University of Groningen Pattern Formation in organic

University of Groningen

Pattern Formation in organic monolayersSchuurmans, Norbert

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2006

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Schuurmans, N. (2006). Pattern Formation in organic monolayers. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-amendment.

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 15-06-2022

https://research.rug.nl/en/publications/b6950c9b-aff4-4e90-bf7a-937559943c0a

Stellingen:

1. Refereren naar niet gepubliceerd werk is weinig inzichtelijk.

2. Het gebruik van triviale nomenclatuur op de verpakking van genees- en voedingsmiddelen

heeft louter tot doel de consument te imponeren.

3. De discrepantie tussen het grote belang dat aan citatiescores wordt toegekend en de bekendheid

van daadwerkelijk veel geciteerde onderzoekers is frappant. Lowry, O.H., Rosebrough, N.J., Farr, A.L., Randall, R.J. J. Biol. Chem. 1951, 193, 265-275

4. Het wederzijds onbegrip tussen man en vrouw kan in een evolutionaire context geduid worden

door in acht te nemen welk een efficientie aan seks wordt toegedicht bij het beslechten van

echtelijke ruzies.

5. De discussie m.b.t. de depletie van de fossiele brandstoffen gaat te vaak voorbij aan de

gevolgen hiervan voor de organische bulk-chemie.

6. De bottom-up benadering wordt binnen sociale structuren nog te weinig gehanteerd.

7. Zo het vocabulaire van de natuur bestaat uit moleculen, zo kunnen haar grammatica en

syntaxis gevonden worden in de supramoleculaire chemie.

8. Als je niet kijkt, bestaat de maan niet. A.Pais, Rev. Mod. Phys. 1979, 51, 863-914, p 907.

Aspect, A., Grangier, P., Roger, G. Phys. Rev. Lett. 1982, 49, 91-94

9. Overdaad schaadt, zeker ook in het geval van illustraties bij artikelen.

Percec, V.; Peterca, M.; Sienkowska, M.J.; Ilies, M.A.; Aqad, E.; Smidrkal, J.; Heiney, P.A. J.Am.Chem.Soc. 2006, 128, 3324-3334.

10. Moleculair modelleren zonder experimentele structurele informatie is een hachelijke

bezigheid. Dit proefschrift.

11. Het toekennen van een eredoctoraat zou moeten geschieden op basis van wetenschappelijk

meetbare criteria. Het promotieregelement van deze universiteit

12. Irreduceerbare complexiteit in levende organismen verhoudt zich slecht met de redundantie in

de genetische code. Darwin's black box: the biochemical challenge to evolution, by M.J.Behe, The Free Press, New York 1996,

Trial version of ABC Amber PDF Merger, http://www.processtext.com/abcpdfmg.html


Pattern Formation in Organic Monolayers

About Molecular Origami in 2 Dimensions

Norbert Schuurmans



PhD thesis Groningen University

ISBN: 90367-2659-X

ISBN: 90367-2660-3 (elektronisch)

© Norbert Schuurmans, Groningen, 2006

The work described in this thesis was carried out at the

department of Organic and Molecular Inorganic Chemistry,

Stratingh Institute, University of Groningen, The Netherlands.

The work described in this thesis was financially supported by

MSCplus

.

MSCplus

PhD thesis series 2006-10

ISSN: 1570-1530



RIJKSUNIVERSITEIT GRONINGEN

Pattern Formation in Organic Monolayers

About Molecular Origami in 2 Dimensions

Proefschrift

ter verkrijging van het doctoraat in de

Wiskunde en Natuurwetenschappen

aan de Rijksuniversiteit Groningen

op gezag van de

Rector Magnificus, dr. F. Zwarts,

in het openbaar te verdedigen op

vrijdag 23 juni 2006

om 13.15 uur

door

Norbert Schuurmans

geboren op 12 oktober 1975

te Vlissingen



Promotores: Prof. Dr. B.L. Feringa

Prof. Dr. J.H. van Esch

Beoordelingscommissie: Prof. Dr. J.B.F.N. Engberts

Prof. Dr. J.C. Hummelen

Dr. S. De Feyter



Cover: The doors of perception

www.coolwallpaper.com

There is plenty of room at the surface.

(free interpretation of Richard Feynman’s adage)



Contents: Chapter 1: General introduction 9

1.1 Nanotechnology, monolayers and patterning 10

1.2 Outline 12

1.3 References 13

Chapter 2: Monolayers and (molecular) patterning; an overview 15

2.1 Classification of monolayers 15

2.2 Pattern formation in monolayers 16

2.3 Chemisorbed and physisorbed systems 17

2.4 An overview of top-down methodologies 20

2.5 Large scale patterning (10 nm - 1 µm) 22

2.6 Patterning in physisorbed monolayers 24

2.7 Conclusions and outlook 35

2.8 References 36

Chapter 3: Incorporation of functionality 45

3.1 Introduction 45

3.2 About the platform. A modular approach 46

3.3 About the complexant 47

3.4 Design and synthesis of the adsorbents 49

3.5 Formation of monolayers 53

3.6 Complexation experiments 55

3.7 Bipyridine derivatives without urea groups 56

3.8 Complexation experiments with the new bipyridine derivatives 59

3.9 The monoalkylated bipyridine 64

3.10 Can the template be used? 65

3.11 Conclusions 67

3.12 Experimental section 69

3.13 References 75



Chapter 4: Mixing and phase separation in binary systems 77

A means to control periodicity

4.1 Introduction 77

4.2 Intermixing and periodicity 78

4.3 About the components 80

4.4 Synthesis of the components 82

4.5 STM images of the individual components 82

4.5 The extremes in mixing behavior 85

4.6 Molecular modeling 88

4.7 Tolerance towards mismatch 89

4.8 Mixing with the shorter alkyl bisureas 91

4.9 Molecular length or H-bonding complementarity? 92

4.10 Conclusions 94


4.12 References 98

Chapter 5: Toward folded structures at the interface. Turnmimics 101

5.1 Introduction 101

5.2 Foldamers in 2D 106

5.3 The turn element 107

5.4 Catechol as the turn element 108

5.5 Towards functional turn mimics 111

5.6 Synthesis of the compounds 113

5.7 Monolayer formation of compounds 5.1 and 5.2 114

5.8 Further modeling studies 118

5.9 Towards a new turn element 120

5.10 Modeling and monolayer formation of compound 5.3 122

5.11 Conclusions. 124

5.12 Experimental Section 124

5.13 References 128



Chapter 6: Minimal foldamers in 2D 131

6.1 Introduction 131

6.2 Towards a new turn mimic 132

6.3 Monolayers of compound 6.1 134

6.4 Discussion 135

6.5 Towards 2-D foldamers 137

6.6 Foldamers of class 2 140

6.7 Foldamers of class 3 145

6.8 Conclusions 148


6.10 References 157

Summary 159

Samenvatting 160

Dankwoord 164



9

Chapter 1 General introduction

These days Nanotechnology has become a buzzword throughout the scientific community. The proposed miniaturization of engineering to the molecular scale is highly enticing and intellectually stimulating. Continuous cross-fertilization between knowledge from hitherto separate disciplines as physics, chemistry and biology promises to spawn a new paradigm, changing the way we think about matter and life itself. Ideas are plenty. In his seminal work, engines of creation, K. Eric Drexler described a universal assembler, a kind of synthetic ribosome, equipped with a genetic algorithm, allowing it to self-replicate. This ultimate machine would be able to construct any conceivable material structure, in any number of copies. Matter would be like software today. Whether this dream will ever be realized remains to be seen,

1 but if one thing, history taught us that revolutionary technologies have

always been under –, rather than overestimated regarding their scope and width.



General Introduction.

10

1.1 Nanotechnology, monolayers and patterning

Nanos (νανος) means dwarf in Greek. This indicates that one is dealing with very small entities. The term is usually taken to mean manipulation of matter on a scale of 0.1 nm (1 Å) to 1000 nm (1 µm). As this world is inhabited by entities like atoms, molecules, and molecular assemblies, this requires a methodology to control the relative location, and possibly the dynamics, of each entity (atom, molecule, assembly). In fact nanotechnology more accurately refers to a set of techniques to achieve this manipulation. A dichotomy can be discerned between top-down methodology and bottom-up methodology. The former refers to the continuous miniaturization of known technology with respect to construction and engineering, a process that brought down dimensions from millimeters to micrometers (1970’s) to nanometers. The latter refers to the process of actually building up nanostructures atom-by- atom, or molecule-by-molecule, preferably by a self-assembly process. As technology develops, the (length) scales that can be reached by either approach are converging.

Today nanoscience is fed from a number of subdisciplines. Nanoelectronics, nanomaterials, molecular nanotechnology, bionanotechnology and development of tools for analysis, especially scanning probe microscopies all contribute. The development of nanoelectronics and – materials mostly relies on top-down methodology. Molecular – and bionanotechnology have a more bottom-up approach. One very important goal in nanotechnology is bridging the gap between the subdisciplines, as prompted by the rush towards miniaturization in the electronics industries, which demands ever smaller components. Flexibility in the manipulation of the basic components of future devices will be imperative to a successful implementation of a next generation technology. As these future components are widely regarded to be of molecular scale, attaining control, at the molecular level, over their spatial distribution and organization will hold the key. A pure self-assembly approach towards spatial organization, to provide a complementary paradigm to the currently available top-down methodologies,2 including lithographic techniques, would in itself be a very desirable goal.

3 Other

parameters like cost and ease of fabrication only seem to be speaking more in favor of such an approach. Novel concepts and strategies for bottom-up construction of molecular assemblies might also find application in next generation materials and smart medicine. SPM (scanning probe microscopy, vide supra) will provide an invaluable tool for study and manipulation of these architectures.

Self-assembly:

Self-assembly, defined as the spontaneous, thermodynamically controlled, organization of individual molecules into a (meta) stable and (spatially) well-defined aggregate usually is a complicated and not very well understood phenomenon.



Chapter 1

11

Figure 1.1 A fine example of lithographic patterning.

It critically relies on a delicate interplay of many interactions, often of different nature. Nonetheless it is evident that living nature could not exist in its absence, as it is the driving force for the formation of the basic components in living organisms. It is exactly for this reason that self-assembly approaches have been put forward as an alternative to top-down methodology in nanotechnology.4 Dimensionality is poised to be a key parameter for the development of an understanding of self-assembly approaches at the nanoscale, as it is often more easy to control aggregation phenomena in two dimensions then in three dimensions. Reduction of dimensionality for the sake of simplification does not discard the relevance with respect to application though. Two-dimensional structures are typically encountered in nanoelectronics, where specific features are assembled onto a substrate.

5 Many new materials arise from a layer-by-layer build-up. A

nanostructured plane can also be used as a template to build structures in 3-D. Self-assembly promises to be a cheap and potentially extremely easy methodology to decorate surfaces with all kinds of interesting templates.

6 In general it would be

highly desirable to be able to control interfacial properties, as interfaces play such a prominent role in many processes, e.g. in catalysis, molecular recognition, and as sites of nucleation, both in natural and synthetic systems.

This thesis will be concerned with developing new self-assembly approaches towards nanostructured monolayers on surfaces. The monolayers will form at equilibrium at the solid – liquid interface. These monolayers will specifically alter the properties of the surface, due to periodically distributed chemical functionality. In this thesis we will describe an exploratory study to control the packing of these functionalities at a molecular level. Its main goal is to come to an integrated approach towards the design of adsorbents, such that they assemble into well- defined patterns, hence its title. Several aspects of these self-assembled surface patterns will be explored.



General Introduction.

12

1.2 Outline

After a literature survey on self-assembled monolayers and some existing methodologies to create nanoscale patterns in chapter 2, the focus will be on the introduction of reactive chemical groups into a monolayer, self-assembled by physisorption. This is the topic of chapter 3. It will be shown that variation of the chemical composition along the alkyl chains of the adsorbents allows for the introduction of functional entities.

7 When the molecules form a regular

superstructure at the interface, these functionalities are intrinsically organized into a pattern. The role of additional interactions, whose introduction can either add stability, or drive the formation of the aggregates into a specific shape, will be explored. Formation of a two-dimensional aggregate, a monolayer, can be inferred from imaging the system with the scanning tunneling microscope (STM).

In the next chapters, interactions in the lateral plane will be explored in more depth. This work aims to control the periodicity of the functionality in the plane. Novel concepts to influence the spacing, and thus the spatial distribution of functionality have been developed, adding to the versatility of the self-assembly approach. In chapter 4 it will be shown that multi-component monolayers can be formed, simply from a solution containing all the components. This emphasizes the generality of the approach.

In chapters 5 and 6 an entirely new idea will be introduced. The use of 2-D foldamers in the context of surface patterning will be explored. To this extent oligomeric molecules will be used, that absorb onto the surface and fold into a well-defined conformation and shape, and thus form regular 2-D patterns by self-assembly. The basic knowledge that results is directly relevant for protein absorption at interfaces, but is also combines aspects from folding and crystal design. The 2-D confinement, greatly reduces the degrees of freedom which will facilitate rational design by molecular modeling. In chapter 5 the design of a model system will be described, whereas in chapter 6 potential synthetic routes towards higher homologues, that can truly be regarded foldamers, will be worked out. In principle, the foldamer approach can be used for synthetic as well as natural oligomers, like peptides, and it can be applied to many different substrates by adjusting the foldamer - substrate interactions. The approach offers the advantage that the precise spatial positioning of functional groups can be programmed at the sequence level.



Chapter 1

13

1.3 References

1 The topic is rather controversial. See the Drexler – Smalley debate. Sci. Am. 2001, 285,

68-69 and concurrent rebuttals.

2 (a) Tseng, A.A., Notargiacomo, A. J. Nanosci. Nanotech. 2005, 5, 683-702 (b) Shimomura, M., Sawadaishi, T. Curr. Opinion in Coll. Interf. Sci. 2001, 6, 11-16

3 Science, 2002, 295, March 29, special issue devoted to self-assembly (b) Proc. Natl. Acad.Sci. 2002, 99, 8, special issue. 4 Whitesides, G.M., Grzybowski, B. Self-assembly at all scales, Science, 2002, 295, 2418-

2421

5 Zhirnov, V.V., Herr, D.J.C. New frontiers: self-assembly and nanoelectronics. Computer, 2001, 34, 34-43

6 Comprehensive Supramolecular Chemistry (Eds: Atwood, J.L., Davies, J.E.D., MacNicol,

D.D., Vogtle, F.) Pergamon, New York, 1996

7 Lewis, P.A., Donhauser, Z.J., Mantooth, B.A., Smith, R.K., Bumm, L.A.., Kelly, K.F., Weiss, P.S. Nanotechnology 2001, 12, 231, and references therein.



15

Chapter 2 Monolayers and (molecular) patterning; an overview

Monolayers (composed of molecules) have received a lot of attention in the last two decades, because of their remarkable versatility and ease of formation. Their potential in device fabrication has not gone unnoticed. A monolayer is defined as a layer of one molecule thickness, assembled at an interface. Self-assembled monolayers are already known since 1917, when Langmuir experimented with the formation of thin films on water.

1 Merging of knowledge from the chemical and

physical disciplines since has created an opportunity to construct monolayers on solid substrates. The development of surface analysis techniques has been crucial in this endeavour.

2

2.1 Classification of monolayers

Monolayers can be broadly classified in two groups, with respect to the nature of the interaction with the substrate. In the first group this interaction is chemical in nature, i.e. covalent bonds are formed. These are called chemisorbed monolayers. Alkylthiols on a gold substrate are a prominent member of this group (vide infra).

In the second group the interaction does not involve the formation of chemical bonds, and these are called physisorbed monolayers. Linear alkanes on a graphite (HOPG) substrate are a prominent member of this group (vide infra).

Usually Zisman is credited with the first report concerning monolayer formation on solid substrates,3 but the first report concerning a chemisorbed monolayer, on a



Monolayers and Molecular Patterning

16

gold interface, dates back to the early eighties.4 Later on, this kind of aggregate was popularized by Whitesides.

5,6 Typically, such a monolayer consists of one

layer of alkane thiols on the (111) gold surface. This layer can be regarded as a crystalline phase, templated on the solid substrate, extending in two dimensions. Only the thiol part of these molecules interacts with the gold. The part of the molecules exposed to the solution can take, in principle, any identity.

In the early nineties, however, it was demonstrated that covalent attachment is not a necessary prerequisite for monolayers to form. In 1991 Rabe and Bucholtz showed that alkanes assemble into a 2-dimensional crystalline structure (a monolayer) at the basal plane of highly oriented pyrolytic graphite.

7 The absence of

stable (covalent) interactions implies that something else must provide for the interaction to keep the adsorbing molecules in place at the interface.8 Weaker (intermolecular) forces, typically van der Waals interactions, can achieve this, but only if the contact area is large enough.

2.2 Pattern formation in monolayers

The relevance of monolayer structures (physisorbed as well as chemisorbed) has tremendously increased because of the possibilities to generate spatially defined substructures (patterns) into pre-aggregated monolayers. Nanoscale patterning is one of the most important areas of nanoscience, not in the least because of its relevance for the electronics- and biotechnnology industries.9 It therefore generates an enormous input from widely varying disciplines. As most of this work is largely outside the scope of this thesis only some recent developments will be highlighted here. The focus will be on the various length scales that have been explored. Presently, monolayers (either physisorbed or chemisorbed) with periodically recurring motifs ranging from sub-nanometer to micrometer can be generated. A continuous race towards increasingly smaller structures has resulted in very small patterns, but µm–sized structures are important in their own right, especially in the growing field of bio-nanotechnology. The distinction between chemisorbed and physisorbed monolayers, in the context of this chapter, is maintained to stress the different methodologies employed to affect patterning. The approaches are complementary to a large extent (see chapter 1). In physisorbed systems, patterning is intrinsically connected to the self-assembly process (and therefore relies on bottom-up approaches), whereas in chemisorbed systems, the material in which the pattern is generated is the result of a self-assembly process, but an external means is invoked to create the pattern itself (and therefore relies on top-down approaches).

Physisorbed systems are indispensable for generating very small (molecular scale) regular patterns in the formation of a 2-D crystal. The reversibility of the process dramatically contrasts the stable aggregates that can be formed by chemisorption. Chemisorbed monolayers however are, on account of their characteristic stability, probably more interesting from the point of view of device construction.



Chapter 2

17

2.3 Chemisorbed and physisorbed systems.

Chemisorbed monolayers are formed by immersion of a freshly cleaned substrate into a solution of the adsorbate (fig.2.1). If the solution contains a mixture of adsorbates, a mixed monolayer will be formed. The adsorbate molecules are then covalently attached to the surface. The thiol group can be considered to be an “anchor” because it has specific affinity for the substrate (gold). Thioethers

10 and

dithiol moieties can also be used as anchor groups.

Other substates, like glass11

and mica12,13

have also been explored as a template to form self-assembled monolayers. These substrates require their own respective adsorbate anchoring groups. For glass, for example, this is a silane group.

In these systems, the common feature is that the molecules comprising the monolayer have their long axes oriented perpendicular to the surface, although this long axis usually makes an angle with the surface (fig.2.1). This in turn implies that the intrinsic periodicity (the molecule – molecule distance) is determined by the periodicity of the gold lattice (fig.2.2). Composition and structure of these chemisorbed monolayers depends in the first place on the strong and directional interaction of the anchor with the substrate. Additionally, non-covalent interactions between the alkyl moieties are important to determine the angle that the molecules make with the substrate.

Figure 2.1 Formation of self-assembled monolayers on gold substrate. Image taken from www. ifm.liu.se/applphys/biomaterial/research/sam.html.

Figure 2.2 Typical periodicity in a chemisorbed monolayer. From www.cms.llnl.gov/s-t/surface.html




18

Chemical modification of alkane thiols is possible, which can introduce complementary interactions. Usually, functional groups, other than methyl, are introduced at the terminal position in bifunctional alkylthiols, such that these are in contact with the solution, upon formation of the monolayer. The variable chemical nature of the terminal groups has been shown to influence various characteristics of the interface (eg. wettability). Incorporation of hydroxy,

14 carboxylic acid

15 or

amine16

functionalities as the terminal group, has been reported. These derivatives all form self-assembled monolayers, with the same basic characteristics as the parent alkylthiols.

2.3.1 Physisorbed monolayers.

In physisorbed monolayers, van der Waals interactions provide for the affinity with the substrate. For alkanes on graphite (HOPG), the composition and structure of the monolayers depends on the formation of a commensurate lattice17 on top of the basal plane of the graphite, because the C-C bond length in the alkanes, in their extended zig-zag conformation is similar to the intrinsic atom-to-atom distances in the graphite lattice.

As a consequence, molecules in physisorbed monolayers usually have a coplanar orientation with respect to the substrate (fig. 2.3; alkyl bisureas are the topic of chapter 4, hence this example). A number of papers reviewing the topic of physisorbed monolayers has appeared.18

2.3.2 Methods of analysis

Chemisorbed monolayers are usually analyzed with a number of techniques, including contact angle goniometry,

19 IR techniques

20 (ATR-IR, IRRAS) surface

plasmon resonance (SPR),21 X-ray photoelectron spectroscopy (XPS),22 and scanning probe microscopies (STM, AFM).23 The relatively large tunneling barrier due to the orientation of the molecules makes it difficult to study these monolayers with STM though. Some results have been obtained with short molecules.

24

0.5 nm S

S

S

S

0.5 nm

0.5 nmU

U

U

U

U = urea

variable, up to 20 nm

Figure 2.3 Comparing periodicities in typical physisorbed (left) and chemisorbed (right) monolayers.



Chapter 2

19

Physisorbed monolayers are most easily characterized with the STM (scanning tunneling microscope).

2.3.2.1 The Scanning Tunneling Microscope25

The scanning tunneling microscope,26 since its development in 1982, had a tremendous impact on the study of interfacial phenomena. The technique depends on the flow of a current between a tip and a conductive sample.

27 The tunneling

current is inversely proportional to the distance between the substrate and the tip.27

Therefore STM is able to give a topographic profile of the surface under study. But because the technique is electronic in nature, electronic information within the surface (the surface density of states) is incorporated in the final images obtained.

28 According to Tersoff and Hamann STM basically gives a contour map

of the local density of states at the Fermi level of the surface at the position of the tip.28

Observation and study of adlayers on a substrate is possible if the adsorbate sufficiently modifies the (tunneling) current. This is caused by either a change in the tunneling barrier29 or by a change in the density of states within the substrate, due to electronic interactions with the adsorbate.30 Typically the contrasts are different from plain graphite, the exact shading being determined by electronic properties of the individual atoms.

This allows probing the chemical composition and discerning various functional groups at the interface. Size, geometry and electronic structure of the functionalities are all important.

18d

Interestingly the observed contrasts are also dependent on applied bias voltage31,32 and even scan direction. The degree of electronic coupling between frontier orbitals of the adsorbate and the substrate seems to determine the relevancy of geometry (i.e. the degree to which the shape of the adsorbate is reflected in the STM image).

The tips, which are cut from Pt/Ir wire, are ideally only one atom thick at the point of interaction. Positioning of the tip is accomplished by means of an electronic feedback mechanism, directed by a piezo element.

Figure 2.4 STM set-up.




20

2.4 An overview of top-down methodologies

A number of techniques have been successfully developed for patterning organic substances onto solid substrates such as metals, glass, and silicon. A first class of techniques is photochemical in nature. Examples include photolithography,

33 laser

ablation.34

ion and e-beam writing,35

photoimmobilization,36

and surface initiated polymerization.37

A second class utilizes mechanical means. Inkjet printing,38 microcontact printing39 and micromachining

40 can be mentioned here. It is possible to write patterns by the

application of mechanical stress with an AFM tip.41

Recently scanning probe (dip-pen) lithography has joined (fig. 2.5).42 This technique is different from conventional pattern generation with an AFM tip, in that the features are actually written.

These techniques have in common that the patterns are generated by an external means. In photolithography this is a mask, which determines the spots on the sample that will be irradiated. The printing techniques make use of a pattened master. The size of the patterns that can be created thus critically depends on the accuracy in the fabrication process of these masks/masters. Although this has greatly improved in recent years (vide infra), molecular dimensions are not in reach yet. Also the variation in the geometries of the patterns is limited. The most important methodologies, and their recent improvements will be briefly discussed now, to provide a context for the self-assembly approaches towards patterning, discussed in paragraph 2.6 onwards.

2.4.1 Photolithography (10 µm – 100 nm)

Photolithography, for a long time, has been the favorite technique for pattern-generation, and is applied commercially on a large scale in the electronics industry in the fabrication of integrated circuits.

43

Figure 2.5 Cartoon representatation of the set-up for dip pen lithography.

42



Chapter 2

21

The technique relies on irradiation of a sample through a mask, exposing specific areas of the sample to be patterned (fig. 2.6). Photo-labile groups can be cleaved off, exposing different chemical functionalities on places that were irradiated or not irradiated. Typically, micrometer-sized patterns are created,

44 but smaller features

are possible. 250 nm has been the limit until the new millennium,45 but concurrent developments have reduced the scales since. Features as small as 100 nm can now be routinely produced46 by means of deep UV light, using surface plasmons

47 or with a UV curable

mold.48

Another promising approach is scanning near field optical lithography (SNOL). Features below the diffraction limit of light can be generated.

49

Photo-patterning techniques have been applied to a wide range of substrates and functionalities. Thiolate monolayers on gold50 as well as organosilane monolayers51 have been used as photoresists. Several functional entities, in a variety of contexts, have been organized following this approach, including zeolites

52 and

metal oxide (semiconductor) particles.53

An interesting paper by the Whitesides group54 describes how photopatterning can be combined with other techniques. The authors used photolithography to burn a pattern into a SU 8 photoresist supported on a silica substrate. The resulting structure then looks like a set of molecular dominoes. With the aid of a PDMS slab, shear stress can be applied, causing the dominoes to topple over. How they fall, can be controlled by the shape of the dominoes and the direction of the stress applied. Patterns of specific size and morphology are produced by the collapsed ensemble, and these are immediately transferred to the PDMS slab.

It should be stressed at this point that self-assembled monolayers are not the only substrates used to generate patterns.

Other soft materials (other than monolayers), which can be applied as thin layers have frequently been used in photolithography. Many of these are polymeric in nature, for example polychloromethylstyrene films.55 Surface photografting / polymerization is basically a lithographic technique. Photografting methods are based upon incorporation of photoactive monomers into the polymeric substrate. The usual procedure involves forming a layer of a second polymer on top of the

Figure 2.6 Typical lithography approach. Image taken from www.dbanks.demon.co.uk/ueng/plith.html.




22

photoactive substrate and generating the desired pattern by exposure to UV radiation through a mask.

56

2.4.1.1. Other lithography techniques

Not only photons are used in lithography. Low-energy electrons have been used as well.57,58 E-beam patterned monolayers have been used to locally deposit nanotubes.

59 Also soft (1.5 keV) and hard (10keV) X-ray sources

60 can be applied.

These techniques are mainly used to pattern polymeric substrates though. The dimensions are comparable to conventional photolithography.

2.4.2 Soft lithography (1 µm – 20 nm)

Soft lithography has been proposed as an alternative to photolithography

to surmount the 100 nm barrier.

61 This

term is understood to comprise a set of microfabrication methods, that all make use of patterned elastomers. Micro-contact printing (µ-CP),

62,63 is the most prominent

member, but related techniques like replica molding,64

or capillary micromolding65 also fall in the category. Also phase-shift lithopgraphy61 and embossing61 can be categorized as such. A schematic representation of the process is shown in fig. 2.7. Although the µ in the name seems to indicate that typically micrometer-sized patterns can be generated, much smaller features are within reach today, due to improved techniques to construct the masters themselves. Features as small as 20 nm can now be patterned, almost overlapping with self-assembly approaches in physisorbed systems. Masters used in soft lithography can now be made with a commercial microscope (projection lithography).

66 The technique has

been applied to various substrates, to organize for example molecular printboards67 and lipid vesicles.68 Even spherical surfaces have been patterned.

69

2.5 Large-scale patterning (10 µm - 1 µm)

To stress the complementarity between (molecular) self-assembly and the top-down approaches just discussed, this paragraph is added. The large-scale periodicities are difficult to access via molecular self-assembly. Furthermore, this is a rapidly expanding field, and it is especially important in bio-nanotechnology.70 Various biochip platforms have found commercial application in high-throughput assays.

71

At the top end of scales, efforts have focused mainly on cells and proteins. Surfaces can be locally altered to resist adsorption by proteins. Standard photolithography on inorganic substrates directs the formation of a first adlayer,

Figure 2.7 Outline of the contact printing process.

62



Chapter 2

23

which in turn drives adsorption of proteins.72 Patterned substrates, locally exposing carboxylic acid groups selectively bind antigens, which in turn bind antibodies. Scanning probe lithography can also be applied to the fabrication of protein patterns.73 Recent developments have focused on even larger structures. Two examples will be discussed.

Neurons are attractive candidates for patterning because of their responsiveness to electrical stimuli. Thus a report on (patterned) hybrid neuron-semiconductor chips has emerged.74 Such chips will eventually consist of neural networks directly interfaced to electronic circuits. Another interesting application involves the lateral organization of the neurons. Experimental control over connectivity within functional neuronal networks is a promising approach in research on signal transduction and processing by the nervous system. Rat embryonic cortical neurons have been grown on patterns of extracellular matrix proteins applied to polystyrene substrates by microcontact printing.

Cells comply well with the pattern and form synaptic connections along the experimentally defined pathways.75

Another interesting and potentially important application of biomolecular nanotechnology involves organized placement of bacteria on a surface. Recently predesigned microarrays of living bacteria on a surface have been developed (fig. 2.8).76 Bacteria have also been patterned using hydrogel stamps.77 Applications in biodiagnostics have been reviewed.78

Figure 2.8 Arrays of bacteria, grown on lithographically patterned substrate.




24

2.6 Patterning in physisorbed monolayers

This section deals with physisorbed systems. Because the patterns are intrinsically formed in the assembly process, the dimensions are typically smaller then in chemisorbed monolayers. This means that molecular dimensions are achieved in the patterned monolayers. Usually the periodicities are smaller then 10 nm. Another aspect of physisorbed monolayers is their small extension towards the solution. This makes this kind of monolayer very amenable to investigation by means of the scanning tunneling microscope (STM). A distinction is made (this and the next paragraph) between adsorbents of different shape. They are classified as tile-like or rod-like. Shape of the adsorbents drastically influences the morphology of the lattice that will be formed. If a space-filling organization can be attained, this usually results in a periodic distribution.

2.6.1. 2-D crystals formed from adsorbents with high aspect ratio (rods)

Monolayers formed by this type of adsorbent are treated separately, because they generally lead to different structures. Lamellar superstructures are prominent. Many of these molecules incorporate alkyl fragments, following the original work of Rabe.

7 The protoype is a linear

alkane without functional groups; the oldest report of which dates back to 1991.

7 The periodicity in one

dimension is defined by the length of the alkanes (e.g C16H32; l = 2 nm), whereas in the other direction it is of atomic dimensions; 2 – 3 Å. A representation of the typical lamellar organization is given in fig.2.9. By introduction of functional groups the orientation of the molecules in the lattice can be altered. Herringbone packings are driven by H-bond formation. The individual lamellae make an angle of about 120 ° with each other in monolayers of aliphatic alcohols (fig. 2.10).79 Carboxylic acid groups seem to be favorable to effect specific packing geometries, as was shown by a number of groups.

x

y

z

2.46 Å 2.51 Å

x

y

z

2.46 Å 2.51 Å

Figure 2.9 Schematic representation of the ordering of n-alkanes on graphite. The molecules lie with their molecular axes parallel to the graphite lattice directions; the methylene hydrogens occupy the hollow sites of the surface. The molecules organize into rows. Note the similarity of the distances between neighboring atoms in the graphite lattice and the alkanes. Therefore a commensurate lattice can be formed. The image is adapted from ref. 18d.



Chapter 2

25

Surfactants like STS (sodium tridecyl sulfate), which incorporate a sulfonate head-group show comparable packing motifs,

80 as do triazines with long aliphatic tails.

81

Bifunctional compounds in turn show different packing geometries then monofunctionalized ones. In the case of hydroxyhexadecanoic acid the packing is dominated by interactions between the carboxylic acid functionalities.82

Because alkanes tend to form a commensurate lattice on top of graphite (fig.2.9), introduction of long alkane fragments is one of the best ways to form laterally organized monolayers. The approach has widely been employed to organize functionality in a plane. The functionalities are typically arranged in rows (lamellar structures), which can thus be regarded as a one-dimensional mode of organization. Sometimes additional interactions are introduced to stabilize the aggregates.

83 Examples of functionality organized in this way include

quinacridones,84 coumarins85 and (oligo)thiophenes.87 Oligothiophenes also organize on graphite without alkyl substituents, but the resolution is much better in the former case.

86 Organization of various dyes on surfaces has been recently

reviewed.87

2.6.1.1 Alkane monolayers on Au (111)

Alkanes not only form crystalline structures on graphite though. They also organize on Au (111).

88 Attention is generally more directed towards development of an

understanding of the organization process in studies on gold substrates. It has been proposed that ordered overlayers can only be formed on reconstructed gold.89 Initially it was postulated that only alkanes of specific length (number of methylene groups) could form commensurate lattices,

90 which would be related to

a non-linear variation in the molecule – substrate interaction for alkanes with increasing length. A similar argument was invoked to explain the formation of tilted and rectangular lattices respectively, for various alkanes.91 Recently, however, it has been shown that all linear alkanes, with an even number of C atoms, form

OH OH OH OH

OH

HOHOHOHO

HO

60°

60°

120°

90°

OH OH OH OH

OH

HOHOHOHO

HO

60°

60°

120°

90°

Figure 2.10 Influence of functional groups on the morphology of monolayers STM images and schematic diagrams for the organization of alkanes (left ) and alkanols (right). To optimize the geometry for H-bond formation alkanols make a 60° angle with the lamellar axis whereas alkanes make a 90° angle. This leads to the observed herringbone structure for alkanols.




26

organized 2-D lattices on Au (111).92 A variation in tilted and rectangular lattices was still observed.

2.6.2 Applications

Many of the systems described here have been developed not solely for the purpose of pattern generation (a 2-D crystal). Specific aspects of these systems have been looked at. These include chirality, crystallization dynamics, organization of functionality, phase behavior in multicomponent systems and reactivity. Some relevant examples will be discussed.

2.6.2.1 Chirality

Chirality, the property of non-superimposibility of an object and its mirror image, is an extremely important concept in organic chemistry. It also plays a major role in aggregate formation at interfaces. Because of reduced dimensionality it has an even wider scope than in conventional chemistry, which operates in a 3-dimensional world. A molecule can be intrinsically chiral in a 2-dimensional environment without typical asymmetric C atoms. Referring to the definition, it can be easily shown that many objects cannot be superimposed on their mirror image, if confined to a plane. This means that molecules that are not chiral in solution, can become 2D-chiral when adsorbed, if one of the enantiotopic faces has preferential interaction with the substrate. As a consequence, mixtures of (2-D) enantiomers will tend to separate in enantioresolved domains. In the case of alkyl fragments, which adopt a zig-zag conformation on graphite, the addition of one methylene group reverts the orientation of a functionality in the plane. This can make the difference between a chiral and a non-chiral compound, and thus greatly influence the respective packings. These odd-even effects are therefore very important.93 In the arrangements of bifunctional compounds hydroxypentadecanoic acid and hydroxyhexadecanoic acid such an odd-even effect was observed.

94 A number of

studies towards alkanoic acids have focused on odd-even effects, and 2-D chiral phase separation.95 The driving forces for the formation of a specific structure can be rather subtle; an odd-even effect was responsible for the formation of either homofacially or alternating heterofacially adsorbed monolayers (the 2-dimensional equivalents of conglomerates and racemates, respectively) in the crystallization of thioether substituted anthracenes (fig. 2.11).96

But also introduction of an asymmetric C atom is known to promote chiral phase separation in physisorbed monolayers. With formamide derivatives molecular tapes have been created giving enantiomorphous monolayers for the (S) – and (R) forms, respectively, while the racemate gave a distinct structure.97 Some strange effects have been observed. The phase separation of a racemic mixture of (S)- and (R)- bromohexadecanoic acid was brought about by achiral hexadecanoic acid.

98 In a previous study on these enantiomers the use of STM in determination

of absolute configuration of asymmetric centers was established.99



Chapter 2

27

2.6.2.2 Other studies

Studies focusing on the details of the 2-D crystallization process (as a model for 3-D crystallization) are an important application of the scanning tunneling microscope. Some interesting observations have been made. Contrary to 3-D crystals, physisorbed monolayers usually show only one orientation within a domain. In a study centering on an aromatic diketone however it was shown that multiple inequivalent packings of the same molecule within a unit cell are indeed possible.100 This study has been extended to a series of 1,3-disubstituted benzene derivatives.101 The packing motifs were shown to critically depend on the functional groups (ester, thioester, ketone) attached to the benzene moiety. These studies are very important to understand the relation between molecular structure and morphology of the aggregates. Compared to crystallization in 3-D, complexity of the problem is reduced because only 17 symmetry groups describe all possible symmetry elements, as opposed to the 230 space groups necessary to describe packing in 3-D.

102

Multi-component systems have been studied too. Phase separation of two non-mixing components was observed when fluorinated and non-fluorinated alkanes were co-deposited.

103 Because fluorinated derivatives have a specific contrast,

SC11H23

SC11H23

SC12H25

SC12H25

Figure 2.11 Formation of different packings due to odd-even effect. Note the orientation of the anthracene moieties.




28

they are easily identified in the STM images. This property is not unique to fluorocarbons, it holds true for brominated alkanes too. It has been utilized in studies where some brominated molecules were introduced in a non-brominated matrix. The brominated derivatives act as chemical markers, which facilitates study of the dynamics in the monolayers.104

Manipulation of physisorbed monolayers is also possible. This is quite remarkable given the dynamic nature of these systems, as a consequence of which any post-modification might seem to perturb the monolayer. Especially photochemistry is a good tool to affect reactions in the plane. With an isophthalic acid derivative, bearing a diacetylene side chain, photochemical polymerization resulted in new contrasts being observed, indicative of a monolayer polymerized along the lamellar direction.105 Another example is given by the photodimerization of cinnamates, bearing two long alkyl chains. The process could be observed by comparing the molecule – molecule distances, which are shorter for the dimers (fig. 2.12).

106 On

the other hand direct manipulation using the STM tip can even effect chemical reaction without any other external stimulus.107 One example is provided by the group of Fujihara, who were able to remove a protecting group from an α,ω-aminooctanethiol embedded in an octanethiol monolayer.108 Pure mechanical manipulation is also possible.109

C18H37O

CO2C18H37

C18H37O

CO2C18H37C18H37O

CO2C18H37

Figure 2.12 Change of packing as a result of a chemical reaction.



Chapter 2

29

2.6.2.3 Non-lamellar packings

Not every molecule with a high aspect ratio and a tendency to adsorb on graphite, ends up in a lamellar structure. Specific interactions can be introduced to drive the formation of other packings. This is nicely illustrated by the different motifs that are formed by OPV derivatives, equipped with either a quadruply H-bonding ureidotriazine moiety

110 (lamellar packing, due to

dimerization) or a diaminotriazine moiety,111 which gives a beautiful flower shaped arrangement on the surface (fig. 2.13). Interaction with the graphite is largely provided by the OPV unit, which can as a result be regarded as an alternative to alkane fragments. The viability of this approach has been further explored in a study where a perylene group was organized by two OPV moieties.112 This monolayer was used to demonstrate bias-dependent contrast.

2.6.3. 2-D crystals formed from adsorbents with low aspect ratio (tiles)

2.6.3.1 Periodicities of one nm or smaller

Sub-nanometer sized features are found in the domain of small organic molecules. Small molecules have been shown to form ordered monolayers on transition metal surfaces. Especially aromatic ones like benzene, naphthalene, anthracene,113 pyrene and perylene

114 have well been imaged on Cu(111). Also heteroaromatic

compounds pyridine, pyrazine and triazine have been imaged on copper.115

Naphthalene – and perylene dicarboxylic anhydides give good surface coverages on inorganic substrates.116 Carbocyanine dyes form herringbone structures on Ag (111).

117 The organization makes photophysical studies very interesting. There is a

long- standing interest in the organization of the component bases of DNA. A, C, T and G, all organize regularly onto Au (111).118

Substituents can alter the geometries of the assemblies formed. Due to the small size, additional interactions between the molecules, to stabilize the aggregate as a whole, are frequently employed. Trimesic acid was shown to assemble on Cu(111) forming a honeycomb structure.119 The same authors published that these structures can be modulated by adding Fe, forming a mixed organic-inorganic hybrid structure at the solid-liquid interface.

120 Phthalic acid derivatives show a zig-

zag arrangement due to the H-bonding pattern.121

In another study phthalic acid

N

N

N

NH2

NH2

C12H25O

C12H25O

C12H25O

O

O

O

O

Figure 2.13 Typical flower motif formed by OPV derivative due to triazine head group.




30

derivatives, with one122 or two chiral side chains, were adsorbed onto HOPG. Assembly was shown to be dominated by the aromatic part of the molecule, while the chiral side chain could adopt several conformations, leading to polymorphism. Enantiospecific domain formation was maintained though.123 Self-assembly of tricarboxylic acids has been observed on Au(111). Two types of structures can be formed (polymorphism) a hexagonal packing (D3 symmetry) or a close-packed structure where the molecules interdigitate.

124

2.6.3.2. Periodicities of 1 – 2 nm

Here we find regular structures formed by medium sized organic molecules. This entry is mentioned to stress the importance of a specific class of molecules: Porphyrins and phthalocyanins. These are probably the best-studied molecules with respect to formation of 2-D crystals. The group of Bai and in China has done a lot of work on phtalocyanine (abbreviated Pc) derivatives and porphyrins, on various substrates, but most on HOPG.

125 Assembly on graphite is facilitated by

appending alkyl chains on the periphery of the phthalocyanines. Recent work has allowed interpretation of the intramolecular contrast variations in naphthocyanines (as the free base) by comparison to the frontier orbitals, as calculated for individual molecules. The internal structure appears to be sensitive to the bias voltage.

126

Other groups as well worked on the system. It was shown in an early stage, these molecules form ordered structures on various substrates.127 The flat geometry is favorable with respect to the 2-D ordering and formation of supramolecular architectures. Substituting the periphery with various functional groups can have profound effects on the supramolecular structures that are formed.

This was aptly

shown in a study by Yokoyama,128 where the periphery was substituted with cyanophenyl groups. The number and relative position of these groups directed the

N

HNN

NH N

HNN

NH

N

HNN

NH N

HNN

NH

CN

CN

CN

CNNC

Figure 2.14 Tuning assembly at the surface by introducing interaction points.



Chapter 2

31

formation of trimers, tetramers and molecular wires respectively (fig. 2.14).

Alternatively, various metals can be caught inside the core, dramatically changing the electronic properties of these molecules.

129 This leads for example to the

observation of a depression or a protrusion respectively in the STM images of nickel-Pc and cobalt-Pc. Copper,130 cobalt,131 palladium132 (for investigation of the contribution of d-orbitals to the STM images), iron,

130 nickel,

133 and tin

134

(nonplanar) phtalocyanines, as well as the free base126

have been imaged on various substrates up till now.

Combining the approaches, the group of Hipps codeposited fluorofunctionalized cobalt phthalocyanine (CoPc) and parent nickel phthalocyanine to find the (partial) formation of a new species: a 1:1 complex, presumably held together by F----H interactions.131 A mixture of copper tetraphenyl porhyrins and cobalt phthalocyanines also formed regular 2-D binary structure. The individual components can again be discerned by their contrasts, which is different for the participating species due to differences in the mode of occupation of d-orbitals.

133

Even in the mixed monolayer of cobalt phthalocyanine and cobalt tetraphenylporphyrine (which have the same metal complexed) the two species could be discerned because the phenyl rings in the latter are twisted a bit, which causes the metals in the two species to have different distances to the substrate, giving different contrasts in the STM.135

Co-assembly of Pc derivatives can also be achieved together with haloalkanes on HOPG, which is remarkable because the geometry of these two species is so different. The groove between the lamellae of the haloalkanes can be occupied by the phthalocyanines;136 so one might say that the alkanes act as a template. In another experiment copper phthalocyanines were deposited on top of a previously formed alkane monolayer on graphite.

137 This seemed to actually enhance the

stability. Even porphyrins with an axial ligand attached (pyridine; rhodium as the metal) have been imaged.138 The apparent height was compared to the species without the ligand. The optoelectronic properties of this kind of compounds have been exploited in an STM set-up. It proved possible to induce excitation with the STM tip, which led to fluorescence.

139 This is only possible when using a small

isolating organic multilayer on top of the substrate (Cu(111)) to decouple the excited states of the molecules from the metal substrate in the tunneling regime.

2.6.3.3. Periodicities in the order of 2 – 10 nm.

This is the size domain of large organic molecules. A few important categories that have been looked at will be discussed here. 3-fold-symmetry and 6-fold symmetry are common for the members of this group.

The first category features macromolecular, but constitutionally well-defined structures. Especially macrocycles are amenable to form organized superstructures at this length scale. Bauerle et al. have synthesized a macrocycle comprising three or four terthiophene moieties,140 joined by diacetylene bridges. These molecules retain their structural integrity, due to high rigidity, and adopt an




32

annular conformation, upon adsorption onto HOPG, with lattice constant (translational repeat) of 2.6 nm (fig. 2.15).141

Shape–persistent macrocycles, composed of aromatic and acetylene moieties form similar structures on HOPG.142 Although the parent macrocycle in this study has a 6-fold symmetry, the crystal structure was found to be oblique with cell parameters of 3.6 x 5.6 nm. In another example macrocycles were adsorbed onto a negatively charged layer of cloride anions, which in turn had been assembled onto a positively charge copper substrate.143 Macrocycles can provide a basis for further self-assembly. A three-component nanobasket

144 (whose base is a macrocycle) has been constructed. This base was

imaged with STM; the entire structure (which is, of course, not flat) was not imaged.

A second category, forming similar superstructures is comprised of polycyclic aromatic hydrocarbons (PAH). Benzocoronene and higher homologues form 2-dimensional crystals at the liquid-solid interface onto HOPG. These molecules are flat disks145 and are thus very suitable for assembly on graphite. They can be regarded as two-dimensional subsections of graphite.

146 Solubility is obviously a

major issue, but this problem can be circumvented by attaching aliphatic chains to the disks. Members of this class of increasing size have been imaged with STM.147, 148 Additional information regarding their organization was obtained from angle-resolved ultraviolet photoelectron spectroscopy (UPS).

149 These molecules

have additional advantages: a large number of side chains can be introduced, influencing the properties. In one study varying contrasts have been observed in a monolayer of one derivative. This was ascribed to the formation of a molecular “staircase” (fig. 2.16).

146 The authors claim that the disks can float on three

different equilibrium heights above the surface, rather than being stacked. This counterintuitive explanation was backed up by a detailed analysis of the contrasts, and is thought to arise from steric hindrance. Apart from the suitability to assemble on graphite, benzocoronenes (HBC’s) also have interesting electronic properties due to their extended π-conjugated system. Therefore it is not surprising that more advanced systems, where the coronene donor moiety was covalently connected to one or more acceptor (in the form of anthraquinone) moieties have been

Figure 2.15 Model of Bauerle’s self-assembling thiophene macrocycles.



Chapter 2

33

synthesized with the aim of studying their supramolecular assembly behavior at interfaces.150,151 The unit cell in these crystalline monolayers is in the order of 2 nm.

Discotic liquid crystals, based on triphenylene can be considered as small PAH’s. It is not too surprising that these molecules form organized structures.152 If suitably substituted, with aliphatic chains, supramolecular rows are formed, characterized by pairwise interactions. The spacing is about 3 nm. Depending on side chains and substrate, other geometries have been observed.

153

Alternatively, dendrimers have been used to generate periodicities at this length scale. Higher generation Frechet-type dendrons organize into supramolecular ribbons.

154,155 Polyphenylene dendrimers show polymorphism, giving rise to a

variety of superstuctures, which can be characterized as granular, diffuse or nanorod. Usually sub-molecular resolution is lost due to conformational disorder. The lower-generations form 2-D crystals with lattice constants of 2.5 x 2.3 nm.

156,157

In the above examples mono-component self-assembly has been discussed. The organization is driven by weak intermolecular forces (typically van der Waals interactions). Special attention should be given to multi-component self-assembly, which can also produce organized structures at this length scale. It has been found that coronene, when imaged in heptanoic acid forms a 2-dimensionally organized structure, with six solvent molecules surrounding a central coronene. The spacing is 1.4 nm between adjacent coronenes.158 Alternatively, assembly of the coronenes is guided by orthogonal self-assembly of trimesic acid. Trimesic acid assembles into a hexagonal superstucture, which doubles as a template, as it leaves precisely enough space in the core to accommodate coronene molecules; the binary self-assembly is largely driven by shape complementarity.159 An interesting structure

C12H25

C12H25C12H25

C12H25

C12H25

C12H25

Figure 2.16 Benzcoronenederivative and resultant structure on graphite. Note the various contrasts.




34

was observed when perylene tetracarboxylic diimide (PTCDI) and melamine were codeposited on a silver surface (fig. 2.17). Supramolecular hydrogen bonding directs the formation of an open honeycomb structure.

160 Within the pockets of this

supramolecular lattice fullerenes can be trapped. Also other intermolecular interaction types have been utilized. Lehn et al. used coordination interactions in the formation of molecular grids.

161 These are composed of pyridines and

pyrazines. The molecules are glued together at the interface by means of transition metal ions like cobalt and zinc.

A final example that deserves mention comes from the group of Nolte. By attaching six

162 or twelve

163 porphyrins to an aromatic core, disc-shaped supermolecules can

be synthesized, which assemble into “cables”. These cables can be visualized by means of STM; they assemble edge-on with respect to graphite, which is quite remarkable. A parallel mode of assembly though (of the individual discs) can be induced by addition of extra ligands for the zinc porphyrins. The periodicity is close to 5 nm in that case.

2.6.3.4. Periodicities over 10 nm

In this range one can find polymeric structures or large oligomers. These large entities usually do not form well-organized structures. Still they are able to cover substrates in a variety of ways.

164 Quite some work has been done on colloids. The

work is aimed for example at influencing interfacial properties,165 or study of the aggregation behavior of the colloids.166 Interesting work has been done with DNA-tiles, which at least in one dimension exceed 10 nm.

167,168 These tiles are

composed of anti-parallel double-crossover (DX) DNA, which is of high rigidity.169

Four Strands combine, via complementary base pairing, to form the tile (repeat unit, fig. 2.18a). Each tile features so called sticky ends, which allows for (programmable) recognition (fig. 2.18b). The tiles self-assemble onto mica, forming a woven fabric of DNA strands, the lattice being held together by non-covalent interactions. Incorporation of a DNA hairpin into the tiles, introduces a marker,

Figure 2.17 Multi-component self-assembly by PTCDI and melamine, and entrapment of fullerenes.



Chapter 2

35

which allows the creation of striped patterns that can be imaged with AFM (fig 2.18c). Self-assembly with DNA tiles may signpost a route towards bridging the gap between self-assembly and the top-down methodology described before. Very interesting applications of these structures have already been explored.170, 171

It should be noted, finally, that self-assembly is not a particular molecular property or even a property of entities the size of molecules. The same principles apply to mesoscopic objects. This notion has been cleverly exploited, also in 2-D, with small hexagonal metallic plates that interact by shape complementarity and capillary interactions.

172

2.7 Conclusions and outlook

A range of techniques can be applied to enforce periodic distributions (patterns) of chemical functionality in a plane. A distinction can be made with respect to the methodology according to which this is achieved. Top-down methodology is frequently being applied to metal or semiconductor surfaces, possibly modified by chemisorbed monolayers, or alternatively polymer brushes.

Figure 2.18 Seeman’s assembling DNA tiles. (a) composition, (b) surface topology, (c) AFM image of self-assembled tiles; scale-bar measures 300 nm.




36

On the other hand patterns are intrinsically formed in physisorbed systems (bottom-up methodology). The applicable scales are overlapping ever more. Lithographic techniques in general are ever more refined so as to allow the creation of patterns many micrometers in size down to 20 nm. Bottom-up (self-assembly) techniques typically operates on a molecular scale (~ 1nm), but it is possible these days to manipulate macromolecules, or even (supra)molecular assemblies (colloids, DNA) in such a way that they form periodic structures with dimensions well over 10 nm.

It is therefore appropriate to identify a convergence of these methodologies. Self-assembly with supramolecular entities poses a problem however. How to control the exact chemical nature of for example a colloidal structure? Unit cells up to 50 nm can be generated in a plane, but control over the chemical functionality within the unit cell is remains a challenge. Self-assembly with supramolecular building blocks, that are exactly defined with respect to their chemical composition, is therefore an interesting next step. Input from synthetic chemistry will be necessary to create these next generation adsorbents for surface patterning.

2.8 References

1 Langmuir, I. J. Am. Chem. Soc. 1917, 39, 1848 – 1906.

2 Ulman, A. An introduction to ultrathin organic films: from Langmuir-Blodgett to self-

assembly. 1991, Academic Press. 3 Bigelow, W.C., Pickett, D.L., Zisman, W.A. J.Coll. Sci. 1946, 1, 513-538.

4 Nuzzo, R.G., Allara, D.L. J. Am. Chem. Soc. 1983, 105, 4481-4483.

5 Bain, C.D., Whitesides, G.M. J.Am.Chem.Soc. 1989, 111, 321-335.

6 Bain, C.D., Whitesides, G.M. Angew. Chem. Int. Ed. Engl. 1989, 28, 506 – 512.

7 Rabe, J.P., Bucholz, S. Phys. Rev. Lett. 1991, 66, 2096-2099 (b) Rabe, J.P., Bucholz, S.

Science 1991, 253, 641 – 644.

8 Tetradecane (C14H30) seems to be the smallest alkane that forms organized monolayers

on graphite. Chen, Y.J., Zhao, R.G., Yang, W.S. Acta Phys. Sin. 2005, 54, 284-290. 9 Weigl, B.H., Bardell, R.L., Cabrera, C.R. Adv. Drug Delivery Rev. 2003, 55, 349-377.

10 De Jong, M.R., Huskens, J., Reinhoudt, D.N. Chem. Eur. J. 2001, 7, 4164-4170.

11 Rozkiewicz, D.I., Ravoo, B.J., Reinhoudt, D.N. Langmuir, 2005, 21, 6337-6343.

12 Crampton, N., Bonass, W.A.., Kirkham, J., Thomson, N.H. Langmuir, 2005, 21, 7884-7891.

13 Kim, S., Cho, K.W., Curry, J.E. Langmuir, 2005, 21, 8290-8296.

14 Auletta, T., van Veggel, F.C.J.M., Reinhoudt, D.N. Langmuir, 2002, 18, 1288-1293.



Chapter 2

37

15

Cecchet, F., Rudolf, P., Rapino, S., Margotti, M., Paolucci, F., Baggerman, J., Brouwer, A.M., Kay, E.R., Wong, J.K.Y., Leigh, D.A. J. Phys. Chem. B 2004, 108 15192-15199.

16 Sahoo, R.R., Patnaik, A. Appl. Surf. Sci. 2005, 245, 26-38.

17 Groszek, A.J. Nature, 1962, 196, 531.

18 (a) Giancarlo, C., Flynn, G.W. Annu. Rev. Phys. Chem. 1998, 49, 297 - 336 (b) Giancarlo, C., Flynn, G.W. Acc. Chem. Res. 2000, 33, 491 – 501 (c) De Feyter, S., De Schryver, F.C., Chem. Soc. Rev. 2003, 32, 139 – 150 (d) Cyr, D.M., Venkataraman, B., Flynn, G.W. Chem. Mater. 1996, 29, 1600 – 1615 (e) De Feyter, S., Gesquiere, A., Abdel-Mottaleb, M.M., Grim, P.C.M., De Schryver, F.C., Acc. Chem. Res. 2000, 33, 520 – 531.

19 Shaporenko, A., Cyganik, P., Buck, M., Ulman, A., Zharnikov, A. Langmuir, 2005, 21, 8204-8213.

20 (a) Bieri, M., Burgi, T. J. Phys. Chem. B, 2005, 109, 22476-22485 (b) Swanson, S.A.,

McClain, R., Lovejoy, K.S., Alamdari, N.B., Hamilton, J.S., Scott J.C. Langmuir, 2005, 21, 5034-5039. 21

Lee, J.W., Sim, S.J., Cho, S.M., Lee, J. Biosens. Bioelectron. 2005, 20, 1422-1427.

22 Petoral, R.M., Uvdal, K. J. .Phys. Chem. B 2005, 109, 16040-16046.

23 Erts, D., Polyakov, B., Olin, H., Tuite, E. J. Phys. Chem. B 2003, 107, 3591-3597.

24 Schafer, A.H., Seidel, C., Chi, L., Fuchs, H. Adv. Mater. 1998, 10, 839 – 842.

25 For a good textbook, see Magonov, S.N., Whangbo, M-H. Surface analysis with STM and

AFM : experimental and theoretical aspects of image analysis, 1996, Weinheim VCH.

26 Binnig, G., Rohrer, H., Gerber, Ch., Weibel, E. Appl. Phys. Lett. 1982, 40, 178.

27 (a) Binnig, G., Rohrer, H. Helv. Phys. Acta 1982, 55, 726 (b) Binnig, G., Rohrer, H., Gerber, Ch., Weibel, E. Appl. Phys. Lett. 1982, 49, 57.

28 Tersoff, J., Hamann, D.R. Phys. Rev. Lett. 1983, 6, 313.

29 Spong, J.K. Nature,1989, 338, 187.

30 Ohtani, H., Wilson, J., Chiang,S., Mate, C.M. Phys.Rev.Lett., 1988, 60, 2398.

31 Giancarlo, L., Cyr, D., Muyskens, K., Flynn, G.W. Langmuir, 1998, 14, 1465 – 1471.

32 Lee, H.S., Iyengar, S., Musselman, I.H. Langmuir, 1998, 14, 7475 – 7483.

33 (a) Xia, Y., Whitesides, G.M., Annu. Rev. Mater. Sci. 1998, 28, 153 (b) Pritchard, D.J.,

Morgan, H., Cooper, J.M. Angew. Chem. Int. Ed. Engl. 1995, 34, 91.

34 Vaida, R., Tender, L.M., Bradley, G., O’Brien, M.J., Cone, M., Lopez, G.P. Biotechnol.

Prog. 1998, 14, 371.

35 (a) Matsui, S., Kojima, Y., Ochiai, Y. Appl. Phys. Lett. 1988, 53, 868 (b) Ada, E. T., Hanley, L., Etchin, S., Melngailis, J., Dressick, W.J., Chen, M.S., Calvert, J.M. J. Vac. Sci. Technol. B 1995, 13, 2189.

36 Delamarche, E., Sudarababu, G., Biebuyck, H., Michel, B., Gerber, C., Sigirst, H., Wolf,

H., Ringsdorf, H., Xanthopoulos, N., Mathieu, H.J. Langmuir, 1996, 12, 1997.




38

37

Husemann, M., Morrison, M., Beniot, D., Frommer, J., Mate, C.M., Hinsburg, W. D., Hedrick, J.L., Hawker, C.J. J. Am. Chem. Soc. 2000, 122, 1844.

38 Calvert, P. Chem. Mater. 2001, 13, 3299.

39 (a) Quist, A.P., Pavlovic, E., Oscarsson, S. Anal. Bioanal. Chem. 2005, 381, 591– 600 (b)

Xia, Y., Whitesides, G.M. Angew. Chem. Int. Ed. 1998, 37, 550.

40 Abbot, N.L., Folkers, J.P., Whitesides, G.M. Science 1992, 257, 1380.

41 Zhao, J., Uosaki, K. Langmuir, 2001, 17, 7784 – 7788.

42 Ivanisevic, A., Mirkin, C.A. J. Am. Chem. Soc. 2001, 123, 7887 – 7889.

43 Moreau, W.M. Semiconductor lithography: principles and materials. 1988, Plenum, New

York.

44 Yang, Y., Huang, S., He, H., Albert, W.H.M., Dai, L. J. Am. Chem. Soc. 1999, 123, 10832 – 10833.

45 Levensson, M.D. Solid State Technol. 1995, Febr: 57 – 66.

46 Goodberlet, J.G., Appl. Phys. Lett. 2000, 76, 667 – 669.

47 Luo, X., Ishihara, T. Jap. J. Appl. Phys. 2004, 43, 4017 – 4021.

48 Choi, S-J., Yoo, P.J., Baek, S.J., Kim, T.W., Lee, H.H. J. Am. Chem. Soc. 2004, 126, 7744 – 7745.

49 Naber, A., Fischer, U.C., Kirchner, S., Dziomba, T., Kollar, G., Chi, L.F., Fuchs, H. J.

Phys. Chem. B 1999, 103, 2709 – 2717.

50 Chang, W., Choi, M., Kim, J., Cho, S., Whang, K. Appl. Surf. Sci. 2005, 240, 296 – 304.

51 Sugimura, H., Hanji, T., Takai, O., Masuda, T., Misawa, H. Electrochim. Acta, 2001, 103 –

107.

52 Ha, K., Lee, Y-J., Sung, Y., Park, Y.S., Lee, G.S., Yoon, K.B. Adv. Mater. 2001, 13, 594 – 596.

53 Choi, C.H., Buriak, J.M. Chem. Comm. 2001, 1614 – 1615.

54 Gates, B.D., Xu, Q.B., Thalladi, V.R., Cao, T.B., Knickerbocker, T., Whitesides, G.M.

Angew. Chem. Int. Ed. Engl. 2004, 43, 2780 – 2783.

55 Brandow, S.L., Chen, M-S., Fertig, S.J., Chrisey, L.A., Dulcey, C.S., Dressick, W.J. Chem. Eur. J. 2001, 7, 4495 – 4499.

56 Cheng, X., Guo, L.J., Fu, P.F. Adv. Mater. 2005, 17, 1419.

57 Zhang, G.J., Tanii, T., Zako, T., Hosaka, T., Miyake, T., Kanari, Y., Funatsu, T.W.,

Ohdomari, I. Small 2005, 1 (8-9), 833-837.

58 Harnett, C.K., Satyalakshmi, K.M., Craighead, H.G. Langmuir, 2001, 17, 178 – 182.

59 (a) Liu, J., Casavant, M.J., Cox, M., Walters, D.A., Boul, P., Lu, W., Rimberg, A.J., Smith, K.A., Colbert, D.T., Smalley, R.E. Chem. Phys.Lett. 1999, 303, 125.



Chapter 2

39

60

Di Fabrizio, E., Fillipo, R., Cabrini, S., Kumar, R., Perennes, F., Altissimo, M., Businaro, L., Cojac, D., Vaccari, L., Prasciolu, M., Candeloro, P. J. Phys. Condensed Matter 2004, 16, S3517-S3535.

61 Xia, Y.N., Whitesides, G.M. Angew. Chem. Int. Ed. Engl. 1998, 37, 551-575.

62 Kumar, A., Whitesides, G.M. Appl. Phys. Lett. 1993, 63, 2002 – 2004.

63 Quist, A.P., Pavlovic, E., Oscarsson, S. Anal.Bioanal.Chem. 2005, 381, 591-600.

64 Xia,Y., Kim, E., Zhao, X-M., Rogers, J.A., Prentiss, M., Whitesides, G.M. Science, 1996,

273, 347 – 349.

65 Kim, E., Xia, Y., Whitesides, G.M. Nature, 1995, 376, 581 – 584.

66 Love, J.C., Wolfe, D.B., Jacobs, H.O., Whitesides, G.M. Langmuir, 2001, 17, 6005 – 6012.

67 Bruinink, C.M., Nijhuis, C.A., Peter, M., Dordi, B., Crespo-Biel, O., Auletta, T., Mulder, A.,

Schonherr, H.M., Vasco, G.J., Huskens, J., Reinhoudt, D.N. Chem. Eur. J. 2005, 11, 3988 – 3996.

68 Jenkins, A.T.A., Neumann, T., Offenhauser, A. Langmuir, 2001, 17, 265 – 267.

69 Paul, K.E., Prentiss, M., Whitesides, G.M. Adv. Funct. Mater. 2003, 13, 259 – 263.

70 Curtis, A., Wilkinson, C. Trends Biotech. 2001, 19, 97-101.

71 Mrksich, M., Whitesides, G.M. Trends Biotechnol. 1995, 13, 228.

72 Michel, R., Lussi, J.W., Csucs, G., Reviakine, I., Danuser, G., Kettere, B., Hubbell, J.A., Textor, M., Spencer, N.D. Langmuir, 2002, 18, 220-230.

73 Kenseth, J.R., Harnisch, J.A., Jones, V.W., Porter, M,D. Langmuir, 2001, 17, 4105 –

4112.

74 Jenker, M., Muller, B., Fromherz, P. Biol. Cybern. 2001, 84, 239-249.

75 Vogt, A.K., Wrobel, G., Meyer, W., Knoll, W., Offenhausser, A. Biomater. 2005, 26, 2549-2557.

76 Rozhok, S., Shen, C.K-F., Littler, P-L.H., Fan, Z ., Liu, C., Mirkin, C.A., Holz, R.C. Small,

2005, 1, 445 – 451.

77 Weibel, D.B., Lee, A., Mayer, M., Brady, S.F., Bruzewicz, D., Yang, J., Di Luzio, W.R.,

Clardy, J., Whitesides, G.M. Langmuir, 2005, 21, 6436 – 6442.

78 Rosi, N.L., Mirkin, C.A. Chem. Rev. 2005, 105, 1547 – 1562.

79 Rabe, J.P., Bucholz, S. Angew. Chem. Int. Ed. Engl. 1992, 31, 189-191.

80 Yin, X-L., Wan, L-J., Yang, Z-Y., Yu, J-Y. Appl. Surf. Sci. 2005, 240, 13-18.

81 Thalacker, C., Atsushi, M., De Feyter, S., De Schryver, F.C., Würthner, F. Org. Biomol.

Chem, 2005, 3, 414-422.

82 Xu, S-L., Yin, S-X., Liang, H-P., Wang, C., Wan, L-J., Bai, C-L. J. Phys .Chem. B 2004,

108, 620 – 624.




40

83

(a) Gesquiere, A., Abdel-Mottaleb, M.M.S., De Feyter, S., De Schryver, F.C., Schoonbeek, F., van Esch, J., Kellogg, R.M., Feringa, B.L., Calderone, A., Lazzaroni, R., Bredas, J.L. Langmuir, 2000, 16, 10385 - 10391 (b) Gesquiere, A., De Feyter, S., De Schryver, F.C., Schoonbeek, F., van Esch, J., Kellogg, R.M., Feringa, B.L. Nano Lett. 2001, 1, 201 – 206.

84 Mu, Z., Wang, Z., Zhang, X., Ye, K., Wang, Y. J. Phys .Chem. B 2004, 108, 19955 –

19959.

85 Pan, L-l., Zeng, Q-D., Lu, J., Wu, D-X., Xu, S-D., Tan, Z-Y., Wan, L-J., Wang, C., Bai, C-

L. Surf.Sci. 2004, 559, 70 – 76.

86 Stecher, R., Drewnick, F., Gompf, B. Langmuir, 1999, 15, 6490 – 6494.

87 De Feyter, S., De Schryver, F. Top.Curr.Chem. 2005, 258, 205-255.

88 Uosaki, K., Yamada, R. J. Am. Chem. Soc. 1999, 121, 4090 – 4091.

89 (a) Xie, Z.X., Xu, X., Tang, J., Mao, B.W. J. Phys. Chem. B 2000, 104, 11719 – 11722 (b)

Xie, Z.X., Xu, X., Mao, B.W., Tanaka, K. Langmuir, 2002, 18, 3113 – 3116.

90 Marchenko, O., Cousty, J. Phys. Rev. Lett. 2000, 84, 5363 – 5366.

91 Marchenko, O., Cousty, J., Pham Van, L. Langmuir, 2002, 18, 1171 – 1175.

92 Zhang,H-M., Xie, Z-X., Mao, B-W., Xu, X. Chem. Eur. J. 2004, 10, 1415 – 1422.

93 Yablon, D.G., Wintgens, D., Flynn, G.W. J. Phys. Chem. B 2002, 106, 5470-5475.

94 Wintgens, D., Yablon, D.G., Flynn. G.W. J. Phys. Chem. B 2003, 107, 173 – 179.

95 Yablon, D.G., Giancarlo, L.C., Flynn, G.W. J.Phys.Chem.B. 2000, 104, 7627 – 7635.

96 Wei, Y., Kannappan, K., Flynn, G.W., Zimmt, M.B. J. Am. Chem. Soc. 2004, 126, 5318 –

5322.

97 De Feyter, S., Gesquière, A., Wurst, K., Amabilioni, D.B., Veciana, J., De Schrijver, F.C. Angew. Chem. Int. Ed. Engl. 2001, 40, 3217 – 3220.

98 Yablon, D.G., Giancarlo, L.C., Flynn.G.W. J. Phys. Chem. B 2000, 104, 7627 – 7635.

99 Fang, H., Giancarlo, L.C., Flynn, G.W. J. Phys. Chem. B 1998, 102, 7311 – 7315.

100 Matzger, A.J., Kim,K. J. Am. Chem. Soc. 2002, 124, 8772 – 8773.

101 Plass, K.E., Kim, K., Matzger, A.J. J. Am. Chem. Soc. 2004, 126, 9042 – 9053.

102 Schattschneider, D. Amer.Math Monthly, 1978, 85, 439-450.

103 Gesquière, A., Abdel-Mottaleb, M.M., De Feyter, S., De Schryver, F.C., Sieffert, M.,

Mullen, K., Calderone, A., Lazzaroni, R., Bredas, J-L. Chem. Eur. J. 2000, 6, 3739 – 3746.

104 Giancarlo, L.C., Flynn, G.W. Acc. Chem. Res. 2000, 33, 491 – 501.

105 Grim, P.C.M., De Feyter, S., Gesquière, A., Vanoppen, P., Rucker, M., Valiyaveettil, S., Moessner, G., Mullen, K., De Schryver, F.C. Angew. Chem. Int. Ed. Engl. 1997, 36, 2601 – 2603.



Chapter 2

41

106

Abdel-Mottaleb, M.M.S., De Feyter, S., Gesquière, A., Sieffert, M., Klapper, M., Mullen, K., De Schryver, F.C. Nano Lett. 2001, 1, 353 – 359.

107 Hla, S-W., Rieder, K-H. Annu. Rev. Phys. Chem. 2003, 54, 307 – 330.

108 Kudo, H., Okamoto, Y., Kwak, K.J., Fujihara, M. Ultramicroscopy, 2004, 100, 353 – 358.

109 Shigekawa, H., Miyake, K., Sumaoka, J., Harada, A., Komiyama, M. J. Am. Chem. Soc. 2000, 132, 5411 – 5412.

110 Gesquière, A., Jonkheijm, P., Hoeben, F.J.M., Schenning, A.P.H.J., De Feyter, S., De

Schryver, F.C., Meijer, E.W. Nano Lett. 2004, 7, 1175 – 1179.

111 Miura, A., Jonkheijm, P., De Feyter, S., Schenning, A.P.H.J., Meijer, E.W., De Schryver,

F.C. Small, 2005, 1, 131-137.

112 Miura, A., Chen, C., Uji-I, H., De Feyter, S., Zdanowska, M., Jonkheijm, P., Schenning, A.P.H.J., Meijer, E.W., Wurthner, F., De Schryver, F.C. J. Am. Chem. Soc. 2003, 125, 14968 – 14969.

113 Wan, L.J., Itaya, K. Langmuir , 1997, 13, 7173 – 7179.

114 Wang, D., Wan, L.J., Xu, Q.M., Wang, C., Bai, C.L. Surf.Sci. 2001, 478, L320 – L326.

115 Wang, D., Xu, Q.M., Wan, L.J., Wang, C., Bai, C.L. Langmuir, 2002, 18, 5133 – 5138.

116 Strohmaier, R., Petersen, J., Gompf, B., Eisenmenger, W. Surf. Sci. 1998, 418, 91 – 104.

117 Janssens, G., Touhari, F., Gerritsen, J.W., van Kempen, H., Callant, P., Deroover, G.,

Vandenbroucke, D. Chem. Phys. Lett. 2001, 344, 1 – 6.

118 Tao, N.J., De Rose, J.A., Lindsay, S.M. J. Phys. Chem. B 1993, 97, 910 – 919.

119 Dmitriev, A., Lin, N., Weckesser, J., Barth, J.V., Kern, K. J. Phys. Chem. B 2002, 106,

6907 – 6912.

120 Dmitriev, A., Spillmann, H., Lin, N., Barth, J.V., Kern, K. Angew. Chem. Int. Ed. Engl. 2003, 42, 2670 – 2673.

121 Lackinger, M., Griessl, S., Markert, T., Jamitzky, F. J. Phys. Chem. B 2004, 108, 13652-

13655.

122 (a) De Feyter, S., Gesquière, A., Grim, P.C.M., De Schryver, F.C. Langmuir, 1999, 15,

2817 – 2822 (b) De Feyter, S., Grim, P.C.M., Rucker, M., Vanoppen, P., Meiners, C., Sieffert, M., Valiyaveettil, S., Mullen, K., De Schryver, F.C. Angew. Chem. Int. Ed. Engl. 1998, 37, 1223 – 1226.

123 De Feyter, S., Gesquière, A., De Schryver, F.C. Langmuir, 2000, 16, 9887 – 9894.

124 Yan, H-J., Wan, L-J., Bai, C.L. J. Phys. Chem. B 2004, 108, 11251 – 11255.

125 a) Lei, S.B., Wang, C., Yin, S.X., Wang, H.N., Xi, F., Liu, H. W., Xu, B., Wan, L. J., Bai, C.L. J. Phys. Chem. B 2001, 105, 10838 – 10841 b) Qiu, X.H., Wang, C., Zeng, Q.D., Xu, S., Yin, X., Wang, H. N., Xu, S. D., Bai, C. L. J.Am. Chem. Soc. 2000, 122, 5550 - 5556 c) Liu, Y.H., Lei, S.B., Yin, S.X., Xu, S.L., Zheng, Q.Y., Zeng, Q.D., Wang, C., Wan, L.J., Bai, C.L., J. Phys. Chem. B 2002, 106, 12569 - 12574; d) Qiu, X.H., Wang, C., Yin, S.X., Zeng, Q.D., Xu, B., Bai, C.L. J. Phys. Chem. B 2000, 104, 3570 – 3574 e) Lei, S.B., Yin, S.X., Wang, C., Wan, L.J., Bai, C.L. Chem. Mater. 2002, 14, 2837 – 2838.




42

126

Lackinger, L., Muller, T., Gopakumar, T.G., Muller, F., Hietschold, M., Flynn.G.W. J. Phys. Chem. B 2004, 108, 2270 – 2284.

127 Gimzewski, J.K., Stoll, E., Schlitter, R.R. Surf.Sci. 1987, 181, 267 – 277.

128 Yokoyama, T., Yokoyama, S., Kamikado, T., Okuno, Y., Mashiko, S. Nature, 2001, 413,

619 – 621.

129 Gopakumar, T.G., Lackinger, M., Hackert, M., Muller, F., Hietschold, M. J. Phys. Chem.B

2004,108, 7839 – 7843.

130 (a) Lippel, P.H., Wilson, R.J., Miller, M.D., Woll, C., Chiang, S. Phys. Rev. Lett. 1989,

62, 171 – 174 (b) Ludwig, C., Strohmaier, R., Peterson, J., Gompf, B., Eisenmenger, W. J. Vac. Sci. Technol. B 1994, 12, 1963 (c) Kanai, M., Kawai, T., Motai, K., Wang, X.D., Hashizume, T., Sakura, T. Surf.Sci. 1995, 329, L619 – L623.

131 Hipps, K.W., Scudiero, L., Barlow, D.E., Cooke, M.P. J. Am. Chem. Soc. 2002, 124, 2126

– 2127.

132 Xing, L., Hipps, K.W.J. J. Phys. Chem. B 1997,101, 5391 – 5396.

133 (a) Xing, L., Hipps, K.W., Wang, X.D., Mazur, U. J. Am. Chem. Soc. 1996, 118, 7197 –

7202 (b) Suto, K., Yoshimoto, S., Itaya, K. J. Am. Chem. Soc. 2003, 125, 14976 – 14977.

134 Lackinger, M., Hietschold, M. Surf.Sci. 2002, 520, L619 – L624.

135 Barlow, D.E., Scudiero, L., Hipps, K.W. Langmuir, 2004, 20, 4413 – 4421.

136 Lei, S.B., Wang, C., Yin, S.X., Bai, C.L. J. Phys. Chem. B 2001, 105, 12272 – 12277.

137 Xu, B., Yin, S., Wang, C., Qiu, X., Zeng, Q., Bai, C.L. J. Phys. Chem. B 2000, 104,

10502 – 10505.

138 Ikeda, T., Aksakawa, M., Goto, M., Miyake, K., Ishida, T., Shimizu, T. Langmuir, 2004, 20, 5454 – 5459.

139 Guo, X-L., Dong, Z-C., Trifonov, A.S., Miki, K., Kimura, K., Mashiko, S. Surf Sci. 2005,

241, 28 – 32.

140 Kromer, J., Carreras-Rios, I., Fuhrmann, G., Musch, C., Wunderlin, M.,

Debaerdemaeker, T., Mena-Osteritz, E., Bauerle, P. Angew. Chem. Int. Ed. Engl. 2000, 39, 3481 – 3486.

141 Mena-Osteritz, E., Bauerle, P. Adv. Mater. 2001, 13, 243 – 246.

142 Hoger, S., Bonrad, K., Mourran, A., Beginn, U., Moller, M. J. Am. Chem. Soc. 2001, 123,

5651 – 5659.

143 Safarowsky, C., Merz, L., Rang, A., Broekmann, P., Hermann, B.A., Schalley, C.A.

Angew. Chem. Int. Ed. Engl. 2004, 43, 1291 – 1294.

144 Venkateshwarlu, K., Ammon, H., Jackel, F., Rabe, J.P., Schmittel, M. Chem. Eur. J. 2004, 10, 5481 – 5492.

145 Clar, E. Aromatische Kohlenwasserstoffe: polycyclische systeme. 1952, Springer, Berlin.

146 Samori, P., Fechtenkotter, A., Jackel, F., Mullen, K., Rabe, J.P. J. Am. Chem. Soc. 2001,

123, 11462 – 11467.



Chapter 2

43

147

Iyer, V.S., Yoshimura, K., Enkelmann, V., Epsch, R., Rabe, J.P., Mullen, K. Angew. Chem. Int. Ed. Engl. 1998, 37, 2696 – 2699.

148 Ito, S., Wehmeier, M., Brand, J.D., Kubel, C., Epsch, R., Rabe, J.P., Mullen, K. Chem.

Eur. J. 2000, 6, 4327 – 4323.

149 Samori, P., Keil, M., Friedlein, R., Birgeson, J., Pakula, T., Watson, M.D., Mullen, K., Salaneck, W.R., Rabe, J.P. J. Phys. Chem.B 2001, 105, 11114 – 11119.

150 Jackel, F., Yin, X., Tchebotereva, N., Watson, M.D., Venturini, A., Mullen, K., Rabe, J.P.

Synthetic Metals, 2004, 147, 5 – 9.

151 Samori, P., Yin, X., Tchebotareva, N., Wang, Z., Pakula, T., Jackel, F., Watson, M.D.,

Venturini, A., Mullen, K., Rabe, J.P. J. Am .Chem. Soc. 2004, 126, 3567 – 3577.

152 Katsonis, N., Marchenko, A., Fichou, D. Synthetic Metals, 2004, 147, 73 – 77.

153 Wu, P., Zeng, Q., Xu, S., Wang, C., Yin, S., Bai, C-L. ChemPhysChem, 2001, 12, 750 –

754.

154 Scherer, J., Merz, L., Constable, E.C., Housecroft, C., Neuburger, M., Hermann, B.A. J.

Am. Chem .Soc. 2005, 127, 4033 – 4041.

155 Merz, L., Guntherodt, H-J., Scherer, J., Constable, E.C., Housecroft, C., Neuburger, M., Hermann, B.A. Chem. Eur. J. 2005, 11, 2307- 2318.

156 Loi, S., Wiesler, U-M., Butt, H-J., Mullen, K. Macromolecules, 2001, 34, 3661 – 3671.

157 Loi, S., Butt, H-J., Hampel, C.H., Bauer, R., Wiesler, U-M., Mullen, K. Langmuir, 2002,

18, 2398 – 2405.

158 Gyarfas, B.J., Wiggins, B., Zosel, M., Hipps, K.W. Langmuir, 2005, 21, 919 – 923.

159 Griessl, S.J.H., Lackinger, M., Jamitzky, F., Markert, T., Hietschold, M., Heckl, W.M. Langmuir, 2004, 20, 9403 – 9407.

160 Theobald, J.A., Oxtoby, N.S., Phillips, M.A., Champness, N.R., Beton, P.H. Nature, 2003,

424, 1029 – 1031.

161 Semonov, A., Spatz, J.P., Moller, M., Lehn, J-M., Sell, B, Schubert, D., Weidl, C.H.,

Schubert, U.S. Angew. Chem. Int. Ed. Engl. 1999, 38, 2547 – 2550 (b) Ziener, U., Lehn, J-M., Mourran, A., Moller, M. Chem. Eur. J. 2002, 8, 951 – 956.

162 Elemans, J.A.A.W., Lensen, M.C., Gerritsen, J.W., van Kempen, H., Speller, S., Nolte,

R.J.M., Rowan, A.E. Adv. Mater. 2003, 15, 2070 – 2073.

163 Lensen, M.C., van Dingenen, S.J.T., Elemans, J.A.A.W., Dijkstra, H.P., van Klink,

G.P.M., van Koten, G., Gerritsen, J.W., Speller, S., Nolte, R.J.M. Chem. Commun. 2004, 762 – 763.

164 Mena-Osteritz, E., Meyer, A., Langeveld-Voss, B.M.W., Janssen, R.A.J., Meijer, E.W., Bauerle, P. Angew. Chem. Int. Ed. Engl. 2000, 39, 2680 – 2684.

165 Zhang, G., Wang, D.Y., Gu, Z.Z. Langmuir, 2005, 21, 9143-9148.

166 Ray, M.A., Kim, H., Jia, L. Langmuir, 2005, 21, 4786-4789.

167 Endo, M., Seeman, N.C., Majima, T. Angew. Chem. Int. Ed. Engl. 2005, 44, 6074-6077.




44

168

Le, J.D., Pinto, Y., Seeman, N.C., Musier-Forsyth, K., Taton,T.A., Kiehl, R.A. Nano Lett. 2004, 4, 2343-2347.

169 Winfree, E., Liu, F., Wenzler, L.A., Seeman, N.C. Nature, 1998, 394, 539-544.

170 Mao, C.D., LaBean, T.H., Reif, J.H., Seeman, N.C. Nature, 2000, 407, 493-496.

171 Rothemund, P.W.K., Papadakis, N., Winfree, E. PloS Biology, 2004, 2, 2043 – 2053.

172 Whitesides, G.M., Boncheva, M. Proc. Natl. Acad. Sci. 2002, 8, 4769 – 4774.



45

Chapter 3 Incorporation of functionality into physisorbed monolayers

This chapter is devoted to the development of new approaches for decorating surfaces at the supramolecular level. The resulting patterns can be used for the spatial organization of functional groups, such as addressable groups, catalytically active groups, and so on. Functionality will preferably allow for post-modification, as this renders the aggregate suitable as a template.

3.1 Introduction

As described in chapter 2, formation of molecular assemblies at surfaces can be driven either by chemisorption or physisorption. Relying on physisorption, functionality can be incorporated in the adsorbents, by a ‘simple’ variation of the chemical composition along the alkyl chains. The functionality will be patterned on (periodically distributed over) a surface, when these molecules form a regular supramolecular aggregate at the interface.

1 The approach naturally contrasts to a

top-down type approach, in which the sharp tip of an STM or an atomic force microscope (AFM) can be used to form patterns on the nanoscale.

2

Metal centers would seem to be interesting candidates to be patterned, because they can be very specifically addressed, and they are so often the active center in catalytic reactions.

3 Their incorporation into a self-assembled monolayer

furthermore is proposed as a means to endowe the monolayer with interesting chemical properties, as the rigidity and directionality of the metal coordination allows for specific interactions. Formation of a monolayer from adsorbents, incorporating a complexation scaffold is described. This will allow the controlled 2D



Incorporation of Functionality

46

spatial disposition of metal centers that can act as anchor points for further functionalization.

3.2 About the platform. A modular appoach

The above implies that the envisioned molecules can be constructed in a modular way, yielding great (synthetic) versatility. Due to alkyl fragments in the molecules, favorable interaction with the substrate (graphite) is ensured. Additional interactions are introduced normal to the long axis of the molecules to add stability to the aggregates. A functional group, capable of complexation to a metal center, will be incorporated in the molecular design. Due to good experience in the past (vide infra) a choice was made for an ensemble of two urea groups per molecule. Urea groups form strong, bifurcated hydrogen bonds between each other, which strongly drives the formation of one-dimensionally organized tapes.

4 This principle

has been explored extensively by our group5 and others in the formation of low

molecular weight organogelators and their corresponding gels. A pictorial representation of the aggregate formed is shown in fig. 3.1.

Validity of this concept has been demonstrated with bisurea dervatives spaced by various alkyl chains.

4 Functional (non-alkyl) groups

6 have been shown to be

compatible with the strategy, in the sense that two-dimensionally organized monolayers can still be formed under the typical conditions, as can be seen in fig. 3.2. The image shown, together with a molecular model, belongs to a monolayer

F

F N N

O

H H

R

R

HH

O

NN

R

HH

O

NNR

HH

O

NN

R

HH

O

NN

N N

O

H H

R

R

HH

O

NN

R

HH

O

NN

R

HH

O

NN

R

HH

O

NN F

FF

Figure 3.1 Modular approach to 1-D assembly.



Chapter 3

47

formed from molecules of the general type (fig. 3.1), incorporating a bisthiophene moiety as the functional entity. It is expected that introduction of a moiety related in aromatic nature will lead to a monolayer with a comparable morphology (i.e. a lamellar structure).

3.3 About the complexant

The strategy will thus be to marry the formation of well-defined monolayers, which can be imaged with the STM, with functional groups that allow for post-modification. This is necessary if the monolayer is to be regarded as a template. In this context the organization of supramolecular coordination arrays, both at surfaces and in solution was instructive.

7,8 Based on metal complexation,

metallogrids were found to possess interesting electronic and magnetic properties.7

Metal coordination has also been exploited in the formation of supramolecular polymers,

9 showing the potential as a structurally defining element. Other

examples include the self-assembly of metal-organic coordination complexes and networks.

10 Furthermore, coordination complexes allow for electrochemical

manipulation, sometimes with structural implications, as the preferred coordination number or geometry may vary with the oxidation state of the metal.

11 In the context

of STM imaging, these features become even more attractive as the STM tip might provide a tool to locally address only parts of the monolayer. The rigidity and directionality of the metal coordination make it ideally suitable for the purpose of

Figure 3.2 STM image and model of a bisurea- bisthiophene derivative (molecular structure on top).

H25C12

NH

NH

O

S

SHN

HN

C12H25

O




48

further elaborating a spatially organized template to a three-dimensional aggregate.

Obviously one has to consider which specific ligand – metal pair can best be used. The metal coordination modes of oligopyridines render them the most ubiquitously used components in supramolecular systems.

12 Coordination is relatively strong in

these ligands due to the bidendate effect. The aforementioned metallogrids are based on bipyridine, terpyridine and quaterpyridine moieties.

7 Exchanging metals

can interconvert grid-type and helix-type superstructures,13

because the coordination mode varies with the metal. Bipyridine, being the parent compound of this family, constitutes an ideal candidate to start the exploration. Synthetic accessibility, structural resemblance to bisthiophene (both are composed of two planar, aromatic disks), and a wide variety of bipyridine complexes being known from literature

14 further point to the use of bipyridine. Bipyridines have been widely

used for the realization of discrete, highly ordered nanostructures based on transition metal coordination.

15 Based on their rich chemical properties we have

chosen bipyridine-derived molecules as a basis for template formation, that is, the controlled 2-D spatial disposition of metal centers that can act as anchor points for further functionalization. The conceptual step is oulined in fig. 3.3.

3.3.1 About the metal

Because there are a couple of assembly steps in coming to the final structure, there should be a means to elucidate complex formation (apart from literature reference). Many coordination complexes are electrochemically active. The oxidation potential(s) for the bare 2,2’-bipyridine are usually beyond the scope of simple cyclovoltammetry (CV), but a comparison between the characteristics of the metal source and the complex is quite instructive, to infer complexation. Another criterion is the stoichiometry of the complex. The best situation is unambiguous 1:1 complexation of a metal (with additional ligands) to the bipyridines, organized at the surface. This mode of coordination allows for subsequent manipuation by

S

S

N NN N

M

LL

S

S

N NN N

M

LL

Figure 3.3 Bipyridine instead of bisthiophene.



Chapter 3

49

replacing (exchanging) the additional ligands. A 2:1 stoichiometry (2 bipyridines coordinated to a metal) will possibly impose too much rigidity onto the supramolecular structure, either preventing formation of the monolayer in the first place, or seriously hindering the conceived addressing. Furthermore the complex must retain electrical neutrality, if it is to be imaged by means of STM. Electrical charges, or even dipoles near the substrates’ surface will create great problems by introducing noise to the STM images. This arises from the introduction of competing tunneling paths for the electrons in the substrate, and it is for the same reason that apolar solvents are used in STM experiments.

16 This introduces a final

criterion for the complexation in this context: the metal source must be soluble in apolar solvents, and the complexation must be possible in an apolar environment, and still proceed fast without heating. These factors introduce a number of constraints regarding the choice of the metal in the construction of a bipyridine-based template.

3.4 Design and synthesis of the adsorbents.

Possible target molecules combine the alkyl fragments with urea groups and a central bipyridine moiety. The connectivity between the bipyridine and the urea groups is an issue. Three connection schemes can be envisaged: meta, meta (relative to the bipyridine biaryl bond); para, para; or meta, para. The mixed combination would seem to make the synthesis unnecessarily complicated, and is thus discarded. The meta, meta combination will probably force the bipyridine moiety to stand up from the surface, whereas the para, para combination would be likely to impose a flat, coplanar conformation on the bipyridine (fig. 3.4). The former option looks most promising, as it is likely to facilitate reaction with the metal centers coming in from solution.

Spacers are a second issue. Length (number of methylene groups) of the spacers will affect the rigidity of the adsorbents, and thus the fixation of the functionality. For reasons of synthetic convenience spacers of intermediate length are

Figure 3.4 Presumed conformations of bipyridine-bisurea derivatives described in the text. The bipyridine moiety is indicated with an oval.

3.1

3.2




50

preferable. Electronic conjugation of the urea groups with the bipyridine ring-system is undesirable, because of unpredictable delocalization, leading to either altered behavior of the bipyridines or altered hydrogen-bonding properties of the ureas. In an early stage of the synthesis a synthetic handle should be appended to the bipyridines, which can subsequently be elaborated. Attaching an alkyl-spacer, terminated by an amine group, allows for easy conversion to a urea group. The reaction of primary amines with isocyanates is known to produce urea groups.

17 A

number of isocyanates is commercially available, including dodecyl isocyanate, which would introduce a long enough alkyl fragment for effective interaction with the graphite.

3.4.1 Urea groups in the meta position

4,4’-Dimethyl bipyridine has the meta, meta substitution pattern. The compound is commercially available, or it can be synthesized by reductive coupling of p-methyl pyridine using Raney nickel. Functionalization with several spacers has been considered. A 2-C spacer can be introduced by functionalizing dimethylbipyridine with a leaving group (reaction with NBS generates the dibromide). Subsequent reaction with cyanide anion, followed by reduction will give a 2-C terminal amine. A 3-C spacer can be introduced by generating the dimethylbipyridine dianion (from dimethyl bipyridine). This can be accomplished with LDA or n-BuLi. The reactive dianion can then be treated in situ with oxirane to generate a 3-C terminal alcohol. The alcohol can be converted to the amine in a few steps. A 4-C or a 5-C spacer can be introduced by reacting the aforementioned dianion with 1,3-bromochloropropane or 1,4-bromochlorobutane, respectively. Selectivity towards the more reactive halogen in these reactions should be sufficient. Again, the conversion of the 4-C - and 5-C terminal chlorides to the corresponding amines requires only a few steps (compare scheme 3.1). The approaches mentioned above have all been tested.

The synthesis of the 2-C amine proved to be difficult because the reduction of the bis-nitrile with various methods did not work out. Bromination of dimethyl bipyridine with NBS works fine (60% yield), and subsequent reaction with NaCN provided

N N

N N

NH

NH

NH

NH

C12H25

HN

HN

C12H25

HN

HN

C12H25

O O

C12H25

O O

3.1

3.2

Figure 3.5 Chemical structure of target compounds 3.1 and 3.2.



Chapter 3

51

the nitrile in 90% yield. But the reduction of the nitriles to amines requires rather drastic conditions, employing metal catalysts (Raney nickel), which gave unseparable reaction products, presumably because of coordination of these metals with bipyridine.

The synthesis of the 3-C alcohol (bipyridine functionalized with propanol) has been accomplished, albeit in low yield (10%). Reproducibility proved to be a problem. Reaction of the in situ generated dimethyl bipyridine dianion with oxirane (the gas was dissolved in THF, the amount established by weight) requires carefully controlled conditions. Formation of oligomers thwarts easy access to the desired product, and purification is quite difficult, due to the physical appearance (like tar) of the product mixture, and comparable polarities of the products.

The reactions with the bromo chloroalkanes were more successful. The assumed selectivity was indeed observed. Reaction of 4,4’-dimethylbipyridine dianion with 1,3-bromo chloropropane gave the desired chloride product with n-butyl spacer 3.5 (see experimental section) in 60% yield. This compound was converted into the azide 3.6 (idem, 95% yield), and subsequently into the amine 3.7 (98% yield) by catalytic hydrogenation over Pd-C. The amine was reacted with n-dodecylisocyanate, which led to the bisurea bipyridine product (3.1, fig. 3.5) in 30% yield. Care must be taken to keep the intermediate mono-ureas in solution (boiling toluene). The sequence is analoguous to the synthesis of compound 3.2, which is shown in scheme 3.1.

3.4.2 Urea groups in the para position

5,5’-Dimethylbipyridine has the para, para substitution pattern. Synthetic modification of this derivative is more cumbersome. This results from the distribution of electron density in the mono – and dianion. It is easily understood when one considers the possible electron flows of the 4,4’- and 5,5’ isomers (fig. 3.6). In the former the nitrogen atoms act as a sink for the increased electron density.

Figure 3.6 Comparison of resonance contributions in 4,4’-dimethyl bipyridine and (hypothetical) 5,5’-dimethylbipyridine dianion.

N N

N N




52

3.8

3.9

3.10

3.11

3.12

3.13

3.2

N N

Br O O

LDATHF

N N

OTHP

Br O O

LDA,THF

N N

OTHPTHPO

35%N N

OHHO

78% N N

BrBr

NaN3 , DMSO

99% N N

N3N3

H2, Pd-C

99% N N

NH2H2N

C12H25NCO

30%

N N

NHCONHC12H25C12H25HNOCHN

HBr-H2SO4

90%

88%

toluene

HCl, H2O

2

Scheme 3.1 Synthesis of compound 3.2.



Chapter 3

53

As a result of the accumulating charge in 5,5’-dimethylbipyridine, the dianion cannot be generated. Therefore, elaboration of the spacers has to proceed in two steps. First one spacer has to be introduced, followed by the other. This methodology limits the number of derivatives that are reasonably accessible for the para, para substituted bipyridine. Bromochloropropane cannot be used in the sequence, because the monofunctionalized bipyridine-alkylchloride would be reactive towards itself (chloride is a leaving group) when in the second step the anion is generated. To circumvent these problems a strategy avoiding potential leaving groups has to be invoked. Obviously there still has to be a suitable handle at the terminus. The problem can be attacked by introducing a protected alcohol (as THP ether). This ether is supposed to be base-stabile, and indeed proved to be so. The mono-functionalized THP derivative 3.8 was obtained in good yield (90%). Next, the second leg can be appended without problems. The bis-THP derivative 3.9 was obtained in 88% yield as well. After deprotection (35% yield), the alcohol 3.10 was converted to a bromine 3.11 in 78% yield, and the rest of the synthesis proceeded as described above via the azide 3.12 and the amine 3.13 (quantitative yields) towards the bisurea product 3.2, which was obtained in 30% yield. The route is outlined in scheme 3.1. All compounds were characterized by 1H-NMR and

13C-NMR. Products 3.1 and 3.2 have been characterized by MS and

elemental analysis as well.

3.5 Formation of monolayers.

The aggregation behavior of compound 3.1 was first investigated with STM. A droplet of a solution (~ 1mg/ml) in phenyloctane was applied to the basal plane of HOPG (highly oriented pyrolytic graphite). Imaging this compound was quite cumbersome. A representative image is shown in fig. 3.7. Although the alkyl fragments are clearly visible; every second atom is visible (which is expected because of the zig-zag conformation of the alkyl fragment at the interface), the most important part of the molecule, the core, is not very well resolved. This observation might be explained by the conformation that the molecules are likely to adopt on the surface. If the bipyridines are not flat on the

Figure 3.7 STM image of a monolayer formed by compound3.1 on HOPG. Iset = 0.4 nA, Vset = - 0.304 V.

1.5 nm.




54

surface, but dangling a little above it, obtaining decent resolution will be complicated. Though it was actually anticipated that the bipyridines would bend away from the surface, such poor resolution is not encouraging. Absence of submolecular resolution will probably hamper the observation of complexation. The difficulty in imaging might result from mobility of the bipyridine moieties, which is introduced by the increased rotational and vibrational degrees of freedom in the spacer as compared with a situation where the spacers interact closely with the substrate. The thiophene bisureas (vide supra) were never observed to partially lift from the surface.

5

Compound 3.2 was imaged next. The compound did indeed form a physisorbed monolayer. A representative image is shown in fig. 3.8, together with a molecular model (constructed in hyperchem). The model is added to corroborate ideas about the structure. Resolution over the entire structure is much improved as compared to the image in fig. 3.7. In this image the bright tape in the middle is attributed to the bipyridine cores. The bipyridines are expected to show up brightly, because they consist of aromatic parts. Aromatic groups are associated with a higher density of states (MO’s) then aliphatic groups, thus providing more pathways for the electrons to tunnel. The respective urea groups to the left and the right can be discerned with differing contrasts; one is darker, one is brighter. This is a feature that has been previously observed in similar systems

18 and which could be related

to the opposite directionality, as exemplified by the orientation of the carbonyl, of the hydrogen-bonding arrays. To the far ends of the lamellae the alkyl chains can be seen.

Figure 3.8 STM image (left) and model (right) of monolayer formed by compound 3.2 The image size is 10.8 x 10.8 nm

2; Iset = 0.4 nA, Vset = - 0.304 V.



Chapter 3

55

As before, half the number of carbon atoms is visible. The bipyridine moieties are assumed to be not completely parallel to, but tilted off the surface (vide infra). The supramolecular organization is lamellar. The borders between the lamellae are clearly visible as a dark trench. Individual parts of the molecules are specified by their relative orientation with respect to the long axis of a lamella. The bipyridine moieties make an angle of 70.2 ± 2.0° with respect to the lamella axis; the alkyl tails make an angle of 81.1 ± 1.0°. The observed width of the lamellae is 6.5 nm, which corresponds with the length of the molecules in their extended zig-zag conformation. The inter-molecular distance within a lamella is 0.46 ± 0.03 nm. This distance is indicative of hydrogen-bond formation between the molecules, as this is close to the ideal value (0.46 nm) for hydrogen-bonded urea groups.

19

The supramolecular aggregate seems to be dominated by the H-bonding interactions, which forces a tilted orientation onto the bipyridines. There is, however, no information concerning the conformation of the bipyridines. The pyridine nitrogen atoms can either be located towards the graphite surface, or towards the supernatant solution; either the bipyridines adopt a cisoid or a transoid conformation. In solution bipyridines have a slightly skewed transoid conformation but this cannot be a priori extrapolated to this situation. It is likely that a combination of conformations will be found, when looking in more detail, as any conformation (arising from rotation around the bridging bond) will be locked if a sufficiently large number of molecules is glued together by hydrogen bonding. In connection with the complexation that should be performed, a cisoid - nitrogens up conformation would be ideal.

3.6 Complexation experiments

On the basis of the criteria given in paragraph 3.3 a number of metal salts from various sources had been selected for addition to the system. Unambiguous formation of 1:1 complexes proved to be not as abundantly reported as initially thought. Especially the d

8 metals (palladium, nickel and mercury) seem amenable

to formation of 1:1 square planar complexes with bipyridine.20

Square planarity is deemed advantageous, because of reduced spatial constraints in the monolayer. Reported electrochemical activity

21 advocated the use of palladium acetate

(Pd(OAc)2) or palladium chloride (PdCl2). Solubility and reactivity towards bipyridine has been tested in phenyloctane, a typical solvent in STM measurements. The reaction proceeded smoothly and instantaneously (the product precipitated and

1H-NMR indicated formation of the complex). In the STM

cell however, addition of either of these species (Pd(OAc)2 and PdCl2) never led to the observation of restructuring of the monolayer, or to any new features attributable to a complexation event.

Typically, these experiments were performed on a preformed monolayer. The metal source was then added in a dilute solution. A drop of this solution was added




56

to the already applied drop containing the adsorbents. Usually the monolayer even disappeared, leaving only the graphite surface. The failure to affect complexation in this system can be attributed to either the conformation of the bipyridine, as this could not be unequivocally established from the STM images before, or the spatial requirements of the complexed species. Influencing the conformation in situ (on the surface) however would a priori seem to be too challenging a task.

On the other hand it is conceivable that the supramolecular structure is too densely packed, so that the dynamics that are usually prominent in physisorbed monolayers

22 are greatly reduced. This situation mainly arises because of the

dominance of the hydrogen-bonding in the total of interactions. Therefore hydrogen-bonding might actually preclude the formation of complexes in the monolayer. Either the close-packed structure does not fulfill the spatial requirements to accommodate the metals (i.e. the gain in free energy that would be achieved by metal complexation does not compensate for the loss of eight hydrogen bonds), or a complexation event in situ, onto the pre-arranged bipyridines, is disfavored compared to the process in solution. The observed disappearance of the monolayer upon addition of the metal species might indicate that the complexation takes indeed place in solution. Spatial constraints subsequently prevent readsorption and formation of a complexed monolayer.

Too strong interactions stabilize the monolayer to the extent that lateral diffusion of the complexants is on the surface is diminished, so there is also no possibility for the aggregate to rearrange on the surface and thus make room for the complexed species. Both interpretations advocate the introduction of more flexibility and less stringent control over the intermolecular interactions.

3.7 Bipyridine derivatives without urea groups

Without urea groups, the interaction between the adsorbents will be controlled by weaker van der Waals interactions. These will be in the same order of magnitude as the interactions between the adsorbents and the substrate. It is hard to forecast how this new balance will influence the structure that will be formed. Nevertheless it seems plausible that a monolayer, if formed, will be more dynamic and more likely to be in an equilibrium situation. From a synthetic point of view it is not a very demanding task to functionalize a bipyridine with two long alkyl tails. These chains should at least have a length of 12 C-atoms because interaction with the graphite should be provided.



Chapter 3

57

Synthesis involved deprotonation of 5,5’-dimethylbipyridine and a substitution reaction with bromooctadecane, which is commercially available. As before, the reaction proceeded in two steps, first generating the monoalkylated species (compound 3.3, in 55% yield), and subsequently the dialkylated species (compound 3.4; 40% yield) was obtained (scheme 3.2). The compounds were characterized by means of

1H-

NMR, 13

C-NMR, MS and elemental analysis.

3.7.1 Surface assembly

Upon applying a drop of a solution of compound 3.4 (dialkylated bipyridine) in phenyloctane to a graphite surface a physisorbed monolayer is spontaneously formed. The STM image can be seen in fig. 3.9. On the right side a molecular model is shown to corroborate the conclusions made. This image is sub-molecularly resolved. The supramolecular structure is again characterized by a lamellar morphology, not unlike the one observed for the bisurea derivatives. The lamella is defined by two black troughs, which are characteristic for terminal methyl groups. The bipyridine moieties form an angle of 64.0 ± 2.8° while the aliphatic chains form an angle of 49.0 ± 2.8° with respect to the lamellar axis. For clarity one molecular model of 3.4 has been superimposed on the STM image. Above the image and model the chemical structure of compound 3.4 is shown. From the contrast variations along the lamella axis it can be concluded that all the molecules are equivalent (same orientation). The distance between two neighboring molecules within a lamella measured along the lamella axis is 0.69 ± 0.03 nm and the distance between equivalent points in abutting lamellae ∆L is 5.1 ± 1.6 nm. This corresponds to the length of the molecules if the alkyl chains adopt a fully extended zig-zag conformation. Compared to the monolayers formed by compound 3.2, the intermolecular (between molecules in the same lamella) distance is about 50% larger in the monolayers of 3.4. This corroborates the hypothesis of hydrogen-bond domination in the interaction scheme leading to formation of monolayers by compound 3.2. The packing parameters acquired from the STM image, indicate that the bipyridine moieties are adsorbed parallel to the graphite plane. Though well resolved, again there is no information on the conformation of the bipyridine moieties. If the molecules in monolayers of 3.4 are indeed less tightly bound, this might allow for considerable dynamics, thus favoring

Scheme 3.2 Synthesis of compounds 3.3 and 3.4.

N N

N N

C19H39

1. LDA

2.C18H37Br

55%

1. LDA

2.C18H37Br

40%

N N

C19H39 C19H39

3.3

3.4




58

complexation. Another indication that this is a less tightly bound aggregate is the observation of irregularities in some of the images.

This is demonstrated in fig. 3.10. Lamellae with different widths (∆L1 and ∆L2) are observed. Whereas ∆L1 is about 5.2 nm, the same distance as measured before (fig. 3.9), ∆L2 is larger, about 5.9 ± 1.5 nm. It is believed that the wide lamella is made up from the same adsorbents, which here have partially desorbed. In this case one alkyl chain has moved out of the packing regime, presumably dangling in the supernatant solution. This process has created empty space, which the system seeks to fill up. In this aggregate this is achieved when molecules in the abutting lamella make the same move. What looks like a wide lamella is actually composed of two molecules that have their respective bipyridine moieties oriented towards each other. Because the aliphatic chains make up for the largest part of the molecules, measured along the long axis, desorption of one chain creates a species whose length along the surface is slightly more then half that of the original. As a result the widths of the respective lamellae are not dramatically different. The width of the cores (the part showing up brightly) in the wide lamella in fig. 3.10 is 1.82 ± 0.13 nm, which is consistent with the above explanation.

N N

Figure 3.9 STM image (left) and model (right) of monolayer formed by compound 3.4. The image area is 9.1x 9.1 nm2; Iset = 1.2 nA, Vset = - 0.244 V.



Chapter 3

59

3.8 Complexation experiments with the new bipyridine derivatives

Once the two-dimensional ordering of the bipyridine derivatives had been established, experiments were continued by addition of the metal sources. The same sources as previously described (Pd(OAc)2 and PdCl2) were used. A drop of a concentrated solution of the chosen metal source, in phenyloctane, was added in situ to the preformed bipyridine monolayer (a drop was added to a drop already present on top of the graphite). The addition of palladium acetate to a monolayer of 3.4 led almost immediately to a dramatic change in the packing and morphology of the monolayer. Fig. 3.11 shows the STM image and the model of the situation after addition. The image is sub-molecularly resolved, which again enables identification of the aliphatic chains as well as the complexation sites. The bipyridine moieties cannot be discerned; instead, well-defined large bright structures appear. This presents the most striking change in the monolayer, but closer inspection reveals that also the distances have changed. The distance ∆L between two successive bright structures measured along the lamella axis is 0.94 ± 0.01 nm, which is significantly larger than the distance of 0.69 ± 0.03 nm between bipyridine moieties

Figure 3.10 Single and double bipyridine moieties stacked along the lamella direction are indicated by big arrows. The image size is 10.6x 10.6 nm

2; Iset = 1.2 nA, Vset = -0.476 V.




60

before the addition of Pd(OAc)2. A change in the packing pattern of the aliphatic chains occurs too, and the chains appear to be interdigitated. The data suggest that one bipyridine unit complexes to each metal complex. This can be explained as follows: the distance between the neighboring molecules increases to 0.94 nm in order to accommodate the Pd(OAc)2 moieties, causing the aliphatic chains to interdigitate in order to reduce the free space in the monolayer, as is revealed by the large decrease in distance between adjacent rows of bipyridine units (∆L2 = 3.51 ± 0.13 nm as compared to 5.10 ± 0.16 nm previously). It must be noted that the orientation of the aliphatic chains with respect to the lamellar axis changes from 49.0 ± 2.8° to 87.0 ± 2.8° after addition of Pd(OAc)2. In the model presented here the acetate moieties are believed to be still attached to the metals on the surface. To further clarify the packing pattern, two complexed molecules have been superimposed on the STM image. These results are interpreted in terms of a rearrangement on the surface.

Figure 3.11 STM image and model of a monolayer of 3.4 after addition of Pd(OAc)2. The image area is 10.2 x 10.2 nm

2 ; Iset = 1.2 nA, Vset = - 0.486 V.

Figure 3.12 Interdigitation. The absence of H-bonding allows the monolayer to rearrange.



Chapter 3

61

The observed packing in the complexed monolayer belongs to an interdigitated structure. Interdigita- tion,

23 i.e. the protrusion of molecules

in neighboring lamellae, so that the lamellae partially overlap, is a common phenomenon in physisorbed systems (fig. 3.12). Formation is largely entropy-driven, minimizing the free (not covered) space at the surface. It is especially prevalent in the absence of strong interactions. The result corroborates presumptions about flexibility and dynamics. Because of the freedom to rearrange, without enthalpic penalty, the spatial constraints imposed by the newly formed bipyridine-palladium complexes can be met. It is remarkable that the area per molecule is about the same in both cases (0.69 x 5.10~ 0.94 x 3.50 nm). The dynamic nature of these bipyridine monolayers is apparent. Directly upon addition usually a coexistence of complexed and non-complexed species is observed, a situation that transforms after a while in a uniformly distributed fully complexed monolayer. This is shown in fig. 3.13. Two domains can be identified, domain A with a packing pattern indicative of bipyridines without metals, and domain B with a packing pattern indicative of bipyridine-palladium complexes. At the domain boundary, increased mobility of the molecules in some locations is evident (indicated by arrows). Clearly this image represents a transient situation.

3.8.1 Bipyridine monolayer is a generic scaffold for complexation

The complexation is not limited to one metal or source. This is an important finding, as it demonstrates the generality of the concept. Fig. 3.14 shows representative images of the monolayers after addition of the respective metals. Images 3.14 A and 3.14 B show the monolayer after addition of PdCl2.

Palladium chloride is an interesting metal-ligand combination, because it is, more than palladium acetate, amenable to reduction to a Pd

0 state.

24 Pd

0 is known to be

a good nucleation agent for electroless deposition of copper.25

So the use in this methodology potentially opens up routes towards molecular scale gratings, from which nanoelectrodes can be built.

Figure 3.13 Coexistence of complexed (area B) and non-complexed species

(area A). ∆L1 = 4.75 ± 1.10 nm and

∆L2 = 3.51 ± 0.13 nm.The image size is 21.5 x 21.5 nm

2; Iset = 0.6 nA, Vset = -

0.418 V. The arrows indicate the phase boundary.

A

B

A

B




62

It should be noted that PdCl2 is not soluble in phenyloctane. This problem was overcome by adding a drop of acetonitrile to the suspension of palladium chloride in phenyloctane. Although acetonitrile itself, on behalf of its polarity, would disturb the measurement, it was anticipated that it evaporates quickly enough to not effectively interfere. The images in fig. 3.14 A and B are well resolved. The structural reconstitution is the same as the one upon addition of palladium acetate. That means interdigitation, to create extra space. The characteristic distances, as described above are also found here (~ 3.5 nm). This is not too surprising as the size of PdCl2 is almost the same as that of Pd(OAc)2. The angle the alkyl fragments make to the long axis of the lamella is slightly smaller (82°) as before. The metal centers show up quite distinctly, but the extent to which this reveals the chemical nature of PdCl2 as compared to Pd(OAc)2 remains a matter of speculation.

Not only the source of the metal has been varied, but also the metal itself. This is illustrated by fig. 3.14C – F. In these experiments copper was the metal to be added to the bipyridine monolayer, either in the form of Cu(OAc)2.H2O or CuCl2.xH2O. Copper has a more complicated behavior in its complexation with bipyridines, because more coordination geometries and stoichiometries are known, including polymeric species.

26 It was therefore a bit surprising that the results

shown could be obtained. The monolayers again show the same periodicities as before, and the images are well resolved. The angles (between the molecular and lamellar axes) are a bit smaller as compared to the Pd-complexed monolayers, (78° for Cu(OAc)2 ; 72° for CuCl2).



Chapter 3

63

Figure 3.14 Results of complexation experiments with different metals and sources. STM images of monolayers of 3.4 complexed with: (A, B) PdCl2; (C, D) Cu(OAc)2; (E, F) CuCl2 . A) The image area is 11.2 x 11.3 nm

2; Iset = 1.2 nA, Vset = - 0.142 V. B) The

image area is 50.0 x 50.0 nm2; Iset = 1.4 nA, Vset = -

0.160 V. C) The image area is 12.9 x 12.9 nm2; Iset =

0.8 nA, Vset = - 0.450 V. D) The image area is 50.0 x 50.0 nm

2; Iset = 0.8 nA, Vset = - 0.472 V. E). The image

area is 13.1 x 13.1 nm2; Iset = 1.0 nA, Vset = - 0.460 V.

F) The image area is 50.0 x 50.0 nm2; Iset = 0.8 nA,

Vset = - 0.440 V.




64

3.9 The monoalkylated bipyridine

To make the picture complete also the monoalkylated bipyridine was imaged. An important reason for this experiment was to provide a control for the irregularities observed in the monolayers of 3.4.

The detailed conclusions of this work are described elsewhere.

27 Suffice it to

say that the results were in accordance with the model presented in paragraph 3.7. A representative image is seen in fig. 3.15. Due to symmetry breaking two different widths are observed. There is actually only one kind of lamella, but the distances between the bright markers (bipyridines) are variable. This is because tail-to-tail as well as head-to tail orientations are possible. The image is not so well resolved because only one alkyl chain is available now for interaction with the graphite. The intermolecular distance between neighboring molecules along a lamella axis is 0.65 ± 0.04 nm, indicating that the alkyl chains are not interdigitated. It transpired that also the monolayers of compound 3.3 could be complexed to palladium acetate (fig. 3.16). A spontaneous change in the packing was observed. The molecules are again packed in a lamellar-type structure, in which the alkyl chains are interdigitated; two arrows indicate a lamella. The intermolecular distance between neighboring molecules along the lamella is 0.92 ± 0.05 nm. This value is identical (within experimental error) to that observed for the complexed monolayer of 3.4. Remarkably, after complexation, no different packings are observed; all domains show the same packing, and the lamellae are oriented head-to-head. Addition of the metal salts and consequent complexation leads to the formation of a unique 2D pattern.

Figure 3.15 Chemical structure and STM image of monolayer of

compound 3.3 (∆L1 = 2.55 ± 0.12 nm

(head-to-tail); ∆L2 = 3.57 ± 0.09 nm (tail-to-tail). Six molecular structures are superimposed on the STM image. The image area is 15.2 x 15.2 nm

2; Iset

= 0.8 nA, Vset = - 0.436 V.

N N

C19H39



Chapter 3

65

3.10 Can the template be used ?

After establishing that stable monolayers can be formed from compounds 3.3 and 3.4, and that in situ addition of metals leads to complexation, which is indicated by changes in the packing motif of the monolayers, and the appearance of hitherto absent bright spots, one should address the issue of how to use this nanoscale template. From the outset it had been anticipated that the metals would complex to the bipyridines, keeping their original ligands (the coordination number of the metals is changed from two to four in the process, though their formal oxidation state does not change).

The STM images do not give evidence of this actually being the case in the monolayers, but the metal-bipyridine complexes have been chosen such that literature data

28 makes it likely that indeed the additional ligands are still on the

surface. The trick is now to exchange (displace) these additional ligands with a second equivalent of bipyridine, or another ligand with strong affection for the respective metal. The second, incoming bipyridine can be synthetically modified, to be appended to another functional group. This can be a bioactive (biotin) -, or a catalytically active group. It can also be another bipyridine, which gives way, in a third step, to repeat the complexation, and so on and so forth. This would present a layer-by-layer approach towards nanostructured materials.

Attempts were made to effectuate such a displacement in the case of the monolayer formed by dialkylbipyridine 3.4, complexed with palladium acetate. A ligand with higher affinity for palladium then acetate was chosen to affect the displacement. A bidentate ligand is presumably necessary, and a choice was made for dibromomaleic acid

29. This ligand will coordinate to palladium with the

Figure 3.16 Large scale and smaller scale STM images of monolayers of 3.3, after addition of Pd(OAc)2. The image areas are 24.1x 24.1 nm

2

on the left and 10.8 x 10.8 nm2 on the right; A) Iset = 1.3 nA, Vset = - 0.5

V. B) Iset = 1.2 nA, Vset = - 0.6 V.




66

carboxylate groups, thus presenting a similar mode of binding as the acetates, but it will add a bidendate effect. It was hoped that the bromines, in case of a successful exchange, would show up with a pronounced contrast in the STM images.

30 Disappointingly, upon addition of a dilute solution of this ligand to the

preformed complexed monolayer, in all cases the only thing observed was the original, non-complexed monolayer. This gives the opportunity to switch back-and-forth between a complexed and a non-complexed state, but that is not really interesting. This result, though, does add information to the general picture, as it can be interpreted as follows: instead of an exchange with the acetates, an exchange with the bipyridines was affected. This means that the bipyridines must have desorbed from the graphite, and the reaction took place in solution, creating a malonate-diacetate palalladium species and the free bipyridines, which can subsequently reassemble onto the graphite. Extrapolating this view, it seems likely that also the original complexation takes place in solution, onto a desorbed bipyridine, and the newly formed complex assembles into a new type of monolayer. This is fully consistent with the dynamic nature that has been observed for monolayers of 3.4.

3.10.1 Working at the solid-air interface

A way around the problem of dynamics is to work at the solid-air interface. This can be achieved by generating a complexed monolayer, as described above, at the liquid-solid interface, followed by slowly evaporating the solvent in the STM cell. Phenyloctane is not very suitable to do this, due to its high boiling point and low volatility. The procedure was therefore repeated in 1-heptanol. 5,5’-Dialkylbipyridine was applied to the graphite as a solution in heptanol (1 mg/ml). The system was left to dry for 2 days under ambient conditions, and then imaged with STM. Fig. 3.17 shows the image obtained. It is not so well resolved, but has all the characteristics of the monolayer that is formed at the liquid-solid interface. This proves that it is possible to construct a nanotemplate outside a wet environment. It should be noted that the image in fig. 3.17 probably does not show a monolayer. It is more likely to be the top (or the bottom) of a multilayer. One can now think about complexation from the gas phase. Gaseous metal complexes, also reactive towards bipyridine can be envisaged (e.g. pentacarbonyliron).

31 This

experiment has not been performed though, because of possibly dangerous situations (toxicity of the gas coupled with lack of ventilation).



Chapter 3

67

3.11 Conclusions

It was demonstrated that nanoscale organization of chemically relevant functionality can be programmed at the molecular level by smart design of complexants forming a physisorbed monolayer at solid-liquid interface. The influence of hydrogen-bonding has been addressed, and was found to be dramatic. Thus, alkylated 2,2’-bipyridines can be used as complexation scaffolds for metal binding, both at the liquid/solid and liquid/air interfaces, whereas the bisurea derivatives do not participate in any complexation event. The intermolecular distance within the non-complexed monolayer is of crucial importance in determining whether or not complexation can occur. If the intermolecular distance is too small, as in the case of the urea derivative, no complexation is observed. This is thought to be due to insufficient space between the adsorbed molecules to allow for the accommodation of the metal ions on the surface – as a result of the stability of the hydrogen-bonded array – or to the desorption of complexed molecules. Addition of a metal complex, either ex situ or in situ, to both symmetric alkylated and asymmetric alkylated bipyridine derivatives leads to complexation. The situation is summarized in the cartoon in fig. 3.18. Complexations with different metal ions led to the formation of an identical 2-D ordering (i.e. interdigitated structures with a distance between the complexed centers of about

Figure 3.17 STM image of 3.4 at the graphite-air interface. The image area is 10.2x 10.2 nm

2 ; Iset = 1.00 nA, Vset =

-0.498 V.




68

3.5 nm; with a small variation in the angles made with respect to the long axes of the lamellae).

A delicate balance of interactions (van der Waals and possibly H-bonding), between the adsorbent molecules, as well as between adsorbent and substrate, has to be taken into account. Also the contribution of the solvent should not be ignored. It is not claimed that the process is fully understood. Clearly, too strong interactions between the adsorbents make it difficult for the aggregate (monolayer) to restructure. Therefore, complexation was not observed. Chemical reactions onto the monolayer in general might be difficult. Too weak interaction between the adsorbents allow for complexation, but make it difficult to modify this structure, presumably due to high mobility. More knowledge about the delicate interplay of forces in these supramolecular systems on surfaces is necessary to exploit its full potential. It will be necessary to test and investigate different complexes with different bonding strengths with respect to the bipyridine molecules or other methods of changing the ligands.

The long-term goal of these experiments is to form stable templates, and then to use such templates for the assembly of other molecules or nanostructures in the third dimension. The stability of such templates in air is a requisite. As a

steric hindrance/ electrostatic repulsion

Figure 3.18 Cartoon representation of the different situation for H-bonding complexants and non-H-bonding complexants with respect to complexation of metal centers.



Chapter 3

69

demonstration of such stability, STM investigation before and after evaporation of the solvent revealed apparently identical monolayers in the case of alkylated bipyridines. These potential templates are sufficiently stable to be used in building 3-D nanostructures.

3.12 Experimental section

Materials and methods 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-

300 spectrometer (at 300 MHz) with samples in CDCl3 or [d6] DMSO; chemical

shifts are given in ppm relative to CDCl3 (δ = 7.24 ppm) or DMSO (δ = 2.50 ppm). 13

C NMR spectra were recorded on a Varian VXR-300 spectrometer (at 75.48 MHz) with samples in CDCl3 or [d

6] DMSO; chemical shifts are given relative to

CDCl3 (δ = 77 ppm) or DMSO (δ = 39.5 ppm) The splitting patterns in the 1H NMR

spectra are designated as follows: s (singlet), d (doublet), dd (double doublet), t (triplet), m (multiplet), br (broad). Melting points were measured on a Stuart Scientific SMP1 apparatus. Infrared spectra were recorded on a Nexus FTIR spectrometer. HRMS was performed on a JEOL JMS 600H spectrometer in EI+ ionization mode. Elemental analyses were carried out in the Microanalytical Department of the Stratingh Institute, University of Groningen (The Netherlands).

4-(4-chlorobutyl)-2-(4-(4-chlorobutyl)pyridin- 2-yl)pyridine (3.5): A solution of LDA (30 mmol) was freshly prepared from diisopropylamine (1.6 mL) and n-BuLi (7 mL, 1.6M in hexanes) in THF (4 mL) at - 80 ºC. A solution of 4,4’-dimethyl-2,2’-bipyridine (1.0 g,

5.4 mmol) in THF (150 mL) was then added at the same temperature. The reaction mixture was stirred for 1.5 h at -80ºC. It was then allowed to warm to room temperature, whereupon bromochloropropane (1.2 ml, 1.9 g, 12 mmol) was added slowly over 30 min. The mixture was stirred at rt for 24 h, and subsequently heated to reflux for 8 h. The mixture was quenched with MeOH and poured into ice-water (300 mL). The resulting mixture was extracted with Et2O (3 x 50 mL), and the combined organic layers were dried (Na2SO4) and concentrated in vacuo. The product was further purified by column chromatography (silica; CH2Cl2). Yield: 1.1 g (3.3 mmol, 60%).

1H-NMR (300 MHz, CDCl3): δ = 8.55 (d, J = 4.9 Hz, 2H), 8.22

(s, 2H), 7.14 (d, J = 5.0 Hz, 2H), 3.55 (t, J = 6.6 Hz, 4H), 2.73 (t, J = 6.5 Hz, 4H), 1.83 (m, 8H) ppm.

13C-NMR (300 MHz, CDCl3): δ 154.6, 150.3, 147.7, 122.4,

119.7, 43.1, 33.1, 30.5, 26.0 ppm.

N N

ClCl




70

4-(4-azidobutyl)-2-(4-(4-azidobutyl)pyridin-2-yl)pyridine (3.6): NaN3 (650 mg, 10 mmol, 3 equiv.) was added to a solution of the above 4,4’-bis(chlorobutyl)- 2,2’-bipyridine (1.1 g, 3.3 mmol) in DMSO (25 mL). The reaction mixture was stirred at 50ºC for 12 h. It was then poured

into water (100 mL), and the resulting mixture was extracted with Et2O (3 x 40 mL). The combined organic phases were dried (Na2SO4) and concentrated in vacuo to provide the product as a yellow oil. Yield: 1.1 g (3.2 mmol, 95%);

1H-NMR (300

MHz, CDCl3): δ = 8.55 (d, J = 4.9 Hz, 2H), 8.22 (s, 2H), 7.14 (d, J = 5.0, 2H), 3.24 (t, J = 6.4 Hz, 4H), 2.73 (t, J = 6.5 Hz, 4H), 1.83 (m, 8H) ppm.

4-(2-(4-(4-aminobutyl)pyridin-2-yl)pyridin-4-yl)butan-1-amine: The above 4,4’-bis(azidobutyl)-2,2’-bipyridine (1.0 g, 2.8 mmol) was dissolved in EtOH (20 mL), and 10% Pd/C (100 mg) was added. A balloon filled with H2 gas was attached to the top of the flask, and the mixture was vigorously

stirred for 12 h under the H2 atmosphere. The mixture was subsequently filtered through Celite and the filtrate was dried (Na2SO4) and concentrated in vacuo. The product was further purified by acid-base extraction to give a yellow oil that solidified on standing. Yield: 0.95 g (2.8 mmol, 98%);

1H-NMR (300 MHz, CDCl3) δ

= 8.52 (d, J = 4.9 Hz, 2H), 8.20 (s, 2H), 7.10 (d, J = 5.0 Hz, 2H), 2.80 – 2.60 (m, 8H), 2.73 (t, J = 6.5 Hz, 4H), 1.80 – 1.40 (m, 8H) ppm.

13C-NMR (300 MHz,

CDCl3): δ 154.6, 150.9, 147.5, 122.4, 119.8, 43.1, 33.1, 30.5, 26.0 ppm. MS: m/z 298, 254.

4,4’-Bisdodecylureidobutyl bipyridine (3.1): 4,4’-Bis(aminobutyl)-2,2’-bipyridine (0.9 g, 3.0 mmol) was added to CH2Cl2 (50 mL). The mixture was heated to reflux, until all material had dissolved. At this point, a

solution of dodecyl isocyanate (1.3 g, 6.2 mmol, 2 equiv) in CH2Cl2 (10 mL) was added. The mixture was stirred at rt until a white solid started to precipitate.The white precipitate was collected by suction filtration, and further purified by resuspending it, on the filter, in acetone, methanol, and diethylether, respectively, followed by suction. The product was dried at 80 ºC, 3 mmHg. Yield: 0.7 g (0.95 mmol, 30%); white solid; m.p. 225 ºC.(decomp);

1H-NMR (300 MHz, [d

6]DMSO): δ

= 8.52 (d, J = 5.1 Hz, 2H), 8.20 (s, 2H), 7.22 (d, J = 4.8 Hz, 2H), 5.52 (br, 2H), 5.45 (br, 2H), 3.04 (t, J = 6.6 Hz, 4H), 2.99 (t, J = 7.0 Hz, 4H), 2.70 (t, J = 7.7 Hz, 4H), 1.66 (m, 4H),1.45 (m, 4H), 1.35 – 1.20 (m, 40H), 0.86 (t, J = 2.9 Hz, 6H) ppm.

13C-

NMR (300 MHz, [d6]DMSO) = 157.6, 155.1, 151.5, 148.3, 123.2, 120.0, 31.4, 31.1,

29.8, 29.4, 28.8, 28.5, 28.4, 27.3, 26.8, 26.2, 21.7, 13.4 ppm. MS: m/z 721.7. Anal: calcd. for C44H76N6O2 : C 73.75, H 10.62; found: C 73.74, H 10.63.

N N

N3N3

N N

NH2H2N

N N




Chapter 3

71

N N

OTHP

1,3-Bromopropanol-THP ether: 3,4-2H-dihydropyran (12 mL; excess) was added to a solution of bromopropanol (15

g, 0.11 mol) in CH2Cl2 (60 mL) at 0 ºC. The mixture was stirred for 4 h at room temperature. It was then washed with water (100 mL), and the aqueous layer was extracted wih CH2Cl2 (2 x 100 mL). The combined organic layers were washed with brine, dried (MgSO4), and concentrated in vacuo. The brown-yellow oil was purified by column chromatography (silica; hexane/ diethyl ether (10%)) to yield a colorless oil. Yield 18.9 g (0.085 mol, 80%);

1H-NMR (300 MHz, CDCl3): δ = 4.51 (t,

J = 2.9 Hz, 1 H), 3.82 - 3.74 (m, 2 H), 3.47 - 3.39 (m, 4H), 2.05 (quintet, 1J = 6.2

Hz, 3J = 12.5 Hz, 2H), 1.70 - 1.40 ppm (m, 6 H).

32

5-Methyl,5’((tetrahydropyranyloxy) butyl)-2,2’-bipyridine (3.8): A solution of LDA (30 mmol) was freshly prepared from

diisopropylamine (4.6 mL) and n-BuLi (19 mL, 1.6 M in hexanes) in THF (10 mL) at - 80 ºC. A solution of 5,5’-dimethyl-2,2’-bipyridine (5.0 g, 27 mmol) in THF (150 mL) was then added at the same temperature. The reaction mixture, which immediately turned black, was stirred for 2 h at -80°C. It was then allowed to warm to room temperature, whereupon a solution of bromopropanol THP ether (7.8 g, 35 mmol) in THF (40 mL) was added. The mixture was stirred at rt for 100 h, then quenched with MeOH and poured into ice-water (300 mL). The resulting mixture was extracted with Et2O (4 x 100 mL), and the combined organic layers were dried (Na2SO4) and concentrated in vacuo. The product was further purified by column chromatography (silica; elution with 2% MeOH in CH2Cl2) to yield the pure material as an orange brown oil. Yield 7.8 g (24 mmol, 90%);

1H-NMR (300 MHz, CDCl3): δ

= 8.46 (s, 2 H), 8.22 (dd, J = 3.9 Hz, 2H), 7.62 - 7.55 (m, 2 H), 4.55 (s, 1H), 3.82 – 3.72 (m, 2 H), 3.49 - 3.41 (m, 2H), 2.67 (t, J = 6.9 Hz, 2H), 2.35 (s, 3 H), 1.77 - 1.50 ppm (m, 12H).

5,5’-Bis((tetrahydropyranyloxy)butyl)-2,2’-bipyridine (3.9): The same procedure was

followed as for the synthesis of the monosubstituted derivative. A solution of the monosubstituted derivative (7.8 g, 24 mmol) in THF (150 mL) was added to freshly prepared LDA solution (30 mmol) in THF. Subsequently, a solution of the bromopropanol THP ether in THF (9.0 g, 0.04 mol) was added. After work-up, the product was obtained as a brown oil. Yield: 9.6 g (0.021 mol, 88%);

1H-NMR: (300

MHz, CDCl3): δ = 8.46 (s, 2H), 8.23 (d, J = 8.1 Hz, 2H), 7.59 (d, J = 8.3 Hz, 2H), 4.54 (s, 2H), 3.86 -3.70 (m, 4H), 3.53 – 3.33 (m, 4H), 2.67 (t, J = 6.6 Hz, 4 H), 1.81 – 1.50 ppm (m, 20H);

13C-NMR: (300 MHz, CDCl3): δ = 152.5, 147.7, 136.1, 135.3,

118.9, 65.6, 61.4, 60.8, 31.4, 29.1, 27.7, 26.3, 24.0, 18.1 ppm.

N N

OTHPTHPO

B r O T H P




72

N N

OHHO

5,5’-Bis(hydroxybutyl)-2,2’-bipyridine (3.10): The protected bipyridine (5.0 g, 10.7 mmol) was dissolved in EtOH (120 mL) and p-

toluenesulfonic acid (0.5 g) was added. The mixture was refluxed for 24 h, than neutralized with triethylamine. The ethanol was evaporated and the crude product was redissolved in CHCl3. This solution was filtered through neutral aluminum oxide, the filtrate was concentrated to a viscous oil, and this was suspended in toluene. The product separated as a white solid upon sonication. Yield: 1.1 g (3.7 mmol, 35%);

1H-NMR (300 MHz, CDCl3): δ = 8.50 (s,2H), 8.26 (d, J = 8.3 Hz, 2 H),

7.64 (d, J = 8.3 Hz, 2 H), 3.66 (t, J = 6.4 Hz, 4H), 2.69 (t, J = 6.6 Hz, 4H), 1.79, 1.58 ppm (m, 8H).

13C NMR (300 MHz, CDCl3): d = 152.3, 147.4, 136.1, 135.3,

119.3, 60.9, 31.0, 30.6, 25.7 ppm; MS: m/z: 300, 255. HRMS: calcd for C18H24N2O2

300.18376; found 300.18438.

5,5’-Bis(bromobutyl)-2,2’-bipyridine (3.11): Concentrated H2SO4 (0.25 mL, 4 mmol) was

added to a solution of the diol (0.5 g, 1.8 mmol) in 48% aqueous HBr (5 mL). The reaction mixture was refluxed for 12 h, then diluted with ice-water (80 mL), and neutralized with aqueous Na2CO3 solution (20 mL). The yellow mass that was liberated was extracted with CHCl3 (50 mL) and the combined organic phases were dried (Na2SO4) and purified by column chromatography (CH2Cl2/toluene, 2:8). Evaporation of the solvents gave the dibromide as a white solid. Yield: 0.6 g (1.4 mmol, 78%);

1H-NMR (300 MHz, CDCl3): δ = 8.48 (s, 2H), 8.26 (d, J = 8.0 Hz, 2 H),

7.62 (d, J = 8.1 Hz, 2 H), 3.42 (t, J = 7.1 Hz, 4 H), 2.68 (t, J = 6.9 Hz, 4 H), 2.67 (t, J = 6.9 Hz, 2 H), 2.02 - 1.66 ppm (m, 8H);

13C-NMR (300 MHz, CDCl3): δ 147.8,

135.3, 130.5, 127.1, 119.1, 34.6, 30.8, 26.9, 20.0 ppm; MS: m/z: 426, 277; HRMS: calcd for C18H22N2Br2: 424.01489; found 424.01361.

5,5’-Bis(azidobutyl)-2,2’-bipyridine (3.12): NaN3 (400 mg, 6.2 mmol, 4 equiv.) was added to a solution of the aforementioned

5,5’-bis- (bromobutyl)- 2,2’-bipyridine (0.6 g, 1.4 mmol) in DMSO (10 mL). The reaction mixture was stirred at 50°C for 12 h. It was then poured into water (50 mL), and the resulting mixture was extracted with Et2O (30 mL). The combined organic phases were dried (Na2SO4) and concentrated in vacuo to provide the product as a yellow oil. Yield: 0.48 g (1.4 mmol, 100%);

1H-NMR (300 MHz,

CDCl3): δ = 8.49 (s, 2 H), 8.29 (d, J = 8.1 Hz, 2 H), 7.62 (dd, 1J = 8.1 Hz,

2J = 2.2

Hz, 2H), 3.31 (t, J = 6.6 Hz, 4H), 2.71 (t, J = 6.9 Hz, 4 H), 1.76 - 1.66 (m, 8H) ppm; 13

C-NMR (300 MHz, CDCl3): δ 147.8, 135.3, 130.5, 127.1, 119.1, 49.7, 30.8, 26.9, 20.0 ppm.

5,5’-Bis(aminobutyl)-2,2’-bipyridine (3.13): The aforementioned 5,5’-bis(azidobutyl)-2,2’-bipyridine (0.5 g, 1.4

N N

BrBr

N N

N3N3

N N

NH2H2N



Chapter 3

73

mmol) was dissolved in EtOH (20 mL), and 10% Pd/C (50 mg) was added. A balloon filled with H2 gas was attached to the top of the flask, and the mixture was vigorously stirred for 12 h under the H2 atmosphere. The mixture was subsequently filtered through Celite and the filtrate was dried (Na2SO4) and concentrated in vacuo. The product was further purified by acid-base extraction to give a yellow oil that solidified on standing. Yield: 0.41 g (1.4 mmol, 100%);

1H-NMR (300 MHz,

CDCl3 + [d6]-DMSO): δ = 8.19 (s, 2H), 7.96 (d, J = 8.1 Hz, 2H), 7.35 (dd,

1J = 8.1

Hz, 2J = 2.2 Hz, 2 H), 2.45, 2.29 (m, 8H, CH2NH2 + CH2Pyr), 1.39 (t, J = 6.6 Hz,

4H), 1.24 ppm (t, J = 7.0 Hz, 4H); 13

C-NMR (300 MHz, CDCl3): δ 146.6, 134.1, 130.5, 127.1, 117.7, 39.2, 30.5, 29.9, 25.7 ppm.

5,5’-Bis(dodecylureidobutyl)-2,2’- bipyridine (3.2): 5,5’-Bis(aminobutyl)-2,2’-bipyridine (0.3

g, 1.1 mmol) was added to toluene (30 mL). The mixture was heated to reflux, until it became a smooth and slightly transparent suspension. At this point, a solution of dodecyl isocyanate(800 mg, 3.8 mmol, >3 equiv) in toluene (5 mL) was added. Almost immediately,a white solid started to precipitate. The mixture was stirred while slowly (over a period of 1 h) cooling down to room temperature. The white precipitate was collected by suction filtration, and further purified by resuspending it, on the filter, in acetone, methanol, and diethyl ether, respectively, followed by suction. The product was dried at 80 ºC, 3 mmHg. Yield: 0.2 g (0.3 mmol, 30%); off-white solid; m.p. 220ºC (dec.);

1H- NMR (300 MHz, [d] TFA/[d

6] DMSO): δ =

8.54 (s, 2H), 8.24 (d, J = 8.1 Hz, 2H), 8.05 (d, J = 8.1 Hz, 2H), 3.09 (t, J = 6.6 Hz, 4H), 2.99 (t, J = 7.0 Hz, 4H), 2.70 (t, J = 7.0 Hz, 4H), 1.59 (m, 4H), 1.46 (m, 4H), 1.30 - 0.99 (m, 40H), 0.70 ppm (t, J = 6.6 Hz, 6H);

13C-NMR (300 MHz, [d

6]

DMSO/TFA): δ 145.5, 139.6, 123.7, 121.9, 116.1, 31.4, 31.1, 29.8, 29.4, 28.8, 28.5, 28.4, 27.3, 26.8, 26.2, 21.7, 13.4 ppm; IR (KBr): ν˜ = 3350, 2956, 1625, 1575, 1466 cm

-1; MS: m/z = 721.7. Anal: calcd. for C44H76N6O2 : C 73.29, H 10.62; found:

C 72.98, H 10.43.

5’-Methyl-5-nonadecyl-2,2’-bipyridine (3.3): A mixture of diisopropylamine (800 mL, 580 mg, 5.7 mmol) and nBuLi (1.6 M in hexanes; 3.5 mL, 5.6 mmol) in THF (10 mL) was first stirred at -

78°C for 15 min. A solution of 5,5’-dimethyl-2,2’-bipyridine (1.0 g, 5.45 mmol) in THF (50 mL) was then added from a dropping funnel over a period of 30 min. The reaction mixture was stirred for an additional 2 h, while the temperature was slowly raised to 0ºC. A solution of n-octadecyl bromide (2.0 g, 6 mmol, 1.1 equiv.) in THF (5 mL) was then added by means of a syringe. The mixture was stirred for 48 h at room temperature, and then ice was added. The product was extracted with Et2O, and the combined extracts were washed with aqueous NaHCO3 solution and water, and then concentrated. The residue was recrystallized twice from CHCl3. Yield: 1.3 g (3.0 mmol, 55%); white powder; m.p. 74 - 75 ºC;

1H-NMR (300 MHz,

CDCl3): δ = 8.43 (s, 2H), 8.20 (d, J = 8.0 Hz, 1 H), 8.19 (d, J = 8.0 Hz, 1H), 7.55 (d, J = 8.0 Hz, 2 H), 2.59 (t, J = 7.8 Hz, 2H), 2.33 (s, 3H), 1.57 (m, 2H), 1.19 (m, 32H),

N N


N N

C19H39




74

0.82 ppm (t, J = 6.8 Hz, 3H); 13

C-NMR (300 MHz, CDCl3): δ 154.0, 149.6, 149.3, 138.0, 137.5, 136.8, 133.1, 120.5, 120.4, 32.9, 32.0, 31.5, 29.8, 29.4, 29.2, 22.7, 14.1 ppm; MS: m/z: 436, 407, 393, 379, 365, 351, 337, 323, 309, 295, 281, 267, 253, 239, 211,197, 184. Anal: calcd. for C30H48N2 : C 82.51, H 11.08, N 6.41; found: C 82.33, H 11.73, N 6.30.

5,5’-Bis(nonadecyl)-2,2’-bipyridine (3.4): The procedure described above was repeated, but with diisopropylamine (350 mL, 2.4 mmol) and n-

BuLi (1.5 mL, 2.4 mmol) in THF (10 mL), 5’-methyl-5-nonadecyl-2,2’-bipyridine (1 g, 2.35 mmol) in THF (50 mL), and octadecyl bromide (850 mg, 2.55 mmol). Upon addition of ice in the work-up, the product precipitated. It was collected by suction filtration and recrystallized twice from CHCl3 to provide a white powder. Yield: 820 mg (1.2 mmol, 40%); m.p. 99 - 100 ºC;

1H-NMR (300 MHz, CDCl3): δ = 8.46 (s,

2H), 8.23 (d, J = 8.0 Hz, 2H), 7.59 (dd, 1J = 8.1 Hz, 3J = 1.8 Hz, 2H), 2.63 (t, J = 7.8 Hz, 2 H), 1.61 (m, 4H), 1.23 (m, 64H), 0.82 ppm (t, J = 6.8 Hz, 6H);

13C-NMR

(300 MHz, CDCl3): δ 156.3, 149.3, 138.0, 136.8, 120.5, 32.9, 32.0, 31.2, 29.8, 29.5, 29.4, 29.2, 22.8, 14.2 ppm; MS: m/z: 688, 449, 321. Anal: calcd. for C48H84N2: C 83.65, H 12.28, N 4.06; found: C 82.90, H 12.26, N 4.36. STM: Prior to imaging, all compounds to be investigated were dissolved in 1-octanol or 1-phenyloctane, and a drop of this solution was applied to a freshly cleaved surface of highly oriented pyrolytic graphite (HOPG). The STM images were acquired in the variable current mode (constant height) under ambient conditions with the tip immersed in the liquid. In the acquired STM images, white corresponds to the highest and black to the lowest measured tunneling current. STM experiments were performed using a Discoverer scanning tunneling microscope (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 a 2n KOH/ 6n NaCN solution in water. All complexation reactions were performed in situ, unless stated otherwise. Therefore, a drop of the concentrated solution of the metal complex (in 1-octanol or 1-phenyloctane) was added to the sample. For solubility reasons, in some reactions a cosolvent was used (see text). These cosolvents were selected such that they had low boiling points and that the metal ions would be known to complex to the bipyridine in the resulting medium. The experiments were repeated in several sessions using different tips to check for reproducibility and to avoid artefacts. Different settings of the tunneling current and the bias voltage were used, ranging from 0.3 nA to 1.0 nA and -10 mV to -1.5 V, respectively. All STM images are derived from raw data and have not been subjected to any manipulation or image processing. STM images were obtained in the laboratory of prof. De Schryver and dr. De Feyter at the university of Leuven, Belgium.

N N

C19H39 C19H39



Chapter 3

75

3.13 References.

1 Lewis, P.A., Donhauser, Z.J., Mantooth, B.A., Smith, R.K., Bumm, L.A., Kelly, K.F., Weiss,

P.S. Nanotechnology 2001, 12, 231, and references therein.

2 Nyffenegger, R.M., Penner, R.M. Chem. Rev. 1997, 97, 1195.

3 Aqueous-phase organometallic catalysis : concepts and applications, Cornils, B.,

Herrmann, W.A., Wiley-VCH Verlag, New York, 2004.

4 van Esch, J., De Feyter, S., Kellogg, R.M., De Schrijver, F.C., Feringa, B.L. Chem. Eur. J.

1997, 3, 1238 – 1242.

5 van Esch, J., Schoonbeek, F., de Loos, M., Kooijman, H., Spek, A.L., Kellogg, R.M.,

Feringa, B.L. Chem. Eur. J. 1999, 5, 937-950.

6 Schoonbeek, F.S., van Esch, J.H., Wegewijs, B., Rep, D.B.A., de Haas, M.P., Klapwijk,

T.M., Kellogg, R.M., Feringa. B.L. Angew. Chem. Int. Ed. Engl. 1999, 38, 1393-1397.

7 Semenov, A., Spatz, J.P., Moller, M., Lehn, J.-M., Sell, B., Schubert, D., Weidl, C.H.,

Schubert, U.S. Angew. Chem. Int. Ed. 1999, 38, 2547.

8 Ziener, U., Lehn, J.-M., Mourran, A., Moller, M. Chem. Eur. J. 2002, 8, 951.

9 Hofmeier, H., Hoogenboom, R., Wouters, M.E.L., Schubert, U.S. J. Am. Chem. Soc. 2005,

127, 2913-2921.

10 Dmitriev, A., Spillman, H., Lin, N., Barth, J.V., Kern, K. Angew. Chem. Int. Ed. 2003, 42,

2670.

11 Rapenne, G., Dietrich-Buchecker, C., Sauvage, J.P. J. Am. Chem. Soc. 1999, 121, 994-1001.

12 Fujita, M. Chem. Soc. Rev. 1998, 27, 417.

13 Baxter, P.N.W., Lehn, J-M., Baum, G., Fenske, D. Chem. Eur. J. 2000, 6, 4510-4517.

14 Anbalagan, V. J. Coord. Chem. 2003, 56, 161-172.

15 Zaworotko, M.J. Angew. Chem. Int. Ed. Engl. 2000, 39, 3052.

16 Cyr, D.M., Venkataraman, B., Flynn, G.W. Chem. Mater. 1996, 8, 1600-1615.

17 Patai, S. The chemistry of amino, nitroso and nitro coumpounds and their derivatives.

1982, Part of: The chemistry of functional groups, ISSN 0069-3146; supplement F. Wiley, New York.

18 De Feyter, S., Grim, P.C.M., van Esch ,J., Kellogg, R.M., Feringa, B.L., De Schryver, F.C.

J. Phys. Chem. B, 1998,102, 8981-8987.

19 Examination of 28 crystal structures of noncyclic urea compounds deposited in the

Cambridge Crystallographic Database revealed that the average distance between two successive hydrogen-bonded urea groups is 0.46 nm.




76

20

Examining the Gmelin database for example reveals for tetrasubstituted metal- bipyridine complexes: Rh 5 entries; Pd 275 entries; Ag 3 entries.

21 Kamath, S.S., Uma, V., Srivastava, T.S. Inorg. Chim. Acta, 1989, 161, 49 – 56.

22 Giancarlo, L.C., Flynn, G.W. Annu. Rev. Phys. Chem. 1998, 49, 297.

23 Abdel-Mottaleb, M.M.S., De Feyter, S., Sieffert, M., Klapper, M., Mullen, K., De Schryver,

F.C. Langmuir, 2003, 19, 8256-8261.

24 Kamath, S.S., Uma, V., Srivastava, T.S. Inorg. Chim. Acta 1989, 161, 49 – 56.

25 Oh, Y.J., Cho, S.M., Chung, C.H. Electrochem. and Solid State Lett. 2005, 8, C1-C3.

26 (a) Dhar, S.K. Inorg. Chim. Acta 1995, 240, 609 – 614 (b) Carlucci, L., Ciani, G.,

Proserpio, D.M., Rizzato, S. Chem. Commun. 2001, 1198 - 1199 (c) Gao, Y., Wang, Y., Zhu, Y., Shi, Q. Polyhedron, 1991, 10, 1893 – 1896.

27 De Feyter, S., Abdel-Mottaleb, M.M.S., Schuurmans, N., Verkuijl, B.J.V., van Esch, J.H.,

Feringa, B.L., De Schryver, F.C. Chem. Eur. J. 2004, 10, 1124 – 1132.

28 (a) Iwasawa, T., Tokunaga, M., Obora, Y., Tsuji, Y. J. Am. Chem. Soc. 2004, 126, 6554 –

6555 (b) Potenzo, G., Ravasio, N., Aresta, M. J. Organomet. Chem. 1993 , 451, 243 – 248.

29 Palladium – malonate complexes have been described. Wolkowski, J.P., Hartwig, J.F.

Angew. Chem. Int. Ed. 2002, 41, 4289 – 4291.

30 Giancarlo, L.C., Flynn, G.W. Acc. Chem. Res. 2000, 33, 491 – 501.

31 (a) Nayak, S.K., Farrell, G.J., Burkey, T.J. Inorg. Chem. 1994 , 33, 2236 - 2242 .

(b) Calderazzo, F., Falaschi, S.M., Fabio, P. J. Organomet. Chem. 2002, 662, 137 – 143.

32 Bohlmann, F. et al. Chem. Ber. 1960, 93, 1931-1937.



77

Chapter 4 Mixing and phase separation in binary systems. A means to control periodicity

In this chapter two-component self-assembly to affect pattern formation in organic monolayers at the nanoscale is described. The components are symmetrically substituted bisurea derivatives. One of the components incorporates a thiophene functionality. Variations in the relative positions of the urea (H-bonding) groups along the long molecular axis of the second component are shown to influence intermolecular H-bonding, as well as length and shape complementarities of the components. This in turn is crucial for the two-dimensional phase behavior of the mixtures. Insight into these parameters allows the formation of either randomly intermixed systems or phase separation.

4.1 Introduction

The control of the lateral assembly and spatial arrangement of micro- and nano-objects at interfaces is a prerequisite when it comes to potential applications in the field of nanoscience and technology. Self-assembly methods provide one approach to make defined structures with dimensions on the nanometer scale. Self-assembly is a natural phenomenon that can be observed in many biological, chemical, and physical processes.1 Self-assembled monolayers (SAMs) are ordered molecular assemblies formed by the adsorption of an active surfactant on a solid surface,

2 as described in the introduction. Whereas the previous chapter

addressed incorporation of functionality into physisorbed monolayers, this chapter will focus on the spatial disposition of such functionality.

An a priori attractive route to achieve spatial arrangement is the formation of distinctive domains, exposing one functionality in specific areas and another (or none) in others. Several research groups have taken advantage of the



Mixing and phase separation

78

spontaneous formation of chemisorbed alkane thiolate monolayers and the formation of nanometer-sized domains by the coadsorption of two or more adsorbates.

3,4 If more than one adsorbate is involved in the self-assembly process,

one must consider the interactions between the different components. In studies involving mixtures of molecules differing only in the nature of the end groups, it was shown that the difference in their polarity drives the extent of phase separation.

5 In addition to phase separation induced by differences in the alkyl

chain length,6 it was demonstrated that phase separation can be driven by intermolecular interactions buried within the film.7

Less attention, however, has been given to coadsorption in the self-assembly of physisorbed layers at surfaces. These physisorbed adlayers are suitable model systems to investigate the interplay between molecular structure and the formation of ordered assemblies in two dimensions and can be studied in great detail with STM.

8 It still remains a challenge to control the ordering in multi component

mixtures at the supramolecular level in these layers. In principle three situations can arise when mixing two components: one of the components preferentially adsorbs, both components adsorb but in separate domains (phase separation), or both components adsorb in a supramolecular aggregate of some form (intermixing). The latter situation is the most interesting. Most binary mixtures investigated so far, however, show phase separation on the nanometer scale9 or the formation of randomly mixed monolayers.10 Highly ordered bimolecular two-dimensional adlayers are only formed in a few cases.

11 In this chapter the results

of a series of studies will be reported, in which the interplay between molecular structure, supramolecular interactions, and occurrence of phase separation or mixing in two-dimensional monolayers was addressed. The results clearly show that it is possible to direct pattern formation in monolayers via intermolecular interactions, which can potentially be exploited in future studies on the patterning of two-dimensional monolayers at the supramolecular level.

4.2 Intermixing and periodicity

Bisurea derivatives can form physisorbed monolayers with a lamellar morphology (chapter 3). These aggregates are largely characterized by three parameters:

• The width of the lamellae.

• The angle the molecules make with the main axis of the lamella.

• The molecule-to-molecule distance within a lamella.



Chapter 4

79

It has been demonstrated that the lamellar width can be modified by altering the length of the molecules (tailoring the chemical composition). This provides an element of control in the construction of the monolayers. The lamellar morphology can be modified by interdigitation. The molecule-to-molecule distance within a lamella is governed by the type of interaction between the molecules. Also this aspect can be programmed at the molecular level.

Introduction of H-bonding interactions can be used to tune the molecule-to-molecule distance as compared to the situation where mere van der Waals interactions are operational.12 However, the window of variation is quite narrow. It is therefore a major challenge to develop an approach that will ensure control over all the relevant distances between functional entities in a 2- dimensional space (fig. 4.1). Ideally, functionalized molecules will arrange themselves, preprogrammed in one unique pattern. Controlling one dimension (y direction in fig. 4.1) is easily done for this type of structure, because the periodicity in 1-D is governed by the lamellae, whose width in turn corresponds directly to the length (and angle) of the molecules, as measured along their long axes. The functionality-to-functionality distance within a lamella can only be influenced to a minor extent. Furthermore a

Figure 4.2 Diluting functionality in the monolayers. Indentations indicate urea groups.

a) b)

x

y

x1

x2

y1

y2

y3

Figure 4.1 Control over the position of various functionalities. Ovals indicate arbitrary functionality. x1 and x2 are (variable) distances along the lamellar axis. y1 , y2 and y3 are (variable) distances along the molecular axes.




80

new level of complexity is met in fig. 4.1. There are several distinct molecular structures involved in this hypothetical monolayer. The type of (supramolecular) structure envisaged in fig. 4.1 can for instance be obtained if several components can be made to effectively intermix within the lamellae.

The degree to which the components intermix is dependent on their complementarity with respect to size, shape and position of interacting groups. Fig. 4.2 provides a representation of the idea. The distance between the functionalities in 4.2 b) is considerably larger than in 4.2 a). Because only a part of the molecules is functionalized, functionality is diluted. The second component can incorporate a simple alkyl spacer between the urea groups. The basic geometrical features of the components are the same in that case. The role of the hydrogen-bonding urea groups is expected to be prevalent as it provides the dominant interaction. If the position of the urea groups in the two components is matched, good recognition can be expected. But to which extent is a mismatch tolerated? And is H-bonding the only relevant parameter? Systematic variation of the alkyl spacer should give an answer to these questions. It is expected that mixing a functional bisurea derivative with a range of alkyl bisureas, will show either phase separated – or intermixed monolayers, and possibly intermediate forms.

4.3 About the components

What functionality would one incorporate into the bisurea derivatives? It has been observed that simple alkyl and oligothiophene bisurea compounds can form mixed bimolecular adlayers, a feature which was used to study the electronic properties of the oligothiophene moieties and its supramolecular arrays by means of scanning tunneling spectroscopy.13 The different contrast from the thiophene and alkyl moieties greatly facilitates the observation of phase-separated or mixed adlayers. The bisthiophene unit plays the role of a marker, because it will show up brightly in the STM images, like the bipyridine (described in chapter 3), due to its aromatic character. The patterns that are formed in the monolayers can therefore be studied in great detail.

The other components are symmetrically alkylated bisurea derivatives (R1-urea-spacer-urea-R2; R1, R2 = alkyl). The alkylated bisureas can be varied with respect to: i) the position of the hydrogen-bond forming urea groups along the molecule, and ii) the length of the alkyl chains.

Following this approach, the effect of hydrogen bonding, molecule length, odd/even effects, and shape complementarity can be shown. The molecules that were used in this study are outlined in table 4.1. Note that compounds featuring a spacer with an odd number of methylene groups have a slightly different shape (slightly bow-shaped, as determined by molecular modeling) than derivatives with an even number of methylene groups. The two represent two classes, which are listed separately in table 4.1. The length of the spacing moiety in the bisthiophene



Chapter 4

81

derivative corresponds to a total of 14 methylene groups. The structure of the bisthiophene derivative is shown on top of table 4.2. It is designated T2, which is the name to which it will be referred in the remainder of this chapter. The bisureas starring in the role of the second component are shown on the bottom. Their nomenclature is based on the number of C-atoms in the spacer, and the tails respectively, and indicated in general as Cx-y. Thus C6-12 has an n-hexyl spacer in between the two urea groups, and has two n-dodecyl tails.

Cx-y n m

C9-12 7 11

C15-12 13 11

C9-15 7 14

C6-12 3 11

C8-15 5 14

C12-12 9 11

C14-12 11 11

C14-6 11 5

C16-12 13 11

Table 4.1 Compounds used in the mixing experiments. x = n+2 or n+3; y = m+1.

HN

HN

HN

HN

O O

mnm

NH

NH

HN

HN

O

O

mnm

H25C12NH

NH

O

S

SHN

HN

C12H25

O

T2




82

4.4 Synthesis of the components

The synthesis of bisurea bisthiophene T2, has been described previously.14

Synthesis of C6-12, C9-12 and C14-12 has been reported elsewhere.

15 Synthesis

of the other molecules is described in this paragraph. All bisureas with C12 tails were obtained from the coupling of the appropriate diamine precursor with dodecyl isocyanate. C14-6 was obtained by the reaction of n-hexyl isocyanate and diamino tetradecane in 75% yield. The diamines were synthesized in two steps from dibromide - or dichloride precursors, via azidation (90 % yield) and catalytic hydrogenation (90 % yield, scheme 4.1). Dibromo tetradecane is commercially available. Synthesis of C16-12 started with reduction of tetradecane dicarboxylic acid (thapsic acid) to the corresponding C16-diol (46 % yield). The diol was converted to the dichloride with SOCl2 (60%). Azidation, hydrogenation and formation of the bisurea proceeded as before (scheme 4.1). For the synthesis of C15-12 pentadecalactone was used as a starting material. This macrocycle was immediately reduced to the diol as before (45 % yield). Subsequent reaction steps proceeded as before. Synthesis of C8-15 and C9-15 requireed reaction of diamino octane and diamino nonane, respectively, with isocyanato pentadecane (85% yield and 90 % yield). The latter compound was synthesized from palmitoyl chloride by means of the Curtius rearrangement, a reaction that went smoothly in 99% yield. The reactions are summarized in scheme 4.1. All compounds were characterized by 1H-NMR and 13C-NMR. The final products were additionally characterized by IR spectroscopy and elemental analysis.

4.5 STM images of the individual components

The individual components were investigated first. As can be seen in fig. 4.3, they form characteristic monolayers. In the single-component monolayers the dominating interaction is clearly the hydrogen-bonding between the urea groups.

15

In every lamella molecules are bound by eight hydrogen bonds to the neighboring molecules. This determines the intermolecular distance of 0.46 nm, characteristic for urea groups.

16 As a result of this small intermolecular distance, the T2

molecules have their thiophene rings tilted with respect to the graphite substrate17

to conform to the situation (otherwise they would impose too much spatial constraint to the aggregate). This orientation allows the possibility of π − π interactions between adjacent molecules in a stack. The thiophene rings can easily be recognized as the brightest spots in the images; this is caused by the enhanced tunneling current associated with them. The alkyl chains appear less bright. In CX-

Y- type molecules (table 4.1) the contrast associated with the urea groups is often



Chapter 4

83

BrC14H28Br

1. NaN3

DMSO95%

2. H2 ; Pd-CMeOH90%

H2NC14H28NH2

C6H13NCO

C12H25NCO

toluene

toluene

75%

65%

C14-6

C14-12

HO2CC14H28CO2H HOC16H32OH ClC16H32Cl

LAHTHF

46 %

(4.3)

SOCl2CH2Cl2

90 %

C15COCl

NaN3p-xylene

C15NCO99 %

O

O

as above

85 % over 2 steps

C12H25NCO

toluene

78 %

C16-12

C15-12

as above

25 % over 5 steps

H2NC8H16NH2

H2NC9H18NH2

85 %

toluene

toluene

C8-15

C9-15

90%

(4.7)

Scheme 4.1 Syntheses of the compounds C14-6, C14-12, C16-12, C15-12, C8-15 and C9-15.

4.1 4.2 4.4 4.5 4.6

4.8

4.9




84

quite different from the alkyl chains so that they can be easily located (fig. 4.3).In monolayers of T2 molecules, this is often more difficult because of the adjacent bright bisthiophene groups. The number of carbon atoms of the alkyl group linking both urea groups (called the spacer), being odd or even, determines the shape of the molecule. If the number of carbon atoms is even the molecule adopts an extended zigzag shape, with the urea groups pointing in opposite directions. If the alkyl spacer contains an odd number of carbon atoms, the molecule is bowl-shaped, and the urea groups point in the same direction. This is illustrated in figs. 4.3 a) and b).

Figure 4.3 Monolayers formed by (A) C12-12 (B) C9-12 and (C) T2. The superimposed cartoon represents the alkyl chains (lines) and the urea groups (arrow points). (d) interaction between the urea groups. The scale bar measures 2 nm; molecular models are based on semi-empirical calculations.



Chapter 4

85

4.5 The extremes in mixing behavior

The components can be mixed in the STM cell by applying a drop of a solution containing equimolar amounts of the adsorbents. The images obtained from mixing C6-12 and T2 are indicative of the situation where phase separation occurs. Because C6-12 is a smaller molecule than T2 the interactions between them are not ideal. This is represented in the cartoon in fig. 4.3 a). Whereas a row of T2 (or C6-12) molecules can engage in double hydrogen bonding, the combination can only form one hydrogen-bonding array.

S S

H N N H C 1 2 H 2 5

N H H N H 2 5 C 1 2 O

O N H H N H 2 5 C 1 2

O H N N H

C 1 2 H 2 5 O

S S

H N N H C 1 2 H 2 5

N H H N H 2 5 C 1 2 O

O N H H N H 2 5 C 1 2

O H N N H

C 1 2 H 2 5 O

S S

H N N H C 1 2 H 2 5 N H H N H 2 5 C 1 2

O

O H N N H

C 1 2 H 2 5 N H H N H 2 5 C 1 2 O

O

S S

H N N H C 1 2 H 2 5 N H H N H 2 5 C 1 2

O

O H N N H

C 1 2 H 2 5 N H H N H 2 5 C 1 2 O

O

a) b)

Figure 4.4 Images of T2/C6-12 (left) and T2/C14-12 (right) mixtures; the cartoons above the images explain the preference for the observed phase separation in theT2/C6-12 mixture: intermixing would cause quantitative loss of H-bonding interactions. T2 molecules are highlighted with solid white lines; the alkyl bisureas with dotted lines. The scale bars measure 4 nm.




86

Furthermore, by matching the positions of these molecules in order to do so, a significant amount of empty space is left. This may be filled up with molecules from a neighboring lamella, but this complicates the formation of a regular lattice. Lateral diffusion is costly process, as the carbon atoms in the molecules have to transverse atoms in the top-layer of the graphite. As a consequence the formation of mixed lamellae would seem to be disfavored. Indeed do the STM images show that monolayers composed of T2 and C6-12 phase separate (fig.4.4a). In fig. 4.4 the lamellae composed of T2 molecules are identified by the long white stripe; this is the location of the thiophenes. The darker regions correspond to the lamellae composed of C6-12. The urea groups can be discerned in this region, they show up as small protrusions from the monolayer. In fig. 4.4a) the arrangement is further clarified as such: a T2 molecule is highlighted with a solid line, a C6-12 is highlighted with a dashed line. At the boundary between lamellae defects can be seen (white arrow in fig. 4.4a). These are caused by the non-complementarity of the molecular lengths. The gaps are filled up with molecules, which are perpendicularly oriented with respect to the lamellae.

The situation is quite different in the case where T2 is mixed with C14-12. These two molecules are complementary with respect to two parameters (cartoon fig. 4.4 b): their molecular lengths are matched and the urea groups are in registry (have the same relative locations in the molecules). Intermixing of these components within a lamella allows the formation of two hydrogen-bonding arrays along the lamella. The STM image (fig. 4.4b) shows this arrangement, corroborating the ideas. The situation can be described as the T2 molecules being dissolved into a matrix of C14-12 molecules. The bisthiophene moieties seem to be protruding from the monolayer, due to their bright contrast (vide supra), indicating the position of the T2 molecules. In fig. 4.4b) the arrangement is further clarified as such: a T2 molecule is highlighted with a solid line, a C14-12 is highlighted with a dashed line. The intermolecular distances were found to be 0.46 nm in both experiments, which underscores the prominence of the H-bond interactions in the formation of these aggregates.

At first glance one can discern blocks of T2’s in the image (fig. 4.4b). Some of these consist of only one molecule (a monomer), some consist of more molecules (two, three). Blocks with a size exceeding five were found to be rare (vide infra). The phase separation in the case of T2/C6-12 can in this context be described as the formation of large blocks (multimers; measuring more than 5 consecutive identical molecules).

In these experiments the components were mixed in a 1:1 molar ratio. This figure is approximately represented in the composition in the monolayers. The T2/C14-12 deposition ratio and average lamella length are 1.7 and 19.2 (molecules) respectively. This means that on average there are 1.7 T2 molecules in an array (aggregate) for every C14-12 molecule. The over-presence of T2’s is possibly caused by an additional interaction between the thiophenes and the graphite, not present in the C14-12 coadsorbents. Also an interaction between the thiophenes in neighboring molecules (π−π stacking) is possible. Deviation from unity is not



Chapter 4

87

dramatic however, indicating reasonably similar nucleation behavior of the components.

The distribution of block sizes found in T2/C14-12 mixtures was plotted in a histogram (fig. 4.5). The observed distribution can be compared with ideal statistic (stochastic) mixing. For statistic mixing exactly similar interactions between all components are supposed to be operational. The statistic simulation used experimental findings as input. An aggregate was taken to be composed of 19 molecules, corresponding to the average length of the lamellae, and the over-presence of T2’s (58%) was taken into account. As can be seen in fig. 4.5, the correspondence is striking. Only the number of multimers is underestimated in the simulation, probably due to interactions between the thiophene groups (vide supra).

From the point of view of controlling the interfunctionality distances this result clearly indicates the limitations of the mixing approach. Although intermixed monolayers can be formed, functionality will be statistically distributed over the plane. An alternating pattern is not seen. Also the co-deposition ratio is not easy to forecast or influence. Other approaches will have to be employed to gain higher degrees of control over the periodicities in 2D physisorbed systems. These ideas are the topic of the next two chapters. Nevertheless this experiment constitutes a useful first approach towards non-trivial packings in physisorbed monolayers.

Figure 4.5 Histograms of observed and simulated blocksizes in binary monolayers of T2 and C14-12.

0

5

10

15

20

25

30

35

40

45

50

1 2 3 4 5 5+

n-mer

% n

-mer

datasimulation

0

5

10

15

20

25

30

35

40

45

50

1 2 3 4 5 5+

n-mer

% n

-mer

datasimulation




88

4.6 Molecular modeling

To increase the knowledge about the mixing process, the interactions between C6-12, C14-12 and T2 were modeled on HOPG. Simplified models were generated in Materials Studio.18 By optimizing the energies of two molecules, far enough apart so that they do not interact, and subsequently bringing the molecules in close proximity, i.e. to the equilibrium distance found in the experiments (about 0.46 nm), the favorable enthalpic contribution due to the electrostatic interaction can be estimated.

The process is outlined in fig. 4.6 for one C14-12 and one T2 molecule. The same approach has been taken for all possible combinations. C6-12/C6-12; C6-12/T2; C14-12/C14-12; C14-12/T2 and T2/T2. The results are summarized in table 4.2.

The computational results indeed correspond to the experimental findings, as might be expected. The small energy differences between the T2 - C14-12 pair, and the C14-12 – C14-12 pair indicates that it is appropriate to consider the mixing process as a statistical process. The slightly better interaction for the T2-T2 pair explains why T2 is more prominent in equimolarly deposited mixed monolayers. The excess of multimers, as observed in the experiments can also be rationalized with this result: as ∆ET2-T2 > ∆ET2-C14-12 the preference for mixing is somewhat diminished, and an excess of larger blocks is observed. The preferential interaction between T2 molecules probably originates from π−π stacking.

Figure 4.6 Modeling the interactions between components in binary systems. Bringing the components in close proximity estimates the favorable contribution due to all electrostatic interactions (van der Waals as well as H-bonding).



Chapter 4

89

4.7 Tolerance towards mismatch

The question arises how much the position of the urea groups can be out-of line in order to still have good intermolecular H-bonding, and intermixing. To address this question mixing experiments of T2 with C15-12 and C16-12 were performed. The total molecular lengths of these co-adsorbents are not very different from T2.

The relative position of the urea groups would seem to be a problem however, as efficient H-bonding is not possible. In C15-12 odd-even effects play a role, as the directionality of one of the urea groups is opposite, which would prevent efficient recognition altogether. Furthermore the shape of this molecule is slightly different, it is slightly bent (like a bow, cf. fig. 4.3b), as compared to C14-12.

Therefore it had been anticipated that in aggregates of these components only one H-bonding array can be effectively formed along the lamellae (comparable to the situation for mixing C6-12 and T2), upon applying a mixture of these species to the basal plane of HOPG. This in turn would provide a considerable driving force for phase separation. STM images, however, indicate that T2 and C15-12 intermix rather well, as can be seen in figs. 4.7a) and b). The block sizes of T2 are on average slightly larger then those found in the mixing with C14-12.

Interestingly the co-deposition ratio of T2/C15-12 is smaller then unity (0.8) and the average length of the lamellae is shorter (15 molecules). These results indicate that something special is happening (the co-deposition ratio’s for T2/C14-12, as well as T2/C16-12 are larger then unity). A possible explanation is that T2

Pair of molecules Energy gain

C6-12 + C6-12 -37.6

C6-12 + T2 -30.9

C14-12 + C14-12 -43.1

C14-12 + T2 -42.8

T2 + T2 -48.5

Table 4.2 Energy gain ∆E (kcal/mol) upon approach to observed equilbrium distance in the aggregates.




90

undergoes a conformational rearrangement, making a rotation of 90º about the bond joining the thiophenes, under the influence of the C15-12 coadsorbents.

This would bring the urea groups back in registry, explaining why two H-bonding arrays can be formed, and the components can intermix in the lamellae. Because a parallel orientation of the thiophenes is not the preferred one this can also explain why the codeposition rate is lower. The T2 molecules are less prone to assemble onto the graphite in this conformation. This would also be in accordance with the shorter average length of the lamellae. There is however no direct evidence for this interpretation from the STM images, as the bisthiophene moieties are not sub-molecularly resolved.

Upon extension of the spacer by one more C-atom, one ends up with C16-12. Odd-even effects do not apply here. Formation of two H-bonding arrays along the

a) b)

c) d)

Figure 4.7 STM images of T2 mixed with C15-12 (a and b) and C16-12 (c and d). The scale bar measures 4 nm.



Chapter 4

91

lamellae is severely hindered. This is reflected in the STM images (fig. 4.7 c and d). Much more large blocks of T2 or C16-12 can be seen. However, phase separation is not as dramatic as was seen for the mixing of T2 with C6-12. This indicates cooperation of length complementarity and hydrogen bonding in these systems. There is a trend in extending the spacers of the co-adsorbents, every methylene group added to the spacer induces more phase separation. From the experiments described up till now emerges a picture of hydrogen bonding as the principle driving force, as it can even modify the conformation of one of the components to drive mixing. An intermixed monolayer must be able however to reach good spatial coverage. Excluded space is a second(ary) parameter influencing the mixing. It should be appreciated that the dynamic nature of these physisorbed systems generates the thermodynamically favored structure, which implies there is an entropic component, which favors mixing. Phase separation is brought about by the enthalpic component.

4.8 Mixing with the other alkyl bisureas

Codeposition of C9-12 or C12-12 and T2 yields aggregates, which are similar to the ones observed for the mixing of C16-12 and T2. That means mainly blocks of T2 dissolved in the matrix of the alkylbisureas.

19 Comparison of the block sizes (1

to 5+ T2 molecules in a row) for the respective mixtures (T2 with C6-12, C9-12,

C12-12, C14-12, C15-12 and C16-12) shows a clear trend (fig. 4.8). C14-12 is complementary to T2 in total size and length of the spacer. This molecule allows for effective intermixing to form multi-component lamellae. The more the length of

Figure 4.8 Observed block sizes of T2 for the respective mixtures of T2 with alkylbisureas.

0

20

40

60

80

100

1 2 3 4 5 5+

n -mer

% n

-mer




92

the spacer deviates from the optimum length, the more the mixture tends to phase separate. Thus, for the mixture of C14-12 and T2, monomers are observed more frequently then dimers, which are more observed then trimers, etc. This trend is still visible for C12-12 and C15-12, though less pronounced. C16-12 shows the reversed trend. Both C12-12 and C16-12 are clearly prone to the formation of multimers. C6-12 and C9-12 almost exclusively form multimers. This underscores the importance of length complementarity of the spacers. However, in these experiments, extending the spacer, implies extending the total length of the molecules.

4.9 Molecular length or H-bonding complementarity ?

To further investigate the role of H-bonding and length complementarity, T2 has been mixed with C14-6, C8-15, and C9-15. These co-adsobents more specifically address the role of these parameters. C14-6 is complementary to T2 because the location (and directionality) of the urea groups is the same, but the total length of these molecules differs. On the other hand, in C8-15 and C9-15 the total molecular lengths of the components is the same, but the position of the urea groups does not match.

Applying a mixture of T2 with C14-6 upon the basal plane of HOPG gave monolayers where intermixing in the lamellae occurs, as shown in fig. 4.9. The C14-6 molecules can easily be identified by the short terminal hexyl chains and the long spacer connecting both urea groups. In fig. 4.9 one molecule is highlighted by the dotted line. T2 molecules are co-deposited (highlighted by the solid line). In fig. 4.9 isolated T2 molecules are

-

U U

U U

Figure 4.9 STM image of monolayer composed of T2 and C14-6. One T2highlighted with solid line, one C14-6with dashed line. Scale bar measures 4 nm. The arrow indicates a new domain, where the main axes of the lamellae have a different orientation. The cartoon on top shows how the components are thought to interact. U = urea group.



Chapter 4

93

visible. The bright structures, corresponding to the bisthiophene unit, are centered and located between the urea groups of adjacent C14-6 molecules. This indicates that the isolated T2 molecules are anchored by hydrogen bonding. Two H-bonding arrays are formed between T2 and the neighboring C14-6 molecules. As a result, the terminal methyl groups of the dodecyl chains of the T2 molecules cannot be in line with the terminal methyl groups of the hexyl chains of the C14-6 molecules. Evidently, in terms of packing efficiency, this is not the best solution; however, it demonstrates the effect of hydrogen bonding on the two-dimensional ordering. Based upon the large difference in size, T2 and C14-6 are expected to exclusively show phase separation. The observation of T2’s trapped in the matrix of C14-6 corroborates the dominance of hydrogen bonding in the mixing process.

In C8-15 or C9-15 the urea groups are designed to be in a position such that formation of two H-bonding arrays when intermixing with T2 is not possible. The experiment is done with one derivative featuring an even number of methylene groups in the spacer (C8-15) and one with an odd number (C9-15) on purpose. Hereby a possible odd-even effect can be probed. In principle, three different situations are possible for the mixtures:

- -

Figure 4.10 STM images of mixed monolayers. T2 + C8-15 (left) and T2 + C9-15 (right). The scale bars measure 4 nm. T2 and alkylbisurea are highlighted with solid and dotted lines respectively. The white arrow refers to situation i. The striate arrow refers to situation ii. (see the text).




94

i. Phase separation: the molecules stack in rows with alternating blocks of T2 and CX-15. Both types of molecules are perfectly aligned, lamellae are straight and hydrogen bonding at the boundary between T2 and CX-15 blocks is not possible.

ii. Phase separation: the molecules align in rows and phase separate, but in such a way that one hydrogen bond can be formed at the boundary between T2 and CX-15 lamellae.

iii. Mixing with and without formation of hydrogen bonds.

Fig. 4.10 shows the STM images of the monolayers formed by the respective mixtures. It is clear that phase separation is observed exclusively, regardless of the C8 or C9 spacer length. The molecules stack in rows with alternating blocks of T2 and CX-15. Close inspection reveals that both situation i. and situation ii. occur. On the left side of fig. 4.10 the lamella in the center (white arrow) shows a mismatch of the urea groups of the T2 and C8-15 blocks. On the right side of fig. 4.10 (striate arrow), the urea groups of the T2 and C9-15 blocks appear to be in line. The occurrence of both packings marks a trade-off between the enthalpically driven contribution, and the excluded space (resulting in monolayer defects) contribution in the process. For situation i, straight lamellae are formed, limiting the number of monolayer defects formed, at the expense of hydrogen-bond formation. For situation ii, one hydrogen bond can be formed at the domain boundary between a T2 and CX-15 block, inducing a small lateral offset, which could potentially lead to an enhanced formation of monolayer defects. Again the importance of hydrogen bonding is stressed. Without H-bond formation no intermixing is possible.

4.10 Conclusions

This study was dedicated to the investigation of the relationship between molecular parameters like molecular size and shape, and the influence of these parameters on the two-dimensional phase behavior. It was found that in multi-component systems phase-behavior is strongly determined by the degree of complementarity of the components. Overall molecular size, as well as the location of functional groups have to be complementary if proper intermixing is to occur. Thus, phase separation is promoted by an increase in the difference in molecular size, while random intermixing is optimal when the size of both components is identical. In addition, the presence and the location of the hydrogen bonding units in the molecules plays an important role. Hydrogen bonding can counteract the effect of the difference in molecule length on the two-dimensional phase behavior. Moreover, in the combinations that were examined hydrogen bonding interactions are clearly the dominant contribution to the palette of interactions governing the



Chapter 4

95

phase behavior. Odd/even effects that might arise from reversal of the orientation of functional groups were found to be not very prominent.

Although it is not possible to control the size of the aggregates, it is possible to influence to a large extent the phase behavior leading to optimal intermixing or phase separation by paying attention to the possible intermolecular interactions.


Materials and methods 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-

300 spectrometer (at 300 MHz) in TFA -CD3OD (10% v/v), chemical shifts are given in ppm relative to methanol (δ = 3.35 ppm).

13C-NMR spectra were recorded

on a Varian VXR300 spectrometer (at 75.48 MHz) in TFA-D2O (10% v/v), chemical shifts are given relative to TFA (δ =154.3 ppm). The splitting patterns in the 1H-NMRspectra are designated as follows: s (singlet), d (doublet), t (triplet), m (multiplet), br (broad). Melting points were measured on Stuart scientific SMP1 apparatus. Infrared spectra were recorded on a Nexus FTIR spectrometer. Elemental analyses were carried out in the Microanalytical department of the Stratingh Institute, University of Groningen (The Netherlands).

1-Hexyl-3-[14-(3-hexylureido) tetradecyl] urea (C14-6; 4.1): n-Hexyl isocyanate

(250 mg, 2 mmol) was slowly added to a stirred solution of 1,14-diaminotetradecane (180 mg, 0.8 mmol) in hot toluene (20 mL). An offwhite suspension formed immediately. After stirring for 2 h, the mixture was poured into diethyl ether, and the product precipitated as a white solid. After sonication for 1 h, the precipitate was collected by filtration and washed with diethyl ether. The product could be purified by repeated precipitation from p-xylene. Yield: 0.3 g (0.6 mmol, 75%); m.p. 163 -165 ºC (dec.);

1H-NMR (TFA-CD3OD): δ 2.70 (t,

J=6.9 Hz, 8H), 1.01 (m, 8H), 0.67(m, 32H), 0.23 ppm (s, 6H); 13C NMR (TFA-D2O): δ 152.5, 35.6, 24.4, 22.8, 22.7, 22.6, 22.3, 21.9, 19.7, 19.4, 15.5, 6.0 ppm; IR(KBr): 3336, 1614, 1576 cm-1; C28H58N4O2 (482.79): m/z: 482. Anal: calcd: C 69.66, H 12.11, N 11.60; found: C 69.53, H 12.02, N 11.52.

1-Dodecyl-3-[14-(3-dodecylureido) tetradecyl]urea (C14-12; 4.2): This

compound was synthesized as described for C14-6, starting from 1,14-diaminotetradecane (200 mg, 0.8 mmol) and dodecylisocyanate (400 mg, 1.9

HN

HN

O

NH

NH

O

HN

HN

O

NH

NH

O




96

mmol). Yield: 0.34 g (65%); m.p. 159 -160 ºC (dec.); 1H-NMR (300 MHz, TFA -CD3OD): δ 2.71 (t, J=6.6 Hz, 8H), 1.02 (m, 8H), 0.67 (m, 56H), 0.23 (s, 6H) ppm; 13C- NMR (TFA -D2O,): δ 36.1, 25.9, 23.5, 23.3, 22.9, 22.6, 20.4, 16.5, 7.0 ppm; IR (KBr): 3336, 1611, 1574 cm-1; Anal: calcd for C40H82N4O2 (651.12): C 73.79, H 12.69, N 8.60; found: C 73.60, H 12.64, N 8.37.

Hexadecane-1,16-diol (4.3): To a suspension of LAH (420 mg) in THF (30 ml) was added at 0°C a solution of thapsic acid

(1.55 g, 5.4 mmol) inTHF (75 ml). The mixture was stirred for 2h at rt. The mixture was poured into a solution of Na,K-tartrate (aq.), extracted (Et2O), washed (brine, water), dried and concentrated in vacuo. Yield: 0.65 g (2.5 mmol, 46%). Waxy solid.

1H-NMR (CDCl3): δ 3.58 (t, J=6.6 Hz, 4H), 1.45 – 1.50 (m, 8H), 1.25 (m,

20H).

1,16-dichlorohexadecane (4.4): Compound 3 (0.65 g, 2.5 mmol) was suspended in CH2Cl2 (25 ml). SOCl2 (600 mg, 5mmol) was

added dropwise at 0°C. The mixture was stirred and heated at reflux for 2h. A little ice, Na2CO3 (aq) and Et2O were added. The organic layer was separated, dried and concentrated in vacuo. Yield: 0.58 g (2.3 mmol, 90%). 1H-NMR (CDCl3): δ 3.47 (t, J=7.0 Hz, 4H), 1.60 (m, 4H), 1.25 (m, 24H).

1-Dodecyl-3-[16-(3-dodecylureido) hexadecyl]urea (C16-12; 4.5): This

was synthesized as described for C14-6, starting from 1,16-diaminohexadecane (200 mg, 0.8 mmol) and dodecylisocyanate (400 mg, 1.9 mmol). Yield: 430 mg (78%); m.p.164 - 165 ºC (dec.); 1H NMR (300 MHz, TFA -CD3OD): δ2.71 (t, J=6.6 Hz, 8H), 1.02 (m, 8 H), 0.67 (m, 60H), 0.23 ppm (s, 6H); 13C NMR (TFA - D2O): δ (152.7), 35.5, 25.2, 22.9, 22.8, 22.4, 22.3, 22.1, 19.9, 16.0, 6.3 ppm; IR (KBr ): 3331, 1612, 1569 cm

-1. Anal: calcd for C41H84N4O2 (679.17): C 74.28, H 12.76, N

8.25; found: C 74.09, H 12.81, N 8.13.

1-Dodecyl-3-[15-(3-dodecylureido) pentadecyl]urea (C15-12; 4.6): This was

synthesized as described for C14-6, starting from 1,15-diaminopentadecane (200 mg, 0.8 mmol) and dodecylisocyanate (400 mg, 1.9 mmol). Yield: 0.38 g (71%); m.p. 162-164 °C (decomp);

1H NMR (300 MHz, TFA-CD3OD): δ 2.71 (t, J=6.6 Hz,

8 H), 1.02 (m, 8H), 0.67 (m, 58H), 0.22 ppm (s, 6H); 13C NMR (75.48 MHz, TFA -D2O): δ152.7, 35.7, 23.0, 22.9, 22.8, 22.4, 22.1, 19.9, 16.0, 6.4 ppm; IR(KBr ): 3336, 1611, 1574 cm-1; C41H84N4O2 (665.14): m/z: 665.8. Anal: calcd C 74.04, H 12.73, N 8.42; found: C 73.88, H 12.69, N 8.45.

1-Isocyanatopentadecane (4.7): Palmitoyl chloride (5 g, 0.018 mol)

was dissolved in p-xylene (40 mL). Sodium azide (1.5 g, 0.023 mol) was added.

HN

HN

O

HN

HN

O

NCO

NH

HN

HN

NH

O

O

HOC16H32OH

ClC16H32Cl



Chapter 4

97

The reaction mixture was heated at reflux for 2 h under a continuous flow of nitrogen, after which conversion to the isocyanate was complete. The hot reaction mixture was filtered to remove insoluble residue. The solvent was removed in vacuo to yield the product as a semisolid material. Yield: 4.5 g (0.018 mol, 99%); 1H NMR (300 MHz, CDCl3): δ3.22 (t, J=6.6 Hz, 2H), 1.55 (m, 2 H), 1.21 (m, 24 H), 0.82 t, 3H) ppm;

13C NMR (75.48 MHz, CDCl3, TMS): δ 134.0, 42.1, 31.2, 30.5,

29.9, 29.8, 29.4, 28.8, 26.4, 22.1, 20.2, 16.0, 13.5 ppm.

1-Pentadecyl-3-[8-(3-pentadecyl ureido)octyl]urea (C15-8; 4.8): This

was synthesized as described for C14-6, starting from 1-isocyanato-pentadecane (0.4 g, 1.6 mmol) and 1,8-diaminooctane (100 mg, 0.7 mmol) in hot toluene (20 mL). Yield: 380 mg (85%); m.p. 167 - 169 ºC (decomp);

1H NMR (300 MHz, TFA -

CD3OD): δ 2.70 (t, J=6.6 Hz, 8H), 1.02 (m, 8H), 0.66 (m, 56 H), 0.22 (s, 6H) ppm; 13C NMR (75.48 MHz, TFA - D2O): δ (152.7), 36.3, 26.1, 23.7, 23.6, 23.5, 23.4, 23.1, 22.9, 22.8, 20.6, 20.5, 16.7, 7.3 ppm; IR(KBr): 3336, 1611, 1576 cm

-1;

C40H82N4O2 (651.12): m/z: 651.6. Anal: calcd: C 73.79, H 12.69, N 8.60; found: C 73.41, H 12.57, N 8.35.

1-Pentadecyl-3-[9-(3-pentadecyl ureido)nonyl]urea (C15-9; 4.9):

This was synthesizedas described for C14-6, starting from 1-isocyanato-pentadecane (0.35 g, 1.4 mmol) and 1,9-diaminononane (100 mg, 0.65 mmol) in hot toluene (20 mL). Yield: 390 mg (90%); m.p. 161 -163 ºC;

1H NMR (300 MHz,

TFA - CD3OD): δ 2.72 (t, J=6.6 Hz, 8H), 1.03 (m, 8H), 0.67 (m, 58 H), 0.24 (s, 6H) ppm; 13C NMR (75.48 MHz, TFA - D2O): δ (153.1), 35.9, 25.8, 23.5, 23.4, 23.3, 23.2, 23.1, 22.8, 22.7, 20.2,16.4, 7.0 ppm; IR (KBr): 3336, 1611, 1574 cm

-1;

C41H84N4O2 (665.14): m/z: 665.6. Anal: calcd C 74.04, H 12.73, N 8.42; found: C 73.68, H 12.75, N 8.39.

STM: Prior to imaging, all compounds under investigation were dissolved in 1-octanol and a drop of this solution was applied to a freshly cleaved surface of highly oriented pyrolytic graphite (HOPG). The molar ratio in solution of the mixtures of T2 and CX-Y was 1:1. The STM images were acquired in the variable-current mode (constant height) under ambient conditions with the tip immersed in the liquid. In the acquired STM images, white corresponds to the highest and black to the lowest measured tunneling current. STM experiments were performed with a Discoverer scanning tunneling microscope (Topometrix Inc., Santa Barbara, CA) along with an external pulse/function generator (Model HP8111 A), with negative sample bias. Tips were electrochemically etched from Pt/Ir wire (80%/20%, diameter 0.2 mm) in a 2M KOH/6M NaCN solution in water. The experiments were repeated in several sessions with different tips to check for reproducibility and to avoid artifacts. Different settings for the tunneling current and the bias voltage were used, ranging from 0.3 nA to 1.0 nA and –1.0 mV to -1.5 mV, respectively. All STM

HN

HN

NH

NH

O

O

HN

HN

HN

HN

O O




98

images contain raw data and are not subjected to any manipulation or image processing. STM images were obtained in the laboratory of prof. De Schryver and dr. De Feyter at the university of Leuven, Belgium. Molecular Modeling: Molecular modeling calculations were carried out using the compass forcefield, as implemented in Materials Studio

20, a product of Accelrys,

San Diego, CA, USA. The energy minimizations were carried out in the gas phase with a dielectric constant of 1. All energy-terms were included with the exception of an explicit hydrogen-bonding term. For the non-bonding interactions a cut-off radius of 12.5 Ǻ was used, with a spline width of 3 Ǻ, and a buffer width of 1.0 Ǻ. A graphite sheet, 20 x 30 atoms in size, with fixed cartesian position for the carbon atoms was used as the substrate. All structures were subjected to energy minimization using the Fletcher-Reeves algorithm, to a final gradient with maximum derivative of 0.001 kcal/mol.

4.12 References

1 Comprehensive supramolecular chemistry (Eds: Atwood, J.L., Davies, J.E.D., MacNicol,

D.D., Vogtle, F.) Pergamon, New York, 1996.

2 Ulman, A. Chem. Rev. 1996, 96, 1533 -1554, and references therein.

3 Bain, C.D., Whitesides, G.M. J. Am. Chem. Soc. 1988, 110, 6560 - 6561.

4 Lewis, P.A., Donhauser, B.Z., Mantooth, J.A., Smith, R.K., Bumm, L.A.., Kelly, K.F., Weiss, P.S. Nanotechnology, 2001, 12, 231 237, and references therein.

5 Stranick, S.J., Atre, S.V., Parikh, A.N., Wood, M.C., Allara, D.L., Winograd, N., Weiss,

P.S. Nanotechnology , 1996, 7, 438 – 442.

6 Tamada, K., Hara, M., Sasabe, H., Knoll, W. Langmuir, 1997, 13, 1558 – 1566.

7 Yokoyama, T., Yokoyama, S., Kamikado, T., Okuno, Y., Mashiko, S. Nature, 2001, 413,

619 - 621.

8 Lorenzo, M. O., Baddeley, C. J., Muryn, C., Raval, R. Nature, 2000, 404, 376 - 379.

9 (a) Venkataraman, B., Breen, J.J., Flynn, G.W. J. Phys. Chem. 1995, 99, 6608 - 6619 (b)

Poulin, J.-C. Microsc. Microanal. Microstruct. 1994, 5, 351 (c) Elbel, N., Roth, W., Gunther, E., von Seggern, H. Surf. Sci. 1994, 303, 424 - 432 (d) Baker, R.T., Mougous, J.D., Brackley, A., Patrick, D.L. Langmuir 1999, 15, 4884 – 4891.

10 (a) Padowitz, D.F., Messmore, B.W. J. Phys. Chem. B 2000, 104, 9943 - 9946 (b) Padowitz, D.F., Sada, D.M., Kemer, E.L., Dougan, M.L., Xue, W.A. J. Phys. Chem. B 2002, 106, 593 - 598 (c) Gesquiere, A., Abdel-Mottaleb, M.M.S., De Schryver, F.C., Sieffert, M., Mullen, K. Langmuir, 1999, 15, 6821 - 6824 (d) Giancarlo, L.C., Flynn, G.W. Acc. Chem. Res. 2000, 33, 491 –501.



Chapter 4

99

11

(a) Qian, P., Nanjo, H., Yokoyama, T., Suzuki, T.M., Akasaka, K., Orhui, H. Chem. Commun. 2000, 2021 – 2022 (b) Hipps, K.W., Scudiero, L., Barlow, D.E., Cooke, J.M. J. Am. Chem. Soc. 2002, 124, 2126 – 2127 (c) Lei, S.B., Wang, C., Yin, S.X., Bai, C.L. J. Phys. Chem. B 2001, 105, 12272 - 12 277.

12 Cf. (a) Xie, Z.X., Xu, X., Mao, B. W., Tanaka, K. Langmuir, 2002, 18, 3113 – 3116 (b) De

Feyter, S., Grim, P.C.M., van Esch, J., Kellogg, R.M.,Feringa, B.L., De Schryver, F.C. J.Phys.Chem. B, 1998, 102, 8981 – 8987.

13 Gesquiere, A., De Feyter, S., De Schryver, F.C., Schoonbeek, F., van Esch, J., Kellogg,

R.M., Feringa, B.L. Nano Lett. 2001, 1, 201 – 206.

14 Thesis F.S.Schoonbeek, 2002, R.U.Groningen.

15 van Esch, J., De Feyter, S., Kellogg, R.M., De Schryver, F.C., Feringa, B.L. Chem. Eur. J.

1997, 3, 1238 – 1243.

16 Examination of 28 crystal structures of noncyclic urea compounds deposited in the

Cambridge Crystallographic Database revealed that the average distance between two successive hydrogen-bonded urea groups is 0.46 nm.

17 Gesquiere, A., Abdel-Mottaleb, M.M.S., De Feyter, S., De Schryver, F.C., Schoonbeek, F., van Esch, J., Kellogg, R.M., Feringa, B.L., Calderone, A., Lazzaroni, R., Bredas, J.L. Langmuir, 2000,16, 10385-10391.

18 Materials Studio provides a poweful software environment to perform a range of

simulation methodologies. www.accelrys.com.

19 De Feyter, S., Larsson, M., Schuurmans, N., Verkuijl, B.V., Zoriniants, G., Gesquiere, A., Abdel-Mottaleb, M.M., van Esch, J., Feringa, B.L., van Stam, J., De Schryver, F. Chem.Eur.J. 2003, 9, 1198 – 1206.

20 www.accelrys.com.




100



Chapter 5 Toward folded structures at the interface. Turnmimics

The 2D spatial positioning of functional groups at 0.2 - 20 nm length scales is still a major challenge in nanotechnology and surface science.

1 A powerful approach is

the decoration of interfaces with organic molecules that form regular 2-D patterns by self-assembly, and recently this strategy has been exploited to organize discrete molecular features.

2 However, most systems reported up to date have

only limited possibilities to include functionality and are not flexible with regard to size and symmetry.

3 In chapter 4, a methodology was proposed to influence the

spatial disposition of functional entities by means of mixing complementary adsorbents. This approach proved to be valid with respect to the assumptions that had been made, but is in practice not very useful. This was caused by the formation of aggregates, characterized by a statistic distribution of the components. In this chapter a new approach towards pattern formation in physisorbed monolayers will be explored, with the aim to provide a new strategy for organizing functionality in a plane.

5.1 Introduction

In view of the problems encountered in pattern formation via self-assembly it is instructive to admire and be inspired by the solutions that have been found to this problem in nature. In many biological systems exact positioning of functional groups is of key importance for the correct operation of large molecular ensembles.4 Biopolymers such as DNA and proteins assemble with precision and accuracy into well-defined conformations. From trillions of possible conformations, the necessary one is seemingly easily found in a split second. The ability to achieve this remarkable feat is of fundamental importance to life as we know it. Because only correctly folded proteins are able to fulfill their functions, precise control over the position of every residue in 3-D is necessary. Only then (chiral) recognition is possible, enabling enantioselective catalysis and transport of small molecules.5 Misfolded proteins are prone to aggregation, which can cause serious illness.6



Turnmimics

102

Proteins are linear oligomeric strands, composed of α-amino acids. Their primary structure however is very different from the structure adopted in living cells. Some sequences of amino acid form secondary structural motifs, notably (α) helices and (β) sheets. The ensemble of secondary elements in turn folds into a specific conformation, fixing the positions of every residue in 3-D space.7 The formation of this tertiary structure is a very complicated process, intensively studied in the biochemical and biophysical communities,8 and is well outside the scope of this thesis. Suffice it to make two remarks here:

(1) A plethora of long-range interactions (H-bonding, sulfur bridges, polar interactions, and hydrophobic interactions) cumulate in the determination of the 3-D shape.

(2) In the folding process, from a random coil (or random conformation) to a folded one, entropy will be lost to make an enthalpic gain. The process can in general be examined evaluating the partition function, from which the Gibbs energy of folding can be derived as ε + kTN ln ω , with ε being the enthalpy gain, ω the number of conformations available, N the number of segments, and kT the Boltzmann constant x temperature. The free energy surface (FES) for this process will resemble a smooth funnel, directing the process towards its minimum.

The secondary structural elements however are less complex and still highly interesting motifs, formed in a self-assembly process. This has rendered them attractive model systems for synthetic chemists. En route towards protein mimetics9 a number of non-natural systems, capable of adopting specific conformations have been developed. Many, but not all, are based on amino acid sequences. J. Moore has coined a term to describe this kind of man-made oligomers: foldamers. A definition was added: any oligomer that folds into a conformationally ordered (collapsed) state in solution, the structures of which are stabilized by a collection of non-covalent interactions between non-adjacent monomer units.

10

Figure 5.1 Examples of folded architectures. Adapted from ref. 9.



Chapter 5

103

Not surprisingly, helical structures can be formed by helicenes.11 But helices have also been formed from phenylene ethynylene oligomers.12 The twist sense of these helices can even be altered.13 These helical foldamers have been used for molecular recognition.14 The formation of the helix structure relies on solvophobic interactions. In other examples aromatic π−π stacking plays an important role in the formation of the helix.15 Most examples of helices are actually oligopeptides.16 These can either be composed of α-amino acids (the naturally occurring), or β, γ, or δ peptides. Some of these oligopeptide foldamers are indeed capable of recognition and transformation of molecular targets.17 Formation of the secondary architecture in these systems relies heavily (but not exclusively) on hydrogen bonding of the amide groups, not unlike the situation in natural proteins. Comparably folded structures have been obtained with hydrazine18 - and urea19 moieties providing the H-bonding interactions.

In the context of this thesis, however, the β-sheet structure is the most interesting. Note that the β-sheet can be regarded as a folded ribbon (fig. 5.1), the two-dimensional analogue of a helix. It therefore is intrinsically compatible with the assembly process at interfaces. A β-sheet is typically composed of individual oligopeptide strands, which are connected by means of a turn (β-turn). The turn, in fact, promotes a 2-dimensional reversal of direction in a longer sequence. In natural β-sheets the turn is frequently formed by a two-residue dimer.20 The strands can have a parallel or an anti-parallel orientation in the sheet (fig. 5.2).21

Figure 5.2 General structure of parallel and anti-

parallel β-sheets. Image from www.nku.edu.

Anti-parallel

ββββ-sheet

Parallel

ββββ-sheet



Turnmimics

104

Surprisingly little work has been done towards artificial β-sheets. Maybe this arises from the fact that β-sheets have a supporting role in the structure of proteins, and are less prominent in the active sites, which renders them less interesting for medical purposes. For natural sequences of α-amino acids, the propensity to form either helices or β-sheets has been experimentally determined for a large number of residue combinations,22 which has allowed the development of commercial software, which accurately predicts the structure of many sequences. Some is available free of charge on the internet.23 For peptide sequences incorporating non-natural residues the situation is more complicated.

β-turn mimics for incorporation into artificially de novo designed peptides have been developed. Incorporation of a β-turn mimic has been used to program a temperature dependent folding event into a designed sequence.24 The folding triggers the formation of a hydrogel. In aqueous solutions D-proline (enantiomer of the naturally occurring amino acid) seems to be effective in driving the formation of β-hairpins.25 This property has been exploited by Tirrell et al. to construct a peptide sequence that forms a β-sheet monolayer at the air-water interface (fig. 5.4).26

On the other hand some nylons (synthetic polyamides) are known to crystallize in plates, where the individual molecules adopt a folded conformation, without turn elements.27 Formation of β-sheet type structures on solid substrates is virtually unknown. Alternatively, elements can be incorporated into a peptide sequence that forces the formation of a closed loop. This kind of structure can be regarded as a molecular tile. This principle has been exploited to form monolayers on gold.28

5.1.1. Bridging the gap between surface patterning and ββββ-sheets

In the previous chapters it has been demonstrated that linear alkyl fragments featuring amide or urea groups form lamellar structures upon adsorption at the solid-liquid interface of HOPG. These lamellae are structurally reminiscent of the β-sheets found in nature.

Figure 5.3 Oligomeric sequences of alkyl bisureas, connected with a turn element. Folding creates a tile-like morphology, which allows effective coverage of the surface upon adsorption.



Chapter 5

105

It should thus be clear from the above that untold opportunity exists to use the concept of folded oligomers, and the formation of artificial β-sheet type structures at the solid-liquid interface in the context of surface patterning. The general definition of foldamers can then be extended to include interfaces (cf. solution) as well.

The idea is illustrated in fig. 5.3. The individual alkyl bisurea strands, encountered in chapter 4 might be connected, with a β-turn mimic. The urea groups will in that situation be responsible for intramolecular as well as intermolecular interactions (fig. 5.3), not unlike the situation in artificial β-amino acids. Mind the similarity of the proposed structure with Tirrell’s artificial peptide, that folds into a β-sheet type structure at the air-water interface (fig. 5.4).

The goal of the work described in this chapter is to arrive at a synthetic analogue of the oligopeptides that adopt folded conformation to form β-sheets in proteins. This analogue should additionally be capable to give effective coverage on a graphite surface, in order to form a monolayer. The aggregate will probably look like the one proposed in fig. 5.3.

The central topic of this chapter is the development of a model system, incorporating such a turn-mimic, and the study of the monolayers it forms. The first task is the design a turn-mimic, which is compatible with the strands that are normally used in physisorbed monolayers, i.e. alkyl fragments, possibly featuring H-bonding moieties (ureas or amides).

In the future, this system can serve as a platform, that can be elaborated. Tailoring the chemical composition along the alkyl fragments, as established previously seems to be not a priori incompatible with the proposed introduction of a turn

Figure 5.4 Formation of β-sheets at the air water interface. The letters are standard abbreviations for amino acid residues. Image taken from ref. 26.



Turnmimics

106

element. Altogether, it is a conceptually simple, but practically audacious idea, adding new levels of complexity to both the self-assembly process and the synthesis of the adsorbents.

5.2 Foldamers in 2-D

The idea looks simple on paper, but it requires a high degree of control over the conformation of extended oligomeric chemical species.The targeted molecules must fold into a precisely defined conformation at the solid-liquid interface. In order to do so, a number of prerequisites will have to be fulfilled. The basic requirements are:

• The entire structure should be a two-dimensional object, i.e. it should be planar.

• There have to be turning points, so that a bend is created that positions the linear strands in a plane, in a parallel fashion.

• The strands are to have intermolecular interaction points, to glue the individual stretches together.

• These interaction points are kept in registry, with respect to position as well as orientation.

The interactions can be purely van der Waals in nature, but the most obvious choice would be the use of H-bonding units. In previous chapters the introduction of urea groups with the aim to provide intermolecular interactions has been discussed. In this context however the use of amide moieties is more obvious, the concept being to a large extent inspired by architectures that are built from amino acids, thus utilizing amide groups. The amide group is less symmetric, allowing for greater versatility (CONH vs. NHCO). Also the synthetic methodology for making amide groups is much better documented. The strength of the H-bonds formed by amide groups is slightly less then that of the ones formed by ureas,29 and consequently the (intermolecular) distance between two amide groups held together by H-bonds is larger. The typical distance between urea groups is 4.6 Å, for amide groups this is 5.0 Å. This implies that the spacing affected by the turn elements should correspond to this value.

Not only do the amide groups introduce intramolecular H-bonding, also intermolecular H-bonding is provided. This renders the foldamers capable of forming extended two-dimensional structures (monolayers; cf. fig. 5.3).



Chapter 5

107

5.3 The turn element

The turn element itself is preferably flat and quite rigid, if it is to comply with the aforementioned criteria. A degree of unsaturation helps herein. In fig. 5.5 are listed some potential candidates. The (aromatic) moieties shown are flat, bifunctional, and derivatives are synthetically accessible; the respective substitution patterns endow these structures with the necessary geometry for making turns. It should be noted that the intramolecular distances between the substituents are variable, and usually smaller than 5 Å. Appending the amide groups (vide supra) with a small spacer can introduce the flexibility to place them 5 Å apart. To get an idea about the interplay between spacers, amide groups and turn elements, a systematic conformational search is essential. The confinement in 2D space greatly reduces conformational degrees of freedom, which facilitates rational design by means of molecular modeling. Details on the molecular modeling procedure are given in the experimental section. It should be stressed here that comparisons between different molecules can not a priori be made. Introduction of additional groups or atoms can give rise to very different interactions, and yields uncomparable data for respective energy-minimized structures. Deconvolution is necessary and some of the contributing terms may be compared, maybe after calibrating for a homologues series. Different conformations of one molecule however can be compared. This is possible because the algorithm30 used identifies local minima on the potential energy surface. A model of the compound can be drawn, and subsequently minimized in different conformations. The potential energies of these conformations can be compared.

Figure 5.5 Possible scaffolds for the construction of a turn-mimic.



Turnmimics

108

If the lowest-energy conformation indeed places the amide groups in a plane and in registry there is a hit. The result is illustrated in fig. 5.6. for a resorcinol derivative (meta substitution). On the left side is the conformation that would assist in the formation of a turn. In this conformation the fragment is flat and the distance between the two carbonyl oxygen atoms of the amide groups is 5.3 Å.

This model was cleaned (assigning standard values for bond lengths and angles), removing impossible bond angles, etc. Its energy was calculated as ca. 232 kcal/mol. Minimizing the structure, in a compass force field using an atom-based summation, gave the model shown on the right in fig. 5.6. Its potential energy was actually much lower, ~ -52 kcal/mol, indicating a more stable structure. The absolute values are not relevant, nor the fact that the values differ in sign (+ 232 vs. – 52 kcal/mol). This arises from the summation. The important message is that

a synthetic foldamer incorporating this ”turn” element is not likely to lie flat on top of a graphite surface, because this conformation does not comply with any of the requirements as stated above. The result can be rationalized considering the hydrogen-atom in between the substituents. This creates electrostatic repulsion in the plane, which causes the substituents to bend away. Although the use of longer spacers will probably reduce this negative contribution it is clear that resorcinol derivatives are not the ideal scaffolds to build synthetic foldamers from.

5.4 Catechol as the turn element

The analysis as described above has been made for all the scaffolds depicted in fig. 5.5. If the spacers are connected via an ortho-substitution pattern (with oxygens the parent compound is called catechol) potentially a turn can be made. A model based on catechol with short spacers and two amide groups was generated in Materials Studio.31 The model is shown in fig. 5.7 (left). This is an isomer of the

Figure 5.6 Model of resorcinol derivative as scaffold for foldamers. Desired (left) and energy minimized (right).



Chapter 5

109

resorcinol derivative shown in fig. 5.6. It can be seen immediately that this is not a favorable conformation, because the substituents are too much in proximity. The potential energy is correspondingly high, mainly due to a very unfavorable repulsive van der Waals term. Minimizing this model under the same conditions as before generates the model shown on the right side of fig. 5.7. The substituents have bent away, out of the plane; the amide groups are not in registry, and the distance between them is about 6.3 Å. The potential energy is –56 kcal/mol. This is lower than the value for the resorcinol derivative. Although one should be cautious comparing the two, here the difference is illustrative, as they are isomers.

Because the other scaffolds did not give more encouraging results, other parameters were varied. Taking catechol as the scaffold, dissymmetry with respect to the spacers can be introduced. An important objective for doing this can be seen in fig. 5.7. Directionality of the amide moieties is opposite, because the substituents are connected to the aromatic part in a divergent way. Introducing an extra methylene unit in one of the spacers will bring the amide groups back in registry again. A model was generated in Materials Studio (fig. 5.8). This structure has one C2 spacer and one C3 spacer. The initial structure is obviously unfavorable, but the minimized conformer (fig. 5.8 b) is not far off the envisaged properties. The amide groups are still not in registry, and the distance is 5.3 Å, a little too high for optimal H-bonding. The calculated potential energy is –41 kcal/mol, a value that should not be compared to the one found before (because this compound is larger by one methylene group).

The structure in fig. 5.8b gives a hint to how in a catechol-based derivative the correct spacing between the amide groups can be realized. The clue is in the dihedral angles defining the angle between the planes of the substituents and aromatic moiety respectively (indicated in fig. 5.8a with arrows). Because one of the spacers bends away, the distance between the chains is close to 5 Å, without the introduction of torsional strain close to the amide groups. This situation can be idealized in a model.

Figure 5.7 Models of catechol-based turn element. Unfavorable (left) and energy-minimized (right).



Turnmimics

110

The initial conformation, from which the minimization will be carried out, is characterized by dissymmetry in spacer length as well as dihedral angles. The shorter spacer makes a 0º dihedral angle with the aromatic plane; the longer spacer makes a 180º dihedral. The energy-minimized structure (fig. 5.8 c) complies with the demands. The distance is almost 5 Å. The calculated potential energy for this structure is -48 kcal/mol. The extra stability is mainly due to the better disposition of the amide groups, which allows for efficient H-bonding. The dihedrals are 140º and -40º respectively (difference is 180º). These results underscore the importance of using multiple initial conformations in a modeling session, so that local minima can be compared and evaluated. A sound chemical intuition herein is indispensable.

A final in silico experiment involved minimization of the structure in fig. 5.8c in the proximity of a sheet of graphite, to see if it would maintain its structural properties. Several modeling sessions established an equilibrium distance of approximately 3.5 Å between the molecular fragment and the graphite. Taking this initial value, the minimum-energy conformation hardly changed. The graphite in these sessions was represented as a sheet of 50 x 50 atoms, under the constraint of fixed cartesian coordinates. No charges were taken into account for the graphite.

Figure 5.8 Bis-amide catechol derivative with dissymmetrized spacers in not-optimized (a) energy-minimized (b) and desired (c) conformation. The arrows mark the first and last atom of the sequence over which the dihedral angle is determined.

a) b)

c)



Chapter 5

111

5.5 Towards functional turn mimics

Because it is known that alkyl chains interact favorably with the surface of graphite32, it was anticipated that introduction of longer spacers would assist in assembly of the molecules onto the graphite. Moreover, the van der Waals interactions between the alkyl fragments will influence the folding process favorably, because a conformation with the chains in close proximity is additionally stabilized. On the down side a negative entropic contribution is to be expected, because the amide groups are further apart (more atoms in between; lower effective molarity).

The situation has been evaluated for two homologues of the aforementioned models: catechol derivatives with two C10 spacers and two C12 tails (compound 5.1, symmetric), and with one C10 spacer and one C11 spacer with two C12 tails (compound 5.2, dissymmetric). The specific lengths of the spacers were chosen for reasons of synthetic accessibility. The respective structures were generated in Materials Studio and energy-minimized to test if the results described in paragraph 5.4 can be extended to a generic principle for this class of molecules. The result for compound 5.1 is shown in fig.5.9. The structure was minimized in the proximity of a sheet of graphite. The most striking aspect of the conformation this molecule presumably will adopt when assembling at the graphite is the twist along one of the spacers. Contrary to its short homologue this molecule does allow placement in registry of the amide groups; this gives the molecule an overall planar appearance. The penalty for realizing this conformation is the twist, which induces quite some strain. Moreover, distortion from an extended all-trans local conformation is expected to hamper interaction with the graphite. Comparing the two spacers, one seems to be rotated by 90º with respect to the other.

Figure 5.9 Energy-minimized conformation of a model of compound 5.1. Minimized in the presence of a stash of graphite, which is omitted for clarity.



Turnmimics

112

The result for compound 5.2 is shown in fig. 5.10. In this structure the amide groups can be placed in registry, in an overall flat arrangement, without introducing additional strain. Because it is difficult to compare calculated potential energies for non-isomeric compounds, comparisons were made between the respective differences in energy for the folded (figs. 5.9 and 5.10) and “extended” conformations near a graphite slab.

The extended conformations have the substituents bent away, and are consequently not capable of forming intramolecular H-bonds. On the account of this effect both the folded conformations are supposed to be more stable than the

extended ones, and it is expected that the asymmetric compound 5.2 shows a greater energy gain in going from the extended to the folded conformation than the symmetric compound 5.1. The results are summarized in table 5.1 (energies are given in kcal/mol). The two compounds are indicated as compound 5.1 and compound 5.2, respectively. The potential energies for the extended en folded conformations were calculated for isolated molecules, representing the situation in solution (sln) and in the presence of graphite (grf). The energy difference between the folded and extended conformation in solution for the respective compounds is given in the second; the energy difference on graphite in the third column. It can be seen that compound 5.1, in going from the extended to the folded conformation in solution gains 23 kcal/mol; and compound 5.2 gains 22.5 kcal/mol. These values are almost the same. But in the presence of graphite compound 5.1 gains only 19.3 kcal/mol, whereas compound 5.2 gains 23.6 kcal/mol. This reflects the arguments discussed above. Compound 5.1 has to pay a penalty to allow for efficient in-plane H-bonding; compound 5.2 does not, and is in fact assisted by the graphite in adopting the preferred conformation. The third column reports on the difference between graphite and solution. It can be seen that the folded conformation of compound 5.1 is less stabilized by stacking on the graphite than the extended one; for compound 5.2 the folded conformation is slightly more stabilized.

Figure 5.10 Energy minimized conformation of a model of compound 2. Minimized in the presence of a stash of graphite, which is omitted for clarity.



Chapter 5

113

These data indicate that the dissymmetric catechol derivative is better disposed to organize itself on a graphite-surface than the symmetric one. It should be noted however that these models oversimplify the real situation. Entropic contributions are not treated; only single molecules have been modeled, ignoring intermolecular interactions, both in solution and in the aggregates; the dielectric constant of the medium was fixed at 1.00, instead of a value corresponding to a real solvent. Despite these assumptions and approximations, and despite the limited number of initial conformations generated, the modeling sessions established a consistent picture indicating that folding is facilitated by a dissymmetry with respect to the spacers.

5.6 Synthesis of the compounds

Having explored the theoretical background in some detail, the ideas must be tested. In order to do so compounds 5.1 and 5.2 were synthesized. It was decided to synthesize the amide-bearing substituents first. Commercially available bromo undecanoic acid was converted to its acyl chloride (quantitative), and subsequently reacted with n-dodecylamine. The same procedure was followed starting with bromododecanoic acid. These reactions proceeded in 75% yield and 70% yield respectively, and produced ω-bromo amides 5.4 and 5.5. Synthesis of the symmetric derivative proceeded with catechol, excess base and ω-bromo-amide 5.4. The product precipitated from the reaction mixture, and was collected in 50% yield. For the synthesis of the dissymmetric derivative 2, catechol was reacted with a substoichiometric quantity of 5.4. After work-up dialkylated product was removed by precipitation and monoalkylated intermediate 5.6 was collected as an oil, in 55% yield. Reaction of 5.6 with bromo amide 5.5, under the same conditions produced dissymmetric derivative 5.2 in 67% yield (scheme 5.1). All compounds were characterized with 1H-NMR and 13C-NMR. Compounds 5.1 and 5.2 were additionally characterized by HRMS.

Entry (compound)

∆∆∆∆Efolded – extended

in solution (kcal/mol).

∆∆∆∆Efolded – extended

on graphite (kcal/mol).

5.1 -22.9 -19.3

5.2 -22.5 -23.6

Table 5.1 Summary for calculated energy differences between the folded and extended conformations of compounds 5.1 and 5.2 in different environments.



Turnmimics

114

5.7 Monolayer formation of compounds 5.1 and 5.2

Upon applying a drop of a dilute solution of 5.1 on the top of the basal plane of graphite, a monolayer spontaneously formed, as observed by STM (fig.5.11). The bright spots correspond to the catechol groups and the darker rods to the alkyl chains. The sites of H-bonding, via the amide groups, appear as bright spots in the middle of the alkyl chains in the monolayer of 5.1. The apparent molecular lengths, indicated by the lamella widths (∆L1 = 3.5 nm) are in agreement with the models. The distance between adjacent alkyl chains was measured to be 0.47 ± 0.01 nm.

To analyze what this image tells about the individual molecular conformations at the interface a line profile can be made. This is illustrated with the line in fig. 5.11. Along such a line the number of groups can be counted. Because the aromatic and aliphatic regions can be identified, and the amide groups and catechol moieties show up as bright spots, it is possible to count the number of spots along lines parallel to the long lamellar axis, traversing the amide groups and the catechol groups in the molecules respectively. The result of this analysis is shown in fig. 5.12. A bright spot is associated with a high tunneling current, so the spots in the STM image show up as spikes in the profile. Fig 5.12 a) shows the profile over the amide groups, and fig 5.12 b) is the frequency Fourier transformation of this

BrC10H20CO2H1. SOCl2

2. C12H25NH2

75%

BrC10H20CONHC12H25

OH

OH

K2CO3acetone

OC10H20CONHC12H25

OC10H20CONHC12H25

5.1

5.4

BrC11H22CO2H1. SOCl2

2. C12H25NH2

70%

BrC11H22CONHC12H25

5.5

OH

OH

K2CO3

acetone

1.2 equivalent

0.4 equivalent

OC10H20CONHC12H25

OH

5.6

K2CO3acetone

OC10H20CONHC12H25

OC11H22CONHC12H25

5.4

5.2

50%

55%

67%

Scheme 5.1 Synthesis of compounds 5.1 and 5.2.



Chapter 5

115

analysis, indicating a peak at 2 spots/spikes per nanometer. Fig.5.12 c) shows the profile over the aromatic moieties, and 5.12 d) is the frequency Fourier transformation, again indicating a peak at 2 spots/spikes per nanometer. Comparing 5.12 a) and c) shows a similar number of spikes, which is elucidated in fig. 5.12 b) and d). The similarity of the line profiles carries one important message:

a) b)

c) d)

Figure 5.12 Line profile analyses (a and c) and FFT (b and d) over amide groups (a and b) and catechol groups (c and d).

Figure 5.11 STM images of the monolayer formed by compound 5.1 (left) and magnification (right). Iset = 0.6 nA; Vbias = -0.90 V.

2 nm



Turnmimics

116

for each molecule there is one amide group, and one catechol group on the surface! This means that the molecules do not adopt a folded conformation in the monolayer at the graphite surface. In the aggregate, every molecule has one arm to interact with the surface. This corresponds to the calculated twist in one of the spacers, and diminished interaction of the alkyl chain with the graphite in this conformation. Although the calculations suggested that a folded conformation is more favorable than an extended (or random) one, folding is not observed.

Furthermore, the monolayer is quite irregular. In the magnified STM image in fig. 5.11 it can be seen, on the right, that the bright row suddenly stops. This is also reflected in the line profiles in fig. 5.12, where other frequencies than 2 are encountered (fig. 5.12 b and d).

This observation can be explained if the molecules arrange themselves in head-to-head as well as head-to-tail orientations. It is thought therefore that the situation in the aggregate is like the representation in fig. 5.13

It is possible that the molecules are not organized coplanar to the surface but edge-on. The molecules can still be folded then (have intramolecular H-bonding) but the second arm is off the surface and thus not observed. Another explanation is that the molecules are not folded, but have the second arm dangling in the supernatant solution, an argument that has been previously invoked (chapter 3).

5.7.1 The case of compound 5.2

The STM experiment has been repeated with compound 5.2. Also this compound formed monolayers. Typical STM images are shown in fig. 5.14. The images appear to be better resolved and show a more regularly organized monolayer. In contrast to the monolayers formed by 5.1, the catechol groups appear more isolated in monolayers of 5.2. The line profile analyses (dotted lines in 5.14 left, and solid / dotted lines in 5.14 right) are shown in figs. 5.15a and b. It is immediately apparent that the frequencies of amide groups and catechol groups are different. In fact the number of amide groups is twice the number of catechol groups in fig. 5.15b. This means that for compound 5.2, molecules do adopt a folded conformation in the monolayer. More detailed analysis of the line profiles however indicated that the situation is more complex. Fig. 5.15 gives a more

Figure 5.13 Proposed model explaining the

orientation in the aggregate for compound 5.1.



Chapter 5

117

accurate picture: folded as well as unfolded conformations are observed in coexistence on the surface. The line profile can be rationalized in terms of the model shown in fig. 5.15c. Both the alkyl chains of (arbitrary) molecules i, iii, and iv are adsorbed on the surface, while one of the alkyl chains of molecule ii points into the solution. As before, head-to-head as well as head-to-tail orientations in the lamellae are observed (fig. 5.14; the white dots seem scattered over the surface, and do not organize in long rows).

Figure 5.14 STM image (left) and magnification (right) of monolayer formed by compound 5.2. Iset = 1.0 nA; Vbias = -0.82 V.

a)

Figure 5.15 Line profile analyses over the STM images in fig. 5.14. a) left. b) right). c) Model fitting the profile in 5.15 a).

b)

c)



Turnmimics

118

The situation is thought to be like the representation in fig. 5.16.

The orientational degeneracy is probably related to the symmetry around the amide groups. Since these groups are located in the middle of the aliphatic part of the molecule, there is no preference for the catechol groups to be aligned at one side, as intermolecular hydrogen bonding is still possible when molecules flip over.

The STM results, to a considerable extent, corroborated the ideas that were put forward on the basis of molecular modeling. The asymmetric compound indeed seems to have a higher propensity to fold in a two-dimensionally constrained environment. The compound is clearly not ideal, as non-folded conformations are also observed. Furthermore is the flip over of molecules along a lamella an unwanted side effect.

5.8 Further modeling studies

From the above it is clear that an improved turn element should be designed. Till so far only one possibility to push these catechol derivatives into an intramolecularly H-bonded conformation has been looked at, i.e. the class where the spacers have length Cn and Cn+1 respectively. The question arises whether this case represents the full scope, and therefore the influence of various spacer combinations was explored in more detail. First consider the model compound with short spacers, obeying the CnCm (n=m+1) rule again. Hitherto it has been assumed that intramolecular H-bond formation is facilitated if the longer spacer (the one with most methylene groups) adopts a gauche conformation (fig. 5.17 a). But what if the shorter spacer adopts the gauche conformation (fig. 5.17 b)? The model indicates that this conformation does not facilitate H-bond formation (as measured by the distance between amide-NH in one substituent, and amide-carbonyl in the other) and consequently has more potential energy. But is this a generic trend? Consider the homologue in fig. 5.17 c) and d). Both spacers are extended by one C-atom. In 5.17 c) the longer spacer adopts the gauche conformation, whereas in 5.17 d) the shorter spacer does. The models indicate that both conformations are well disposed for H-bond formation, and consequently their potential energies are very similar. Summarizing, there seems to operate an odd-even effect in deciding about the number of H-bonding conformations. This is an unexpected finding with potentially interesting consequences.

Figure 5.16 Model for the organization of compound 5.2 in the monolayer.



Chapter 5

119

Both conformations are equally likely, but mutually exclusive in aggregate formation, as the directionality of the respective amide groups is opposite, and the mode of H-bonding is different: the H-bonding array is normal to the long axis of

the molecules in 5.17c; while it makes an angle with this axis in 5.17d. Therefore it can be expected that monolayers will be made up of separate domains. This indicates that supramolecular polymorphism can arise from dynamics (conformational interconversion) at the molecular level. It is expected that monolayer formation is complicated by this phenomenon, so the original choice for a C10-C11 combination was a lucky one.

A closer look at the symmetric derivatives (Cn-Cn) revealed that that also in the case of n = odd no stable H-bonded conformations could be identified. The case of Cn-Cn+2 showed the same tendency (for n = odd as well as n = even). The case of Cn-Cn+3 however revealed the existence of one H-bonded conformation for n = odd, whereas none were found for n = even. So this situation is different than Cn-Cn+1, where one H-bonded conformation was found for n = even. The results are summarized in table 5.2 (stable foldamer refers to a conformation which allows intramolecular H-bonding).

Figure 5.17 Comparison of conformations for catechol derivates with spacer combinations Cn-Cn+1. Note the distances between the amide groups. (a) n = even, long = gauche (b) n = even, short = gauche (c) n = odd, short = gauche (d) n = odd, long =gauche. The arrows identify the directionality of the H-bonding.

a) b)

c) d)



Turnmimics

120

5.9 Towards a new turn element

Armed with this new information, a number of interesting targets were selected. These represented the three classes that were found to be able to give H-bonded conformations: Cn-Cn+1(n = even), Cn-Cn+1(n = odd) and Cn-Cn+3 (n = odd). It was anticipated that shorter spacers would be required to undo the symmetry around the amide groups, thus eliminating “flip-over” along the lamellae. Attempts were made to synthesize catechol-bisamides with C2 - C3 spacers, with C3 - C4 spacers and with C3 - C6 spacers. Questions concerning the sequence arise. Which substituent should be placed first? And at which stage should the amide groups be introduced? For the derivative with C2 - C3 spacers it was arbitrarily decided to introduce the C2 fragment first, and to introduce the spacers as methylesters, which could then be converted to acylchlorides, and subsequently amides together in other same step. Commercially available bromopropionic acid was quantitatively converted to its methyl ester via a standard procedure. The coupling between this compound and catechol however could not be affected. The lack of reactivity is attributed to the basic conditions (vide supra) in this step. Elimination of the components of HBr produces methylacrylate as the main product (fig. 5.18a). Because this problem is in a reactant that has to be used, reversing the sequence will not help much here. First converting bromopropionic acid to the corresponding n-dodecylamide will also not resolve the problem. Therefore this target was deemed synthetically not easily accessible, and attention was focused on the derivative with C3 - C4 spacers.

onen,n+3nonen,n+3

twon,n+1onen,n+1

nonen,nnone n,n

Number of stable foldamers

n = oddNumber of stable foldamers

n = even

onen,n+3nonen,n+3

twon,n+1onen,n+1

nonen,nnone n,n

Number of stable foldamers

n = oddNumber of stable foldamers

n = even

Table 5.2 Summary of the occurrence of intramolecular H-bond formation in catechol derivatives with various spacer combinations.



Chapter 5

121

Monofunctionalization of catechol with either methyl bromobutyrate or methyl bromovalerate proved to be difficult in either acetone or cyclohexanone. The complex reaction mixtures could not be separated. Because compound 5.2 could be synthesized from catechol and the bromo-amides (vide supra), an attempt was made to convert the esters first into the amides. This was not successful however due to lactam formation (the amides, upon formation, undergo an intramolecular reaction, eliminating bromide; fig. 5.18b). Use of microwaves, and a hindered base did not improve the results (complex reaction mixtures).

Instead of attempting to synthesize homologues attention was now directed to the derivative with C3 – C6 spacer combination. Experience in the difficulties of the coupling reaction prompted a more careful approach. It turned out that catechol can be protected as the monoacetate. In the reaction of catechol with acetic anhydride, the mono-protected intermediate separated from the reaction mixture (as an oil). The acetate was collected and crystallized (40% yield). Catechol monoacetate was reacted with 7-bromo methyl heptanoate, obtained from the hydrolysis of commercially available 7-bromo hepanenitrile (60% yield) and subsequent methyl esterification (quantitative yield). This coupling was successful and, as it turned out, the acetate protective group was not stable towards purification over a silica column. Compound 5.7 was obtained in 54% yield, and subsequently reacted with methyl bromobutyrate. This reaction yielded compound 5.8, a catechol derivative with differentiated spacers, in 52% yield. The methyl esters were subsequently, in a few steps, converted to the n-dodecyl amides to give compound 5.3 (30% overall). The sequence is illustrated in scheme 5.2. Intermediates were characterized by 1H-NMR and 13C-NMR; compound 5.3 additionally by HRMS.

CO2Me

Br

O

NHR

NR

O

Br

CO2Me

Figure 5.18 Problems in the synthesis of derivatives with C2 – C3 spacers and C3 – C4 spacers. (a) formation of methylacrylate (b) formation of valerolactam.



Turnmimics

122

5.10 Modeling and monolayer formation of compound 5.3

Before investigating the monolayer formation of compound 5.3 a model was generated in Materials Studio31 (fig. 5.19). The model shows an almost ideal disposition of the amide groups for H-bond formation. The distance between the carbonyl-oxygen and amide-nitrogen is 1.9 Å (the distance between the carbonyl groups is 5.3 Å). The choice of spacers also dictates the directionality of the H-bonding array, which is opposite compared to compounds 5.1 and 5.2 (the carbonyl groups are pointing down with respect to the long axis).

Figure 5.19 Energy-minimized conformation of compound 5.3. Minimized in the proximity of a stash of graphite, which is omitted for clarity.

OH

OH

OH

OAc

OC6H12CO2Me

OH

5.7

OC6H12CONHC12H25

OC3H6CONHC12H25

OC6H12CO2Me

OC3H6CO2Me

Ac2O

KOH

60%

BrC3H6CO2Me

K2CO3

acetone

54%

K2CO3

acetone

52%

1. NaOH2.SOCl2

3. C12H25NH2 Et3N

30%5.8 5.3

BrC6H12CO2Me

Scheme 5.2 Synthesis of catechol derivative 5.3.



Chapter 5

123

Modeling indicated that the minimum-energy conformation for 5.3 obeyed all the criteria. As can be seen the all-trans conformation is similar to that of compound 5.2. Although this is a smaller molecule (less methylene groups), and can therefore not a priori be compared to compounds 5.1 and 5.2, folding of 5.3 is again favorable with respect to extended conformations (∆E = 18 kcal/mol).

It was gratifying to see that not only did this compound form monolayers, the monolayers were very stable and well-ordered (fig. 5.20). The lamellae display a head-to-head type interaction exclusively. This proves that symmetry elements (or

the lack thereof!) in the molecules can be an influential parameter for the morphology of the aggregate.

Here the amide groups appear dark, which suggests that the in-plane orientation of the H-bonding site is different from the other derivatives. The line profile analysis (fig. 5.20 right) clearly indicates that every molecule adopts the folded conformation on the surface as illustrated with a model (superimposed on the image in fig 5.20; two alkyl chains per catechol group).

curr

ent

1 nm / div.cu

rren

t1 nm / div.

Figure 5.20 STM image (left) and line profile analysis (right) of monolayer formed by compound 5.3. Iset = 0.8 nA. Vbias = - 0.93 V. Some models of the compounds are superimposed on the image. The red crosses indicate the main axes of the graphite lattice.

3 nm



Turnmimics

124

5.11 Conclusions

The result implies that a good turn mimic has been designed and synthesized, and that it behaves the way forecasted by molecular modeling. The length of the spacers between the catechol and amide moieties plays an important role in the folding process. Again it should not be overlooked that modeling is merely a guiding principle, though an effective one. No claim is made that modeling at this level of accuracy can predict the outcome of a process as complicated as aggregate formation, not even in a 2-dimensionally constrained environment. Aggregates with different symmetries can be generated, and minimized in materials studio but their potential energies are very close, which indicates that other (kinetic) parameters are equally important in the process.

In conclusion, a 2D turn element for oligo-amide sequences has been successfully designed. Molecular modeling is a neat tool to gain insight in some aspects of the molecular structure that are relevant for aggregate formation. To the best of our knowledge the findings described in this chapter constitutes the first systematic design of artificial β-turn mimics in the context of physisorbed monolayers. Its potential for organizing functionality in a plane, at a molecular scale has yet to be explored, but the initial results are promising.

5.12 Experimental Section

Materials and methods

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 spectrometer (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 VXR400 spectrometer (at 400 MHz) in CDCl3, chemical shifts are given relative to CDCl3 (77). The splitting patterns in the 1H-NMRspectra 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.

Bromo-undecanoic acid dodecylamide (5.4): Thionylchloride (0.7 ml, 1.1 g, 1.2 equiv.) was added

drop wise to a stirred solution of 11-bromo-undecanoic acid (2.0 g, 7.5 mmol) in CH2Cl2 (50 ml). After addition was complete the reaction mixture was refluxed for 2h, cooled to rt, and concentrated in vacuum to remove excess thionylchloride. The acid chloride (2.0 g) was dissolved in CH2Cl2 (50 ml) and dodecylamine (1.5 g, 8 mmol) and Et3N (0.9 g, 9 mmol) were slowly added. After addition was complete

BrC10H20CONHC12H25



Chapter 5

125

the reaction mixture was refluxed for 12h, cooled to RT, washed (NaHCO3 (aq. sat), H2O; 2 x 50 ml), dried over Na2SO4, filtered over silica and concentrated in vacuum to give a white powder. Yield 2.4 g (5.6 mmol, 75 %); 1H- NMR (400 MHz, CDCl3): δ 5.42 (br, 1H), δ 3.37 (t, 3J = 7.0 Hz, 2H), δ 3.20 (q, J = 7.3, 2H), δ 2.12 (t, 3J = 7.7 Hz, 2H), δ 1.81 (t, 3J = 7.3 Hz, 2H), δ 1.58 (t, 3J = 6.9 Hz, 1H), δ 1.45 (m, 2H), δ 1.22 (m, 36H), δ 0.85 (t, 3J = 6.6 Hz, 3H).13C- NMR (400 MHz, CDCl3): δ 173.3, 39.8, 37.2, 34.4, 33.1, 32.2, 31.2, 30.0, 29.9, 29.9, 29.8, 29.8, 29.6, 29.5, 29.0, 28.4, 27.2, 26.1, 23.0, 14.4.

12-Bromo-dodecanoic acid dodecylamine (5.5): This compound was synthesized as described above

for 5.4, starting from 12-bromo-dodecanoic acid (2.0 g). Yield 2.2 g (4.9 mmol, 70%). White powder. 1H-NMR (400 MHz, CDCl3): δ 5.37 (br, 1H), δ 3.38 (t, 3J = 7.0 Hz, 2H), δ 3.21 (q, J =7.3 Hz, 2H), δ 2.12 (t, 3J = 7.7 Hz, 2H), δ 1.83 (t, 3J = 7.3 Hz, 2H), δ 1.59 (t, 3J=6.9 Hz 1H), δ 1.44 (m, 2H), δ 1.39 (t, 3J=6.8 Hz 1H) δ 1.22 (m, 36H), δ 0.85 (t, 3J = 6.6 Hz, 3H).13C- NMR (400 MHz, CDCl3): δ 173.0, 39.5, 36.9, 34.0, 32.8, 31.9, 30.9, 29.7, 29.6, 29.6, 29.5, 29.5, 29.4, 29.4, 29.4, 29.3, 28.7, 28.1, 26.9, 25.8, 22.7, 14.1.

11-[2-(10-Dodecylcarbamoyl-decyloxy)-phenoxy]-undecanoic acid dodecylamide (5.1): Potassium carbonate (0.5 g, 3.6 mmol) was added to a solution of catechol (0.04 g, 0.36 mmol) in acetone (40 ml) under an atmosphere of N2, After refluxing the

suspension for 10 min, 11-bromo-undecanoic acid dodecylamide 5.4 (0.4 g, 0.9 mmol) was added and refluxing was continued for 48 h. The mixture was filtered, the residue of the filtration was collected and resuspended in CHCl3. This suspension was filtered again, and the filtrates were collected and concentrated in vacuo. The oily remains were supended in MeOH (10 ml) from which the product precipitated as a white powder that was collected on a filter and dried under vacuum. Yield 150 mg (0.18 mmol, 50 %); mp.115-116˚; 1H- NMR (400 MHz, CDCl3): δ 6.86 (s, 4H), δ 5.45 (br, 2H), δ 3.96 (t, 3J = 6.6 Hz, 4H), δ 3.21 (q, J = 6.8 Hz, 4H), δ 2.12 (t, 3J = 7.1 Hz, 4H), δ 1.77 (m, 4H), δ 1.59 (m, 4H), δ 1.44 (m, 4H), δ 1.23 (m, 72H), δ 0.84 (t, 3J = 6.6 Hz, 6H). 13C- NMR (400 MHz, CDCl3): δ 173.0, 149.2, 120.9, 114.0, 69.2, 39.5, 36.9, 31.9, 29.7, 29.6, 29.6, 29.6, 29.5, 29.5, 29.4, 29.4, 29.3, 26.9, 26.1, 25.9, 22.7, 14.1. EI-MS: m/z 812, 586, 474. HRMS: calcd for C52H96N2O4 812,737, found 812.732.

11-(2-Hydroxy-phenoxy)undecanoic acid dodecyl amide (5.6): Potassium carbonate (0.2 g, 1.4 mmol) was added to a solution of catechol (0.13 g, 1.2 mmol) in acetone (40 ml) under an atmosphere of N2, After refluxing the suspension for 10 min, 11-bromo-undecanoic acid dodecylamide 5.4 (0.5 g, 1.1

mmol) was added and refluxing was continued for 48 h. The salts were removed

BrC11H22CONHC12H25

OC10H20CONHC12H25

OC10H20CONHC12H25

OC10H20CONHC12H25

OH



Turnmimics

126

by filtration and the filtrate was concentrated vacuum. The oily residue was purified by column chromatography (silica, eluent pentane-EtOAc 50/50) to give the product as a colorless oil. Yield 300 mg (0.65 mmol, 55%); 1H- NMR (400 MHz, CDCl3): δ 6.76 - 6.94 (m, 4H), δ 5.69 (s, 1H), δ 5.42 (br, 1H), δ 4.03 (t, 3J = 6.6 Hz, 2H), δ 3.24 (q, J=6.8 Hz, 2H), δ 2.16 (t, 3J = 7.1 Hz 2H), δ 1.80 (t, 3J = 7.6 Hz, 2H), δ 1.45 (m, 2H) δ 1.25 (m, 36H), δ 0.84 (t, 3J = 6.6 Hz, 3H). 13C- NMR (400 MHz, CDCl3): δ 172.8, 145.5, 120.9, 119.7, 114.1, 111.3, 68.5, 39.2, 36.5, 31.6, 29.3, 29.3, 29.2, 29.2, 29.1, 29.0, 28.9, 28.9, 28.8, 26.6, 25.7, 25.5, 22.4, 13.8.

12-[2-(10-Dodecylcarbamoyl-decyloxy)-phenoxy]-dodecanoic acid dodecylamide (5.2): Potassium carbonate (0.1 g, 0.7 mmol) was added to a solution of 5.6 (80 mg, 0.27 mmol) in acetone (30 ml) under an atmosphere of N2, After refluxing the suspension

for 10 min, the 12-bromo-dodecanoic acid dodecylamine 5 (0.12 g, 0.27 mmol) was added and refluxing was continued for 5 days. The salts were removed by filtration, and washed with CHCl3. The filtrates were combined, concentrated in vacuum to give a yellow oil. Upon stirring of this oil in MeOH (10 ml), the product precipitated as white powder that was collected on a filter and dried under vacuum. Yield 100 mg (0.12 mmol, 67%). mp. 114-115˚; 1H- NMR (400 MHz, CDCl3):

δ 6.83

(s, 4H), δ 5.46 (br, 2H), δ 3.93 (t, 3J = 6.6 Hz, 4H), δ 3.18 (q, J = 6.8 Hz, 4H), δ 2.09 (t, 3J = 7.1 Hz, 4H), δ 1.75 (m, 4H), δ 1.52 (m, 4H), δ 1.40 (m, 4H), δ 1.20 (m, 72H), δ 0.82 (t, 3J = 6.6 Hz, 6H). 13C- NMR (400 MHz, CDCl3): δ 173.5, 149.7, 121.4, 118.9, 114.6, 69.7, 39.9, 37.3, 32.3, 30.1, 30.0, 30.0, 29.9, 29.9, 29.9, 29.8, 29.7, 27.4, 26.5, 26.3, 25.9, 23.1, 14.5. EI-MS: m/z 826, 600, 474. No peak at m/z 812. HRMS: calcd for C54H98N2O4 826,753, found 826.748.

7-(2-Hydroxy-phenoxy)-heptanoic acid methyl ester (7): This compound was prepared as described above for 5.6, starting from acetic acid 2-hydroxy-phenylester (= mono acetyl catechol)33 (1.2 g, 8 mmol), potassium carbonate (1.4 g, 10 mmol), and 7-bromoheptanoic acid

methylester34 (1.8 g, 8 mmol). The crude product was purified by column chromatography (silica, eluting with 1 % MeOH in CHCl3) to give a yellow oil that solidified on standing. Yield 1.1 g (4.3 mmol, 54 %); 1H- NMR (400 MHz, CDCl3): δ 6.75 - 6.85 (m, 4H), δ 5.77 (s, 1H), δ 3.95 (t, 3J = 6.6 Hz, 2H), δ 3.61 (s, 3H), δ 2.26 (t, 3J = 7.3 Hz, 2H), δ 1.75 (t, 3J = 7.0 Hz , 2H), δ 1.60 (t, 3J = 7.1 Hz, 2H), δ 1.43 (m, 2H), δ 1.32 (m, 2H); 13C- NMR (400 MHz, CDCl3): δ 175.0, 143.7, 118.5, 117.1, 112.4, 109.9, 66.1, 48.8, 31.2, 26.4, 26.1, 23.0, 22.1; EI-MS: m/z 252, 220, 110.

OC10H20CONHC12H25

OC11H22CONHC12H25

OC6H12CO2Me

OH



Chapter 5

127

7-[2-(3-Methoxycarbonyl-propoxy-phenoxy]-hepta noic acid methyl ester (5.8). This compound was prepared as described above for 5.2, starting from 5.7, (1.0 g, 3.9 mmol), potassium carbonate (1.4 g, 10 mmol), and 3-bromobutyric acid methylester35

(0.9 g, 5 mmol), to give an oil, which was used in the next step without further

purification. Yield 0.7 g (2 mmol, 52 %); 1H-NMR (400 MHz, CDCl3): δ 6.86 (s, 4H), δ 4.01 (t, 3J = 6.2 Hz, 2H), , δ 3.95 (t, 3J = 6.6 Hz, 2H), δ 3.66 (s, 3H), δ 3.65 (s, 3H), δ 2.54 (t, 3J = 7.3 Hz, 2H), δ 2.30 (t, 3J = 7.0 Hz , 2H), δ 2.12 (t, 3J = 7.1 Hz, 2H), δ 1.79 (t,3J = 7.0 Hz, 2H), δ 1.64 (m, 2H), δ 1.47 (m, 2H), δ 1.38 (m, 2H); 13C- NMR (400 MHz, CDCl3): δ 173.4, 172.9, 148.6, 148.2, 120.8, 120.5, 113.9, 113.3, 68.3, 67.4, 50.8, 50.7, 33.3, 29.8, 28.5, 28.2, 25.1, 24.2, 24.1. HRMS: calcd for C19H28O6 352.189, found 352.191.

7-[2-(3-Dodecylcarbamoyl-propoxy)-phenoxy]-hepta noic acid dodecylamide (5.3): Compound 5.8 (0.35 g, 1 mmol) was saponified by dissolution in NaOH (2.5 M) in water/ethanol 1/1, followed by refluxing for 2h. The mixture was left to stand for 12 h, poured into NaCl (aq, sat. 50 ml) and washed with EtOAc (50 mL). The

aqueous layer was acidified to pH 2 and extracted again with EtOAc (2 x 50 ml) The organic layer was dried over Na2SO4, concentrated in vacuum to give a dark oil. 1H NMR indicated that saponification of the ester groups was complete. The crude product was dissolved in CH2Cl2 (10 ml) and thionylchloride (300 µl, 500 mg) was added. The reaction mixture was refluxed for 2 h, and concentrated in vacuum to remove excess thionylchloride. The oily residue was dissolved in CH2Cl2 (10 ml) and dodecylamine (600 mg, 3.2 mmol) and Et3N (350 mg, 3,5 mmol) were added. The reaction mixture was refluxed 12 h and concentrated in vacuum to give the crude product that was purified by column chromatography (silica, eluting with 2% MeOH in CHCl3). Yield: 0.2 g (0.3 mmol, 30 % over 3 steps) of a white solid; mp. 94-96˚; 1H-NMR (400 MHz, CDCl3): δ 6.85 (s, 4H), δ 6.02 (br, 1H), δ 5.64 (br, 1H), δ 3.99 (t, 3J = 6.2 Hz, 2H), δ 3.95 (t, 3J = 6.6 Hz, 2H), δ 3.16 (m, 4H), δ 2.38 (t, 3J = 7.3 Hz, 2H), δ 2.12 (m, 4H), δ 1.76 (t, 3J = 7.0 Hz , 2H), δ 1.62 (t, 3J = 7.7 Hz, 2H), δ 1.20-1.47 (m, 44H), δ 0.83 (t, 3J = 7.0 Hz, 6H); 13C-NMR (400 MHz, CDCl3): δ 173.2, 172.8, 149.3, 148.9, 121.7, 121.4, 114.7, 114.0, 69.1, 68.4, 39.9, 39.8, 36.8, 33.3, 32.2, 29.9, 29.9, 29.9, 29.9,, 29.8, 29.6, 29.5, 29.4, 29.2, 27.3, 26.1, 25.9, 25.7, 23.0, 14.4. EI-MS: m/z 658, 405, 254. HRMS: calcd for C41H74N2O4

658.565, found 658.566.

Molecular Modeling: Molecular modeling calculations were carried out using the compass forcefield, as implemented in Materials Studio31, a product of Accelrys, San Diego, CA, USA. The energy minimizations were carried out in the gas phase

OC6H12CO2Me

OC3H6CO2Me

OC6H12CONHC12H25

OC3H6CONHC12H25



Turnmimics

128

with a dielectric constant of 1. All energy-terms were included with the exception of an explicit hydrogen-bonding term. For the non-bonding interactions a cut-off radius of 12.5 Ǻ was used, with a spline width of 3 Ǻ, and a buffer width of 1.0 Ǻ. A graphite sheet, 20 x 30 atoms in size, with fixed cartesian position for the carbon atoms was used as the substrate. All structures were subjected to energy minimization using the Fletcher-Reeves algorithm, to a final gradient with maximum derivative of 0.001 kcal/mol. The folding abilities for the derivatives were expressed by comparison of the total potential energy for the energy-minimized 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. No direct comparison between derivatives was made.

5.13 References

1 (a) Kim, J., Swager, T.M., Nature, 2001, 411, 1030-1034 (b) Joachim, C., Gimzewski, J.K., Aviram, A. Nature, 2000, 408, 541 (c) Watson, M.D., Jäckel, E., Severin, N., Rabe, J.P., Müllen, K. J. Am. Chem. Soc. 2004, 126, 1402 (d) Miura, A., Chen, Z.J., Uji-I, H., De Feyter, S., Zdanowska, M., Jonkheijm, P., Schenning, A.P.H.J., Meijer, E.W., Wurthner, F., De Schryver, F.C. J. Am. Chem. Soc. 2003, 125, 1496 (e) Hla, S.W., Rieder, K.H. Annu. Rev. Phys. Chem. 2003, 54, 307, and references therein. 2 (a) Gesquière, A., De Feyter, S., De Schryver, F.C., Schoonbeek, F., van Esch, J., Kellogg, R.M., Feringa, B.L. Nano Lett. 2001, 1, 201 (b) De Feyter, S., Larsson, M., Schuurmans, N., Verkuijl, B., Zoriniants, G., Gesquiere, A., Abdel-Mottaleb, M.M., van Esch, J., Feringa, B.L., van Stam, J., De Schryver, F.C. Chem. Eur. J. 2003, 9, 1198 – 1206 (c) De Feyter, S., Abdel-Mottaleb, M.M., Schuurmans, N., Verkuijl, B.J.V., van Esch, J.H., Feringa, B.L., De Schryver, F.C. Chem. Eur. J. 2004, 10, 1124 – 1132. 3 (a) Uji-i, H., Yoshidome, M., Hobley, J., Hatanaka, K., Fukumura, H. Phys. Chem. Chem. Phys. 2003, 5, 4231 (b) Eichhorst-Gerner, K., Stabel, A., Moessner, G., Declercq, D., Valiyaveettil, S., Enkelmann, V., Müllen, K., Rabe, J.P. Angew. Chem. Int. Ed. Engl. 1996, 35, 1492 (c) Wintgens, D., Yablon, D.G., Flynn, G.W. J. Phys. Chem. B 2003, 107, 173. 4 (a) Saven, J.G. Chem. Rev. 2001, 101, 3113 – 3130 (b) Kooning, E.V., Wolf, Y.I., Karev, G.P. Nature, 2002, 420, 218-223. 5 Fersht, A. Structure and Mechanism in Protein Science: A Guide to Enzyme Catalysis and Foldin. Freeman, New York, 1999. 6 Selkoe, D.J. Nature, 2003, 426, 900 –904. 7 Stryer, L. Biochemistry, Freeman, New York, 1988. 8 (a) Socolich, M., Lockless, S.W., Russ, W.P., Lee, H., Gardner, K.H., Ranganathan, R. Nature, 2005, 437, 7058, 512-518 (b) Deechongkit, S., Nguyen, H., Powers, E.T., Dawson, P.E., Gruebele, M., Kelly, J.W. Nature, 2004 430, 6995, 101-105 (c) Dobson, C.M. Nature, 2003, 426, 884-890.



Chapter 5

129

9 Voyer, N. The Development of Peptide Nanostructures, Vol. 184, Springer Berlin, Heidelberg, 1996. 10 Hill, D.J., Moi, M.J., Prince, R.B., Hughes, T.S., Moore, J.S. Chem. Rev. 2001, 101, 3893-4011. 11 Stone, M.T., Fox, J.M., Moore, J.S. Org.Lett. 2004, 6, 3317 – 3320. 12 Nelson, J.C., Saven, J.G., Moore, J.S., Wolynes, P.G. Science, 1997, 277, 1793 – 1796. 13 Prince, R.B., Brunsveld, L., Meijer, E.W., Moore, J.S. Angew. Chem. Int Ed. Engl. 2000, 39, 228 – 230. 14 Prince, R.B., Barnes, S.A., Moore, J.S. J. Am. Chem. Soc. 2000, 122, 2758-2762. 15 Gabriel, G.J., Sorey, S., Iverson, B.L. J. Am. Chem. Soc. 2005, 127, 2637-2640. 16 Hill, D.J., Moi, M.J., Prince, R.B., Hughes, T.S., Moore, J.S. Chem. Rev. 2001, 101, 3893-4011. 17 Licini, G., Prins, L.J., Scrimin, P. Eur. J. Org. .Chem. 2005, 70, 969 – 977. 18 Saluan, A., Potel, M., Roisnel, T., Gall, P., Le Grel, P. J. Org. Chem. 2005, 70, 6499 – 6502. 19 Violette, A., Averlant-Peteit, M.C., Semetey, V., Hemmerlin, C., Casimir, R., Graff, R., Marraud, M., Briand, J-P., Rognan, D., Guichard, G. J. Am. Chem. Soc. 2005, 127, 2156 – 2164. 20 Sibanda, B.L., Thornton, J.M. Nature, 1985, 316, 170 – 174. 21 Hahn, S., Kim, S.S., Lee, C., Cho, M. J. Chem. Phys. 2005, 123, 084905. 22 Minor, D.L., Kim, P.S. Nature, 1994, 367, 660 – 663. 23 e.g. psipred. http://bioinf.cs.ucl.ac.uk/psipred. 24 Pochan, D.J., Schneider, J.P., Kretsinger, J., Ozbas, B., Rajagopal, K., Haines, L. J. Am. Chem. Soc. 2003, 125, 11802-11803. 25 Stanger, H., Gellman, S.H. J. Am. Chem. Soc. 1998, 120, 4236 – 4237. 26 Rapaport, H., Moller, G., Knobler, C.M., Jensen, T.R., Kjaer, K., Leiserowitz, L., Tirrell, D.A. J. Am. Chem. Soc. 2002, 124, 9342 – 9343. 27 Atkins, E.D.T., Hill, M.J., Jones, N.A., Sikorski, P. J. Mater. Sci. 2000, 35, 5179 – 5186. 28 Baas, T., Gamble, L., Hauch, K.D., Castner, D.G., Sasaki, T. Langmuir, 2002, 18, 4898 – 4902. 29 A urea group forms two hydrogen bonds per moiety, and an amide group only one. The strength of hydrogen bonds, however, is very much dependent on environment and geometry. (a) Gavezzotti, A. Acc. Chem. Res. 1994, 27, 309 - 314. (b) Vachal, P., Jacobsen, E.N. J. Am. Chem. Soc. 2002, 124, 10012-10014. 30 Calculations were performed using Discover. This program uses 3 minimization algorithms: steepest descent, conjugate gradient and Newton-Raphson. For a general



Turnmimics

130

discussion, see Fletcher, R. (ed.), Optimization, Academic Press, New York and London, 1969. 31 www.accelrys.com. 32 Discussed in chapter 2. 33 Olcott, H.S. J. Am. Chem. Soc, 1937, 59, 392. 34 This compound was synthesized from the corresponding nitrile according to a literature procedure: Woolford, R.G. Can. J. Chem. 1962, 40, 1846-1850. 35 Miyaoka, H., Tamura, M., Yamada, Y. Tetrahedron, 2000, 56, 41, 8083 – 8094.



131

Chapter 6 Minimal foldamers in 2-D

In the previous chapter the successful design and synthesis of a turnmimic has been described. Such a turnmimic is a first prerequisite for the construction of a real 2-D foldamer. In the next step the connection of several single-turnmimics should be realized. This step however is not trivial. It will be much more difficult to control the molecular structure at several points at the same time. Also the synthetic routes towards the desired molecules will be longer.

6.1 Introduction

The aim of the work described in this chapter is to arrive at a (small) oligomer incorporating multiple turn-mimics of the kind described in chapter 5. Intramolecular H-bonding should direct the conformation of such a molecule into a folded ribbon (fig. 6.1). Connecting multiple turnmimics asks for a modified design though, because one would like to employ a sequential coupling scheme for the various components. Design and synthesis of this new turnmimic will be discussed.

O OAA O O A

O

O

A O

A

OA

O OAA O O A

O

O

A O

A

OA

Figure 6.1 Formation of a folded ribbon from an oligomer featuring 2 turn mimics.



Minimal Foldamers

132

The behavior of a model compound with respect to monolayer formation is shown to be markedly different from that described in chapter 5. Possible reasons will be discussed. Finally, though superfluous with the results obtained, some possibilities (design and synthesis) for actual oligomers are considered. Several turn mimics can be connected directly or with spacers. This leads to interesting variations.

6.2 Towards a new turn mimic

With the results established so far one should consider the next step: connecting multiple turn elements to synthesize a true minimal foldamer. However successful the turn elements described in the previous chapter may be, from a synthetic point of view it is clear that they are not the most ideal components for the synthesis of larger foldamers.

Because both the amide groups in compound 6.1 have the same side (the carbonyl) closer to the catechol, an extension strategy would necessarily involve two building blocks: one containing the acids, and one containing the amines. An orthogonal protection-deprotection protocol has to be invoked, as only one of the acids is supposed to react with only one of the amines every coupling step. Although this is not impossible, it would seem to be cumbersome. Instead a coupling scheme involving only one building block, with a (protected) acid and a (protected) amine in the same molecule is much more attractive. Such a procedure bears resemblance to conventional peptide chemistry. In a way the proposed foldamers can be considered as peptidomimetics.

Synthesis of peptidomimetics has been reported, some of which also incorporate natural or non-natural turn fragments.1,2 Recent work involved e.g. the use of carbohydrate –derived scaffolds of γ − and δ-amino acids,3 azabicycloalkane amino acids as dipeptide mimics,4 and studies on the properties of β-amino acids (which form a range of secondary structures).5 The importance of β-turns in the design of various peptidomimetics for many diseases has been reviewed. 6

Figure 6.2 Compound 6.1. A turn mimic with the amide groups oppositely oriented to allow for easy synthetic elaboration. Mind the conformational equilibrium.

O

O NH

NH

O

O

O

NH

O

NH

O

O



Chapter 6

133

6.2.1. Design and synthesis

In order to be able to follow the procedures of peptide chemistry, a new turn mimic (compound 6.1, fig. 6.2) is required. Compound 6.1 is an α,ω-amino acid. Molecular modeling established that conformations allowing for intramolecular H-bond formation are favored. In fact, because the amide groups are non-equivalent (they have a reverse orientation), dissymmetry with respect to the spacers is no longer necessary. Both spacers can be of the same length. No odd-even effects were found, and in every case a conformational equilibrium exists (fig. 6.2). A model with C3 spacers will serve as a representative example.


OH

OH

O

OH

Cl O

OH

N3

O

OH

NH2O

OH

NHZ

O NHZ

O CO2Me

O NHZ

O CO2H

O NHZ

O CONHC12H25

O NHCOC13H27

O CONHC12H25

BrC3H6Cl

K2CO3

acetone

NaN3

DMSO

H2

Pd-C

O Cl

O

KOHTHF - H2O

BrC3H6CO2Me

K2CO3

acetone

LiOH

MeOH

C12H25NH2

EDC/HOBTHunig base

DMF

2. C13H27CO2H

EDC, HOBT, iPr2NEt, DMF

27 % 6.5 6.6

6.7 6.8

6.96.10

6.116.1

53 %

65%46 %

20 %

82 %

35 %

1. H2 Pd-C

30%



Minimal Foldamers

134

The model compound 6.1, with C3 spacers and C12 tails has been synthesized according to scheme 6.3. 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 6.5. It was eventually obtained in 27% yield. Compound 6.5 was easily converted to the azide 6.6, in modest yield of 53%. Subsequently this compound was converted to the amine (6.7, via catalytic hydrogenation, 65% yield). Protection of this amine with benzyl chloroformate was cumbersome and required extensive column chromatography again. The Z-protected amine 6.8 was eventually obtained in 44% yield. Compound 6.8 was coupled to methyl bromobutyrate as before, appending the second arm, in 20% yield. The ester 6.9 was now deprotected by saponification (82%). The Z-protected mono-acid 6.10 was crystallized. A more advanced protocol for generating the amides was used as before.7 The mono-acid was activated with EDC and HOBT, and then reacted with dodecylamine. This reaction generated compound 6.11 in 35% yield. The Z-protected amine 6.11 was now deprotected by catalytic hydrogenation and subsequently reacted with myristic acid, preactivated with EDC/HOBT. This reaction produced compound 6.1 in 30% yield. All compounds were characterized by 1H-NMR, 13C-NMR and MS.

6.3 Monolayers of compound 6.1

The new model compound 6.1 was tested for monolayer formation. The results were quite unexpected. Compound 6.1 did form monolayers on the basal plane of HOPG. The images are submolecularly resolved, as shown in fig. 6.3. Contrary to what had been assumed, compound 6.1 did not adopt a folded conformation at the interface. Two structures were observed. One structure seems to have a head-to-head type interaction, and the distance between the bright spots is 0.485 nm (fig. 6.3 top). The distance to the end of the lamella is 2.1 nm. Although it was not possible to do a line-profile analysis, it looks like only one chain is adsorbed on the surface. This is not unlike the situation encountered for compound 5.1. Sometimes another structure (polymorph) was observed (fig. 6.3 bottom). This structure seems to have only one bright spot per lamella, which probably indicates that it actually adopts an extended conformation. This orientation has not been previously encountered. The tails on the two sides of the bright spot are non-equivalent in length (2.1 nm and 2.4 nm) respectively corresponding to the different tails (C12 and C13) in the adsorbents. The total width of the lamellae is 4.45 nm. The distance between the bright spots is 0.54 nm.



Chapter 6

135

6.4 Discussion

Because this result was unexpected, the modeling was reevaluated. Compound 6.1 was compared to compound 5.3. Arrays of 5 x 2 adsorbents were generated in Materials Studio (fig. 6.4) in both the folded and the extended conformation for these 2 compounds. Compound 5.3 convincingly adopted a folded conformation in the monolayers, whereas compound 6.1 adopted (sometimes) an extended conformation. The results of molecular modeling are not in correspondence with the STM results. According to the models, folded conformations should be more favorable for both compound 5.3 and compound 6.1 in the monolayers. Actually,

Figure 6.3 STM images of compound 6.1 on HOPG. The compound gives rise to two different lamellar structures (above and below). I = 0.6 nA, U = -0.422 V (upper right). I =

0.6 nA, U = -0.714 V (upper left). I = 0.6 nA, U = -0.400 V (lower images).



Minimal Foldamers

136

the relative energy gain for compound 6.1 is much larger than for compound 5.3 (table 6.1). Obviously the conformational equilibrium has not been taken into account in these calculations, and as a matter of fact all the dynamics have been omitted. It is believed that these play a very important role in the process. These results also clearly establish the limitations of molecular modeling in this work.

∆∆∆∆E (folded-extended) for the arrays.

6.1 -265 kcal/mol

5.3 -36 kcal/mol

Table 6.1 Energy difference for folded and extended conformers of compounds 6.1 and 5.3 in a 5 x 2 aggregate. The quoted differences relate to the 6 molecules in the middle of each aggregate.

O

O

NH

NH

O

OO

O

NH

NH

O

O

6.1 5.3

Figure 6.4 Calculated conformations for the folded and extended conformations of compound 6.1 (right) and compound 5.3 (left).



Chapter 6

137

6.5 Towards 2-D foldamers

It is relevant to have a look at the kind of structures that can be formed, using this approach, because with small adaptations in the design it might still be feasible to arrive at 2-D foldamers. The considerations below, as well as the synthetic plan can still apply. Furthermore it is interesting to see whether derivatives incorporating multiple turn elements behave similarly as model compound 6.1.

6.5.1 Role of the turn element

The molecules are thought to be in a random coil state in solution, and aided by the graphite template they will adopt a specific conformation, as directed by the turn elements, the amide groups, and the spacers. It is important that the H-bonding units are in registry. Because of the 2-dimensional organization, odd-even effects in the spacers are playing an important role. For the model systems it was shown that these effects are operational in determining the number of conformations that allow intramolecularly H-bonded structures. Now they are also playing a role in determining how individual fragments are combined. Reviewing the situation, one can identify a number of variable parameters:

• position of the H-bonding moieties.

• directionality of the H-bonding moieties.

• the number of carbon atoms in the spacers (odd-even effects).

These parameters are all related.

The substituents have a divergent relation because of the connectivity to the aromatic ring. Whether the amide groups are pointing up or down is (in 2-D) governed by the number of carbon atoms in the spacer, on behalf of the zig-zag conformation. Because of the symmetry of the amide group, the mode of construction is important. Coupling an amine fragment to a carboxylic acid, appended to the catechol, yields a different situation than the process the other way around. The latter is rotated by 180º relative to the former in a 2-D environment. This implies that the direction of the carbonyl group is opposite in the respective cases, given an equal number of carbon atoms in the spacers. Extending the spacer by one C atom has the same effect as reversing the mode of construction. The coupling of multiple fragments introduces new parameters that can be varied:

• the number of H-bonding moieties along a strand.

• complementarity effects (shape and size).

• chirality.



Minimal Foldamers

138

6.5.2 Number of H-bonding arrays

Introducing multiple H-bonding moieties will improve the stability of the structures. Some possibilities are represented in figs. 6.5 – 6.7. In fig. 6.5 a conceptual model with one H-bonding array is displayed. Such a system can be constructed by directly coupling two turn elements. These molecules are designated as class 1 in this chapter. Whether the amide groups will be pointing up or down is in principle governed by the parameters discussed before. This system has three principle axes of rotation and reflection. Because there is no facial preference in the assembly of such a molecule (either side is equally likely to adsorb onto a graphite surface) directionality of the H-bonding array is not fixed. Rotation by 180º around the x-axis or the z-axis will reverse the directionality of the array, whereas rotation around the y-axis leaves it the same. Desymmetrization with respect to axis y-axis, can either be achieved by coupling fragments of different size, i.e. with longer and shorter spacers respectively, or by introduction of tetrahedral chirality. In the former case it should be noted that coupling two different spacers should yield a total number of C-atoms, which is even. Otherwise the direction of the turn is changed, and the next strand folds back into the foldamer. Introduction of chirality also takes care of desymmetrization with respect to the x-axis. Desymmetrization with respect to the z-axis can for instance be achieved by introducing turn elements of a different nature, e.g. based on another scaffold than catechol.

This is a relevant analysis because desymmetrization reduces the number of packing modes available for the molecules in a plane, which allows more control over the aggregate that will eventually be formed. Therefore it is in principle possible to program the directionality of the H-bonding array, given a smart choice of the spacers.

Another approach might be to connect two turn elements by means of a linear spacer (fig. 6.6). In such foldamers there are two H-bonding arrays. These

Figure 6.5 Rotational symmetry operations in class 1 foldamers. Z-axis is normal to the plane of the drawing.

C2 (y)

C2 (x)

C2(z)



Chapter 6

139

molecules are designated as class 2 foldamers in this chapter. Most of the parameters discussed above also apply here. There is a greater deal of control however because the relation between the directionalities of the H-bonding arrays can be synthetically programmed, independent of symmetry operations to the entire molecule. Again odd-even effects determine whether an up-down relation or an up-up (= down-down) relation exists. Also the total number of C-atoms in all the spacers (even or odd) determines what direction the turns make relative to each other. It should be stressed that the suggested programmability only holds in the case of unequivocal relation between directionality and position of the amide groups. Enthalpic conformational degeneracy, as encountered in the turn-mimic discussed in paragraph 6.2, thwarts this element of control.

In this model, along a lamella in the aggregate, strands of different nature alternate. They are either covalently bound over the entire stretch, or composed of two tails, which should be designed such that all the empty space can be filled (complementarity). Another element of supramolecular control is added therefore.

From the point of view of monolayer formation, it should be clear that introduction of a spacer, provided that it is of sufficient length, will help in the assembly process, as it introduces a moiety with affinity for graphite at a central location in the molecule.

A third alternative model for foldamers can combine features of the above two models. The first model has approximately the shape of a square tile. Apart from conformational interconversion this structure is quite rigid, whereas the second model has “loose ends” but adds more control over the conformation, by introducing more H-bonding moieties. Conceptually combining these parameters, one ends up with a structure like that shown in fig. 6.7. These molecules are designated class 3 in this chapter. The shape is approximately as a rectangle; two H-bonding arrays can be formed. In every strand a spacer is introduced. This should also provide a driving force for the folding process, as well as the interaction with a graphite surface. From a synthetic point of view the number of coupling steps is increased in going from the first to the second to the third model.

===

Figure 6.6 Class 2 foldamers.



Minimal Foldamers

140

Finally, the sequences have been shown for two turn elements in each model, but this number can be extended indefinitely (given endless time and material). It is also possible to combine features of different models in one molecule. Together with the combination of several spacers, differing in length, in various parts of the

foldamers, more complex shapes can be constructed, which in theory allows for a true molecular jigsaw puzzle (fig. 6.7 right). All in all, a huge variety of structures can be envisaged.

Synthesis and monolayer formation of class 2 foldamers as well as a synthetic route towards class 3 foldamers will now be discussed.

6.6 Foldamers of class 2

In this family of molecules the turn elements are connected by a linear spacer. This characteristic provides a good opportunity to introduce a long alkyl fragment, that can drive the formation of monolayers. Like before, the turn elements will have C3 spacers. Because the total number of C-atoms has to be even, the linear spacer should also have an even number of C-atoms. It will have to be a bifunctional reactant, featuring an amino group, and a carboxylic acid group. For reasons of commercial availability, a choice was made for 11-amino undecanoic acid. Because the carboxylic acid group is counted as a functionality, and the carbonyl part will end up as an amide, this molecule can be regarded to have a C10 (10 = even) spacer between the functionalities. In the proposed foldamer no direct interaction between the turn fragments exists. Therefore the conformational interconversion that was found to exist for the mimic also exists here (in duplicate).

Figure 6.7 Class 3 foldamers (left). Foldamers as a molecular jigsaw puzzle (right).



Chapter 6

141

Molecular modeling, however, showed a (small, ~1.0 kcal/mol)) preference for the conceptual conjoint of one up and one down direction of the amide groups (fig.6.8 e). Recapitulate that in the mimic either one of the two substituents can adopt a gauche conformation relative to the aromatic ring, in order to achieve an intramolecular spacing of 5 Å, and that this process inverts the directionality of the amide groups. In fig. 6.8 the combination of the two conformations cannot be judged from the relative directionality of the amide groups, which is always parallel, as dictated by the spacer, but rather from the combination of the bends. In fig. 6.8e two gauche-bends are connected, indicating two different conformations. It is this conformation which has the lowest potential energy. Only one derivative has been synthesized (6.3, fig. 6.8a). The compound was synthesized according to scheme 6.2.

The alkyl spacer (commercially available 11-amino undecanoic acid) was Z-protected with benzylchloroformate to give intermediate 6.15 in 31 % yield. Onto the turn element 6.10 was appended a short alkyl tail by means of the EDC/HOBT mediated coupling, to give intermediate 6.13 in 45% yield. After deprotection, this

O

O

NH

O

NH

O

NH

OO

O

NH

O

a) c)

d)

e)

b)

Figure 6.8 (a) Compound 6.3 (b) Proposed aggregation. (c)-(e) Possible conformations of 6.3, optimized with materials studio.

6.3



Minimal Foldamers

142

HO2CC10H20NH2 HO2CC10H20NHZ

O NHZ

O CO2Me

O NHZ

O CONHC4H9

HO2CC10H20NHZ

O NHCOC10H20NHZ

O CONHC4H9

OC3H6NHCOC10H20NHCOC3H6O

OC3H6CONHC4H9

6.15

benzyl chloroformate

KOHTHF - H2O

31%

6.10

1. H2 Pd-C

2. C4H9NH2

EDC, HOBT

iPr2NEt

DMF

45%6.13

1. H2 Pd-C

2.


6.16

1. H2 Pd-C

2. compound 10 (deprotected)

EDC, HOBT, iPr2NEt, DMFZHNC3H6O

6.17

70%

50 %

6.3

1. H2 Pd-C

2. C3H6CO2H





Chapter 6

143

compound was coupled with compond 6.15 by means of EDC/HOBT. This reaction produced intermediate 6.16 in 70% yield. Compound 6.16 was deprotected, and coupled to another batch of the turn fragment 6.10 in 50% yield. This gave an intermediate with 2 turns: 6.17, in 50% yield. Compound 6.17 was deprotected again and endcapped with butyric acid to give the foldamer 6.3. Compound 6.3 was tested for its ability to form monolayers.

Compound 6.3 does form organized structures on graphite. The presumed positive contribution of the alkyl group directed the formation of a periodic structure at the graphite interface (fig. 6.9). The resolution of the image however is not enough to say anything about the conformation of individual molecues. It would seem that the bright lines, protruding from the surface, correspond to the catechol units. The distance between the bright structures corroborates this view. The bars in fig. 6.9 approximately correspond to 4 nm. This is the calculated distance of the long axis in compound 6.3. In between the bright structures can sometimes a substructure be seen (cirkel, and inset). This might indicate the alkyl spacers.

A better resolution is desirable, and it was eventually obtained (fig. 6.10). The image shown on the left side of fig.6.10 gives a large-scale view of the monolayer. The full surface coverage is apparent. A lamellar structure can be seen, with alternating bright rows and darker troughs. The small-scale magnification (fig. 6.10 right) shows the organization of the alkyl spacers. There seems to be an alternating contrast modulation (bright, well resolved chains, followed by darker

Figure 6.9 Formation of organized structures by compound 6.3.

Scale bar = 4 nm



Minimal Foldamers

144

chains) in traversing a lamella. The resolution of the aromatic part (the turn elements) is not very good. This can indicate that the turn elements are not flat, on the surface, but bent away. Alternatively it can indicate a high degree of mobility. Probably the best interpretation is that the molecules do not adopt a flat folded conformation in the images in fig. 6.10. The lamellar width, corresponding to the length of the molecules is 3.5 nm. The molecule-to-molecule distance within a lamella is about 0.45 nm, corresponding to fragments kept in place on behalf of H-bonding. The distance between two neighboring lamellae amounts to 0.7 – 0.8 nm (fig.6.11a).

These data, together with the lower resolution of the turn elements, contribute to a packing model where the central spacers are kept in place by the amide groups but the turn fragments bend away from the surface. The short chains are probably not on the surface. Comparison of the two models (fig. 6.11 b and c) shows that interaction between the amide groups in the short tails is still possible, if these chains are not on the surface. Consequently the potential energy of both structures is about the same. The observed packing is more easily formed because a smaller part of the molecules is organized in the aggregate. Depending on the dynamics of the system, which is high, judging from the difficulty to image this system, it is possible that the originally proposed structure (fig 6.11 b) will form over time.

One of the main problems is the low contribution of the small tails to the interaction of the molecules with the graphite. Because it seems that H-bonding (which probably occurs) does not keep the turn elements flat on the surface, a larger bulk in all the strands will be required to drive the formation of a flat, and folded conformation on the surface.

Figure 6.10 Large area and inset of monolayer formed by compound 6.3. I = 1 nA; U = -0.504 V.



Chapter 6

145

6.7 Foldamers of class 3

Class 3 foldamers would be ideally suited for this task. Every strand in these molecules incorporates a spacer. Furthermore additional H-bonding interactions are introduced. The penalty for strands for not lying on the surface is much bigger now, as dangling in solution goes at the expense of more H-bonding. The end-fragments will be slightly longer than before to match the entire length of the central strands, including the catechol moiety. The analogy to peptide synthesis (sequential deprotection and coupling to a next protected fragment) should be apparent in this strategy. The use of amino acid-like fragments in these longer sequences fully pays of. To drive the analogy even further, eventually one might consider a solid-phase based synthetic protocol. Presumably the Z-protection will have to be replaced by an FMOC-protective methodology if such an approach would be followed.

The synthesis is summarized in schemes 6.3, 6.4 and 6.5. There are a lot of coupling steps, but the synthesis is highly convergent, which means that it can be achieved in a reasonable amount of time.

The alkyl spacer returns in all strands. Therefore it was anticipated that it would be the most economical to generate extended fragments, already incorporating both the spacer and the turn. These fragments would then be coupled together in the final step. The two fragments can both be obtained from the same starting materials. Synthesis of the outer strands (cf. fig. 6.7) is shown in scheme 6.3.

O N H

O N H

O O O

N H O

O H N O O N

H O

N H

O O O

N H O

O H N

O 3 . 4 - 3 . 6

n m

0 . 4 5 - 0 . 4 8 n m

0 . 7 2 - 0 . 7 7 n m

O N H

O N H

O O O

N H O

O H N

O O N H

O N H

O O O

N H O

O H N

O 3 . 4 - 3 . 6

n m

0 . 4 5 - 0 . 4 8 n m

a)

b) c)

Figure 6.11 (a) Proposed arrangement of 6.3 in the monolayers. (b) Optimized folded conformation on the surface (c) Optimized conformation bending away from the surface.



Minimal Foldamers

146

11-Amino undecanoic acid was protected as the methyl ester 6.18 on the acid side, in 95% yield, whereas another batch was protected on the amine side, with a Z-group. This reaction proceeded in 30% yield, and gave compound 6.15. These intermediates were extended on their free sites. Compound 6.18 was reacted with octanoic acid according to the EDC/HOBT protocol, to form the amide 6.19 in 50% yield; compound 6.15 was reacted with hexylamine to form the amide on the other side in 55% yield, yielding compound 6.20. Intermediates 6.19 and 6.20 were both deprotected thereupon according to the respective standard methodologies.

The same strategy was followed for the turn element. This is illustrated in scheme 6.4. The Z-ester 6.9 was split into two batches. One batch was deprotected on the acid side to give intermediate 6.10 in 82% yield; the other batch was deprotected on the amine side, giving intermediate 6.21, in 85% yield. The free acid (from compound 6.19) was coupled to intermediate 6.21 in 50% yield. This step afforded intermediate 6.22, which is the first fragment to be used in the final step.

After deprotection, the free amine (from compound 6.20) was coupled to free acid intermediate 6.10. This step afforded extended turn fragment 6.23, in 45% yield. Intermediate 6.23 was first Z-deprotected, and subsequently coupled to another equivalent of compound 6.15 to generate intermediate 6.24, in 35% yield. Compound 6.24 is the second fragment.

C10H20HO2C NH2C10H20MeO2C NH2 C10H20HO2C NHZMeOH

H+KOH

THF-H2O

C7H15CO2H

EDC, HOBT, iPr2NEt

DMF

C6H13NH2

EDC, HOBT, iPr2NEt

DMF

O

NHC6H13C8H16ZHN

MeO2C C18H16 NH

O

C7H15

6.18 6.15

6.19 6.20

benzylchloro formate

30%

95%

50% 55%

Scheme 6.3 Synthesis of intermediates 6.19 and 6.20.



Chapter 6

147

In the final step compounds 6.24 (after deprotection) and 6.22 (after deprotection) were coupled. Hereby was the synthesis of compound 6.4 completed in 50 % crude yield. The crude product was characterized with 1H-NMR, and 13C-NMR spectroscopy.

O NHCOC10H20CONHC6H13

O CO2Me

O NHZ

O CONHC10H20NHCOC7H15

6.22

O NHCOC10H20NHZ

O CONHC10H20NHCOC7H15

6.24

O NH2

O CO2Me

6.21

deprot. 6.19

EDC, HOBT, iPr2NEt

DMF

6.23

O NHZ

O CO2H

6.10

deprot. 6.20

EDC, HOBT, iPr2NEt

DMF

LiOHMeOH

C10H20HO2C NHZ

EDC, HOBT, iPr2NEt

DMF

50%

35 %

53%

Scheme 6.4 Synthesis of intermediates 6.22 and 6.24.

H2Pd/C

OC3H6CONHC10H20CONHC7H15

OC3H6NHCOC10H20NHCOC3H6O

H13C6HNOCH20C10HNOCC3O

6.4

6.24

6.22

EDC, HOBT, iPr2NEt

DMF

50%




Minimal Foldamers

148

Compound 6.4 was purified by preparative RP- HPLC (symmetry RP 18, CH3CN–H2O). Because of the low insolubility of these compounds in conventional HPLC solvents, they were dissolved in a mixture of acetic acid and acetonitrile, the pH of which was adjusted to 4.0 with triethylamine. The compounds were eluted with conventional acetonitrile – water mixtures, with gradients ranging from 50-50 to 80-20. Going higher in the organic phase would risk precipitation on the column.

2 mg of compound 6.4 was isolated. It was characterized with maldi-TOF. MALDI demonstrated large parent peaks (M+H+, M+Na+, M+K+) at 1261, 1283 and 1300. Impurities were still present though. Because purity could not be unequivocally established monolayer formation was not investigated.

6.8 Conclusions

An audacious goal was formulated. With a slight modification on the turn-mimic it seemed possible to synthesize a variety of oligomers, capable of folding at an interface. The unexpected behavior of a model compound featuring this new mimic, however, makes it unlikely that the postulated approach will lead to such structures. Using the α,ω-amino acid building blocks, as in compound 6.1, complicates the folding process, as a result of which the molecules aggregate in their extended conformation on HOPG. Molecular modeling though has indicated that the folded conformations should also be more stable for these derivatives. The folded conformation in compound 6.1 was shown to undergo conformational interconversion. This process is believed to play a role in the aggregation behavior. The exact mechanism of the monolayer formation, and why a particular conformation of the adsorbents is preferred, is not known. MD simulations might provide insight into this process. It is very well possible that small changes in the design will lead to the desired (folded) structures. New results seem to indicate that compound 6.1 indeed adopts a folded conformation at the gold-water interface. This would stress the importance of the environment for the monolayer formation of the kind of molecules described in this chapter.

The synthetic accessibility of homologues of compound 6.1 has been addressed. The homologues incorporate multiple turn-mimics, like compounds 6.3 and 6.4. The synthesis was at several stages found to be cumbersome. An improved synthetic route towards intermediate 6.10 would be desirable. The reaction with bromochloropropane does not seem to be an ideal method to mono-functionalize catechol. The EDC/HOBT coupling protocol, though effective, gives average yields of 50%; for longer sequences this might be too low.

Compound 6.3, the smallest double-turn member of the family, does form monolayers. Monolayer formation is driven by the long alkyl fragment between the turn-mimics. Again, the turn moiety does not behave as anticipated, i.e. it does not adopt the intramolecularly H-bonded conformation in the lateral plane, parallel to the surface.



Chapter 6

149


Methods and materials

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 spectrometer (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 VXR400 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 were added one equivalent of diisopropylethyl amine. The mixtures were stirred and were cooled to 0ºC. HOBT (hydroxybenzotriazole) and EDC (a water soluble version of DCC) were added in equimolar 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 to dryness in vacuo at 40ºC. CH2Cl2 and water were added. The organic phase was extracted 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 chromatography (silica, 2% MeOH in CHCl3). The amides were the first fractions to come of the column.

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 was evaporated. The amines were used without further purification.

2-(3-Chloro-propoxy)-phenol (6.5): Catechol (100 g, 0.9 mol) was dissolved in acetone (500 ml) and the solution was stirred at reflux. Pre-dried (120°C) 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%). Yellow oil. 1H-

O

OH

Cl



Minimal Foldamers

150

O

OH

N3

NMR (CDCl3): δ 6.80 - 6.87 (m, 4H), 5.53 (s, 1H) 4.16 (t, 3J = 11.6 Hz, 2H), 3.68 (t, 3J = 11.7 Hz, 2H), 2.23 (m, 2H). 13C-NMR (CDCl3): δ 121.8, 120.2, 114.7, 111.9, 65.5, 41.3, 32.0. MS (CI+): m/z 187. HRMS : calcd for C9H11ClO2 : 186.045, found 186.051.

2-(3-Azido-propoxy)-phenol (6.6): Compound 6.5 (44 g, 0.24 mol) was dissolved in DMSO (300ml). The mixture was stirred and NaN3 (20 g, 0.3 mol) was added in small portions. The mixture was stirred at 60 °C for 12h. The mixture was then poured into water (500 ml),

and extracted with EtOAc (2x). The organic layers were washed with water (2x), dried (Na2SO4) and concentrated in vacuo to yield the product as a yellow oil. It was purified by trituration with pet. ether. Yield: 25 g (0.13 mol, 53 %). 1H-NMR (CDCl3): δ 6.84 - 6.91 (m, 4H), 5.69 (s, 1H), 4.12 (t, 2H, 3J = 6.2 Hz), 3.50 (t, 3J = 6.6 Hz, 2H), 2.07 (m, 2H). 13C-NMR (CDCl3): δ 121.6, 120.3, 117.2, 111.7, 68.2, 44.0, 32.0. HRMS: cacd for C9H11N3O2

193.085, found 193.089.

2-(3-Amino-propoxy)-phenol (6.7): The azide 6.6 (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 12h 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): δ 6.81 - 6.92 (m, 4H) δ 5.81 (br, 2H) δ 4.01 (t, 3J = 5.1 Hz, 2H), 3.02 (m, 2H), 1.96 (m, 2H). 13C-NMR (CDCl3): δ 125.0, 120.3, 119.7, 117.7, 73.0, 39.8, 31.4.

[3-(2-Hydroxy-phenoxy)-propyl]-carbamic acid benzyl ester (6.8): The amine 6.7 (12 g, 72 mmol) was dissolved in THF - H2O (5:1, 200 ml). The mixture was cooled to 0° and KOH (30 g, 0.2 mol, excess) was

added. The mixture was stirred 5 min and benzyl chloroformate (12.5 ml, 15 g, 85 mmol, 1.2 eq.) was added dropwise. The mixture was stirred for 2h at 0° C, and allowed to stand overnight at room temperature. Water (100ml) was added, and the organic layer was separated. The aqueous layer was extracted with CH2Cl2 (3x). The combined 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): δ 7.32 (m, 5H), 6.73 – 6.85 (m, 4H), 5.43 (br, 1H), 5.02 (s, 2H), 3.96 (t, 3J = 5.8, 2H), 3.31 (t, 3J = 4.8Hz, 2H), 1.87 (t, 3J=6.2 Hz, 2H). 13C-NMR (CDCl3): δ 156.7, 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, 192. HRMS: calcd for C17H19NO4: 301.131, found 301.139.

O

OH

NH2

O

OH

NHCO2CH2C6H5



Chapter 6

151

4-[2-(3-Benzyloxycarbonylamino-propoxy)-phenoxy] -butyric acid methyl ester (6.9): The Z-protected amine 6.8 (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 portionwise. The mixture was stirred 30 min. at reflux 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 (3x : silica, CHCl3; EtOAc; CHCl3 + 5% MeOH + 1 % NH3). Yield: 2.4 g (6 mmol, 20%). 1H-NMR (CDCl3): δ 7.35 (m, 5H), 6.88 – 6.92 (m, 4H), 6.73 – 6.85 (m, 4H), 5.02 (s, 2H), 4.08 (t, 3J = 5.7 Hz, 2H), 3.97 (t, 3J = 5.8 Hz, 2H), 3.65 (s, 3H), 3.42 (m, 2H), 2.26 (t, 3J = 7.3 Hz, 2H ), 1.84 (m, 4H). 13C-NMR (CDCl3): δ 156.0, 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.

4-[2-(3-Benzyloxycarbonylamino-propoxy)-phenoxy]-butyric acid (6.10): The ester 6.9 (2.5 g, 6 mmol) was dissolved in MeOH (50 ml) and LiOH (200 mg) was added. The mixture was stirred for 2h at rt. The mixture

was concentrated 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°C. 1H-NMR confirmed removal of the methoxy group. MS (EI+): m/z 192. MS (EI+): m/z 387. HRMS: calcd for C21H25NO6 387.168, found 387.171.

4-[2-(3-Amino-propoxy)-phenoxy]-butyric acid methyl ester (6.21): The Z group was removed from diprotected compound 6.9 (2.5 g, 6.5 mmol) using the standard procedure. 1H-NMR indicated removal of the Cbz-group. Yield: 1.5 g (5.6 mmol, 85 %). White solid. MS (EI+): m/z 267, 101.

{3-[2-(3-Dodecylcarbamoyl-propoxy)-phenoxy]-propyl}-carbamic acid benzyl ester (6.11): Compound 6.10 (mono-Z acid, 200 mg, 0.5 mmol) was dissolved in

DMF (20 ml). The solution was stirred and diisopropylethylamine (DIEA, Hunig’s base, 100 µl) was added. The solution was stirred for 10 min at room temperature. EDC (1-(3-dimethylaminopropyl)-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. Dodecylamine (100 mg, 0.55 mmol) was added and the mixture was stirred for 24 h at rt. The mixture was concentrated in vacuo at 80 °C and CHCl3 (50 ml) was added. This solution was extracted with water, HCl (aq. 0.6 M) and again water. The aqueous layers were extracted with CHCl3. The

O

O

NHCO2CH2C6H5

CO2Me

O

O

NHCO2CH2C6H5

HN

O

O

O NHZ

CO2H

O NH2

O CO2Me



Minimal Foldamers

152

combined organic layers were dried (Na2SO4) and concentrated in vacuo. The crude product was purified by column chromatography (silica, CHCl3 + 5% MeOH). Yield: 90 mg (0.16 mmol, 35%). 1H-NMR: δ 7.26 – 7.24 (m, 5H), 6.77 – 6.81 (m, 4H), 5.99 (br, 1H), 5.58 (br, 1H), 5.00 (s, 2H), 3.98 (t, 3J=5.6 Hz, 2H), 3.89 (t, 3J = 5.7 Hz, 2H), 3.38 (m, 2H), 3.05 (m, 2H), 2.19 (t, 3J = 7.3 Hz, 2H), 1.94 (m, 4H), 1.60 (m, 2H), 1.16 (m, 20H), 0.82 (t, 3J = 6.4 Hz, 6H). 13C-NMR: δ 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, 254.

Tetradecanoic acid {3-[2-(3-dodecylcarbamoyl-propoxy)-phenoxy]-propyl}-amide (6.1):

The above compound (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 over celite, the filtrate concentrated in vacuo. 1H-NMR confirmed 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 µl). Yield: 25 mg (0.04 mmol, 30 % over 2 steps). 1H-NMR: δ 6.85 – 6.95 (m, 4H), 4.04 (t, 3J = 3.4 Hz, 2H), 4.02 (t, 3J =2.6 Hz, 2H), 3.51 (m, 2H), 3.19 (m, 2H), 2.45 (t, 3J = 8.3 Hz, 2H), 2.16 (m, 4H), 1.98 (t, 3J = 5.8 Hz, 2H), 1.58 (m, 2H), 1.43 (m, 2H), 1.22 (m, 42H), 0.86 (t, 3J = 6.4 Hz, 6H). 13C-NMR: δ 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.

11-Benzyloxycarbonylamino-undecanoic acid (6.15): 11-amino undecanoic acid (5.0 g,

0.025 mol ) was dissolved in THF-H2O (4:1, 100ml). The mixture was cooled to 0ºC, and KOH (2.8 g, 50 mmol, 2 eq.) was added. The mixture was stirred and benzylchloroformate (5g, 4 ml, 30 mmol, 1.2 eq.) was added carefully. The mixture was stirred for 2h at 0ºC and allowed to stand overnight at room temperature. The aqueous phase was separated and extracted with CHCl3. The combined organic phases were evaporated and redissolved in CHCl3. The organic phase was extracted with H2O, and evaporated. The material was purified over silica (CHCl3). Yield: 2.6 g (7.8 mmol, 31%). white solid. mp. 125-127°C. 1H-NMR (CDCl3): δ = 7.20 (m, 5H), 5.05 (s, 2H), 4.80 (br, 1H), 3.06 (t, 3J=6.6 Hz, 2H), 2.24 (t, 3J =7.1 Hz, 2H), 1.60 (m, 2H), 1.40 (m, 2H), 1.20 (m, 12H). 13C-NMR: δ 179.2, 156.4, 137.3.128.4, 63.8, 40.5, 34.5, 29.8 – 29.0. HRMS: calcd for C18H27NO4 321.194, found 321.196.

HO2C

NHZ

O

HN

ONH

O

O



Chapter 6

153

11-Amino-undecanoic acid methyl ester (6.18): 11-amino undecanoic acid (5.0 g, 0.025

mol) was dissolved in MeOH (100 ml). A drop of H2SO4 was added and the mixture was stirred for 12h at reflux. The solvent was evaporated, the residue taken up in CHCl3, extracted with water, dried and the solvent was removed in vacuo. Yield: 4.9 g (0.022 mol, 95%). white solid. mp.185-186°C. 1H-NMR (CDCl3): δ = 3.62 (s, 3H), 2.59 (t, 3J = 7.5 Hz, 2H), 2.28 (t, 3J = 7.0 Hz, 2H), 1.60 (m, 2H), 1.40 (m, 2H), δ 1.20 (m, 12H). 13C-NMR: δ 174.2, 51.3, 42.6, 34.6, 29.9 – 29.3, 27.4, 24.8. HRMS: calcd for C20H31NO4 349.225, found 349.228.

11-Octanoylamino-undecanoic acid methyl ester (6.19): The methyl ester (6.18, 2.0 g, 0.01

mol) and octanoic acid (1.6 g, 0.011 mol) were coupled according to the standard procedure. Yield: 1.8 g (5 mmol, 50 %). white solid. mp. 106-107°C. 1H-NMR (CDCl3): δ = 5.40 (br, 1H), 3.65 (s, 3H), 3.25 (t, 3J=6.7 Hz, 2H), 2.21 (t, 3J = 7.3 Hz, 2H), 2.08 (t, J =6.9 Hz, 2H), δ 1.58 (m, 4H), δ 1.42 (m, 2H), δ 1.19 (m, 20H), δ 0.83 (t, 3J=7.0 Hz, 3H). 13C-NMR (CDCl3): δ = 177.0, 173.9, 50.8, 38.9, 35.8, 33.4, 33.3, 30.9, 28.7, 28.6, 28.5, 28.4, 28.3, 28.2, 28.1, 26.1, 25.2, 24.3, 21.8, 13.1. MS (EI+): m/z 341, 268, 184. HRMS: calcd for C21H41NO3 355.309, found 355.306.

(10-Hexylcarbamoyl-decyl)-carbamic acid benzyl ester (6.20): The Z-protected acid

(6.17, 2.5 g, 7.5 mmol) was coupled to hexylamine (0.8 g, 7.6 mmol) according to the standard procedure. Yield: 1.6 g (4 mmol, 55%). white solid. mp. 112-113°C. 1H-NMR (CDCl3): δ 7.25 – 7.20 (m, 5H), 5.38 (br, 1H), 4.62 (br, 1H), 3.12 (t, 3J = 5.9 Hz, 2H), δ 3.06 (t, 3J= 5.7 Hz), 2.03 (t, 3J = 7.2 Hz, 2H), 1.51 (m, 2H), 1.38 (m, 4H), 1.20 (m, 20H), 0.84 (t, 3J = 7.1 Hz, 3H). 13C-NMR (CDCl3): δ 173.1, 156.5, 136.8, 128.5, 128.0, 66.2, 41.1, 39.5, 36.9, 31.5, 29.9, 29.6, 29.4, 29.3, 29.2, 29.2, 29.1, 26.6, 26.5, 25.8, 22.5, 13.1. MS (EI+): m/z 418, 311, 283. HRMS: calcd for C24H40N3O4 418.320, found 418.318.

{3-[2-(3-Butylcarbamoyl-propoxy)-phenoxy]-propyl}-carbamic acid benzyl ester (6.13): Compound 6.10 (380 mg, 1mmol) was coupled to butylamine(100mg, 1.3 mmol, freshly distilled over CaH2) according to the standard

procedure. The crude product was purified by column chromatography (silica, CHCl3 + 5% MeOH). Yield: 200 mg (0.45 mmol, 45%) yellow oil. 1H-NMR: δ 7.29 (m, 5H), δ 6.88 – 6.88 (m, 4H), δ 6.09 (br, 1H), δ 5.67 (br, 1H), δ 5.04 (s, 2H), δ 4.03 (t, J=6.8, 2H), δ 3.96 (t, J=6.4, 2H), δ 3.43 (m, 2H), δ 3.12 (m, 2H), δ 2.25 (t, J=5.6, 2H), δ 1.97 (m, 4H), δ 1.34 (m, 2H), δ 1.21 (m, 2H), δ 0.81 (t, J=6.3, 3H). 13C-NMR: δ 171.8, 155.6, 147.7, 135.9, 127.6, 127.1, 120.8, 112.4, 66.8, 66.5, 59.2, 38.4, 38.1, 31,5, 30.7, 28.3, 24.2, 19.0, 12.9. MS (EI+): m/z 442, 142.

O

O NHZ

O

NHC4H9

MeO2C

NH2

MeO2C

HN

O

NHZHN

O



Minimal Foldamers

154

(10-{3-[2-(3-Butylcarbamoylpropoxy)-phenoxy]-propylcarbamoyl}-decyl)-carbamic acid benzyl ester (6.16): Compound 6.13 (0.9 g) was deprotected according to the standard procedure. Yield: 0.65 g.

The free amine (0.6 g, mmol) was coupled to compound 6.15 (0.75 g, mmol,) according to the standard procedure. The crude product was purified by column chromatography (silica, CHCl3 + 5% MeOH). Yield: 200 mg (0.25 mmol, 70%), yellow oil. 1H-NMR: δ 7.30 (m, 5H), 6.86 – 6.88 (m, 4H), 6.54 (br, 1H), 6.06 (br, 1H), 5.04 (s, 2H), 4.02 (t, 3J = 6.8, 2H), 3.46 (m, 2H), 3.13 (m, 4H), 2.39 (t, 3 J =5.6, 2H), 2.11 (m, 4H), 1.96 (t, 3J = 5.8 Hz, 2H), 1.54 (m, 2H), 1.41 (m, 4H), 1.21 (m, 14H), 0.81 (t, 3J = 6.3, 3H). 13C-NMR: δ 173.7, 172.8, 159.1,149.3, 148.7, 136.9, 128.9,128.8, 128.5, 128.4, 122.1, 121.9, 114.4, 114.2, 68.3, 67.6, 41.4, 39.6, 37.6, 37.1, 33.2, 32.0, 30.2, 29.7 – 29.4, 26.9, 25.6, 20.4, 14.1. MS (EI+): m/z 625, 484, 375. HRMS: calcd for C36H55N3O6 625.409, found 625.512.

(3-{2-[3-(10-{3-[2-(3-Butylcarbamoyl-propoxy)-phenoxy]-propylcarbamoyl}-decylcarbamoyl)-propoxy]-phenoxy}-propyl)-carbamic acid benzyl ester (17):

Compound 6.15 (200 mg, 0.32 mmol) was deprotected according to the standard procedure. Yield: 150 mg. The free amine was subsequently coupled to compound 6.10 according to the standard procedure. The product was purified by column chromatography (silica, CHCl3). Yield: 100 mg (0.12 mmol). yellow oil. 1H-NMR: δ 7.30 (m, 5H), 6.82 – 6.86 (m, 8H), 5.02 (s, 2H), 3.98 – 4.02 (m, 4H), 3.38 (m, 6H), 3.09 (m, 2H), 2.40 - 2.00 (m, 14 H), 1.50 – 1.20 (m, 20H), 0.81 (t, 3J = 6.3 Hz, 3H). 13C-NMR: δ 173.8, 172.9, 158.1, 149.1, 148.7,128.8,128.4, 121.9, 121.4, 114.1, 113.8, 68.3, 67.5, 39.6, 37.1, 33.2, 31.9, 29.7 – 29.4, 26.1, 25.6, 20.4, 14.1.

11-{4-[2-(3-Butyrylamino-propoxy)-phenoxy]-butyrylamino}-undecanoic acid {3-[2-(3-butylcarbamoyl-propoxy)-phenoxy]-propyl}-amide (6.3):

Compound 6.17 (90 mg, 0.1 mmol) was deprotected according to the standard procedure. Yield: 60 mg. The free amine was subsequently reacted with butyric acid (standard procedure). The product was purified with repetitive column chromatography (silica, CHCl3 + 1% MeOH; 4x). Yield: 8 mg (0.015 mmol, 15%). 1H-NMR: δ 7.30 (m, 5H), 6.85 – 6.88 (m, 8H), 6.62 (br, 2H), 6.18 (br, 2H), 4.04 (dt,

O

O NHCOC10H20NHZ

O

NHC4H9

O O

NH

NH

O

OO

O

NH

O

NHZ

O ONH

NH

O

OO

O

NH

O

HN

O



Chapter 6

155

4H), 3.46 (m, 2H), 3.20 (m, 4H), 3.09 (m, 2H), 2.42 (m, 2H), 2.30 (m, 4H), 2.13 (m, 2H), 1.43 (m, 8H), 1.23 (m, 14H), 0.84 – 0.88 (m, 6H). 13C-NMR: δ 174.4, 173.5, 149.9, 149.5, 122.6, 122.1, 114.9, 114.6, 69.0, 40.6, 40.3, 36.9, 36.5, 33.9, 32.7, 30.6, 30.4, 30.3, 30.2, 27.9, 26.8, 26.4, 21.2, 19.7, 16.4, 15.0. Maldi-TOF: calcd 796.6, found: 819.8 (M + Na+).

4-{2-[3-(11 Octanoyl amino-undecanoylamino)-pro poxy]-phenoxy}-butyric acid methyl ester (6.22):

The ester (6.18), 1.8 g, 5 mmol) was deprotected with LiOH (0.5 g) in MeOH (50ml). The mixture was stirred for 12 h at rt. The solvent was removed in vacuo, the residue taken up in CHCl3

(50 ml), extracted with water, and dried (Na2SO4). The solvent was removed to yield the free acid. Yield: 1.1 g (3.3 mmol, 66%). 1H-NMR indicated removal of the methoxy group. The acid (0.8 g, 2.5 mmol) and compound 6.21 (0.6 g, 2.2 mmol) were coupled according to the standard procedure. The product was purified by column chromatography (silica, CHCl3). Yield: 0.6 g (1.1 mmol, 50 %). white powder. mp. 117-121°C. 1H-NMR (CDCl3): δ 6.89 (m, 4H), 4.06 (m, 4H), 3.66 (s, 3H), 3.49 (t, 3J = 6.2 Hz, 2H), 3.19 (t, 3J = 6.4 Hz, 2H), 2.52 (t, 3J =6.9 Hz, 2H), 2.16 (m, 6H), 1.97 (m, 2H), 1.46 (m, 4H), 1.28 (m, 22H), 0.83 (t, 3J = 7.0 Hz, 3H). 13C-NMR (CDCl3): δ 173.7, 173.5, 149.3, 121.7, 114.6, 68.7, 68.2, 52.0, 39.9, 38.3, 37.2, 32.0, 30.0 – 29.2, 27.2, 26.1, 24.9, 22.9, 14.5. MS (EI+): m/z 576, 367. HRMS: calcd for C33H56N2O6 576.808, found 577.102.

(3-{2-[3-(10-Hexylcarbamoyl-decylcarbamoyl)-propoxy]-phenoxy}-propyl)-carbamic acid benzyl ester (6.23):

Compound 6.20 (1.5 g, 4 mmol) was deprotected according to the general procedure. Yield: 0.9 g (3 mmol, 75%). 1H-NMR indicated removal of the CbZ- group. The free amine (1.2 g, 3 mmol) and compound 6.10 (0.85 g, 3 mmol) were coupled according to the standard procedure. The product was purified by column chromatography (silica, CHCl3). Yield: 0.7 g (1.1 mmol, 35 %). white solid. mp. 131-132°C. 1H-NMR (CDCl3): δ 7.30 (m, 5H), δ 6.88 (m, 4H), δ 4.07 (t, J= 6.4 Hz, 2H), δ 3.97 (t, J=6.5 Hz, 2H), 3.43 (q, J=5.8 Hz, 2H), δ 3.18 (m, 4H), δ 3.08 (m, 2H), δ 2.28 (t, J=6.9, 2H), δ 2.18 (m, 2H), δ 1.98 (m, 4H), δ 1.41 (m, 4H), δ 1.26 (m, 20H), δ 0.83 (t, J=7.0 Hz, 3H). 13C-NMR (CDCl3): δ 174.1, 173.4, 157.6, 149.7, 149.3, 137.6, 129.6, 129.2, 122.5, 121.9, 114.3, 114.1, 68.7, 68.6, 40.5, 37.9, 33.8, 32.5, 30.7, 30.6, 30.4, 30.3, 30.3, 27.9, 27.6, 26.9, 23.6, 15.0. MS (EI+): m/z 653, 545, 353. HRMS: calcd for C38H59N3O6: 653.440, found 653.442.

OHN

HN

O CO2Me

O O

O NHZ

O

O

HN

HN

O



Minimal Foldamers

156

[10-(3-{2-[3-(10-Hexylcarbamoyl-decylcarbamoyl)-propoxy]-phenoxy}-propylcarbamoyl)-decyl]-carbamic acid benzyl ester (6.24):

Compound 6.23 (0.65 g, 1 mmol) was deprotected according to the standard procedure. Yield: 0.5 g (0.95 mmol, 95 %). The free amine was coupled with the spacer (360 mg, 1.05 mmol) according to the standard procedure. The product was purified by column chromatography (silica, CHCl3

+ 1% MeOH) and reprecipitated from EtOAc. Yield: 0.4 g (0.5 mmol, 53 %). White solid. 1H-NMR (CDCl3): δ = 7.30 (m, 5H), δ 6.85 (m, 4H), δ 5.04 (s, 2H) δ 4.01 (m, 4H), δ 3.48 (m, 2H), δ 3.24 – 3.06 (m, 6H), δ 2.43 (m, 2H), , δ 2.11 (m, 6H), δ 1.95 (m, 2H), δ 1.56 (m, 4H), δ 1.42 (m, 4H), δ 1.26 (m, 28H), δ 0.83 (t, J=7.0 Hz, 3H). 13C-NMR (CDCl3): δ 173.7, 173.5, 156.7, 149.1, 148.7, 144.7, 128.8, 128.4, 121.9, 121.4, 114.1, 113.8, 68.3, 67.5, 41.4, 39.9, 37.1, 37.0, 31.8, 30.2, 29.9, 29.8, 29.7, 29.6, 29.5, 29.2, 27.2, 27.0, 26.9, 25.7, 22.9, 14.3.

11-[4-(2-{3-[11-(4-{2-[3-(11-Octanoylamino-undecanoylamino)-propoxy]-phenoxy}-butyrylamino)-undecanoylamino]-propoxy}-phenoxy)-butyrylamino]-undecanoic acid hexylamide (6.4):

The ester 6.22 (0.5 g, 0.9 mmol) was deprotected wth LiOH /MeOH. Yield: 0.4 g (0.7 mmol, 78%). Compound 6.24 (0.4 g, 0.45 mmol) was deprotected according to the standard procedure: Yield: 0.3 g (0.4 mmol). Both compounds were briefly flushed over silica (CHCl3 + 2% MeOH) before use in the final coupling step. The free acid (90 mg, 0.16 mmol) was coupled to the free amine (100 mg, 14 mmol) according to the standard procedure. The product was purified with column chromatography (silica, CHCl3 + 2% MeOH). Yield: 20 mg (0.016 mmol, 10 %). The compound was further purified by RP-HPLC (symmetry column). The compound was dissolved in a mixture of acetic acid and acetonitrile (50/50), brought to pH 4 with Et3N (5% v/v). The compound was eluted with acetonitrile/water (gradient from 50/50 to 80/20). The main peak was collected at

O

O

HN

HN

O

ONH

O

ZHN

O

O

HN

HN

O

ONH

O

NH

O

O

OHN

O

HN

O



Chapter 6

157

70 min. The combined fractions were evaporated, taken up in CHCl3 and washed with water. The organic phase was dried and evaporated. Yield: 2mg. 1H-NMR (CDCl3): δ = 6.87 (m, 8H), δ 4.04 (m, 8H), δ 3.48 (m, 4H), δ 3.21 (m, 8H), δ 2.45 (m, 2H), δ 2.13 (m, 12H), δ 1.98 (m, 2H), δ 1.45 (m, 8H), δ 1.23 (m, 58 H), δ 0.85 (m,6H). 13C-NMR (CDCl3): δ 173.5, 148.8, 121.4, 113.5, 67.9, 39.6, 36.9, 31.6, 29.8 – 28.8, 26.8, 25.8, 22.5, 14.0. Maldi-TOF: calcd. 1261.9 found: 1283.9 (M + Na).

6.10 References

1 For an excellent review on design of folded peptides see: Venkatraman, J., Shankaramma, S.C., Balaram, P. Chem.Rev. 2001, 101, 3131 − 3152. 2 Hill, D.J., Moi, M.J., Prince, R.B., Hughes, T.S., Moore, J.S. Chem. Rev. 2001, 101, 3893-4011. 3 Trabocchi, A., Guarna, F., Guarna, A. Current Org. Chem. 2005, 9, 1127 – 1153. 4 Belvisi, L., Colombo, L., Manzoni, L., Potenza, D., Scolastico, C. Synlett. 2004, 9 1449 – 1471. 5 Lelais, G., Seebach, D. Biopolymers 2004, 76, 206 – 243. 6 Suat, K., Jois, S.D.S. Current Pharm. Design, 2003, 9, 1209 – 1224. 7 Ienaga, K., Higashiura, K., Toyomaki, Y., Matsuura, H., Kimura, H. Chem. Pharm. Bull. 1988, 36, 70 – 77.



Summary In supramolecular chemistry the interaction between molecules is a central theme. Just like two molecules

with the same gross formula can have very different properties, molecular aggregates, composed of the

same molecules can be very different.

In this thesis a special kind of aggregate is the central topic: this aggregate is called a monolayer.

As the name indicates a monolayer is a film, of one molecule thickness, extending in two dimensions.

Because we live in a 3-dimensional world, a 2-dimensional structure is quite remarkable; it tends to form

preferentially at the interface between two phases. These can be liquid-liquid interfaces or liquid-solid

interfaces.

The monolayers described in this thesis form at the liquid-solid interface by means of a process known as

physisorption. This name signifies that no covalent bonds are formed between the solid phase (the

substrate) and the molecules in the aggregate. The substrate can be regarded a template, on which the

monolayer can grow. This growth process is known under the name self-assembly.

In this thesis, a number of methods to generate self-assembling molecular systems at the liquid-solid

interface are described. The importance of interfaces, in chemistry, physics and biology underscores the

relevance of the project. The title of this thesis is ”Pattern formation in organic monolayers.” Monolayers

have been introduced above. Organic refers to the kind of molecules that are used in the process. Pattern

formation is what makes the work described in this thesis interesting. The patterns are intrinsically formed

when the molecules adsorb parallel (coplanar) to the surface. The process is in a way comparable to tiling

a wall, but on the nanoscale. The molecules (adsorbents) can in principle be laid in a multitude of patterns,

depending on their shape.

What makes this approach so powerful is the fact that the kind of pattern is intrinsically connected to the

molecular properties. As a consequence, the (geometric) properties of the aggregate can be programmed

at the molecular level. Therefore changing the molecular features intrinsically transcends a level of

hierarchy. Upon applying a solution of the adsorbents, the pattern is generated spontaneously.

In chapter 2, the scientific literature concerning molecular patterning is scrutinized. Two approaches, top-

down and bottom-up (self-assembly), are compared. Usually bottom-up approaches are invoked to create

patterns in chemisorbed monolayers (with covalent connections to the substrate), whereas self-assembly

is used in physisorbed monolayers (vide supra).

The residual part of this thesis deals with several aspects of the intrinsic pattern formation in physisorbed

monolayers. In chapter 3, monolayers formed from functionalized adsorbents are discussed. By

incorporation of a bipyridine moiety, the formation of complexes is made possible, in situ, at the interface.

Parameters involved with the formation of complexes, or the absence thereof, will be identified.

In chapter 4 a different aspect of the monolayers is treated. The central topic of this chapter is how

different components can be mixed into the monolayer. The question arises whether the components can

be programmed to lie in an alternating sequence in the lamellae. It will be shown that the self-assembly

approach is limited with respect to this endeavour.



Confronted with the problems identified in chapter 4, in chapters 5 and 6 a possible solution is outlined. It

is proposed that by means of covalent connection of the individual strands, in a way that facilitates folding

of the molecules in such a way that all the strands lie into the plane of the monolayer, alternating

sequences can be realized. The envisaged structure is reminiscent of a motif found in nature: the β-sheet.

Relevant structural variations will be examined, and the synthesis of such oligomers (foldamers) is

discussed.



Samenvatting Chemie is de wetenschap der materie. Chemici gaan er van uit dat materie is opgebouwd uit molekulen,

die weer zijn opgebouwd uit atomen. Echter, de stap naar allerhande nuttige materialen vereist een vaak

zeer specifieke constitutie van deze moleculen. Dit is chemie voorbij het molekuul, voorbij de covalente

binding. Het besef dat er in dit bereik veel te ontdekken valt, is vanaf de jaren tachtig steeds sterker

geworden, en er is zelfs een aparte naam voor bedacht: supramoleculaire chemie.

In de supramoleculaire chemie draait het om de wisselwerking tussen de molekulen. Zoals twee

molekulen met dezelfde brutoformule, zeer verschillende eigenschappen kunnen hebben, zo kunnen

molekulaire aggregaten, bestaande uit dezelfde molekulen, zeer verschillend zijn. In dit proefschrift staat

een speciaal soort aggregaat centraal: de monolaag.

Zoals de naam zegt is dit een film, slechts een molekuul dik, die zich in twee dimensies uitstrekt.

Aangezien wij ons in een (minimaal) drie dimensionale wereld bevinden, is een 2-dimensionale struktuur

vrij bijzonder. Zij vormt zich bij voorkeur aan een grensvlak tussen twee fasen. Dit kan een vloeibaar-

vloeibaar grensvlak zijn, of een vast-vloeibaar grensvlak.

De monolagen, beschreven in dit proefschrift vormen zich op het vloeibaar-vaste grensvlak d.m.v.

physisorptie. Dat wil zeggen dat er geen covalente bindingen gevormd worden tussen de vaste fase (het

substraat) en de molekulen in het aggregaat. Het substraat vormt als het ware een sjabloon, waarop de

monolaag kan groeien. Dit groeiproces staat bekend onder de de naam zelf-assemblage.

Dit proefschrift beschrijft een aantal methodes om zelf-assemblerende moleculaire systemen te maken op

het vloeibaar-vaste grensvlak. Het grote belang van grensvlakken in chemie, fysica en biologie maakt dit

tot een uitdagend project.

De titel van dit proefschrift luidt ”Pattern formation in organic monolayers”, ofwel patroonvorming in

organische monolagen. Monolagen zijn hierboven geintroduceerd. Organisch slaat op het soort molekulen

dat hierbij gebruikt wordt. De meest interessante term uit de titel is echter “patroonvorming”. Indien de

Figuur 1. De vorming van physisch gesorbeerde

monolagen.



molekulen evenwijdig aan een oppervlak adsorberen, kan de vorming van een monolaag gezien worden

als een proces gelijkaardig aan het betegelen van een muurtje, maar dan op nanoschaal.

Zoals tegeltjes in een veelheid aan patronen gelegd kunnen worden, zo kunnen molekulen (adsorbenten)

in een veelheid aan patronen assembleren op een oppervlak. De vorm van de molekulen is hierbij zeer

belangrijk. Een voorbeeld is de assemblage van alkyl fragmenten (figuur 2).

Het aardige is nu dat deze patronen intrinsiek reeds besloten liggen in de molekulaire eigenschappen.

Dientengevolge kunnen de structurele eigenschappen van het aggregaat geprogrammerd worden op het

moleculaire niveau. Dit is derhalve een hierarchie overschrijdende bezigheid. Na het aanbrengen van een

oplossing met de adsorbenten hoeft er niets gedaan te worden van de kant van de chemicus: het patroon

ontstaat vanzelf.

Om de vergelijking met het bedekken van een muur nog wat verder te trekken: men kan er voor kiezen in

plaats van tegeltjes, een egaal behang te gebruiken, en daar vervolgens een patroon op te schilderen.

Zoiets is ook mogelijk bij het vormen van monolagen. Indien de molekulen in de monolaag wel covalente

bindingen vormen met het substraat, men spreekt dan van een chemisch gesorbeerde laag, verkrijgt men

een egaal gevormd tapijtje, waarin vervolgens patroontjes gemaakt kunnen worden.

In dit proefschrift worden beide methoden kort met elkaar vergeleken in hoofdstuk 2. Met name het

verschil tussen zelf-assemblage (een laag-hoog benadering) en de hoog-laag benaderingen voor patroon

vorming in chemisch gesorbeerde lagen wordt hier behandeld. De rest van het proefschrift behandelt

diverse aspecten van de intrinsieke patroonvorming in fysisch gesorbeerde lagen.

In hoofdstuk 3, worden monolagen behandeld die gevormd worden door gefunctionaliseerde adsorbenten.

Vergelijk deze molekulen met latjes, die een haakje bevatten, waaraan vervolgens iets opgehangen kan

worden. In het tweede deel van hoofdstuk 3 zal de modificatie (het ophangen!) besproken worden. Het zal

duidelijk worden dat de manier waarop deze molekulen ontworpen worden, een grote invloed heeft op het

uiteindelijk welslagen van de operatie.

Figuur 2. Adsorptie van alkylfragmenten, met interacties

tussen de ketens, op het grafiet rooster.



In hoofdstuk 4 gaat het om een ander aspect. Hoe kunnen latjes van verschillende vorm samen gelegd

worden, zodanig dat ze netjes om en om liggen. Er zal aangetoond worden dat het mengproces tot op

zekere hoogte gestuurd kan worden door op moleculaire schaal de potentiele interacties op een slimme

manier in te programmeren. Tevens zal het duidelijk worden dat het vormen van een patroon met 2

constituenten, gedeeltelijk buiten de mogelijkheden valt, hetgeen terug te voeren is op de manier (zelf-

assemblage) waarop het lattenbodempje tot stand komt.

In het aanschijn van deze problematiek, zal een mogelijke oplossing worden aangedragen, die vervolgens

in de hoofdstukken 5 en 6 verder uitgewerkt wordt. Deze oplossing behelst het covalent verbinden van de

latjes, zodanig dat deze kunnen vouwen in hele specifieke vormen. De gevouwen latjes, zouden

vervolgens weer moeten kunnen assembleren op een oppervlak. Deze structuur vertoont de nodige

gelijkenis met een assemblage motief, dat reeds bekend is uit de levende natuur: de β-sheet.

Dit patroon komt men onder andere tegen in eiwitten. Aangezien de vouwing van eiwitten van cruciaal

belang is voor de manier waarop zij hun werk doen, en dit proces nog allerminst in detail begrepen wordt,

kunnen zelf-assemblerende gevouwen structuren op een grensvlak als een model-systeem beschouwd

Figuur 3. Monolaag van gefunctionaliseerde adsorbenten. Onderwerp van

hoofdstuk 3.

+

+

2

?

Nee ! Statistische verdeling der componenten op het oppervlak!

Figuur 4. Monolaag gevormd met meerdere componenten. Er is geen controle over de precieze distributie van

de componenten.



worden, hetgeen de relevantie van het onderzoek beschreven in dit proefschrift in een bredere context

plaatst.

Tenslotte, wil ik de lezer er op wijzen dat het onderzoek beschreven in dit proefschrift op meerdere nivo’s

zich op een grens bevindt. Niet alleen vormen de structuren zich op de grens van een vloeibare en een

vaste fase, ook worden aspecten van meerdere disciplines verenigd. Zo bezien speelt dit onderzoek op

het grensvlak van chemie, fysica en biologie. Organische synthese, fysische karakterisatie, en gebruik van

computermodellen, dienen op elkaar afgestemd te worden teneinde deze fascinerende aggregaten te

realiseren. Naar mijn stellige overtuiging derhalve komt het werk beschreven in dit boekje aan het motto

van deze universiteit: “werken aan de grenzen van het weten” ten volle tegemoet.

fragmentenverbinden

Figuure 5. De laterale distibutie van

functionaliteit kan, in principe,

gecontroleerd worden door middel van

vouwende adsorbenten (foldameren).



Dankwoord

Het is volbracht. Men aanschouwe het resultaat van 5 jaar noeste arbeid. Het was zeker niet altijd

gemakkelijk. Zo nu en dan sloeg de wanhoop toe, maar uiteindelijk is er wel wat bereikt. Vanuit een pril

begin zijn een aantal aardige resultaten geboekt. Zoals die dingen gaan, heeft idealisme gaandeweg

plaats gemaakt voor realisme. Ja, pragmatisme name de overhand. Maar goed het doel heiligt de

middelen, en het doel is bereikt.

Rest mij een aantal mensen te bedanken. Dit proefschrift is tot stand gekomen dankzij de contributie van

velen. In willekeurige volgorde: Jan Engberts en Kees Hummelen, leden van de leescommissie. Bedankt

voor de snelle correctie van het manuscript.

Ben, mijn promotor. We hebben elkaar misschien iets minder vaak gesproken dan jou lief was, maar ik

hoop dat het eindresultaat je toch kan bekoren. Je enthousiasme voor de chemie is aanstekelijk, en heeft

mij meer dan eens doen doorzetten, als het weer eens lastig was.

Jan, je was verantwoordelijk voor de dagelijkse begeleiding, hetgeen je naar mijn mening goed is

afgegaan. Ik geloof dat wij elkaar redelijk aanvoelen (aanvullen ?). Jouw analyse m.b.t. het promotietraject

in termen van uithuilen en weer door gaan, heb ik als zeer accuraat ervaren.

Steven, zonder de cooperatie van de spectroscopie groep in Leuven had hier nooit een boekje gelegen.

Bedankt voor de immer rappe correspondentie, en ook voor het corrigeren van het manuscript.

Also many thanks to the PhD students and postdocs in Leuven, who proved to be so capable in making

the images of my compounds. I realize some of them were quite nasty to handle. Thanks Mohamed,

Hiroshi en Andrey.

Nathalie, I would also like to thank you for proofreading the manuscript, and for your assistance with some

of the work, that hasn’t made the thesis.

Ik heb het genoegen mogen smaken een student te begeleiden tijdens een stage in Leuven. Bas, zoals je

ziet, hebben een aantal van jouw resultaten het boekje gehaald.

Ook mijn directe collegae in de supramoleculaire chemie groep: Maaike, Niek, Jaap en Joost, Kjelt en

Arianne. Bedankt voor de leuke samenwerking en de stimulerende discussies.

De collegae op zaal, Hans en Rob, hebben bijgedragen aan een goede atmosfeer. Voorts natuurlijk dank

aan de mensen die de diverse analyses uitvoerden of daarbij behulpzaaam waren: Theodora voor de

HPLC, Albert voor de massa’s en Wim voor de NMR. Tenslotte dank aan alle goede vrienden en

kennissen binnen en buiten het lab die het mogelijk maakten de nodige afstand te nemen van de chemie,

als het allemaal iets te veel werd. Als laatste, paps en mams. Ik heb me altijd zeker geweten van jullie

onvoorwaardelijke steun.



Documents

University of Groningen Pattern Formation in organic