Shape-Switchable Azo-Macrocycles

Article abstract of DOI:10.1002/ejoc.200801282

The synthesis of four shape-switchable macrocycles comprising different peripheral substituents is described. The macro- cycles 1-4 consist of m-terphenyl semicircles interlinked by two azo joints. These macrocycles were assembled from ni- tro-functionalized m-terphenyl moieties through reductive dimerization. The semicircles were assembled through Suzuki cross-coupling reactions. The molecular weights of the macrocycles were determined by vapour pressure osmome- try, because mass spectrometry failed in the cases of 2 and 3. The E - Z photoisomerization reactions were analysed by UV/Vis spectroscopy complemented by JH NMR studies. A very slow thermal back-reaction indicated considerable stabilization of the Z isomer. The reduced efficiency of the thermal back-reaction probably arises from the reduced degree of freedom due to the mechanical interlinking of the two azo groups. The photostationary state consisted of all-Z (85%) and all-E isomers (15 %). The E - Z transformation induced by irradiation displayed simple exponential kinetics, which indicates pairwise switching of the two azo groups in a macrocycle, at least on the timescale under investigation.

Full text of DOI:10.1002/ejoc.200801282

FULL PAPER
DOI: 10.1002/ejoc.200801282
Shape-Switchable Azo-Macrocycles
Marcel Müri,[a] Klaus C. Schuermann,[b] Luisa De Cola,[b] and Marcel Mayor*[a,c]
Keywords: Azo compounds / Macrocycles / Isomerization / Cross-coupling
The synthesis of four shape-switchable macrocycles compris-
ing different peripheral substituents is described. The macro-
cycles 1–4 consist of m-terphenyl semicircles interlinked by
two azo joints. These macrocycles were assembled from ni-
tro-functionalized m-terphenyl moieties through reductive
very slow thermal back-reaction indicated considerable sta-
bilization of the Z isomer. The reduced efficiency of the ther-
mal back-reaction probably arises from the reduced degree
of freedom due to the mechanical interlinking of the two azo
groups. The photostationary state consisted of all-Z (85%)
dimerization. The semicircles were assembled through Su- and all-E isomers (15%). The E Ǟ Z transformation induced
zuki cross-coupling reactions. The molecular weights of the
macrocycles were determined by vapour pressure osmome-
try, because mass spectrometry failed in the cases of 2 and 3.
The E Ǟ Z photoisomerization reactions were analysed by
UV/Vis spectroscopy complemented by 1H NMR studies. A
by irradiation displayed simple exponential kinetics, which
indicates pairwise switching of the two azo groups in a
macrocycle, at least on the timescale under investigation.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2009)
pendent due to mechanical interlinkage. Macrocyclic model
compounds consisting of two azobenzene subunits inter-
linked by alkyl chains have already been synthesized to in-
vestigate the influence of steric restrictions on the photo-
chemical properties of these photoactive subunits.[2] Re-
cently, Tamaoki and co-workers reported the synthesis of
a (3,3Ј)-azobenzophane displaying inverted thermodynamic
stabilities.[18] Owing to steric strain, the Z isomer of this
azobenzophane was more stable, whereas steric repulsion
usually favours the E isomers of azo derivatives. Thus, the
thermal back-reaction of (3,3Ј)-azobenzophane was from
the E to the Z isomer, and under irradiation the amount of
E isomer was increased.
Synthetically, only a few different strategies have been
reported for the synthesis of azo derivatives. However, ow-
ing to their importance as readily available and inexpensive
dyes in industry, various procedures for their synthesis have
been developed and patented.[19,20] Most of these pro-
cedures are based on so-called azo coupling,[21] which is an
electrophilic substitution of aryldiazonium salts with elec-
tron-rich aromatic compounds.[22,23] However, the moderate
electrophilicity of aryldiazonium cations requires rather nu-
cleophilic arenes, which limits the scope of the synthetic
strategy. Furthermore, symmetrical azo compounds have
been assembled by the oxidation of primary aryl-
amines[24–26] or by the reduction of nitrobenzene deriva-
tives.[27] Both strategies, which alter the oxidation state of
the aryl nitrogen atom, are expected to have a common ni-
troso intermediate that reacts with an arylamine to provide
the targeted azo group. Thus, to form azo derivatives com-
prising two different aryl subunits, a nitroso derivative as
precursor of the azo group is isolated that subsequently re-
Introduction
The E Ǟ Z photoisomerization of azobenzene is a well-
understood photochromic transformation reaction[1–5] and
has already been used in the development of light-driven
functional molecules.[5–7] Chemical substances exhibiting
photoinduced structural changes are not only suitable can-
didates for the storage of light energy, but also for the con-
version of light energy into mechanical motion.[5,8–10] The
photoinduced reversible E Ǟ Z isomerization of azo-
benzene derivatives causes considerable structural
changes.[5,7,11,12] The azobenzene motive is thus an ideal
building block for integrating light-induced conformational
changes into molecular structures and for creating patterns
and nanostructures in polymeric and glass substrates.[13–17]
Bulky substituents or cyclic structures cause steric hin-
drance and thus may influence the isomerization behaviour
of azo-functionalized systems. Of particular interest are
macrocycles comprised of several azo groups as their struc-
tural rearrangement upon isomerization becomes interde-
[a] University of Basel, Department of Chemistry,
St. Johanns-Ring 19, 4056 Basel, Switzerland
Fax: +41-61-267-1016
E-mail: marcel.mayor@unibas.ch
[b] Westfälische Wilhelmsuniversität Münster, Institut of Physics,
Mendelstrasse 7, 48149 Münster, Germany
Fax: +49-251-980-2834
E-mail: decola@uni-muenster.de
[c] Forschungszentrum Karlsruhe GmbH, Institute for Nanotech-
nology,
P. O. Box 3640, 76021 Karlsruhe, Germany
Fax: +49-7247-82-5685
E-mail: marcel.mayor@kit.edu
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
2562
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 2562–2575

Shape-Switchable Azo-Macrocycles
acts with an arylamine in the Mills reaction.[28,29] More re- asymmetric macrocycle 4, the coupling between a nitroso-
cently, an alternative strategy towards asymmetric azo de- and an amino-functionalized semicircle (Mills reaction),
rivatives based on the palladium-catalysed coupling of aryl which would enable two different precursors to be distin-
hydrazides with aryl halides followed by an oxidation was guished, might be more favourable. This consideration was
reported.[30,31]
even more supporting of the general strategy, as the precur-
Herein we report the synthesis and shape-switching be- sors for the Mills reaction should be available by reduction
haviour of macrocycles 1–4 consisting of two rigid m-ter- of the original starting materials, namely the corresponding
phenyl semicircles interlinked by two azo groups. Although nitro-functionalized semicircles. A Suzuki cross-coupling
the central photochemically active macrocycle is common reaction should provide the m-terphenyl semicircles. 3,5-Di-
to all the macrocycles 1–4, they differ considerably in the bromoanisole as the starting material for the Suzuki reac-
substituents on the central ring of their terphenyl subunits. tion will provide a phenolic hydroxy group on the m-ter-
With the variation of these substituents not only are chemi- phenyl semicircle after deprotection, which would enable its
cal properties like solubility, processability or adhesion fea- further functionalization by nucleophilic substitution reac-
tures expected to differ, but also physical properties like tions. Finally, to obtain a suitable 2,3,4-trialkyl-1,5-dibro-
size, shape, or momentum of inertia. Whereas the target mobenzene precursor for the Suzuki reaction, 2,6-diiodo-4-
structures 1–3 are symmetric azobenzene macrocycles con- nitroaniline was considered as a starting material. A Sand-
sisting of two identical semicircles, in the asymmetric mayer reaction will provide 3,4,5-triiodo-1-nitrobenzene
macrocycle 4 the central substituents on both semicircles and subsequent substitution of the iodine atoms by alkynyl
are different.
chains in a Sonogashira–Hagihara reaction should yield the
The shape-switchable macrocycles 1–4 are paving the way required carbon chains. Hydrogenation should reduce both
towards photoswitchable macrocycles, the isomers of which the triple bonds and the nitro group to provide 3,4,5-trialk-
have improved thermodynamic stability features as well as ylaniline, a suitable precursor for introducing bromine
considerably different spatial appearances. However, the atoms at the 2- and 6-positions. Subsequent deamination
combination of efficient photochemical switching with large should give the desired 2,3,4-trialkyl-1,5-dibromobenzene
differences in appearance requires a subtle balance between precursor.
the rigidity and flexibility of the macrocycle subunits. Al-
though a macrocyclic backbone that is too rigid will handi-
cap the isomerization reaction, subunits that are too flexible
will not stabilize the thermodynamically less favoured con-
figuration sufficiently.
Synthesis
The dibromo intermediate 7 was obtained from commer-
cially available 4-hexylaniline. Treatment of 4-hexylaniline
(5) with N-bromosuccinimide (NBS) in DMF gave the dou-
bly brominated product 6 as a dark-red solid in a good yield
of 86%.[32] Deamination with concentrated sulfuric acid
and sodium nitrite at 75 °C provided after 3 h the desired
Results and Discussion
Synthetic Strategy
An overview of the synthetic strategy is shown in dibromo compound 7, which was isolated by column
Scheme 1. All target structures 1–4 are macrocycles con- chromatography (CC) in 55% yield over both steps. Treat-
sisting of two m-terphenyl semicircle subunits and two azo ment of 2.6 equiv. of the commercially available (3-ni-
groups on opposite sides, and thus similar synthetic strate- trophenyl)boronic acid and the dibromoaryl compound 7
gies were considered in all four cases. Owing to the syn- with 8 mol-% of tetrakis(triphenylphosphane)palladium
thetic availability of its precursors, a reductive dimerization [Pd(PPh3)4] in toluene/ethanol (3:1) at 85 °C for 3 h pro-
between nitro-functionalized m-terphenyl semicircles was vided the desired semicircle 8 as a light-green solid in a
favoured as the ring-closing reaction. In the case of the good yield of 95% after CC.[33] To assemble the macrocycle
Eur. J. Org. Chem. 2009, 2562–2575
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2563

M. Müri, K. C. Schuermann, L. De Cola, M. Mayor
FULL PAPER
Scheme 1. Retrosynthetic scheme for the synthesis of symmetric macrocycles. (a) Reductive dimerization, (b) Suzuki cross-coupling reac-
tion, (c) bromination and deamination, (d) hydrogenation, (e) nucleophilic substitution, (f) Sonogshira–Hagihara cross-coupling, (g)
Sandmayer reaction.
1 (Scheme 2), the dinitro semicircle 8 was treated with lith- already been reported by Höger et al.[34] Increased adsorp-
ium aluminium hydride (LiAlH4) in dry THF at room tem- tion properties of these macrocycles with increasing number
perature (r.t.) for 75 min. Interestingly, precipitation indi- of alkyl chains were observed, an interesting aspect in view
cating the formation of insoluble polymers was not ob- of a future scanning probe investigation of switching
served in spite of the bifunctional building blocks. The macrocycles.
crude product was collected by filtration of the precipitate
The synthetic route to the target macrocycle 2 is shown
formed after the addition of water. Subsequent extraction in Scheme 3. The commercially available 1,3,5-tribromob-
by CH2Cl2 and purification by gel permeation chromatog- enzene (13) was converted in an aromatic nucleophilic sub-
raphy (GPC) provided the macrocycle 1 as a bright-orange stitution reaction[35] with 1.4 equiv. of sodium methoxide
solid in 37% yield. With the pure compound in hand, an in DMF at 80 °C. After CC, the desired monosubstituted
alternative purification procedure was found. Dissolution product 14 was isolated in a good yield of 70%. Subsequent
of the crude extract in hot toluene/ethanol (6:5) and collec- Suzuki cross-coupling[36,37] turned out to be challenging.
tion of the oily residue after storage at –20 °C for 10 d gave With [Pd(PPh3)4] as catalyst and K2CO3 as base, 14 was
the macrocycle 1 in 50% yield and in a pure form according treated with 2.6 equiv. of (3-nitrophenyl)boronic acid in tol-
to analytical GPC.
uene/ethanol (8:3) at 85 °C for 150 min. The doubly aryl-
To further clarify the surprising efficiency of the cycliza- ated product 15 was obtained in a moderate yield of 33%
tion reaction, the crude reaction mixture was also investi- after recrystallization from hot toluene. Although shorter
gated by analytical GPC. And indeed, besides the dominant reaction times provided the monocoupled compound as the
main peak corresponding to the macrocycle 1 with a reten- main product, considerable degradation of the starting ma-
tion time of 11.85 min, only a few rather weak peaks at terial was observed for elongated reaction times. After ne-
shorter retention times, probably arising from larger oligo- arly quantitative deprotection to the free alcohol 16 with
mers, and a single small peak at a longer retention time, BBr3, the semicircle 17 was assembled in an SN2 reaction
probably due to the fully reduced diamino semicircle, were with the electrophile 12. The polyalkylated chlorometh-
observed. However, these side-products were not further in- ylbenzene 12 was synthesized according to the literature[38]
vestigated.
by starting from the commercially available methyl 3,4,5-
To increase the molecular weights of the subunits of the trihydroxybenzoate (9) in three steps in an excellent overall
two semicircles moving relative to each other and thus to yield of 85%.[39] The subsequent SN2 reaction[31] with the
increase the spatial difference of the two isomers, the phenol 16 was performed with potassium carbonate in
macrocycles 2 and 3 were envisaged. Shape-persistent DMF at 65 °C. After CC, the semicircle 17 was isolated as
macrocycles comprising similar polyalkyl substituents have a white powder in a good yield of 89%. For the macrocycli-
2564
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 2562–2575

Shape-Switchable Azo-Macrocycles
Scheme 2. Synthetic route to target macrocycle 1. Reagents and conditions: (a) NBS, DMF, 0 °C Ǟ r.t., 86%; (b) NaNO2, H2SO4, EtOH,
75 °C, 60%; (c) [Pd(Ph3)4], K2CO3, C6H5CH3, EtOH, 85 °C, 95%; (d) LiAlH4, THF, r.t., 37–50%.
zation reaction similar conditions to those described above was isolated by CC in pure hexane in a good yield of 95%.
for the macrocycle 1 were applied. Thus, after treating 17 A Suzuki cross-coupling reaction[36] with commercially
in THF with LiAlH4, the macrocycle 2 was isolated in a available 3-nitrophenylboronic acid allowed the triply alkyl-
yield of 34% by size-exclusion chromatography (SEC) as an ated semicircle 24 to be prepared in 81% yield. The desired
intensely orange-coloured resin. A single peak in the analyt- macrocycle 3 was obtained by a reductive dimerization with
ical GPC (10.8 min) further confirmed the purity of the iso- LiAlH4 in THF at room temp. After purification with SEC
lated macrocycle 2.
by using toluene as eluent, the pure product was isolated in
As a more compact macrocycle consisting of trialkyl- 49% yield as an intensely orange-coloured resin. Analytical
functionalized semicircles, 3 was envisaged with three long GPC confirmed the purity of the compound with a single
alkyl chains directly connected to the central ring of the m- peak after 10.87 min.
terphenyl subunit (Scheme 4). With three alkyl chains per
To investigate the role of symmetry, an asymmetric
semicircle the solubility and adsorption properties of the macrocycle 4 was envisaged (Scheme 5). The Mills reac-
macrocycle 2 should be maintained, but the floppiness of tion[43] has previously turned out to be very efficient for
the semicircle subunits should be reduced due to the ab- constructing asymmetric azo compounds.[28,44] Because of
sence of the flexible –O–CH2 linker.
the different functionalities required as precursors (R–NO
The commercially available 2,6-diiodo-4-nitroaniline (18) and RЈ–NH2), it is ideally suited for the assembly of two
was first converted through a Sandmayer reaction[40] into different building blocks. Amino-functionalized semicircles
triiodonitrobenzene 19 in 80% yield. According to the lit- were readily available by reducing the corresponding dinitro
erature,[41] intermediate 21 was synthesized by a Sonog- semicircles with tin powder in concentrated hydrochloric
shira–Hagihara cross-coupling reaction followed by a palla- acid. However, a dinitroso-functionalized semicircle as the
dium-catalyzed hydrogenation in 53% yield over both steps. second precursor of the Mills reaction could not be ob-
The moderate yield of the hydrogenation reaction was due tained in reasonable yields, neither by reduction of a dinitro
to incomplete hydration hydrogenation of the double bond semicircle nor by oxidation of a diamino semicircle. Ac-
in the para position to the nitro group, probably as a result cording to a second strategy, an amino semicircle was
of the sterically hindered formation of the π complex in the treated with potassium peroxymonosulfate in a CH2Cl2/
1
catalytic cycle.[22,42] However, neither increased hydrogen H2O solvent mixture.[45] Both H NMR and MALDI-TOF
pressure nor an increased concentration of the catalyst pro- MS investigations revealed the presence of the desired com-
vided a better yield.
pound in the crude reaction mixture. However, all attempts
The subsequent assembly of the macrocycle 3 was similar to isolate this labile nitroso semicircle by CC or by crystalli-
to the synthesis described above for macrocycle 1. The reac- zation failed. As an alternative, an in situ preparation of
tion of the amine 21 with NBS in DMF gave the doubly the dinitroso semicircle was envisaged. However, treatment
brominated product 22 as a yellowish oil in 70% yield.[32] of the reaction mixture comprising the dinitroso semicircle
Deamination with concentrated sulfuric acid and sodium with a diamino semicircle only provided poor amounts of
nitrite provided the desired dibromo compound 23, which an inseparable mixture, probably of various oligomers.
Eur. J. Org. Chem. 2009, 2562–2575
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2565

M. Müri, K. C. Schuermann, L. De Cola, M. Mayor
FULL PAPER
Scheme 3. Synthetic route to target macrocycle 2. Reagents and conditions: (a) NaOMe, DMF, 80 °C, 70%; (b) [Pd(Ph3)4], K2CO3,
C6H5CH3, EtOH, 85 °C, 33%; (c) BBr3, DCM, –78 °C Ǟ r.t., 95%; (d) K2CO3, DMF, 65 °C, 89%; (e) LiAlH4, THF, r.t., 34%, (f)
BrC12H25, DMF, K2CO3, KI, 70 °C, 90%; (g) (1) LiAlH4, Et2O, r.t.; (2) NaOH, H2O, r.t., 96%; (h) SOCl2, DMF, DCM, r.t., 98%.
Scheme 4. Synthetic route to target macrocycle 3. Reagents and conditions: (a) (1) AcOH, NaNO2, H2SO4, 15 °C Ǟ r.t.; (2) KI, H2O,
CH3COOH, H2SO4, 95 °C, 80%; (b) HCCC10H21, [Pd(PPh3)2Cl2], CuI, NEt3, 85 °C, 89%; (c) H2, Pd/C, EtOH, AcOEt, r.t., 59%; (d)
NBS, DMF, 0 °C Ǟ r.t., 70%; (e) H2SO4, NaNO2, EtOH, 85 °C, 95%; (f) [Pd(Ph3)4], K2CO3, C6H5CH3, EtOH, 85 °C, 85%; (g) LiAlH4,
THF, 40 °C, 49 %.
An alternative strategy towards an asymmetric azo- scribed above. Equimolar amounts of the two different
macrocycle consists of the statistical assembly of two dif- semicircles 15 and 17 were dissolved in THF. A 1  LiAlH4
ferent dinitro semicircles. Thus, the two semicircles 15 and solution in THF was slowly added at r.t. over a period of
17 were subjected to the macrocyclization conditions de- 20 min. After keeping the reaction mixture at 40 °C for
2566
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 2562–2575

Shape-Switchable Azo-Macrocycles
nal-to-noise ratio for the relatively insensitive 13C NMR
spectra turned out to be too small to obtain meaningful
one-dimensional 13C{1H} NMR spectra.[46] However, the
1
more sensitive H-detected HMBC and HMQC spectra al-
lowed full characterization of the carbon resonances of
both macrocycles 2 and 3. Attempts to record the one-di-
mensional 13C NMR spectra of these two macrocycles in
their more compact Z,Z form by illumination at 313 nm
prior to recording the 13C NMR spectra were not success-
ful.
Of particular importance is the determination of the pu-
rity and the molecular weight of each of the isolated macro-
cycles, as the synthetic strategy based on the dimerization
of two bifunctional building blocks allows the formation
of larger cyclic oligomers with comparable NMR spectra.
Analytical gel permeation chromatograms were recorded to
document the purity of the macrocycles 1–4 and are dis-
played in the Supporting Information. All four compounds
displayed well-defined single peaks with retention times
Rt(1) = 11.85 min, Rt(2) = 10.8 min, Rt(3) = 10.87 min and
Rt(4) = 11.01 min in the expected order of decreasing mo-
lecular weight. However, owing to the particular shape and
thus hydrodynamic volume of the macrocycles 1–4, GPC is
not suited to determining their molecular weights. Surpris-
ingly, similar GPC conditions applied to polystyrene stan-
dards (see the Supporting Information) displayed systemati-
cally shorter retention times for the more compact macro-
cycles. Fortunately, the molecular signals of the macrocycles
1 and 3 were also detected in mass spectrometry by electron
impact ionization for 1 and by matrix-assisted laser desorp-
tion ionization (MALDI) in the case of 3, which enabled
the peaks detected by analytical GPC to be calibrated. All
attempts to ionize the macrocycles 2 and 4 for subsequent
detection by mass spectrometry failed. Neither the molecu-
lar ions nor characteristic decomposition fragments were
observed. As a complementary method for determining the
molecular weights of the macrocycles 1–4 vapour pressure
osmometry (VPO) was considered. Thus, the molecular
weights of the macrocycles 1–4 were determined by VPO.
All four compounds were measured in CHCl3 at 42 °C, and
the results are displayed in Table 1. Within the limits of ex-
perimental accuracy, the expected molecular weights were
recorded for all macrocycles 1–4, further corroborating
their structures.
Scheme 5. Synthetic route to the asymmetric macrocycle 4. Rea-
gents and conditions: (a) LiAlH4, THF, 40 °C, 44 %.
110 min, it was poured into water and extracted with
CH2Cl2. Owing to the statistical nature of the reaction, not
only the desired macrocycle 4 but most probably also the
two symmetric macrocycles were formed. Fortunately, the
molecular weights of the three macrocycles are sufficiently
different to allow their separation by SEC. Although ana-
lytical GPC was not able to separate the crude reaction mix-
ture (see the Supporting Information, ESI 9), preparative
SEC divided the sample into three well-separated main
bands, of which the most intense central one was isolated.
Owing to the limited amounts formed, the two ac-
companying bands were not further analyzed. The analyti-
cal data (1H NMR, GPC, VPO) of the isolated central band
indicated a pure chemical compound with the structure of
macrocycle 4. The isolated intensely orange-coloured resin
corresponded to a yield of 44% of the non-symmetrical
compound 4 and displayed a single symmetric peak with a
retention time of 11.01 min in analytical GPC. The effi-
ciency of the formation of the asymmetric macrocycle 4 re-
mains to some extent puzzling. It might be related to the
restricted formation of the symmetric macrocycle 2 com-
prising two sterically more demanding semicircles.
The chemical structures of all four macrocycles 1–4 were
fully confirmed by the analytical investigations described
above. Additional supporting evidence was obtained from
Characterization
Although the precursors 6–24 were fully characterized their UV/Vis and IR spectra. For example, the UV/Vis ab-
1
by H and 13C NMR spectroscopy, mass spectrometry and sorption in the region of 320 nm and the visible band at
elemental analysis, the characterization of the macrocycles about 450 nm, responsible of the bright-orange colour of
1–4 turned out to be more challenging.
the compounds, are characteristic features of the azo deriv-
Although the 1H NMR spectra corroborated the sug- atives. The cyclization reaction was further monitored by
gested structures of all the macrocycles 1–4, one-dimen- comparison of the IR spectra of the macrocycles 1–4 and
sional 13C NMR spectra could only be obtained for the their precursors. The disappearance of the characteristic
macrocycles 1 and 4. The size and the relatively high inertia aromatic nitro band in the region of 1520 cm–1 together
of each of the larger macrocycles 2 and 3 cause long rota- with the appearance of the signals at 1450 and 1240 cm–1,
tional correlation times τ. Because τ is proportional to the which were assigned to a (Z)-azo subunit, further document
linewidths of the resonances in the NMR spectra, the sig- the transformation of the nitro groups of the semicircles
Eur. J. Org. Chem. 2009, 2562–2575
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2567

M. Müri, K. C. Schuermann, L. De Cola, M. Mayor
FULL PAPER
Table 1. Molecular weights of 1–4 measured by VPO.
Macrocycle Measured molecular weight [g/mol] Calculated molecular weight [g/mol] Concentration [mg/g] Standard deviation[a]
1
2
3
4
709.1
1739.7
1500.7
1174.4
680.4
1830.8
1522.5
1201.7
3.93
5.22
4.69
1.44
0.06
0.05
0.07
0.09
[a] Standard deviation of the measured samples derived from at least three independent measurements.
into the azo groups in the macrocycles. Furthermore, even
To investigate the E Ǟ Z isomerization mechanism, all
linear oligomers could be excluded by IR spectroscopy, as the macrocycles 1–4 were irradiated at 313 nm. For the irra-
neither signals expected for aromatic nitro functions (1660– diation, a 200 W mercury lamp with a dichromic filter
1440 cm–1) nor those expected for aromatic amines (3500– (280–400 nm) and a 320 nm band-pass filter to select only
3300 cm–1) were observed.
the peak at 313 nm were used. All four macrocycles 1–4
displayed very similar spectral changes. As a typical exam-
ple, the behaviour of macrocycle 1 in THF at a concentra-
tion of 1.7ϫ10–5 mol/L is shown in Figure 2.
UV/Vis Measurements
All four macrocycles 1–4 showed absorption bands at
about 450 nm as well as at about 320 and 285 nm. In par-
ticular, the high-energy bands were attributed to πǞπ*
transitions and the broad low-energy absorption was attrib-
uted to nǞπ* transitions.[9,47]
All the macrocycles 1–4 displayed similar spectra that
differ mainly in the UV region, that is, at about 250 nm.
This has been attributed to the different substituents on the
central phenyl ring, which cause a twisting of the phenylene
units resulting in electronic decoupling of the aromatic moi-
eties. Thus, a blueshift of this band was observed for the
sterically more hindered compounds 2 and 3 (Figure 1).
Emission was not detected for any of these macrocycles,
neither at r.t. nor at 77 K.
Figure 2. Changes in the absorption spectra of macrocycle 1 in
THF upon irradiation at 313 nm. Inset: absorbance change versus
irradiation time.
The decrease in the absorbance of the πǞπ* band and
the increase in the nǞπ* absorbance upon light irradiation
over time was attributed to the formation of the Z isomer.
The photoisomerization reaction presented two isosbestic
points at 285 and 395 nm and reached the photostationary
state within 8 min of irradiation under the experimental
conditions applied.
The inset in Figure 2 displays the absorbance at 321 nm
plotted against the irradiation time. The evolution of the
321 nm signal (see inset) followed an exponential trend.
This suggests a switching dynamic for the pair of azo chro-
mophores in the macrocycle 1, which should involve, at le-
ast within the temporal resolution of the optical investiga-
tion, the switching of both azo groups.
The reaction was monitored by H NMR spectroscopy
in order to determine the amounts of Z and E isomers and
to calculate the molar extinction coefficient of the Z isomer.
Figure 1. Absorption spectra of the macrocycles 1–4 (1: solid black
line; 2: dashed line; 3: solid grey line; 4: dotted line) in THF solu-
tion.
1
Upon exposure to ambient light, considerable changes Isomerization was performed in a deuterated THF solution,
in the UV/Vis spectra were observed, as expected for azo and measurements were taken at different irradiation times.
1
derivatives. Photoisomerization reactions were investigated The H NMR spectra and the corresponding UV/Vis ab-
for all compounds 1–4 in THF by observing the changes in sorptions of macrocycle 1 are displayed in Figure 3; the
their absorption spectra. Furthermore, to quantify the E/Z other compounds 2–4 behave similarly, and their results are
ratio of isomers, the concentrations of the isomers formed summarized in Table 2.
were determined in deuteriated THF by 1H NMR spec-
troscopy.
The Z isomers displayed signals in the H NMR spectra
at frequencies different to those of the E isomers. Both the
1
2568
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 2562–2575

Shape-Switchable Azo-Macrocycles
Figure 3. (A) UV/Vis spectra of macrocycle 1 of the corresponding E/Z ratio (black: thermally stable state; dark grey: 50% isomerized;
1
light grey: photostationary state). (B) Corresponding H NMR spectra (markers indicate the peaks corresponding to the Z isomer).
Table 2. Quantum yields of photoisomerization and molar extinc-
tion coefficients (at 321 nm) of macrocycles 1–4.
The back Z Ǟ E back-isomerization was either triggered
by light or was observed as a thermal back-reaction. For
the light-induced back-reaction, the sample was irradiated
at the maximum of the peak of the Z isomer (450 nm). To
monitor the thermal back-reaction with time, the sample
was kept in the dark until the azo compound 1 had
switched back to its thermodynamically more stable E
form. The dynamics of the thermal back-reaction were in-
vestigated by monitoring the increasing absorbance at
321 nm (Figure 4). The thermal back-isomerization turned
out to be a very slow reaction, which indicates a very stable
Z isomer. The re-equilibration of the Z-enriched solution
back to a solution comprising exclusively the thermody-
namically favoured E form took several days. The rate con-
Macrocycle ΦE–Z ΦZ–E
εE
εZ
[dm3/molcm]
[dm3/molcm]
1
2
3
4
0.29
0.33
0.28
0.26
0.43
0.56
0.45
0.51
50800
49300
48300
44000
7100
15900
8600
14500
1
UV/Vis and H NMR spectra were used to calculate the
concentration of the Z compound formed. On the timescale
of the 1H NMR measurements, no detectable thermal back-
isomerization was observed, as shown by measuring the ab-
sorption spectra before and after the NMR experiments.
By integration of the corresponding 1H NMR signals,
the amount of E isomer in the photostationary state was
determined to be 15%. Furthermore, the ε values of the Z
isomers of the macrocycles 1–4 were determined and are
listed in Table 2.
stant
for
the
thermal
back-isomerization
was
1.15ϫ10–6 s–1, which is of a magnitude comparable to
those of similar systems with multiple azobenzene moie-
ties.[47]
By using the acquired data, the photoisomerization
quantum yields for the E Ǟ Z reactions were calculated
(see Table 2) according to the procedure reported by Zim-
merman et al.[48] (see the Exp. Sect.).
The ε values differed considerably for the macrocycles
and were slightly higher than those found in comparable
systems.[2] However, the calculated quantum yields were
comparable considering that both azo moieties underwent
photoisomerization. Despite the fact that the two azoben-
zene groups are linked together, the formation of an inter-
mediate E,Z isomer upon irradiation with UV light was not
1
observed. The UV/Vis and H NMR spectra gave no evi-
Figure 4. Thermal back-reaction Z Ǟ E of macrocycle 1 monitored
dence of an E,Z intermediate, as all the spectra can be ex-
pressed as the sum of the absorbances of the Z and E iso-
mers. The macrocycles 1–4 are extremely rigid, and there-
over time.
To obtain a faster Z Ǟ E isomerization and to monitor
fore a small motion of half of the macrocycle forces the the reversibility of the photoisomerization reaction, the
other half to switch accordingly. This explains the absence nǞπ* band at 450 nm was irradiated by using a xenon
of E,Z intermediates, which have been observed in compar- lamp and an appropriate set of filters (water filter and 450
able, but less rigid azo systems.[2]
BP). After the thermal back-reaction, the macrocycle 1 was
Eur. J. Org. Chem. 2009, 2562–2575
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2569

M. Müri, K. C. Schuermann, L. De Cola, M. Mayor
FULL PAPER
Figure 5. Photoinduced back-reaction of macrocycle 1, alternatingly irradiating with UV (313 nm) and blue (450 nm) light to switch
between the two photostationary states (PSS).
completely switched back to the E isomer, whereas in the methods. So far, all attempts to immobilize these macro-
photochemical back-reaction in the photostationary state cycles on graphite have failed in spite of their long alkyl
15% of the macrocycle still remained in its Z form. To gain chains.
information on the stability and fatigue resistance of the
system, macrocycle 1 was cycled five times. Reversible
Experimental Section
switching between the two photostationary states was ob-
served (Figure 5). Figure 5 shows data obtained by irradiat-
ing the macrocycle 1 at 313 nm for the E Ǟ Z isomerization
and at 450 nm for the Z Ǟ E conversion. The data is repre-
sentative for all the target compounds 1–4.
Reagents and Solvents: All chemicals were directly used for the syn-
thesis without further purification unless otherwise stated. Solvents
for chromatography and crystallization were distilled once before
use; technical grade solvents were used for extraction. Dry tetra-
hydrofuran (THF) was distilled from sodium and potassium, and
dry dichloromethane (DCM) was distilled from calcium hydride or
purchased over molecular sieves from Fluka.
Conclusions
Synthesis: All reactions with reagents that are easily oxidized or
Four shape-switchable macrocycles 1–4 consisting of m-
hydrolysed were performed under argon by using Schlenk tech-
terphenyl semicircles interlinked by two azo joints have niques; only dry solvents were used, and the glassware was heated
prior to use.
been synthesized and characterized. All four macroycles
were obtained by reductive dimerization of the correspond-
ing dinitro semicircles. Considering the polymerization po-
Analyses and Instruments: Bruker DPX NMR (400 MHz) and BZH
NMR (250 MHz) instruments were used to record H NMR spec-
1
tential of the synthetic strategy, all four azo-macrocycles tra. Chemical shifts (δ) are reported in parts per million (ppm)
relative to residual solvent peaks or trimethylsilane (TMS), and
coupling constants (J) are reported in Hertz (Hz). NMR solvents
were obtained from Cambridge Isotope Laboratories, Inc. (And-
over, MA, USA). The spectra were recorded at room temperature.
The multiplicities are denoted as the following: s = singlet, d =
doublet, q = quartet, quint = quintet, m = multiplet and br. =
were isolated in surprisingly good yields, ranging from 34%
for 2, to 44 and 49% for 4 and 3, respectively, up to 50%
1
for 1. These macrocycles were characterized by H and 13C
NMR spectra, profiting from the more sensitive 1H-de-
tected HMBC and HMQC techniques in the cases of 2 and
3. Although the purity of 1–4 was shown by analytical
GPC, their molecular weights were determined by VPO.
Mass spectra were also obtained for 1 and 3.
1
broad. In addition, H NMR spectra during isomerization studies
were recorded with a Bruker AM 400 MHz spectrometer. A Bruker
DPX (101 MHz) instrument was used to record 13C NMR spectra.
Chemical shifts (δ) are reported in parts per million (ppm) relative
All four macrocycles 1–4 displayed comparable E Ǟ Z
photoisomerization upon irradiation at 313 nm. An E/Z ra- to residual solvent peaks. NMR solvents were obtained from Cam-
bridge Isotope Laboratories, Inc. (Andover, MA, USA). The mea-
surements were performed at room temperature. The carbon atom
assignments are classified by the following: Cp = primary, Cs =
secondary, Ct = tertiary, Cq = quaternary. The peaks were related
to the carbon atoms by estimations and calculations of incremental
values. A Bruker (600 MHz) instrument was used to record 2D
NMR spectra. HMQC spectra were recorded with an acquisition
time in f2 of 85 ms and 1 K data points and an acquisition time in
f1 of 20 ms and 0.5 K data points. HMBC spectra were recorded
with an acquisition time in f2 of 170 ms and 2 K data points and
tio of 15:85 of the isomers was determined in their photo-
stationary states. The full thermal back-reaction required
several weeks. Hence, the thermodynamically less favoured
Z isomer showed substantial stabilization by the rigid
macrocyclic structure. Despite having two azo groups in the
macrocyclic periphery, an intermediate mixed E,Z isomer
was not observed; the stiff terphenyl semicircles probably
disfavour a strained mixed state.
These shape-switchable macrocycles 1–4 were designed to
investigate their spatial transformations by scanning probe an acquisition time in f1 of 34 ms and 1 K data points.[49] Mass
2570
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 2562–2575

Shape-Switchable Azo-Macrocycles
spectra were recorded with a Bruker Esquire 3000 Plus for electron
spray ionization (ESI), a Finnigan MAT 95Q for electron impact
(EI), a Finnigan MAT 8400 for fast atom bombardment (FAB) or
a Voyager-De Pro for MALDI-TOF. The peaks were measured
in m/z (%). Elementary analyses were carried out with a Perkin-
Elmer Analysator 240. The samples for UV analysis were irradiated
by using a Lot-Oriel 200 W high-pressure mercury lamp with a
280–400 nm dichroic mirror (to remove IR and visible light) and a
Perkin-Elmer 320 nm band-pass filter to allow only light from the
313 nm Hg line. The intensity of the light source was determined
by chemical actinometry using a 6.4ϫ10–4 molL–1 azobenzene
(Sigma–Aldrich, 99%) solution in methanol. All samples were con-
stantly stirred upon irradiation. Spectra were recorded with a Var-
ian Cary 5000 two-beam spectrophotometer with baseline correc-
tion and pure solvent as the reference. Spectroscopic grade THF
and methanol were used as obtained from Merck. Samples for IR
spectroscopy were measured neat with a Shimadzu FTIR 8400S
Fourier transform IR spectrometer. A Shimadzu LC-8A instrument
was used to record gel permeation chromatograms. The measure-
ments were performed at room temperature with an OligoPore
300ϫ7.5 mm column (particle size 6 µm) from Polymer Laborato-
ries by eluting with toluene at a flow rate of 0.5 mL/min at λ =
220 nm. Calibrations were performed with polystyrene from Poly-
mer Laboratories under the same conditions as used for the analy-
sis. Vapour pressure osmometry (VPO) experiments were per-
formed with a Knauer K-7000 vapour pressure osmometer. For the
calculations the software EuroOsmo 7000, Version 1.5, was used.
For column chromatography (CC), silica gel 60 (40–63 µm) from
Merck or silica gel 60 (40–63 µm) from Fluka was used. Preparative
size-exclusion chromatography (SEC) was performed with Bio-Be-
ads® S-X Beads from BIO RAD with distilled toluene as eluent.
For thin-layer chromatography (TLC), silica gel 60 F254 glass plates
with a thickness of 0.25 mm from Merck were used. Detection was
carried out by using a UV lamp at 253 or 366 nm.
MS (EI): m/z (%) = 318 (17), 320 (36), 322 (17), 250 (100), 169
(51), 89 (22).
1-Hexyl-3,5-bis(3-nitrophenyl)benzene (8): 3-Dibromo-5-hexylben-
zene (7; 1.00 g, 3.150 mmol), 3-nitrophenylboronic acid (1.41 g,
8.180 mmol), tetrakis(triphenylphosphane)palladium (290 mg,
0.252 mmol) and potassium carbonate (3.48 g, 25.200 mmol) were
added to a solution of dry toluene (100 mL) and dry ethanol
(30 mL) under argon. The potassium carbonate did not completely
dissolve. The reaction mixture was stirred vigorously at 85 °C for
3 h. After cooling to room temperature, the reaction mixture was
filtered through a 1 cm pad of silica, eluting with toluene and etha-
nol. The residue was absorbed with dichloromethane on silica and
then purified by chromatography (silica gel, ethyl acetate/hexane,
1:5, Rf = 0.32) to give the product 8 as a light-yellow oil, which
precipitated to give a light-green solid after standing (1.21 g, 95%,
1
4
m.p. 86–89 °C). H NMR (400 MHz, CDCl3): δ = 8.50 (t, JHH
=
3
4
5
2.0 Hz, 2 H), 8.24 (ddd, JHH = 8.1, JHH = 2.0, JHH = 1.0 Hz, 2
3
4
5
H), 7.98 (ddd, JHH = 8.1, JHH = 2.0, JHH = 1.0 Hz, 2 H), 7.66
(t, 3JHH = 8.1 Hz, 2 H), 7.65 (t, 4JHH = 1.5 Hz, 1 H), 7.50 (d, 4JHH
3
3
= 1.5 Hz, 2 H), 2.79 (t, JHH = 7.6 Hz, 2 H), 1.73 (quint, JHH
=
3
7.6 Hz, 2 H), 1.42 (quint, JHH = 7.6 Hz, 2 H), 1.35 (m, 4 H), 0.91
(t, JHH = 7.1 Hz, 3 H) ppm. 13C NMR (101 MHz, CDCl3): δ =
3
149.2 (s, 2 C, Cq), 145.6 (s, 1 C, Cq), 143.0 (s, 2 C, Cq), 140.2 (s,
2 C, Cq), 133.6 (s, 2 C, Ct), 130.2 (s, 2 C, Ct), 127.9 (s, 2 C, Ct),
123.9 (s, 1 C, Ct), 122.8 (s, 2 C, Ct), 122.5 (s, 2 C, Ct), 36.5 (s, 1
C, Cs), 32.1 (s, 1 C, Cs), 32.0 (s, 1 C, Cs), 29.5 (s, 1 C, Cs), 23.0 (s,
1 C, Cs), 14.5 (s, 1 C, Cp) ppm. MS (EI): m/z (%) = 404.2 (69),
405.2 (18), 334 (100), 287 (25), 239 (43), 226 (23). C24H24N2O4
(404.46): calcd. C 71.27, H 5.98, N 6.93; found C 71.10, H 5.96, N
6.87.
Macrocycle 1: 1-Hexyl-3,5-bis(3Ј-nitrophenyl)benzene (8; 300 mg,
0.742 mmol) was dissolved under argon in dry THF (3 mL). A 1 
lithium aluminium hydride solution in THF (3.0 mL, 2.969 mmol)
was carefully added over a period of 15 min. After complete ad-
dition, the reaction mixture was stirred at room temperature for
70 min and at 55 °C for an additional 20 min. After cooling to
room temperature, water (7 mL) was added carefully, and, after
stirring for 5 min, the reaction mixture was filtered. The filtrate was
extracted with dichloromethane (2ϫ10 mL), and the combined or-
ganic layers were dried with MgSO4 and the solvents evaporated.
The almost pure crude product was dissolved in a warm mixture
of ethanol/toluene (6:5, 5 mL). After 10 d in the freezer at –20 °C,
the solvent was decanted, and the dark-orange oily residue was
washed with ethanol and dried at high vacuum for 5 h to give the
macrocycle 1 as a bright-orange solid (504 mg, 50%, m.p. 155–
157 °C). TLC (silica gel, DCM/hexane, 2:1): Rf = 0.32 (due to the
limited solubility of 1 in the eluent, the spot was observed to have
2,6-Dibromo-4-hexylbenzenamine (6): 4-Hexylaniline (5; 4.520 g,
25.480 mmol) was dissolved in DMF (30 mL) and cooled to 0 °C.
N-Bromosuccinimide (11.34 g, 63.710 mmol) in DMF (30 mL) was
added dropwise over a period of 30 min while stirring. The reaction
mixture was stirred at 0 °C for 30 min and at room temperature for
another 3 h. Afterwards, the DMF was removed by rotary evapora-
tion. The residue was purified by CC (silica gel, toluene/DCM, 2:1,
Rf = 0.73) to afford the product 6 (7.27 g, 86%) as dark-red crystals
after standing overnight. 1H NMR (250 MHz, CDCl3): δ = 7.19 (s,
3
2 H), 4.39 (br. s, 2 H, NH2), 2.45 (t, JHH = 7.4 Hz, 2 H), 1.53 (m,
3
2 H), 1.28 (m, 6 H), 0.88 (m, JHH = 6.0 Hz, 3 H, CH3) ppm.
1,3-Dibromo-5-hexylbenzene (7): 2,6-Dibromo-4-hexylbenzenamine
(6; 7.26 g, 21.800 mmol) was dissolved in ethanol (130 mL) and
concd. sulfuric acid (10 mL). Sodium nitrite (4.50 g, 65.400 mmol)
was slowly added in small portions while stirring. Afterwards, the
reaction mixture was stirred at 75 °C for 2 h. The reaction mixture
was then cooled with an ice bath to 5 °C, water (50 mL) was added,
and the suspension was extracted with DCM (2ϫ30 mL). The
combined organic layers were dried with MgSO4 and the solvents
evaporated. The black oily residue was dissolved in hexane (2 mL)
and ethyl acetate (1 mL) and purified by column chromatography
(silica gel, hexane, Rf = 0.56) to afford the product 7 (4.170 g, 60%)
1
a long tail). H NMR (400 MHz, CDCl3): δ = 8.28 (s, 4 H), 7.96
(m, 4 H), 7.80 (m, 6 H), 7.60 (m, 4 H), 7.53 (s, 4 H), 2.77 (m, 4
H), 1.74 (m, 4 H), 1.41 (m, 4 H), 1.33 (m, 8 H), 0.89 (m, 6 H) ppm.
13C NMR (101 MHz, CDCl3): δ = 153.6 (s, 4 C, Cq), 144.7 (s, 4
C, Ct), 142.8 (s, 4 C, Cq), 141.5 (s, 4 C, Cq), 130.3 (s, 2 C, Cq),
130.0 (s, 4 C, Ct), 127.3 (s, 4 C, Ct), 124.1 (s, 4 C, Ct), 122.6 (s, 2
C, Ct), 122.0 (s, 4 C, Ct), 36.7 (s, 2 C, Cs), 32.2 (s, 2 C, Cs), 32.1
(s, 2 C, Cs), 29.6 (s, 2 C, Cs), 23.1 (s, 2 C, Cs), 14.6 (s, 2 C, Cp)
ppm. IR (ATR, neat): ν = 3051 (w), 3024 (w), 2924 (s), 2852 (m),
˜
1
as a light-yellow liquid. H NMR (400 MHz, CDCl3): δ = 7.47 (t,
1593 (m), 1504 (w), 1445 (4), 1441 (m), 1155 (m) cm–1. MS (EI):
m/z (%) = 680.3 (6), 681 (3), 344 (100), 85 (11), 71 (17), 57 (33), 44
(35).
4
3
4JHH = 2.0 Hz, 1 H), 7.24 (d, JHH = 2.0 Hz, 2 H), 2.53 (t, JHH
=
3
3
7.6 Hz, 2 H), 1.57 (tt, JHH = 7.6, JHH = 6.6 Hz, 2 H), 1.29 (m, 6
H), 0.88 (t, JHH = 7.1 Hz, 3 H) ppm. 13C NMR (101 MHz,
3
CDCl3): δ = 147.3 (s, 1 C, Cq), 131.7 (s, 1 C, Ct), 130.7 (s, 2 C,
1,3-Dibromo-5-methoxybenzene (14): Sodium methoxide (1.20 g,
Ct), 123.1 (s, 2 C, Cq), 35.8 (s, 1 C, Cs), 32.0 (s, 1 C, Cs), 31.4 (s, 22.24 mmol) was added to a solution of 1,3,5-tribromobenzene (13;
1 C, Cs), 29.2 (s, 1 C, Cs), 23.0 (s, 1 C, Cs), 14.5 (s, 1 C, Cp) ppm. 5.10 g, 15.88 mmol) in DMF (50 mL). The reaction mixture was
Eur. J. Org. Chem. 2009, 2562–2575
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2571

M. Müri, K. C. Schuermann, L. De Cola, M. Mayor
FULL PAPER
4
stirred at 80 °C for 8 h, quenched with a 10% aq. HCl solution
(20 mL) and extracted with tert-butyl methyl ether (3ϫ30 mL).
The combined organic layers were washed with brine (35 mL),
dried with magnesium sulfate and the solvents evaporated. The
crude product was purified by CC (silica gel, hexane, Rf = 0.3) to
1.5 Hz, 1 H), 7.29 (d, JHH = 1.5 Hz, 2 H), 6.69 (s, 2 H), 5.11 (s, 2
H), 4.00 (m, 6 H), 1.81 (m, 6 H), 1.48 (m, 6 H), 1.25 (m, 48 H),
0.87 (m, 9 H) ppm. 13C NMR (101 MHz, CDCl3): δ = 160.4 (s, 1
C, Cq), 153.9 (s, 2 C, Cq), 149.2 (s, 2 C, Cq), 142.6 (s, 2 C, Cq),
141.5 (s, 2 C, Cq), 138.6 (s, 1 C, Cq), 133.6 (s, 2 C, Ct), 131.6 (s, 1
C, Cq), 130.3 (s, 2 C, Ct), 123.0 (s, 2 C, Ct), 122.5 (s, 2 C, Ct),
119.2 (s, 1 C, Ct), 114.2 (s, 2 C, Ct), 106.6 (s, 2 C, Ct), 73.9 (s, 1
C, Cs), 71.3 (s, 1 C, Cs), 69.6 (s, 2 C, Cs), 32.3(s, 3 C, Cs), 29.8 (m,
afford the title compound 14 (2.95 g, 70%) as a white solid (m.p.
42–43 °C). 1H NMR (400 MHz, CDCl3): δ = 7.25 (t, JHH
=
4
4
1.8 Hz, 1 H), 6.99 (d, JHH = 1.8 Hz, 2 H), 3.78 (s, 3 H) ppm. 13C
NMR (101 MHz, CDCl3): δ = 161.2 (s, 1 C, Cq), 126.8 (s, 1 C, Ct), 21 C, Cs), 26.5 (s, 3 C, Cs), 23.1 (s, 3 C, Cs), 14.5 (s, 3 C, Cp) ppm.
123.5 (s, 2 C, Cq), 116.9 (s, 2 C, Ct), 56.1 (s, 1 C, Cq) ppm. MS
(EI): m/z (%) = 265.8 (100), 263.8 (49), 266.8 (7), 267.8 (47), 235.8
(19). C7H6Br2O (265.93): calcd. C 31.62, H 2.27; found C 31.49, H
2.27.
IR (ATR, neat): ν = 3089.8 (w), 2916.2 (s), 2848.7 (m), 1593.1 (m),
˜
1521.7 (s), 1342.4 (s), 1244.0 (m), 1195.8 (s), 1114.8 (s), 1012.6 (m),
808.1 (s) cm–1. MS (FAB): m/z (%) = 978.4 (1.4), 979.4 (1.1), 980.2
(0.6), 643.4 (7.6), 383.0 (13), 138.9 (31), 90.9 (60), 57.0 (92), 55.0
(80), 43.0 (100). C61H90N2O8 (979.38): calcd. C 74.81, H 9.26, N
2.86; found C 74.87, H 9.00, N 2.65.
5-Methoxy-1,3-bis(3-nitrophenyl)benzene (15): (3-Nitrophenyl)bo-
ronic acid (816 mg, 4.89 mmol), tetrakis(triphenylphosphane)palla-
dium (217 mg, 0.19 mmol) and potassium carbonate (2.07 g,
15.04 mmol) were added to a solution of 14 (500 mg, 1.88 mmol)
in dry toluene (40 mL) and dry ethanol (15 mL) under argon. The
reaction mixture was stirred vigorously at 85 °C for 150 min. After
cooling to room temperature, the reaction mixture was filtered
through a 1 cm pad of silica gel, the pad was washed with toluene
(150 mL), and the solvents were evaporated to dryness. The dark
crude product was recrystallized from hot toluene to afford the title
compound 15 (219 mg, 33%) as a white powder (m.p. 203–205 °C).
1H NMR (400 MHz, [D6]DMSO): δ = 8.57 (t, 4JHH = 2.0 Hz), 8.28
Macrocycle 2: A 1  lithium aluminium hydride solution in THF
(0.6 mL, 0.601 mmol) was slowly added to a solution of 17 (58 mg,
0.059 mmol) in dry THF (10 mL) under argon at room tempera-
ture. The reaction mixture was stirred at room temperature for
2.5 h, and then the black suspension was quenched with water
(10 mL). After extraction with toluene (3ϫ15 mL), the combined
organic layers were washed with brine (20 mL), dried with magne-
sium sulfate and the solvents evaporated to dryness. The crude
product was purified by size-exclusion CC (Bio-Beads® S-X Beads,
toluene) to afford the macrocycle 2 (18 mg, 34%) as an orange resin
(m.p. 98–101 °C). TLC (silica gel, DCM/hexane, 2:1): Rf = 0.52
(due to the limited solubility of 2 in the eluent, the spot was ob-
served with a long tail). 1H NMR (400 MHz, CDCl3): δ = 8.26 (m,
4 H), 7.95 (m, 4 H), 7.77 (m, 4 H), 7.60 (m, 6 H), 7.34 (m, 4 H),
6.69 (m, 4 H), 5.10 (m, 4 H), 3.96 (m, 12 H), 1.75 (m, 12 H), 1.44
(m, 12 H), 1.22 (m, 96 H), 0.85 (m, 18 H) ppm. 13C NMR (CDCl3,
determined by HMBC and HMQC): δ = 159.8 (2 C, Cq), 153.4 (4
C, Cq), 153.2 (4 C, Cq), 142.4 (4 C, Cq), 142.0 (4 C, Cq), 138.1 (2
C, Cq), 131.6 (2 C, Cq), 130.0 (4 C, Ct), 129.7 (4 C, Ct), 122.2 (4
C, Ct), 121.9 (4 C, Ct), 119.0 (4 C, Ct), 113.3 (2 C, Ct), 106.1 (4
C, Ct), 73.5 (2 C, Cs), 70.9 (2 C, Cs), 69.2 (4 C, Cs), 32.0 (6 C, Cs),
30.4 (6 C, Cs), 29.6 (36 C, Cs), 26.1 (6 C, Cs), 22.7 (6 C, Cs), 14.1
3
3
4
(d, JHH = 7.8 Hz), 8.23 (dd, JHH = 7.8, JHH = 2.0 Hz), 7.77 (t,
3JHH = 7.8 Hz, 2 H), 7.71 (t, JHH = 1.4 Hz, 1 H), 7.39 (d, JHH
=
4
4
1.4 Hz, 2 H), 3.93 (s, 3 H) ppm. 13C NMR (101 MHz, [D6]DMSO):
δ = 161.4 (s, 1 C, Cq), 149.3 (s, 2 C, Cq), 142.3 (s, 2 C, Cq), 141.1
(s, 2 C, Cq), 134.7 (s, 2 C, Ct), 131.3 (s, 2 C, Ct), 123.4 (s, 2 C, Ct),
122.6 (s, 2 C, Ct), 119.3 (s, 1 C, Ct), 113.7 (s, 2 C, Ct), 56.5 (s, 1
C, Cp) ppm. IR (ATR, neat): ν = 3072 (w), 2939 (w), 1595 (m),
˜
1518 (s), 1456 (m), 1344 (s), 1278 (m), 1209 (m), 1072 (m), 1041
(m), 854 (m) cm–1. MS (EI): m/z (%) = 350.1 (100), 351.1 (21),
352.1 (3), 258.1 (13), 215.1 (17). C19H14N2O5 (350.32): calcd. C
65.14, H 4.03, N 8.00; found C 64.80, H 4.21, N 7.93.
3,5-Bis(3-nitrophenyl)phenol (16): BBr3 (174 µL, 1.84 mmol) was
slowly added to a solution of 15 (150 mg, 0.43 mmol) in dry DCM
(15 mL) at –78 °C. The orange reaction mixture was stirred for
2.5 h at room temperature and poured onto ice (10 g). After extrac-
tion with DCM (3ϫ10 mL), the combined organic layers were
washed with brine (20 mL), dried with magnesium sulfate and the
solvents evaporated to dryness to afford the desired compound 16
(136 mg, 95%) as a pale-yellow solid (m.p. 190–192 °C). The prod-
uct 16 was used in the next step without further characterization.
(6 C, Cp) ppm. IR (ATR, neat): ν = 3059 (w), 3024 (w), 2920 (s),
˜
2851 (m), 1591 (m), 1504 (w), 1466 (m), 1437 (m), 1334 (m), 1112
(s) cm–1.
Macrocycle 4: A 1  lithium aluminium hydride solution in THF
(1.25 mL, 1.250 mmol) was slowly added to a solution of 15
(43.6 mg, 0.125 mmol) and 17 (122 mg, 0.125 mmol) in dry THF
(30 mL) over a period of 20 min. After complete addition, the reac-
tion mixture was heated to 40 °C and stirred for 110 min. The mix-
ture was cooled to room temperature and quenched with water
(20 mL). After extraction with DCM (2ϫ15 mL), the combined
organic layers were washed with brine (20 mL), dried with magne-
sium sulfate and the solvents evaporated to dryness. The orange
crude product was purified by size-exclusion CC (Bio-Beads® S-X
4
1H NMR (400 MHz, CDCl3): δ = 8.48 (t, JHH = 1.9 Hz, 2 H),
3
8.28 (m, 2 H), 7.98 (m, 2 H), 7.67 (t, JHH = 8.1 Hz, 2 H), 7.41 (t,
4
4JHH = 1.5 Hz, 1 H), 7.17 (d, JHH = 1.5 Hz, 2 H) ppm. MS
(MALDI-TOF): m/z (%) =336.7 (30), 335.7 (100).
1,3-Bis(3-nitrophenyl)-5-{[3,4,5-tris(dodecyloxy)phenyl]methoxy}- Beads, toluene) to afford macrocycle 4 (66 mg, 44%) as an orange
benzene (17): Potassium carbonate (245 mg, 1.78 mmol) was added resin (m.p. 120–125 °C). TLC (silica gel, DCM/hexane, 2:1): Rf =
to a solution of 16 (60 mg, 0.18 mmol) and 12 (133 mg, 0.20 mmol) 0.33 (due to the limited solubility of 4 in the eluent, the spot was
1
in DMF (25 mL). The orange reaction mixture was stirred at 65 °C
for 22 h. The violet reaction mixture was cooled to room tempera-
ture and quenched with water (20 mL). After extraction with DCM
(3ϫ20 mL), the combined organic layers were washed with water
(2ϫ20 mL), dried with magnesium sulfate and concentrated to
dryness. The crude product was purified by CC (silica gel, ethyl
acetate/hexane, 1:7, Rf = 0.25) to afford the title compound
(155 mg, 89%) as a white powder (m.p. 88–90 °C). 1H NMR
observed with a long tail). H NMR (400 MHz, CDCl3,): δ = 8.26
(m, 4 H), 7.95 (m, 4 H), 7.77 (m, 4 H), 7.59 (m, 6 H), 7.35 (m, 2
H), 6.70 (m, 2 H), 5.11 (m, 2 H), 3.96 (m, 9 H), 1.77 (m, 6 H), 1.46
(m, 6 H), 1.25 (m, 48 H), 0.87 (m, 9 H) ppm. 13C NMR (101 MHz,
CDCl3): δ = 160.9 (s, 1 C, Cq), 160.0 (s, 1 C, Cq), 153.8 (s, 4 C,
Cq), 153.5 (s, 2 C, Cq), 142.9 (s, 4 C, Cq), 142.4 (s, 4 C, Cq), 138.4
(s, 1 C, Cq), 132.1 (s, 1 C, Cq), 130.3 (s, 4 C, Ct), 130.0 (s, 4 C,
Ct), 122.4 (s, 4 C, Ct), 122.3 (s, 4 C, Ct), 113.5 (s, 2 C, Ct), 112.7
(s, 4 C, Ct), 106.7 (s, 2 C, Ct), 73.9 (s, 1 C, Cs), 71.2 (s, 1 C, Cs),
67.1 (s, 2 C, Cs), 56.0 (s, 1 C, Cq), 32.4 (s, 3 C, Cs), 30.2 (s, 21 C,
4
(400 MHz, CDCl3): δ = 8.49 (t, JHH = 1.9 Hz, 2 H), 8.25 (m, 2
3
4
H), 7.95 (m, 2 H), 7.65 (t, JHH = 8.0 Hz, 2 H), 7.44 (t, JHH
=
2572
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 2562–2575

Shape-Switchable Azo-Macrocycles
Cs), 26.6 (s, 3 C, Cs), 23.1 (s, 3 C, Cs), 14.6 (s, 3 C, Cp) ppm. IR
(60), 288.2 (9), 134.1 (12). C42H79N (598.08): calcd. C 84.34, H
13.31, N 2.34; found C 84.36, H 13.12, N 2.26.
(ATR, neat): ν = 3063 (w), 2920 (s), 2851 (m), 1591 (m), 1500 (w),
˜
1437 (m), 1337 (m), 1211 (w), 1113 (s) cm–1.
2,6-Dibromo-3,4,5-tri(dodecyl)aniline (22): A solution of dry DMF
(20 mL) and 3,4,5-tri(dodecyl)aniline (21; 1.90 g, 3.18 mmol) was
cooled to 0 °C under argon. N-Bromosuccinimide (1.36 g,
7.62 mmol) in dry DMF (20 mL) was added at 0 °C over a period
of 20 min. After the complete addition, the reaction mixture was
stirred at room temperature for another 120 min. The dark-brown
reaction mixture was quenched with water (30 mL) and extracted
with tert-butyl methyl ether (3ϫ20 mL); the combined organic lay-
ers were dried with magnesium sulfate and the solvents evaporated
to dryness. The black crude product was purified by CC (silica gel,
ethyl acetate/hexane, 1:30, Rf = 0.44) to afford the title compound
22 (1.68 g, 70%) as a pale-yellow oil. 1H NMR (250 MHz, CDCl3):
δ = 4.56 (s, 2 H), 2.71 (m, 4 H), 2.55 (m, 2 H), 1.53–1.27 (m, 60 H),
0.88 (t, 3JHH = 6.5 Hz, 9 H) ppm. 13C NMR (101 MHz, CDCl3): δ
= 140.5 (s, 1 C, Cq), 140.3 (s, 2 C, Cq), 130.6 (s, 1 C, Cq), 110.6
(s, 2 C, Cq), 34.2 (s, 3 C, Cs), 32.4 (s, 3 C, Cs), 30.1 (m, 24 C, Cs),
1,2,3-Triiodo-5-nitrobenzene (19): 2,6-Diiodo-4-nitroaniline (18;
10.3 g, 25.65 mmol) was dissolved in acetic acid (60 mL) at 15 °C,
and then a solution of sodium nitrite in sulfuric acid (15 mL) was
added dropwise at room temperature. The reaction mixture was
stirred for 90 min until everything had dissolved. The yellow reac-
tion mixture was poured into ice/water (200 mL), stirred vigorously,
quenched with urea (4.0 g) and filtered. The filtrate was diluted
with water (30 mL), potassium iodide (6.0 g, 36.23 mmol) in water
(30 mL) was added and the mixture stirred vigorously. Afterwards,
the brown reaction mixture was heated to 95 °C and stirred for 2 h.
The cold mixture was then quenched with sodium bisulfite (5.0 g)
and the precipitate filtered. The crude product was dried at high
vacuum and recrystallized from 2-methoxyethanol (20 mL) to af-
ford the desired product 19 (10.3 g, 80%) as brown crystals (m.p.
1
166–168 °C). H NMR (400 MHz, CDCl3): δ = 8.60 (s, 2 H) ppm.
13C NMR (101 MHz, CDCl3): δ = 147.68 (s, 1 C, Cq), 133.06 (s, 2
23.1 (s, 3 C, Cs), 14.5 (s, 3 C, Cp) ppm. IR (ATR, neat): ν = 3485
˜
C, Ct), 131.26 (s, 1 C, Cq), 107.20 (s, 2 C, Cq) ppm. IR (ATR, (w), 3387 (w), 2954 (w), 2920 (s), 2852 (m), 1599 (m), 1466 (m),
neat): ν = 3082 (w), 3049 (w), 1503 (s), 1498 (s), 1326 (s), 874 (s), 1430 (w) cm–1. MS (EI): m/z (%) = 753.3 (51), 754.4 (24), 755.3
˜
734 (s) cm–1. MS (EI): m/z (%) = 500.8 (100), 501.8 (6.7), 454.9 (100), 756.3 (45), 757.3 (56), 758.3 (23), 759.3 (5), 600.2 (6), 521.3
(15), 327.9 (15), 201.0 (23), 74.0 (30). C6H2I3NO2 (500.80): calcd.
C 14.39, H 0.40, N 2.80; found C 14.41, H 0.42, N 2.85.
(7), 433.0 (7). C42H77Br2N (755.88): calcd. C 66.74, H 10.27, N
1.85; found C 66.78, H 10.23, N 1.70.
1,2,3-Tri(dodec-1-ynyl)-5-nitrobenzene (20): 1,2,3-Triiodo-5-nitro-
benzene (19; 2.0 g, 3.99 mmol), 1-dodecyne (2.85 mL, 13.18 mmol),
1,5-Dibromo-2,3,4-tri(dodecyl)benzene (23): 2,6-Dibromo-3,4,5-tri-
(dodecyl)aniline (22; 83 mg, 0.110 mmol) was dissolved in ethanol
(5 mL) by sonication. After the addition of sulfuric acid (2 mL)
and sodium nitrite (23 mg, 0.331 mmol), the reaction mixture was
stirred at 85 °C for 4 h, cooled to room temperature and quenched
with water (10 mL). The water phase was extracted with DCM
(3ϫ10 mL), the combined organic layers were dried with magne-
sium sulfate and the solvents evaporated to dryness. The crude
product was purified by CC (silica gel, hexane, Rf = 0.58) to afford
the title compound 23 (78 mg, 95%) as a yellow oil. 1H NMR
(400 MHz, CDCl3): δ = 7.57 (s, 1 H), 2.62 (m, 6 H), 1.51–1.24 (m,
bis(triphenylphosphane)palladium(II)
dichloride
(140 mg,
0.199 mmol) and copper(I) iodide (76 mg, 0.399 mmol) were added
to degassed triethylamine (60 mL) under argon. After 150 min at
85 °C, the black reaction mixture was cooled to room temperature,
filtered, the filtrate washed with water (30 mL) and dried with mag-
nesium sulfate. The crude product was purified by CC (silica gel,
ethyl acetate/hexane, 1:20, Rf = 0.41) to afford the title compound
1
20 (2.18 g, 89%) as a black oil. H NMR (400 MHz, CDCl3): δ =
3
3
8.07 (s, 2 H), 2.54 (t, JHH = 7.0 Hz, 2 H), 2.46 (t, JHH = 7.0 Hz,
3
3
4 H), 1.63 (m, 6 H), 1.48 (m, 6 H), 1.27 (m, 36 H), 0.87 (t, JHH
=
60 H), 0.85 (t, JHH = 6.5 Hz, 9 H) ppm. 13C NMR (101 MHz,
7.0 Hz, 9 H) ppm. 13C NMR (101 MHz, CDCl3): δ = 146.0 (s, 1
C, Cq), 134.8 (s, 1 C, Cq), 128.4 (s, 2 C, Ct), 125.4 (s, 2 C, Ct),
104.5 (s, 1 C, Cq), 97.4 (s, 2 C, Cq), 78.6 (s, 1 C, Cq), 78.5 (s, 2 C,
Cq), 32.3 (s, 3 C, Cs), 30.1–29.0 (m, 18 C, Cs), 23.1 (s, 3 C, Cs),
CDCl3): δ = 142.7 (s, 1 C, Cq), 140.1 (s, 2 C, Cq), 134.2 (s, 1 C,
Ct), 122.8 (s, 2 C, Cq), 33.5 (s, 3 C, Cs), 32.4 (s, 3 C, Cs), 30.1 (m,
24 C, Cs), 23.1 (s, 3 C, Cs), 14.5 (s, 3 C, Cp) ppm. MS (EI): m/z
(%) = 738.4 (49), 739.4 (22), 740.4 (100), 741.4 (44), 742.4 (55),
743.4 (23), 744.4 (5), 418.1 (15), 276.9 (33), 197.0 (10), 71.1 (14),
20.1 (s, 3 C, Cs), 14.5 (s, 3 C, Cp) ppm. IR (ATR, neat): ν = 3002
˜
(w), 2921 (s), 2853 (m), 1521 (m), 1465 (w), 1347 (m) cm–1. MS 57.1(19). C42H76Br2 (740.86): calcd. C 68.09, H 10.34; found C
(EI): m/z (%) = 615.4 (35), 616.3 (16), 617.3 (0.6), 585.4 (16), 530.3
(18), 474.2 (22), 446.2 (19), 404.2 (20), 390.2 (30), 378.2 (49), 83.1
(45), 69.1 (51), 55.0 (67), 43.0 (100). C42H65NO2 (615.97): calcd. C
81.89, H 10.64, N 2.27; found C 81.73, H 10.69, N 1.93.
68.02, H 10.11.
1,3-Bis(3-nitrophenyl)-4,5,6-tri(dodecyl)benzene (24): (3-Nitrophen-
yl)boronic acid (46 mg, 0.274 mmol), tetrakis(triphenylphosphane)-
palladium (12 mg, 0.011 mmol) and potassium carbonate (116 mg,
0.842 mmol) were added to a solution of 1,5-dibromo-2,3,4-tri(do-
decyl)benzene (23; 78 mg, 0.105 mmol) in dry toluene (10 mL) and
dry ethanol (3 mL) under argon. The reaction mixture was stirred
vigorously at 85 °C for 150 min. After cooling to room tempera-
ture, the reaction mixture was filtered through a 1 cm pad of silica,
eluted with toluene (15 mL), and the solvents were evaporated to
dryness. The dark crude product was purified by CC (silica gel,
toluene/hexane, 1:1, Rf = 0.59) to afford the title compound 24
3,4,5-Tri(dodecyl)aniline (21): 1,2,3-Tri(dodec-1-ynyl)-5-nitroben-
zene (20; 1.28 g, 2.08 mmol) and palladium on activated charcoal
(440 mg, 0.416 mmol) were dissolved in dry ethanol (80 mL) and
dry ethyl acetate (20 mL). The reaction mixture was hydrogenated
(two hydrogen balloons) at room temperature for 28 h while stirring
vigorously. After filtration through Celite, the concentrated crude
product was purified by CC (silica gel, DCM/hexane, 1:1, Rf =
0.32) to afford the title compound 21 (711 mg, 59%) as a pale-
orange oil. 1H NMR (250 MHz, CDCl3): δ = 6.38 (s, 2 H), 3.43
(73 mg, 85%) as
a
white solid (m.p. 37–39 °C). 1H NMR
3
(br. s, 2 H, NH2), 2.49 (m, 6 H), 1.56 (m, 6 H), 1.27 (m, 54 H),
(400 MHz, CDCl3): δ = 8.20 (m, 4 H), 7.64 (d, JHH = 8.8 Hz, 2
3
3
0.88 (t, JHH = 6.5 Hz, 9 H) ppm. 13C NMR (101 MHz, CDCl3): H), 7.55 (t, JHH = 8.8 Hz, 2 H), 6.79 (s, 1 H), 2.70 (m, 2 H), 2.53
3
δ = 144.0 (s, 1 C, Cq), 142.4 (s, 1 C, Cq), 129.5 (s, 2 C, Ct), 114.5
(s, 2 C, Ct), 33.7 (s, 3 C, Cs), 32.3 (s, 3 C, Cs), 30.1 (m, 24 C, Cs),
(m, 4 H), 1.26–1.12 (m, 60 H), 0.87 (t, JHH = 6.7 Hz, 9 H) ppm.
13C NMR (101 MHz, CDCl3): δ = 148.3 (s, 2 C, Cq), 144.6 (s, 2
C, Cq), 139.4 (s, 2 C, Ct), 138.1 (s, 1 C, Cq), 135.9 (s, 2 C, Cq),
23.1 (s, 3 C, Cs), 14.5 (s, 3 C, Cp) ppm. IR (ATR, neat): ν = 3004
˜
(w), 2920 (s), 2852 (m), 1620 (w), 1466 (w), 847 (br. w), 631 (m) 130.1 (s, 2 C, Cq), 129.3 (s, 2 C, Ct), 124.7 (s, 2 C, Ct), 122.2 (s, 2
cm–1. MS (EI): m/z (%) = 597.6 (100), 598.6 (42), 599.6 (9), 442.4 C, Ct), 32.3 (s, 3 C, Cs), 30.0 (m, 27 C, Cs), 23.1 (s, 3 C, Cs), 14.5
Eur. J. Org. Chem. 2009, 2562–2575
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2573

M. Müri, K. C. Schuermann, L. De Cola, M. Mayor
FULL PAPER
(s, 3 C, Cp) ppm. IR (ATR, neat): ν = 3080 (w), 2920 (s), 2851 (m),
acknowledge Dr. Daniel Häussinger for recording the two-dimen-
sional NMR spectra.
˜
1526 (s), 1466 (m), 1346 (m), 1099 (w) cm–1. MS (EI): m/z (%) =
824.5 (100), 825.5 (55), 826.5 (17), 827.5 (4), 794.6 (18), 502.2 (22),
361.1 (17), 315.1 (10). C54H84N2O4 (825.26): calcd. C 78.59, H
10.26, N 3.39; found C 78.56, H 10.09, N 3.22.
[1] F. Cisnetti, R. Ballardini, A. Credi, M. T. Gandolfi, S. Masiero,
F. Negri, S. Pieraccini, G. P. Spada, Chem. Eur. J. 2004, 10,
2011–2021.
Macrocycle 3: A 1  lithium aluminium hydride solution in THF
(0.48 mL, 0.485 mmol) was slowly added to a solution of 24
(80 mg, 0.097 mmol) in dry THF (20 mL) under argon at room
temperature. The reaction mixture was stirred at 40 °C for 3 h and
the black suspension quenched with water (10 mL). After extrac-
tion with DCM (2ϫ15 mL), the combined organic layers were
washed with brine (20 mL), dried with magnesium sulfate and the
solvents evaporated to dryness. The crude product was purified by
size-exclusion CC (Bio-Beads® S-X Beads, toluene) to afford
macrocycle 3 (36 mg, 49%) as an orange resin (m.p. 67 –69 °C).
TLC (silica gel, DCM/hexane, 2:1): Rf = 0.92; (silica gel, DCM/
hexane, 1:1): Rf = 0.66 (due to the limited solubility of 3 in the
eluents, the spots were observed with long tails). 1H NMR
(400 MHz, CDCl3): δ = 7.91 (m, 4 H), 7.85 (m, 4 H), 7.49 (m, 4
H), 7.44 (m, 4 H), 6.96 (s, 2 H), 2.71 (m, 4 H), 2.60 (m, 8 H), 1.57
(m, 12 H), 1.41 (m, 12 H), 1.26–1.12 (m, 96 H), 0.84 (m, 18 H)
[2] Y. Norikane, K. Kitamoto, N. Tamaoki, J. Org. Chem. 2003,
68, 8291–8304.
[3] N. Tamaoki, K. Koseki, T. Yamaoka, Tetrahedron Lett. 1990,
31, 3309–3312.
[4] N. Tamaoki, T. Yamaoka, J. Chem. Soc. Perkin Trans. 2 1991,
873–878.
[5] A. Khan, C. Kaiser, S. Hecht, Angew. Chem. Int. Ed. 2006, 45,
1878–1881.
[6] V. Balzani, A. Credi, M. Venturi, Molecular Devices and Ma-
chines – A Journey into the Nanoworld, 1st ed., Wiley-VCH,
Weinheim, 2003.
[7] M. V. Peters, R. S. Stoll, A. Kuehn, S. Hecht, Angew. Chem.
Int. Ed. 2008, 47, 5968–5972.
[8] Y. Shirai, T. Sasaki, J. M. Guerrero, B.-C. Yu, P. Hodge, J. M.
Tour, ACS Nano 2008, 2, 97–106.
[9] Y. Norikane, N. Tamaoki, Eur. J. Org. Chem. 2006, 5, 1296–
1302.
ppm. 13C NMR (CDCl3, determined by HMBC and HMQC): δ = [10] V. Ferri, M. Elbing, G. Pace, M. D. Dickey, M. Zharnikov, P.
Samori, M. Mayor, M. A. Rampi, Angew. Chem. Int. Ed. 2008,
152.2 (4 C, Cq), 143.9 (4 C, Cq), 139.2 (4 C, Cq), 138.4 (4 C, Cq),
135.8 (2 C, Cq), 132.1 (4 C, Ct), 129.7 (2 C, Ct), 128.6 (4 C, Ct),
123.7 (4 C, Ct), 121.5 (4 C, Ct), 31.9 (6 C, Cs), 31.4 (6 C, Cs), 29.7
47, 3407–3409.
[11] W. R. Brode, J. H. Gould, G. M. Wyman, J. Am. Chem. Soc.
1952, 74, 4641–4646.
[12] D. L. Beveridge, H. H. Jaffe, J. Am. Chem. Soc. 1966, 88, 1948–
1953.
[13] R. Hagen, T. Bieringer, Adv. Mater. 2001, 13, 1805–1810.
[14] K. G. Yager, C. J. Barrett, Polym. Nanostruct. Their Appl. 2007,
2, 243–280.
[15] N. Landraud, J. Peretti, F. Chaput, G. Lampel, J. P. Boilot, K.
Lahlil, V. I. Safarov, Appl. Phys. Lett. 2001, 79, 4562–4564.
[16] B. Darracq, F. Chaput, K. Lahlil, Y. Levy, J.-P. Boilot, Adv.
Mater. 1998, 10, 1133–1136.
(48 C, Cs), 22.7 (6 C, Cs), 14.1 (6 C, Cp) ppm. IR (ATR, neat): ν
˜
= 3057 (w), 2953 (w), 2920 (s), 2851 (m), 1597 (w), 1458 (m), 1047
(br. s) cm–1. MS (MALDI-TOF): m/z (%) = 1523.1 (100), 1431.3
(25).
VPO Measurements: The pure compounds 1–4 were separately dis-
solved in chloroform. The apparatus was calibrated with a benzil
standard. Each sample was measured at 42 °C at least three times
and the average value calculated. All concentrations were higher
then the critical lower concentration of about 0.005 mol/kg indi-
cated for the VPO apparatus (Knauer VPO K-7000).
[17] L. A. Ingerman, M. L. Waters, J. Org. Chem. 2009, 74, 111–
117.
[18] Y. Norikane, R. Katoh, N. Tamaoki, Chem. Commun. 2008,
Determination of the Quantum Yields of Photoisomerization: Start-
ing from Equation (1), in which Φct and Φtc are the quantum yields
of the Z Ǟ E and E Ǟ Z photoisomerization reactions, f is the
optical density, N is the number of incident photons (in Einstein/
cm2 s), the factor 3 stems from the volume of the cuvette, being
3 cm3, and 1000 from the conversion of L into cm3, we can easily
obtain a differential equation that correlates the relative concentra-
tion and change thereof to the absorbed light and from that obtain
the quantum yields from the measurements.
1898–1900.
[19] C. S. Sheppard, Encycl. Polym. Sci. Eng. 1985, 2, 143–157.
[20] H. Zollinger, Diazo Chemistry I. Aromatic and Heteroaromatic
Compounds, VCH, New York, 1994.
[21] A. F. Hegarty, The Chemistry of Diazonium and Diazo Groups,
Wiley, New York, 1978.
[22] R. Brückner, Reaktionsmechanismen – Organische Raktionen,
Stereochemie, moderne Synthesemethoden, 2nd ed., Spektrum
Akademischer Verlag, Heidelberg, Berlin, 2003.
[23] M. J. Jones, Organic Chemistry, 2nd ed., W. W. Norton & Com-
pany, New York, 2000.
[24] E. Haselbach, Helv. Chim. Acta 1970, 53, 1526–1543.
[25] M. Hedayatullah, J. P. Dechatre, L. Denivelle, Tetrahedron Lett.
1975, 16, 2039–2042.
(1)
[26] B. Ortiz, P. Villanueva, F. Walls, J. Org. Chem. 1972, 37, 2748–
2750.
[27] E. Tauer, R. Machinek, Liebigs Ann. 1996, 1213–1216.
[28] G. Pace, V. Ferri, C. Grave, M. Elbing, C. von Haenisch, M.
Zharnikov, M. Mayor, M. A. Rampi, P. Samori, Proc. Natl.
Acad. Sci. USA 2007, 104, 9937–9942.
[29] H. Feuer, The Chemistry of the Nitro and Nitroso Groups, Inter-
science, New York, 1969, vol. 1.
[30] Y.-K. Lim, K.-S. Lee, C.-G. Cho, Org. Lett. 2003, 5, 979–982.
[31] H.-M. Kang, H.-Y. Kim, J.-W. Jung, C.-G. Cho, J. Org. Chem.
2007, 72, 679–682.
The ε values for the E and Z isomers used for this analysis were
calculated by correlating the height of the S1ǟS0 (π–π*) band and
the corresponding relative concentrations of the E and Z isomers
1
obtained from the H NMR analysis.
Supporting Information (see footnote on the first page of this arti-
cle): GPC charts and 1H NMR spectra of the macrocycles 1–4,
synthesis and characterization of the building block 12 and its pre-
cursors.
[32] L. Shu, M. Mayor, Chem. Commun. 2006, 4134–4136.
[33] Y. Teki, S. Miyamoto, M. Nakatsuji, Y. Miura, J. Am. Chem.
Soc. 2001, 123, 294–305.
[34] S. Höger, K. Bonrad, A. Mourran, U. Beginn, M. Moeller, J.
Am. Chem. Soc. 2001, 123, 5651–5659.
Acknowledgments
We thank the Swiss National Science foundation (SNSF) and the
Innovation Promotion Agency (CTI) for financial support. We also
2574
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 2562–2575

Shape-Switchable Azo-Macrocycles
[35] S. Tisserand, R. Baati, M. Nicolas, C. Mioskowski, J. Org.
Chem. 2004, 69, 8982–8983.
[36] A. de Meijere, F. Diederich, Metal-Catalyzed Cross-Coupling
Reactions, 1st ed., Wiley-VCH Verlag GmbH, Weinheim, 2004,
vol. 1.
[37] N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457–2483.
[38] H. Hayashi, K. Kamata, J. Abe, H. Yoshida, S. Asaoka, T.
Iyoda, Trans. Mater. Res. Soc. Jpn. 2003, 28, 569–572.
[39] Synthetic protocols for the synthesis of 12 are available in the
Supporting Information.
[43] K. H. Schuendehuette, Methoden der organischen Chemie, 1st
ed., Thieme, Stuttgart, 1965, vol. 3.
[44] M. Elbing, A. Błaszczyk, C. von Haenisch, M. Mayor, V. Ferri,
C. Grave, M. A. Rampi, G. Pace, P. Samori, A. Shaporenko,
M. Zharnikov, Adv. Funct. Mater. 2008, 18, 2972–2983.
[45] B. Priewisch, K. Rueck-Braun, J. Org. Chem. 2005, 70, 2350–
2352.
[46] S. Berger, S. Braun, 200 and More NMR Experiments, Wiley-
VCH, Weinheim, 2004, vol. 1.
[47] D. Roettger, H. Rau, J. Photochem. Photobiol., A 1996, 101,
205–214.
[40] C. Niemann, C. E. Redemann, J. Am. Chem. Soc. 1941, 63,
1549–1552.
[48] G. Zimmermann, L. Y. Chow, U. I. Paik, J. Am. Chem. Soc.
1958, 80, 3528.
[49] R. Wager, S. Berger, J. Magn. Reson. 1996, 123, 119–121.
Received: December 23, 2008
[41] F. Wurthner, Z. Chen, V. Dehm, V. Stepanenko, Chem. Com-
mun. 2006, 1188–90.
[42] F. A. Carey, R. J. Sundberg, Organische Chemie, 3rd. ed.,
Wiley-VCH, Weinheim, 1995.
Published Online: April 9, 2009
Eur. J. Org. Chem. 2009, 2562–2575
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2575