A Kelvin-Probe Force Microscopy Study - Wiley Online Library

2 downloads 0 Views 265KB Size Report
Oct 11, 2005 - ... Palermo,[a] Matteo Palma,[a, e] Zˇeljko Tomovic,[b] Mark D. Watson,[b ... [3] The electrical and electronic characterization of these films can be.
DOI: 10.1002/cphc.200500181

Influence of Molecular Order on the Local Work Function of Nanographene Architectures: A Kelvin-Probe Force Microscopy Study Vincenzo Palermo,[a] Matteo Palma,[a, e] Zˇeljko Tomovic´,[b] Mark D. Watson,[b, c] Rainer Friedlein,[d] Klaus M#llen,*[b] and Paolo Samor&*[a, e] We report a Kelvin-probe force microscopy (KPFM) investigation on the structural and electronic properties of different submicronscale supramolecular architectures of a synthetic nanographene, including extended layers, percolated networks and broken patterns grown from solutions at surfaces. This study made it possible to determine the local work function (WF) of the different pconjugated nanostructures adsorbed on mica with a resolution below 10 nm and 0.05 eV. It revealed that the WF strongly depends on the local molecular order at the surface, in particular on the delocalization of electrons in the p-states, on the molecular orientation at surfaces, on the molecular packing density, on the presence of defects in the film and on the different conforma-

tions of the aliphatic peripheral chains that might cover the conjugated core. These results were confirmed by comparing the KPFM-estimated local WF of layers supported on mica, where the molecules are preferentially packed edge-on on the substrate, with the ultraviolet photoelectron spectroscopy microscopically measured WF of layers adsorbed on graphite, where the molecules should tend to assemble face-on at the surface. It appears that local WF studies are of paramount importance for understanding the electronic properties of active organic nanostructures, being therefore fundamental for the building of high-performance organic electronic devices, including field-effect transistors, light-emitting diodes and solar cells.

Introduction Organic thin films of p-conjugated molecules with tailor-made properties are promising candidates for the fabrication of displays, transistors, memories and solar cells.[1, 2] The performance of these devices strongly depends on the structural arrangement of the components at the supramolecular level.[3] The electrical and electronic characterization of these films can be challenging, due to the complexity of their morphologies, their high electrical resistance and the presence of defects. In this frame, the determination of the work function of the adsorbate (WF, that is, the energy difference of an electron between the vacuum level and the Fermi level[4]) on a local scale is of fundamental importance. It is, in fact, essential to tune differences in energy levels between molecular components and electrodes properly, as they influence the rates of electron- and holetransfer processes in a hybrid metal–organic contact, which in turn determine the nature of coherent conductance through a junction.[5] The electronic structure of organic thin films is often studied by X-ray (XPS) and ultraviolet photoelectron spectroscopies (UPS).[6] Since these techniques usually average over a rather large microscopic area and since charges are produced, the application of these methods is often not appropriate for the study of different film types, including films with a nominal thickness larger than 20 nm, thin discontinuous layers and films with a thickness smaller than 10 nm supported on insulating substrates. Charging problems can be avoided by performing Kelvinprobe measurements that allow the determination of the WF ChemPhysChem 2005, 6, 2371 – 2375

in a noninvasive way without any limitation on film thickness.[7] In the last ten years, the Kelvin-probe technique has been coupled with scanning force microscopy to achieve a spatial resolution on the scale of tens of nanometers.[8] The simplest version of this technique is electrostatic force microscopy (EFM), which allows the mapping of charges on a sample surface.[9] A conductive tip scans the surface while an alternating current

[a] Dr. V. Palermo, Dr. M. Palma, Dr. P. Samor Istituto per la Sintesi Organica e la Fotoreattivit Consiglio Nazionale delle Ricerche, via Gobetti 101 40129 Bologna (Italy) Fax: (+ 39) 051-639-9844 E-mail: [email protected] [b] Dr. Zˇ. Tomovic´, Dr. M. D. Watson, Prof. Dr. K. M=llen Max-Planck Institute for Polymer Research, Ackermannweg 10 55124 Mainz (Germany) Fax: (+ 49) 6131-379-350 E-mail: [email protected] [c] Dr. M. D. Watson Department of Chemistry, University of Kentucky, Lexington KY 40506-0055 (USA) [d] Dr. R. Friedlein Department of Physics (IFM), Linkçping University 581 83 Linkçping (Sweden) [e] Dr. M. Palma, Dr. P. Samor Nanochemistry Laboratory, Institut de Science et d’IngFnierie SupramolFculaires (ISIS), UniversitF Louis Pasteur 8, allFe Gaspard Monge , 67083 Strasbourg (France) Supporting information for this article is available on the WWW under http://www.chemphyschem.org or from the author.

D 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

2371

K. M=llen, P. Samor et al. (AC) voltage is applied between the tip and the sample surface. The tip oscillation, due to electrostatic interactions during the scan, maps regions of higher or lower surface-charge distribution on the sample surface. These are directly related to work-function variations.[10, 11] In Kelvin-probe force microscopy (KPFM), the electrostatic interaction between the tip and the surface is minimized by applying a variable offset potential voltage (Vdc) to the cantilever tip. When the first harmonic component (Fw) of the electrostatic force interaction [12] is zero, Vdc is equal to the surface potential, that is, it corresponds to the difference in WF between the tip and the sample surface.[13] In this way, the surface potential can be measured with a resolution of a few millivolts on areas of some tens of square nanometers. KPFM has been widely employed to study conventional semiconductors,[14] organic thin films—including chemisorbed monolayers of either alkanethiols on Au[15] or phenylacetylene on SiO2,[16] ordered doped polythiophene monolayers [17] and Langmuir–Blodgett monolayers of hexa-peri-hexabenzocoronene on SiO2,[18] with a lateral resolution down to some tens of nanometers. Moreover, it provided insight into the electrostatic potential decay in field-effect transistors (FETs) based on oligo[19] and poly-alkylthiophenes[20] as well as pentacenes.[21] Polycyclic aromatic hydrocarbons (PAHs) are nanographenes possessing attractive electronic properties.[22, 23] The high structural order which can be achieved in the bulk and at surfaces,[24–29] together with a large functional electronic component, the PAH core, gives rise to exceptional one-dimensional transport properties,[30] which have been exploited, for example, in the development of solar cells.[31] We have chosen an alkylated PAH, C96-C12 (Figure 1 a) having 96 carbons in the aromatic core. This molecule exhibits a remarkably high solubility in different organic solvents,[32] which can be explained by an inefficient side-chain packing around self-assembled columns. Moreover bulk samples of C96-C12 revealed hexagonally packed columnar mesophases with the discs planes oriented perpendicular to the columnar main axes; such liquid crystalline character persists from 100 8C up to 550 8C. Here we describe a KPFM study of the electronic properties of different nanoscale morphologies of C96-C12 on an electrically insulating surface. By recording the surface potentials we determined the work functions of these architectures, and compared them to the work functions of standard crystalline surfaces such as highly oriented pyrolytic graphite (HOPG) and gold.

Results and Discussions C96-C12 has been processed on freshly cleaved muscovite mica surfaces by varying systematically the solvent, concentration, temperature, and deposition method, leading to three different surface morphologies: i)

A continuous layer, with a thickness of (2.5  0.5) nm (Figure 1 b) obtained by drop-casting a 10 5 mol l 1 solution in 1,2-dichlorobenzene at 135 8C;

2372

www.chemphyschem.org

Figure 1. a) Chemical formula of C96-C12. A rough determination of the width of the molecule gave a diameter of the aromatic core of 1.7 nm and dodecyl side chains with a contour length of 1.5 nm in the case that they adopt the ideal fully stretched conformation. b)–d) SFM topographical images of thin films of C96-C12 adsorbed on mica by: b) drop-casting a 10 5 mol l 1 solution in 1,2-dichlorobenzene at 135 8C, c) immersion of the substrate in a 10 6 mol l 1 solution in 1,2-dichlorobenzene at 85 8C, d) dropcasting a 10 5 mol l 1 solution in CHCl3 at 45 8C. The z scales are indicated in the top left of each SFM image.

A percolated network of fibrelike objects, having widths of (11  2) nm and heights of (2.7  0.4) nm (Figure 1 c). This film was prepared by immersion of the mica in a 10 6 mol l 1 solution in 1,2-dichlorobenzene at 85 8C; iii) Broken patterns of elongated clusters with widths of (10  3) nm, heights of (2.6  0.8) nm and lengths of (200  100) nm (Figure 1 d). Such a surface was produced by drop-casting a 10 5 mol l 1 solution in CHCl3 at 45 8C. ii)

The height of these three assemblies is comparable to the molecule diameter, which suggests an edge-on packing of the molecules at the surface. This is in good agreement with the well-known tendency of conjugated molecules bearing aliphatic side groups to pack edge-on on the basal plane of an insulating mica surface, and also because of the different hydrophobic and hydrophilic character of the admolecule and of the substrate, respectively.[33] The widths of the fibres and of the elongated clusters suggest a lateral packing of about three to five molecules forming the anisotropic architectures. The formation of these three different assemblies is governed by the interplay of intramolecular, intermolecular and interfacial interactions. Key roles can be ascribed to the strong p–p interactions among the conjugated cores[26, 27] and to the mechanism of solvent evaporation at the surface.[34, 35] A systematic study of this growth process will be reported elsewhere.[36] Figure 2 displays the simultaneously recorded topographical (Figure 2 a) and EFM (Figure 2 b) images of C96-C12 deposited on mica by immersion of the substrate in a 10 5 mol l 1 solu-

D 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemPhysChem 2005, 6, 2371 – 2375

Nanographene Architectures: A Force Microscopy Study

Figure 2. SFM topographical (a) and EFM (b) images of C96-C12 networks deposited on mica by immersion of the substrate in a 10 5 mol l 1 solution in CHCl3. The EFM measurements were performed applying a bias of 8 V to the sample. Arrows indicate impurities/co-adsorbates on the network. [z scale in (a) = 7 nm].

From the topographical profiles in the KPFM images it was possible to determine the WF for each type of nanostructure. The layers exhibited a surface potential of DF = (1.00  0.02) eV. Consequently, the work function value for the C96-C12 layers (i) was determined to be F(C96 layer) = F(tip)-DF = (4.88–1.00) eV = (3.88  0.03) eV. Similarly, the WF of the networks (ii) and of the elongated clusters (iii) was found to be F(C96 network) = (3.53  0.05) eV and F(C96 clusters) = (3.47  0.06) eV, respectively. The WFs of the three different types of C96-C12 architectures are summarized in Figure 4. While the WF of the network was 350 mV smaller than that estimated for the layers, the WF of

tion in chloroform. The topographical image shows a percolated network of fibres with a height of  2.6 nm and a typical width of (11  2) nm. Small globular co-adsorbates with a height of 5 nm are visible on top of the fibre branches (see arrows in Figure 2). In the EFM image, the C96-C12 network appears brightFigure 3. KPFM image of C96-C12 architectures on mica. a) Layers formed by drop-casting a 10 5 mol l 1 solution in er than the surrounding flat sub- 1,2-dichlorobenzene at room temperature. b) Network of fibres obtained by immersion of mica in a 10 6 mol l 1 strate areas, indicating that the solution in 1,2-dichlorobenzene at 85 8C. c) Elongated agglomerates produced by spin-coating a 10 5 mol l 1 in former is more charged. On the chloroform solution at 45 8C. Different profiles have been traced on the KPFM images to estimate the local work other hand, the globular co-ad- function (not shown). A white line is shown in each image as an example of a location where the profiles were traced. sorbates are darker. This suggests a lower surface charge and, importantly, provides unambiguous evidence that in contrast to previous reports,[37] the contribution of the topographical signal in the EFM is not pivotal. Since the EFM technique just offers a qualitative map of the surface charges, the study was extended to KPFM measurements. In our experiments we have used a p-type silicon tip and a freshly cleaved HOPG reference sample for the set-up calibration. The HOPG work function in air amounts to 4.65 eV.[14] We measured a surface potential of a freshly cleaved HOPG (ZYH grade, Advanced Ceramics, USA) as: DF(tipHOPG) = (0.23  0.02) eV. From this value, we determined the silicon tip work function as F(tip) = F(HOPG) + DF = (4.88  0.02) eV, which matches well with that of a p-type silicon, that Figure 4. Comparison of work functions for different reference samples, C96is, (4.85  0.05) eV.[38] After calibration, the work functions of C12 architectures (i)–(iii) on mica and C96-C12 ultrathin films prepared by spin-coating on HOPG. other surfaces, including gold and silicon, were measured to test the reliability of the methodology. The agreement between measured and expected values is good (see top of the elongated clusters was found to be even lower. This differFigure 4).[12] ent electronic behaviour might be explained by assuming a We focused our attention on the three different types of higher delocalisation of the electrons along the p-stacks in the C96-C12 architectures formed on mica shown in Figures 1 b–d. layer architecture compared to the network. It is indeed reaThe KPFM images of: i) layers, ii) network of fibres and iii) elonsonable to consider that the continuous morphological strucgated clusters are shown in Figures 3 a, 3 b and 3 c, respectiveture of the former architectures will possess more extended ply. The high lateral resolution in the KPFM measurements states caused by a substantial intermolecular p–p overlap of made it possible to resolve their structural motifs properly, as the electronic molecular wavefunctions[26] that will contribute confirmed by comparing the KPFM images to the simultaneously recorded topographical scanning force microscopy (SFM) to an increase of the depth of the attractive potential inside images. the solid. This, in turn, is known to affect the value of the work ChemPhysChem 2005, 6, 2371 – 2375

D 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemphyschem.org

2373

K. M=llen, P. Samor et al. function.[39] Such an interpretation is confirmed by the higher WF of the (0001) face of HOPG, 4.65 eV, compared to the different WFs for the C96-C12 nanostructures, which is due to the large out-of-plane p–p overlap of the electronic wavefunctions in the graphite crystal, and the exclusive surface termination by p-orbitals instead of the rather localised s states of aliphatic side chains that might cover the aromatic cores in the PAH films. Nevertheless, it should be pointed out that, due to the longrange nature of electrostatic forces used to map a surface by KPFM, the resolution and accuracy in the KPFM results obtained is likely to be higher for the nanostructure (i) compared to (ii) and (iii). This is because the measured potential is actually a weighted average over the potential distribution on the surface, the derivatives of the capacitances being the weighting factors.[40] Therefore the resolution of the work function determination decreases with the lateral size of the analysed object. Furthermore, it is worth noting that a direct contribution of the substrate to the electrostatic interaction between the tip and the surface, as measured by KPFM, might also be possible.[13] The C96-C12 WF values of the nanostructures (i)–(iii) have also been compared to the macroscopic work function of an  2-nm-thick (very flat) C96-C12 film on HOPG (see the SFM topographical image in the Supporting Information) prepared by spin-coating a solution in 1,2-dichlorobenzene. From the secondary electron cut-off in the UPS spectra,[41] a value of F(C96 spin-coated) = (4.2  0.1) eV was obtained, which is significantly larger than the WFs of the nanostructured films supported on mica. This difference can be ascribed to the known tendency of the nanographenes to assemble parallel to the basal plane of the substrate on HOPG,[42] which differs from the edge-on packing on mica. Furthermore, the influence of the substrate on the obtained WF values cannot be neglected. In particular, the HOPG/C96-C12 interaction can induce a re-hybridization of the p states of the adsorbed molecule, altering its electronic properties.[25] The important role of the molecular orientation at surfaces, the molecular packing density, the presence of defects in the film and the different conformation of the aliphatic peripheral chains that might cover the conjugated core is thus obvious. Similar to the case of inorganic crystals,[7] the results obtained here clearly demonstrate that the measured work function strongly depends on the surface considered and thus is not a bulk parameter. The ionisation potential of the C96-C12 film spun on HOPG was 4.5–4.8 eV, as measured by UPS, and about 4.6 eV for the molecules in solution, as determined by cyclic voltammetry.[43] This indicates that the charge-injection barrier of the spincoated C96-C12 film, with respect to a conventional metallic indium tin oxide (ITO) electrode (WF = 4.7 eV), can be very small. With the significantly reduced work functions of about 3.5–3.9 eV of the morphologies (i)–(iii), a slightly larger injection barrier should be expected. This is a clear example of how the confinement and ordering on the nanoscale of a material deeply influences its electronic structure and the energy-level

2374

www.chemphyschem.org

alignment at inorganic–organic interfaces. Consequently, the choice of the ideal contact electrodes for a C96-based nanostructure or for the 3D microscopic bulk will be different, since a small charge-injection barrier at the metal–organic interface is fundamental for the fabrication of hybrid devices.

Conclusions In summary, we have shown that KPFM is a powerful method to explore, simultaneously, the structural and electronic properties of nanoscale organic architectures grown on surfaces with a resolution below 10 nm and 0.05 eV. We have determined with high precision the work function of different pconjugated nanographenes, that is, C96-C12 assemblies adsorbed on insulating mica. The measured WF of these morphologies, characterised by a preferential edge-on packing of the molecules on the basal plane of the substrate, was found to be markedly smaller than that of spin-coated films supported on HOPG, where the conjugated discs should tend to be oriented parallel to the substrate surface plane. These results indicate that the estimated work function of a nanosized molecular assembly depends on many factors, which include the degree of electron delocalisation in the p states, the molecular orientation at surfaces, the molecular packing density, the presence of defects in the film and the different conformations of the aliphatic peripheral chains that might cover the conjugated core. These findings demonstrate that local work-function studies are of great importance for understanding the electronic functions of active organic self-assembled structures, being therefore essential for the fabrication of high-performance supramolecularly engineered electronic devices including FETs, light-emitting diodes and solar cells.

Experimental Section The C96-C12 molecule was assembled at surfaces, systematically changing the relevant experimental parameters, which include the solvent, concentration, deposition method, and the temperature of the substrate and of the solution. Solutions have been prepared in chloroform, 1,2-dichlorobenzene, 1,2,4- trichlorobenzene and toluene (Aldrich) at various concentrations (ranging from 10 4 down to 10 9 mol l 1). The different solutions were deposited on freshly cleaved muscovite mica (Ted Pella Inc., USA) either by drop-casting, spin-coating or substrate immersion. The solution was heated before deposition at temperatures of 45–150 8C and the sample was left to dry, after deposition, on a heating plate for 2 h at the same temperature of the solution. The SFM topographical imaging, as well as the EFM and KPFM measurements, were carried out using a commercial apparatus (CP AutoProbe Research, Thermomicroscopes, Veeco) operating in noncontact mode at room temperature in an air environment. Silicon p-doped tips with a force constant of k = 3.2 N m 1 were used. Scan rates were 0.2–0.5 line s 1. Kelvin-probe and the topographical signals were simultaneously detected at two different frequencies of oscillation of the tip. KPFM exploits the first harmonic component (Fw) of the electrostatic force interaction to determine the local work function on the sample surface ,[12] where w is the frequency of the applied AC voltage signal.

D 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

ChemPhysChem 2005, 6, 2371 – 2375

Nanographene Architectures: A Force Microscopy Study Image processing was performed using the scanning probe image processor (SPIP, Version 2.000, Image Metrology ApS, Lyngby, Denmark). Thin C96-C12 films with a nominal thickness of 2 nm were spun on freshly cleaved highly oriented pyrolytic graphite (ZYH grade, Advanced Ceramics, USA) using 1,2-dichlorobenzene as solvent. Their macroscopic work function and the ionisation potential were measured in a home-built spectrometer under ultra-highvacuum conditions.[44]

Acknowledgments Financial support from ESF-SONS-BIONICS, the EU through the Integrated Project NAIMO, the Marie Curie EST project SUPER (MEST-CT-2004–008128), the ForceTool project (NMP4-CT-2004– 013684), the bilateral programme CNR-CNRS and the Swedish Science Foundation (VR) under contracts 12252003 and 12252020 is gratefully acknowledged. Keywords: conjugation · Kelvin-probe force microscopy · organic electronics · thin films · work function [1] R. H. Friend, R. W. Gymer, A. B. Holmes, J. H. Burroughes, R. N. Marks, C. Taliani, D. D. C. Bradley, D. A. Dos Santos, J. L. BrQdas, M. Lçgdlund, W. R. Salaneck, Nature 1999, 397, 121. [2] S. R. Forrest, Nature 2004, 428, 911. [3] M. Van der Auweraer, F. C. De Schryver, Nat. Mater. 2004, 3, 507. [4] C. Kittel, Introduction to Solid State Physics, 8th edition, Wiley, 2005. [5] J. R. Heath, M. A. Ratner, Physics Today 2003, 56(5), 43 – 49. [6] W. R. Salaneck, S. Stafstrçm, J. L. BrQdas, Conjugated Polymer Surfaces and Interfaces: Electronic and Chemical Structure of Interfaces for Polymer Light Emitting Devices, Cambridge University Press, Cambridge, 2003. [7] H. Ishii, K. Sugiyama, E. Ito, K. Seki, Adv. Mater. 1999, 11, 605. [8] M. Nonnenmacher, M. P. Oboyle, H. K. Wickramasinghe, Appl. Phys. Lett. 1991, 58, 2921. [9] B. D. Terris, J. E. Stern, D. Rugar, H. J. Mamin, Phys. Rev. Lett. 1989, 63, 2669. [10] T. Q. Nguyen, M. L. Bushey, L. E. Brus, C. Nuckolls, J. Am. Chem. Soc. 2002, 124, 15 051. [11] C. H. Lei, A. Das, M. Elliott, J. E. Macdonald, Appl. Phys. Lett. 2003, 83, 482. [12] H. O. Jacobs, A. Stemmer, Surf. Interface Anal. 1999, 27, 361. [13] H. O. Jacobs, H. F. Knapp, S. Muller, A. Stemmer, Ultramicroscopy 1997, 69, 39. [14] C. Sommerhalter, T. W. Matthes, T. Glatzel, A. Jager-Waldau, M. C. LuxSteiner, Appl. Phys. Lett. 1999, 75, 286. [15] T. Ichii, T. Fukuma, K. Kobayashi, H. Yamada, K. Matsushige, Nanotechnology 2004, 15, S30. [16] N. Saito, K. Hayashi, H. Sugimura, O. Takai, Langmuir 2003, 19, 10 632. [17] T. Hassenkam, D. R. Greve, T. Bjçrnholm, Adv. Mater. 2001, 13, 631.

ChemPhysChem 2005, 6, 2371 – 2375

[18] N. Reitzel, T. Hassenkam, K. Balashev, T. R. Jensen, P. B. Howes, K. Kjaer, A. Fechtenkçtter, N. Tchebotareva, S. Ito, K. M#llen, T. Bjçrnholm, Chem. Eur. J. 2001, 7, 4894. [19] T. Miyazaki, K. Kobayashi, K. Ishida, S. Hotta, T. Horiuchi, H. Yamada, K. Matsushige, Japanese J. Appl. Phys. Part 1 2003, 42, 4852. [20] L. B#rgi, H. Sirringhaus, R. H. Friend, Appl. Phys. Lett. 2002, 80, 2913. [21] K. P. Puntambekar, P. V. Pesavento, C. D. Frisbie, Appl. Phys. Lett. 2003, 83, 5539. [22] M. D. Watson, A. Fechtenkçtter, K. M#llen, Chem. Rev. 2001, 101, 1267. [23] F. JTckel, M. D. Watson, K. M#llen, J. P. Rabe, Phys. Rev. Lett. 2004, 92. [24] P. Samor&, A. Fechtenkçtter, F. JTckel, T. Bçhme, K. M#llen, J. P. Rabe, J. Am. Chem. Soc. 2001, 123, 11 462. [25] P. Samor&, N. Severin, C. D. Simpson, K. M#llen, J. P. Rabe, J. Am. Chem. Soc. 2002, 124, 9454. [26] R. Friedlein, X. Crispin, C. D. Simpson, M. D. Watson, F. Jackel, W. Osikowicz, S. Marciniak, M. P. de Jong, P. Samor&, S. K. M. Jonsson, M. Fahlman, K. M#llen, J. P. Rabe, W. R. Salaneck, Phys. Rev. B 2003, 68, 195 414. [27] A. M. van de Craats, N. Stutzmann, O. Bunk, M. M. Nielsen, M. Watson, K. M#llen, H. D. Chanzy, H. Sirringhaus, R. H. Friend, Adv. Mater. 2003, 15, 495. [28] J. S. Wu, M. D. Watson, L. Zhang, Z. H. Wang, K. M#llen, J. Am. Chem. Soc. 2004, 126, 177. [29] P. Samor&, X. M. Yin, N. Tchebotareva, Z. H. Wang, T. Pakula, F. JTckel, M. D. Watson, A. Venturini, K. M#llen, J. P. Rabe, J. Am. Chem. Soc. 2004, 126, 3567. [30] M. G. Debije, J. Piris, M. P. de Haas, J. M. Warman, Z. Tomovic, C. D. Simpson, M. D. Watson, K. M#llen, J. Am. Chem. Soc. 2004, 126, 4641. [31] L. Schmidt-Mende, A. Fechtenkçtter, K. M#llen, E. Moons, R. H. Friend, J. D. MacKenzie, Science 2001, 293, 1119. [32] Z. Tomovic, M. D. Watson, K. M#llen, Angew. Chem. Int. Ed. 2004, 43, 755. [33] P. Samor&, V. Francke, K. M#llen, J. P. Rabe, Chem. Eur. J. 1999, 5, 2312. [34] A. Sharma, G. Reiter, J. Colloid Interface Sci. 1996, 178, 383. [35] E. Rabani, D. R. Reichman, P. L. Geissler, L. E. Brus, Nature 2003, 426, 271. [36] V. Palermo, M. Palma, Z. Tomovic, K. M#llen, P. Samor&, unpublished results. [37] K. Okamoto, Y. Sugawara, S. Morita, Jpn. J. Appl. Phys. Part 1-Regular Papers Short Notes & Review Papers 2003, 42, 7163. [38] H. W. Wolf, Semiconductors, Wiley-Interscience, New York, 1971. [39] M. Pruton, Introduction to Surface Physics, Oxford University Press, New York, 1994. [40] H. O. Jacobs, P. Leuchtmann, O. J. Homan, A. Stemmer, J. Appl. Phys. 1998, 84, 1168. [41] C. S. Fadley in Electron Spectroscopy: Theory, Techniques, and Applications (Eds.: C. R. Brundle, A. D. Baker), Academic Press, London, 1978. [42] P. Samor&, M. Keil, R. Friedlein, J. Birgerson, M. Watson, M. M#llen, W. R. Salaneck, J. P. Rabe, J. Phys. Chem. B 2001, 105, 11114. [43] S. Goddard, unpublished results. [44] R. Friedlein, X. Crispin, M. Pickholz, M. Keil, S. Stafstrçm, W. R. Salaneck, Chem. Phys. Lett. 2002, 354, 389.

Received: April 4, 2005 Published online on October 11, 2005

D 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

www.chemphyschem.org

2375