

reacts with 6 to form the CA-RE adduct 8 in good yield, but

with the opposite regioselectivity to that of 7 (Figure 8). This is

the ﬁrst example of substituent-directed regioselectivity in the

CA-RE reaction. In the case of CA-RE reactions with TCNQ,

the donor aniline moiety is conjugated to the quinone moiety

(Figure 7); similarly, adduct 7 has the donor aniline conjugated

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

Fig. 10. Three adducts formed from the reaction of push–pull-substituted DCF 4 and electron-rich alkynes 6 and

13. Given also are the electrochemical data, recorded in CH₂Cl₂containing 0.1 m nBu₄PF₆versus Fc/Fc⁺, and the

[16]

maximum of the longest-wavelength band,

λ

_max, in CH₂Cl_2.

Quinoid character = δr = (δ_a+ δ_a′+ δ_c+ δ_c′)/

4 − (δ_b+ δ_b′)/2.^[24]For benzene, δr = 0; for fully quinoid rings, δr = 0.08–0.10 Å.^[24]

to the fulvene. Both CA-RE adduct 7 and the TCNQ adduct

have a “proaromatic” resonance structure, which we believe is a

driving force for the observed regioselectivity.

6 onto DCFs 1 and 9 and the intermediate zwitterion were

calculated. The computational results corroborated our empiri-

cal ﬁndings; for tetraalkynyl-DCF 9, Pathway 1 was lower in

energy than Pathway 2 for both TS (ΔΔG^‡= 1.3 kcal mol⁻¹)

and intermediate formation (ΔΔG = 3.4 kcal mol⁻¹), whereas

for tetraphenyl-DCF 1, Pathway 2 was lower in energy than

Pathway 1 for both TS (ΔΔG^‡= 4.0 kcal mol⁻¹) and interme-

diate formation (ΔΔG = 16.9 kcal mol⁻¹) (Figure 9). The

results for tetraphenyl-DCF 1 are particularly striking.

Although we assumed that the formation of an aromatic, zwit-

terionic intermediate would lead to a lower-energy species, this

was emphatically not the case. As it turns out, such a scenario

is ruled out by both steric and electronic considerations, and

thus, the “unexpected” regioisomer is formed instead.^[3]

The unusual results obtained from the CA-RE reactivity of

symmetrical DCFs were nothing compared with the CA-RE

reactivity of donor–acceptor-substituted DCF 4.^[16]This DCF,

when reacted with alkyne 6, gave three adducts: CA-RE

adduct Z-10, heptafulvene 11, and cyclobutene-

fusedtetrahydropentalene 12 formed from the reaction with two

equivalents of 6 (Figure 10).

So, the question was raised: why do we not observe the

formation of the “proaromatic” regioisomer from the CA-RE

reaction of 1 with 6, and observe the formation of 8 instead?

We tackled this problem with a combined experimental and

computational approach.^[3]We knew from previous mechanis-

tic experiments that the rate-determining step (RDS) of the

CA-RE reaction on sterically rather congested oleﬁns was the

nucleophilic attack of the alkyne on the oleﬁn to yield a zwit-

terionic intermediate.^[21]In the case of DCFs, this led to two

possibilities: 1) attack at C(6) of the DCF, and 2) attack at C(1)

(Figure 9).

Thus, understanding the role of the stabilization of charge

for each intermediate is critical to understanding the mecha-

nism. It was clear from the EPR studies that the stabilization of

electron density was quite different between 1 and 5; however,

to understand the role charge stabilization had in the

divergence of regioselectivity required computational investi-

gations. Utilization of a density functional with dispersion

interactions was critical, so the ωB97XD/6-31G* method^[23]

was used for our computational studies. The difﬁculty of

working with heavy silicon atoms with DFT was overcome by

studying tetraethynyl-DCF 9 instead of 5. The energies of the

transition state (TS) in the initial nucleophilic attack of alkyne

The staggeringly high introduction of complexity into

each scaffold is unprecedented for DCFs. The ratio of forma-

tion of each product is sensitive to solvent polarity and stoi-

chiometry (Table 1); the use of more polar MeCN leads to the

formation of a higher proportion of CA-RE adduct Z-10 than

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

Table 1. Conditions for CA-RE reaction of 4 with 6.

Ratio [%]^[a]

11

6

Entry

Conditions

[equiv] Z-10

12

1

2

3

CH₂Cl₂, 20 °C, 60 h

CH₂Cl₂, 20 °C, 24 h

CH₃CN, 20 °C, 60 h

1

2

1

34

33 (29) 26 (23) 41 (36)

57 24 19

34

32

^[a]Ratios determined by HPLC and ¹H NMR spectroscopy; isolated yields are

given in parentheses.

when CH₂Cl₂is used. The use of excess alkyne 6 leads to more

tetrahydropentalene 12; however, the mechanism of formation

of 12 is distinct from those of Z-10 and 11, which means the

monoadducts are not an intermediate for the formation of 12.

The proposed mechanisms of 11 and 12 are shown in

Figure 11.

While the properties of the novel adducts 11 and 12 are

unspectacular, those of the CA-RE adduct Z-10 are outstand-

ing. Its absorption spectrum is dispersed throughout the entire

visible region and into the near-IR (λ_max= 765 nm,

λ_end= 1150 nm (1.59 eV); ε = 15800 m⁻¹cm⁻¹) due to its

strong intramolecular CT band. In addition, adduct Z-10 also

displays interesting electrochemical behavior: fairly anodically

shifted reduction potentials for a push–pull chromophore

(E_red= −0.67 V and −0.89 V vs. Fc/Fc⁺in CH₂Cl₂), and a very

low ﬁrst oxidation potential ((E_ox,1= +0.27 V vs. Fc/Fc⁺in

CH₂Cl₂); this suggests a signiﬁcant change in geometry

between neutral 10 and its radical cation. The crystal structure

of Z-10 shows the strong quinoidal character^[24]of the DMA

ring attached to C(6) of the fulvene (δ_r= 0.05 Å; Figure 10), as

well as a very long exocyclic fulvene C=C bond (1.41 Å). This

indicates that there is strong conjugation between DMA and

the fulvene rings; however, although the long exocyclic C=C

bond may suggest isomerization behavior, solution and solid-

state studies show that the Z isomer of 10 predominates. By

contrast, adduct 14, the exclusive product from the reaction of

DCF 4 and alkyne 13, has dynamic E/Z isomerization behav-

ior in solution, with a pronounced solvent dependence (E/Z

ratio of 3:1 in CD₂Cl₂and 1:1 in CD₃CN), but reverts exclu-

sively to the E isomer in the solid state.^[16]

5. Polar DCF-to-Benzene Rearrangements

Fig. 11. Mechanisms for the formation of 11 and 12.^[16]

Unsubstituted pentafulvene, C₆H₆, is a nonvalence isomer of

benzene; in other words, it cannot rearrange to form benzene

through pericyclic reactions. The fulvene-to-benzene rear-

rangement occurs via diradicaloid intermediates formed from

ﬂash vacuum pyrolysis (T > 500 °C)^[25]or irradiation.^[26]Little

was known about the rearrangement chemistry of substituted

fulvenes, aside from methyl derivatives.

Despite this, it turns out that many DCFs rearrange spon-

taneously to their benzene isomers under very mild conditions.

We ﬁrst discovered that push–pull DCF 4 underwent rear-

rangement to tetracyanobenzene (TCB) 15 in excellent yield

through the treatment of 4 with wet silica gel in MeCN.^[16]We

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

Fig. 12. Proposed mechanism of the SiO₂/MeCN-promoted rearrangement of DCF 4 to give TCB 15.^[16]

Table 2. Rearrangement of tetraphenyl-DCFs to DCBs.^[27]

E_red[V]^[b]

Reaction time

Ratio 1,3:1,4

[a]

Entry

R

σ_p

1

2

3

4

5

CN (17)

Br (18)

H (1)

Me (19)

OMe (20)

0.66

0.23

0

-0.17

-0.27

−0.57, −1.07

−0.79, −1.07

−0.94, −1.40

−1.01, −1.49

−1.03, −1.48

10 min

1 h

6 h

82:18

99:1

90:10

95:5

8 h

16–21 h

75:25

^[a]Hammett sigma parameter for para-substituted benzenes. Values taken from Ref. [28]. ^[b]Reduction potentials of the parent DCFs in CH₂Cl₂+ 0.1 M nBu₄PF₆

versus Fc/Fc⁺.^[27]

believe that MeCN acts as an electrophile in the course of the

rearrangement; the postulated mechanism is shown in

Figure 12. The rearrangement conditions for 4 seem to be

unique because we have yet to ﬁnd another DCF that under-

goes rearrangement under these conditions.

However, many other DCFs can undergo rearrangement

to dicyanobenzenes (DCBs) under mild conditions without

any additional reagents. We found that simply heating

tetraphenyl-DCFs, such as 1, in a dipolar aprotic solvent led to

the quantitative formation of 1,3-DCB 16a and 1,4-DCB

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

16b.^[27]The choice of solvent is critical; dipolar aprotic sol-

vents, such as dimethylformamide (DMF), dimethylsulfoxide

(DMSO), N,N-dimethylacetamide, and N-methylpyrrolidine

(NMP), promote rearrangement, whereas less polar solvents,

such as MeCN, CH₂Cl₂, pyridine, and acetic acid, are ineffec-

tive. Pendant groups on the aryl rings also affect the rearrange-

ment (Table 2); electron-donating groups, such as methyl (19)

and methoxy (20), lead to slower rates of rearrangement,

whereas electron-accepting groups, such as bromo (18) and

cyano (17), lead to enhanced rates. In addition,

acenaphthylene-fused DCF 21 also undergoes quantitative

rearrangement to 1,3-DCB 22 after 1 h at 160 °C in DMF; no

1,4-DCB isomer is detected (Figure 13).

There is also a strong effect of additives: the addition of

one equivalent KCN/[18]crown-6 in DMF leads to the rear-

rangement of DCF 1 to give DCBs 16a/b in a matter of

minutes at 160 °C; ¹³C enrichment of cyano groups was

observed when ¹³C-labeled KCN was used; this indicated

replacement of the cyano groups. Indeed, the release of cyanide

Fig. 13. Rearrangement of DCF 21.^[27]

Fig. 14. The ring-walk mechanism for the DCF-to-DCB rearrangement.^[27]

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

appears to be critical for the reaction; mixing ¹⁵N-labeled DCF

1 with unlabeled methyl-substituted 19 led to scrambling of

the labeled cyano groups, as shown by ¹⁵N NMR spectroscopy.

We propose a “ring-walk” mechanism that conforms to the

mechanistic data we have gathered (Figure 14).

[3] A. D. Finke, O. Dumele, M. Zalibera, D. Confortin, P. Cias,

G. Jayamurugan, J.-P. Gisselbrecht, C. Boudon, W. B.

Schweizer, G. Gescheidt, F. Diederich, J. Am. Chem. Soc. 2012,

134, 18139–18146.

[4] C. Dahlstrand, K. Yamazaki, K. Kilså, H. Ottosson, J. Org.

Chem. 2010, 75, 8060–8068.

[5] R. B. King, M. S. Saran, J. Chem. Soc. Chem. Commun. 1974,

851–852.

[6] Y. Shvo, D. Czarkie, Y. Rahamim, D. F. Chodosh, J. Am.

Chem. Soc. 1986, 108, 7400–7402.

[7] H. Junek, G. Uray, G. Zuschnig, Liebigs Ann. Chem. 1983,

154–158.

[8] H. Junek, G. Uray, G. Zuschnig, Dyes Pigm. 1988, 9, 137–

152.

[9] A. R. Katritzky, W.-Q. Fan, D.-S. Liang, Q.-L. Li, J. Heterocycl.

Chem. 1989, 26, 1541–1545.

[10] I. Paci, J. C. Johnson, X. Chen, G. Rana, D. Popovic´, D. E.

David, A. J. Nozik, M. A. Ratner, J. Michl, J. Am. Chem. Soc.

2006, 128, 16546–16553.

[11] a) H. Möllerstedt, M. C. Piqueras, R. Crespo, H. Ottosson, J.

Am. Chem. Soc. 2004, 126, 13938–13939; b) H. Ottosson, K.

Kilså, K. Chajara, M. C. Piqueras, R. Crespo, H. Kato, D.

Muthas, Chem. Eur. J. 2007, 13, 6998–7005; c) M.

Rosenberg, C. Dahlstrand, K. Kilså, H. Ottosson, Chem. Rev.

2014, 114, 5379–5425.

6. Summary and Outlook

DCFs were largely forgotten after their discovery. However,

after their potential in advanced materials applications was

recognized, an astonishing array of reactivity and properties

associated with them was soon reported. It is clear that the

work detailed herein has only scratched the surface of DCF

chemistry. The electronic tunability of DCFs and their facile

preparation can lead to materials of functional utility, or even

scaffolds of unprecedented complexity. In addition, DCFs

display a rich reactivity proﬁle exempliﬁed by the CA-RE reac-

tivity of DCFs and their facile rearrangement chemistry. As we

learn more about the potential of DCFs, exciting new chem-

istries will undoubtedly follow.

Acknowledgements

[12] T. L. Andrew, J. R. Cox, T. M. Swager, Org. Lett. 2010, 12,

5302–5305.

[13] a) W. Lehnert, Tetrahedron Lett. 1970, 11, 4723–4724; b) W.

Lehnert, Tetrahedron 1972, 28, 663–666; c) W. Lehnert, Tet-

rahedron 1973, 29, 635–638.

[14] M. A. Fox, K. Campbell, G. Maier, L. H. Franz, J. Org. Chem.

1983, 48, 1762–1765.

[15] G. Jayamurugan, J.-P. Gisselbrecht, C. Boudon, F.

Schoenebeck, W. B. Schweizer, B. Bernet, F. Diederich, Chem.

Commun. 2011, 47, 4520–4522.

[16] G. Jayamurugan, O. Dumele, J.-P. Gisselbrecht, C. Boudon,

W. B. Schweizer, B. Bernet, F. Diederich, J. Am. Chem. Soc.

2013, 135, 3599–3606.

[17] T. Shoji, S. Ito, T. Okujima, N. Morita, Org. Biomol. Chem.

2012, 10, 8308–8313.

We are indebted to our colleagues and collaborators who

worked with us on these projects, in particular, Dr.

Govindasamy Jayamurugan, who discovered the synthesis of 4

and its fascinating properties, and led us into the world of DCF

chemistry; Sophie Haberland, who discovered the rearrange-

ment of 1; and Oliver Dumele, who supported our work with

DFT calculations. EPR studies on the radical anions were

performed in the laboratory of Prof. Georg Gescheidt (TU-

Graz, Austria) by Dr. Michal Zalibera, Dr. Daria Confortin,

and Dr. Pawel Cias. Electrochemistry was measured by Dr.

Jean-Paul Gisselbrecht and Prof. Corinne Boudon, Université

de Strasbourg, France. A.D.F. acknowledges the NSF-

International Research Fellowship Program (USA) for a fellow-

ship. This project was supported by the ERC Advanced Grant

No. 246637 (“OPTELOMAC”).

[18] N. Jux, K. Holczer, Y. Rubin, Angew. Chem. Int. Ed. Engl.

1996, 35, 1986–1990.

[19] Y. Tobe, K. Kubota, K. Naemura, J. Org. Chem. 1997, 62,

3430–3431.

REFERENCES

[20] a) T. Michinobu, J. C. May, J. H. Lim, C. Boudon, J.-P.

Gisselbrecht, P. Seiler, M. Gross, I. Biaggio, F. Diederich,

Chem. Commun. 2005, 737–739; b) T. Michinobu, C.

Boudon, J.-P. Gisselbrecht, P. Seiler, B. Frank, N. N. P.

Moonen, M. Gross, F. Diederich, Chem. Eur. J. 2006, 12,

1889–1905; c) M. I. Bruce, Austr. J. Chem. 2011, 64,

77–103; d) X. Tang, W. Liu, J. Wu, C.-S. Lee, J. You, P.

Wang, J. Org. Chem. 2010, 75, 7273–7278; e) M.

Morimoto, K. Murata, T. Michinobu, Chem. Commun.

2011, 47, 9819–9821.

[1] a) K. Hafner, K. H. Häfner, C. König, M. Kreuder, G. Ploss,

G. Schulz, E. Sturm, K. H. Vöpel, Angew. Chem. Int. Ed. Engl.

1963, 2, 123–134; b) E. D. Bergmann, Chem. Rev. 1968, 68,

41–84; c) R. D. Brown, F. R. Burden, J. E. Kent, J. Chem. Phys.

1968, 49, 5542; d) H. Sauter, H. Prinzbach, Angew. Chem. Int.

Ed. Engl. 1972, 11, 296–298; e) M. Neuenschwander in The

Chemistry of Double-Bonded Functional Groups (Ed.: S. Patai),

Wiley, London, 1989, pp. 1131–1268.

[2] a) M. Kivala, F. Diederich, Acc. Chem. Res. 2009, 42, 235–248;

b) S. Kato, F. Diederich, Chem. Commun. 2010, 46, 1994–

2006.

[21] Y.-L. Wu, P. D. Jarowski, W. B. Schweizer, F. Diederich, Chem.

Eur. J. 2010, 16, 202–211.

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

P e r s o n a l A c c o u n t

T H E C H E M I C A L R E C O R D

[22] M. Chiu, B. Jaun, M. T. R. Beels, I. Biaggio, J.-P. Gisselbrecht,

C. Boudon, W. B. Schweizer, M. Kivala, F. Diederich, Org.

Lett. 2011, 14, 54–57.

[23] J.-D. Chai, M. Head-Gordon, Phys. Chem. Chem. Phys. 2008,

10, 6615–6620.

1984, 106, 6751–6755; d) G. B. M. Kostermans, M.

Hogenbirk, L. A. M. Turkenburg, W. H. D. Wolf, F.

Bickelhaupt, J. Am. Chem. Soc. 1987, 109, 2855–2857; e) G.

Zimmerman, M. Remmler, B. Ondruschka, F.-D. Kopinke, B.

Olk, Chem. Ber. 1988, 121, 1855–1860.

[24] C. Dehu, F. Meyers, J. L. Bredas, J. Am. Chem. Soc. 1993, 115,

6198–6206.

[25] a) T. J. Henry, R. G. Bergman, J. Am. Chem. Soc. 1972, 94,

5103–5105; b) B. Hankinson, H. Heaney, A. P. Price, R. P.

Sharma, J. Chem. Soc. Perkin Trans. 1973, 1, 2569–2575; c) S.

Oikawa, M. Tsuda, Y. Okamura, T. Urabe, J. Am. Chem. Soc.

[26] a) I. Jano, Y. Mori, Chem. Phys. Lett. 1968, 2, 185–188; b) J.

E. Kent, P. J. Harman, M. F. O’Dwyer, J. Phys. Chem. 1981,

85, 2726–2730.

[27] A. D. Finke, S. Haberland, W. B. Schweizer, P. Chen, F.

Diederich, Angew. Chem. Int. Ed. 2013, 52, 9827–9830.

[28] C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–195.

Chem. Rec. 2014, ••, ••–••

www.tcr.wiley-vch.de

Article Doi

6,6-Dicyanopentafulvenes: Teaching an old dog new tricks

DOI: 10.1002/tcr.201402060

Source and publish data:

Authors:

Article abstract of DOI:10.1002/tcr.201402060

Full text of DOI:10.1002/tcr.201402060

Products guided by the article

R&D Labs maybe for 1422962-61-5

Relevant to this article

Hot Product