Two-dimensional oligo(phenylene-ethynylene-butadiynylene)s: All-covalent nanoscale spoked wheels

DOI: 10.1002/chem.200801939

Two-Dimensional Oligo(phenylene-ethynylene-butadiynylene)s:

All-Covalent Nanoscale Spoked Wheels

Shengbin Lei,^[a]An Ver Heyen,^[a]Steven De Feyter,*^[a]Mathieu Surin,^[b]

Roberto Lazzaroni,*^[b]Sabine Rosenfeldt,*^[c]Matthias Ballauff,^[c]Peter Lindner,^[d]

Dennis Mçssinger,^[e]and Sigurd Hçger*^[e]

Dedicated to Professor Fritz Vçgtle on the occasion of his 70th birthday

Abstract: The synthesis and characteri-

zation of a shape-persistent two-dimen-

sional (2D) organic compound is de-

scribed in detail. In a rational modular

synthesis of a dodecaacetylene precur-

sor and its subsequent template-aided

cyclization, we obtained a molecularly

defined, stable, C₆-symmetric, rigid,

spoked wheel. Peripheral tert-butyl

groups and alkyl chains attached to the

plane of the molecule provide suffi-

cient solubility, so that the 2D oligomer

can be fully characterized by MALDI-

MS, GPC, and ¹H NMR, UV/Vis ab-

sorption, and fluorescence spectrosco-

py. Molecular mechanics and dynamics

simulations indicate that the most

stable conformer of the molecule in

vacuum is a shallow boat conformation

with a small dihedral angle. Compari-

sons with the precursor as well as a

ring-only structure clearly reveal the

high rigidity of the title compound.

Small-angle neutron scattering (SANS)

experiments in [D₈]THF and CDCl₃

affirm the rigid backbone structure in

solution, that is, a radius of about

0.22 nm. STM investigations illustrate

that the wheel molecules adsorb with

their molecular plane parallel to the

surface and can form hexagonal crys-

talline domains (unit cell parameters

are a=b=6.0ꢀ0.2 nm and q=60ꢀ28),

with the tert-butyl groups on the

apexes staggered. Such staggering indu-

ces chirality in the organized domains.

AFM investigations demonstrate that

the wheel molecules inside overlayers

organize in the same way as in the

layer directly in contact with the sur-

face. This indicates an epitaxial growth

characteristic of the film.

2.7 nm and

a

thickness of about

Keywords: molecular dynamics

molecular spoked wheels · scanning

probe microscopy neutron

scattering · template synthesis

·

[a] Dr. S. Lei, Dr. A. Ver Heyen, Prof. Dr. S. De Feyter

Department of Chemistry

[d] Dr. P. Lindner

Institut Laue-Langevin, B.P. 156X

38042 Grenoble Cedex (France)

Laboratory of Photochemistry and Spectroscopy, and

Institute for Nanoscale Physics and Chemistry

Katholieke Universiteit Leuven

Celestijnenlaan 200-F, 3001 Leuven (Belgium)

Fax : (+32)16327-990

[e] D. Mçssinger, Prof. Dr. S. Hçger

Kekulꢀ-Institut fꢁr Organische Chemie und Biochemie

Universitꢂt Bonn, Gerhard-Domagk-Strasse 1

53121 Bonn (Germany)

E-mail: Steven.DeFeyter@chem.kuleuven.be

Fax : (+49)228-73-5662

[b] Dr. M. Surin, Prof. Dr. R. Lazzaroni

Laboratory for Chemistry of Novel Materials

Universitꢀ de Mons-Hainaut, Place du Parc 20

7000 Mons (Belgium)

E-mail: hoeger@uni-bonn.de

Supporting information for this article is available on the WWW

under http://dx.doi.org/10.1002/chem.200801939.

Fax : (+32)65373-861

E-mail: Roberto@averell.umh.ac.be

[c] Dr. S. Rosenfeldt, Prof. Dr. M. Ballauff

Institut fꢁr Physikalische Chemie, Universitꢂt Bayreuth

Universitꢂtsstrasse 30, 95440 Bayreuth (Germany)

Fax : (+49)921-55-2780

E-mail: Sabine.Rosenfeldt@uni-bayreuth.de

2518

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 2518 – 2535

FULL PAPER

Introduction

platelets have been analyzed in order to understand funda-

mental physical and chemical aspects of polymer–clay com-

posite materials. By nature, inorganic silicates are not com-

patible with nonpolar organic polymers and have therefore

to be processed and made “organophilic” to prevent micro-

phase separation.^[11]

Large shape-persistent molecules of defined constitution,

size, and shape have gained increasing importance in materi-

al science, both in fundamental research and various fields

of application. Precise control of the structure and definite

extensions rely on the moleculesꢄ internal order, which is

achieved by the use of rigid building blocks and well-de-

signed architecture. Apart from several supramolecular ap-

proaches that result in monodisperse organic or organome-

tallic nanostructures, most synthetic monodisperse macro-

molecules are produced in a stepwise fashion by covalent

bond-forming reactions provided by organic synthesis.^[1]Al-

though oligomer synthesis is generally more tedious than

the preparation of the corresponding polymers, linear oligo-

mers are well investigated as models for the corresponding

(polydisperse) polymers, since they can, in principle, be ob-

tained free from defects. Physical, optical, and electronic

properties can be correlated with the chain length and allow

an estimation of a specific property of the pure polymer. In

addition, thin-film deposition from the gas phase as well as

lower intrinsic viscosities of oligomer solutions have opened

up a particular field of applications for defined linear oligo-

mers, mainly in optoelectronics.^[2]

Rigid 2D organic polymers and oligomers with similar lat-

eral expansions could be employed as reinforcing compo-

nents instead of inorganic platelets, be it as model com-

pounds or even for the formulation of new composites.^[12]

2D polymers are highly intriguing systems, although only

few of them were prepared with a defined structure.^[13–14]

However, for these macromolecules defined structure does

not necessarily mean defined size and shape, some of them

behave rather like flexible membrane sheets than rigid

disks.^[15]There have also been reports on single-lamella

polyethylene crystallites, which are shape-persistent due to

intramolecular interactions,^[16]but the problem of size con-

trol remains. By contrast, rigid, defined 2D oligomers offer

several evident advantages over the inorganic counterparts.

Beside better polymer compatibility, they are well character-

izable, can be obtained in different, yet defined sizes, and

are not limited to special structural motifs. Moreover, they

allow the introduction of functionality at the perimeter as

well as above and below the molecular plane. As also neces-

sary for linear rigid oligomers, these structures require the

attachment of flexible side groups in order to render them

soluble and meltable (Figure 1).

Although the aforementioned criteria hold mostly for

linear (1D) (and rarely for cyclic) oligomers, similar synthet-

ic approaches lead to perfectly hyper-branched molecules,

which can form three-dimensional (3D) structures. In the

field of dendrimer chemistry, two synthetic strategies, con-

vergent and divergent, are routinely used for compound

preparation. Both approaches deploy multifunctional build-

ing blocks, either starting from a multifunctional core

(spherical growth) or by attaching the ready-made modules

(dendrons) to a multifunctional hub in the last step.^[3]If

rigid building blocks such as phenylene or phenylene–eth-

ACHTUNGTRENNUNG

ynylene^[4]are used, the branches are prevented from back-

folding. Therefore, stiff dendritic macromolecules are not

only defined in constitution but also in macroconforma-

tion.^[5]Beside those aspects, concerning the importance of

defined molecules, the share of aesthetics due to high sym-

metry should not be understated in this context.^[6]

Figure 1. Schematic drawing of a defined shape-persistent 2D oligomer:

A rigid scaffold in the shape of a spoked wheel and flexible side groups

attached within the molecular plane and at the periphery.

As it holds for 1D and 3D representatives mentioned

above, defined rigid two-dimensional (2D) structures are apt

for purposes that require strict size- and shape regularity.

Such 2D nano objects of defined size and shape are predom-

inantly applied in the field of surface patterning and func-

tionalization, in many cases involved with host–guest

chemistry.^[7]Further areas of interest are defined layers for

light emission and harvesting,^[8]the formation of discotic

liquid crystals^[9]and the formulation of nanocomposite ma-

terials.^[10]Intensely explored oblate objects in the nanometer

regime are disk-shaped synthetic clay platelets (e.g., Lapon-

ite RD^ꢅ, dꢁ25 nm) applied as rigid components in poly-

mer–clay nanocomposites.^[11]Layered silicate polymer fillers

enhance stiffness, strength, and thermal stability and reduce

gas and liquid permeability of the resulting composite mate-

rials.^[10]More recently, nanoscopic disk-shaped synthetic clay

Although approaches towards oblate objects in a size

range of the inorganic counterparts can be rarely found,

some defined rigid 2D structures have already been de-

scribed. Starting from the oldest carbon-based 2D functional

material, graphite itself, research focussed on processable

graphite cutouts like hexabenzocoronenes (HBCs), trinaph-

tho-HBCs, super-acenes, super-phenalenes and even C₂₂₂

-

graphene.^[17]Moreover, artificial ethynylene and butadiyny-

lene extensions of the graphite carbon network have been

investigated. Although the all-carbon networks graphyne

and graphdiyne have so far not been accomplished, a limited

number of cutouts were synthesized and investigated. Such

benzodehydroannulene systems include triangles, trefoils,

bow ties, rhombi, and other edge- or angle-fused triangles,^[18]

Chem. Eur. J. 2009, 15, 2518 – 2535

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

2519

S. Hçger, S. De Feyter, R. Lazzaroni, S. Rosenfeldt et al.

phthalocyanine- and subphthalocyaninedehydroannulenes,^[19]

as well as heterocycle-fused polycyclic aromatic hydrocar-

bons.^[20]These investigations also lead to the synthesis of

macrocyclic and spoked wheel structures, that is, macrocy-

cles with opposite apexes connected by rigid linear elements.

Beside purely covalent examples,^[21]there are also supra-

molecular spoked-wheel structures known in literature with

even larger size expansions. In those cases the noncovalent

interior scaffold is either intended as a reinforcement of a

flexible macrocycle^[22]or—more commonly—as a temporary

template in order to facilitate

Scheme 1. Covalent-template principle: hub and rim modules (A) under-

going a 1:6 coupling to form a star-shaped molecule (B) and subsequent

cyclization to a molecular spoked wheel (C).

cyclization.^[23]

We applied the benefits of

covalent template synthesis by

using a star-shaped precursor

molecule

as

a

template.

Scheme 1 presents the straight-

forward synthetic pathway to

all-covalent 2D oligo(phenyl-

ACHTUNGTRENNUNGene-ethynylene-butadiynylACHUTNGTRENeNUGN ne)s

(2D-OPEBs) relying on a high-

yield template-directed sixfold

acetylene dimerization in the

final step.^[24]Our approach

paves the way to molecularly

defined 2D structures that are

well characterizable, allow in-

troduction of functionality at

the rim as well as above and

below the molecular plane, and

can be obtained in different

sizes. Unlike the abovemen-

tioned cutouts of graphite,

graphyne, or graphdiyne, our

approach is not restricted to

structural ideals, but can be

adapted to synthetic feasibility

as well as improved solubility

and surface patterning proper-

ties. Here we describe the de-

tailed synthesis of 1 (Figure 2),

the first member of this family

of molecules, along with its

characterization. Our afore-

Figure 2. Molecular spoked wheel: structure formula of shape-persistent 2D oligomer 1.

mentioned proposal on the

shape-persistence of this 2D oli-

precursor 2 by sixfold Glaser coupling, in which the central

“hub” acts as a template that guarantees a high yield due to

the intramolecular nature of the reaction. Second, the for-

mation of 2 from two reasonable modules functioning as

“hub” and “rim” by 1:6 Sonogashira-Hagihara coupling.

Scheme 2 displays the synthesis of 1. The central “hub”

module was prepared starting from commercially available

hexaphenylbenzene. Hexabromination^[25]yielded 3, which

was radially extended with the silylated diethylphenylboron

pinacolate 4.^[26]In this sixfold Suzuki coupling reaction, we

employed an activated NHC catalyst system^[27]in anhydrous

THF and with KOH as base. If the reaction was performed

gomer is supported by molecular modeling simulations and

small-angle neutron scattering in solution. The organization

of these rigid structures at the solid/liquid and solid/air inter-

face is investigated by means of STM and AFM.

Results and Discussion

Synthesis of 2D oligo(phenylene-co-ethynylene-co-butadiy-

nylene) (1): The synthesis of 2D oligomer 1 relies on two

strategic disconnections. First, the cyclization of the open

2520

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 2518 – 2535

Template Synthesis

FULL PAPER

Scheme 2. TIPS=triisopropylsilyl, CPDMS=(3-cyanopropyl)dimethylsilyl. a) PEPPSI catalyst,^[27]KOH, THF, 2 d reflux, 35%; b) ICl, CHCl₃, 0–208C,

24 h, 91%; c) Ac₂O, 1508C, 5 h, 70%; d) [Pd

(PPh₃)₄], Cs₂CO₃, H₂O, THF, 3 d reflux, 26%; e) nBuLi, C₂H₄I₂, THF, ꢂ788C, 2 h, 20 8C; f) CPDMS acety-

lene, [PdCl₂A(PPh₃)₂], CuI, PPh₃, THF, piperidine, 208C, 24 h, 61% (2 steps); g) K₂CO₃, THF/MeOH (2:1), 208C, 3 h, 98%; h) [PdCl₂A(PPh₃)₂], CuI, PPh₃,

THF, piperidine, 208C, 4 h, 64%; i) K₂CO₃, THF/MeOH (2:1), 208C, 16 h, 91%; j) [PdCl₂A(PPh₃)₂] CuI, PPh₃, THF, piperidine, 708C, 3 d, 59%; k) excess

TBAF=[NBu₄]F, 1.0m in THF, 5% H₂O, 20 8C, 3 d, 76%; l) CuCl, CuCl₂, pyridine, 508C, 4 d, 59%.

AHCTUNGTRENNUNG

C

CHTUNGTRENNUNG

CTHUNGTRENNUNG

Chem. Eur. J. 2009, 15, 2518 – 2535

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

2521

S. Hçger, S. De Feyter, R. Lazzaroni, S. Rosenfeldt et al.

with nine equivalents of base, crude 5 was obtained in about

65% yield after stirring at 708C for 5 h. However, the

MALDI-TOF mass spectrum revealed up to 25% of four-

and fivefold coupling products, which could not be separat-

ed by column chromatography. On the other hand, if the re-

action was stirred at the same temperature for 48 h and up

to 24 equivalents of base were added during the reaction,

pure sixfold coupling product could be easily isolated in

35% yield after column chromatography (clean NMR spec-

tra, single peak in MALDI-MS). Subsequently, the hexakis-

the synthesis. Since R_fvalues on TLC were similar, we first

assumed exclusive sixfold coupling.^[32]However, the crude

product composition of 19 could be depicted by MALDI-

MS. The spectrum showed a distribution of four-, five-, and

sixfold coupling products (and their corresponding matrix

adduct peaks) of approximately 0.7:1:1. To our surprise,

after the removal of the triisopropylsilyl (TIPS) groups with

TBAF (TBAF=[NBu₄]F), the corresponding desilylated

five- (and fourfold) coupling products have larger R_fvalues,

thus 2 could be purified by radial chromatography. The four-

and fivefold coupled side products were isolated and charac-

ACHTUNGTRENNUNG

1

G

terized by MALDI-MS, GPC analysis, and H NMR spec-

Due to the twelve ethyl groups, its solubility was sufficient

for recrystallization from chloroform. It is worth mentioning

that in preliminary experiments the unsubstituted analogue

of 5 could neither be properly characterized nor purified.

However, purity is a prerequisite for the success of the fol-

lowing reaction sequence.

troscopy.

This last method allows an easy distinction of the different

species. As depicted in Figure 3, the two signals for the

methylene moieties of the hub ethyl groups (d=2.67 and

2.31 ppm in 6) shift downfield upon the coupling reaction

Pyrylium salt 7 was used as precursor for the corner piece

9 of the “rim” module.^[28]Condensation with ten equivalents

of sodium (4-bromophenyl)acetate 8 in acetic anhydride

gave the bromodiiodo compound 9 in 70% yield.^[29]In the

next step we utilized the iodo selectivity of the Suzuki con-

densation protocol for the elongation of the m-terphenyl

arms with two equivalents of pinacol boronic ester 10. How-

ever, chromatographic separation of compound 11 from its

byproducts was tedious, resulting in mediocre yields. The

bromo function of 11 was transformed into the more reac-

tive iodide by lithiation with n-butyllithium and addition of

diiodoethane. The crude iodination product 12 was coupled

with (3-cyanopropyl)dimethylysilyl acetylene (CPDMS acet-

ylene, 13) providing 14. Generally speaking, the CPDMS

group is a polar analogue of the trimethylsilyl group, is read-

ily available, and dramatically simplifies the chromatograph-

ic separation of coupling products, starting material, and de-

halogenated side products.^[30]Base-catalyzed deprotection of

14 gave the free acetylene 15, which was coupled with 16.

This building block appreciably enhanced the solubility of

the “rim” module as well as of the final 2D oligomer 1.

Here again, purification of 17 by column chromatography

was straightforward even on a hundred-milligram scale due

to the polar CPDMS group. After base-catalysed removal of

the CPDMS group, 18 could be also obtained in high purity.

As confirmed by GPC analysis, the consecutive polarity

change of the products in the reaction sequence ensured

that the ready “rim” module was not contaminated even

with traces of side product, an absolute requirement for the

next severalfold coupling steps.^[31]

Figure 3. ¹H NMR spectra (left) and cartoons (right) of the separated 1:6

(top), 1:5 (middle) and 1:4 (bottom) Sonogashira coupling products oc-

curring in the reaction towards 2. During the coupling reaction the ethyl

CH₂signals of 6 (2.67 and 2.31 ppm) shift to 2.85 and 2.39 ppm (b and c),

respectively. The terminal acetylene signal (a, 3.16 ppm) is split into two

or three signals, depending on the number of neighboring vacancies the

acetylene protons “sense”. Arrows indicate signals not present in the

spectrum of the desired 1:6 coupling product.

(2.85 and 2.39 ppm in 2). While the desired sixfold coupling

product 2 merely shows shifted signals (Figure 3 top, b and

c), integration of shifted and non-shifted signals gives a 5:1

ratio for the fivefold- (middle) and a 4:2 ratio for the four-

fold-coupled species (bottom). Surprisingly, the terminal

acetylene protons (solid circles in the cartoons) turn out to

be very sensitive to their chemical environment, so they

make an even better monitor. Given unhindered rotation

around the spokes, all alkyne protons in the 1:6 coupling

product experience the same environment, resulting in one

singlet signal “a” (d=3.16 ppm). In the 1:5 species, there

are four acetylene protons (rotation!) close-by to a vacancy,

while only six protons are adjacent to other “corner pieces”

as in the 1:6 coupling product. This results in a shifted

second signal. The situation becomes even more complex

The assembly of the “hub” and “rim” subunits had to be

achieved in a sixfold Sonogashira–Hagihara coupling under

rather harsh conditions. In first experiments, the hexaiodide

6 was coupled with nine equivalents of 18 at 20–508C and

5 mol% Pd⁰catalyst per reactive site (i.e. 0.3 equivalents

based on 6) for two days. While the homocoupled side prod-

uct (diacetylene) could be removed from the mixture by

silica gel chromatography, the separation of four-, five-, and

sixfold coupling products was not possible at this point of

2522

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 2518 – 2535

Template Synthesis

FULL PAPER

for the 1:4 species, that is, with two vacant sites. Depending

on the position of the vacancies (three permutations, see

cartoons in Figure 3), the alkyne hydrogen atoms can expe-

rience two, one, or no neighboring vacancies at all, resulting

in three distinct singlets. Their intensity distribution can be

approximated by simple statistics that support the above in-

terpretation.^[33]In successive coupling experiments we were

able to optimize the sixfold Sonogashira reaction towards

19. Doubling the catalyst load (0.6 equiv), increasing the

overall concentration of the reaction partners and stirring at

708C for three days gave good yields of the sixfold coupling

product (70% after desilylation towards 2).

The attachment of the six corner pieces to the central hub

leads to a high local acetylene concentration. When 2 was

added slowly to a suspension of CuCl/CuCl₂(10:1, 500 mo-

lar equiv of CuCl) in pyridine, an oxidative acetylene cou-

pling took place.^[34]Due to the linkage of the acetylenes at

the hub (“covalent template”), the intramolecular coupling

is strongly favored over the intermolecular reaction.^[30b,35]

However, GPC analysis of the crude product reveals that

the reaction does not run completely intramolecular at

room temperature and the formation of the dimer of 1

could be observed (ca. 16000 gmol^ꢂ1, Figure 4). The mono-

Figure 4 also displays the molecular weight distributions

of pure 1 (black curve) and the open precursor molecule 2

(gray curve). Although both species only differ by about 12

units in molecular mass, a clear shift of the maxima can be

observed. This nicely illustrates the decrease of the hydrody-

namic volume: in the course of cyclization the star-shaped

precursorꢄs twigs become coplanar and lose their rotational

degrees of freedom around the spoke axes. According to

GPC interpretation in terms of molecular weight, 1 and 2

should differ by approximately 500 gmol^ꢂ1. Evidently this

difference is a gross overestimation; we also have to be cau-

tious with absolute molecular mass values obtained from

GPC analysis as we used common polystyrene calibration as

a reference. However, PPE and PS fundamentally differ in

rigidity, as expressed by the persistence length.^[37]Previously,

linear PPEs have been examined regarding the overestima-

tion of their molar mass by established methods such as

GPC and end group analysis.^[38]Several groups were in

search of an empirical PPE–PS conversion factor, which

turned out to depend on sample properties and determina-

tion method and hence spread from 1.5 to 3. Astonishingly,

on our 2D-OPEB system, the molecular weight measured

by GPC in THF matches better with the absolute molecular

mass than for linear PPEs. We find a molar mass overesti-

mation of less than 10%, that is, our conversion factor is

smaller than 1.1. In the case of 1, the second dimension ap-

parently compensates the difference in persistence length,

so that the molecule has only a slightly larger hydrodynamic

volume than a 3D polystyrene random coil of the same mo-

lecular mass.^[39]

By far the most powerful method to prove complete cycli-

zation towards 1 is MALDI-TOF mass spectrometry. The

mass difference between 1 and 2 of about 12 amu is clearly

resolved in the mass spectra as shown in Figure 5 (non high-

resolution spectra). In addition to matrix adduct peaks that

are shifted by 250 and 500 m/z units, the spectra display dis-

tinct molecule peaks at m/z 8779.5 (2, calcd.: 8779.7) and

Figure 4. GPC analysis of 2D oligomer 1 (black) and open precursor 2

(gray) versus PS standard in THF. The dotted line shows a typical crude

product composition with dimer signal around 16000 gmol^ꢂ1. Intensity

signals were obtained from UV absorption at l=254 nm.

mer–dimer ratio estimated from GPC data of the crude cyc-

lization product was 5:1 (UV detector signals, relative ab-

sorption at l=254 nm). As we have shown previously on

other systems,^[36]slightly elevated reaction temperatures

(508C) favor the intramolecular reaction resulting in a prod-

uct distribution of about 12:1 under otherwise identical con-

ditions. Assuming that the extinction coefficient of the side

product is twice as large as that of 1, less than 5% of dimer

was produced under these conditions. Purification of 1 by

column chromatography (to remove low molecular weight

impurities) and subsequent purification by preparative GPC

yielded 1 in 59%.

Figure 5. Cut-outs of MALDI-TOF mass spectra of 2D oligomer 1 and

open precursor molecule 2 (m/z values indicated). Minor peaks at higher

m/z ratios originate from matrix adducts (DCTB, 250.34 gmol^ꢂ1). See

Supporting Information for full spectra.

Chem. Eur. J. 2009, 15, 2518 – 2535

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

2523

S. Hçger, S. De Feyter, R. Lazzaroni, S. Rosenfeldt et al.

m/z 8767.2 (1, calcd.: 8767.6). The difference between the

two distribution maxima of 12.3 (ꢀ 0.5) m/z units indicates

the loss of twelve acetylene protons in the course of cycliza-

tion. Minor peaks in the mass spectrum of 1, which are not

shown in Figure 5, originate from doubly charged molecules

(m/z 4383.9) as well as dimers and trimers (see Supporting

Information).^[40]

1 were performed on a

500 MHz spectrometer (Figure 6). At ambient temperature

we obtained a spectrum of broadened signals, even in good

High-temperature experiments in TCE were carried out

in order to overcome these rotation barriers. Measurements

of 1 at 908C gave a resolution comparable to the room-tem-

perature experiments of 2, that is, the doublet signals of p-

phenylene units a and d were merged. Good resolution of

all signals was achieved at 1208C (Figure 6 middle). Though

we observed a temperature shift of the solvent signal, we set

it as reference in both spectra. For that reason the spectra at

25 and 1208C also vary in chemical shift. By comparison

with NMR measurements of several precursor molecules—

including DEPT and 2D techniques—we were able to assign

the signals. The aromatic region of the high-temperature

spectrum consists of only 14 partly overlapping signal

groups, which correspond to different phenylene moieties

(Figure 6 top, a–h). Actually the molecule bears 204 aromat-

ic hydrogen atoms, so this feature reflects its high symmetry,

as expected for a perfect wheel structure. Furthermore, the

precursor characteristic signal at d=3.2 ppm (terminal

alkyne units) is absent in the full spectrum of 1 (Supporting

Information). Both results support our assumption that the

cyclization is complete. We decided against measuring

¹³C NMR spectra at 1208C for reasons of low sensitivity at

elevated temperatures, limited substance quantity, and a re-

sulting long-term high-temperature strain of the measuring

head.

¹H NMR measurements of

The 2D oligomer 1 is a stable, slightly yellow solid that

does not melt below 2508C (decomp). Due to solubility-en-

hancing alkyl groups at the rim and inside the molecular

plane, the molecule can be handled and processed in tolu-

ene (solubility ꢁ0.1 mgmL^ꢂ1), THF (ꢁ1.5 mgmL^ꢂ1),

chloroform (ꢁ7.5 mgmL^ꢂ1), and tetrachlororethane (TCE;

>15 mgmL^ꢂ1). As illustrated in Figure 7 for measurements

in dichloromethane, 2D oligomer 1 displays a broad and

structured absorption band stretching from 290 nm to

about 415 nm. Beside two maxima at 324 nm (e_l

ꢁ764000 Lmol^ꢂ1cm^ꢂ1)

and

347 nm

(e_l

Figure 6. ¹H NMR spectra of 1 showing the aromatic region at 258C and

1208C. The corresponding assignment of signals is illustrated on top. In

both spectra deuterated TCE was used as solvent and reference peak

(d=6.00 ppm). See Supporting Information for full spectra.

solvents such as deuterated TCE (Figure 6 bottom). This

finding is good evidence for the 2D oligomer rigidity and

can be ascribed most probably to hindered rotation of the p-

phenylene units, especially the units close to the hub and

the corners of the wheel (signals a and d, respectively).^[41]To

a smaller extent, this effect can already be observed in the

spectrum of 2 (Supporting Information), although the mole-

cule conformational strain is expected to be far less (see mo-

lecular modeling).

Figure 7. Normalized UV/Vis absorption (full lines) and emission

(dashed lines) spectra of 1 (black curves) and 2 (gray curves) measured

in CH₂Cl₂. The emission spectra for

1 (l_exc=350 nm) and 2 (l_exc=

379 nm) are very similar in shape and spectral range.

2524

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 2518 – 2535

Template Synthesis

FULL PAPER

ꢁ729000 Lmol^ꢂ1cm^ꢂ1), there are two shoulders at larger

wavelengths. Oligomers 1 and 2 can be easily distinguished

1-2) around 128 (5 and 2 are the “bow” and “stern” atoms,

to remain in the marine lexicon). This dihedral angle is com-

parable to the boat conformations of other unsubstituted

phenylene–ethynylene macrocycles.^[35g]Notably for the

wheel, we find no stable chair conformation, while it is the

most stable conformer for an unsubstituted phenylene–ethy-

nylene macrocycle.^[44]The size of the hexagonal frame of the

wheel imposes that the spokes are not strictly linear, be-

cause they are slightly too long for that frame in their fully-

extended linear conformation. In Figure 9, we compare the

behavior of the entire wheel with that of the open precursor,

that is, the same wheel “cut” in the middle of the hexagon

segments (Figure 8, left). Starting from a relatively planar

conformation, the spokes are revealed to be much more

flexible than the wheel. This is illustrated in Figure 9, show-

ing the evolution of geometric features for the wagon wheel

1 (blue line) and the open precursor (red line). In the

former, the distance between opposite atoms (e.g. atoms 2

and 5) remains constant, between 48 and 50 ꢆ, while for the

latter this distance shows large variations between 50 ꢆ and

36 ꢆ. This is also found for the angle 1-7-4, showing large

variations in the case of the precursor (between 1208 and

1808), while in the wheel it only evolves from boat (1608) to

planar (1808) conformations. This difference in flexibility

can also be illustrated by the self-diffusion coefficient^[45](D)

for all the atoms during the MD: D=0.5ꢇ10^ꢂ3ꢆ²psatom^ꢂ1

for the wheel versus D=0.3ꢇ10^ꢂ1ꢆ²psatom^ꢂ1for the pre-

cursor. This difference of two orders of magnitude clearly

indicates that the wheel is very rigid compared to the open

precursor.

by their absorption spectra, but upon excitation (1: l_exc

=

350 nm, 2: l_exc=379 nm) both species exhibit a violet–blue

fluorescence that originates from a relatively sharp emission

band at 417 nm and vibronic progression. While the absorp-

tion spectra of 1 and 2 are merely comparable, the emission

spectra are nearly identical. Hence, the chromophore from

which radiant relaxation emanates must already be present

in 2, that is, before cyclization.^[42]

Molecular modeling: Molecular mecahnics (MM) and mo-

lecular dynamics (MD) simulations can provide information

on the individual molecule level, such as conformation anal-

ysis and bond strain. Extensive simulations have been car-

ried out for shape-persistent macrocycles based on phenyl-

ene, phenylene–ethynylene, or phenylene–butadiynylene

backbones.^[35g]These calculations indicate that, despite the

rigid building blocks, these macrocycles adopt various mac-

roconformations and are merely shape-persistent in the

sense, that they do not fold or collapse.^[43]

To focus on the structure of the p-conjugated wheel (hub,

spokes, and rim) the long hexadecyloxy chains on the

spokes were replaced by methoxy groups. The MD run re-

veals a rigid structure, with no change in the hexagonal

shape of the molecule (Figure 8, right). The distance be-

tween opposite atoms 1–4 (or 2–5 or 3–6; Figure 8, left) only

slightly fluctuates between 48 and 50 ꢆ during the MD run,

and the distance between two opposite, most far apart hy-

drogen atoms of the tert-butyl groups is on average 68 ꢆ.

The molecule shape can fluctuate close to the fully planar

conformation, the angle (atoms 1-7-4) typically evolves be-

tween 1808 and 1608, that is, between the planar and a shal-

low boat conformation. The most stable conformations are

those showing a boat-type structure (Figure 8, right), but the

energy difference between the boat conformation and the

planar conformation is small, that is, around 2.3 kcalmol^ꢂ1.

This conformation shows a small dihedral angle (atoms 5-6-

MD of a hypothetical ring structure (Figure 8, right) re-

veals a relatively flexible structure. The self-diffusion coeffi-

cient during the MD run is D=0.1ꢇ10^ꢂ1ꢆ²psatom^ꢂ1, to be

compared with 0.5ꢇ10^ꢂ3ꢆ²psatom^ꢂ1for the wheel and

0.3ꢇ10^ꢂ1ꢆ²psatom^ꢂ1for the open precursor. The “flexibili-

ty” of the ring is therefore comparable to that of the open

precursor, while for the wheel the spokes maintain a rigid

conformation. This is also illustrated in Figure 9, showing

Figure 8. Left: Structure formulas of the three models used for MD simulations. The “wheel” (blue) corresponds to 1 bearing methoxy instead of hexade-

cyloxy groups, while the “open precursor” is derived from 2. The “ring” represents a hypothetical macrocyclic compound consisting of the same phenyl-

ene–butadiynylene frame as 1. Right: CPK models of the boat conformer of the wheel, with the hexagonal rim shown parallel (top) and perpendicular

(bottom) to the view.

Chem. Eur. J. 2009, 15, 2518 – 2535

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

2525

S. Hçger, S. De Feyter, R. Lazzaroni, S. Rosenfeldt et al.

major role in the formation of a stable rigid conformation.

The high flexibilities of the open precursor and the hypo-

thetical ring are not surprising if we take into account the

compoundsꢄ size and the relatively small persistence length

of phenylene–ethynylene structures in general.^[37]Even

single-crystal X-ray structures of phenylene–acetylene rings

exhibit considerable deviations from linear phenylene–acet-

ylene bonds.^[46]

Small-angle neutron scattering (SANS) investigations: To

get an experimental insight into the structure of the com-

pound 1, SANS investigations were performed. Coherent

elastic neutron scattering provides information on the spa-

tial structure of dissolved species with a typical size range of

1–200 nm.^[47]Molecules that locally vary in scattering length

density can be investigated in different solvents or solvent

mixtures, resulting in different scattering contrasts.^[47a,48]

SANS data taken at high contrast between the solvent and

the solute can be used to explore the shape of the entire

molecule. At intermediate contrast the details of the mole-

cule can be detected. Scattering intensities can be fitted by

applying models with appropriate geometry and scattering

length density distribution.^{[11b,16,47,49]}

Small-angle neutron scattering was performed on several

samples of 1 in CDCl₃and in a contrast series of THF/

[D₈]THF mixtures. As a result of the limited solubility of 1,

the investigations were restricted to low concentrations. For

the evaluation, any scattering contribution due to interpar-

ticular interaction was neglected; hence, any concentration

influence was disregarded. Due to the restricted resolution

of SANS, the experimental data are described by using the

simple model of a stack of three disks (“hamburger model”,

Figure 10).^[16]

The inner disk corresponds to the backbone that deter-

mines the molecule radius. The outer dimension of the ham-

burger is given by the overall thickness of the attached alkyl

chains. At high contrast, neutron scattering is most sensitive

to the outer dimensions (the radius R and the thickness L1).

The best description of the experimental data was obtained

using a radius of 2.7 nm. However, the overall thickness of

the molecule is governed by the hexadecyloxy chains, the

conformation of which depends on the solvent: we found

L1=2.5 nm in CDCl₃and L1=1.2 nm in [D₈]THF. This dif-

Figure 9. Graphs showing the evolution of geometrical features (see la-

beled atoms in Figure 8, left) during the MD run. Blue line: wheel 1; red

line: open precursor 2; green line: ring.

the evolution of the distance between opposite atoms (e.g. 2

and 5). For the ring (green line) that distance shows large

variations, from 35 to almost 55 ꢆ. The energy-minimized

structures of several conformations extracted from the MD

show relatively planar conformations. Boat and fully planar

hexagonal conformations are two stable conformers, with a

very small energy difference (few tenths of kcalmol^ꢂ1). The

boat conformation is very simi-

lar to that of the wheel: the dis-

tance between atoms 2 and 5

(between bow and stern atoms)

is around 49 ꢆ and the dihedral

angle (5-6-1-2) is around 78.

This is similar to the behavior

of the wheel described above.

Therefore, the ring, though

being much more flexible than

Figure 10. Schematic drawing of 2D oligomer 1 illustrating the rigid backbone with orthogonal alkyl chains

(left). The molecule can be modelled as “hamburger”, that is, a hard disk between two soft layers (right), since

the rigid backbone and the orthogonal alkyl chains inside the plane differ in scattering length density. Varying

the ratio of protonated to deuterated solvent changes the scattering contrast of the hamburger. The hamburger

model requires only three parameters: radius R, overall thickness L1 and rigid disk thickness L2.

the spoked wheel, exhibits the

same stable conformers. This

suggests that the phenylene–bu-

tadiynylene frame plays

a

2526

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 2518 – 2535

Template Synthesis

FULL PAPER

ference can be ascribed to the solubility of the side chains in

the different solvents. At intermediate contrast, the scatter-

ing intensity is also sensitive to the variation of the scatter-

ing length density distribution. Thus, by fitting different con-

trasts, the thickness of the backbone L2 is available. From

the contrast variation and model calculations using the ex-

perimental contrasts, the radius of 2.7 nm, and the deter-

mined overall dimension L1 gave a thickness of the inner

disk of approximately 0.22 nm, underlining the rigidity of

the backbone (see Supporting Information). In Figure 11,

priate method to investigate the particle morphology in so-

lution and to prove the rigidity of the backbone experimen-

tally. The conformation of the hexadeyloxy chains

determines the thickness of the soft shell and therefore the

overall dimension of the molecule in solution.

Scanning tunelling microscopy (STM) and atomic force mi-

croscopy (AFM) investigations: To gain insight into the ad-

sorption bahavior of 1 on atomically flat solid surfaces, STM

is an extremely powerful tool, either under ambient condi-

tions or at a liquid–solid interface.^[52]At present, interest is

directed towards the development of an understanding of

the dynamic and static problems associated with the forma-

tion of 2D crystals of these large and well-defined mole-

cules. This aspect relates to the wide area of epitaxial

growth, with all its implications for material science. Infor-

mation on 3D multilayer formation, be it random stacking

or crystal growth, can be obtained by AFM investigations.

This method does not depend on substrate or surface con-

ductivity, which makes it suitable to study the topology of

blank and patterned surfaces.

Molecule 1 can form close-packed 2D assemblies both at

the air–solid (deposited from toluene) or liquid–solid (de-

posited from octanoic acid) interface. The structure of the

assemblies is identical in both environments, indicating that

the solvent has a negligible influence on the assembling of

this spoked wheel molecule. The ordered 2D assembly could

only be imaged at large tip–sample distances (high bias volt-

age and low current) and was easily destroyed with de-

creased tip–sample distance. This highlights the weak inter-

action between molecule and substrate. Figure 12a–c shows

a well-ordered 2D crystalline domain formed by 1 at the oc-

tanoic acid/graphite interface. In high-resolution images, the

internal structure of the molecule, the hub, spoke, and rim

subunits could be clearly revealed. The spokes normally

appear with a much higher contrast than the rim. The clear

sixfold symmetry and 5.7 nm distance between two opposite

apexes matches well the optimized molecular model of a

molecule lying flat on the surface. The six edges as well as

the spokes appear with equal contrast in the STM image, in-

dicating that the molecule is adsorbed with its molecular

plane parallel to the surface. The STM observations provide

clear evidence for the rigidity and shape-persistence of this

giant molecular spoked wheel. The spokes play a crucial

role in maintaining the rigidity of the monolayer. Interest-

ingly, sometimes a different contrast could be observed with

the hub and six apexes appearing brighter than the spoke

and rim subunits (Figure 12c). In this case, a cross composed

by four bright spots is revealed for each apex. These bright

spots could possibly be attributed to the phenylene rings

connected to the apex.

Figure 11. Measured coherent scattering intensities at highest contrast

(symbols) and theoretical scattering intensities based on the hamburger

model (lines) of two samples of 1 in [D₈]THF (circles) and CDCl₃

(squares). Fit parameters in [D₈]THF: R=2.7 nm, L1=1.2 nm, L2=

0.22 nm. Fit parameters in CDCl₃: R=2.7 nm, L1=2.5 nm, L2=0.22 nm.

the scattering intensities of the 2D oligomer at highest con-

trast are compared with theoretical intensities. For the

sample in [D₈]THF the upturn of the scattering intensities at

scattering vectors q<0.35 nm^ꢂ1is attributed to a small frac-

tion of aggregates.^[16,50]The coherent scattering intensities

could be successfully fitted by theoretical scattering curves

obtained from the hamburger model.

Evidently, the radius found in solution compares well to

the value obtained from STM investigations (see below).

Furthermore, the small thickness found for the central disk

implies a rigid framework with a planar geometry. Consider-

ing that 1) the shallow boat conformation shown in Figure 8

has an effective atomic-density-weighted thickness of

0.33 nm,^[51]2) the actual structure is expected to fluctuate

between the boat and the fully planar conformation due to

the very small energy difference (2.3 kcalmol^ꢂ1), and 3)

SANS probes the position of the nuclei, while the molecular

representation in MD simulation comprises the electronic

cloud, the results of the simulations and the SANS data are

reasonably consistent with each other. Consequently, the 2D

oligomer 1 morphology can be well described by the ham-

burger model. Due to SANS evaluation, the first informa-

tion concerning the structure of the spoked wheel 1 in solu-

tion is now available. We can state that SANS is an appro-

The unit cell contains one molecule (Figure 12a). Using

the graphite lattice as a calibration grid, the unit cell param-

eters of the ordered domains could be determined as a=b=

6.0ꢀ0.2 nm and q=60ꢀ28, which indicates C₆symmetry.

On the basis of these parameters, a molecular model was

built reflecting the packing of 1 in ordered domains. In the

Chem. Eur. J. 2009, 15, 2518 – 2535

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

2527

S. Hçger, S. De Feyter, R. Lazzaroni, S. Rosenfeldt et al.

In the lower part of the image isolated giant wheel mole-

cules are located randomly on top of the bottom crystalline

layer, while in the upper part of the image the molecules in

the top layer also form crystalline domains. Those mole-

cules, which are also resolved with submolecular resolution,

sit exactly on top of the molecules in the bottom layer with-

out any significant shift. In contrast, the isolated giant wheel

molecules in the lower part of the image appear more like

an “O” ring: only the rim is clearly visible, while the spoke

and hub units are hardly visible. The sixfold symmetry is not

pronounced. This suggests that the isolated giant wheel mol-

ecules are more mobile in comparison with those forming a

2D lattice. Nevertheless, since some congruent stacking

occurs, the multilayer formation reveals the possibility of

the giant wheel molecule to form 3D crystals at the interface

through epitaxial growth. However, the role in the assembly

process and the actual position of the alkyl chains in the

crystal remains unclear. There are two possibilities: the

alkyl chains fill the voids between the spokes of their own

wheel or extend out of the plane and point into the voids of

neighboring wheels. The STM contrast is known to be a

combination of both topographic and electronic contribu-

tions,^[53]and thus the height measured by STM is not reli-

Figure 12. STM images of the 2D organization of 1 at the octanoic acid/

graphite interface. a) A unit cell is indicated in white. b) and c) reveal a

different contrast of the molecule in the 2D lattice. Tunneling parame-

ters: a) V_bias=ꢂ1.50 V, I_set=10 pA, b) and c) V_bias=ꢂ1.75 V, I_set=16 pA.

d) STM image revealing clearly the formation of a multilayer structure of

1 (V_bias=ꢂ1.50 V, I_set=11 pA). In the upper part of the image an ordered

multilayer is observed, while in the lower part the molecules in the top

layer are disordered. e) A tentative model reflecting the random posi-

tion/orientation of molecules on top of a well-ordered bottom layer.

AHCTUNGTREGaNNUN ble to distinguish between those two cases. For this purpose

AFM is a better tool.

Figure 13 shows typical AFM topography and phase

images obtained on the epitaxial crystalline film of 1 on the

graphite surface. The topography image indicates that 1

model, the giant wheel molecules are aligned side-by-side

with their molecular plane parallel to the surface. Adjacent

molecules are slightly shifted to

allow staggering of the tert-

butyl groups. This results in an

intermolecular

distance

of

6.0 nm, in agreement with that

determined experimentally, but

slightly smaller than the 6.8 nm

distance between two farthest

apart hydrogen atoms deter-

mined by molecular modeling.

Therefore, the wheels must

pack tightly, with staggered tert-

butyl groups as deduced from

the STM data. By imaging the

graphite lattice underneath, the

epitaxial orientation of these

chiral domains could be deter-

mined. Our observation indi-

cates that the unit cell vectors

of the chiral domain are orient-

ed either clockwise or counter-

clockwise by 18ꢀ38 with re-

Figure 13. a) Large-scale and b) high-resolution AFM topography and phase images (c, corresponding to b) of

the as-prepared crystalline film of the giant molecular wheel, revealing the 3D crystallization on the graphite

surface. Domains of different orientations could be clearly revealed and the angle between these domains is

determined to be 308. d) Large-scale and e) and f) high-resolution AFM images of the film after annealing at

508C for 45 min. Height profiles in the topography images allow the determination of the film and terraces

thickness. The total thickness of the film is about 7 nm, corresponding to more than ten layers. The phase

image in f) clearly demonstrates the single-crystal nature of the area probed, which is confirmed by the Fourier

transform shown in g). h) Zoom of the area marked in f) with the white rectangle, the same area is also

marked in e). The two white lines indicate the orientation of the molecular rows in the top layer and the layer

underneath: the molecules in the top layer sit on top of the molecules in the second layer without any notice-

¯

spect to the h1120i direction of

graphite (see Supporting Infor-

mation).

Occasionally,

multilayers

could be revealed by STM. Fig-

ure 12d shows a typical image

of such a multilayer structure. able shift.

2528

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 2518 – 2535

Template Synthesis

FULL PAPER

forms a layered structure. However, in some parts of the

film, triangular shapes are revealed (Figure 13a), in line with

the threefold symmetry of the crystalline film. The molecu-

larly resolved phase and topography images highlight the

crystalline characteristics of the film: each molecule appears

as a disk with a diameter of about 6 nm, which is in line

with the model that the giant wheel molecules are lying with

their molecular plane parallel to the surface, in agreement

with STM results. The films are polycrystalline in nature

with domain sizes ranging from several tens to hundreds of

nanometers. From the analysis of the high-resolution topog-

raphy images, the molecular height is estimated to range

from 0.5 to 0.7 nm (see also Figure 13d–h). Considering the

thickness of the aromatic rigid backbone (ꢁ0.3 nm) and the

alkyl chains (ꢁ0.4 nm), the 0.5 to 0.7 nm thickness of the

molecular layer indicates that some alkyl chains stay within

or close to the molecular plane. More probably they fill the

voids or lie parallel on top of the backbone, rather than ex-

tending away from the plane. As discussed before, SANS

characterization indicates that the conformation of alkyl

chains depends heavily on the solvent (L1=2.5 nm in

chloroform, L1=1.2 nm in THF). Thus the 0.5 to 0.7 nm

thickness seems in line with the SANS data. Considering

there is no solvent anymore in the crystalline layer, an even

smaller L1 value is expected in comparison with 1 in THF.

The domain symmetry directions do not always follow the

three main symmetry directions of the substrate: domains

are often rotated with respect to each other by about 308,

which is in line with the STM observation of chiral domains

within the monolayers. The existence of chiral domains in

the upper layers highlights the epitaxial growth characteris-

tics of the film: the in-plane ordering in 3D crystals is identi-

cal to and directed by the 2D monolayer assembly. Anneal-

ing of the film at 508C for about 45 min significantly in-

creases the domain size. Single crystalline domains of more

than 500 nm are commonly observed. The height differences

between terraces, as revealed by height profiles is typically

1.1 and 1.6 nm, so approximately two and three times the

height of an individual molecule, indicating the tendency of

the giant wheel to form double and triple layers. The triple

layer always forms large islands (Figure 13d), while the

double layer structure most frequently appears as small is-

lands on top of the triple layer. Occasionally, at the site at

which a double layer joins a triple layer island, single molec-

ular steps are observed with a height difference around

0.5 nm (Figure 13d). Figure 13e and f show molecular reso-

lution topography and phase images. The unit cell parame-

ters (a=b=6.0ꢀ0.2 nm, and q=60ꢀ28) are in good agree-

ment with the STM and SANS results, again indicating that

the structure of 2D giant wheel layers (in the bulk) is inde-

pendent of the film thickness.

phenylene–ethynylene macrocycles.^[44]This has also been

confirmed by our STM observations, in which the six rim

segments as well as the spokes show equal contrast in the

image. However, it is not clear whether the long alkoxy

groups on the spokes (OC₁₆H₃₃) will entirely adsorb on the

surface, within a triangle formed by two spokes and a seg-

ment of the rim (e.g., triangle formed by atoms 1, 2, 7 in

Figure 8, left). Calculations show that, due to the restricted

dihedral angles in alkoxy chains, it is not possible to have all

the methylene groups of two OC₁₆H₃₃chains in the plane of

the cavity. Figure 14 illustrates that the available van der

Figure 14. Model showing the van der Waals volume of atoms (dotted

spheres) in a triangle formed between two spokes, with n-octyloxy groups

at the substitution position.

Waals surface is already filled by two octyloxy (OC₈H₁₇)

groups (note that one OC₉H₁₉plus one OC₈H₁₇group can

still fit in this cavity). This shows that at least half of the

whole hexadecyloxy chain is not adsorbed on graphite in the

plane of the cavity, but either on top of the wheel and/or at

the periphery, outside of the rim. In a multilayer film, in

which the molecular planes are parallel to the surface, both

arrangements would increase the stacking distance between

the wheels compared to a typical p-stacking distance. If part

of the alkoxy groups were located on top of the wheel, con-

sidering the van der Waals radius of a methylene group, that

is, 1.85 ꢆ, this would lead to an increased thickness of at

most 0.3–0.4 nm. This is consistent with the fact that the

thickness of the layers, as observed by AFM on multilayer

films, ranges between 0.5–0.7 nm, slightly larger than the

typical p-stacking distance (0.4 nm).

Thus, the 2D assembly is stabilized solely by the stagger-

ing of the tert-butyl groups and the interaction with the sub-

strate. This implies that the intermolecular interaction is

weak, which has been confirmed both by the STM and

AFM observations. The assembly is easily affected by the

tip scanning. Nevertheless, the molecule is so big that its dif-

fusion on the surface is quite slow at room temperature, and

even isolated molecules could be observed by STM with

reasonable resolution.^[24]

From the molecular modeling simulations we know that

the dihedral angle of the boat conformer is rather small.

Moreover, the energy difference between the planar and

boat conformation is around 2.3 kcalmol^ꢂ1. It is very likely

that the wheel planarizes when adsorbing on a graphite sur-

face to maximize the p-stacking with the surface, as in other

Chem. Eur. J. 2009, 15, 2518 – 2535

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

2529

S. Hçger, S. De Feyter, R. Lazzaroni, S. Rosenfeldt et al.

Conclusion

molecules inside the top layer organize in the same way as

in the layer directly in contact with the surface. This indi-

cates an epitaxial growth characteristic of the film.

In summary, we described a representative of a novel family

of shape-persistent two-dimensional oligomers in detail. In a

rational modular synthesis and subsequent template-aided

cyclization, we obtained a molecularly defined, stable,

highly symmetrical, rigid, wheel-shaped compound. Periph-

eral tert-butyl groups and alkyl chains attached to the planes

of the molecule provide sufficient solubility such that the

compound can be fully characterized. We were able to ac-

complish the critical steps with respect to the covalent-tem-

plate synthesis principle, such as sixfold Suzuki and Sonoga-

shira–Hagihara couplings and the final Glaser–Eglinton cou-

pling in good yields. Molecular masses of the target mole-

cule and its free acetylene precursor could be determined by

MALDI-MS with perfect match. Other than for linear

OPEs, polystyrene calibrated GPC analysis in THF gave a

good estimation of the molecular weight. As expected, the

molecule conformational constraint observed in ¹H NMR

spectra at ambient temperature could be overcome by heat-

ing. At 1208C a good resolution is obtained, as illustrated

for the aromatic region.

Since the covalent-template synthesis principles have now

been verified, the synthetic scope of the 2D oligomer at

hand enables us to introduce functional groups onto the

wheel, either on the spokes or on the rim. Thus, the assem-

bly of these molecules on surfaces does not only lead to

nanopatterned surfaces, but can also be enhanced to func-

tionalized surfaces. These 2D functionalized assemblies

could be used as platform for the directed growth of supra-

molecular 3D structures.^[54]Investigations using the self-or-

ganized 2D crystalline adsorption layer at the interface as

template to guide the assembly of guest molecules are ongo-

ing.^[55]Functionalization of the rim with dendrons should

pave the way to meltable or even mesomorphic derivatives

of 1. Alternatively, the rim can be reserved for the construc-

tion of the next generation in a repetitive approach leading

to spider web topologies. Ongoing synthetic efforts towards

more efficient routes to rigid 2D oligomers will also enable

us to investigate high-performance composite materials.

Molecular mechanics and molecular dynamics simulations

indicate that the most stable conformer of the molecule is a

nearly planar boat conformation with a small dihedral angle

(128). During the simulation the distance between “bow”

and “stern” atoms only fluctuates about 0.2 nm. Compared

to the open precursor and a hypothetical void macrocycle of

the same size and structure, the wheel-shaped compound is

utterly constrained in its macroconformations. Upon adsorp-

tion on the surface, the molecule is very likely to planarize,

which has already been observed for other phenylene–eth-

Experimental Section

Molecular mechanics (MM) and molecular dynamics (MD) simulations:

Simulations and analysis of the results were performed using the Materi-

als Studio 4.0 and Cerius²packages from Accelrys. We used the same

methodology as that applied for a previous phenylene–ethynylene macro-

cycle,^[35g]using the COMPASS force field as it provides an accurate de-

scription of the geometry and the torsion barriers of phenylene–ethynyl-

AHCTUNGTRENNUNG

ene oligomers and saturated chains, such as alkoxy or alkyl groups.^[56]

ACHTUNGTRENNUNGynylene macrocycles. The simulations also demonstrate that

The simulations were carried out on an isolated molecule, with methoxy

instead of the long OC₁₆H₃₃groups except when stated. The core was

first built planar and a preliminary molecular mechanics (MM) minimiza-

tion was performed. The nonbonded van der Waals and electrostatic

terms are described using a spline function, with a cut-off at 14.0 ꢆ

(spline width: 3 ꢆ). For MM energy minimization, we used the conjugate

gradient algorithm with an RMS force convergence parameter set to

10^ꢂ3kcalmol^ꢂ1ꢆ. A molecular dynamics (MD) procedure was then ap-

plied to the system using the canonical ensemble (N,V,T) at 298 K with a

Nosꢀ–Hoover thermostat. The time step was set to 1 fs and the duration

of the run was 500 ps, with an output frame every 200 fs. A few minima

of the potential energy during the whole MD run were picked from the

trajectory file, and a final MM procedure was applied to those structures.

at least half of the hexadecyloxy chains cannot be accommo-

dated within the triangle cavity formed by the spokes and

rim. Therefore they have to arrange above the molecular

plane, resulting in an increased thickness of the molecule,

consistent with the SANS, STM, and AFM observations.

Small-angle neutron scattering (SANS) experiments

affirm the rigidity of the molecule in solution. Observed

scattering intensities could be described by a theoretical

hamburger model, which assigns a radius of about 2.7 nm

and a thickness of about 0.22 nm to the rigid backbone. The

thickness of the soft shell, that is, the conformation of the

orthogonal hexadecyloxy chains, depends on the solvent.

The supramolecular organization of the compound has

been investigated both at the solid/liquid interface and in

thin films with STM and AFM. STM characterization dem-

onstrates that the wheel molecules adsorb with their molec-

ular plane parallel to the surface and organize into hexago-

nal crystalline domains both at the octanoic acid/graphite

and air/graphite interface with the tert-butyl groups on the

apexes staggered with those of neighboring molecules. The

staggering characteristic induces chirality to the organized

domains: the tert-butyl groups staggered either clockwise or

counterclockwise on the surface. AFM investigations of a

thin film grown from toluene demonstrate that the wheel

SANS measurements: 2D oligomer 1, protonated (THF, p.a., Riedel-de

Haꢈn) and deuterated solvents ([D₈]THF, 99.5%, CDCl₃, 99.8%, Deu-

tero GmbH) were used as received. The investigations were restricted to

low concentrations (ca. 1 gL^ꢂ1in THF and ca. 7 gL^ꢂ1in chloroform) due

to the limited solubility of the compound in these solvents. The solvents

used have high vapor pressures, leading to a marked loss of the solvent

during SANS measurements. This resulted in high concentration errors.

Therefore, the number density of the particles was a free fitting parame-

ter during evaluation of the data. The specific volume of 2D oligomer 1

was measured by using a DMA-60-densitometer (Paar, Graz, Austria).

The specific volume of 1 was determined to be 1.0 cm³g in chloroform.

Due to limited amount of 1, we assumed the equal value in THF. All

SANS data were obtained using the instrument D11 of the Institut Laue–

Langevin in Grenoble.^[57]To obtain the absolute intensities, the data were

corrected by the use of the software provided at the instrument. For all

data sets, the rates of incoherent scattering caused by the protons were

determined at high scattering vector, set as a constant and subtracted

2530

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 2518 – 2535

Template Synthesis

FULL PAPER

from the crude data. Further data treatment was carried out according to

literature procedures.^[58]

tions). Data integration was undertaken of the UV absorbance at l=

254 nm.

STM and AFM characterization: The compound was either dissolved in

octanoic acid or toluene with a concentration of about 0.1 mgg^ꢂ1. Sam-

ples for STM and AFM characterizations are prepared by drop casting a

solution (1 to 3 mL) on a freshly cleaved graphite substrate (HOPG,

grade ZYB, Advanced Ceramics Inc., Cleveland, OH). All STM experi-

ments were performed at 20 to 228C using a PicoSPM (Agilent Technolo-

gies, Santa Clara, CA). Pt/Ir (80%/20%, diameter 0.2 mm) STM tips

were prepared by mechanical cutting. The STM images were acquired in

the constant-current mode. The experiments were repeated in several

sessions by using several tips to check for reproducibility and to avoid ar-

tifacts. The molecules could be imaged both under positive and negative

bias voltage with submolecular resolution, both under ambient conditions

(from toluene) and at the liquid–solid interface (from octanoic acid). For

analysis purposes, the graphite substrate was used as a calibration grid.

The images were processed by scanning probe image processor (SPIP)

software (Image Metrology ApS). The imaging parameters (bias voltage

and set point of the tunnelling current) are indicated in the figure cap-

tions. The STM sample prepared from toluene was also used for AFM

characterization on a Nanoscope IV (Veeco Metrology, USA) operating

in tapping mode. A J scanner was used for large scale imaging and an A

scanner for high-resolution imaging.

Synthesis of 5: A mixture of 4 (1.90 g, 5.72 mmol), 3 (0.480 g, 0.48 mmol),

PEPPSI catalyst^[27](39.0 mg, 57.2 mmol), and KOH (321 mg, 5.72 mmol,

12.0 equiv) were placed in a Schlenk flask and then evacuated and

purged with Ar three times. After adding dry THF (6 mL), the reaction

mixture was stirred at 708C for 24 h. Additional powdered KOH

(12.0 equiv) was added and the reaction was stirred on for additional

24 h under argon. The ochre suspension was transferred into a round

flask with THF (10 mL) and dried under reduced pressure. Column chro-

matography (PE/CH₂Cl₂=5:1, R_f=0.60) yielded a pure white powder

(0.296 g, 0.168 mmol, 35%). M.p.>3008C; ¹H NMR (400 MHz, CD₂Cl₂,

3

258C, TMS): d=7.26 (s, 6H), 7.02 (d, J

(H,H)=8.2 Hz, 12H), 6.86 (s, 6H), 2.69 (q, ³J

(q, ³J(H,H)=7.5 Hz, 12H), 1.17 (t, 18H, ³J(H,H)=7.5 Hz), 0.84 (t, ³J-

AHCTUNGTRENNUNG

(H,H)=7.5 Hz, 18H), 0.33 ppm (s, 54H); ¹³C NMR (75 MHz, CD₂Cl₂,

ACHTUNGTRENNUNG

A

ACHTUNGTRENNUNG

A

ACHTUNGTRENNUNG

258C, TMS): d=147.20, 142.65, 140.59, 139.44, 139.17, 138.09, 136.45,

134.99, 131.42, 129.47, 127.47, 28.63, 26.03, 16.62, 15.77, 0.25 ppm; MS

(MALDI-TOF) C₁₂₀H₁₅₀Si₆(1759.04): m/z: 1760.1 [M⁺].

Synthesis of 6: Compound 5 (230 mg, 0.131 mmol) was placed in a

heated and Ar-flushed flask, dissolved in CHCl₃(75 mL, degassed with

Ar), and cooled down to 08C. Under strict exclusion of light, a 0.1m solu-

tion of ICl in CH₂Cl₂(23.5 mL, 2.35 mmol) was added over 30 min. The

dark solution was allowed to warm to RT and stirred overnight. Excess

halogen was decomposed by adding aqueous Na₂S₂O₅(40 mL, 10%). The

aqueous phase was extracted three times with CHCl₃, and the combined

organic layers were washed with water and brine, and dried over MgSO₄.

Removal of the solvent under reduced pressure yielded the crude prod-

uct, which was purified by recrystallization from CHCl₃. Compound 6

was obtained as white powder in several precipitations (249 mg,

Synthesis and characterization: The syntheses of compounds 3, 4, 7, 8, 10,

13, and 16 as well as their precursors were carried out according to litera-

ture methods and will be described in the Supporting Information. The

syntheses of compounds 1, 2, 5, 6, 9, 11, 12, 14, 15, 17, 18 and 19 are de-

scribed below. For synthesis, all experiments were performed under strict

exclusion of air and humidity. Therefore all glassware was evacuated,

heated, and cooled down in a flow of argon. Unless otherwise indicated,

commercially available chemicals were used as received. All solvents

were at least “p.a.” quality. Dry solvents were either purchased in

<50 ppm quality (toluene, dichloromethane) or dried, distilled, and

stored under argon according to standard methods (THF, piperidine, pyr-

idine). Prior to characterization and further processing, all solids and oils

0.120 mmol, 91%). M.p.>3008C; ¹H NMR (400 MHz, CD₂Cl₂, 258C,

3

TMS): d=7.64 (s, 6H), 7.01 (d, J

A

(d, ³J

(H,H)=7.5 Hz, 12H), 1.14 (t, 18H, ³J

³J

(H,H)=8.1 Hz, 12H), 2.65 (q, ³J

ACHUTGTNRENNUG CAHTUNGTRENNUNG

G

ACHTUNGTRENNUNG

AHCTUNGTRENNUNG

1

were dried overnight at RT under vacuum. H and ¹³C NMR spectra were

142.19, 141.40, 140.16, 139.58, 139.20, 137.77, 131.43, 129.80, 127.29, 98.98,

33.71, 25.60, 15.47, 14.91 ppm; MS (MALDI-TOF) C₁₀₂H₉₆I₆(2082.18):

m/z (%): 2206.5 (12), 2082.3 (100) [M⁺], 1957.4 (18).

recorded on a Bruker DPX 300, DPX 400, and DRX 500 (300, 400, and

500 for ¹H and 75, 100 and 125 MHz for ¹³C, respectively). Chemical

shifts are given in parts per million (ppm) referring to tetramethylsilane

(TMS). Mass spectra were measured on a Finnigan ThermoQuest MAT

95 XL (EI-MS), Kratos Concept 1H (FAB-MS; matrix material: 3-nitro-

benzyl alcohol, no salts added) and a Bruker Daltronics autoflex TOF/

TOF, (MALDI-MS; matrix material: DCTB, no salts added). m/z peaks

smaller than 10% (compared with the most intensive signal) are not re-

ported. UV/Vis spectra were recorded on a Shimadzu UV-2100 spectro-

photometer. Unless otherwise indicated, solutions were prepared with in-

definite concentration of the sample in dichloromethane. Quartz cuvettes

measuring 1 cm in diameter were used for sample solutions as well as sol-

vent references (dual-beam measurements). Fluorescence experiments

were run on a Horiba Jobin Yvon FluoroMax-4 spectofluorometer in all-

transparent quartz cuvettes and by monochromatic excitation at the indi-

cated wavelength. Melting points were determined using a Leica DMLB

microscope with resistive heating socket controlled by a Leica LMW

transformator and a Testo 925 digital thermometer. Thin-layer chroma-

tography was conducted on silica gel coated aluminium plates (Ma-

cherey-Nagel, Alugramm SIL G/UV 0.25 mm, silica gel with fluorescence

indicator); detection either by fluorescence quenching of the luminous

indicator (l=254 nm) or autofluorescence of substance spots (l=

366 nm). For column chromatography, silica gel 60 (40–63 mm) from

Merck was used as stationary phase, suspended in the respective solvent.

Radial thin-layer chromatography was performed with a “chromatotron”

(Harrison Research) with silica gel 60 PF254 containing gypsum and fluo-

rescence indicator (designed for preparative thin layer chromatography)

as stationary phase. Gel permeation chromatograms (GPC) were mea-

sured in THF (stabilized with 2.5 ppm BHT) at RT using a UV detector

operating at l=254 nm (flow rate: 1 mLmin^ꢂ1in the analytical mode;

4 mLmin^ꢂ1for preparative separations). The molecular weight was ob-

tained from polystyrene-calibrated SEC columns (porosity 10², 10³, 10⁵

and 10⁶ꢆ for analytical measurements and 10³ꢆ for preparative separa-

Synthesis of 9: Compounds

7 (17.63 g, 25.04 mmol) and 8 (58.29 g,

0.246 mol) were ground in a mortar, then suspended in Ac₂O (210 mL)

and heated to reflux for 5 h at 1508C bath temperature (no more CO₂de-

veloping). The reaction mixture was allowed to cool down and stirred

overnight at RT. The precipitated solid was filtered, washed with distilled

water and MeOH, recrystallized from CHCl₃(90 mL), and dried in

vacuo. The product was obtained as white powder in several precipita-

tions (13.48 g, 17.52 mmol, 70%). M.p.>2508C; ¹H NMR (250 MHz,

C₂D₂Cl₄, 258C, TMS): d=7.63 (d, ³J

G

3

7.56 (d, J

A

N

(H,H)=8.4 Hz, 2H), 6.86 (d, ³J

A

8.4 Hz, 2H), 1.37 ppm (s, 9H); ¹³C NMR (100 MHz, C₂D₂Cl₄, 258C,

TMS): d=151.06, 141.20, 141.06, 140.39, 137.72, 137.04, 136.64, 136.05,

133.23, 131.84, 131.77, 130.94, 128.32, 126.78, 126.03, 120.63, 92.72, 34.62,

31.46 ppm; MS (EI) C₃₄H₂₇BrI₂(767.94): m/z (%): 770.0 (100) [⁸¹M⁺],

768.0 (100) [⁷⁹M⁺], 755.0 (50) [⁸¹M⁺ꢂCH₃], 753.0 (50) [⁷⁹M⁺ꢂCH₃].

Synthesis of 11: Compounds

9 (1.70 g, 2.20 mmol) and 10 (1.69 g,

4.40 mmol), Cs₂CO₃(2.50 g, 7.67 mmol), THF (10 mL) and distilled water

(4 mL) were mixed and the suspension was degassed three times (“pump

and freeze”), [PdACTHNUGTRNEUNG(PPh₃)₄] (40 mg, 35 mmol) was added and the mixture

was degassed once again. The reaction was stirred at 708C for three days

before it was diluted with water (50 mL) and Et₂O (100 mL). The organic

layer was separated and the aqueous phase was extracted with Et₂O. The

combined organic layers were washed with diluted HCl and brine, and

dried over MgSO₄. Removal of the solvent in vacuo gave the crude prod-

uct, which was purified by column chromatography (PE/CH₂Cl₂=15:1;

TLC in 5:1, R_f=0.32). Compound 11 was obtained as white solid

1

(0.580 g, 0.56 mmol, 26%). M.p. 171–1748C; H NMR (250 MHz, CD₂Cl₂,

258C, TMS): d=7.71–7.62 (m, 4H); d=7.72 (s, 2H), 7.68 (d, ³J

ACHTUNGTRENNUNG

8.5 Hz, 2H), 7.56 (s, 4H), 7.55 (s, 4H), 7.52 (d, ³J

ACHTUNGTRENNUNG

Chem. Eur. J. 2009, 15, 2518 – 2535

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

2531

S. Hçger, S. De Feyter, R. Lazzaroni, S. Rosenfeldt et al.

3

7.49 (d, J

A

G

E

0.264 mmol, 64%). ¹H NMR (400 MHz, CDCl₃, 258C, TMS): d=7.72 (s,

G

2H), 7.65 (d, ³J(H,H)=8.5 Hz, 2H), 7.52 (s, 8H), 7.50 (d, ³J

AHTCUNGTRENNUGN ACHTUNGTRENNUNG

1.16 ppm (s, 42H); ¹³C NMR (100 MHz, C₂D₂Cl₄, 258C, TMS): d=

151.04, 142.01, 141.22, 140.42, 140.36, 138.60, 137.21, 136.53, 133.50,

132.43, 130.69, 130.55, 128.35, 126.76, 126.70, 126.31, 126.01, 122.48,

120.43, 106.97, 91.47, 34.56, 31.15, 18.53, 11.57 ppm; MS (EI) C₆₈H₇₇BrSi₂

(1028.47): m/z (%): 1029.6 (100), 1028.5 (75) [M⁺], 766.4 (100).

8.4 Hz, 2H), 7.44 (d, ³J

2H), 7.21 (d, ³J(H,H)=8.3 Hz, 4H), 6.93 (d, ³J

(s, 2H), 3.94 (dt, ³J

(H,H)=8.3 Hz, 4H), 7.22 (d, ³J

ACHUTGTNRENNUG CAHTUNGTRENNUNG

N

ACHTUNGTRENNUNG

N

ACHTUNGTRENNUNG

1.90–1.73 (m, 6H), 1.52–1.41 (m, 4H), 1.38 (s, 9H), 1.37–1.19 (m, 48H),

1.14 (s, 42H), 0.90–0.82 (m, 8H), 0.26 ppm (s, 6H); ¹³C NMR (100 MHz,

CD₂Cl₂, 258C, TMS): d=154.18, 153.41, 151.03, 142.01, 141.27, 140.42,

140.39, 140.04, 138.24, 137.26, 137.17, 132.40, 131.95, 130.71, 130.57,

128.34, 126.76, 126.70, 126.28, 126.00, 122.45, 121.07, 119.78, 117.14,

116.68, 114.37, 113.25, 106.99, 102.48, 97.99, 95.00, 91.43, 86.15, 69.63,

69.45, 34.57, 32.02, 31.16, 29.79, 29.74, 29.66, 29.47, 29.45, 29.40, 29.35,

29.27, 26.07, 26.05, 22.78, 20.75, 20.45, 18.52, 15.71, 13.96, 11.44,

ꢂ2.05 ppm; MS (MALDI-TOF) C₁₁₆H₁₅₇NO₂Si₃(1680.15): m/z: 1681.1

[M⁺+H].

Synthesis of 14: At ꢂ788C, nBuLi (0.56 mL, 1.6m in hexane) was added

dropwise to a stirred solution of 11 (910 mg, 0.88 mmol) in THF (10 mL)

and stirred for 15 min. Then a solution of diiodoethane (250 mg) in THF

(10 mL) was added slowly. The reaction mixture was stirred for 1 h at

ꢂ788C, the cooling bath was removed, and the mixture was allowed to

warm to RT. After stirring for another 30 min the reaction was diluted

with Et₂O (100 mL), washed twice with sodium thiosulfate solution

(10%) and brine, and dried over MgSO₄. Removal of the solvent under

reduced pressure gave crude 12 as a white solid (0.890 g, 0.83 mmol) that

was used without further purification. Under Ar, a dry Schlenk flask was

charged with the crude product 12 (0.502 g, 0.47 mmol), CuI (10 mg, 52.5

Synthesis of 18: The compound was synthesized by the same procedure

described for 15 from 17 and purified by column chromatography (PE/

CH₂Cl₂=2:1, R_fꢁ0.8). Compound 18 was obtained as a slightly yellow

resin (0.183 g, 0.117 mmol, 91%). ¹H NMR (300 MHz, CD₂Cl₂, 258C,

mmol), PPh₃(20 mg, 76 mmol), [PdCl₂ACHTUNTGRNEUNG(PPh₃)₂] (20 mg, 28.5 mmol) and 13

(0.106 g, 0.70 mmol). The reagents were dissolved in THF (8 mL) and pi-

peridine (4 mL) and stirred for 24 h at RT. CH₂Cl₂(70 mL) and water

were poured into the cooled-down reaction mixture, the separating or-

ganic layer was washed with diluted HOAc, water, NaOH solution

(10%), water, and brine. The crude product was obtained after drying

over MgSO₄and removing the solvent in vacuo. Purification by column

chromatography (PE/CH₂Cl₂=4:1, R_f=0.07) yielded 14 (0.330 g,

0.30 mmol, 61% for both steps). M.p. 153–1568C; ¹H NMR (250 MHz,

TMS): d=7.74 (s, 2H), 7.69 (d, ³J

G

3

14H), 7.25 (d, J

(H,H)=8.2 Hz, 4H), 7.22 (d, J

(d, ³J(H,H)=8.1 Hz, 2H), 6.94 (s, 1H), 6.93 (s, 1H), 3.94 (t, ³J

ACTHNUGTRENNUGN ACHTUNGTRENNUNG

AHCTUNGTRENNUNG

6H); ¹³C NMR (100 MHz, CD₂Cl₂, 258C, TMS): d=154.16, 153.39,

151.02, 142.00, 141.27, 140.41, 140.38, 140.05, 138.23, 137.25, 137.15,

132.39, 131.95, 130.71, 130.57, 128.33, 126.76, 126.69, 126.27, 125.99,

122.44, 121.04, 117.69, 116.73, 114.48, 112.44, 106.98, 94.90, 91.42, 86.03,

82.19, 80.01, 69.64, 69.62, 34.56, 32.01, 31.15, 29.77, 29.74, 29.64, 29.44,

29.41, 29.38, 29.31, 29.24, 26.02, 25.97, 22.77, 18.51, 13.96, 11.43 ppm; MS

CD₂Cl₂, 258C, TMS): d=7.72 (s, 2H), 7.67 (d, ³J

7.56 (s, 4H), 7.54 (s, 4H), 7.51 (d, ³J(H,H)=8.6 Hz, 2H), 7.48 (d, ³J-

(H,H)=8.4 Hz, 4H), 7.22 (d, ³J(H,H)=8.3 Hz, 4H), 7.16 (d, ³J

(H,H)=

8.4 Hz, 2H), 6.94 (d, ³J(H,H)=8.4 Hz, 2H), 2.38 (t, ³J

(H,H)=7.0 Hz,

2H), 1.81–1.69 (m, 2H), 1.37 (s, 9H), 1.15 (s, 42H), 0.81–0.74 (m, 2H),

0.19 ppm (s, 6H); ¹³C NMR (100 MHz, CD₂Cl₂, 258C, TMS): d=151.04,

141.98, 141.26, 140.41, 140.38, 140.34, 138.18, 137.21, 137.05, 132.42,

131.85, 131.10, 130.54, 128.32, 126.75, 126.68, 126.25, 126.00, 122.47,

120.54, 119.80, 106.95, 106.23, 92.58, 91.47, 34.55, 31.14, 20.66, 20.43,

18.15, 15.62, 11.41, ꢂ2.17 ppm; MS (MALDI-TOF) C₇₆H₈₉NSi₃(1099.63):

m/z (%): 1313.8, 1099.6 [M⁺], 1065.2, 1056.6 (100) [M⁺ꢂC₃H₇].

Synthesis of 15: Compound 14 (0.489 g, 0.44 mmol) was dissolved in

THF/MeOH (2:1; 15 mL, purged for 15 min with Ar). In a flow of Ar,

anhydrous powdered K₂CO₃(0.607 g, 4.4 mmol) was added in small por-

tions and the resulting suspension was vigorously stirred for 3 h under

Ar. The reaction was diluted with water (50 mL) and Et₂O (100 mL).

The aqueous phase was extracted with Et₂O, the combined organic layers

were washed with water and brine, and were dried over MgSO₄. After re-

moving the solvents under reduced pressure, the crude product was dis-

solved in CH₂Cl₂(3 mL) and precipitated by the addition of MeOH

(15 mL). The white solid was filtered and dried in vacuo (0.420 g,

0.43 mmol, 98%). M.p. 2408C (decomp); ¹H NMR (250 MHz, CD₂Cl₂,

ACHTUNGTREN(NUGN H,H)=8.5 Hz, 2H),

AHCTUNGTRENNUNG

A

E

(MALDI-TOF)

1555.1 [M⁺]; GPC (PS calibration): M_w=1800 gmol^ꢂ1

Synthesis of 19: Compound 6 (17.8 mg, 8.5 mmol), CuI (2 mg, 10.5 mmol),

PPh₃(4 mg, 15 mmol) and PdCl₂A(PPh₃)₂(4 mg, 5.7 mmol) were dissolved

C

₁₁₀H₁₄₆O₂Si₂(1555.09): m/z: 1806.3 [M⁺+DCTB],

.

CTHUNGTRENNUNG

in piperidine (2 mL) under Ar and 18 (0.120 g, 77.1 mmol) in dry THF

(4 mL) was added by syringe. The reaction mixture was heated to 708C

and stirred for 3 d. The reaction was diluted with water (20 mL) and

Et₂O (50 mL), the separated aqueous phase was extracted three times

with Et₂O (20 mL each). Subsequently, the combined organic layers were

washed with diluted H₂SO₄(10%), water, and brine. The crude product

was obtained after drying over Na₂SO₄and removing the solvent in

vacuo. Column chromatography (PE/CH₂Cl₂=4:1 to 3:1; TLC in 2:1,

R_f=0.55) gave 19 (53.6 mg, 5.0 mmol, 59%) as a yellow solid slightly con-

taminated with the fivefold coupling product (<5%).¹H NMR

(400 MHz, CD₂Cl₂, 258C, TMS): d=7.74 (s, 12H), 7.69 (d, ³J

8.4 Hz, 12H), 7.62–7.47 (m, 84H), 7.35 (s, 6H), 7.30–7.21 (m, 36H), 7.06

ACHTUNGTRENNUNG(H,H)=

3

(d, J

(dt, ³J

1.88–1.74 (m, 24H), 1.57–1.43 (m, 24H), 1.38 (s, 54H), 1.35–1.18 (m,

306H), 1.14 (s, 252H), 0.90 (t, ³J(H,H)=7.5 Hz, 18H), 0.84 ppm (t, ³J-

(H,H)=5.6 Hz, 36H); ¹³C NMR (125 MHz, CD₂Cl₂, 258C, TMS): d=

ACHTUNGTRENNUNG

AHCTUNGTRENNUNG

258C, TMS): d=7.72 (s, 2H), 7.68 (d, ³J

4H), 7.54 (s, 4H), 7.51 (d, ³J(H,H)=8.3 Hz, 2H), 7.48 (d, ³J

8.3 Hz, 4H), 7.22 (d, ³J(H,H)=8.4 Hz, 4H), 7.18 (d, ³J

(H,H)=8.3 Hz,

2H), 6.96 (d, ³J

(H,H)=8.3 Hz, 2H), 3.05 (s, 1H), 1.37 (s, 9H), 1.15 ppm

ACHTUNGTRENNUNG

G

ACHTUNGTRENNUNG

G

153.53, 153.52, 151.01, 143.91, 142.18, 142.00, 141.26, 140.52, 140.37,

139.90, 139.61, 139.23, 138.74, 138.23, 137.26, 137.17, 132.39, 132.23,

131.93, 131.47, 130.68, 130.56, 129.51, 128.33, 127.50, 127.49, 126.75,

126.68, 126.27, 125.99, 122.44, 121.29, 121.21, 116.75, 116.28, 114.15,

113.54, 106.98, 94.75, 93.83, 91.39, 89.23, 86.46, 69.60, 69.41, 34.56, 32.01,

31.99, 31.14, 29.79, 29.74, 29.70, 29.66, 29.50, 29.44, 29.42, 29.38, 27.28,

26.14, 26.07, 25.79, 22.77, 18.51, 15.41, 15.06, 13.98, 13.97, 11.40 ppm; MS

ACHTUNGTRENNUNG

(s, 42H); ¹³C NMR (100 MHz, CD₂Cl₂, 258C, TMS): d=151.05, 142.00,

141.23, 140.49, 140.44, 140.38, 138.23, 137.23, 137.04, 132.41, 131.87,

131.28, 130.54, 128.32, 126.77, 126.70, 126.27, 126.00, 122.47, 119.80,

106.99, 91.46, 83.59, 77.21, 34.57, 31.16, 18.52, 11.44 ppm; MS (ESI)

C₇₀H₇₈Si₂(974.56): m/z: 1013.5 [M⁺+K].

(MALDI-TOF)

C₇₆₂H₉₆₆O₁₂Si₁₂(monoisotopic: 10646.23, distr. max.:

Synthesis of 17: Compounds 15 (0.395 g, 0.405 mmol) and 16 (0.370 g,

0.44 mmol), CuI (7 mg, 37 mmol), PPh₃(15 mg, 57 mmol), and [PdCl₂-

ACHTUNGTRENNUNG(PPh₃)₂] (15 mg, 21 mmol) were dissolved in THF (10 mL) and piperidine

(5 mL) and stirred for 4 h at RT under Ar. The reaction was diluted with

Et₂O (70 mL) and water (50 mL) and the separated aqueous layer was

extracted several times with Et₂O. The combined organic layers were

washed with diluted HOAc, water, NaOH solution (10%), water, and

brine. After drying over MgSO₄and removing the solvent in vacuo the

crude product was purified by column chromatography (PE/CH₂Cl₂=

10655.25): m/z: 10879.9 [M⁺+dithranol], 10655.1 [M⁺], 10431.8 [M⁺

ꢂC₁₆H₃₁], 10206.7 [M⁺ꢂ2C₁₆H₃₁], 9226.1 [M⁺_(5:1)], 8966.7; GPC (PS cali-

bration): M_w=10600 gmol^ꢂ1

.

Synthesis of 2: TBAF solution in THF (5 mL, 1.0m) was added slowly to

19 (0.113 g, 10.6 mmol) under Ar and the starting material was dissolved

with the aid of an ultrasonic bath (5 min). The reaction mixture was

stirred for 16 h before additional TBAF solution (5 mL, 1.0m) was added.

Again, the reaction flask was treated in an ultrasonic bath, then stirred

overnight and diluted with Et₂O and water. The aqueous phase was ex-

5:2; TLC in 3:2, R_f=0.49) to give 17 as

a yellow solid (0.444 g,

2532

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 2518 – 2535

Template Synthesis

FULL PAPER

tracted with Et₂O, the combined organic layers were washed with water

and brine and dried over Na₂SO₄. After removing of the solvent, the

crude product was purified by radial chromatography (PE/CH₂Cl₂=2:1,

R_f=0.51). Again, after removing of the solvent, the product was dis-

solved in CHCl₃(3 mL), precipitated with MeOH (15 mL) and dried in

vacuo to give 2 as a faintly yellow solid (70.6 mg, 8.04 mmol, 76%).

M.p.>2508C (decomp); ¹H NMR (500 MHz, C₂D₂Cl₄, 258C, TMS): d=

7.68–7.62 (m, 48H), 7.57 (d, ³J(H,H)=8.5 Hz, 12H), 7.54 (d, ³J

AHTCUNGTRENNUGN ACHTUNGTRENNUNG

8.3 Hz, 24H), 7.41 (s, 6H), 7.32 (d, ³J

A

3

N

(H,H)=8.0 Hz, 12H), 7.04 (s, 6H), 7.01–

6.93 (m, 36H), 4.08 (dt, ³J

ACTHUNTGRENNG(U H,H)=6.5 Hz, 24H), 2.98–2.88 (m, 12H),

2.50–2.40 (m, 12H), 1.91–1.81 (m, 24H), 1.60–1.50 (m, 24H), 1.47 (s,

54H), 1.45–1.26 (m, 306H), 1.00–0.90 ppm (m, 54H); MS (MALDI-

TOF) C₆₅₄H₇₁₄O₁₂(monoisotopic: 8759.53, distr. max.: 8767.56): m/z:

17507.5 [2M⁺], 13135.9 [3M²⁺], 9268.3 [M⁺+2DCTB], 9018.0 [M⁺

+DCTB], 8767.2 [M⁺], 4414.4, 4383.9 [M²⁺]; GPC (PS calibration):

7.79 (s, 12H), 7.72 (d, ³J

(H,H)=8.3 Hz, 12H), 7.63 (d, ³J

ACHUTGTNRENNUG ACHUTTGNREN(NUGN H,H)=8.5 Hz,

24H), 7.58 (d, ³J

ACHTUNGTRENNUNG

M_w=9600 gmol^ꢂ1

.

7.32–7.24 (m, 36H), 7.12–7.03 (m, 12H), 7.03–6.92 (m, 42H), 4.08–3.96

(m, 24H), 3.20 (s, 12H), 2.97–2.82 (m, 12H), 2.51–2.35 (m, 12H), 1.89–

1.77 (m, 24H), 1.56–1.46 (m, 24H), 1.41 (s, 54H), 1.37–1.17 (m, 306H),

ACTHNUTRGNEUNG

0.96–0.90 (m, 18H), 0.88 ppm (dt, ³J(H,H)=7.0 Hz, 36H); ¹³C NMR

(125 MHz, C₂D₂Cl₄, 258C, TMS): d=153.60, 153.46, 150.92, 143.34,

142.34, 141.77, 141.28, 140.87, 140.15, 139.71, 139.10, 137.81, 137.04,

136.98, 132.68, 132.39, 131.86, 131.45, 131.12, 130.97, 130.59, 129.47,

128.94, 128.38, 127.36, 126.83, 126.63, 126.41, 126.03, 121.09, 120.90,

120.88, 120.82, 117.11, 116.36, 114.32, 113.59, 95.46, 94.24, 89.40, 86.49,

83.86, 78.33, 69.84, 69.53, 34.63, 32.04, 31.49, 29.87, 29.84, 29.81, 29.75,

29.74, 29.72, 29.70, 29.56, 29.50, 29.45, 27.34, 26.17, 26.15, 25.87, 22.85,

15.56, 15.23, 14.37 ppm; MS (MALDI-TOF) C₆₅₄H₇₂₆O₁₂(monoisotopic:

8770.62, distr. max.: 8779.65): m/z: 9030.2 [M⁺+DCTB)], 8779.5 [M⁺],

Acknowledgements

This work was supported by the Belgian Federal Science policy through

IAP 6/27, the Fund for Scientific Research-Flanders (FWO), the Belgian

National Fund for Scientific Research (F.R.S.-FNRS), the Deutsche For-

schungsgemeinschaft and the Sonderforschungsbereich 624. M.S. is

Chargꢀ de Recherches of the F.R.S.-FNRS. The authors thank the Insti-

tute Laue–Langevin for providing beamtime at the instrument D11.

8600.3, 7663.2 [M⁺₍₅₎], 7405.3; GPC (PS calibration): M_w=10000 gmol^ꢂ1

.

Under non-optimized reaction conditions (concerning the preceeding six-

fold coupling reaction) we found the following properties for the depro-

tected incomplete coupling products.

[1] G. Wegner, K. Mꢁllen, Electronic Materials. The Oligomer Ap-

proach, Wiley-VCH, Weinheim, 1999.

[2] J. M. Tour, Chem. Rev. 1996 , 96 , 53 7 –5 53.

[3] G. R. Newkome, C. N. Moorefield, F. Vçgtle, Dendrimers and Den-

drons. Concepts, Syntheses, Applications, Wiley-VCH, Weinheim,

2001.

1

Data for 1:5 coupling product: H NMR (400 MHz, CD₂Cl₂, 258C, TMS):

d=7.75 (s, 10H), 7.69 (d, ³J

ACHTUNGTRENNUNG

AHCTUNGTRENNUNG

ACHTUNGTRENNUNG

[4] Z. Xu, M. Kahr, K. L. Walker, C. L. Wilkins, J. S. Moore, J. Am.

Chem. Soc. 1994 , 116 , 453 7 –4 550.

(m, 23H), 4.02–3.95 (m, 20H), 3.17, 3.16 (2 s, 10H), 2.90–2.81 (m, 10H),

2.72–2.63 (m, 2H), 2.45–2.27 (m, 12H), 1.86–1.74 (m, 20H), 1.55–1.44 (m,

20H), 1.38 (s, 45H), 1.36–1.10 (m, 255H), 0.93–0.81 ppm (m, 45H); MS

[5] a) S. Rosenfeldt, N. Dingenouts, D. Pçtschke, M. Ballauff, A. J. Ber-

resheim, K. Mꢁllen, P. Lindner, Angew. Chem. 2004 , 116 , 11 1 –1 14;

Angew. Chem. Int. Ed. Engl. 2004 , 43 , 10 9 –1 12; b) S. Rosenfeldt, N.

Dingenouts, D. Pçtschke, M. Ballauff, A. J. Berresheim, K. Mꢁllen,

P. Lindner, K. Saalwꢂchter, J. Lumin. 2005 , 111 , 22 5 –2 38.

[6] M. D. Watson, A. Fechtenkçtter, K. Mꢁllen, Chem. Rev. 2001 , 101 ,

126 7 –1 300.

[7] a) S. De Feyter, F. C. De Schryver, Chem. Soc. Rev. 2003 , 32 , 13 9 –

150; b) J. V. Barth, G. Costantini, K. Kern, Nature 2005 , 437 , 67 1 –

679; c) S. De Feyter, F. C. De Schryver, Top. Curr. Chem. 2005 , 258 ,

20 5 –2 55; d) J. V. Barth, Annu. Rev. Phys. Chem. 2007 , 58 , 37 5 –4 07.

[8] a) P. Ceroni, G. Bergamini, F. Marchioni, V. Balzani, Prog. Polym.

Sci. 2005 , 30 , 45 3 –4 73; b) A. Ajayaghosh, V. K. Peaveen, C. Vijaya-

kumar, Chem. Soc. Rev. 2008 , 37 , 10 9 –1 22.

(MALDI-TOF)

7663.74): m/z: 7914.9 [M⁺+DCTB], 7664.5 [M⁺], 7483.9; GPC (PS cali-

bration): M_w=9300 gmol^ꢂ1

Data for 1:4 coupling product: H NMR (400 MHz, CD₂Cl₂, 258C, TMS):

d=7.75 (s, 8H), 7.69 (d, ³J(H,H)=8.4 Hz, 8H), 7.64 (s, 2H), 7.59 (d, ³J-

(H,H)=8.0 Hz, 16H), 7.54 (d, J

8.4 Hz, 8H), 7.50 (d, ³J(H,H)=8.5 Hz, 16H), 7.34 (s, 4H), 7.26 (d, ³J-

(H,H)=8.4 Hz, 16H), 7.23 (d, ³J

(H,H)=8.2 Hz, 8H), 7.07–7.01 (m,

12H), 7.00–6.96 (m, 14H), 6.95–6.85 (m, 20H), 4.01–3.95 (m, 16H), 3.18,

3.17, 3.16 (3 s, 8H), 2.85 (q, ³J(H,H)=7.0 Hz , 8H), 2.67 (q, ³J

(H,H)=

C₅₆₂H₆₂₁IO₁₀(monoisotopic: 7656.71, distr. max.:

.

1

AHCTUNGTRENNUNG

3

N

ACHUTGTNREN(UNG H,H)=8.0 Hz, 16H), 7.52 (d, JACHTUGNTREN(NUGN H,H)=

ACHTUNGTRENNUNG

A

ACHTUNGTRENNUNG

A

ACHTUNGTRENNUNG

7.3 Hz, 4H), 2.42–2.27 (m, 12H), 1.85–1.74 (m, 16H), 1.56–1.43 (m,

16H), 1.38 (s, 36H), 1.37–1.17 (m, 204H), 0.91–0.80 ppm (m, 36H); MS

(MALDI-TOF)

C₄₇₀H₅₁₆I₂O₈(monoisotopic: 6541.81, distr. max.:

´

[9] a) S. Kumar, Chem. Soc. Rev. 2006 , 35 , 8 3 –1 09; b) Z. Tomovic, J.

6547.83): m/z: 6798.5 [M⁺+DCTB], 6548.5 [M⁺], 6368.2 [M⁺ꢂC₁₅H₃₀];

Van Dongen, S. J. George, H. Xu, W. Pisula, P. Leclꢉre, M. M. J.

Smulders, S. De Feyter, E. W. Meijer, A. P. H. J. Schenning, J. Am.

Chem. Soc. 2007, 129, 16190–16196.

GPC (PS calibration): M_w=8700 gmol^ꢂ1

.

Synthesis of 1: Under Ar, CuCl (96 mg, 0.97 mmol) and CuCl₂(13 mg,

97 mmol) were suspended in dry pyridine (5 mL) and heated to 508C.

Compound 2 (17 mg, 1.94 mmol) in pyridine (10 mL) was added over 4 d

(pseudo high-dilution conditions). The syringe was rinsed with pyridine

(5 mL), which was added slowly (12 h) to the stirred reaction mixture at

508C. After stirring for another 12 h and cooling to RT, the reaction was

diluted by adding CHCl₃(200 mL) and water (50 mL). Subsequently, the

organic layer was separated and washed with aqueous NH₃(25%),

water, HOAc (10%), water, NaOH (10%), and brine. After drying over

Na₂SO₄and evaporation of the solvent the crude product was filtered

through a short silica gel column using CHCl₃as eluent (R_f=0.90). The

solvent was evaporated nearly to dryness and the crude cyclization prod-

uct was precipitated by the addition of MeOH. GPC analysis of the pre-

cipitate (12.3 mg, 72%) indicated the presence of a small amount of the

dimer. The yellowish powder was dissolved in a small amount of THF

and purified by preparative GPC. The dry product fraction was dissolved

in CHCl₃(1 mL) and precipitated by the addition of MeOH (15 mL)

twice. Drying in vacuo gave of the pure 1 as slightly yellow solid

[10] Polymer-Clay Nanocomposites (Eds.: T. J. Pinnavaia, G. W. Beall),

Wiley, Chichester, 2000.

[11] a) A. Nelson, T. Cosgrove, Langmuir 2004 , 20 , 229 8 –2 304; b) L. Li,

L. Harnau, S. Rosenfeldt, M. Ballauff, Phys. Rev. E 2005 , 72 ,

05150 4; c) A. Loiseau, J.-F. Tassin, Macromolecules 2006 , 39 , 918 5 –

9191.

[12] D. J. Voorn, W. Ming, A. M. Van Herk, Macromolecules 2006 , 39 ,

465 4 –4 656.

[13] For nanocomposites with tubular molecules (CNTs) see: A. Beig-

beder, M. Linares, M. Devalckenaere, P. Degꢀe, M. Claes, D. Bel-

jonne, R. Lazzaroni, P. Dubois, Adv. Mater. 2008 , 20 , 100 3 –1 007.

[14] For surface-grown rigid extended 2D structures see, for example:

N. A. A. Zwaneveld, R. Pawlak, M. Abel, D. Catalin, D. Gigmes, D.

Bertin, L. Porte, J. Am. Chem. Soc. 2008 , 130 , 667 8 –6 679.

[15] S. I. Stupp, S. Son, L. S. Li, H. C. Lin, M. Keser, J. Am. Chem. Soc.

1995 , 117 , 521 2 –5 227.

[16] C. H. M. Weber, A. Chiche, G. Krausch, S. Rosenfeldt, M. Ballauff,

L. Harnau, I. Gçttker-Schnetmann, Q. Tong, S. Mecking, Nano Lett.

2007 , 7 , 202 4 –2 029.

(10.0 mg, 1.14 mmol, 59%). M.p.>2508C (decomp); ¹H NMR (500 MHz,

3

C₂D₂Cl₄, 1208C, TMS): d=7.84 (s, 12H), 7.76 (d, J

N

Chem. Eur. J. 2009, 15, 2518 – 2535

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

2533

S. Hçger, S. De Feyter, R. Lazzaroni, S. Rosenfeldt et al.

[17] a) C. D. Simpson, J. D. Brand, A. J. Berresheim, L. Przybilla, H. J.

f) M. Fischer, S. Hçger, Tetrahedron 2003 , 59 , 944 1 –9 446; g) A. Zie-

gler, W. Mamdouh, A. Ver Heyen, M. Surin, H. Uji-i, M. M. S.

Abdel-Mottaleb, F. C. De Schryver, S. De Feyter, R. Lazzaroni, S.

Hçger, Chem. Mater. 2005 , 17 , 567 0 –5 683.

´

Rꢂder, K. Mꢁllen, Chem. Eur. J. 2002 , 8 , 142 4 –1 429; b) Z. Tomovic,

M. D. Watson, K. Mꢁllen, Angew. Chem. 2004, 116, 773–777;

Angew. Chem. Int. Ed. 2004, 43, 755–758; c) A. C. Grimsdale, K.

Mꢁllen, Angew. Chem. 2005 , 117 , 573 2 –5 772; Angew. Chem. Int.

Ed. 2005 , 44 , 559 2 –5 629; d) D. Wasserfallen, M. Kastler, W. Pisula,

A. Hofer, Y. Fogel, Z. Wang, K. Mꢁllen, J. Am. Chem. Soc. 2006 ,

128 , 133 4 –1 339; e) X. Feng, J. Wu, M. Ai, W. Pisula, L. Zhi, J. P.

Rabe, K. Mꢁllen, Angew. Chem. 2007 , 119 , 309 3 –3 096; Angew.

Chem. Int. Ed. 2007 , 46 , 303 3 –3 036.

[36] S. Hçger, K. Bonrad, L. Karcher, A.-D. Meckenstock, J. Org. Chem.

2000 , 65 , 158 8 –1 589.

[37] A. Godt, M. Schulte, H. Zimmermann, G. Jeschke, Angew. Chem.

2006 , 118 , 772 2 –7 726; Angew. Chem. Int. Ed. 2006 , 45 , 756 0 –7 564.

[38] a) V. Francke, T. Mangel, K. Mꢁllen, Macromolecules 1998 , 31 ,

244 7 –2 453; b) S. Huang, J. M. Tour, J. Am. Chem. Soc. 1999 , 121 ,

490 8 –4 909; c) U. H. F. Bunz, Chem. Rev. 2000 , 100 , 160 5 –1 644.

[39] Otherwise, one might argue that the rigid 2D scaffold is wrapped

inside a shell of solubilising substituents that manipulate the hydro-

dynamic volume of the molecule. The influence of long alkyl chains

inside the molecular plane on the hydrodynamic volume will be in-

vestigated, as soon as different 2D oligomer representatives are

available for comparison.

[18] a) W. B. Wan, S. C. Brand, J. J. Pak, M. M. Haley, Chem. Eur. J.

2000 , 6 , 204 4 –2 052; b) J. A. Marsden, M. M. Haley, J. Org. Chem.

2005 , 70 , 1021 3 –1 0226; c) T. Yoshimura, A. Inaba, M. Sonoda, K.

Tahara, Y. Tobe, R. V. Williams, Org. Lett. 2006 , 8 , 293 3 –2 936.

[19] a) E. M. Garcꢊa-Frutos, E. M. M. Fernꢋndez-Lꢋzaro, P. Vꢋzquez, T.

Torres, J. Org. Chem. 2000, 65, 6841–6846; b) R. S. Iglesias, C. G.

Claessens, M. ꢌ. Herranz, T. Torres, Org. Lett. 2007 , 9 , 538 1 –5 384.

[20] Y. Fogel, M. Kastler, Z. Wang, D. Adrienko, G. J. Bodwell, K.

Mꢁllen, J. Am. Chem. Soc. 2007 , 129 , 1174 3 –1 1749.

[40] S. Hçger, J. Spickermann, D. L. Morrison, P. Dziezok, H. J. Rꢂder,

Macromolecules 1997 , 30 , 311 0 –3 111.

[21] J. A. Marsden, M. J. OꢄConnor, M. M. Haley, Org. Lett. 2004 , 6 ,

238 5 –2 388.

[41] T. C. Bedard, J. S. Moore, J. Am. Chem. Soc. 1995 , 117 , 1066 2 –

10671.

[22] M. Schmittel, P. Mal, Chem. Commun. 2008 , 96 0 –9 62.

[23] a) D. W. J. McCallien, J. K. M. Sanders, J. Am. Chem. Soc. 1995 , 117 ,

661 1 –6 612; b) K. Tomizaki, L. Yu, L. Wei, D. F. Bocian, J. S. Lind-

sey, J. Org. Chem. 2003 , 68 , 819 9 –8 207; c) S. Rucareanu, A. Schu-

wey, A. Gossauer, J. Am. Chem. Soc. 2006 , 128 , 339 6 –3 413; d) S.-H.

Jung, W. Pisula, A. Rouhanipour, H. J. Rꢂder, J. Jacob, K. Mꢁllen,

Angew. Chem. 2006 , 118 , 480 1 –4 806; Angew. Chem. Int. Ed. 2006 ,

45 , 468 5 –4 690; e) M. Hoffmann, C. J. Wilson, B. Odell, H. L. Ander-

son, Angew. Chem. 2007 , 119 , 318 3 –3 186; Angew. Chem. Int. Ed.

2007 , 46 , 312 2 –3 125.

[42] Detailed photophysical investigations on spoked wheels and their

precursors are in progress and will be reported elsewhere.

[43] Specifically, the term shape-persistence means that the average di-

ameter of the ring can be estimated from the effective contour

length of the phenylene–acetylene backbone divided by p: see, for

example, reference [35b].

[44] S. Hçger, K. Bonrad, A. Mourran, U. Beginn, M. Mçller, J. Am.

Chem. Soc. 2001 , 123 , 565 1 –5 659.

[45] The self-diffusion constant of an atom or a group of atoms can be

evaluated from the Mean Square Displacement during the MD run,

N

[24] D. Mçssinger, J. Hornung, S. Lei, S. De Feyter, S. Hçger, Angew.

Chem. 2007 , 119 , 692 6 –6 930; Angew. Chem. Int. Ed. 2007 , 46 , 680 2 –

6806.

P

1

d

and is defined as: D=_{6 N}lim _dt

h[r_i(t)ꢂr_i(0)]²i, in which t is the dy-

t!1

namics time and N the number^i¼o¹f atoms.

[46] a) S. Hçger, V. Enkelmann, Angew. Chem. 1996, 107, 2917–2919;

Angew. Chem. Int. Ed. Engl. 1996 , 34 , 271 3 –2 716; b) S. Hçger, D. L.

Morrison, V. Enkelmann, J. Am. Chem. Soc. 2002 , 124 , 673 4 –6 736;

c) S. Hçger, X. H. Cheng, A.-D. Ramminger, V. Enkelmann, A.

Rapp, M. Mondeshki, I. Schnell, Angew. Chem. 2005 , 117 , 286 2 –

2866; Angew. Chem. Int. Ed. 2005 , 44 , 280 1 –2 805.

[47] a) J. S. Higgins, H. C. Benoit, Polymers and Neutron Scattering, Clar-

endon, Oxford, 1994; b) L. Harnau, S. Rosenfeldt, M. Ballauff, J.

Chem. Phys. 2007 , 127 , 014901.

[25] J. Wu, M. D. Watson, L. Zhang, Z. Wang, K. Mꢁllen, J. Am. Chem.

Soc. 2004 , 126 , 17 7 –1 86.

[26] Compound 4 was synthesized according to literature methods; for

example: V. Hensel, A.-D. Schlꢁter, Liebigs Ann. 1997, 303–309.

[27] PEPPSI=pyridine-enhanced precatalyst preparation, stabilization,

and initiation; [1,3-Bis(2,6-diisopropyl-phenyl)imidazol-2-ylidene](3-

chloropyridyl)-palladium(II) dichloride: C. J. OꢄBrien, E. A. B.

Kantchev, C. Valente, N. Hadei, G. A. Chass, A. Lough, A. C. Hop-

kinson, M. G. Organ, Chem. Eur. J. 2006 , 12 , 474 3 –4 748.

[48] a) S. Rosenfeldt, N. Dingenouts, M. Ballauff, N. Werner, F. Vçgtle,

Macromolecules 2002 , 35 , 809 8 –8 105; b) M. Ballauff, C. N. Likos,

Angew. Chem. 2004 , 116 , 306 0 –3 082; Angew. Chem. Int. Ed. 2004 ,

43 , 299 8 –3 020; c) S. Rosenfeldt, N. Dingenouts, M. Ballauff, P. Lind-

ner, C. N. Likos, N. Werner, F. Vçgtle, Macromol. Chem. Phys. 2002 ,

203 , 199 5 –2 004.

[49] K. Henzler, S. Rosenfeldt, A. Wittemann, L. Harnau, S. Finet, T.

Narayanan, M. Ballauff, Phys. Rev. Lett. 2008 , 100 , 158301.

[50] C. N. Likos, S. Rosenfeldt, N. Dingenouts, M. Ballauff, P. Lindner,

N. Werner, F. Vçgtle, J. Chem. Phys. 2002 , 117 , 186 9 –1 877.

[51] The effective thickness has been estimated by plotting the atomic

density (weighted by the atomic mass) as a function of the atomic

position relative to the normal of the main molecular plane.

[52] a) J. P. Rabe, S. Buchholz, Science 1991 , 253 , 42 4 –4 27; b) D. M. Cyr,

B. Venkataraman, G. W. Flynn, Chem. Mater. 1996 , 8 , 160 0 –1 615;

c) S. De Feyter, F. C. De Schryver, Top. Curr. Chem. 2005 , 258 , 20 5 –

255; d) Y. Yang, C. Wang, Int. J. Nanotechnology 2007 , 4 , 4 –2 0.

[53] P. Sautet, Chem. Rev. 1997 , 97 , 109 7 –1 116.

[28] T. Zimmermann, G. W. Fischer, J. Prakt. Chem. 1987 , 329 , 97 5 –9 84.

[29] a) S. Hçger, S. Rosselli, A.-D. Ramminger, V. Enkelmann, Org. Lett.

2002 , 4 , 426 9 –4 272; b) S. Klyatskaya, N. Dingenouts, C. Rosenauer,

B. Mꢁller, S. Hçger, J. Am. Chem. Soc. 2006 , 128 , 315 0 –3 151.

[30] a) S. Hçger, K. Bonrad, J. Org. Chem. 2000 , 65 , 224 3 –2 245; b) K.

Becker, P. G. Lagoudakis, G. Gaefke, S. Hçger, J. M. Lupton,

Angew. Chem. 2007 , 119 , 352 0 –3 525; Angew. Chem. Int. Ed. 2007 ,

46 , 345 0 –3 455; c) G. Gaefke, S. Hçger, Synthesis 2008, 2155–2157.

[31] At present, we are developing a more rational assembly strategy re-

lying on larger but structurally related building blocks.

[32] GPC analysis did only show an envelope of the three peaks with a

polydispersity of only 1.07, similar to some monodisperse com-

pounds, which suggested the presence of pure sixfold coupling prod-

uct.

[33] If we assume that the three 1:4 coupled isomers are equally formed

during the reaction, we expect an acetylene signal ratio of 1:8:3

(Figure 3, bottom, a). By integration of the ¹H NMR signals we

found 2:8:3.3 (two/one/no neighboring void).

[54] a) G.-B. Pan, X. H. Cheng, S. Hçger, W. Freyland, J. Am. Chem.

Soc. 2006 , 128 , 421 8 –4 219; b) E. Mena-Osteritz, P. Bꢂuerle, Adv.

Mater. 2006 , 18 , 44 7 –4 51.

[34] a) C. Glaser, Chem. Ber. 1869, 2, 422–424; b) P. Siemsen, R. C. Liv-

ingston, F. Diederich, Angew. Chem. 2000 , 112 , 274 0 –2 767; Angew.

Chem. Int. Ed. 2000 , 39 , 263 2 –2 657.

[55] For a 2D, monolayered dehydrobenzoannulene host structure that

accepts spoked wheel 1 as a guest, see: K. Tahara, S. Lei, D. Mçs-

singer, H. Kozuma, K. Inukai, M. van der Auweraer, F. C. De

Schryver, S. Hçger, Y. Tobe, S. De Feyter, Chem. Commun. 2008,

33, 3897–3899.

[35] a) S. Hçger, A.-D. Meckenstock, H. Pellen, J. Org. Chem. 1997 , 62 ,

455 6 –4 557; b) S. Hçger, J. Polym. Sci. Part A 1999 , 37 , 268 5 –2 698;

c) S. Hçger, A.-D. Meckenstock, Tetrahedron Lett. 1998, 39, 1735–

1736; d) S. Hçger, A.-D. Meckenstock, Chem. Eur. J. 1999 , 5 , 168 6 –

1691; e) M. Fischer, S. Hçger, Eur. J. Org. Chem. 2003 , 44 1 –4 46;

2534

www.chemeurj.org

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Chem. Eur. J. 2009, 15, 2518 – 2535

Template Synthesis

FULL PAPER

[56] a) D. Rigby, H. Sun, B. E. Eichinger, Polym. Int. 1997 , 44 , 31 1 –3 30;

b) H. Sun, J. Phys. Chem. B 1998 , 102 , 733 8 –7 364; c) S. W. Bunte,

H. Sun, J. Phys. Chem. B 2000 , 104 , 247 7 –2 489.

[57] a) P. Lindner, R. P May, P. A. Timmis, Physica B 1992 , 180/181 ,

96 7 –9 72; b) K. Lieutenant, P. Lindner, R. Gꢂhler, J. Appl. Crystal-

logr. 2007 , 40 , 105 6 –1 063.

lauff, ChemPhysChem 2006 , 7 , 209 7 –2 104; c) P. Lindner in Neu-

trons, X-Rays and Light: Scattering Methods Applied to Soft Con-

densed Matter, Scattering Experiments: Experimental Aspects, Initial

Data Reduction & Absolute Calibration (Eds.: P. Lindner, T. Zemb),

Elsevier, Amsterdam, 2002, pp. 23–48.

[58] a) S. Rosenfeldt, N. Dingenouts, M. Ballauff, N. Werner, F. Vçgtle,

P. Lindner, Macromolecules 2002 , 35 , 809 8 –8 105; b) S. Rosenfeldt,

E. Karuk, M. Lehmann, H. Meier, P. Lindner, L. Harnau, M. Bal-

Received: September 19, 2008

Published online: February 3, 2009

Chem. Eur. J. 2009, 15, 2518 – 2535

ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemeurj.org

2535

Article Doi

Two-dimensional oligo(phenylene-ethynylene-butadiynylene)s: All-covalent nanoscale spoked wheels

DOI: 10.1002/chem.200801939

Source and publish data:

Authors:

Article abstract of DOI:10.1002/chem.200801939

Full text of DOI:10.1002/chem.200801939

Products guided by the article

R&D Labs maybe for 957788-32-8

Relevant to this article

Hot Product