Scanning tunnelling microscopy of a foldamer prototype at the liquid/solid interface: Water/Au(111) versus 1-octanol/graphite

Article abstract of DOI:10.1039/b607660m

We report the design and synthesis of a catechol based foldamer containing amide functionalized alkyl chains, and its monolayer formation at the liquid/solid interface. By scanning tunnelling microscopy, both at the 1-octanol/graphite interface as well as

Full text of DOI:10.1039/b607660m

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Scanning tunnelling microscopy of a foldamer prototype at the
liquid/solid interface: water/Au(111) versus 1-octanol/graphite
Andrey S. Klymchenko,wa Norbert Schuurmans,b Mark van der Auweraer,a
Ben L. Feringa,b Jan van Esch*b and Steven De Feyter*a
Received (in Montpellier, France) 30th May 2006, Accepted 9th August 2006
First published as an Advance Article on the web 4th September 2006
DOI: 10.1039/b607660m
We report the design and synthesis of a catechol based foldamer containing amide functionalized
alkyl chains, and its monolayer formation at the liquid/solid interface. By scanning tunnelling
microscopy, both at the 1-octanol/graphite interface as well as at the water/Au(111) interface, the
self-assembly has been investigated with submolecular resolution, demonstrating clear solvent
dependent effects on the conformation of the ‘foldamer’ and its monolayer formation.
importance for the conformational preference. Recently, we
Introduction
reported on the successful design of 2D turn mimics based on a
catechol bis-amide moiety, that fold into b-turn mimics at the
interface of HOPG and 1-octanol as demonstrated by STM.7
However, it remains unclear to which extend the preference for
the folded conformation is influenced by noncovalent interac-
tions with the substrate and the solvent.
Control of the lateral assembly and spatial arrangement of
micro- and nano-objects at interfaces is often a prerequisite
when it comes to potential applications in the field of nano-
science and technology. Molecules are the smallest objects
which can be ordered by self-assembly on surfaces. In order to
exploit such molecular assemblies for catalysis purposes,1 as
(reactive) templates,2 or in molecular electronics3 to name a
few possible applications, it is absolutely crucial to understand
the factors which lead to the two-dimensional self-assembly
(often 2D crystallization) of molecules in order to control the
outcome of this spontaneous process and to tune the desired
properties.
Scanning tunnelling microscopy (STM) proves to be an
excellent method to probe such layers because of the high
spatial resolution (submolecular), its sensitivity to different
chemical functionalities, and the non-invasive nature of the
experiment.8 Typically, the substrates are conductive and
atomically flat (e.g. highly oriented pyrolytic graphite
(HOPG), gold, etc.). Evidently, the choice of the substrate
affects the molecule–substrate interaction and has an influence
on the 2D ordering. In addition to the choice of the substrate,
also the solvent plays a key role.9 The solvent can be tuned in
function of the particular solute and/or substrate. In case of
organic solvents, the solvent has typically a low vapor pres-
sure, is non-conductive, and shows a lower affinity for the
substrate than the solute. Such physisorbed systems show
adsorption–desorption dynamics at the liquid/solid interface
which promotes the repair of defects in the self-assembled
layers.10 Additional control of the monolayer formation can
be achieved under potential control in aqueous solutions. In
case of physisorbed apolar molecules hydrophobic interac-
tions play a key role, so that the self-assembly could be
significantly different as compared to that in organic solvents.
Under electrochemical conditions, adsorbate–substrate inter-
actions can be modulated by the surface charge density.
Electrochemical environments offer therefore additional pos-
sibilities to control surface dynamics and monolayer structure
via the surface charge and to image those structures by
means of electrochemical scanning tunnelling microscopy
(EC-STM).11
Particularly interesting is the self-assembly of (organic)
molecules at the liquid/solid interface which often leads to
the formation of physisorbed monolayers.4 In this respect, the
formation of 2D adlayers on surfaces can be compared not
only with 3D crystallization but also with, for instance, the
formation of 1D systems based upon supramolecular interac-
tions. A first step in the spatial control of molecules on
surfaces is controlling their conformation.5 In many biological
systems for example, the exact positioning of functional
groups is of key importance for the correct operation of large
molecular ensembles. Currently, a lot of research is dealing
with artificial foldamers and their self-assembly.6 These folda-
mers are designed to fold into a predefined conformation due
to specific intramolecular interactions and conformational
restrictions. In many instances folding and unfolding can be
controlled by tuning of the solvent, showing that also inter-
actions between the foldamers and the solvent are of crucial
a Department of Chemistry, Laboratory of Photochemistry and
Spectroscopy, Institute of Nanoscale Physics and Chemistry,
Katholieke Universiteit Leuven, Celestijnenlaan 200-F, 3001 Leuven,
Belgium.
In this contribution, we discuss the self-assembly of a
catechol derivative (Fig. 1), by concept related to a previous
study,7 which has been designed to fold on a surface and
compare its self-assembly at two different liquid/solid inter-
faces. The folding of molecule 1 is based upon a functionalized
catechol unit. Alkyl chains guarantee the interaction with the
E-mail: Steven.DeFeyter@chem.kuleuven.be; Fax: +32 16 327990
b Stratingh Institute and Material Science Center, University of
Groningen, Nijenborg 4, 9747 AG Groningen, The Netherlands.
E-mail: j.h.van.esch@rug.nl; Fax: +31 50 363 4296
w Present address: Department of Pharmacology and Physical Chem-
istry, UMR 7175, Institut Gilbert Laustriat, Universite Louis Pasteur
´
(Strasbourg I), BP 60024, 67401 Illkirch, France.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
1420 | New J. Chem., 2006, 30, 1420–1428

View Article Online
Fig. 1 Chemical structure of 1, indicating two conformations (A and B) leading to a folded structure. Intra- and intermolecular hydrogen bonding
between amide groups is possible in the physisorbed monolayers. (C) Model of a row of 1 on a sheet of graphite based upon molecular mechanics
calculations.
substrate and carefully positioned amide groups should both
optimize intramolecular hydrogen bonding, leading to an ideal
folded structure, and intermolecular hydrogen bonding, lead-
ing to 2D crystals. However, in order to realize such assem-
blies, it is crucial that the molecules indeed adopt the folded
conformation they have been designed for and that both alkyl
chains of the molecule are adsorbed on the surface.
Via molecular modeling, it was established that for this
molecule conformations allowing for intramolecular H-bond
formation are favored and that also intermolecular stabiliza-
tion via van der Waals interactions and H-bonding can be
realized (Fig. 1). Two conformers are possible which inter-
convert by a conformational equilibrium resulting in an over-
all difference in the length of the molecules of 1.3 A (2.53 nm
(a) vs. 2.4 nm (b) in Fig. 1). This small but nevertheless
significant difference is due to the dissymmetry in the length
of the two groups attached to the catechol moiety.
Results and discussion
Compound 1, with C3 spacers and C12 and C13 tails has been
synthesized according to Scheme 1. The essential intermediate
is highlighted in the scheme. Catechol was reacted with bromo
chloropropane to give a mixture of mono- and difunctional
chlorides. Extensive column chromatography was necessary to
isolate and purify the mono-chloride 2. It was eventually
obtained in 27% yield. Compound 2 was easily converted to
the azide 3, in a modest yield of 53%. Subsequently this
compound was converted to the amine (4, via catalytic hydro-
genation, 65% yield). Protection of this amine with benzyl
chloroformate was cumbersome and required extensive col-
umn chromatography again. The Z-protected amine 5 was
eventually obtained in 44% yield. Compound 5 was coupled to
methyl bromobutyrate as before, appending the second arm,
in 20% yield. The ester 6 was then deprotected by saponifica-
tion (82%). The Z-protected mono-acid 7 was crystallized. A
more advanced protocol for generating the amides was used as
described before.12 The mono-acid was activated with EDC
and HOBT, and then reacted with dodecylamine. This reac-
tion generated compound 8 in 35% yield. The Z-protected
amine 8 could now be deprotected by catalytic hydrogenation
and subsequently reacted with myristic acid, preactivated with
EDC/HOBT. This reaction produced compound 1 in 30%
yield. All compounds were characterized by 1H-NMR,
13C-NMR and MS.
Design and synthesis
Despite our previous success in designing a turn element
involving the catechol unit,7 a new design and simpler syn-
thetic approach was followed which should allow a less
cumbersome approach for the synthesis of extended foldamers
at a later stage. Therefore, from a synthetic point of view, a
coupling scheme involving only one building block, with a
(protected) acid and a (protected) amine in the same building
block is an attractive alternative for the orthogonal protec-
tion–deprotection protocol invoked by the previously reported
procedure.7
The new procedure bears a resemblance to conventional
peptide chemistry. In a way the proposed foldamers can be
considered as peptidomimetics. In order to be able to follow
the procedures of peptide chemistry, a slight modification of
the previously reported turn mimic7 is required.
In the previous foldamer, both amide groups were equiva-
lent (both amides having the carbonyl group connected to the
alkyl linker). As a result, to achieve optimized intramolecular
hydrogen bonding and alkyl chain conformation, both alkyl
chains linking the catechol unit with the amide group differed
in length. However, as a result of the new synthetic approach,
both amide groups are non-equivalent (they have a reverse
orientation), and dissymmetry with respect to the alkyl linkers
is no longer an absolute requirement.7 Both spacers can be of
the same length. The total alkyl chain length is also of
importance due to the stabilizing effect at the level of the
interaction between the molecule and the substrate. Com-
pound 1 was regarded as a good candidate.
STM experiments at the liquid/solid interface
The 1-octanol/graphite interface. To reveal the conformation
and self-assembly of this molecule, in a first stage, the molecule
was dissolved in 1-octanol and a drop of this solution was
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 1420–1428 | 1421

View Article Online
Scheme 1 Synthesis of compound 1.
applied on a freshly cleaved surface of HOPG. After a while,
highly ordered patterns were observed (Fig. 2A). Typically, the
2D patterns showed V shape structures. An accurate analysis
of the patterns was possible by calibrating the image based
upon the atomic structure of the substrate underneath, which
was revealed by lowering the bias voltage, leading to a
decreased distance between the tip and the substrate. The
distance between the centers of two adjacent rows is 4.5 Æ
0.1 nm (double arrow). The distance between equivalent points
parallel to the lamella axis is 0.48 nm. The angle between the
‘legs’ of the V-shaped features and the lamella axis is about
651. These results are in line with the adsorption of 1 at the
liquid/solid interface according to the model indicated in Fig.
2B. The brightest features in the STM image correspond to the
location of the aromatic catechol groups (at the position of the
arrow heads) while the ‘legs’ correspond to the alkyl chains
containing the amide functions. The amide functions are not
well resolved. However, according to the model, intermolecu-
lar hydrogen bonding is possible.
lamella rows. This interpretation is supported by the presence
of streaky regions between rows of alkyl chains, suggesting a
high degree of mobility which involves the catechol unit and a
desorbed alkyl chain. It is very unlikely that solvent molecules
(1-octanol) are co-adsorbed, as the length of the ‘legs’ does not
fit with the size of the 1-octanol molecules.13
In view of previous results,7 the lack of folded structures was
unexpected. Also modeling (in the absence of solvent) of 5 by 2
clusters on graphite, allowing the alkyl groups to be in epitaxy
with the substrate, predicts the foldamer conformation to be
thermodynamically more stable than the V-shaped one. As
monolayer formation is also affected by kinetic effects, model-
ing studies have only a limited predictive power. In addition,
the solvent, which can not be taken into account in the
calculations, can affect the molecular conformation and inter-
molecular interactions, both in solution and at the liquid/solid
interface. Overall, the rate for ordering of these molecules at
the 1-octanol/graphite interface was slow (41 h) and in
general, the self-assembled layers could be easily destroyed
by the STM tip.
For this image type, another possibility was considered
where only one alkyl chain for a given molecule was adsorbed
while the other one was pointing into the liquid phase. This
option, however, was ruled out as far as head-to-head or tail-
to-tail arrangements are concerned, based upon the lack of
agreement in the dimensions of the molecular model and the
STM images.
The water/Au(111) interface. Clearly, the 1-octanol/graphite
interface is not the appropriate environment to ‘stimulate’
folding of 1. Motivated by the solvent dependence of the
stability of bisurea-based organogels,14 revealing an increase
in the thermal stability and strength of the gels in going from
1-octanol to 1-propanol due to solvophobic effects, we decided
to use ethanol as solvent. Ethanol, however, is not a good
solvent for performing STM experiments at the liquid/solid
interface due to its high vapor pressure and residual water
content.15 Despite this disadvantage, ethanol can still be used
to apply the molecules on a substrate.16 Thus, the molecules
However, a number of STM images, regularly observed
during several measuring sessions at various sites on the
substrate, reveal another type of arrangement where a one-
chain adsorption is most likely (Fig. 3): the distance between
equivalent points in adjacent rows measures B2.6 nm and the
alkyl chains separated by 0.48 nm run perpendicular to the
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
1422 | New J. Chem., 2006, 30, 1420–1428

View Article Online
Fig. 3 (A) STM image of a monolayer of 1 at the 1-octanol/graphite
Fig. 2 2D ordering of 1 at the 1-octanol/graphite interface. (A) STM
image. It = 0.6 nA, Vt = À0.45 V. (B) Tentative molecular model.
interface with only one alkyl chain adsorbed per molecule. It
0.6 nA, Vt = À0.30 V. (B) Tentative molecular model.
=
were drop-cast from an ethanol solution on Au(111) and the
sample was rinsed several times with ethanol to remove the
excess of 1 prior to transferring the sample to the electro-
chemical cell of the EC-STM , which was covered by 0.1 M
HClO4 at À100 mV (vs. SCE electrode). Au(111) was selected
as substrate due to its proven performance in EC-STM
experiments.16 Under these conditions, the molecules are
completely insoluble and are fully adsorbed on the gold
surface.
shown). All the EC-STM measurements were performed at a
work potential of À100 mV (vs. SCE), which was fixed after
recording the corresponding CV curves. Therefore, we expect
that the ordered layers observed in the present study are
formed after the potential was fixed at À100 mV.
In comparison to the 1-octanol/graphite interface, mono-
layers are much more stable and less dynamic (at least at the
work potential of À100 mV vs. SCE), most likely due to the
hydrophobic effect.
Cyclic voltammetry (CV) measurements show an increase of
the current at positive voltages as well as shifts of the anion
adsorption–desorption peaks (in the region from +100 to
+400 mV vs. SCE) to higher positive work potentials (Fig.
4), which confirms the presence of the adsorbate on Au(111).
We did not perform scans at higher work potentials in order to
avoid possible electrochemical oxidation of the catechol resi-
due of molecule 1.
STM images reveal several two-dimensional polymorphs.
One of them resembles the V-shaped structures observed at the
1-octanol/HOPG interface. Indeed, the lamellae are composed
of bright features and each of them has two chains (Fig. 5A
and Fig. 6A). However, the angle between the chains is larger,
ca. 1501 and the lamellae are not straight but wavy.17
Another monolayer structure (Fig. 5B and 6B), which was
also frequently observed on the same sample but at different
locations, is characterized by perfectly aligned lamellae formed
by perpendicularly oriented alkyl chains. Moreover, we ob-
serve single and double rows of bright spots. The distance
By EC-STM, ordered structures were observed only at
potentials below +100 mV (vs. SCE), while at higher poten-
tials disordered aggregates of molecules were detected (not
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 1420–1428 | 1423

View Article Online
In addition to the head-to-head arrangement, we also observe
head-to-tail arrangements of folded molecules (Fig. 5C, D and
6C). The latter indicates the absence of a specific interaction
between lamellae. Some alkyl chains appear longer than others
and some lamellae differ slightly in width by about 0.15 nm,
which corresponds well to the difference expected based upon
the two possible folded conformations (Fig. 1). Despite this
observation, no firm conclusions can be drawn regarding the
specific conformation of the molecules, except for the fact that
upon folding, intramolecular hydrogen bonding is optimized.
No indication was found that molecules adsorb with only
one alkyl chain on the substrate. The molecules are always
adsorbed with both alkyl chains adsorbed on the substrate,
often adopting a folded conformation. This folded conforma-
tion can intuitively be understood due to the solvophobic
effect of the aqueous subphase. By bringing these molecules,
which are water-insoluble, at the water/gold interface, we have
restricted the number of unfolded conformations. Note that at
this stage we have not investigated specifically the influence of
the substrate (graphite versus Au(111)) which should, in gen-
eral, not be neglected. For instance, the angle between the
alkyl chains of the V-like structures is larger on Au(111) and
molecules are less well aligned. A full investigation of different
Fig. 4 CV curves of bare Au(111) in 0.1 mM HClO4 (solid line) and
in presence of adsorbed molecule 1 (dashed line). Note that both
curves were obtained in different experimental sessions.
between the bright spots (catechol groups) is B1.0 nm, which
is two times as large as the distance between the alkyl chains.
This kind of packing is clear evidence that the molecules fold
and form lamellae via a head-to-head arrangement (Fig. 6B).
Fig. 5 EC-STM images of molecular ordering of 1 on Au(111) in 0.1 mM HClO4. (A) VW = À100 mV (vs. SCE), Vt = À500 mV, It = 1 nA.
(B–D) VW = À100 mV (vs. SCE), Vt = À600 mV, It = 1 nA.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
1424 | New J. Chem., 2006, 30, 1420–1428

View Article Online
tion observed by EC-STM. Prior to imaging, the formation of
an adsorbate layer is checked by taking a CV curve. During
the recording of the voltammogram, at high positive poten-
tials, the highly ordered phase is destroyed possibly also
affecting the molecular conformation.
Conclusion
In conclusion, the water/Au(111) interface proves to be a good
environment to support the folding of water-insoluble catechol
type foldamers and their ordering into 2D patterns, which has
been revealed by EC-STM under potential control. Folding of
the present molecule is driven by a combination of solvopho-
bic and H-bonding interactions, which is analogous to folding
of biomolecules in aqueous media, and is as such promoted at
the water/substrate interface. In general, electrified surfaces
hold great promises to (affect) the conformation/ordering of
molecules, not only water-soluble ones, but also water-inso-
luble ones, which opens many possibilities for applications.
Experimental
Synthesis
All solvents were dried according to standard procedures.
Starting materials were purchased from Aldrich or Acros.
1H-NMR spectra were recorded on a Varian VXR-400 spec-
trometer (at 100.57 MHz) in CDCl3 chemical shifts are given
in ppm relative to CDCl3 (7.24). 13C-NMR spectra were
recorded on a Varian VXR-400 spectrometer (at 400 MHz)
in CDCl3, chemical shifts are given relative to CDCl3 (77). The
splitting patterns in the 1H-NMR spectra are designated as
follows: s (singlet), d (doublet), t (triplet), m (multiplet), br
(broad). Melting points were measured on Stuart scientific
SMP1 apparatus. HRMS was performed on a JEOL JMS
600H spectrometer in EI+ ionization mode.
General procedure for the amide couplings. The carboxylic
acids were dissolved in DMF. To the solutions was added one
equiv. of diisopropylethyl amine. The mixtures were stirred
and were cooled to 0 1C. HOBT (hydroxybenzotriazole) and
EDC (a water soluble version of DCC) were added in equi-
molar amounts. The mixtures were stirred for 10 min and the
amines were added. The mixtures were stirred for 24 h at room
temperature. The mixtures were evaporated in vacuo at 40 1C.
CH2Cl2 and water were added. The organic phase was ex-
tracted with water, dilute HCl (aq) and again water. The
organic phase was dried on Na2SO4 and evaporated in vacuo.
The crude products were further purified by column chroma-
tography (silica, 2% MeOH in CHCl3). The amides were the
first fractions to come off the column.
Fig. 6 Zoomed STM images of (A) Fig. 5A, (B) Fig. 5B, and (C) Fig.
5C with tentative molecular models of the molecular packing.
solvent-substrate combinations will be necessary to get a
conclusive answer on the relative importance of solvent versus
substrate. However, we believe that it is mainly the choice of
solvent—the solvophobic effect—which for this particular
system is the driving force for molecular folding. It is unlikely
that a ‘conformational memory effect’ (solvophobic interac-
tion in ethanol) alone is responsible for the folded conforma-
General procedure for the Z-deprotections. The Z-protected
amines were dissolved in MeOH. A balloon filled with H2 gas
was mounted on the flask. A small amount of Pd–C (10%) was
added to the solution. The flask was evacuated, filled with
hydrogen, again evacuated, and brought under H2 atmosphere
again. The mixtures were stirred for 12 h. Upon removal of the
balloon, the catalyst was filtrated over celite, and the filtrate
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 1420–1428 | 1425

View Article Online
was evaporated. The amines were used without further
purification.
organic layers were dried and concentrated in vacuo. The
compound was purified by column chromatography (silica,
CHCl3 + 5% MeOH). Yield: 10 g (33 mmol, 46%). 1H-NMR
(CDCl3): dH 7.32 (5H, m, HAromatic of the benzyl group),
6.73–6.85 (4H, m, HAromatic of the catechol moiety), 5.43
(1H, br, OH), 5.02 (2H, s, NHCO2CH2C6H5),3.96 (2H, t,
2-(3-Chloro-propoxy)-phenol (2). Catechol (100 g, 0.9 mol)
was dissolved in acetone (500 ml) and the solution was stirred
at reflux. Pre-dried (120 1C) K2CO3 (100 g, 0.72 mol) was
slowly added to the stirred solution, which got a dark color.
The mixture was stirred for 10 min and bromo chloropropane
(120 g, 0.76 mol) was added dropwise. The mixture was stirred
at reflux for 48 h. The mixture was filtered hot and the residue
was concentrated in vacuo. CHCl3 (250 ml) was added, this
mixture was stirred and filtrated again. The crude product was
purified by trituration with pentane. The product separated as
a yellow oil. This was collected and further purified by column
chromatography (silica, CHCl3). Yield: 45 g (0.24 mol, 27%).
1H-NMR (CDCl3): dH 6.80–6.87 (4H, m, HAromatic)(1H, s,
3
3J = 5.8 Hz, OCH2), 3.31 (2H, t, J = 4.8Hz, CH2NHCO2),
3
1.87 (2H, t, J = 6.2 Hz, CH2CH2CH2). 13C-NMR (CDCl3):
dC156.7, 148.1, 145.7, 128.3, 128.0, 121.7, 119.9, 115.1, 112.7,
67.3, 66.1, 39.1, 29.4. MS (EI+): m/z 301 (M+, 10%), 192 (20).
HRMS: calcd for C17H19NO4: 301.131, found 301.139.
4-[2-(3-Benzyloxycarbonylamino-propoxy)-phenoxy]-butyric
acid methyl ester (6). The Z-protected amine 5 (9 g, 30 mmol)
was dissolved in acetone (250 ml). The mixture was stirred and
pre-dried K2CO3 (6.2 g, 45 mmol) was added portion-wise.
The mixture was stirred for 30 min at reflux temperature and
methyl bromobutyrate (6.5 g, 36 mmol, 1.2 eq.) was added
slowly. The mixture was stirred at reflux for 72 h. The mixture
was filtered hot and the residue was concentrated in vacuo.
CHCl3 was added, and this mixture was stirred and filtrated
again. The filtrate was concentrated in vacuo and subjected to
extensive column chromatography (3Â: silica, CHCl3; EtOAc;
CHCl3 + 5% MeOH + 1% NH3). Yield: 2.4 g (6 mmol,
20%). 1H-NMR (CDCl3): dH7.35 (5H, m, HAromatic of the
benzyl group), 6.88–6.92 (4H, m, HAromatic of the catechol
3
3
OH), 4.16 (2H, t, J = 5.9 Hz, OCH2), 3.68 (2H, t, J = 6.2
Hz, CH2Cl), 2.23 (2H, m, CH2CH2CH2). 13C-NMR (CDCl3):
dC 121.8, 120.2, 114.7, 111.9, 65.5 (OCH2), 41.3, 32.0. MS
(CI+): m/z 187 (M+, 29), 110 (100). HRMS: calcd for
C9H11ClO2: 186.045, found 186.051.
2-(3-Azido-propoxy)-phenol (3). Compound 2 (44 g, 0.24
mol) was dissolved in DMSO (300 ml). The mixture was
stirred and NaN3 (20 g, 0.3 mol) was added in small portions.
The mixture was stirred at 60 1C for 12 h. The mixture was
then poured into water (500 ml), and extracted with EtOAc
(2Â). The organic layers were washed with water (2Â), dried
(Na2SO4) and concentrated in vacuo to yield the product as a
yellow oil. It was purified by trituration with pet. ether. Yield:
3
group), 5.02 (2H, s, NHCO2CH2C6H5) 4.08 (2H, t, J = 5.7
Hz, OCH2C2H4NHCO2), 3.97 (2H, t, 3J
= 5.8 Hz,
OCH2C2H4CO2Me), 3.65 (3H, s, CO2CH3), 3.42 (2H, m,
3
CH2NHCO2), 2.26 (2H, t, J = 7.3 Hz, CH2CO2CH3), 1.84
(4H, m, 2Â CH2CH2CH2) 13C-NMR (CDCl3): dC156.0, 149.1,
145.8, 128.3, 127.7, 122.2, 117.8, 115.2, 72.6, 67.5, 66.2, 52.2,
39.1, 29.4, 25.8.
1
25 g (0.13 mol, 53%). H-NMR (CDCl3): dH 6.84–6.91 (4H,
m, HAromatic), 5.69 (1H, s, OH), 4.12 (2H, t, 3J = 6.0 Hz,
OCH2), 3.50 (t, 3J = 6.6 Hz, 2H, CH2N3), 2.07 (m, 2H,
CH2CH2CH2). 13C-NMR (CDCl3): dC 121.6, 120.3, 117.2,
111.7, 68.2, 44.0, 32.0. HRMS: calcd for C9H11N3O2
193.085, found 193.089.
4-[2-(3-Benzyloxycarbonylamino-propoxy)-phenoxy]-butyric
acid (7). The ester 6 (2.5 g, 6 mmol) was dissolved in MeOH
(50 ml) and LiOH (200 mg) was added. The mixture was
stirred for 2 h at room temperature. The mixture was con-
centrated in vacuo and CH2Cl2 was added. The resulting
suspension was acidified with HCl (g), in situ generated by
carefully adding H2SO4 to HCl (aq). The Li salts precipitated
and were removed by filtration: the filtrate was concentrated
in vacuo to give the acid as a white solid. Yield: 1.9 g
(4.9 mmol, 82%). Mp 195–196 1C (from CHCl3). 1H-NMR
confirmed removal of the methoxy group. MS (EI+): m/z 387
(M+, 5%), 192 (51). HRMS: calcd for C21H25NO6 387.168,
found 387.171.
2-(3-Amino-propoxy)-phenol (4). The azide 3 (24 g, 0.12 mol)
was dissolved in MeOH (100 ml) in a two-necked flask
equipped with a balloon filled with H2. A small amount (100
mg) of palladium on carbon (Pd–C, 10%) was added. After
evacuating, the mixture was stirred under H2-atmosphere for
12 h at room temperature. The catalyst was removed by
filtration over celite and the filtrate was concentrated in vacuo.
The amine was obtained as an oil, that waxified on standing.
Yield: 13 g (78 mmol, 65%). 1H-NMR (CDCl3): dH 6.81–6.92
(4H, m, HAromatic), 5.81(2H, br, OH) 4.01 (2H, t, 3J = 5.1 Hz,
OCH2), 3.02 (2H, m, CH2NH2), 1.96 (2H, m, CH2CH2CH2).
13C-NMR (CDCl3): dC 125.0, 120.3, 119.7, 117.7, 73.0, 39.8,
31.4.
{3-[2-(3-Dodecylcarbamoyl-propoxy)-phenoxy]-propyl}-car-
bamic acid benzyl ester (8). Compound 7 (mono-Z protected
acid, 200 mg, 0.5 mmol) was dissolved in DMF (20 ml). The
solution was stirred and diisopropylethylamine (DIEA,
Hunig’s base, 100 ml) was added. The solution was stirred
for 10 min at room temperature. EDC (1-(3-dimethylamino-
propyl)-3-ethylcarbodiimide hydrochloride, 100 mg, 0.55
mmol) and HOBT (hydroxybenzotriazole, 90 mg, 0.6 mmol)
were added and the mixture was stirred for 30 min. Dodecy-
lamine (100 mg, 0.55 mmol) was added and the mixture was
stirred for 24 h at room temperature. The mixture was con-
centrated in vacuo at 80 1C and CHCl3 (50 ml) was added. This
[3-(2-Hydroxy-phenoxy)-propyl]-carbamic acid benzyl ester
(5). The amine 4 (12 g, 72 mmol) was dissolved in THF–H2O
(5 : 1, 200 ml). The mixture was cooled to 0 1C and KOH (30 g,
0.2 mol, excess) was added. The mixture was stirred for 5 min
and benzyl chloroformate (12.5 ml, 15 g, 85 mmol, 1.2 eq.) was
added drop-wise. The mixture was stirred for 2 h at 0 1C, and
allowed to stand overnight at room temperature. Water (100
ml) was added, and the organic layer was separated. The
aqueous layer was extracted with CH2Cl2 (3Â). The combined
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
1426 | New J. Chem., 2006, 30, 1420–1428

View Article Online
solution was extracted with water, HCl (aq. 0.6 M) and again
water. The aqueous layers were extracted with CHCl3. The
combined organic layers were dried (Na2SO4) and concen-
trated in vacuo. The crude product was purified by column
chromatography (silica, CHCl3 + 5% MeOH). Yield: 90 mg
(constant height). The measured tunnelling currents are con-
verted into a gray scale: black (white) refers to a low (high)
measured tunnelling current. For analysis purposes, recording
of a monolayer image was followed by imaging the graphite
substrate underneath under the same experimental conditions,
except if stated otherwise.
1
(0.16 mmol, 35%). H-NMR: dH 7.26–7.24 (5H, m, HAromatic
of the benzyl group), 6.77–6.81 (4H, m, HAromatic of the
EC-STM experiments have been performed using a home-
built STM setup designed and constructed in the group of
Prof. Klaus Wandelt (University of Bonn).18 The instrumental
design allows performing simultaneously cyclic voltametry
(CV) and STM measurements. A platinum wire was used as
a pseudo-reference electrode, while all potentials were rescaled
to the saturated calomel electrode (SCE) reference electrode.
Prior to each STM experiment, a Au(111) single crystal
(MaTeck company, Julich, Germany) was electrochemically
etched and flame annealed. For this purpose, the Au(111)
crystal was immersed into 0.1 M sulfuric acid and an anodic
potential of 10 V was applied between the crystal and a
platinum foil for about 30 s. Then, after rinsing with Milli-Q
water (Milli-Q purification system 418 MO cm) the sample
was immersed into 0.1 M hydrochloric acid in order to reduce
the Au(111) surface. Again, the Au substrate was rinsed with
Milli-Q water and flame annealed for another 2 min. To
deposit 1 on the gold substrate, a drop of an ethanol solution
of 1 (1 mg mlÀ1) was placed on the Au(111) surface. After
solvent evaporation the surface was thoroughly rinsed with
ethanol and dried. EC-STM imaging was started after com-
pletion of the in situ CV measurements. The STM tips were
electrochemically etched from a 0.25 mm tungsten wire in 2 M
KOH solution and subsequently isolated by passing the tip
through a drop of hot-glue. The EC-STM measurements were
performed in 0.1 M perchloric acid (HClO4) as supporting
electrolyte, which was deoxygenated with argon gas one hour
before use. This electrolyte solution does not hydrolyze the
amide bonds of 1 adsorbed on gold. No hydrolysis products
were observed in the STM images. All EC-STM imaging was
performed at a work potential (VW) of À100 mV vs. SCE
electrode. At positive work potentials only a disordered phase
was observed.
catechol group), 5.99 (1H, br, CONH), 5.98 (1H, br,
3
NHCO2CH2), 5.00 (2H, s, NHCO2CH2C6H5), 3.98 (2H, t, J
= 5.6 Hz, OCH2C2H4NHCO2), 3.89 (2H, t, 3J = 5.7 Hz,
OCH2C2H4CONH), 3.38 (2H, m, CONHCH2), 3.05 (2H, m,
CH2NHCO2), 2.19 (2H, t, 3J = 7.3 Hz, CH2CONH), 1.94
(4H, m, 2Â CH2CH2CH2), 1.60 (2H, m, CONHCH2CH2),
3
1.16 (18H, m, (CH2)9), 0.82 (3H, t, J = 6.4 Hz, CH2CH3).
13C-NMR: dC 172.7, 156.8, 148.9, 148.5, 136.8, 128.8, 128.7,
121.8, 121.2, 113.6, 113.4, 67.8, 66.9, 65.4, 40.2, 39.7, 33.0,
32.2, 30.0, 29.9, 29.8, 29.8, 29.7, 29.3, 28.0, 27.2, 26.8, 25.4,
23.0, 14.4. MS (EI+): m/z 554 (M+, 1%), 254 (100).
Tetradecanoic acid {3-[2-(3-dodecylcarbamoyl-propoxy)-phe-
noxy]-propyl}-amide (1). Compound 8 (80 mg) was dissolved in
MeOH (20 ml). The solution was placed in a 2-necked flask,
mounted with a balloon filled with H2 gas. Pd–C (10%, 50 mg)
was added. After evacuation the solution was stirred for 12 h.
under H2 atmosphere. The catalyst was removed by filtration
1
over celite, the filtrate concentrated in vacuo. H-NMR con-
firmed removal of the Z-group. Yield: 60 mg. The free amine
was coupled to myristic acid (60 mg) as described above,
activating the acid with EDC (60 mg), HOBT (50 mg) and
DIEA (60 ml). Yield: 25 mg (0.04 mmol, 30% over 2 steps).
1H-NMR: dH 6.85–6.95 (4H, m, HAromatic of the catechol
3
3
moiety), 4.04 (2H, t, J = 3.4 Hz, OCH2), 4.02 (2H, t, J =
2.6 Hz, OCH2), 3.51 (2H, m, OC2H4CH2NHCO), 3.19 (2H, m,
3
CONHCH2), 2.45 (2H, t, J = 8.3 Hz, OC2H4CH2CONH),
2.16 (2H, m, NHCOCH2), 1.98 (4H, t, 3J = 5.8 Hz, 2Â
CH2CH2CH2), 1.58 (2H, m, CONHCH2CH2), 1.43 (2H, m,
NHCOCH2CH2), 1.22 (m, 38H, (CH2)9 + (CH2)10), 0.86 (6H,
t, 3J = 6.4 Hz, 2Â CH2CH2CH3). 13C-NMR: dC 179.1, 174.7,
149.6, 122.6, 122.5, 115.0, 114.9, 69.2, 68.9, 38.9, 37.8, 35.1,
33.0, 31.5, 30.7, 30.6, 30.6, 30.5, 30.4, 30.4, 30.3, 30.2, 29.9,
26.9, 25.9, 25.6, 23.7, 15.1. MS (Maldi-TOF): calcd. 631,
found: 631. HRMS: calcd for C39H70N2O4 630.984, found
630.991.
Tunnelling bias (Vt) and current (It) parameters were ad-
justed to achieve the highest possible resolution: Vt = À0.50
to À0.60 V, It = 1.0 nA for EC-STM experiments at the
water/gold interface; Vt = À0.30 to À0.50 V, It = 0.60 nA for
STM experiments at the 1-octanol/graphite interface.
Scanning tunnelling microscopy
STM images were analyzed using WSxM 4.0 (Nanotec
Electronica S. L.) and SPIP 4.1 (Image Metrology) software.
STM experiments at the 1-octanol/graphite interface were
performed using a Discoverer Scanning Tunnelling Micro-
scope (Topometrix Inc., Santa Barbara, CA) along with an
external pulse/function generator (Model HP 8111 A), with
negative sample bias. Tips were electrochemically etched from
Pt–Ir wire (80% : 20%, diameter 0.2 mm) in 2 M KOH : 6 M
NaCN solution in water. Prior to imaging, the compound
under investigation was dissolved in 1-octanol at a concentra-
tion close to saturation, and a drop of the solution was applied
onto a freshly cleaved surface of highly oriented pyrolytic
graphite (HOPG, grade ZYB, Advanced Ceramics Inc., Cleve-
land, OH). Then, the STM tip was immersed in the solution
and images were recorded at the liquid/solid interface. The
STM images were acquired in the variable-current mode
Molecular modeling
Molecular modeling calculations were carried out using the
compass force field, as implemented in Materials Studio, a
product of Accelrys, San Diego, CA, USA. The energy mini-
mizations were carried out in the gas phase with a dielectric
constant of 1. All energy-terms were included with the excep-
tion of an explicit hydrogen-bonding term. For the non-
bonding interactions a cut-off radius of 12.5 A was used, with
a spine width of 3 A, and a buffer width of 1.0 A. A graphite
sheet, 20 Â 30 atoms in size, with fixed Cartesian position for
the carbon atoms was used as the substrate. All structures
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
New J. Chem., 2006, 30, 1420–1428 | 1427

View Article Online
were subjected to energy minimization using the Fletcher-
Reeves algorithm, to a final gradient with maximum derivative
of 0.001 kcal molÀ1. The folding ability was expressed by
comparison of the total potential energy for the energy-mini-
mized intramolecular H-bonded structure, with the optimized
extended conformation without intramolecular hydrogen
bond and with all CH2–CH2 bonds in trans configuration,
and both conformations in close contact with the graphite
substrate.
Abdel-Mottaleb, F. C. De Schryver, S. De Feyter, R. Lazzaroni
and S. Hoger, Chem. Mater., 2005, 17, 5670–5683; (h) B. A.
Hermann, L. J. Scherer, C. E. Housecroft and E. C. Constable,
Adv. Funct. Mater., 2006, 16, 221; (i) M. C. Lensen, S. J. T. van
Dingenen, J. A. A. W. Elemans, H. P. Dijkstra, G. P. M. van
Klink, G. van Koten, J. W. Gerritsen, S. Speller, R. J. M. Nolte
and A. E. Rowan, Chem. Commun., 2004, 762.
6 (a) D. J. Hill, M. J. Mio, R. B. Prince, T. S. Hughes and J. S.
Moore, Chem. Rev., 2001, 101, 3893; (b) D. J. Hill and J. S. Moore,
Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 5053; (c) A. Khan, C.
Kaiser and S. Hecht, Angew. Chem., Int. Ed., 2006, 45, 1878.
7 N. Schuurmans, H. Uji-i, W. Mamdouh, F. C. De Schryver, B. L.
Feringa, J. vanEsch and S. De Feyter, J. Am. Chem. Soc., 2004,
126, 13884.
Acknowledgements
8 S. De Feyter and F. C. De Schryver, J. Phys. Chem. B, 2005, 109,
4290.
The authors thank the Federal Science policy through IUAP-
V-03, the Institute for the promotion of innovation by Science
and Technology in Flanders (IWT), and the Fund for Scien-
tific Research-Flanders (FWO), as well as the Top Institute
Material Science Center (MSC+) of the University of Gro-
ningen. A. S. K. thanks the Fund for Scientific Research-
Flanders for financial support. We would like to thank Dr
Peter Broekmann, Prof. Wandelt and their coworkers (Uni-
versity of Bonn) for their support and stimulating discussions.
9 (a) W. Mamdouh, H. Uji-i, J. Ladislaw, A. Dulcey, V. Percec, F. C.
De Schryver and S. De Feyter, J. Am. Chem. Soc., 2006, 128, 317;
(b) B. Venkataraman, J. J. Breen and G. W. Flynn, J. Phys. Chem.,
1995, 99, 6608; (c) M. Lackinger, S. Griessl, W. M. Heckl, M.
Hietschold and G. W. Flynn, Langmuir, 2005, 21, 4984.
10 A. Gesquiere, M. M. Abdel-Mottaleb, S. De Feyter, F. C. De
Schryver, M. Sieffert, K. Mullen, A. Calderone, R. Lazzaroni and
¨
´
J.-L. Bredas, Chem.–Eur. J., 2000, 6, 3739.
11 (a) S. Yoshimoto, N. Higa and K. Itaya, J. Am. Chem. Soc., 2004,
126, 8540; (b) Z. Li, B. Han, L. J. Wan and T. Wandlowski,
Langmuir, 2005, 21, 6915; (c) C. Safarowsky, L. Merz, A. Rang, P.
Broekmann, B. A. Hermann and C. A. Schalley, Angew. Chem.,
Int. Ed., 2004, 43, 1291; (d) T. Ye, Y. F. He and E. Borguet, J.
Phys. Chem. B, 2006, 110, 6141; (e) Q. H. Yuan and L. J. Wan,
Chem.–Eur. J., 2006, 12, 2808.
References
1 M. O. Lorenzo, C. J. Baddeley, C. Muryn and R. Raval, Nature,
2000, 404, 376.
2 (a) Y. Okawa and M. Aono, Nature, 2001, 409, 683; (b) M. M.
Akai-Kasaya, K. Shimizu, Y. Watanabe, A. Saito, M. Aono and
Y. Kuwahara, Phys. Rev. Lett., 2003, 91, 255501; (c) A. Miura, S.
De Feyter, M. M. S. Abdel-Mottaleb, A. Gesquiere, P. C. M.
Grim, G. Moessner, M. Sieffert, M. Klapper, K. Mullen and F. C.
¨
De Schryver, Langmuir, 2003, 19, 6474; (d) S. Hoeppener, L. F. Chi
and H. Fuchs, Nano Lett., 2002, 2, 459.
3 F. Jackel, W. D. Watson, K. Mullen and J. P. Rabe, Phys. Rev.
¨
¨
Lett., 2004, 92, 188303.
4 (a) J. Rabe and S. Buchholz, Science, 1991, 253, 424; (b) D. M. Cyr,
B. Venkataraman and G. W. Flynn, Chem. Mater., 1996, 8, 1600;
(c) P. Samori, J. Mater. Chem., 2004, 14, 1353.
5 (a) T. A. Jung, R. R. Schlittler, J. K. Gimzewski, H. Tang and C.
Joachim, Science, 1996, 271, 181; (b) T. A. Jung, R. R. Schlittler
and J. K. Gimzewski, Nature, 1997, 386, 696; (c) H. B. Fang, L. C.
Giancarlo and G. W. Flynn, J. Phys. Chem. B, 1999, 103, 5712; (d)
F. Tao, J. Goswami and S. Bernasek, J. Phys. Chem. B, 2006, 110,
4199; (e) R. Azumi, E. Mena-Osteritz, R. Boese, J. Benet-Buchholz
and P. Bauerle, J. Mater. Chem., 2006, 16, 728–735; (f) S. Weigelt,
C. Busse, L. Petersen, E. Rauls, B. Hammer, K. V. Gothelf, F.
Besenbacher and T. R. Linderoth, Nat. Mater., 2006, 5, 112; (g) A.
Ziegler, W. Mamdouh, A. V. Heyen, M. Surin, H. Uji, M. M. S.
12 K. Ienaga, K. Higashiura, Y. Toyomaki, H. Matsuura and H.
Kimura, Chem. Pharm. Bull., 1988, 36, 70.
13 Moreover, most likely, the molecules are not folded with a side-on
orientation. If side on, intermolecular hydrogen bonding will not
be optimized. Furthermore, as the alkyl chains are not very
transmissive for tunnelling electrons, having a double layer of
alkyl chains could hamper high quality imaging of the adsorbed
layer as the tip might interact with the ‘upper alkyl chains’ under
the given experimental conditions.
14 J. Brinksma, B. L. Feringa, R. M. Kellogg, R. Vreeker and J. van
Esch, Langmuir, 2000, 16, 9249.
15 As the sample cell is not closed, the solvent evaporates too fast.
Faradaic currents due to residual water mask the tunnelling
current. Coating of the tip according to the procedure for EC-
STM is also not ideal since ethanol might (partly) dissolve the
coating.
16 S. Yoshimoto, R. Narita and K. Itaya, Chem. Lett., 2002, 356.
17 As no clear images of the Au(111) lattice could be recorded, and
therefore internal calibration is lacking, an accurate analysis
of the parameters describing the molecular ordering is not
provided.
18 M. Wilms, M. Kruft, G. Bermes and K. Wandelt, Rev. Sci.
Instrum., 1999, 70, 3641.
ꢀc
This journal is the Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2006
1428 | New J. Chem., 2006, 30, 1420–1428