Self-Assembly Growth of Organic Thin Films and Nanostructures by

Nov 2, 2001 - ... Johannes V. Barth 2, Lukas Bürgi 2, Olivier Jeandupeux 2, Michael Kiy 1, Ivan Biaggio 1, Ilias Liakatas 1, Klaus Kern 2, and Peter ...
0 downloads 0 Views 2MB Size
Download PDF
Chapter 3

Anisotropic Organic Materials Downloaded from pubs.acs.org by UNIV OF CALIFORNIA SANTA BARBARA on 09/19/18. For personal use only.

Self-Assembly Growth of Organic Thin Films and Nanostructures by Molecular Beam Deposition 1,3

1

1

1,4

Chengzhi Cai , Martin Bösch , Christian Bosshard , Bert Müller , Ye Tao , Armin Kündig , JensWeckesser ,Johannes V. Barth , Lukas Bürgi , Olivier Jeandupeux , Michael Kiy , Ivan Biaggio , Ilias Liakatas , Klaus Kern , and Peter Günter 1,5

1

2

2

2

1

2

1

2

1

1

1

Nonlinear Optics Laboratory, Institute of Quantum Electronics, Swiss Federal Institute of Technology (ΕΤΗ), CΗ-8093 Zurich, Switzerland Institut de Physique Expérimentale, EPFL, PHB-Ecublens, CH-1015 Lausanne, Switzerland Current address: Department of Chemistry, University of Houston, Houston, T X 77204-5641 Current address: Biocompatible Materials Science and Engineering, ΕΤΗ Zürich, Switzerland Current address: Institute for Microstructural Sciences, National Research Council Canada, M-50 Montreal Road, Room 178, Ottawa, Ontario K1A 0R6, Canada 2

3

4

5

The concept o f supramolecular assemblies based on strong and directional hydrogen bonding has been applied to organic molecular beam deposition (OMBD) for growth o f anisotropic nanostructures and thin films. A l i g n e d nanostructures were generated on Ag(111) surfaces. T h i n films with a thickness o f 100-400 n m were grown on glass substrates by oblique incidence OMBD, and studied by second h a r m o n i c generation experiments, i n d i c a t i n g that the average direction o f the dipolar molecules i n the films was parallel to the projection o f the m o l e c u l a r beam on the substrate surface. T h i s i n t r i g u i n g result is r a t i o n a l i z e d b y a p r o p o s e d m e c h a n i s m considering the fundamental processes o f self-assembly on surfaces.

34

© 2002 American Chemical Society

35 Introduction Organic thin films and nanostructures have attracted increasing interest for their potential applications i n a variety o f advanced technologies, i n c l u d i n g nonlinear optics ( N L O ) , microelectronics, nanotechnology, light emitting devices, field-effect transistors, l i q u i d crystals, sensors, a n d solar c e l l s (1-7). F o r m a n y o f these applications, the device performance is crucially dependent o n the orientation o f the functional molecules i n the f i l m or the nanostructure (1-3). Therefore, methods for alignment o f molecules i n the structures are o f great t e c h n o l o g i c a l interest. T h e alignment can be centrosymmetric or non-centrosymmetric. Non-centrosymmetric alignment is more challenging, because the acentric molecules need to be aligned w i t h a directional preference. It is, however, the basic requirement for organic second order nonlinear optics (1,2) w h i c h is our primary interest. Several techniques have been developed for the growth o f anisotropic organic thin films (1-7). F o r the alignment o f molecules perpendicular to the substrate surface, most c o m m o n techniques include L a n g m u i r - B l o d g e t t ( L B ) film formation, h i g h electric field poling, and self-assembly using the layer-by-layer methodology (4). F o r aligning molecules parallel to the substrate surface (in-plane), the anisotropy can be induced b y an anisotropic substrate surface (7). H o w e v e r , i f an isotropic substrate surface is required, the surface alone cannot induce an in-plane preferential alignment over a large area. T o circumvent this problem, external p h y s i c a l means m a y be applied to impose a preference to the alignment. F o r example, w e have demonstrated that, during L B f i l m deposition, the d i p p i n g direction can be used to define the alignment direction o f some second order N L O chromophores (8).

Organic Molecular Beam Deposition (OMBD) In general, organic thin films can be deposited i n solution or gas phase. Solutionbased deposition is relatively easy to set up, and has been intensively studied. W h e n we became i n v o l v e d i n this field, w e were, however, interested i n an instrumentally sophisticated technique based on ultrahigh vacuum, referred to as organic molecular beam deposition ( O M B D ) (7). It is an offshoot o f the p h y s i c a l vapor deposition ( P V D ) technique that has been w i d e l y used i n microelectronic and optical industries for the deposition o f inorganic thin films (9). The O M B D process is illustrated i n Figure 1. The materials i n the effusion cells (d, Figure 1) are evaporated, and the gas molecules rush out o f the s m a l l hole o f the cell into the ultrahigh vacuum chamber, forming a molecular beam (f). Some o f them stick o n a relatively c o l d substrate (a) w h i l e the others are absorbed b y the l i q u i d nitrogen shroud (g). The ultrahigh vacuum (< 10" mbar) is generated b y the turbo molecular pump (h) and the l i q u i d nitrogen shroud (g). The film thickness can be monitored w i t h monolayer sensitivity b y the quartz thickness monitor (b) o r the ellipsometer (c). The deposition (on/off) can be controlled by the shutters (e). A s compared to the solution-based growth techniques, O M B D has several practical advantages. It is carried out i n ultrahigh v a c u u m that is an ultraclean environment. It enables precise aiid in-situ control o f substrate temperature, growth rate, and film thickness. Because the molecules i n the beam under ultrahigh vacuum 8

36 are too far away to interact w i t h each other, they fly straight. I f the dimensions o f the substrate are m u c h shorter than the distance between the effusion c e l l and the substrate, the approaching direction o f the gas molecules should be approximately the same over the w h o l e film area. A s discussed later, this " b e a m - l i k e " feature can be used to align molecules i n the film. F r o m a practical point o f view, it also allows easy fabrication o f w e l l defined and miniature patterns u s i n g masks i n front o f the substrate. In addition, O M B D can combine the deposition o f metals, organics, and semiconductors i n the same chamber w i t h a predetermined sequence, composition, and layer thickness, for preparation o f w e l l defined hetero-layer structures. M o r e o v e r , h i g h growth rates i n the order o f micrometers per hour are possible. C o m b i n i n g a l l these advantages, O M B D appeared to be an ideal t o o l for the fabrication o f e.g. integrated optical and electrical circuits based on conjugated organic molecules.

Figure 1. Illustration of Organic Molecular Beam Deposition (OMBD). a: substrate; b: quartz crystal thickness monitor; c: ellipsometer; d: effusion cells; e: shutters;/: molecular beam; g: liquid nitrogen shroud; h: turbo molecular pump.

Supramolecular Assemblies as Anisotropic Materials for OMBD Despite the above advantages o f O M B D , the technique has been far less developed for growth o f anisotropic organic thin films as compared to the solutionbased techniques. The m a i n reason is probably that the O M B D technique is neither

37 familiar nor accessible to many organic chemists. The shortage o f input from organic chemistry hampers the development o f materials for this technique. Our m a i n interest was to grow N L O thin films where the dipolar molecules are aligned i n the same direction. In general, N L O materials can be grouped into t w o types: the l o w molecular weight crystalline materials and the polymers (1,2). It is extremely difficult to grow a large single crystalline film b y O M B D . It has been demonstrated that an anisotropic p o l y crystalline film o f 4'-nitrobenzylidene-3acetamino-4-methoxy-aniline ( M N B A ) , an efficient N L O material, can be grown on a centrosymmetric organic single crystal o f ethylenediammonium terephthalate (10). Here, the anisotropy o f the films is induced by the lattice-matched substrate surface (heteroepitaxy). H o w e v e r , it is difficult to apply this method to m a n y practical inorganic substrates that are either amorphous or have a distinct lacttice constant from that o f organic crystals. In fact, many w e l l k n o w n l o w m o l e c u l a r weight N L O materials have been tried to grow on silicon and glass substrates, but at best resulting in films consisting o f randomly oriented and μ η ι - s i z e d microcrystals that cause h i g h scattering losses (11). O n the other hand, polymers containing N L O chromophores can be easily processed to thin films by spin coating, and the dipolar chromophores can then be aligned by high electric field poling at the glass transition temperature o f the p o l y m e r (1,2). S u c h films have a m u c h better o p t i c a l q u a l i t y than the polycrystalline films o f conventional l o w molecular weight materials. H o w e v e r , the n o n - v o l a t i l e p o l y m e r s are not suitable for O M B D , and the above m e n t i o n e d advantages o f O M B D are difficult to be realized with the polymer materials. W e reasoned that a compromise might be found by developing a new type o f materials for O M B D based on supramolecular assemblies (12) where the molecules are linked to each other v i a strong and directional interactions such as hydrogen bonds ( Η - b o n d s ) . I f the strong intermolecular bonds can be broken at elevated temperatures w h i l e the molecules are intact, then the materials might be suitable for O M B D . In addition, the grain boundary i n such materials can be reduced (13) and the molecular alignment stabilized. Based on this idea, we designed a series o f molecules having a p y r i d y l group at one end, and a carboxy group at the other, such as 4-[(pyridin-4y l ) v i n y l ] b e n z o i c acid (1) and 4-[(pyridin-4-yl)ethynyl]benzoic acid (2) (Figure 2). T h e y are expected to form strong and linear head-to-tail intermolecular Η - b o n d s i n the solid states (14-17).

1

2

—-N - H - -N

. The data were obtained at 77K. Β is reproduced with permission from ref. 18. 2

2

2

41 large crystallites can be attributed to the directional head-to-tail Η - b o n d i n g w h i c h dominates the other intermolecular interactions. T h e strong Η - b o n d i n g is also expected to stabilize the polar order. Indeed, the S H G intensity o f the films decreased only slightly before reaching 190°C for 1 and 180°C for 2. A m o n g the above results, the most intrigue one is that the molecular alignment direction is defined by the projection o f the obliquely incident molecular beam on the substrate surface ( X axis). In fact, it took us a long time to come to this conclusion and to be c o n v i n c e d that the alignment was not due to the possible substrate anisotropy. B u t then h o w can we explain this? The molecules i n the molecular beam can rotate freely although they f l y i n the same direction, then w h y should they preferentially align along the X axis? In addition, w h y the same degree o f order can be kept for hundreds o f layers without accumulation o f errors? 3

3

Self-Assembly of Nanostructures by OMBD W e reasoned that the superstructures o f ultra-thin films o f 1 and 2 on a flat metal surface might be observed directly by scanning tunneling microscopy ( S T M ) . A series o f ultra-thin films o f 1 were then g r o w n on the A g ( l l l ) surface by O M B D , and characterized by i n situ S T M (18). Since A g is a noble metal, the absorbate/substrate interactions are weak, and i n v i e w o f the smoothness o f the close-packed (111) geometry, it was expected that the intermolecular interaction could be reflected by the molecular arrangement at the surface. Indeed, we found that upon deposition at 300 Κ 1 self-assembled into linear lines (the bright lines i n Figure 3 A ) w i t h a length o f up to several μ ι η and a w i d t h o f only about 1 n m . The lines oriented along directions o f the Ag-lattice w i t h mesoscopic ordering at the μ ι η scale, o n l y w e a k l y affected b y the atomic steps o f the substrate surface. T h i s corresponds to a one-dimensional nanograting, noting that the distance between the parallel lines was about 10 n m .

Proposed Mechanism of Self-Assembly During OMBD T o answer the above questions, we need to consider details o f the t h i n film growth process (17). First o f a l l : w h y can the molecules i n ultrahigh v a c u u m be deposited on a surface? O b v i o u s l y , that is because the surface molecules b o n d the i n c o m i n g molecules. T h i s process can occur w h e n the free energy o f b o n d i n g is negative ( A G = Δ Η - T A S < 0), that is, the enthalpy o f bonding ( - Δ Η ) is larger than the entropy term ( - T A S ) w h i c h favors dissociation particularly i n ultrahigh vacuum. F o r 1 and 2, the intermolecular interactions include Η - b o n d i n g , van der Waas forces, and π - π stacking interaction. Without Η - b o n d i n g , the other two bonding interactions appear too weak to keep the molecules from dissociating at r o o m temperature. In fact, without a Η - b o n d , the methyl ester o f 1 has a dramatically lower melting point than 1 (106 vs 3 5 0 ° C ) . Materials w i t h a melting point lower than 130°C usually cannot be

42

Figure 4. A: proposed alignment direction of 1 grown on silylated glass. B: setup for Oblique Incidence OMBD and the observed alignment direction of 1 and 2 on bare and silylated glass substrates. C: sample rotation angle dependent SHG. D: Incident angle dependent SHG. E. Thickness dependent SHG. F: SEM image of a film of 1 grown on glass at 100°C, reproduced with persission from reference 15.

43 deposited at 2 7 ° C b y O M B D (11). T h e fact that films o f 1 and 2 c a n be easily deposited even at 100°C should be due to the H-bonding.

Thermodynamic Aspects of the Hydrogen Bonding F o r 1 and 2, the strongest Η - b o n d s are the tail-to-tail bonding and the head-to-tail bonding (Figure 5), and the former having two O H Ο bonds is stronger than the latter h a v i n g only one Η Ν bond. However, i f many molecules are i n v o l v e d , the head-totail bonding that leads to supramolecular assemblies can be thermodynamically more favored, considering that each molecule i n the chain has also two Η Ν bonds (Figure 5) that should be stronger than the H O bonds. This is i n accord with the above s o l i d state N - N M R studies that indicated the dominance o f head-to-tail Η - b o n d s . T h e enthalpy ( Δ Η ) for d i m e r i z a t i o n o f c a r b o x y l i c a c i d derivatives is t y p i c a l l y - 1 5 k c a l / m o l , a n d the entropy ( A S ) is - 3 6 c a l / m o l ( 2 3 ) . H e n c e the d i s s o c i a t i o n temperature (when A G = 0) for a tail-to-tail bond is 1 4 4 ° C . T h i s is considerably higher than the desorption temperature o f the f i l m ( 1 2 8 ° C ) measured b y i n situ ellipsometry (17). It suggests that the desorption may involves sequentially breaking the head-to-tail Η - b o n d s o f the surface molecules. 1 5

Figure 5. The strongest Η-bonds for 1.

Directional Requirement for Hydrogen Bonding W e have assumed that thin f i l m growth o f 1 and 2 is m a i n l y due to the head-totail and tail-to-tail Η - b o n d i n g . N o w w e need to consider the kinetics o f the bonding. These bonds are most l i k e l y to form when two molecules collide i n the ways shown by the large arrows i n Figure 6, that i s , the carboxy Η - a t o m approaches another molecule along the axes o f the non-bonding electron pairs at the O - or the N - a t o m s (24). I f the collision happens i n the other directions (the small arrow i n Figure 6), the chance for the bonding is lower.

Figure 6. Illustration

of the directional preference

of Η-bonding

of 1.

44 A s mentioned before, during O M B D , the molecules i n the beam translate to the film surface i n the same direction. In this case, the Η-bonding probability should be influenced by the orientation of the surface molecules. This assumption was supported by the growth o f 1 o n P d ( l l O ) surface at 2 7 ° C (25). U n l i k e A g ( l l l ) surface, the P d ( l 10) surface bonds 1 strongly w i t h an estimated bonding energy o f ~65 k c a l / m o l (25). It forces 1 to l a y flat on the P d surface as s h o w n b y the i n situ S T M measurement (Figure 7). Interestingly, when the substrate surface was covered b y a monolayer o f 1 (Figure 7 B ) the growth stopped, that is, the incoming molecules c o u l d no longer b o n d to the already adsorbed molecules. T h i s is probably due to the unfavorable orientations o f the surface molecules that are determined by the substrate surface. T h i s orientation provides the lowest chance for the surface molecules to hydrogen b o n d the i n c o m i n g molecules (Figure 6), w h i l e the other intermolecular forces alone are too weak to keep additional molecules on the film surface. A t l o w temperatures, those forces contribute more to the bonding o f the arriving molecules, but they are far less directional than Η - b o n d i n g , and the randomness is expected to increase. Indeed, the second harmonic intensity o f the films o f 1 g r o w n at - 1 9 0 ° C was only about 10% o f those grown at 30°C (17).

Figure 7. STM images of sub-monolayer(A) and monolayer(B) films oflon Reproduced with permission from reference 17.

Pd(110).

Self-Correcting Effect So far, w e arrive to the conclusion that the highest growth rate is obtained w h e n the end group (the p y r i d y l or carboxy) o f the surface molecules tilts towards the molecular beam direction. The f o l l o w i n g discussion gives a plausible answer to h o w it is possible to achieve a directional alignment that remains over hundreds o f layers (17). W h e n the first layer o f the molecules are deposited on a glass substrate, both the p y r i d y l and carboxy groups can tilt towards the molecular beam direction to capture the arriving molecules (Figure 8). Because the p y r i d y l groups prefer to b o n d to the carboxy groups, after the bonding, the molecular direction w i l l be preserved for the

45 surface molecules w i t h their p y r i d y l group facing to the i n c o m i n g molecules (Figure 8, left). F o r those orienting their carboxy groups towards the i n c o m i n g molecules, although head-to-tail bonding to the p y r i d y l groups o f the arriving molecules is also possible, for both thermodynamic and kinetic reasons (Figure 6), they are more l i k e l y to capture the a r r i v i n g molecules through the tail-to-tail bonding. Hence, after the bonding, more than h a l f o f the carboxy groups on the surface w i l l be changed to the p y r i d y l groups (Figure 8, right). A c c o r d i n g l y , after growth o f η layers, the ratio o f p y r i d y l vs carboxy groups on the surface w i l l be larger than 2 7 1 (Figure 8). T h i s mechanism allows 1 and 2 to "self-correct" errors occurring during the growth, and hence keep the same degree o f order over hundreds o f layers. It also explains w h y the same results were obtained for multi-layer films o f 1 and 2 g r o w n on different glass substrate surfaces. A similar example o f this effect on polar inclusion compounds o f channel-type hydrocarbon crystals and N L O guests was p r o v i d e d b y H u l l i g e r et al (26). H o w e v e r , w h y is the alignment direction i n our case not p a r a l l e l to the molecular beam direction but to its projection on the substrate surface? This can be rationalized by the self-shadowing effect (17).

Figure 8. Illustration

t(

of the self correcting

effect" during OMBD with 1.

Self-Shadowing Effect T h e s e l f - s h a d o w i n g effect is a k i n e t i c p h e n o m e n u m h a p p e n i n g at n o n equilibrium conditions (low substrate temperatures). W h e n a molecule arrives at the film surface, it should choose the nearest and less crowded bonding site on the film

46 surface. After bonding to the surface, it then blocks the subsequent molecules from reaching its shadowing area. This leads to v o i d formation i n the film. The effect is w e l l k n o w n for oblique incidence molecular beam deposition o f inorganic thin films. H o w e v e r , to our knowledge, it has not yet been reported for O M B D . Considering the large size o f organic molecules, the shadowing effect should be more significant for O M B D . Formation o f voids i n the outmost layers during the growth o f 1 is illustrated in F i g u r e 9. The smallest v o i d diameter should be larger than the molecular length. T h e highest surface density o f the p y r i d y l groups and the o p t i m a l H - b o n d i n g geometry are provided to the arriving molecules only when the surface molecules and hence the voids are tilted towards the molecular beam direction. The sum o f the weak van der W a a l s and π - π stacking interactions between the tilted chains increases w i t h the chain length. These interactions attract the molecules to fill the voids for a close packing. The most probable way to fill the voids, w h i l e keeping the chain parallel to the molecular beam direction, is that the molecules i n the inner layers lie i n the X direction on the substrate surface through a two-center hydrogen-bonded intermediate illustrated i n the gray box i n Figure 9. F o r voids along the X i axis, they can be filled through the migration o f the tilted molecular chains along the X i axis (17). 3

Is the Orientation of the Arriving Molecules Random? W h e n the i n c o m i n g molecules are far from the film surface, their orientation should be random. H o w e v e r , when they get close to the film, they start to feel the l o n g range interactions from the film, especially the electro-static interaction. I f the film is anisotropic, such interactions are expected to impose an orientation preference to the arriving molecules. This is more likely to happen i n our system, where the films are not only anisotropic but also dipolar. The distance-dependent electro-static field as a sum over the dipolar molecules i n the film appears to be quite large, although calculations and quantitative measurements remain to be done to evaluate the field strength as compared to the rotational and translational kinetic energy o f the arriving molecules w h i c h favors randomness. A t first sight, the a r r i v i n g molecules seemly prefer to orient anti-parallel to the surface molecules. However, it should be noted that 1 and 2 have a very different dipole moment when they are single molecules i n the gas phase and when they are part o f the supramolecular assemblies i n the films. A c c o r d i n g to our A M I calculations, the dipole moment o f the arriving molecules is about 1 Debye. It mainly arises from the carboxy group and hence is perpendicular to the molecular axis as shown i n Figure 9. F o r those molecules i n the film, their dipole moment was calculated to be around 3 Debye and oriented along the molecular axis as illustrated b y the upper large arrows i n Figure 9. Therefore, the preferred orientation o f the arriving molecules should be the one with the carboxy group facing to the film surface (Figure 9). That is the perfect orientation for Η - b o n d i n g (Figure 6).

Summary O M B D has many technological advantages over solution-based methods for deposition o f organic thin films and nanostructures. It calls for materials designed to take these advantages and to meet the requirements o f specific applications. In this work, we present a new concept to grow anisotropic nanostructures and dipolar m u l t i -

47

Figure 9. Proposed mechanism for alignment of 1 along the X axis, the projection of the molecular beam direction on the film surface. 3

American Chemical Society Library 1155 16th St., N.W. Washington, D.C. 20036

48 layer thin films by O M B D . It began w i t h a simple assumption that the use o f r i g i d rod molecules such as 1 and 2 that can form strong and linear head-to-tail Η - b o n d i n g may reduce randomness. The directional interaction dominates the other non-directional interactions, and thus an anisotropic order c o u l d be generated. T h i s idea was supported by the formation o f centrosymmetric anisotropic nanostructures consisting o f supramolecular assemblies o f 1. Surprisingly, a macroscopic polar order can be generated i n multi-layer films o f 1 and 2 on amorphous glass substrates by oblique incidence O M B D . The polar order direction is parallel to the projection o f the molecular beam direction on the substrate surface. This is remarkable for a molecular beam incident angle o f only 3 0 ° . In addition, the films were grown out o f one source rapidly and continuously. M o r e o v e r , the polar order was independent o f the film thickness i n the range o f 100-400 n m . T o account for these intriguing results, w e proposed that the growth o f 1 and 2 is m a i n l y due to the Η - b o n d i n g o f the arriving molecules to the surface molecules. The bonding probability is determined by the orientation o f the surface molecules relative to the well-defined approaching direction o f the arriving molecules. Besides this effect, the self-correcting and self-shadowing effects that are associated w i t h the shape and bonding features o f 1 and 2, as w e l l as the long range directional interactions o f the film w i t h the arriving molecules, a l l are believed to contribute to the in-plane directional alignment o f the molecules. There are still several open questions. The nonlinear optical coefficients o f the films are rather l o w , around 0.5-1 p m / V . This is expected for 1 and 2 w h i c h have a weak donor and acceptor. The degree o f the ordering is still not clear, since the molecular nonlinearity cannot be measured. W e are trying to address this question by synchrotron X - r a y diffraction and scattering studies. The proposed mechanism also needs more quantitative theoretical support and further experimental verification. A l t h o u g h the nonlinearities obtained w i t h the prototype molecules are still too l o w for practical applications, we have demonstrated that anisotropic films w i t h a p o l a r order can be obtained by oblique incidence O M B D w i t h supramolecular assemblies based on strong and directional intermolecular interactions. O u r results and proposed mechanisms w o u l d lead to a better understanding o f the assembly o f organic molecules on substrates. In the future better nonlinear optical thin films s h o u l d be prepared by i m p r o v i n g both the material design and the processing c o n d i t i o n s . T h e w o r k presented here also demonstrates the n e e d for a m u l t i d i s c i p l i n a r y interaction between organic chemists, materials scientists, and phycisists.

References 1.

2. 3. 4. 5.

B o s s h a r d , C . ; Sutter, K.; P r ê t r e , P . ; H u l l i g e r , J . ; F l ô r s h e i m e r , M.; K a a t z , P . ; Günter, P. Organic Nonlinear Optical Materials, G o r d o n & Breach. Amsterdam, 1995. Organic Thin Films for Waveguiding Nonlinear Optics; Kajzar, F . ; Swalen J . D.; Eds.; G o r d o n & Breach: Amsterdam 1996. T r e d g o l d , R . H . Order in Thin Organic Films, C a m b r i d g e U n i v . Press: Cambridge, 1994. U l m a n . A. Chem. Rev. 1996, 96, 1533-1554. Reed, M. A. Proceedings of the IEEE, 1999, 87, 652.

49 6. 7. 8. 9. 10. 11. 12. 13. 14.

15. 16.

17. 18. 19.

M i y a t a , S.; N a l w a , H. S. E d s . ; Organic Electroluminescent Materials and Devices, G o r d o n & Breach: Amsterdam, 1997. Forrest, S. R . Chem. Rev. 1997, 97, 1793. Decher, G.; Tieke, B.; Bosshard, C.; Günter, P. Ferroelectrics 1989, 91, 193. Ohring, M. The Materials Science of Thin Films; A c a d e m i c Press: Boston, 1992. Dietrich, T.; Schlesser, R . ; Erler, B.; K ü n d i g , Α . ; Sitar, Ζ . ; Günter, P . J. Crystal Growth, 1997, 172, 473. Schlesser, R . P h . D . Dissertation, Ε Τ Η - Z ü r i c h , Switzerland, 1996. L e h n , J.-M. Supramolecular Chemistry, VCH: Weinheim, 1995. Sijbesma, R . P.; Beijer, F . H.; Brunsveld, L.; Folmer, B. J. B.; Hirschberg, J.; Lange, R . F. M.; L o w e , J. K. L.; Meijer, E . W . Science, 1997, 278, 1601. The one-dimensional Η - b o n d networks are shown i n the crystal structure of isonicotinic acid, a lower analogue o f 1 and 2. Takusagawa, F . ; Shimada, A. Acta Crystallogr. 1976, B32, 1925. C a i , C.; B ö s c h , M.; Tao, Y.; M ü l l e r , B.; G a n , Z.; K ü n d i g , Α . ; Bosshard, C.; Liakatas, I.; Jäger, M.; Günter, P. J. Am. Chem. Soc. 1998, 120, 8563. C a i , C.; B ö s c h , M.; M ü l l e r , B.; Tao, Y.; K ü n d i g , Α . ; Bosshard, C.; G a n , Z.; B i a g g i o , I.; Liakatas, I.; Jäger, M.; Schwer, H.; G ü n t e r , P. Adv. Mater. 1999, 11, 745. C a i , C.; M ü l l e r , B.; Weckesser, J.; Barth, J. V.; Tao, Y.; B ö s c h , M.; K ü n d i g , Α . ; Bosshard, C.; Biaggio, I.; Günter, P. Adv. Mater. 1999, 11, 750. Barth, J. V.; Weckesser, J.; Cai, C.; G ü n t e r , P.; B ü r g i , L.; Jeandupeux, O . ; K e r n , K . Angew. Chem. Submitted. C a i , C.; B ö s c h , M.; Tao, Y.; M ü l l e r , B.; K ü n d i g , Α . ; Bosshard, C.; G ü n t e r , P .

Polymer Preprint, 1998, 39(2), 1069. 20. M ü l l e r , B.; Cai, C.; B ö s c h , M.; Jäger, M.; Bosshard, C.; G ü n t e r , P.; Barth, J. V.; Weckesser, J.; K e r n , K. Thin Solid Films 1999, 343-344, 171. 2 1 . M ü l l e r , B.; Cai, C.; K ü n d i g , Α . ; Tao, Y.; B ö s c h , M.; J ä g e r , M.; Bosshard, C.; G ü n t e r , P. Appl. Phys. Lett., 1999, 74, 3110. 22. L i n , W . ; Lin, W . ; W o n g , G. K.; M a r k s , T. J. J. Am. Chem. Soc. 1996, 118, 8034. 23. Curtiss, L. Α . ; Blander, M. Chem. Rev. 1988, 88, 827. 24. L e g o n , A. C.; M i l l e n , D . J. Acc. Chem. Res. 1987, 20, 39. 2 5 . Weckesser, J.; Barth, J. V.; Cai, C.; M ü l l e r , B.; K e r n , K. Surf. Sci., 1999, 431, 168. 2 6 . R o t h , S. W.; L a n g l e y , P . J.; Quintel, Α . ; Wubbenhorst, M.; Rechsteiner, P.; R o g i n , P.; K o n i g , O . ; Hulliger, J. Adv. Mater. 1998, 10, 1543.