A strain induced change of mechanism from a [2 + 2 + 2] to a [2 + 1 + 2 + 1] cycloaddition reaction

Article abstract of DOI:10.1021/ol5007604

While investigating the [2 + 2 + 2] cycloaddition as a tool to build up strained oligophenyl systems with a diyne-ethylene glycol macrocyle, a surprising change of mechanism was observed. Instead of the expected [2 + 2 + 2] para-terphenyl, the ortho-terphenyl product explained by a formal [2 + 1 + 2 + 1] cycloaddition was formed. An η⁴-coordinated metal-cyclobutadiene is proposed as the key structure in the catalytic cycle, which is formed to release the induced strain. The optical properties of the ortho-terphenyl products have been measured as well as the coordination ability of Na⁺ and K⁺.

Full text of DOI:10.1021/ol5007604

Letter
pubs.acs.org/OrgLett
Terms of Use
A Strain Induced Change of Mechanism from a [2 + 2 + 2] to a
[2 + 1 + 2 + 1] Cycloaddition Reaction
Anne-Florence Tran-Van,† Silas Gotz,‡ Markus Neuburger,† and Hermann A. Wegner*,‡
̈
†Department of Chemistry, Basel University, St Johanns-Ring 19, CH-4056 Basel, Switzerland
‡Institut fur Organische Chemie, Justus-Liebig Universitat, Heinrich-Buff-Ring 62, D-35392 Giessen, Deutschland
̈
̈
S
* Supporting Information
ABSTRACT: While investigating the [2 + 2 + 2] cyclo-
addition as a tool to build up strained oligophenyl systems
with a diyne-ethylene glycol macrocyle, a surprising change of
mechanism was observed. Instead of the expected [2 + 2 + 2]
para-terphenyl, the ortho-terphenyl product explained by a
formal [2 + 1 + 2 + 1] cycloaddition was formed. An η4-
coordinated metal-cyclobutadiene is proposed as the key
structure in the catalytic cycle, which is formed to release the
induced strain. The optical properties of the ortho-terphenyl
products have been measured as well as the coordination ability of Na+ and K+.
he [2 + 2 + 2] cycloaddition has been established as one
1
Tof the best methods to access substituted benzene rings.
The high efficiency, combined with a broad tolerance of
functional groups as well as the perfect atom economy,
characterizes this useful process. Therefore, this reaction has
been applied in a number of syntheses of complex molecules
and natural products.2 In this reaction three alkynes are
combined to an aromatic system using a metal catalyst such as
Rh, Co, Ru, or Ir. The mechanism has been studied extensively
for different metal catalysts.3 Driving force for the [2 + 2 + 2]
cycloaddition is the large amount of energy gained from the
aromaticity. Hence, this transformation should be ideal to
overcome energetic barriers introduced in the cycloaddition
product, e.g., by strain.
With this reasoning we envisioned the synthesis of strained
oligophenylenes4 using the [2 + 2 + 2] cycloaddition as the
method of choice, which might serve as a model system for the
assembly of cycloparaphenylenes.5,6 King and co-workers
showed the feasibility of introducing strain via the [2 + 2 +
Figure 1. Design of a macrocyclic precursor for the synthesis of
strained oligophenylenes.
2] cycloaddition in the synthesis of a highly strained
quadranulene.2c
In order to bend the oligophenylene, a suitable tether was
attached in the terminal para-positions (Figure 1). Thus, the
strain introduced could be conveniently adjusted by the tether
length. Because the preparation of such macrocyclic precursors
is notoriously difficult, both tethers had to be carefully chosen.
The two alkynes were joined by a CH2C(CH2OCH3)2CH2
fragment. The quaternary carbon center will force the two
alkyne moieties into closer proximity due to a Thorpe−Ingold
effect. The second tether between the two phenyl rings consists
of an oligoethylene glycol chain. The O atoms in the tether
increase the flexibility dramatically compared to their carbon
analogue. Additionally, the possibility to add a templating metal
further increased its attractiveness.
The synthesis commences with the Sonogashira coupling of
bisalkyne 1 with the iodobenzene derivative 2 to yield the
desired bisbenzylalcohol 3 (Scheme 1). The strategy chosen for
the macrocyclization relied on an ether formation by a
substitution reaction between ditosylated triethylene glycol 4
and the benzylic alcohols of the dialkyne building block 3 using
NaH as the base.7 By taking advantage of the template effect of
the sodium cation, the desired macrocycle could be isolated in
33% yield. A crystal structure of the macrocycle is shown in
Received: March 14, 2014
Published: April 23, 2014
© 2014 American Chemical Society
2410
dx.doi.org/10.1021/ol5007604 | Org. Lett. 2014, 16, 2410−2413

Organic Letters
Letter
Scheme 1. Synthesis of the Macrocyclic Precursor 5
Table 1. [2 + 1 + 2 + 1] Cycloaddition on Strained Precursor
alkyne
Figure 2. The attempt to close the ring via alkene metathesis
was not successful. Also, changing the order in the reaction
sequence, building up first the bisaryl ethylene glycol chain and
then closing the macrocycle via Sonogashira coupling, did not
lead to the desired product.
entry
R1
R2
product
yield
1
2
Et
Et
6a
6b
6c
6d
6e
6f
88%
64%
59%
67%
62%
64%
43%
64%
−
Ph
Ph
3
CH2OH
Ph
CH2OH
H
4
5
C(CH3)2(OH)
CH2CH2CH3
CH2CH2OH
Me
H
6
H
7
H
6g
6h
6i
8
CH2OH
COOMe
TMS
9
COOMe
TMS
10
6j
−
also investigated, but no reaction was observed with any of the
other catalytic systems screened.
To explain the formation of the unexpected product, we
propose the mechanism shown in Figure 3. The first step
involves ligand dissociation and coordination of the two alkyne
moieties of the macrocycle 5 to rhodium. Next, the well
accepted rhodacyclopentadiene B is formed. However, already
at this stage intermediate B experiences the strain induced by
the tether. Therefore, instead of the expected coordination
compound G followed by insertion of the alkyne to the
rhodacycloheptatriene H, the η4 complex C is proposed, which
is formed via reductive elimination.
Figure 2. Crystal structure of the macrocyclic precursor 5 (left) and
the [2 + 1 + 2 + 1] product 6a (ORTEP drawing, hydrogens, solvent,
and disorder are omitted for clarity).
With precursor 5 in hand the [2 + 2 + 2] cycloaddition
reaction was tested. The procedure reported in the literature8
was modified using microwave irradiation to reduce reaction
times. Without microwave irradiation, the reaction required 40
h and the addition of a second batch of catalyst to achieve
completion. Under microwave irradiation, reaction times could
be reduced to 6 h and the catalyst loading down to 10 mol %.
Surprisingly, when 3-hexyne was used as a reaction partner,
an ortho-terphenylene was observed instead of the expected
para-terphenylene product. The structure of the product was
unambiguously proven by NMR experiments as well as by X-
ray analysis (Figure 2). To test the generality of this unexpected
outcome, the reaction was performed with various monoynes
(Table 1). In all cases tested the ortho-terphenyl product was
obtained in moderate to good yield. Besides aromatic and alkyl
substituents, free hydroxyl groups were also well tolerated.
Only ester and TMS substituted alkynes did not give the
desired product. Instead the trimerization of these alkynes was
observed. In the case of dimethyl acetylenedicarboxylate no
reaction was observed (Table 1, entry 9). Unsymmetric
substituted alkynes preferentially gave one regioisomer with
the larger R-group next to the phenyl.
This deviation from the standard [2 + 2 + 2] catalytic cycle
allows reduction of the strain induced by the tether
compensating for the ring strain of the cyclobutadiene, which
is additionally stabilized by the coordination with the metal.
The η4 complex can be represented by two resonance
structures C and D. Insertion of the metal delivers the less
strained complex E which then follows the classical mechanism.
Coordination of the monoalkyne, followed by its insertion,
formed metallacycloheptatriene F. Finally, reductive elimina-
tion leads to the product 6. Hence, the strain induces a change of
the mechanism from the [2 + 2 + 2] mechanism to a formal [2 + 1
+ 2 + 1] cycloaddition. The curvature of the para-terphenyl
product I corresponds roughly to an [8]CPP, for which the ring
strain was calculated to ∼70 kcal/mol,6 translating into a strain
energy of ∼26 kcal/mol for I. Similar reactivity has been
observed as a side reaction by Tanaka and co-workers during
the synthesis of a helicene.9 The usual cycloaddition
mechanism does not involve cyclobutadiene intermediates C
and D and is considered to be nonprogressive for the catalytic
cycle. Numerous cyclobutadiene complexes of cobalt and
rhodium have already been isolated.10 If the strain was the
cause for the observed change in mechanism, increasing the
tether length should influence this parameter and restore the [2
+ 2 + 2] reactivity. Indeed, increasing the number of ethylene
glycol units from three to six lead to the isolation of the [2 + 2
+ 2] cycloadduct as the sole product allowing a straight
Different solvent systems and concentrations were tested in
order to provoke the formation of the para-terphenylene [2 + 2
+ 2] cycloaddition product initially expected. However, the
ortho-terphenyl product was always favored. Traditional heating
led to the same results. The influence of different catalysts
(such as Pd(PPh3)4, Grubbs I, NiBr2(dppe)/Zn, CoI2/Zn) was
2411
dx.doi.org/10.1021/ol5007604 | Org. Lett. 2014, 16, 2410−2413

Organic Letters
Letter
Figure 3. Proposed mechanism for the [2 + 1 + 2 + 1] cycloaddition reaction of strained system (left) and the standard [2 + 2 + 2] cycloaddition
reaction (right) (ligands are omitted for clarity).
arrangement of the para-terphenyl moiety. Not surprisingly, if
no tether was present, only the [2 + 2 + 2] product was
observed.
Noncovalent cation−π interactions are playing an important
role in stabilizing biological and chemical assemblies.12
Therefore, polyaromatic systems combined with a crown
ether cavity could function as electrochemical sensors for
sodium or potassium cations. Such molecules hold potential
application for functional extractants or conductive electrolytes.
As the [2 + 1 + 2 + 1] cycloadducts 6 ideally suit such criteria
the ability to bind cations such as Na+ and K+ was investigated.
The binding of the Na/K cation to the macrocycle was
Photooptical properties were measured for the macrocyclic
molecules. The absorption maximum was observed at 245 nm,
and emission occurred around 350−375 nm. Unsubstituted
linear ortho-terphenyl in contrast absorbs at 260 nm and
presents emission maxima at 450 and 478 nm.11 The tethered
macrocyclic ortho-terphenyls have smaller Stokes shifts.
Notably, the diphenyl substituted molecule showed an even
shorter Stokes shift with absorption at 250 nm and emission at
350 nm (Figure 4).
1
monitored by H NMR spectroscopy in deuterated acetone
(0.02 M) with the incremental addition of a solution of NaPF6
and KPF6 respectively (0.008 M in acetone d6) analogously to
the procedure reported by Rathore and co-workers.7 The
addition of potassium resulted in a dramatic shift of the signals
of both the aromatic and the ethylene glycol protons (by up to
0.1 ppm) indicating a strong coordination of the potassium
cation to both the crown ether part of the molecule and the
phenyl rings. The addition of sodium cations led, however, only
to a shifting of the ethylene glycol protons signals (0.05 ppm),
while the aromatic signals remained unchanged (see SI). The
sodium cation is smaller than the potassium cation and can be
fully accommodated by the triethylene glycol crown ether part
of the molecule further away from the phenyl moieties, as can
be seen in modeled structures (Figure 5). Similar to Rathore, it
was not possible to calculate binding constants based on the
NMR-titration, as the values were too high.7
In conclusion, an unexpected change of reactivity was
observed in the [2 + 2 + 2] cycloaddition when a strained
precursor was subjected to the Rh-catalyzed reaction
conditions. Instead of observing [2 + 2 + 2] cycloaddition, a
[2 + 1 + 2 + 1] cycloaddition took place leading to an ortho-
terphenylene macrocyclic product. This outcome offers new
Figure 4. UV−vis absorption (solid line) and fluorescence spectra
(dashed lines) of [2 + 1 + 2 + 1] cycloaddition products (normalized
spectra, measured in chloroform).
2412
dx.doi.org/10.1021/ol5007604 | Org. Lett. 2014, 16, 2410−2413

Organic Letters
Letter
(3) For two recent studies, see: (a) Sakai, S.; Udagawa, T.; Kato, S.;
Nakada, K. J. Phys. Org. Chem. 2013, 26, 517. (b) Guo, C.-H.; Wu, H.-
S.; Hapke, M.; Jiao, H. J. Organomet. Chem. 2013, 748, 29−35. For a
study on Rh-catalyzed alkyne trimerization, see: (c) Orian, L.; van
Stralen, J. N. P.; Bickelhaupt, F. M. Organometallics 2007, 26, 3816.
(4) (a) Bodwell, G. J.; Fleming, J. J.; Miller, D. O. Tetrahedron 2001,
57, 3577. (b) Bodwell, G. J.; Miller, D. O.; Vermeij, R. J. Org. Lett.
2001, 3, 2093.
(5) For the first synthesis of a cycloparaphenylene (CPP), see:
(a) Jasti, R.; Bhattacharjee, J.; Neaton, J. B.; Bertozzi, C. R. J. Am.
Chem. Soc. 2008, 130, 17646. For a recent review, see: (b) Omachi,
H.; Segawa, Y.; Itami, K. Acc. Chem. Res. 2012, 45, 1378. (c) Yamago,
S.; Kayahara, E.; Iwamoto, T. Chem. Rec. 2014, 14, 84. (d) Hirst, E. S.;
Jasti, R. J. Org. Chem. 2012, 77, 10473. For the smallest CPP
synthesized to date, see: (e) Kayahara, E.; Patel, V. K.; Yamago, S. J.
Am. Chem. Soc. 2014, 136, 2284.
(6) Tran-Van, A. F.; Huxol, E.; Basler, J. M.; Neuburger, M.; Ad-
jizian, J.-J.; Ewels, C. P.; Wegner, H. A. Org. Lett. 2014, 16, 1594.
(7) Shukla, R.; Lindeman, S. V.; Rathore, R. Chem. Commun. 2009,
5600.
(8) (a) McDonald, F. E.; Smolentsev, V. Org. Lett. 2002, 4, 745.
(b) Grigg, R.; Scott, R.; Stevenson, P. J. Chem. Soc., Perkin Trans. 1
1988, 1357.
(9) (a) Tanaka, K.; Kamisawa, A.; Suda, T.; Noguchi, K.; Hirano, M.
J. Am. Chem. Soc. 2007, 129, 12078. (b) For a [2 + 1 + 2 + 1]
cycloaddition catalyzed by Co, see: Gleiter, R.; Kratz, D. Angew.
Chem., Int. Ed. 1990, 29, 276.
Figure 5. Space-filling model (MM2 optimization) of the complex
between the macrocycle 6 and potassium (left) or sodium (right).
insight into the reaction mechanism involved in the [2 + 2 + 2]
cycloadditions and increases the applicability and predictability
of this useful synthetic tool. Finally, the optical properties and
the complexation with alkali metals were studied identifying
different binding modes for K+ and Na+, rationalized by the
different sizes.
ASSOCIATED CONTENT
* Supporting Information
■
S
Syntheses and characterization of all new compounds,
crystallographic data of 5 and 6a. This material is available
free of charge via the Internet at http://pubs.acs.org.
(10) For selected examples of Rh-cyclobutadiene complexes, see:
́
(a) Lamata, M. P.; San Jose, E.; Carmona, D.; Lahoz, F. J.; Atencio, R.;
AUTHOR INFORMATION
Corresponding Author
Oro, L. A. Organometallics 1996, 15, 4852. (b) Rausch, M. D.;
Gardner, S. A.; Tokas, E. F.; Bernal, I.; Reisner, G. M.; Clearfield, A. J.
Chem. Soc., Chem. Commun. 1978, 187. (c) Nixon, J. F.; Kooti, M. J.
Organomet. Chem. 1976, 104, 231. For an example of a Co-
cyclobutadiene complex, see: (d) Werz, D. B.; Schulte, J. H.; Gleiter,
R.; Rominger, F. J. Organomet. Chem. 2004, 689, 3132.
■
*E- mail: Hermann.A.Wegner@org.Chemie.uni-giessen.de.
Notes
The authors declare no competing financial interest.
(11) (a) Ozasa, S.; Fujioka, Y.; Okada, M.; Izumi, H.; Ibuki, E. Chem.
Pharm. Bull. 1981, 29, 370. (b) Garcia Alvarez-Coque, M. C.; Ramis
Ramos, G.; O’Reilly, A. M.; Winefordner, J. D. Anal. Chim. Acta 1988,
204, 247.
(12) (a) Dougherty, D. A. Acc. Chem. Res. 2013, 46, 885.
(b) Mahadevi, A. S.; Sastry, G. N. Chem. Rev. 2013, 113, 2100.
ACKNOWLEDGMENTS
Dedicated to Prof. Dr. Armin de Meijere, Georg-August-
■
Universitat, Gottingen, on the occasion of his 75th birthday.
̈
̈
This work was financially supported by the university of Basel,
Staatssekreteriat fur Bildung and Forschung SBF and the
̈
COST action MP0901 NanoTP.
REFERENCES
■
(1) (a) Luo, Y.; Pan, X.; Yu, X.; Wu, J. Chem. Soc. Rev. 2014, 43, 834.
(b) Broere, D. L. J.; Ruijter, E. Synthesis 2012, 44, 2639. (c) Shibata,
́
Y.; Tanaka, K. Synthesis 2012, 44, 323. (d) Domínguez, G.; Perez-
Castells, J. Chem. Soc. Rev. 2011, 40, 3430. (e) Chopade, P. R.; Louie,
J. Adv. Synth. Catal. 2006, 348, 2307. (f) Saito, S.; Yamamoto, Y. Chem.
Rev. 2000, 100, 2901. (g) Lautens, M.; Klute, W.; Tam, W. Chem. Rev.
1996, 96, 49. (h) Schore, N. E. Chem. Rev. 1988, 88, 1081.
(2) (a) Murayama, K.; Sawada, Y.; Noguchi, K.; Tanaka, K. J. Org.
Chem. 2013, 78, 6202. (b) Sugiyama, Y.-K.; Kato, R.; Sakurada, T.;
Okamoto, S. J. Am. Chem. Soc. 2011, 133, 9712. (c) Bharat; Bhola, R.;
Bally, T.; Valente, A.; Cyranski, M. K.; Dobrzycki, Ł.; Spain, S. M.;
́
Rempała, P.; Chin, M. R.; King, B. T. Angew. Chem., Int. Ed. 2010, 49,
399. (d) Dachs, A.; Torrent, A.; Roglans, A.; Parella, T.; Osuna, S.;
Sola,
̀
M. Chem.Eur. J. 2009, 15, 5289. (e) Wu, Y.-T.; Hayama, T.;
Baldridge, K. K.; Linden, A.; Siegel, J. S. J. Am. Chem. Soc. 2006, 128,
6870. (f) Bonaga, L. V. R.; Zhang, H.-C.; Moretto, A. F.; Ye, H.;
̃
Gauthier, D. A.; Li, J.; Leo, G. C.; Maryanoff, B. E. J. Am. Chem. Soc.
2005, 127, 3473. (g) Shibata, T.; Fujimoto, T.; Yokota, K.; Takagi, K.
J. Am. Chem. Soc. 2004, 126, 8382. (h) Dosa, P. I.; Whitener, G. D.;
Vollhardt, K. P. C.; Bond, A. D.; Teat, S. J. Org. Lett. 2002, 4, 2075.
(i) Eichberg, M. J.; Dorta, R. L.; Grotjahn, D. B.; Lamottke, K.;
Schmidt, M.; Vollhardt, K. P. C. J. Am. Chem. Soc. 2001, 123, 9324.
(j) Araki, T.; Noguchi, K.; Tanaka, K. Angew. Chem., Int. Ed. 2013, 52,
5617.
2413
dx.doi.org/10.1021/ol5007604 | Org. Lett. 2014, 16, 2410−2413