Dialkylated Dihydroazulene and Vinylheptafulvene Derivatives – Synthesis and Switching Properties

Article abstract of DOI:10.1002/ejoc.201700446

Functionalization of molecular photoswitches at specific positions offers a way of tuning their switching properties. The work presented here describes the development of a new protocol towards the synthesis and functionalization of the dihydroazulene/vinylheptafulvene (DHA/VHF) photo-/thermoswitch. The key step is a facile condensation reaction using a functionalized tropone derivative, and the new synthetic method was employed in the regioselective preparation of a novel dimethyl-substituted DHA photoswitch. This compound presents the first accessible derivative with alkyl groups on the seven-membered ring of the system ― at positions C5 and C7. From experimental and theoretical kinetics studies on the thermal VHF-to-DHA electrocyclic reaction, the influence of the electron-donating methyl groups was elucidated. In addition, we subjected a small selection of 4,8a-dialkylated compounds to a computational study to elucidate how steric interactions can alter the relative DHA–VHF stabilities. The synthetic methodology offers access to novel DHA scaffolds and substitution patterns, which in turn can lead to systems that can help to address broader questions concerning the potential applicability of the DHA/VHF system in the context of solar-thermal energy-storage systems or other photochromic applications.

Full text of DOI:10.1002/ejoc.201700446

A Journal of
Accepted Article
Title: Dialkylated Dihydroazulene / Vinylheptafulvene Derivatives -
Synthesis and Switching Properties
Authors: Nickie C. M. Lubrin, Alexandru Vlasceanu, Benjamin N.
Frandsen, Anders B. Skov, Martin Drøhse Kilde, Kurt V.
Mikkelsen, and Mogens Broendsted Nielsen
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Org. Chem. 10.1002/ejoc.201700446
Link to VoR: http://dx.doi.org/10.1002/ejoc.201700446
Supported by

10.1002/ejoc.201700446
European Journal of Organic Chemistry
FULL PAPER
Dialkylated Dihydroazulene / Vinylheptafulvene Derivatives –
Synthesis and Switching Properties
Nickie C. M. Lubrin,[a] Alexandru Vlasceanu,[a] Benjamin N. Frandsen,[a] Anders B. Skov,[a] Martin
Drøhse Kilde,[a] Kurt V. Mikkelsen[a] and Mogens Brøndsted Nielsen*[a]
Abstract: Functionalization of molecular photoswitches at specific
positions offers a way of tuning their switching properties. The work
presented here describes the development of a new protocol
photochromism. Upon irradiation, DHA (1a) undergoes
photoisomerization to the meta-stable VHF (1b) isomer, which in
time thermally reverts to DHA with a simultaneous release of
energy (Scheme 1). The system was initially discovered and
investigated by Daub and co-workers,[3] and in recent years it
has been further tuned by our group in various ways.[4]
Summaries of the current progress towards the synthesis and
functionalization[4h] as well as the properties and behavior[4i] of
the DHA/VHF system have been accounted for in recent reviews.
towards
the
synthesis
and
functionalization
of
the
dihydroazulene/vinylheptafulvene (DHA/VHF) photo-/thermoswitch.
The key step is a facile condensation reaction using a functionalized
tropone derivative, and the new synthetic method was employed in
the regioselective preparation of a novel dimethyl-substituted DHA
photoswitch. This compound presents the first accessible derivative
incorporating alkyl groups in the seven-membered ring of the system
– at positions C5 and C7. From experimental and theoretical kinetics
studies on the thermal VHF-to-DHA electrocyclic reaction, the
influence of the electron-donating methyl groups was elucidated. In
NC
CN
8 NC
7
5
CN
hv
1
8a
6
2
!
3a
4
addition, we subjected
a small selection of 4,8a-dialkylated
3
compounds to calculational study to elucidate how steric
a
DHA (1a)
VHF (1b)
interactions can alter the relative DHA-VHF stabilities. The synthetic
methodology offers access to novel DHA scaffolds and substitution
patterns which in turn can lead to systems that can help address the
broader questions concerning the potential applicability of the
DHA/VHF system in the context of solar-thermal energy storage
systems or other photochromic applications.
Scheme 1. Photo/thermal switching of the parent DHA (1a)/VHF (1b) system,
also showing the commonly used numbering scheme.
Since their discovery, the synthesis of substituted DHA
scaffolds has proceeded through one of two main strategies
(Routes A and B) presented in Scheme 2. Route A provides a
direct path to the desired DHA scaffold through an [8+2]
cycloaddition of 8-methoxyheptafulvene with dicyanoethylenes,
from which methanol is subsequently eliminated using P2O5.[5]
The second general strategy (Route B) furnishes DHA in an
indirect manner via the corresponding vinylheptafulvene (VHF)
intermediate, which is made through the combination of
Introduction
In recent years, the field of molecular photoswitches has
benefited from a particular rise in interest and involvement. This
can be mainly attributed to the numerous applications that have
been envisioned by scientists across all fields of research
including chemistry, biochemistry and physics.[1] Due to such
widespread attention, several books and reviews have
documented the ongoing progress particularly towards the
development of organic molecular switches.[2] The foundation of
any new technology arising from this field, however, is still
dependent on the discovery of new compounds and materials
that exhibit desirable properties for their relevant applications.
Ultimately, the successful preparation of such innovative
materials is grounded on the chemist’s synthetic toolbox.
Among other organic molecular photoswitches, the
tropylium with
a
malononitrile derivative (or with the
corresponding carbonyl derivative and a subsequent reaction
with
malononitrile),
followed
by
an
oxidation
(dehydrogenation).[6] The work presented here outlines Route C,
which offers particular advantages over the other two.
Although routes A and B provide the opportunity for facile
functionalization at positions 2 and 3 of the DHA scaffold (see
Scheme 1), late-stage functionalization of the system at all
positions (1-8) has remained limited. The regioselective
introduction of substituents on the seven-membered ring has
been particularly challenging.[4e, 7] Functionalization of the seven-
membered ring has been investigated at both an early synthetic
stage (before formation of the DHA scaffold) and at a later one
(after DHA formation). The most widely employed late-stage
functionalization method exploits the out-of-plane C7-C8 double
bond of 1a which can be selectively brominated (Scheme 3,
a).[4c,d,j] The subsequent elimination of HBr is then used to yield a
7-bromo-substituted DHA that can subsequently be subjected to
metal-catalyzed cross-coupling reactions.[8]
dihydroazulene/vinylheptafulvene
(DHA/VHF)
photo-
/thermoswitch presents an example of one-way (T-type)
[a]
N. C. M. Lubrin, A. Vlasceanu, B. N. Frandsen, A. B. Skov, M. D.
Kilde, Prof. Dr. K. V. Mikkelsen and Prof. Dr. M. B. Nielsen
Department of Chemistry and Center for Exploitation of Solar
Energy
University of Copenhagen
Universitetsparken 5, DK-2100 Copenhagen Ø, Denmark
E-mail: mbn@chem.ku.dk
http://chem.ku.dk/research_sections/solarenergy/
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201700446
European Journal of Organic Chemistry
FULL PAPER
Route A
In an effort to expand the scope of late-stage
functionalization, oxidation of the DHA scaffold to the azulenium
cation (Scheme 3, b) was successfully conducted; however,
subsequent nucleophilic attack of this system could not be
achieved regioselectively, and a low-yielding tautomerization
reaction was required to form the photoactive 1,8a-DHA.[7]
Alternatively, early functionalization of the seven-
membered ring has also been attempted by employing
substituted tropylium salts as starting materials (Scheme 3, c).[4e]
The substituents of these building blocks, however, were not
found to be sufficiently directing to result in a regioselective
attack of the ring. This method has resulted in product mixtures
containing all possible regioisomers, which proved difficult to
purify by chromatographic techniques.[4e]
O
O
NC
[C7H7]BF4
CN
R1
R1
R1
R2
NC
R2
NC
CN
CN
[8+2]
Cyclo-
NC
CN
R1
addition
(R2 = H)
NC
CN
R1
BF4
OMe
R2
R2
Condensation
R4 = H, Me
R2 = H
Oxidation
R4
We then rationalized that regioselectivity might be
achievable by using a functionalized tropone derivative as the
key building block, which in this work led to the development of
O
NC
NC
CN
R1
R1 = Ph
R4
R4
Route
C shown in Scheme 2. Provided with access to
R2
substituted tropone derivatives, a class of products still under
development, the range of accessible DHA products could
increase significantly. In addition, the key condensation reaction
of Route C may be considered a milder alternative to the
oxidation step of Route B, which has previously been reported to
present certain functional-group incompatibilities and low
reaction yields.[9] Furthermore, whereas cross-coupling reactions
using 7-bromo DHA are limited to the introduction of sp and sp2
hydrbidized carbon substituents, the method presented here
allows for expansion of the substituent scope to include sp3-
hybridized carbon-based R-groups on the seven-membered ring.
Studies on such derivatives allow for elucidation of the influence
of alkyl substitution for the VHF-to-DHA ring closure reaction,
while previous studies have focused on direct conjugation
effects via aryl or arylethynyl substituents.
NC
CN
R1
Route C
!
(This work)
Route B
R4
R4
R4
R2
NC
CN
R1
CN
R1
!
R2
R2
Scheme 2. Established synthetic routes towards functionalized DHA scaffolds
(Routes and B) in addition to this work (Route C). LDA lithium
A
=
diisopropylsilylamide.
a)
NC
CN
Br
NC
CN
Ph
1) Br2, CH2Cl2
2) LiHMDS,THF
Ph
Results and Discussion
Synthesis. In this work, synthetic Route C is explored primarily
with a focus on the regioselective functionalization of the DHA
scaffold. Although seemingly trivial starting materials, the
synthesis of tropones has not been very thoroughly explored in
literature. The lack of detailed investigation into this class of
compounds has perhaps been due, in part, to the relatively
uncommon occurrence of 7-membered, all-carbon atom rings in
natural products.[10]
1a
7-Bromo DHA
cross-
coupling
R
NC
CN
Ph
b)
R3
NC
CN
(i-Pr)3Si
Li
R1
5 Regioisomers
Via
tetrafluoroborate (2), the synthesis of tropone 3 could be
accomplished according to procedure by Reingold and
a
disproportionation
reaction
of
tropylium
THF, -78 °C
BF4
R2
a
DiNardo (Scheme 4).[11] As chromatographic methods result in
product loss, partial purification of the crude tropone could be
achieved by extraction with hot heptane (see SI). The optimal
conditions for a condensation reaction between tropone 2 and
dicyanoethylene derivative 4 were first investigated towards the
synthesis of parent DHA 1a by varying the temperature, solvent,
reaction time, and catalysts (Scheme 4). Investigations were
initially carried out to determine the effect of either acid, base, or
Lewis acid catalysis on the success of the condensation reaction
(see SI for details).
NC
CN
Ph
Et3N,
CH2Cl2,
-78 °C
c)
1)
NC
CN
R
Ph
2) TrBF4, Et3N, DCE
3) 80 °C, PhMe
BF4
R
R = Ph
R = C2Si(i-Pr)3
Scheme 3. Regio- and non-regioselective methods of functionalizing the
seven-membered ring. LiHMDS = lithium hexamethyldisalazide; DCE = 1,2-
dichloroethane.
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201700446
European Journal of Organic Chemistry
FULL PAPER
NC
NC
CN
NH
Bu3SnH, AIBN
OH
4, [Base]
C6H6
21%
X = Cl
Solvent
X
NC
CN
7
O
O
5
X
BF4
NC
CN
CH3(CH2)15NMe3Cl
CHX3, NaOH, H2O
Na2CO3
4
t-Bu
Si
O
8 (X = Cl; 20%)
10 (X = Br; 23%)
6
MeCN
51%
Ac2O, !
MeO
OMe
1a
3
35%
2
AIBN,
9
X = Br
4, [H+]
No reaction
heptane
56%
Solvent
Scheme 4. Conditions for the synthesis of 1a via Route C (see SI).
Scheme 5. Synthesis of 2,7-dimethyltropone 6. AIBN = azabisisobutyronitrile.
Over the course of the screening, basic aldol-type conditions
were consistently observed to promote crotononitrile
Although this method is satisfactory in yielding the desired
2,7-dimethyltropone, the low yielding reactions as well as the
hazardous and toxic reagents employed in the ring expansion
could be circumvented. Recent work by Studer et al.[14] reports
that various silylated 1,4-cyclohexadienes such as 9 can be
employed as superior tin hydride substitutes in a variety of
radical chain reductions. These conditions, however, were found
to be more successful on bromo- or iodo-functionalized starting
materials, and so the synthesis of 2,7-dimethyltropone 6 was
instead achieved in a yield of 56% through the dibromo
derivative 10 (Scheme 5).
dimerization yielding byproduct
5 instead of the intended
condensation with the ketone, which was not observed to any
degree. This is perhaps unsurprising as such reactivity has been
previously reported in literature.[12] In an acidic reaction medium
on the other hand, crotononitrile dimerization could be avoided
completely; however, no reactivity of the substrates could be
observed at all, even under forcing conditions (MW, 200 °C; MW
= microwave heating). Conducting the reaction in a buffered
system under Knoevenagel-like conditions was also
unsuccessful in achieving any reactivity between the starting
materials. Tropone activation using a variety of Lewis acids was
also attempted but this too did not show any trace of the desired
reactivity. The reaction towards 1a was, however, found to
proceed in the presence of acid anhydrides (Table S1, SI).
Gratifyingly, using acetic anhydride as solvent, a yield of 35%
was achieved of 1a.
To test the regioselectivity of the newly developed method,
the synthesis of 2,7-dimethyl-tropone 6 was first undertaken
according to a method reported by Barbier et al.[13] (Scheme 5)
for similar transformations. In the first step, dichlorocarbene is
generated in situ and undergoes a directed addition to 2,6-
dimethylphenol 7 at the ortho position to the OH group to form
product 8. Although initial attack by the carbene at the para
position could also be expected, no such byproduct or any other
major products could be isolated upon chromatographic
purification, indicating that the low yield is likely due to
decomposition. Finally, compound 8 was subjected to a ring
expansion to form 6 in a yield of 21% using tributyltin hydride
and azabisisobutyronitrile (AIBN).
The condensation between dimethyltropone
6
and
dicyanoethene derivative 4 was then successfully conducted,
providing the DHA 11a via VHF 11b (Scheme 6). Further
optimization revealed that the addition of DBU to the reaction
mixture was advantageous. Nucleophilic attack at the tropone
ring is not observed to occur at the carbonyl carbon (C1) but
rather, it can be inferred that the attack must take place at C4 as
the product 11a was isolated. The mechanism of this
condensation is thus hypothesized to occur as described in
Scheme 6. It is thought that the steric effects of the methyl
groups flanking the carbonyl carbon are sufficient to disfavor
attack at this position or at C3. The new synthetic protocol thus
allowed the isolation of one regioisomeric product in high yield
(52%) in contrast to the previous method of using
functionalized tropylium ion as substrate.
a
Finally,
Wittig
and
Horner-Wadsworth-Emmons
olefinations were also attempted towards the synthesis of VHF
1b, but none of these attempts turned out successful (see SI).
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201700446
European Journal of Organic Chemistry
FULL PAPER
6e]
on the seven-membered ring.[4c,d,
Interestingly, we see that
O
O
O
O
O
O
the presence of two methyl groups has a much stronger
influence than expanding the π-conjugated system by a phenyl
substituent. Incorporation of an inductively electron-withdrawing
bromo substituent has the opposite effect on the half-life,
increasing it by a factor of four.
O
O
!
- AcO
6
4, DBU
O
O
NC
CN
40000
NC
CN
+
-
H
O
O
11b/12b
35000
30000
25000
- AcOH
NC
11a
20000
15000
10000
5000
0
CN
NC
CN
!
52%
11b
11a
250
300
350
400
450
500
550
600
650
700
Wavelength (nm)
40000
Scheme 6. Mechanistic proposal for the condensation reaction between 2,7-
dimethyltropone 6 and the dicyanoethene derivative 4 leading to 5,7-dimethyl
DHA 11a. DBU = 1,8-diazabicyclo[5.4.0]undec-7-ene.
35000
30000
25000
20000
15000
10000
5000
0
11b/12b
11a/12a
Optical Properties and Switching Studies. The UV-Vis
absorption spectrum of DHA 11a is shown in Figure 1. It exhibits
a characteristic absorption at λmax 338 nm, blueshifted relative to
that of 1a (λmax 353 nm[6e]). Upon irradiation, the DHA is
converted to VHF 11b with a characteristic absorption at λmax
500 nm, redshifted relative to that of VHF 1b (λmax 470 nm[6e]). In
time, 11b undergoes a thermal ring closure to form 11a, and we
determined a half-life of 36 min (MeCN, 25 °C), whereas that
from 1b to 1a has been previously determined as 218 min.[6e]
The significantly shorter half-life of 11b is considered to be due
to the electron-donating character of the methyl substituents as
electron-donating groups at the seven-membered ring are
known to enhance the thermal back reaction (TBR).[4g] As with
most systems previously investigated,[4c,d] VHF 11b exhibits two
sites of possible ring closure, accessible via double bond
isomerization that may occur both upon irradiation and thermally
(Scheme 7), which can lead to the 6,8-disubstituted DHA/VHF
system 12a/12b. This isomerization is evident by the fact that
the UV-Vis absorption spectrum after one cycle is not identical to
that of DHA 11a as the solution now also contains DHA 12a.
Thus, after one cycle the DHA absorption spectrum presents a
new absorption maximum at 327 nm, assigned to 12a, as well
as a shoulder at the original 338 nm (11a). The different
absorption properties of the two DHA isomers 11a and 12a are
further supported by calculations (vide infra), and the isomeric
mixture was also confirmed by NMR spectroscopic studies (SI,
Figure S4).
250
300
350
400
450
500
550
600
650
700
Wavelength (nm)
Figure 1. UV-Vis absorption spectra of 11a showing conversion to 11b (in
equilibrium with 12b) by irradiation at 338 nm (top) and thermal conversion of
11b/12b into 11a/12a over time at 25 °C in MeCN (bottom).
CN
CN
NC
NC
Ph
11b/12b
t1/2 = 36 min
t1/2 = 206 min
CN
NC
CN
NC
Br
1b
t1/2 = 218 min
t1/2 = 909 min
Figure 2. Differences in half-lives between VHF derivatives, based on electron
donating or withdrawing substituents on the seven-membered ring (measured
in MeCN at 25 °C; this work and references [4c,d, 6e]).
The decay in VHF absorption at 500 nm could be fitted by
one exponential function. In consequence, the two VHFs 11b
and 12b seem to undergo the TBR at similar rates. We achieve
a half-life of 36 min, which is a factor of six smaller than that of
1b (218 min). Figure 2 shows the differences in half-lives for the
TBR of various VHF derivatives, containing different substituents
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201700446
European Journal of Organic Chemistry
FULL PAPER
N
13a/b represent the isomeric couple to 11a/b that would have
formed from nucleophilic attack instead at the carbonyl of the
tropone. For the calculations, the DFT method and basis set
M06-2X/6-311+g(d) were employed, again using the IEFPCM
solvent model, which have previously been shown to provide
accurate results for the parent 1a/1b system.[15] Photochemical
ring opening of DHA (a) initially results in the formation of s-cis
VHF (bs-cis),[16] which exists in thermal equilibrium with the more
stable s-trans VHF (bs-trans). Thus, both conformers should be
considered when estimating thermal ring closure rates. A
representation of the composite reaction coordinate (bs-cis à a
and bs-cis ↔ bs-trans) for the ground state isomerization reactions
of the systems considered here is depicted in Figure 4.
N
N
N
N
NC
CN
hv
N
!
11a
11b s-cis
11b s-trans
NC
CN
hv
N
!
N
12a
12b s-cis
12b s-trans
Scheme 7. One ring opening/closure cycle provides a mixture of DHA isomers
11a/12a. The VHFs 11b/12b correspond to E/Z isomers.
CN
NC
NC
R
The electronic excitation (UV-Vis) spectra of the two DHA
species were computationally modeled (CAM-B3LYP/6-
311+G(d,p) using the IEFPCM method for solvent modeling) and
are shown to accurately reflect the experimental observations
(Figure 3). Thus, partial isomerization from 11a to 12a is
supported by a sligthly blueshifted absorption of the latter.
R
CN
R = H (1a/b)
hv
R = Me (13a/b)
R = i-Pr (14a/b)
R = t-Bu (15a/b)
!
a
b
R
R
Scheme 8. Various DHA/VHF systems subjected to a computational study.
45000
Experimental
40000
11a
11b
11a/12a mix
35000
30000
25000
20000
15000
10000
5000
0
250
300
350
400
450
500
550
600
650
700
Wavelength (nm)
45000
40000
35000
30000
25000
20000
15000
10000
5000
Calculated
11a
11b
12a
Figure 4. Schematic representation of the composite reaction coordinate (bs-
à a and bs-cis ↔ bs-trans) for the ground state isomerization reactions of the
cis
systems considered here.
In theory, a model for the thermal ring closure from VHF
(bs-cis↔bs-trans) to DHA (a) can be constructed in at least two
ways, both employing the Eyring equation to estimate reaction
rates (kcalc). The first considers the bs-cis conformer as the
starting point and the free energy of activation (ΔGs-cis‡) to the
transition state connecting bs-cis to a. For all systems studied
here, two transition states were identified, yet only the one
exhibiting the lowest energy is reported. In the calculation of an
accurate reaction rate constant from bs-cis to a, the pre-
equilibrium, bs-cis↔bs-trans, can be accounted for by factoring
K/(K+1) into the rate equation, where K = [bs-cis]/[bs-trans]. This
factor reflects the limited amount of bs-cis present, and thereby
the contribution of the pre-equilibrium on the reaction rate from
bs-cis to a. The rate equation describing this approach is derived
in the SI. An alternative approach considers bs-trans as the
0
250
300
350
400
450
500
550
600
650
700
Wavelength (nm)
Figure 3. Top: Experimental UV-Vis absorption spectra of 11a, 11b, and an
isomeric mixture of 11a/12a obtained after exposing 11a to a light/heat cycle;
Bottom: Calculated UV-Vis absorption spectra of 5,7-dimethyl DHA 11a, VHF
11b and 6,8-dimethyl DHA isomer 12a. Both experimental and theoretical
spectra were obtained at 25 °C in MeCN.
Computational Study. A theoretical investigation of substituted
systems 11a/11b, 12a/12b, as well as the 4,8a-dialkylated
systems 13-15a/b (Scheme 8) in comparison with 1a/1b was
then conducted in order to estimate DHA-VHF Gibbs free energy
differences (ΔG) and back reaction barriers (ΔG‡). Compounds
‡
starting point and ΔGs-trans as the free energy of activation. One
advantage of this method is that it requires no knowledge of the
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201700446
European Journal of Organic Chemistry
FULL PAPER
reactive conformer and thus avoids any such assumptions that
would otherwise restrict the reaction coordinate to pass through
the b s-cis minimum. This approach can be commonly refered to
as a lowest conformer transition state theory (LC-TST) [17] and
works best when this conformer is predominant, at which point it
also converges with the pre-equilibrium method (SI, Figure S5).
The rate equation describing this approach is also derived in the
SI.
The relative stabilities of VHF conformers bs-cis and bs-trans
,
which on average exhibit an energy difference of approx. 10 kJ
mol-1, translate to a preference of nearly 99% towards the bs-trans
conformation. Although only the pre-equilibrium approach has
the correct behavior for k when K is variable, the LC-TST
method is presented here as a reasonable approximation.
Theoretical estimates of the reaction rate constants (kcalc) based
on both methods are listed in Table 1. The pre-equilibrium
approach yields a rate constant for the conversion of 1b, which
is close to the experimental one, while the LC-TST method
comes within approx. a factor of 2 of this. Experimentally, it was
not possible to differentiate between the rates of ring closure of
11b and 12b, and the overall rate constant was found to be 3.18
x 10-4 s-1, being six times that of 1b. In line with the experimental
results, the calculations do indeed predict a faster TBR of 11b
and 12b relative to that of 1b, but both methods seem to
underestimate the influence of the methyl groups. With respect
to the Gibbs free energy difference between s-trans VHF and
DHA isomers, the calculations predict VHF 12b (s-trans) to be
slightly more stable than DHA 12a.
Computational investigations on systems 13-15a/b (so far
synthetically elusive compounds) show that increasing steric
bulk at the positions of ring closure (C4 and C8a) destabilizes
the DHA isomer relative to the VHF (Table 1). In fact, VHF 15b
(s-trans) is significantly more stable than DHA 15a. With
increasing bulkiness of the alkyl substituent, we observe an
increase in the C8a−C(alkyl) bond length. This increase is more
significant for the DHA species (a) than for the VHF species (b):
1.537 Å (13a) -> 1.580 Å (14a) -> 1.594 Å (15a) versus 1.509 Å
(13b) -> 1.522 Å (14b) -> 1.540 Å (15b). The conformational
distortion resulting from increasing steric bulk at the 4 and 8a
positions can also be clearly observed in the calculated UV-Vis
absorption spectra of s-trans VHF isomers 13b, 14b, and 15b
(Figure 5). Here, a corresponding decrease in intensity and
blueshift of the characteristic VHF absorption band is observed
with increasing steric bulk.
Figure 5. Calculated UV-Vis absorption spectra (CAM-B3LYP/6-311+G(d,p)
using the IEFPCM solvent model) of s-trans VHF isomers 13-15b at 25 °C in
MeCN.
Table 1. Gibbs free energy differences (ΔG, energy difference between s-
trans VHF and DHA) and VHF-to-DHA rate constants calculated for 1a/b as
well as 11-15a/b at 25 °C in MeCN using the M06-2X/6-311+g(d) method and
the IEFPCM solvent model with acetonitrile. Harmonic frequencies and default
approximations in Gaussian 09 were used.
[a]
[b]
[c]
ΔG
kcalc
kcalc
kexp
System
[kJ mol-1]
[10-4 s-1]
0.58
1.39
0.77
0.41
-
[10-4 s-1]
0.28
0.62
0.41
0.21
-
[10-4 s-1]
0.54
3.18
3.18
1a/1b
16
9
-3
18
-2
-21
11a/11b
12a/12b
13a/13b
14a/14b
15a/15b
-
-
-
-
-
[a] Calculated using the pre-equilibrium approach described in the SI. [b]
Calculated using the lowest conformer transition state theory (LC-TST)
described in the SI. [c] Experimental rate constants.
Conclusions
A new method for the synthesis of the DHA scaffold was
developed (Route C), which avoids an occasionally problematic
oxidation step by replacing it with a rather simple condensation
reaction employing tropone as a starting material. This method
was successfully employed in
a
novel regioselective
functionalization of the seven-membered ring using 2,7-
dimethyltropone as substrate, whose synthesis was moderately
optimized. This led to the synthesis of a novel DHA/VHF system
(11a/11b), which was found to exhibit a significantly faster VHF-
to-DHA thermal ring closure than the parent system (1a/1b). The
faster ring closure of 11b was mirrored in a calculational study
although this study did not provide as large a rate enhancement
as observed experimentally. By placing bulky alkyl groups at
positions C4 and C8a, calculations reveal that the relative DHA-
VHF stabilities can be reversed, rendering VHF more stable than
DHA. In all, this work has shown for the first time the influence of
alkyl substitution in the seven-membered ring of the DHA/VHF
system. The work has also shown that DFT calculations provide
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201700446
European Journal of Organic Chemistry
FULL PAPER
NMR (126 MHz, CDCl3) δ 200.92, 138.40, 137.86, 133.25, 123.74, 77.51,
57.70, 24.40, 15.48 ppm. HRMS (ESI, m/z) found 205.01850 [M+H]+,
calc. (C9H11Cl2O)+: 205.01815 [M+H]+.
important guidelines for the design of alkyl-substituted
derivatives with specific switching and optical properties.
6-(Dibromomethyl)-2,6-dimethylcyclohexa-2,4-dien-1-one (10). To a
solution of 2,6-dimethylphenol 7 (20.0 g, 164 mmol) in CHBr3 (47.0 mL,
Experimental Section
537
mmol)
was
added
a
25%
aqueous
solution
of
cetyltrimethylammonium chloride (0.541 mL, 0.423 mmol), and the
reaction mixture was heated to 50 °C under an ambient atmosphere. To
this, a solution of NaOH (50.4 g, 1.26 mol) in H2O (115 mL) was added
over 30 min under mechanical stirring, and the mixture was stirred at this
temperature for a further 4 h. The reaction was then quenched with H2O
(250 mL) and extracted with Et2O (3 x 200 mL). The organic fractions
were combined and washed with H2O (2 x 200 mL) and brine (2 x 200
mL), dried over MgSO4, filtered, and the solvent was removed in vacuo.
The residue was purified by dry column vacuum chromatography (SiO2,
0-50% CHCl3/heptane, 2% increments, 150 mL fractions, then 100%
CHCl3) to yield title compound 10 as a pale yellow oil (11.2 g, 23%). TLC
Rf= 0.37 (20% EtOAc/heptane). 1H NMR (500 MHz, CDCl3) δ 6.90 – 6.87
(m, 1H), 6.60 – 6.57 (m, 1H), 6.41 (dd, J = 9.7, 6.1 Hz, 1H), 5.97 (s, 1H),
1.92 (s, 3H), 1.29 (s, 3H) ppm. 13C NMR (126 MHz, CDCl3) δ 200.65,
139.54, 138.40, 133.38, 123.51, 57.25, 51.01, 25.91, 15.50 ppm. HRMS
(ESI, m/z) found: 314.89903 [M+Na]+, calc. (C9H10Br2ONa)+: 314.89906
[M+Na]+; found: 292.91714 [M+H]+, calc. (C9H11Br2O)+: 292.91712
[M+H]+.
Synthesis and Standard Characterization – General Information.
C6H6 was distilled over Na. All commercially available reagents were
purchased from Sigma Aldrich and used as received. 2-(1-
Phenylethylidene)malononitrile 4 was synthesized according to literature
procedure.[6e] All light-sensitive compounds, reactions and manipulations
were shielded from light by either conducting the procedures in a dimly lit
room or by masking the glassware and equipment with aluminum foil.
Purification of products carried out by flash column chromatography was
conducted using Davisil® LC60A (60 Å pore size; 43-60 µm particle size)
silica. For dry column vacuum chromatography, SilicaFlash G60 (15-40
µm particle size, 400-800 mesh) silica was employed. Nuclear magnetic
resonance (NMR) spectra were acquired on either a Bruker Ultrashield
Plus 500 (1H at 500 MHz, 13C at 126 MHz) equipped with a non-inverse
cryo-probe or a 500 MHz Varian spectrometer (1H at 500 MHz) equipped
with a penta-probe/broad-band probe. Chemical shift values are provided
relative to residual solvent references for 1H and 13C-NMR spectra.
Infrared spectroscopy (IR) data were acquired on a Bruker Alpha FT-IR
spectrometer, equipped with ALPHA‘s Platinum ATR single reflection
diamond ATR module. Samples were loaded directly or by evaporation
2,7-Dimethylcyclohepta-2,4,6-trien-1-one (6). Method using substrate
8; modified literature procedure.[13] To a solution of 8 (3.00 g, 14.6 mmol)
in dry C6H6 (294 mL) under N2 atmosphere was added AIBN (0.241 g,
1.47 mmol) in one portion followed by tributyltin hydride (9.89 mL, 36.7
mmol) via slow addition. The temperature was slowly raised to 80 °C and
the reaction mixture stirred for 14 h. The reaction mixture was then
cooled to rt and the solvent removed in vacuo. The residue was purified
by flash column chromatography (SiO2, 4% EtOAc/heptane) to yield title
compound 6 as a pale yellow oil (0.41 g, 21%). TLC Rf= 0.40 (20%
from
a suitable solvent. IR absorptions are reported in units of
wavenumbers (cm-1); intensities are reported as s = strong; m = medium;
w = weak. Mass spectra were recorded on a Bruker Solarix ESI-MALDI-
FT-ICR instrument equipped with a 7 T magnet (prior to the experiments,
the instrument was calibrated using sodium trifluoroacetate (NaTFA)
cluster ions). Thin layer chromatography (TLC) was conducted using
commercially available, precoated plates (silica 60) with fluorescence
indicator. With respect to the DHA compounds, TLC was carried out in
the absence of light; a color change from yellow to red upon exposure to
UV light indicates a conversion to VHF. All melting points are uncorrected.
1
EtOAc/heptane). H NMR (500 MHz, CDCl3) δ 7.37 – 7.32 (m, 2H), 6.91
– 6.85 (m, 2H), 2.32 (d, J = 0.8 Hz, 6H) ppm. 13C NMR (126 MHz, CDCl3)
δ 186.66, 149.60, 134.84, 131.93, 23.76 ppm. HRMS (ESI, m/z) found:
157.06246 [M+Na]+, calc. (C9H110ONa)+: 157.06239 [M+Na]+. Method
using substrate 10. A solution of 10 (2.00 g, 6.80 mmol), AIBN (0.112 g,
0.682 mmol) and tert-butyl(2,6-dimethoxy-1-methylcyclohexa-2,5-dien-1-
yl)dimethylsilane (2.37 g, 8.83 mmol) in heptane (75 mL) was heated to
reflux and stirred overnight (14 h) under an N2 atmosphere. The mixture
was then cooled to rt and stirred for another 24 h. The reaction mixture
was then concentrated in vacuo and the residue purified by flash column
chromatography (SiO2, 10% EtOAc/heptane) to yield title compound 6 as
a pale yellow oil (0.51 g, 56%).
2-Phenylazulene-1,1(8aH)-dicarbonitrile (1a). To a solution of 2-(1-
phenylethylidene)malononitrile 4 (218 mg, 1.30 mmol) in acetic anhydride
(8 mL) was added tropone 3 (125 mg, 1.18 mmol) under an N2
atmosphere, and the mixture was heated to reflux overnight (16 h). The
reaction mixture was then cooled to rt and diluted with toluene (50 mL) to
assist acetic anhydride evaporation as an azeotrope under reduced
pressure. The residue was purified by flash column chromatography
(SiO2, toluene) to yield title product 1a as a crystalline yellow solid (105
mg, 35%). Characterization data are consistent with those reported in
literature.[6e]
5,7-Dimethyl-2-phenylazulene-1,1(8aH)-dicarbonitrile
(11a).
A
6-(Dichloromethyl)-2,6-dimethylcyclohexa-2,4-dien-1-one
(8).
solution of dimethyltropone (66 mg, 0.49 mmol), 2-(1-
6
Modified literature procedure.[13] To a solution of 2,6-dimethylphenol 7
(20.0 g, 164 mmol) in CHCl3 (44.0 mL, 549 mmol) was added a 25%
aqueous solution of cetyltrimethylammonium chloride (0.541 mL, 0.423
mmol), and the reaction mixture was heated to 50 °C in an ambient
atmosphere. To this, a solution of NaOH (50.4 g, 1.26 mol) in H2O (115
mL) was added over 1 h under mechanical stirring and the mixture stirred
at this temperature for a further 4 h. The reaction mixture was then
diluted with H2O (250 mL) and extracted with Et2O (3 x 200 mL). The
organic fractions were combined and washed with H2O (2 x 200 mL) and
brine (2 x 200 mL), dried over MgSO4, filtered, and the solvent was
removed in vacuo. The residue was purified by flash column
chromatography (SiO2, 4% EtOAc/heptane) to yield title compound 8 as
a pale yellow oil (6.88 g, 20%). TLC Rf= 0.33 (20% EtOAc/heptane). 1H
NMR (500 MHz, CDCl3) δ 6.90 – 6.87 (m, 1H), 6.58 – 6.54 (m, 1H), 6.40
phenylethylidene)malononitrile 4 (126 mg, 0.749 mmol) and DBU (7 µL,
0.047 mmol) in Ac2O (4 mL) was heated to reflux for 4 h under an N2
atmosphere. The reaction mixture was then cooled to rt and diluted with
toluene (20 mL) to assist acetic anhydride evaporation as an azeotrope
under reduced pressure. The residue was dissolved in CH2Cl2 (20 mL),
washed with 2M NaOH(aq) (15 mL), dried over MgSO4, filtered, and the
solvent removed in vacuo. The residue was purified by flash column
chromatography (SiO2, 20% heptane/toluene) to yield the title compound
11a as a yellow film (73 mg, 52%). TLC Rf = 0.34 (40% heptane/toluene).
M.p. 51-53 °C. 1H NMR (500 MHz, CDCl3) δ 7.74 – 7.71 (m, 2H), 7.49 –
7.45 (m, 2H), 7.44 – 7.40 (m, 1H), 6.82 (s, 1H), 6.20 (br s, 1H), 6.12 (br s,
1H), 5.52 (br d, J = 3.3 Hz, 1H), 3.71 (ddd, J = 3.3, 1.9, 1.7 Hz, 1H), 2.06
(d, J = 1.0 Hz, 3H), 1.93 (dd, J = 1.6, 1.5 Hz, 3H) ppm. 13C NMR (126
MHz, CDCl3) δ 140.44, 139.44, 139.18, 135.66, 132.02, 131.59, 130.76,
130.05, 129.36, 126.41, 123.50, 115.54, 114.71, 113.10, 50.50, 45.13,
(dd, J = 9.7, 6.1 Hz, 1H), 6.09 (s, 1H), 1.91 (s, 3H), 1.30 (s, 3H) ppm. 13
C
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201700446
European Journal of Organic Chemistry
FULL PAPER
24.74, 22.90 ppm. IR (thin film): 3064w, 3030w, 2978m, 2917m, 2855m,
2246w, 2207w, 1726w, 1674m, 1637w, 1570w, 1496s, 1446s, 1438s,
1406m, 1379m, 1359m, 1343m, 1302m, 1276m, 1245m, 1213m cm-1.
HRMS (ESI, m/z) found: 285.13961 [M+H]+, calc. (C20H17N2)+: 285.13862
[M+H]+.
8922-8932; g) S. L. Broman, M. Jevric, M. B. Nielsen, Chem. Eur. J.
2013, 19, 9542-9548; h) M. Cacciarini, S. L. Broman, M. B. Nielsen,
ARKIVOC 2014, i, 249-263; i) S. L. Broman, M. B. Nielsen, Phys. Chem.
Chem. Phys. 2014, 16, 21172-21182; j) M. D. Kilde, S. L. Broman, A.
Kadziola, M. B. Nielsen, Synlett 2016, 27, 450-454.
[5]
[6]
a) A. Hasenhündl, K. M. Rapp, J. Daub, Chem. Lett. 1979, 597-600; b)
W. Bauer, I. Betz, J. Daub, L. Jakob, W. Pickl, K. M. Rapp, Chem. Ber.
1983, 116, 1154-1173; c) J. Daub, S. Gierisch, T. Knöchel, E. Salbeck,
Z. Naturforsch. 1986, 41b, 1151-1160.
For synthesis optimizations, switching studies, NMR spectra, and
calculational studies, see SI.
a) T. Watanabe, N. Soma, Chem. Pharm. Bull. 1970, 18, 1595–62; b) I.
D. Reingold, H. A. Trujillo, B. E. Kahr, J. Org. Chem. 1986, 51, 1627-
1629; c) S. Gierisch, J. Daub, Chem. Ber. 1989, 122, 69-75; d) L.
Gobbi, P. Seiler, F. Diederich, Helv. Chim. Acta. 2001, 84, 743–776; e)
S. L. Broman, S. L. Brand, C. R. Parker, M. Å. Petersen, C. G. Tortzen,
A. Kadziola, K. Kilså, M. B. Nielsen, ARKIVOC 2011, ix, 51-67.
A. U. Petersen, M. Jevric, J. Elm, S. T. Olsen, C. G. Tortzen, A.
Kadziola, K. V. Mikkelsen, M. B. Nielsen, Org. Biomol. Chem. 2016, 14,
2403-2412
Acknowledgements
University of Copenhagen is acknowledged for financial support.
We thank Prof. Henrik G. Kjaergaard (University of
Copenhagen) for helpful discussions.
[7]
[8]
Keywords: condensation • electrocyclic reactions • olefination •
a) M. Jevric, S. L. Broman, and M. B. Nielsen J. Org. Chem. 2013, 78,
4348-4356; b) A. Vlasceanu, C. L. Andersen, C. R. Parker, O.
Hammerich, T. J. Morsing, M. Jevric, S. L. Broman, A. Kadziola, M. B.
Nielsen Chem. Eur. J. 2016, 22, 7514-7523. c) A. B. Skov, J. F.
Petersen, J. Elm, B. N. Frandsen, M. Santella, M. D. Kilde, H. G.
Kjaergaard, K. V. Mikkelsen, M. B. Nielsen ChemPhotoChem, 2017, In
press. DOI: 10.1002/cptc.201600046
radical reactions • tropone
[1]
[2]
a) A. J. Kronemeijer, H. B. Akkerman, T. Kudernac, B. J. van Wees, B.
L. Feringa, P. W. Blom, B. de Boer, Adv. Mater. 2008, 20, 1467-1473;
b) E. Marchi, M. Baroncini, G. Bergamini, J. Van Heyst, F. Vögtle, P.
Ceroni, J. Am. Chem. Soc. 2012, 134, 15277-15280; c) M. A. Kienzler,
A. Reiner, E. Trautman, S. Yoo, D. Trauner, E. Y. Isacoff, J. Am. Chem.
Soc. 2013, 135, 17683-17686.
[9]
S. L. Broman, O. Kushnir, M. Rosenberg, A. Kadziola, J. Daub, M. B.
Nielsen Eur. J. Org. Chem. 2015, 4119-4130.
[10] K. T. de Oliveira, B. M. Servilha, L. de C. Alves, A. L. Desidera, T. J.
Brocksom, The Synthesis of Seven-Membered Rings in Natural
Products in Studies in Natural Products Chemistry, Vol. 42, Ch. 14 (Ed.:
Atta-ur-Rahman), Elsevier, 2014, pp. 421-463.
a) M.-M. Russew, S. Hecht, Adv. Mater. 2010, 22, 3348–3360; b)
Molecular Switches, Vol. 1 and 2 (Eds.: B. L. Feringa, W. R. Browne),
Wiley-VCH: Weinheim, Germany, 2011; c) W. Szymanski, J. M. Beierle,
H. A. V. Kistemaker, W. A. Velema, B. L. Feringa, Chem. Rev. 2013,
113, 6114−6178; d) R. Göstl, A. Senf, S. Hecht, Chem. Soc. Rev. 2014,
43, 1982-1996; e) W. A. Velema, W. Szymanski, B. L. Feringa, J. Am.
Chem. Soc. 2014, 136, 2178-2191; f) A. Lennartson, A. Roffey, K.
Moth-Poulsen Tetrahedron Lett. 2015, 56, 1457-1465.
[11] I. D. Reingold, L. J. DiNardo, J. Org. Chem. 1982, 47, 3544–3545.
[12] F. M. Abdelrazek, N. H. Metwally, N. A. Kassab, N. A. Sobhy, P. Metz,
A. Jaeger, J. Het. Chem. 2010, 47, 384-388.
[13] M. Barbier, D. H. R. Barton, M. Devys, R. S. Topgi, Tetrahedron 1987,
43, 5031-5038.
[3]
a) Daub, J.; Knöchel, T.; Mannschreck, A. Angew. Chem. Int. Ed. 1984,
23, 960; b) J. Daub, J. Salbeck, T. Knöchel, C. Fischer, H. Kunkely, K.
M. Rapp, Ang. Chem. Int. Ed. Engl. 1989, 28, 1494-1496; c) J. Daub, S.
Gierisch, J. Salbeck, Tetrahedron Lett. 1990, 31, 3113-3116; d) J.
Achatz, C. Fischer, J. Salbeck, J. Daub, J. Chem. Soc., Chem.
Commun. 1991, 504-507; e) Spreitzer, J. Daub, Liebigs Ann. 1995,
1637-1641; f) L. Gobbi, P. Seiler, F. Diederich, Helv. Chim. Acta 2001,
84, 743-777.
[14] A. Studer, S. Amrein, F. Schleth, T. Schulte, J. C. Walton J. Am. Chem.
Soc. 2003, 125, 5726-5733.
[15] S. T. Olsen, J. Elm, F. E. Storm, A. N. Gejl, A. S. Hansen, M. H.
Hansen, J. R. Nikolajsen, M. B. Nielsen, H. G. Kjaergaard, K. V.
Mikkelsen J. Phys. Chem. A 2015, 119, 896-904.
[16] a) V. De Waele, U. Schmidhammer, T. Mrozek, J. Daub, E. Riedle J.
Am. Chem. Soc. 2002, 124, 2438-2439; b) U. Schmidhammer, V. De
Waele, G. Buntinx, E. Riedle From ultrafast spectroscopy to
bidirectional molecular switches: DHA/VHF in Ultrafast Phenomena XIV
2005, vol. 79, pp. 465-467; c) O. Schalk, S. L. Broman, M. Å. Petersen,
D. V. Khakhulin, R. Y. Brogaard, M. B. Nielsen, A. E. Boguslavskiy, A.
Stolow, T. I. Sølling J. Phys. Chem. A 2013, 117, 3340-3347.
[17] a) L. Vereecken, J. Peeters, J. Chem. Phys. 2003, 119, 5159-5170; b)
K. H. Møller, R. V. Otkjær, N. Hyttinen, T. Kurtén, H. G. Kjaergaard, J.
Phys. Chem. A 2016, 120, 10072-10087.
[4]
a) M. Å. Petersen, L. Zhu, S. H. Jensen, A. S. Andersson, A. Kadziola,
K. Kilså, M. B. Nielsen, Adv. Funct. Mater. 2007, 17, 797-804; b) M. Å.
Petersen, A. S. Andersson, K. Kilså, M. B. Nielsen, Eur. J. Org. Chem.
2009, 1855-1858; c) S. L. Broman, M. Å. Petersen, C. G. Tortzen A.
Kadziola, K. Kilså, M. B. Nielsen, J. Am. Chem. Soc. 2010, 132, 9165-
9174; d) M. Å. Petersen, S. L. Broman, K. Kilså, A. Kadziola, M. B.
Nielsen, Eur. J. Org. Chem. 2011, 1033-1039; e) L. Skov, M. Å.
Petersen, S. L. Broman, A. D. Bond, M. B. Nielsen, Org. Biomol. Chem.
2011, 9, 6498-6501; f) M. Santella, V. Mazzanti, M. Jevric, C. R. Parker,
S. L. Broman, A. D. Bond, M. B. Nielsen, J. Org. Chem. 2012, 77,
This article is protected by copyright. All rights reserved.

10.1002/ejoc.201700446
European Journal of Organic Chemistry
FULL PAPER
Entry for the Table of Contents
FULL PAPER
We here present a novel
Molecular Switches
NC
regioselective synthesis of a dimethyl-
substituted dihydroazulene (DHA)
photoswitch by employing a
CN
Ph
NC
CN
Ph
Nickie C. M. Lubrin, Alexandru
Vlasceanu, Benjamin N. Frandsen,
Anders B. Skov, Martin Drøhse Kilde,
Kurt V. Mikkelsen, Mogens Brøndsted
Nielsen*
+
O
DHA
functionalized tropone as substrate.
Photoisomerization results in isomeric
vinylheptafulvenes (VHFs), which
undergo electrocyclic ring closures
significantly faster than the
unsubstituted derivative, revealing the
influence of electron-donating methyl
substituents.
influence of
alkyl groups
fast
light
CN
Page No. – Page No.
NC
NC
fast
CN
Ph
Ph
Dialkylated Dihydroazulene /
Vinylheptafulvene Derivatives –
Synthesis and Switching Properties
light
DHA
E/Z-VHF
This article is protected by copyright. All rights reserved.