Push-Pull Buta-1,2,3-trienes: Exceptionally Low Rotational Barriers of Cumulenic C=C Bonds and Proacetylenic Reactivity

Article abstract of DOI:10.1002/chem.201406583

A variety of asymmetrically donor-acceptor-substituted [3]cumulenes (buta-1,2,3-trienes) were synthesized by developed procedures. The activation barriers to rotation ΔG^≠ were measured by variable temperature NMR spectroscopy and found to be as low as 11.8 kcal mol^-1, in the range of the barriers for rotation around sterically hindered single bonds. The central C=C bond of the push-pull-substituted [3]cumulene moiety is shortened down to 1.22 ? as measured by X-ray crystallography, leading to a substantial bond length alternation (BLA) of up to 0.17 ?. All the experimental results are supported by DFT calculations. Zwitterionic transition states (TS) of bond rotation confirm the postulated proacetylenic character of donor-acceptor [3]cumulenes. Additional support for the proacetylenic character of these chromophores is provided by their reaction with tetracyanoethene (TCNE) in a cycloaddition-retroelectrocyclization (CA-RE) cascade characteristic of donor-polarized acetylenes.

Full text of DOI:10.1002/chem.201406583

DOI: 10.1002/chem.201406583
Full Paper
&
Proacetylenicity
Push–Pull Buta-1,2,3-trienes: Exceptionally Low Rotational
Barriers of Cumulenic C=C Bonds and Proacetylenic Reactivity
Przemyslaw Gawel,[a] Yi-Lin Wu,[a] Aaron D. Finke,[a] Nils Trapp,[a] Michal Zalibera,[b]
Corinne Boudon,[c] Jean-Paul Gisselbrecht,[c] W. Bernd Schweizer,[a] Georg Gescheidt,*[b] and
FranÅois Diederich*[a]
Abstract: A variety of asymmetrically donor–acceptor-substi-
tuted [3]cumulenes (buta-1,2,3-trienes) were synthesized by
developed procedures. The activation barriers to rotation
length alternation (BLA) of up to 0.17 ꢀ. All the experimental
results are supported by DFT calculations. Zwitterionic transi-
tion states (TS) of bond rotation confirm the postulated pro-
acetylenic character of donor–acceptor [3]cumulenes. Addi-
tional support for the proacetylenic character of these chro-
mophores is provided by their reaction with tetracyanoe-
thene (TCNE) in a cycloaddition-retroelectrocyclization (CA–
RE) cascade characteristic of donor-polarized acetylenes.
¼
DG were measured by variable temperature NMR spectros-
copy and found to be as low as 11.8 kcalmolꢀ1, in the range
of the barriers for rotation around sterically hindered single
bonds. The central C=C bond of the push–pull-substituted
[3]cumulene moiety is shortened down to 1.22 ꢀ as mea-
sured by X-ray crystallography, leading to a substantial bond
Introduction
mulenes feature remarkably low barriers for rotation about the
¼
cumulene axis, with activation free enthalpies DG around
¼
The barrier for rotation DG about the cumulenic axis in [n]cu-
20 kcalmolꢀ1, in the range of the rotational barrier of a peptide
amide bond.[11] Extensive computational studies suggested
a biradical transition state for cis–trans isomerization of asym-
metrically substituted 1,1,4,4-tetraalkynyl[3]cumulenes.[12]
Upon polarization of [3]cumulenes by donors and acceptors
at the terminal C atoms, the BLA increases significantly, with
the central C=C bond adopting nearly the length of an acety-
lenic bond, around 1.21 ꢀ.[7,13] In a preliminary communica-
tion,[14] we showed that the central C=C bond in push–pull-
substituted [3]cumulene 1, with a length of 1.22 ꢀ, indeed fea-
tures characteristic acetylenic reactivity, in addition to olefinic
reactivity.[15] Similar to donor-polarized alkynes, 1 reacted at
room temperature with tetracyanoethene (TCNE) in a formal
[2+2] cycloaddition-retroelectrocyclization (CA–RE)[16] reaction
to give the stable zwitterionic tetracyanobuta-1,3-diene 2 as
the exclusive product (Figure 1). More recently, we proposed
the intermediacy of such zwitterions in the reaction of sym-
metrically substituted tetraaryl[3]cumulenes with TCNE at
higher temperature (90–1408C) in dipolar-aprotic solvents,
leading ultimately to 5,11-dihydrotetracenes.[8,17]
¼
mulenes decreases with increasing n, starting with DG =
65 kcalmolꢀ1 for 1,2-dideuterioethene (n=1),[1] decreasing to
¼
¼
DG =46.2 kcalmolꢀ1 for 1,3-dimethylallene (n=2)[2] and DG
=31.8 kcalmolꢀ1 for 1,4-dimethylbutatriene (n=3),[3] and
reaching approximately 20 kcalmolꢀ1 for 1,6-di(tert-butyl)-1,6-
diphenylhexapentaene (n=5).[4] A large bond-length alterna-
tion (BLA)[5] between the three C=C bonds in the cumulene
moiety is strongly correlated to a decrease in the barrier to ro-
tation. While the central C=C bond in non-polarized [3]cumu-
lenes adopts values of about 1.25 ꢀ,[6] the two peripheral C=C
bonds are quite elongated compared to typical double bonds,
reaching almost 1.40 ꢀ[7] in push–pull [3]cumulenes and about
1.35 ꢀ[8,9] in tetraaryl[3]cumulenes. We reported in the early
1990s the synthesis of 1,1,4,4-tetrakis(trialkylsilylethynyl)buta-
1,2,3-trienes.[10] Later, we found that these peralkynylated [3]cu-
[a] P. Gawel, Dr. Y.-L. Wu, Dr. A. D. Finke, Dr. N. Trapp, Dr. W. B. Schweizer,
Prof. Dr. F. Diederich
Laboratorium fꢀr Organische Chemie, ETH Zꢀrich
Vladimir-Prelog-Weg 3, HCI, 8093 Zꢀrich (Switzerland)
E-mail: diederich@org.chem.ethz.ch
We coined the term “proacetylenic” to describe the acety-
lene-type structure and reactivity of the central C=C bond in
[3]cumulenes,[14] in analogy to the concept of proaromaticity of
quinoid chromophores.[18] According to Chauvin et al., but-3-
en-1-yne is significantly more “aromatic” than buta-1,2,3-
triene.[19] Proacetylenic chromophores such as 1, similar to
their proaromatic counterparts, exhibit very low HOMO–LUMO
gaps, as a result of the “aromaticity” gain in the charge-sepa-
rated excited state.[14,18] Such low-energy gap materials are of
importance in molecular electronics.[20]
[b] Dr. M. Zalibera, Prof. Dr. G. Gescheidt
Institute of Physical and Theoretical Chemistry
Graz University of Technology
Stremayrgasse 9, 8010 Graz (Austria)
[c] Prof. Dr. C. Boudon, Dr. J.-P. Gisselbrecht
Laboratoire d’Electrochimie et de Chimie Physique du Corps Solide
Institut de Chimie-UMR 7177, C.N.R.S., Universitꢁ de Strasbourg
4, rue Blaise Pascal, 67000 Strasbourg (France)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201406583.
Chem. Eur. J. 2015, 21, 1 – 13
1
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
addition of electron-rich terminal olefins 5a,b to the electron-
poor 1,1-dichloroethene derivatives 6a,b,[22] followed by dehy-
drohalogenation in a one-pot reaction (Scheme 1). The sym-
metric olefin 5a was synthesized from Michler’s ketone by re-
action with methylmagnesium bromide.[23] The second olefin
5b, with different amino groups, was obtained by addition of
lithiated aryl bromide 7 to acetophenone 8 and subsequent
elimination of water.
The synthesis of [3]cumulenes 4a–d required the prepara-
tion of the new electron-rich diaminoolefins 9a,b. Condensa-
tion of N-phenylacetamides 10a,b with N-methylaniline (11),
followed by quenching with sodium perchlorate, afforded the
amidinium salts 12a,b, respectively (Scheme 1). Deprotonation
of 12a,b with NaOH (20% w/w aqueous solution) gave the
corresponding 1,1-diaminoethenes 9a,b.[24] A sequence of sub-
stitution and dehydrohalogenation reactions between 9a,b
and 6a,b yielded the [3]cumulenes 4a–d, respectively. In the
preparation of the malononitrile derivatives 4a,c, a stepwise
protocol was required since the reaction of 9a,b with 2-(di-
chloromethylene)malononitrile (6a), even after exhaustive re-
action times, gave exclusively the buta-1,3-diene derivatives
13a,b. Dehydrohalogenation was subsequently performed
with DABCO to provide [3]cumulenes 4a,c.
Figure 1. Cumulene 1 and its conversion with tetracyanoethene (TCNE) to
zwitterion 2.[14]
Here, we provide comprehensive support for the proacety-
lenic structure and reactivity of donor–acceptor-polarized
[3]cumulenes. We report the new, asymmetrically substituted
push–pull [3]cumulenes 3a–c and 4a–d (Figure 2) and their ex-
ceptionally low barriers to rotation about the cumulene axis.
Based on theoretical calculations, we propose that the mecha-
nism of rotation about the cumulenic axis involves a charge-
separated transition state (Figure 2). In addition, we demon-
strate the proacetylenic reactivity of these donor–acceptor-
substituted [3]cumulenes by reacting them with TCNE in the
CA–RE cascade to give new members of an unusual class of
stable zwitterions.
X-ray Crystal Structures and Bond-Length Alternation
The structures of [3]cumulenes 3a,b and 4a–d were confirmed
by single-crystal X-ray diffraction. The crystal structures of 3a
Figure 2. New push–pull-substituted [3]cumulenes investigated in this study
(top) and postulated transition states for rotation about the cumulenic axis
(bottom).
Figure 3. Crystal structures of 3a (top) and 4b (bottom), arbitrary number-
ing. T=100 K. Thermal ellipsoids at 50% probability, hydrogen atoms omit-
ted for clarity. White: C; light grey: O; dark grey: N. The lengths of the cu-
mulenic bonds in 3a and 4b are shown as well as the short distance be-
tween the bridged phenyl rings in 4b. Selected bond lengths [ꢀ]: 3a: C4Aꢀ
N4A 1.355(4), C4BꢀN4B 1.354(4); 4b: C1AꢀN1A 1.441(2), C1BꢀN1B 1.443(2).
The quinoid character dr of the anilino rings is calculated from the bond
lengths using the following formula: dr={[(a+b)/2ꢀc]+[(a’+b’)/2ꢀc’]}/2
(in benzene dr=0 ꢀ; in fully quinoid rings, dr would be around 0.10–0.12 ꢀ).
For 3a: drA =0.041 ꢀ, drB =0.045 ꢀ, for 4b: drA =ꢀ0.004 ꢀ, drB =ꢀ0.002 ꢀ.
Results and Discussion
Synthesis of Push–Pull [3]Cumulenes
Compounds 3a–c were synthesized in good yields using the
previously reported route to polarized [3]cumulenes,[7,13,14,21] by
&
&
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
2
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
stacking geometry (Figure 3).
The shortest distance between
the two bridged aromatic
C
atoms is approximately 2.88 ꢀ,
which is around 0.6 ꢀ less than
the regular p-stacking dis-
tance.[26] For comparison, the
two bridged C(sp2) atoms in
strained [2.2]paracyclophane are
at a distance of 2.751 ꢀ[27] and in
[3.3]paracyclophane
of
3.137 ꢀ.[28] Similar behavior has
been reported for N,N’-dimethyl-
N,N’-diaryl ureas,[29] which even
form
stacked
oligomeric
chains.[30] In the case of 4a–d,
the syn alignment of the two
phenyl rings is probably adopted
to assure best possible push–
pull conjugation between the
amine donors and the nitrile ac-
ceptors along the [3]cumulene
scaffold. The N atoms indeed are
nearly in plane with the cumu-
lene scaffold and the acceptor
groups, ensuring efficient linear
donor–acceptor p-conjugation.
In the anti conformation, the N
atoms would be in conjugation
with the phenyl rings, which
would obviously weaken push–
pull conjugation across the cu-
mulene scaffold.
Scheme 1. Synthesis of push–pull [3]cumulenes 3a–c and 4a–d. DABCO=1,4-diazabicyclo[2.2.2]octane, Ts=4-tol-
uenesulfonyl.
and 4b are depicted in Figure 3; all other crystal structures are
included in Section S3 in the Supporting Information.
Variable-Temperature NMR and Rotational Barriers
The reduced symmetry of compounds 3a–c and 4b,d enables
The influence of substituents on bond length alternation
(BLA)[5] of the buta-1,2,3-triene moiety is substantial and well
documented. The two extremes are symmetric, electronically
neutral cumulenes (average BLA=0.10 ꢀ)[6] and polarized ones
(highest reported BLA=0.18 ꢀ).[13] All new [3]cumulenes re-
ported here display significant BLA. Likewise, the strength of
electron donor and acceptor have an impact on the BLA of the
[3]cumulene inserted between them. Concerning 3a and 3b,
anilines are weaker donors than amines, thus the BLA is in
general lower than in 4a–d, where the amino donors are in-
stalled directly onto the cumulene (Table 1). Surprisingly, there
is little difference between the compounds possessing two
cyano groups (3b and 4a,c), or one cyano and one methyl
ester group (3a and 4b,d), the latter of which is a weaker elec-
tron acceptor.[25] The structural data confirm the proacetylenic
character of push–pull cumulenes and the significant contribu-
tion of the charge-separated mesomeric structure in the
ground state.
us to obtain kinetic parameters for [3]cumulene rotation by
1
means of VT H and 13C NMR spectroscopy (Table 2). Two moi-
eties were investigated: the N-methylamino signal was fol-
Table 1. Bond lengths and bond length alternation (BLA) for cumulenes
3a,b and 4a–d. D=electron-donating group; A=electron-accepting
group. BLA=(C1=C2+C3=C4)/2ꢀC2=C3. Libration effects were tested,
and the changes in bond lengths are negligible (see Section S3 in the
Supporting Information).
C1=C2 [ꢀ]
C2=C3 [ꢀ]
C3=C4 [ꢀ]
BLA [ꢀ]
3a
1.375
1.367
1.394
1.392
1.391
1.392
1.226
1.227
1.216
1.214
1.213
1.217
1.364
1.364
1.378
1.375
1.379
1.378
0.144
0.138
0.170
0.170
0.172
0.168
3b[a]
4a
4b
4c
4d
Unexpectedly, the phenyl rings in cumulenes 4a–d align
themselves parallel to the cumulene axis, adopting a close p-
[a] Average values of two symmetry-unrelated molecules.
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
3
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
1
lowed by H NMR (for 3a,c and
4b,d), while the nitrile signal
was monitored by 13C NMR
spectroscopy (for 3b). All stud-
ied compounds are in fast equi-
librium at 258C, showing
a single peak for both rotamers.
The coalescence points for all
compounds except 4c were met
while cooling samples down,
and splitting the initial singlet
into two singlets of nearly equal
intensity could be observed at
lower temperatures (Figure 4
and Section S6 in the Support-
ing Information). For 4c, we did
not observe signal splitting,
even at ꢀ1008C, possibly due to
the time scale and resolution
limits of the NMR method.
1
Figure 4. VT H NMR spectra (500 MHz) of 3a in CD2Cl2. Peak intensities are normalized for clarity. * NCH3 peak,
evaluated for the determination of the kinetic parameters.
Unlike BLA, the barrier of rota-
tion can be easily compared to
other systems, such as ethene,[1]
allene,[2] or different cumulenes.[3,4,9,11] Even though we predict-
ed low rotational barriers for these systems, the values as low
additional benzene rings between the amine donors and the
[3]cumulene scaffold in 3a,b. The force to maintain the aro-
matic character of these anilino rings diminishes the charge
transfer in 3a,b, which in return leads to a higher barrier of ro-
tation.
¼
as DG =11.8 kcalmolꢀ1 for 4b exceed our expectations. This
low activation free enthalpy, which is in the range of the barri-
er for rotation of a sterically hindered single bond,[31] shows
clearly the extent of the acetylenic character of push–pull sub-
stituted [3]cumulenes. Significant charge separation within the
molecule results in the lengthening and weakening of the pe-
ripheral [3]cumulene bonds up to the length of the aromatic
C=C bond in benzene[32] (Table 1), thereby facilitating the rota-
tion.
Electrochemistry and Electronic Absorption Spectroscopy
Electrochemical measurements were carried out in CH2Cl2 con-
taining 0.1m Bu4NPF6 in a classical three-electrode cell by
cyclic voltammetry (CV) and rotating-disk voltammetry (RDV)
(Table 3, for more details see Section S7 in the Supporting In-
formation).
We note that the entropic contributions to the activation
free enthalpy are very small in the case of 4b,d while they are
much larger for 3a–c. Apparently, the stacking geometry of
the phenyl groups on the amidine donors in 4b,d remains
very much fixed in the transition state, while the two anilino
donors in 3a–c undergo much stronger reorganization.
In general, the measured barriers of rotation correlate well
with the crystallographic data. The barrier of rotation in 4b,d
is significantly smaller (approx. 3 kcalmolꢀ1) than in 3a,b. At
the same time, the BLA in 4b,d is significantly enhanced
(Table 1). These differences are a result of the insertion of the
Cumulenes 3a–c give two reduction steps as well as two ox-
idations. For 3b and 3c, the first reduction is a reversible one-
electron transfer whereas the second step is irreversible at any
scan rate. The behavior of 3a differs from the others, as only ir-
reversible electron transfers could be observed. Such behavior
may occur due to adsorption phenomena at the electrode. We
had previously observed that N,N-diisopropylanilino moieties,
present in 3b and 3c, are more stable against irreversible oxi-
dation than N,N-dimethylanilino (DMA) ones, which are the
only electron donors in 3a.[33] Cumulenes 4a–d behave in
a similar way. The reduction shows a first reversible one-elec-
tron step, followed by a second irreversible electron transfer,
whereas the one oxidation step occurs irreversibly. In 4a and
4c, the presence of two strongly electron-withdrawing cyano
substituents shifts the observed reduction potentials anodical-
ly, compared to 4b and 4d, which bear only one cyano group
per molecule.
Table 2. Kinetic parameters for rotation about the cumulenic axis in
[3]cumulenes 3a–c and 4b,d.
¼
¼
¼
298 K
[kcalmolꢀ1
]
DH
[kcalmolꢀ1
DS
[calKꢀ1 molꢀ1
DG
]
]
3a
3b
3c
4b
4d
9.6ꢁ0.1
4.4ꢁ0.2
8.1ꢁ0.1
13.3ꢁ0.1
11.3ꢁ0.1
ꢀ19.8ꢁ0.1
ꢀ35.2ꢁ1.0
ꢀ24.8ꢁ0.2
5.2ꢁ0.2
15.5ꢁ0.1
14.9ꢁ0.4
15.4ꢁ0.1
11.8ꢁ0.1
12.6ꢁ0.1
The electron-donating character of the amino groups in 4a–
c is clearly stronger than that of the anilino substituents in 3a–
c (Table 3). The first reduction potentials of 4a–c are more neg-
ative by about 500 mV than those of 3a–c. Likewise, the first
ꢀ4.5ꢁ0.2
&
&
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
4
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
390 nm, see below). Another remarkable observation are the
unusually high DEp values between 80 and 200 mV found for
4a–d, pointing to a substantial rearrangement of the molecu-
lar geometry upon electron transfer (Table 3).
Table 3. Electrochemical data of cumulenes 3a–c and 4a–d observed by
cyclic voltammetry (CV) and rotating disk voltammetry (RDV) in CH2Cl2 +
0.1m nBu4NPF6. All potentials are given versus ferrocene, used as internal
standard.
The deep-purple solutions (CH2Cl2) of 3a–c show two strong
absorption bands in the Vis region and one much weaker one
in the UV region. The lowest-energy absorption varies from
lmax =584 nm (2.12 eV) for 3a to lmax =594 nm (2.09 eV) for
3b. Cumulenes 4a–d show one high-intensity band in the Vis
region at lmax around 390 nm (3.19 eV), along with a weaker
one in the UV region. Spectra of 3a and 4b are included as
representative examples in Figure 5; all spectra are shown in
Section S5 in the Supporting Information. The hypsochromic
shift of the longest-wavelength absorption bands of 4a–d rela-
tive to 3a–c is in line with the larger HOMO–LUMO gaps estab-
lished by cyclic voltammetry. Upon introduction of the ben-
zene spacers between amino donors and the cumulene scaf-
fold in 3a–c, push–pull conjugation is reduced and the ener-
gies of HOMO and LUMO resemble more closely those of the
individual donor and acceptor moieties. Such p-spacer effects
on the HOMO–LUMO gap in push–pull chromophores had pre-
viously been fully established.[34]
CV
DEp
[mV]
RDV
[b]
[c]
E8[a]
[V]
Ep
E1/2
[V]
Slope[d]
[mV]
[V]
3a
3b
3c
+0.72
+0.65
ꢀ1.38
+0.72
(1 eꢀ)
+0.60
(1 eꢀ)
ꢀ1.37
(1 eꢀ)
+0.75
(1 eꢀ)
+0.55
(1 eꢀ)
ꢀ1.32
ꢀ1.96
+0.76
(1 eꢀ)
+0.46
(1 eꢀ)
ꢀ1.38
(1 eꢀ)
ꢀ1.99
(1 eꢀ)
+0.82
(1 eꢀ)
ꢀ1.76
(1 eꢀ)
ꢀ2.17
(1 eꢀ)
+0.75
(1 eꢀ)
ꢀ1.89
(1 eꢀ)
ꢀ2.27
(1 eꢀ)
+0.85
ꢀ1.90
ꢀ2.30
+0.80
(1 eꢀ)
ꢀ2.03
(1 eꢀ)
ꢀ2.30
(1 eꢀ)
60
60
100
ꢀ1.76[e]
ꢀ1.93[e]
+0.75
ꢀ1.31
75
75
60
60
60
100
+0.55
ꢀ1.95
+0.74
ꢀ1.36
60
70
60
60
70
110
+0.48
ꢀ1.99
The charge-transfer (CT) character of the lowest-wavelength
absorption for all cumulenes was proven by protonation-re-
neutralization experiments (Section S5 in the Supporting Infor-
mation). In acidified solution, the CT band disappeared in all
cases, but after re-neutralization recovery of the original spec-
tra took place only for 3b and 3c. The high initial strain
energy of cumulenes is probably the reason behind the non-
reversible protonation reactivity.
4a
4b
+0.85
ꢀ2.21
70
100
175
ꢀ1.76
ꢀ1.86
145
90
+0.71
ꢀ2.31
70
75
90
Computational Studies
4c
+0.90
60
60
240
70
100
120
The huge decrease in the rotational barrier from 1,4-dimethyl-
ꢀ1.90
ꢀ1.96
200
80
¼
¼
butatriene (DG =31.8 kcalmolꢀ1 [3]
to 4b (DG =11.8 kcal
)
ꢀ2.12
+0.76
molꢀ1) clearly demonstrates the pronounced effect of molecu-
lar polarization through push–pull conjugation. Computational
results at the B3LYP/6-31G(d) level of theory (Gaussian 09)[35]
show that the rotation of push–pull [3]cumulenes occurs, as
expected, via perpendicular zwitterionic transition states
(Figure 6 and Section S8 in the Supporting Information).
Barriers of rotation for the new [3]cumulenes were comput-
ed. Transition- and ground-state Gibbs free energies were ob-
tained by adding the zero-point energy and thermal correction
to the calculated electronic energies at 298 K. The calculated
values are in good agreement with the experimental ones
(Table 4).
4d
ꢀ2.37
[a] Eo =(Epc +Epa)/2, where Epc and Epa correspond to the cathodic and
anodic peak potentials, respectively. [b] DEp =EpaꢀEpc. [c] Ep =irreversible
peak potential. [d] Logarithmic analysis of the wave obtained by plotting
E versus log[I/(IlimꢀI)]. [e] Small amplitude signals.
oxidation potentials of 4a–d are more positive by about
200 mV compared with 3a–c. Accordingly, 3a–c are both
better electron acceptors and better electron donors than 4a–
d. The larger HOMO–LUMO gap of 4a–d is a consequence of
the more pronounced interaction between the amino group
and the acceptor moieties, compared to the interaction of the
anilino groups and acceptor moieties in 3a–c. The UV/Vis spec-
tra of the corresponding compounds are in very good agree-
ment with the electrochemical data; the lowest-energy absorp-
tion of 3a–c are around 590 nm, whereas the corresponding
absorptions for 4a–d appear at distinctly higher energies (ca.
The changes in geometry between the calculated ground
state (GS) and the transition state (TS) are due to an increase
in polarization of the molecule while the rotation occurs. Bond
length changes in the cumulene moiety are slightly smaller
than the ones obtained for the 1,1,4,4-tetraalkynylbuta-1,2,3-
trienes.[11] Because of the already strong BLA in the ground
state, less change is needed to reach the geometry required
for rotation. The central bond shrinks by approximately
0.014 ꢀ for series 3a–c and approximately 0.010 ꢀ for series
4a–d. The peripheral bonds stretch by approximately 0.042 ꢀ
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
5
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
for series 3a–c and approximate-
ly 0.030 ꢀ for series 4a–d, with
negligible difference between
the two different bonds in the
molecule (Section S8).
Increase in charge separation
during the rotation can be con-
veniently monitored by both the
change in dipole moment and
the electrostatic potential along
the molecule (shown for 3a in
Figure 6; for all other cumulenes,
see Section S8).
A strong in-
crease in the dipole moment is
calculated for all studied [3]cu-
mulenes (Table 4). Such a massive
change confirms our hypothesis
of a zwitterionic transition state
(Figure 2) for the rotation
around the cumulene axis. The
Figure 5. UV/Vis/NIR absorption spectra of compounds 3a, 4b, 14a, and 15b in CH2Cl2 at 298 K, c=2ꢂ10ꢀ5 m.
electrostatic potential map at the isosurface also supports this.
Clear charge accumulation can be seen at both ends of mole-
cule 3a: positive at the amino groups and negative at the
cyano groups (Figure 6).
The electronic transitions for the optimized molecular struc-
tures of cumulenes 3a–c and 4a–d were calculated by time-
dependent density functional theory (TD-DFT; for more details
on the computational methods, see Section S8). The lowest-
energy transition was underestimated for all cumulenes, com-
pared to the experimental results (in the range of 0.5–0.6 eV),
which is a known issue for TD-DFT methods applied to charge-
transfer systems.[36] Nevertheless, particular transitions are well
resolved. Two low-energy transitions in cumulene series 3a–
c are strong HOMO-to-LUMO and (HOMO-1)-to-LUMO transi-
tions, respectively. These can be described as intramolecular
charge-transfer (ICT) bands, with electron density displacement
from the anilino electron donors to the cyano electron accept-
ors. The origin of the lowest-energy transition in the cumulene
series 4a–d is not as clear. Frontier molecular orbitals show
strong displacement along the cumulene, but the tendency is
not so apparent as in series 3a–c (see Section S8 in the Sup-
porting Information).
Figure 6. Top: Gibbs free energy reaction profile for rotation around the cu-
mulene axis of 3a. Bottom: electron density from total SCF density of 3a
(isovalue=0.0004; mapped with electrostatic potential (ESP)). GS=ground
state; TS=transition state.
Proacetylenic Reactivity of Push–Pull [3]Cumulenes and
Properties of Formed Zwitterionic Species
Table 4. Calculated parameters for the rotation of cumulenes 3a–c and
4a–d: rotational barriers (DG ), BLA change (DBLAGS-TS), and dipole
¼
moment change (DPGS-TS). GS=ground state; TS=transition state. The
free enthalpies of activation are given at 298.15 K.
The above sections have shown that cumulenes 3a–c and 4a–
d have remarkable proacetylenic character making them ideal
candidates for cycloaddition reactions with electron-poor ole-
fins, such as TCNE.[14,16] Here, we describe the observed reactiv-
ity and shed some light on the properties of the produced
push–pull-type zwitterionic products.
¼
DG [kcalmolꢀ1
]
DBLAGS-TS [ꢀ]
DPGS-TS [D]
3a
15.6
15.2
15.7
11.4
10.7
10.7
10.6
0.058
0.055
0.056
0.039
0.040
0.038
0.039
9.23
8.64
9.28
5.15
5.15
4.76
5.18
3b
3c[a]
4a
The CA–RE reaction of all [3]cumulenes 3a–c and 4a–d with
TCNE was immediate, and the zwitterionic species 14a–c and
15a–d were isolated in good yields (Scheme 2). In all cases,
the products had paramagnetic character, which hindered
spectral characterization.[37] Using EPR spectroscopy to follow
4b
4c
4d[a]
[a] Average values for E and Z isomers.
&
&
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
6
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
As previously described,[14] we
observed complexation be-
tween the tetraphosphonate
cavitand 16 (Figure 8)[42] and
the zwitterions of series 14a–
c by 1H and 31P NMR in CDCl3.
The formed 1:1 complexes are
stabilized by the strong ion-
dipole interaction of the four
convergent P=O dipoles with
one of the terminal N,N-dime-
thyliminium ions. This binding
mode had been previously con-
firmed by X-ray analysis for the
complex formed by zwitterion
2.[14] Complexation is indicated
by a strong color change, origi-
Scheme 2. Synthesis of zwitterions 14a–c and 15a–d. * Yields are including paramagnetic contaminations.
the concentration of the paramagnetic species during the reac-
tion and investigate the temperature dependence of the EPR
spectra, we found that rather persistent paramagnetic species
are formed during the reaction but the zwitterions themselves
are not paramagnetic (see Section S9 in the Supporting Infor-
mation). Consequently, obtaining NMR spectra of zwitterions
14a–c required several recrystallizations, and in the case of
15a–d, the paramagnetic species were not separable from the
parent compounds, thus X-ray crystallography was the method
of choice for structure confirmation (Section S3 in the Support-
ing Information).
nating from a hypsochromic shift of the ICT band, up to
50 nm. The association strength was determined by titration
experiments (see Section S11). 1H and 31P NMR titrations in
The structures of zwitterions 14a and 15a–c, confirmed by
X-ray crystallography, are similar (Figure 7; for all structures,
see Section S3). The same mechanism for charge separation
holds as has been reported previously for 2.[14] The charges are
permanently separated by the central cross-conjugated dicya-
noethenylidene moiety, with large torsion angles between the
various chromophoric sub-moieties also contributing to the
isolation of charges. The positive charge is located on the
amino donor substituents. In 14a, there is high bond length
alternation in the aniline rings which exhibit high quinoid char-
acter dr, as defined in the caption to Figure 7.[38] The positive
charge is mostly located on the terminal amino centers which
possess high iminium ion character as evidenced by the short
Nsp2ꢀCsp2 bond length of about 1.34 ꢀ.[39] In series 15a–c, the
charge is centered on the two nitrogens in an amidinium-type
system. Again, such amidinium cation character is supported
by the short Csp2ꢀNsp2 bond lengths varying between 1.33 and
1.35 ꢀ.[40] The negative charge in both series is stabilized in
form of a flat tris- or tetracyanoallyl anion possessing almost
equal CꢀC bonds of approximately 1.40 ꢀ length within the
allyl anion moiety.[41] All these bonds lengths are in excellent
agreement with those found in the CSD for similarly charged
species (Section S4 in the Supporting Information). In 15a and
15b, the aniline rings are no longer aligned as they are in 4a–
d and one of the rings is parallel with the polycyanated allyl
anion. The bulkiness of the isopropyl group in 15c, however,
forces the aniline rings to stay p-stacked together as in the
parent cumulene 4c (Section S3).
Figure 7. Crystal structures of 14a (top) and 15b (bottom), arbitrary num-
bering. T=100 K. Thermal ellipsoids at 50% probability, hydrogen atoms
omitted for clarity. White: C; light grey: O; dark grey: N. Selected bond
lengths [ꢀ], angles [8], and torsional angles [8]: 14a C4AꢀN4A 1.342(2), C4Bꢀ
N4B 1.340(2), C3ꢀC4 1.395(3), C3ꢀC3D 1.401(3); C4-C3-C3D 126.0(2), C1-C2-
C3-C4 ꢀ120.2(2); 15b: C1AꢀN1A 1.440(1), C1BꢀN1B 1.450(1), C3ꢀC4
1.408(2), C3ꢀC3D 1.405(2); C4-C3-C3D 121.1(1), C1-C2-C3-C4 ꢀ35.9(1). The
quinoid character dr of the anilino rings is calculated from the bond lengths
in aniline moieties using the following formula: dr={[(a+b)/2ꢀc]+[(a’+b’)/
2ꢀc’]}/2 (in benzene, dr=0; in fully quinoid rings, dr=0.10–0.12 ꢀ).[38] For
14a: drA =0.061 ꢀ, drB =0.062 ꢀ, for 15b: drA =0.001 ꢀ, drB =ꢀ0.001 ꢀ.
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
7
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
CDCl3 yielded Ka values for the
formed 1:1 complexes between
10000 and 40000mꢀ1 whereas iso-
thermal titration calorimetry (ITC) in
nitrobenzene afforded Ka values be-
tween 2600 and 6300mꢀ1 (for de-
tails, see Section S11). In contrast,
zwitterions 15a–d do not bind to
the tetraphosphonate cavitand be-
cause their amidinium groups do
not fit into the cavity.
during the voltammetric scan correlate well with the rise and
decay of optical bands at 591 nm, 400 nm, and 270 nm (Fig-
ure 9a), particularly well separated from the bands of the
parent compound. Additionally, the strong bleaching of the
band at 469 nm indicates an almost complete conversion of
the zwitterion 14a during the experiment, and allows us to
assign also the overlapping absorptions at 830 nm and 307 nm
Cꢀ
to the corresponding radical anion 14a . The intensity of the
Figure 8. Tetraphosphonate
cavitand 16.
EPR signal decreases during the potential scan in the region of
the second electron transfer, and the EPR silent dianion shows
absorptions at 550 nm and 319 nm (Figure 9a). Moreover, for-
mation of dianion 14a2ꢀ results in the complete bleaching of
the strong ICT absorption at 830 nm.
Electrochemical studies on the zwitterions were performed
by CV and RDV in CH2Cl2 +0.1m Bu4NPF6 using ferrocene as in-
ternal standard (Table S7). Zwitterions 14a–c give two reversi-
ble one-electron reduction steps and an irreversible two-elec-
tron oxidation step associated on the reverse scan to a small
amplitude peak at +0.55 V for 14b and +0.41 V for 14c
(Table S7). During the oxidation of 14a, a deposit was formed
on the electrode surface, giving rise, on the reverse scan, to
a redissolution peak at +0.40 V (Figure S62). Similarly to cumu-
lene 3a, the deposit is probably caused by the DMA substitu-
ents which lead to poor solubility compared to the N,N-diiso-
propylanilino derivatives. Two reversible one-electron reduc-
tions and only one irreversible oxidation step were observed
for zwitterions 15a–d. Oxidation of 15c occurs close to elec-
trolyte discharge and could not be resolved.
No new EPR signals were detected during the anodic oxida-
tion of 14a. Additionally, the four isosbestic points (at 283,
303, 326, and 371 nm) found in the simultaneously detected
UV/Vis spectra confirm a clean conversion of the neutral com-
pound to the diamagnetic dication, and its recovery after the
reduction in the CV back scan (Figure 9b). No indication for
the formation of other products was found. The electrochemi-
cally irreversible oxidation that proceeds as a simultaneous
two-electron transfer, followed by a slow heterogeneous elec-
tron transfer, is likely responsible for the 0.3 V positive shift of
the corresponding reduction peak in the CV back scan. The ab-
sorption bands of the dication appear at 452 nm and 275 nm
(shoulder at 327 nm), respectively. Similar to the dianion, for-
mation of the dication results in the complete bleaching of the
ICT absorption.
Zwitterions 14a–c are brown in CH2Cl2 and feature three
broad absorption bands starting in near infrared (lmax approx.
830 nm, 1.49 eV) and extending through the visible region
(lmax approx. 460 nm, 2.69 eV) into the UV (320 nm, 3.87 eV;
see the spectrum of 14a in Figure 5). The strong HOMO–
LUMO bands arise from ICT, as already reported for 2 (Sec-
tion S5).[14] Removal of the phenyl rings from the conjugated
p-system in zwitterions 15a–d leads to a large intensity de-
crease and hypsochromic shift of the ICT band as compared to
14a–c (see the spectrum of 15b in Figure 5). While the intense
high-energy band at lmax approximately 380 nm (3.26 eV) stays
relatively constant within the series 15a–d, the HOMO–LUMO
gap varies. The most sterically hindered zwitterion 15d ab-
sorbs at lmax =709 nm (1.75 eV) and the less crowded 15a at
lmax =799 nm (1.55 eV, Figure S38).
Cꢀ
The EPR spectra of the paramagnetic monoanions 15 show
similar features with a single line at g=2.0027(1), DBpp
=
0.9 mT, and the lineshape indicating an unresolved hyperfine
structure (see Sections S9 and S10 in the Supporting Informa-
tion). The signal characteristics are typical for delocalized or-
ganic radicals, but the low resolution of the spectra does not
allow a clear-cut description of the one-electron reduced spe-
cies. The simultaneously detected optical transitions appear
around 520 and 300 nm, and in comparison with parent 15,
Cꢀ
the band positions for 15 seem to be less influenced by sub-
stituent effects or steric factors. The same holds for the dia-
magnetic dianions 152ꢀ, showing the most intense absorption
around 320 nm. The spectroelectrochemical investigations also
show that the irreversible character of the oxidation of 15 can
be traced back to chemical follow-up reactions. However, more
investigations are needed to clarify details of this reaction.
Zwitterionic Chromophores Followed by EPR and Spectroe-
lectrochemistry
The changes of the electronic structure of the donor–acceptor
(D–A) compounds during redox reactions can be conveniently
followed by UV/Vis spectroscopy and the corresponding para-
magnetic ions characterized by EPR. A detailed in situ EPR/UV/
Vis spectroelectrochemical study of zwitterions 14a and 15a–
d was thus performed to get more insight into their redox
chemistry. Compound 14a was investigated in a more polar
solvent, CH3CN, to avoid the formation of the deposit at the
electrode surface during oxidation. One-electron reduction
Conclusions
We conducted an extensive experimental and theoretical study
which confirms the postulated proacetylenic character of the
central C=C bond in [3]cumulenes.[14] Small HOMO–LUMO gaps
and high bond length alternation (BLA) are in agreement with
theoretical predictions of a strong contribution of charge-sepa-
rated resonance structures in the ground state. Additionally,
we observed exceptionally low rotational barriers around the
¼
Cꢀ
produced paramagnetic 14a giving an unresolved singlet in
the EPR spectrum with DBpp =0.65 mT at g=2.0027(1) (Fig-
[3]cumulene axis. The lowest measured value of DG =
11.8 kcalmolꢀ1 is in the range of the barrier for rotation around
ure 9a bottom). Changes in the intensity of the EPR signal
a sterically crowded CꢀC single bond, and 20 kcalmolꢀ1 lower
&
&
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
8
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
Figure 9. a) Top: UV/Vis spectra recorded during the reduction of 14a on a Pt electrode in CH3CN+0.1m Bu4NClO4. The thick green, red, and blue lines corre-
Cꢀ
Cꢀ
spond to spectra of 14a, 14a , and 14a2ꢀ, respectively. The inset shows the corresponding CV (scan rate=3 mVsꢀ1). Bottom left: EPR spectrum of 14a re-
Cꢀ
corded at E=ꢀ0.5 V vs. Fc+/Fc, (DBpp =0.65 mT, g=2.0027(1)). Bottom right: dependence of the intensity of the EPR signal of 14a (IEPR), DA of selected
bands (DA=AꢀAinitial), and the transferred charge (jqj = jsIdtj) on the potential during the CV (scan direction indicated by arrows). b) Top: UV/Vis spectra re-
corded during the oxidation of 14a on a Pt electrode in CH3CN+0.1m Bu4NClO4. The thick green and magenta lines correspond to the spectra of 14a and
14a2+, respectively. The inset shows the corresponding CV (scan rate=3 mVsꢀ1). Bottom: dependence of DA on selected bands and the transferred charge
jqj on the potential during the CV (scan direction indicated by arrows).
than the barrier measured for 1,4-dimethylbutatriene.[3] Such
ylium-1-ide (14a): A solution of 3a (56 mg, 0.15 mmol) and tetra-
cyanoethene (19 mg, 0.15 mmol) in acetonitrile (10 mL) was stirred
at 258C for 15 min. Evaporation of solvent, flash chromatography
(FC) (CH2Cl2/MeOH 98:2), and recrystallization (EtOAc/CH2Cl2/hex-
rapidly rotating systems could have potential use in molecular
motor construction.[43,44] This facile rotation proceeds most
probably through zwitterionic transition states, stabilized by
anes 2:1:7) gave 14a (58 mg, 76%) as dark brown crystals with
the strong molecular polarization. Comprehensive DFT calcula-
a copper luster. Rf =0.18 (SiO2; CH2Cl2/EtOAc 1:1); m.p. 2628C
(decomp.); 1H NMR (400 MHz, C2D2Cl4, 258C) d=3.38 (s, 12H; 4
tions confirm our experimental findings. In addition, reacting
the cumulenes with TCNE in the CA–RE transformation, which
is characteristic for donor-polarized acetylenes, corroborates
the proacetylenic reactivity. The formed zwitterions were char-
acterized by X-ray analysis and by an in-situ EPR/UV/Vis spec-
troelectrochemical study. They are truly unusual structures
which we also postulated as intermediates in our recently re-
ported new two-step synthesis of tetracenes with a rubrene-
like substitution pattern, starting from tetraarylated [3]cumu-
lenes.[8]
NCH3), 3.78 (s, 3H; OCH3), 6.92 (d, J=9.4 Hz, 4H; 2 H-C(3’’,5’’)), 7.54
(d, J=9.4 Hz, 4H; 2 H-C(2’’,6’’)); 13C NMR (101 MHz, C2D2Cl4, 258C)
d=41.26 (4 NCH3), 47.22 (C(CꢂN)2), 52.31 (OCH3), 78.37 (C(1’)), 111.6
(C(CꢂN)(CO2CH3)), 112.24(C(CꢂN)2), 114.70 (2 C(3’’,5’’)), 120.17 (CꢂN),
129.91 (2 C(1’’)), 140.58 (2 C(2’’,6’’)), 156.88 (2 C(4’’)), 161.49 ppm
(C=O); 4 C hidden by the noise; IR (ATR): n˜ =2919 (w), 2875 (w),
2203 (m), 2188 (m), 2171 (m), 1682 (s), 1617 (m), 1570 (s), 1523 (w),
1455 (s), 1349 (s), 1293 (m), 1264 (s), 1189 (m), 1155 (s), 1119 (br. S),
935 (s), 909 (s), 855 (w), 825 (m), 814 (w), 789 (w), 776 (s), 737 (w),
727 (m), 171 (m), 682 (s), 662 (w) 627 (w), 610 cmꢀ1 (w); UV/Vis
(CH2Cl2): lmax (e)=819 (26500), 456 (21200), 317 nm
(26500mꢀ1 cmꢀ1); HR-ESI-MS: m/z (%): 502.1986 (100, [M+H]+,
calcd for C29H24N7O2+: 502.1980).
Experimental Section
1,1-Dicyano-2,3-bis(dicyanomethylidene)-4-[4-(dimethylamino)-
phenyl]-4-{4-[di(propan-2-yl)amino]phenyl}but-4-ylium-1-ide
(14b): A solution of 3b (32 mg, 0.08 mmol) and tetracyanoethene
(10 mg, 0.08 mmol) in acetonitrile (5 mL) was stirred at 258C for
15 min. Evaporation of solvent, FC (CH2Cl2/MeOH 98:2), and recrys-
tallization (EtOAc/CH2Cl2/hexanes 2:1:7) gave 14b (35 mg, 83%) as
dark brown crystals with a copper luster. Rf =0.39 (SiO2; CH2Cl2/
EtOAc 1:1); m.p. 2378C (decomp.); 1H NMR (500 MHz, C2D2Cl4,
258C) d=1.51 (d, J=6.8 Hz, 12H; 4 CNCH3), 3.38 (s, 6H; 2NCH3),
4.33 (sept, J=6.8 Hz, 2H; 2 NCH), 6.95 (d, J=9.4 Hz, 2H; H-C(3’,5’)),
In the following, the experimental details for synthesis and charac-
terization of zwitterions 14a–c and 15a–d are given. IUPAC names
of zwitterions are given for structures with charge position defined
on a single atom. All other synthetic details and experimental data,
1
X-ray, UV/Vis, H and 13C NMR data, electrochemistry, EPR spectros-
copy, spectroelectrochemistry, binding studies, and theoretical cal-
culations are given in the Supporting Information.
(2E/Z)-1,1-Dicyano-2-(1-cyano-2-methoxy-2-oxoethylidene)-3-(di-
cyanomethylidene)-4,4-bis[4-(dimethylamino)phenyl]but-4-
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
9
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
7.13 (d, J=9.6 Hz, 2H; H-C(3’,5’)), 7.44 and 7.46 ppm (2 d, J
ꢃ8.8 Hz, 4H; 2 H-C(3’,6’)); 13C NMR (101 MHz, C2D2Cl4, 258C) d=
20.88 (4 CHCH3), 41.19 (2 NCH3), 99.49 (2 C(CꢂN)2), 110.73 (C(CꢂN)2),
111.52 (C(CꢂN)2), 114.67 (C(3’,5’), 114.81 (2 CꢂN), 117.64 (2 C(2’,6’)),
117.82 (2 CꢂN), 120.17 (C(1’)), 139.39 (2 C(2’,6’)), 156.48 (C(4’)),
157.97, 158.21 (C(4’)), 165.42 (C=O); 4 C hidden by the noise; IR
(ATR): n˜ =2969 (w), 2934 (w), 2187 (s), 1615 (m), 1573 (s), 1530 (m),
1473 (s), 1432 (m), 1372 (s), 1333 (s), 1306 (S), 1196 (s), 1153 (s),
1111 (S), 1010 (w), 934 (m), 913 (s), 825 (m), 795 (w), 778 (w), 748
(w), 736 (m), 702 (w), 681 (s), 653 (w), 629 (w), 610 cmꢀ1 (w); UV/Vis
(CH2Cl2): lmax (e)=835 (30000), 460 (23800), 319 nm
(30700mꢀ1 cmꢀ1); HR-MALDI-MS: m/z (%): 547.2329 (49, [M+Na]+,
and recrystallization (EtOAc/CH2Cl2/hexanes 2:1:7) gave 15b
(221 mg, 93%) as a black solid. Rf =0.34 (SiO2; cyclohexane/EtOAc
1:1); m.p. 2288C (decomp.); IR (ATR): n˜ =3071 (w), 2953 (w), 2200
(m), 2185 (s), 1666 (s), 1598 (m), 1590 (m), 1556 (s), 1538 (m), 1490
(m), 1446 (s), 1435 (s), 1405 (s), 1336 (w), 1291 (s), 1249 (w), 1291
(s), 1249 (w), 1213 (w), 1189 (m), 1156 (w), 1126 (s), 1104 (m), 1084
(w), 1058 (w), 1021 (w), 1000 (w), 989 (m), 926 (w), 851 (w), 822
(w), 794 (w), 765 (s), 756 (s), 708 (w), 694 (s), 665 (m), 654 (w),
627 cmꢀ1 (m); UV/Vis (CH2Cl2): lmax (e)=790 (850), 395 nm
(20000mꢀ1 cmꢀ1); HR-ESI-MS: m/z (%): 506.1928 (100, [M+MeOH+
H]+, calcd for C28H24N7O3+: 506.1935), 474.1676 (12, [M+H]+, calcd
for C27H20N7O2+: 474.1673).
calcd for C32H28N8Na+: 547.2329), 525.2509 (47, [M+H]+, calcd for
1,1-Dicyano-2,3-bis(dicyanomethylidene)-4-[methyl(phenyl)ami-
no]-4-[phenyl(propan-2-yl)amino]but-4-ylium-1-ide (15c): A solu-
tion of 4c (136 mg, 0.4 mmol) and tetracyanoethene (56 mg,
0.44 mmol) in acetonitrile (10 mL) was stirred at 258C for 15 min.
Evaporation of solvent, FC (CH2Cl2/MeOH 98:2), and recrystalliza-
tion (EtOAc/CH2Cl2/hexanes 2:1:7) gave 15c (180 mg, 96%) as
a black solid. Rf =0.34 (SiO2; cyclohexane/EtOAc 1:1); m.p. 1278C
(decomp.); IR (ATR): n˜ =2981 (w), 2193 (s), 1596 (m), 1553 (s), 1465
(s), 1393 (w), 1372 (w), 1291 (w), 1227 (w), 1194 (w), 1174 (w), 1099
(m), 1081 (w), 1025 (w), 1003 (w), 970 (w), 942 (w), 884 (w), 814
(w), 790 (w) 767 (m), 759 (m), 693 (s), 666 (w), 648 (w), 631 cmꢀ1
(w); UV/Vis (CH2Cl2): lmax (e)=729 (1460), 385 nm (18200mꢀ1 cmꢀ1);
HR-ESI-MS: m/z (%): 501.2143 (63, [M+MeOH+H]+, calcd for
C32H29N8+: 525.2510), 524.2431 (100, [M]+, calcd for C32H28N8
:
+
524.2431).
(2E/Z)-1,1-Dicyano-2-(1-cyano-2-methoxy-2-oxoethylidene)-3-(di-
cyanomethylidene)-4-[4-(dimethylamino)phenyl]-4-{4-[di(pro-
pan-2-yl)amino]phenyl}but-4-ylium-1-ide (14c): A solution of 3c
(86 mg, 0.2 mmol) and tetracyanoethene (26 mg, 0.2 mmol) in ace-
tonitrile (20 mL) was stirred at 258C for 15 min. Evaporation of sol-
vent, FC (CH2Cl2/MeOH 98:2), and recrystallization (EtOAc/CH2Cl2/
hexanes 2:1:7) gave 14c (88 mg, 78%) as dark brown crystals with
a copper luster. Rf =0.37 (SiO2; CH2Cl2/EtOAc 1:1); m.p. 2688C
1
(decomp.); H NMR (500 MHz, C2D2Cl4, 258C) d=1.49 (d, J=6.9 Hz,
12H; 4 CHCH3), 3.35 (s, 6H; 2 NCH3), 3.78 (s, 3H; OCH3), 4.31 (sept,
J=6.9 Hz, 2H; 2 NCH), 6.90 (d, J=9.3 Hz, 2H; H-C(3’’,5’’)), 7.06 (d,
J=9.5 Hz, 2H; H-C(3’’,5’’)), 7.61–7.42 ppm (m, 4H; 2 H-C(2’’,6’’));
13C NMR (101 MHz, C2D2Cl4, 258C; 1:1 mixture of stereoisomers) d=
20.81 (2 CHCH3), 41.03 (2 NCH3), 50.84 (OCH3), 52.29 (2NCH), 78.30
(C(1’)), 111.61 (2 CꢂN), 112.33 (2 CꢂN), 114.22 (C(3’’,5’’)), 115.95 (2
CꢂN), 117.28 (C(3’’,5’’)), 117.72 and 118.27 (C(1’’)), 129.44 and
130.18, 140.16 (2 C(2’’,6’’)), 156.35 (2 C(4’’)), 158.69 and 160.26,
166.05 and 166.77 ppm (C=O); 3 C hidden by the noise; IR (ATR):
n˜ =2974 (w), 2206 (m), 2191 (s), 1678 (s), 1614 (m), 1577 (s), 1524
(m), 1456 (s), 1433 (s), 1382 (s), 1336 (S), 1308 (m), 1272 (s), 1198
(s), 1159 (m), 1114 (s), 1007 (w), 938 (m), 912 (s), 859 (w), 831 (s),
814 (m), 791 (w), 780 (w), 764 (w), 741 (w), 682 (s), 653 (w), 627 (w),
613 cmꢀ1 (w); UV/Vis (CH2Cl2): lmax (e)=828 (29500), 457 (22700),
317 nm (28400mꢀ1 cmꢀ1); HR-MALDI-MS: m/z (%): 580.2431 (38,
[M+Na]+, calcd for C33H31N7NaO2+: 580.2431), 558.2612 (100, [M+
H]+, calcd for C33H32N7O2+: 558.2612), 557.2533 (97, [M]+, calcd for
C33H31N7O2+: 557.2534).
C29H25N8O+: 501.2145), 474.2033 (100, [MꢀCNꢀ +MeOH]+, calcd for
C28H24N7O+: 474.2037), 469.1882 (2, [M+H]+, calcd for C28H21N8
:
+
469.1884).
(2E/Z)-1,1-Dicyano-2-(1-cyano-2-methoxy-2-oxoethylidene)-3-(di-
cyanomethylidene)-4-[methyl(phenyl)amino]-4-[phenyl(propan-
2-yl)amino]but-4-ylium-1-ide (15d): A solution of 4d (187 mg,
0.5 mmol) and tetracyanoethene (70 mg, 0.55 mmol) in acetonitrile
(20 mL) was stirred at 258C for 15 min. Evaporation of solvent, FC
(CH2Cl2/MeOH 98:2), and recrystallization (EtOAc/CH2Cl2/hexanes
2:1:7) gave 15d (223 mg, 89%) as a black solid. Rf =0.29 (SiO2; cy-
clohexane/EtOAc 1:1); m.p. 1298C (decomp.); IR (ATR): n˜ =2952
(w), 2190 (s), 1759 (w), 1673 (s), 1595 (m), 1525 (s), 1490 (m), 1525
(s), 1490 (m), 1452 (s), 1436 (s), 1391 (w), 1372 (w), 1301 (m), 1269
(s), 1238 (w), 1192 (m), 1131 (s), 1099 (s), 1081 (w), 1026 (w), 1003
(W), 972 (w), 925 (w), 883 (w), 848 (w), 820 (w), 792 (w), 764 (s), 694
(s), 647 (w), 632 (w), 618 cmꢀ1 (w); UV/Vis (CH2Cl2): lmax (e)=709
(1610), 370 nm (16500mꢀ1 cmꢀ1); HR-ESI-MS: m/z (%): 534.2236
(100, [M+MeOH+H]+, calcd for C30H28N7O3+: 534.2248), 502.1979
(12, [M+H]+, calcd for C29H24N7O2+: 502.1986).
1,1-Dicyano-2,3-bis(dicyanomethylidene)-4,4-bis[methyl(phenyl)-
amino]but-4-ylium-1-ide (15a):
A solution of 4a (156 mg,
0.5 mmol) and tetracyanoethene (70 mg, 0.55 mmol) in acetonitrile
(20 mL) was stirred at 258C for 15 min. Evaporation of solvent, FC
(CH2Cl2/MeOH 98:2), and recrystallization (EtOAc/CH2Cl2/hexanes
2:1:7) gave 15a (198 mg, 90%) as a black solid. Rf =0.23 (SiO2; cy-
clohexane/EtOAc 1:1); m.p. 2358C (decomp.); IR (ATR): n˜ =3065
(w), 2988 (w), 2216 (w), 2193 (s), 1594 (m), 1558 (m), 1541 (s), 1471
(s), 1452 (s), 1432 (m), 1406 (m), 1336 (w), 1276 (w), 1239 (w), 1214
(w), 1159 (w), 1147 (w), 1107 (m), 1081 (m), 1057 (m), 1044 (w),
1022 (w), 1000 (m), 968 (w), 914 (w), 820 (w), 788 (w), 767 (s), 756
(s), 714 (w), 693 (s), 667 (w), 652 (w), 631 cmꢀ1 (w); UV/Vis (CH2Cl2):
lmax (e)=799 (580), 385 nm (14700mꢀ1 cmꢀ1); HR-MALDI-MS: m/z
(%): 463.1390 (56, [M+Na]+, calcd for C26H16N8Na+: 463.1390),
441.1571 (62, [M+H]+, calcd for C26H17N8+: 441.1570), 440.1492
(100, [M]+, calcd for C27H16N8+: 440.1492).
Acknowledgements
This work was supported by the Swiss National Science Foun-
dation and the ERC Advanced Grant No. 246637 (“OPTELO-
MAC”). G.G. acknowledges financial support from the TU and
NAWI Graz. M.Z. thanks the Stiftung ProBono. We thank Dr.
Bruno Bernet for help with NMR interpretations and Prof. Carlo
Thilgen with the nomenclature of zwitterions. We are grateful
to Prof. Enrico Dalcanale (Univ. of Parma) for providing us tet-
raphosphonate cavitand 16. X-ray services were provided by
the Small Molecule Crystallography Center of ETH Zꢃrich
(http://www.smocc.ethz.ch).
(2E/Z)-1,1-Dicyano-2-(1-cyano-2-methoxy-2-oxoethylidene)-3-(di-
cyanomethylidene)-4,4-bis[methyl(phenyl)amino]but-4-ylium-1-
ide (15b): A solution of 4b (173 mg, 0.5 mmol) and tetracyano-
ethene (70 mg, 0.55 mmol) in acetonitrile (20 mL) was stirred at
258C for 15 min. Evaporation of solvent, FC (CH2Cl2/MeOH 98:2),
&
&
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
10
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
quer, J. Garꢄn, J. Orduna, B. Villacampa, R. Alcalꢅ, J. Casado, M. C. R. Del-
gado, J. T. L. Navarrete, M. Allain, J. Am. Chem. Soc. 2005, 127, 8835–
Keywords: [3]cumulenes
·
bond length alternation
·
ˇ
8845; f) Y.-L. Wu, F. Bures, P. D. Jarowski, W. B. Schweizer, C. Boudon, J.-P.
proacetylenic · push–pull chromophores · rotational barrier ·
Gisselbrecht, F. Diederich, Chem. Eur. J. 2010, 16, 9592–9605.
[19] R. Chauvin, C. Lepetit, V. Maraval, L. Leroyer, Pure Appl. Chem. 2010, 82,
769–800.
zwitterions
[20] a) L. Beverina, J. Fu, A. Leclercq, E. Zojer, P. Pacher, S. Barlow, E. W. Van
Stryland, D. J. Hagan, J.-L. Brꢇdas, S. R. Marder, J. Am. Chem. Soc. 2005,
127, 7282–7283; b) J. M. Hales, S. Zheng, S. Barlow, S. R. Marder, J. W.
Perry, J. Am. Chem. Soc. 2006, 128, 11362–11363; c) Y. Yao, Y. Liang, V.
Shrotriya, S. Xiao, L. Yu, Y. Yang, Adv. Mater. 2007, 19, 3979–3983;
d) S. T. Meek, E. E. Nesterov, T. M. Swager, Org. Lett. 2008, 10, 2991–
2993; e) G. Qian, Z. Zhong, M. Luo, D. Yu, Z. Zhang, Z. Y. Wang, D. Ma,
Adv. Mater. 2009, 21, 111–116.
[21] R. Gompper, U. Wolf, Tetrahedron Lett. 1978, 19, 4263–4264.
[22] R. Gompper, R. Kunz, Chem. Ber. 1966, 99, 2900–2904.
[23] W. Tuemmler, B. Wildi, J. Org. Chem. 1958, 23, 1056–1057.
[24] C. Jutz, H. Amschler, Chem. Ber. 1963, 96, 2100–2108.
[25] Hammet parameter of cyano group equals sp =0.66 and of methyl
ester sp =0.45; C. Hansch, A. Leo, R. W. Taft, Chem. Rev. 1991, 91, 165–
195.
[1] J. E. Douglas, B. S. Rabinovitch, F. S. Looney, J. Chem. Phys. 1955, 23,
315–323.
[2] W. R. Roth, G. Ruf, P. W. Ford, Chem. Ber. 1974, 107, 48–52.
[3] W. R. Roth, H.-D. Exner, Chem. Ber. 1976, 109, 1158–1162.
[4] R. Kuhn, B. Schulz, J. C. Jochims, Angew. Chem. Int. Ed. Engl. 1966, 5,
420; Angew. Chem. 1966, 78, 449.
[5] BLA in cumulenes is reported as the (average) difference in the bond
length between the two central-most bonds of the cumulene chain,
see: J. A. Januszewski, D. Wendinger, C. D. Methfessel, F. Hampel, R. R.
Tykwinski, Angew. Chem. Int. Ed. 2013, 52, 1817–1821; Angew. Chem.
2013, 125, 1862–1867.
[6] Average bond lengths were calculated from CSD entries of [3]cumu-
lenes not bound to transition metals and with the same substituents
on both sides (n=24). F. H. Allen, Acta Crystallogr. Sect. B 2002, 58,
380–388.
[26] a) E. A. Meyer, R. K. Castellano, F. Diederich, Angew. Chem. Int. Ed. 2003,
42, 1210–1250; Angew. Chem. 2003, 115, 1244–1287; b) C. Bissantz, B.
Kuhn, M. Stahl, J. Med. Chem. 2010, 53, 5061–5084.
[27] C. J. Brown, A. C. Farthing, Nature 1949, 164, 915–916.
[28] P. K. Gantzel, K. N. Trueblood, Acta Crystallogr. 1965, 18, 958–968.
[29] G. Lepore, S. Migdal, D. E. Blagdon, M. Goodman, J. Org. Chem. 1973,
38, 2590–2594.
[7] D. Bouvy, Z. Janousek, H. G. Viehe, B. Tinant, J.-P. Declercq, Tetrahedron
Lett. 1993, 34, 1779–1782.
[8] P. Gawel, C. Dengiz, A. D. Finke, N. Trapp, C. Boudon, J.-P. Gisselbrecht,
F. Diederich, Angew. Chem. Int. Ed. 2014, 53, 4341–4345; Angew. Chem.
2014, 126, 4430–4434.
[9] Z. Berkovitch-Yellin, L. Leiserowitz, Acta Crystallogr. Sect. B 1977, 33,
3657–3669.
[30] J. Clayden, L. Lemiꢈgre, M. Helliwell, J. Org. Chem. 2007, 72, 2302–2308.
[31] a) W. T. Ford, T. B. Thompson, K. A. J. Snoble, J. M. Timko, J. Am. Chem.
Soc. 1975, 97, 95–101; b) D. Casarini, L. Lunazzi, A. Mazzanti, J. Org.
Chem. 2008, 73, 2811–2818.
[10] J.-D. van Loon, P. Seiler, F. Diederich, Angew. Chem. Int. Ed. Engl. 1993,
32, 1187–1189; Angew. Chem. 1993, 105, 1235–1238.
[11] a) A. Auffrant, B. Jaun, P. D. Jarowski, K. N. Houk, F. Diederich, Chem. Eur.
J. 2004, 10, 2906–2911; b) V. Maraval, L. Leroyer, A. Harano, C. Barthes,
A. Saquet, C. Duhayon, T. Shinmyozu, R. Chauvin, Chem. Eur. J. 2011, 17,
5086–5100.
[32] E. Cox, Rev. Mod. Phys. 1958, 30, 159–162.
[33] C. Dengiz, O. Dumele, S.-i. Kato, M. Zalibera, P. Cias, W. B. Schweizer, C.
Boudon, J.-P. Gisselbrecht, G. Gescheidt, F. Diederich, Chem. Eur. J. 2014,
20, 1279–1286.
[12] P. D. Jarowski, F. Diederich, K. N. Houk, J. Phys. Chem. A 2006, 110,
7237–7246.
[34] a) N. N. P. Moonen, W. C. Pomerantz, R. Gist, C. Boudon, J.-P. Gissel-
brecht, T. Kawai, A. Kishioka, M. Gross, M. Irie, F. Diederich, Chem. Eur. J.
[13] a) Z.-i. Yoshida, Pure Appl. Chem. 1982, 54, 1059–1074; b) B. Tinant, J.-P.
Declercq, D. Bouvy, Z. Janousek, H. G. Viehe, J. Chem. Soc. Perkin Trans.
2 1993, 911–915.
ˇ
2005, 11, 3325–3341; b) F. Bures, W. B. Schweizer, J. C. May, C. Boudon,
J.-P. Gisselbrecht, M. Gross, I. Biaggio, F. Diederich, Chem. Eur. J. 2007,
13, 5378–5387.
[35] Gaussian 09 (Revision D.01), M. J. Frisch, et al., Gaussian, Inc., Walling-
ford CT, 2013. See the Supporting Information for the full list of the au-
thors.
[14] Y.-L. Wu, F. Tancini, W. B. Schweizer, D. Paunescu, C. Boudon, J.-P. Gissel-
brecht, P. D. Jarowski, E. Dalcanale, F. Diederich, Chem. Asian J. 2012, 7,
1185–1190.
[15] a) M. Iyoda, S. Tanaka, M. Nose, M. Oda, J. Chem. Soc. Chem. Commun.
1983, 1058–1059; b) M. Iyoda, S. Tanaka, H. Otani, M. Nose, M. Oda, J.
Am. Chem. Soc. 1988, 110, 8494–8500.
[16] a) M. I. Bruce, J. R. Rodgers, M. R. Snow, A. G. Swincer, J. Chem. Soc.
Chem. Commun. 1981, 271–272; b) 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; c) S.-i. Kato, F. Diederich, Chem.
Commun. 2010, 46, 1994–2006; d) M. I. Bruce, Aust. J. Chem. 2011, 64,
[36] a) C. J. Jçdicke, H. P. Lꢃthi, J. Chem. Phys. 2002, 117, 4146–4156; b) C. J.
Jçdicke, H. P. Lꢃthi, J. Am. Chem. Soc. 2003, 125, 252–264; c) P. Elliott, K.
Burke, F. Furche, Rev. Comput. Chem. 2009, 26, 91–165; d) E. C. Varkey,
J. Hutter, P. A. Limacher, H. P. Lꢃthi, J. Org. Chem. 2013, 78, 12681–
12689.
[37] For previous examples of paramagnetic CA-RE reaction products, see:
a) M. Kivala, T. Stanoeva, T. Michinobu, B. Frank, G. Gescheidt, F. Dieder-
ich, Chem. Eur. J. 2008, 14, 7638–7647; b) M. Kivala, C. Boudon, J.-P. Gis-
selbrecht, B. Enko, P. Seiler, I. B. Mꢃller, N. Langer, P. D. Jarowski, G. Ge-
scheidt, F. Diederich, Chem. Eur. J. 2009, 15, 4111–4123; c) B. Breiten, Y.-
L. Wu, P. D. Jarowski, J.-P. Gisselbrecht, C. Boudon, M. Griesser, C.
Onitsch, G. Gescheidt, W. B. Schweizer, N. Langer, C. Lennartz, F. Dieder-
ich, Chem. Sci. 2011, 2, 88–93; d) B. Breiten, M. Jordan, D. Taura, M. Zali-
bera, M. Griesser, D. Confortin, C. Boudon, J.-P. Gisselbrecht, W. B.
Schweizer, G. Gescheidt, F. Diederich, J. Org. Chem. 2013, 78, 1760–
1767.
[38] a) C. Dehu, F. Meyers, J. L. Brꢇdas, J. Am. Chem. Soc. 1993, 115, 6198–
6206; b) J. Beckmann, D. Dakternieks, A. Duthie, C. Mitchell, M. Schꢃr-
mann, Aust. J. Chem. 2005, 58, 119–127.
[39] To compare, average CꢀN bond lengths calculated from CSD entries
are: 1.35 ꢀ for N,N-dialkylcyclohexa-2,5-dien-1-iminium cations (n=71)
and 1.38 ꢀ for N,N-dialkylanilines (n=2505). For histograms and more
details, see Section S4.
ˇ
77–103; e) M. Stefko, M. D. Tzirakis, B. Breiten, M.-O. Ebert, O. Dumele,
W. B. Schweizer, J.-P. Gisselbrecht, C. Boudon, M. T. Beels, I. Biaggio, F.
Diederich, Chem. Eur. J. 2013, 19, 12693–12704; f) T. Shoji, M. Maruya-
ma, E. Shimomura, A. Maruyama, S. Ito, T. Okujima, K. Toyota, N. Morita,
J. Org. Chem. 2013, 78, 12513–12524; g) M. Betou, N. Kerisit, E. Meledje,
Y. R. Leroux, C. Katan, J.-F. Halet, J.-C. Guillemin, Y. Trolez, Chem. Eur. J.
2014, 20, 9553–9557.
[17] For the other transformations of tetraaryl[3]cumulenes with TCNE, see:
a) N. Islam, M. Tsukayama, Y. Kawamura, Int. J. Mod. Phys. B 2006, 20,
4619–4624; b) S. Ueta, K. Hida, M. Nishiuchi, Y. Kawamura, Org. Biomol.
Chem. 2014, 12, 2784–2791. For the transformation of tetraaryl[5]cu-
mulenes with TCNE, see: c) J. A. Januszewski, F. Hampel, C. Neiss, A.
Gçrling, R. R. Tykwinski, Angew. Chem. Int. Ed. 2014, 53, 3743–3747;
Angew. Chem. 2014, 126, 3818–3822.
[18] a) R. Gompper, H.-U. Wagner, E. Kutter, Chem. Ber. 1968, 101, 4123–
4143; b) S. Inoue, Y. Aso, T. Otsubo, Chem. Commun. 1997, 1105–1106;
c) T. Kawase, M. Wakabayashi, C. Takahashi, M. Oda, Chem. Lett. 1997,
26, 1055–1056; d) A. S. Batsanov, M. R. Bryce, M. A. Coffin, A. Green,
R. E. Hester, J. A. K. Howard, I. K. Lednev, N. Martꢄn, A. J. Moore, J. N.
Moore, E. Ortꢄ, L. Sꢅnchez, M. Savirꢆn, P. M. Viruela, R. Viruela, T.-Q. Ye,
Chem. Eur. J. 1998, 4, 2580–2592; e) R. Andreu, M. J. Blesa, L. Carras-
[40] To compare, average CꢀN bond lengths calculated from CSD entries
are: 1.33 ꢀ for N,N,N,N-tetraalkylamidinium cations (n=836) and 1.36 ꢀ
for N,N,N,N-tetraalkyldiaminomethylenes (n=22552). For histograms
and more details, see Section S4.
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
11
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Full Paper
[41] To compare, average CꢀC bond lengths calculated from CSD entries are
1.40 ꢀ for allyl anions (n=155). For histograms and more details, see
Section S4.
[43] a) R. A. van Delden, N. Koumura, N. Harada, B. L. Feringa, Proc. Natl.
Acad. Sci. USA 2002, 99, 4945–4949; b) R. A. van Delden, M. K. J. ter
Wiel, M. M. Pollard, J. Vicario, N. Koumura, B. L. Feringa, Nature 2005,
437, 1337–1340; c) J. Wang, B. L. Feringa, Science 2011, 331, 1429–
1432.
[42] a) P. Delangle, J.-P. Dutasta, Tetrahedron Lett. 1995, 36, 9325–9328; b) P.
Delangle, J.-C. Mulatier, B. Tinant, J.-P. Declercq, J.-P. Dutasta, Eur. J. Org.
Chem. 2001, 3695–3704; c) R. M. Yebeutchou, F. Tancini, N. Demitri, S.
Geremia, R. Mendichi, E. Dalcanale, Angew. Chem. Int. Ed. 2008, 47,
4504–4508; Angew. Chem. 2008, 120, 4580–4584; d) E. Biavardi, G. Bat-
tistini, M. Montalti, R. M. Yebeutchou, L. Prodi, E. Dalcanale, Chem.
Commun. 2008, 1638–1640; e) R. M. Yebeutchou, E. Dalcanale, J. Am.
Chem. Soc. 2009, 131, 2452–2453; f) F. Tancini, R. M. Yebeutchou, L. Pir-
ondini, R. De Zorzi, S. Geremia, O. A. Scherman, E. Dalcanale, Chem. Eur.
J. 2010, 16, 14313–14321.
[44] E. R. Kay, D. A. Leigh, F. Zerbetto, Angew. Chem. Int. Ed. 2007, 46, 72–
191; Angew. Chem. 2007, 119, 72–196.
Received: December 20, 2014
Published online on && &&, 0000
&
&
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
12
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Paper
FULL PAPER
&
Proacetylenicity
P. Gawel, Y.-L. Wu, A. D. Finke, N. Trapp,
M. Zalibera, C. Boudon, J.-P. Gisselbrecht,
W. B. Schweizer, G. Gescheidt,*
F. Diederich*
¼
A series of push–pull-substituted [3]cu-
mulenes was synthesized. Extensive UV/
Vis, X-ray, VT NMR, and computational
studies confirmed their proacetylenic
character. The measured rotational bar-
riers of the new cumulenes are in the
range of those of sterically crowded
single bonds and reach values down to
DG =11.8 kcalmolꢀ1. The proacetylenic
&& – &&
reactivity of the central C=C double
bond was proven by the facile, high-
yielding cycloaddition-retroelectrocycli-
zation reaction with tetracyanoethene
at room temperature yielding stable
zwitterions.
Push–Pull Buta-1,2,3-trienes:
Exceptionally Low Rotational Barriers
of Cumulenic C=C Bonds and
Proacetylenic Reactivity
Chem. Eur. J. 2015, 21, 1 – 13
www.chemeurj.org
13
ꢁ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ